literature review type 2 diabetes

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Comput Struct Biotechnol J

Association of risk factors with type 2 diabetes: A systematic review

Leila ismail.

a Intelligent Distributed Computing and Systems Research Laboratory, Department of Computer Science and Software Engineering, College of Information Technology, United Arab Emirates University, Al Ain, Abu Dhabi, 15551, United Arab Emirates

Huned Materwala

Juma al kaabi.

b College of Medicine and Health Sciences, Department of Internal Medicine, United Arab Emirates University, Al Ain, Abu Dhabi 15551, United Arab Emirates

c Mediclinic, Al Ain, Abu Dhabi, United Arab Emirates

Diabetes is the leading cause of severe health complications and one of the top 10 causes of death worldwide. To date, diabetes has no cure, and therefore, it is necessary to take precautionary measures to avoid its occurrence. The main aim of this systematic review is to identify the majority of the risk factors for the incidence/prevalence of type 2 diabetes mellitus on one hand, and to give a critical analysis of the cohort/cross-sectional studies which examine the impact of the association of risk factors on diabetes. Consequently, we provide insights on risk factors whose interactions are major players in developing diabetes. We conclude with recommendations to allied health professionals, individuals and government institutions to support better diagnosis and prognosis of the disease.

1. Introduction

Diabetes Mellitus (DM) commonly referred to as diabetes, is a chronic disease that affects how the body turns food into energy [1] . It is one of the top 10 causes of death worldwide causing 4 million deaths in 2017 [2] , [3] . According to a report by the International Diabetes Federation (IDF) [3] , the total number of adults (20–79 years) with diabetes in 2045 will be 629 million from 425 million in 2017 (48% increase). In 2017, diabetes caused at least 727 billion USD in health expenditure, which is 12% of the total spending on adults [3] . According to the National Diabetes Statistics Report [4] , 30.3 million (9.4% of the US population) people have diabetes, and 84.1 million (29.06% of the population) have pre-diabetes. 1 in 2 people (212 million) with diabetes was undiagnosed in 2017 according to IDF [5] . Diabetes if left untreated can cause serious medical issues, such as cardiovascular disease, stroke, chronic kidney disease, foot ulcers, damage to the eyes, and prolonged kidney ailment. To date, there is no permanent cure for diabetes and the patients have to rely on healthy lifestyle and timely medication [6] .

There are three main types of diabetes: type 1, type 2, and gestational diabetes (diabetes while pregnant) [1] . Type 1 diabetes mostly occurs in children and adolescents. 1,106,500 children were suffering from type 1 diabetes in 2017 [3] . The symptoms of type 1 diabetes include abnormal thirst and dry mouth, frequent urination, fatigue, constant hunger, sudden weight loss, bed-wetting, and blurred vision. Type 2 diabetes is mostly seen in adults, but it is increasing in children and adolescents due to the rising level of obesity, physical inactivity and unhealthy diet [5] . 372 million adults were at the risk of developing type 2 diabetes in 2019 [3] . In 2017, more than 21 million live births were affected by diabetes during pregnancy [3] . In this paper, we focus on type 2 diabetes due to the alarming numbers.

Type 2 Diabetes is thought to prevail in an individual from an interaction between several lifestyle, medical condition, hereditary, psychosocial and demographic risk factors such as high-level serum uric acid, sleep quality/quantity, smoking, depression, cardiovascular disease, dyslipidemia, hypertension, aging, ethnicity, family history of diabetes, physical inactivity, and obesity [6] . In this paper, we present a systematic review of the literature on the association of these risk factors with the incidence/prevalence of type 2 diabetes. We give insights on the contribution of independent risk factors in the development of type 2 diabetes along with possible solutions towards a preventive approach.

We conduct a systematic literature search using CINAHL, IEEE Xplore, Embase, MEDLINE, PubMed Central, ScienceDirect, Scopus, Springer, and Web of Science databases. Our search criteria does not include a time bound. Its main objective is to retrieve all the studies which examine the association between individual risk factors and the incidence/prevalence of type 2 diabetes. Table A1 shows the search string used for each risk factor. The relevant studies have to meet the following inclusion criteria: 1) published in the English language, 2) prospective cohort or cross-sectional study, 3) type 2 diabetes as a specified risk, 4) one of its risk factors, 5) findings in terms of Odds Ratio (OR), Risk Ratio/Relative Risk (RR), or Hazard Ratio (HR), and the corresponding 95% Confidence Intervals (CIs) for the association between the risk factor and type 2 diabetes. To assess the quality of the studies, we use the National Institutes of Health (NIH) quality assessment tool [7] . The tool consists of 14 questions to evaluate the validity and bias risk of a study. We answered each question by either yes, no, cannot be determined, not applicable, or not reported. The tool then classifies each study as high quality (Good), moderate quality (Fair) and low quality (Poor).

Fig. 1 shows the result of our systematic approach that is used to screen the relevant studies. Irrelevant studies that do not meet the inclusion criteria mentioned in the previous section were excluded after screening titles, abstracts and full texts. At last, 106 papers are considered for this review. These papers are divided into ten categories based on the risk factor under study ( Fig. 1 ). Our review reveals that there is no study that examines the association of age or physical inactivity as an independent risk factor with type 2 diabetes. Table A2 shows the quality assessment results for the studies included in this paper. For smoking, cardiovascular disease and hypertension risk factors, the majority of the studies are of high quality. For serum uric acid, sleep quantity/quality, depression, dyslipidemia, ethnicity, family history of diabetes and obesity, the majority of the studies are of moderate quality.

Flowchart of the selection of relevant studies.

3.1. Serum uric acid

Serum uric acid, a common component of urine generated by the metabolic breakdown of purines, have been associated with insulin resistance and type 2 diabetes [8] . High serum uric acid level in an individual leads to: 1) nitric-oxide mediated vasoconstriction (contraction of blood vessels) leading to impaired glucose uptake in the muscles [9] , 2) increase in oxidative stress [10] and 3) increase in inflammation leading to a decrease in adiponectin [11] , [12] . Consequently, the blood glucose level increases leading to dysfunctional and eventually dead beta-cells [13] . As a result, the individual develops type 2 diabetes. Table 1 shows the characteristics and findings of the work in the literature studying the association between high serum uric acid level and type 2 diabetes.

Characteristics and findings of the studies examining the association between high level serum uric acid and type 2 diabetes.

RS-Random Sample, MONICA-Multinational MONItoring of trends and determinants in CArdiovascular disease, ARIC-Atherosclerosis Risk in Communities, FDPS-Finnish Diabetes Prevention Study, CSCCS-Chin Shan Community Cardiovascular study, MRFIT-Multiple Risk Factor Intervention Trial, NHANES-National Health and Nutrition Examination Survey, QFS-Quebec Family Study, M-Men, W-Women, PCS-Prospective Cohort Study, CSS-Cross-Sectional Study.

Perry et al. [14] found that an individual having a uric acid level of more than 411 μ mol/l is at 1.5 times more risk of developing type 2 diabetes compared to an individual having uric acid level less than 302 μ mol/l. Niskanen et al. [15] also confirmed that change in uric acid levels is associated with a 2 times increase in the risk of incidence type 2 diabetes. Dehghan et al. [16] in their study showed that individuals having uric acid level > 370 μ mol/l are at high risk of incidence type 2 diabetes (HR 1.68, 95% CI 1.22–2.30) compared to those having uric acid level ⩽ 267 μ mol/l. The authors concluded that lowering uric acid level can be a novel approach for diabetes prevention. Xu et al. [17] found that the association between high serum uric acid level and diabetes is the same in both men and women (RR 1.131, 95% CI 1.084–1.179). The association (RR 1.17, 95% CI 1.09–1.25) is also examined by Kodama et al. [18] . Nakagawa et al. [19] showed that uric acid is a significant and independent risk factor in predicting hyperinsulinemia. The authors observed that serum uric acid level ⩾ 5.5 mg/dl is associated with the development of hyperinsulinemia after 6 months (OR 5.47, 90% CI 1.6–1.77) and 12 months (OR 3.4, 90% CI 1.1–10.4). However, the cohort was controlled for gender and age ( > 60 years). Consequently, it can not be concluded whether uric acid is an independent risk factor or there is an integrated effect of uric acid, gender and age.

Several studies argue that high-level uric acid is not an independent risk factor and it only emphasizes the association between independent risk factors such as age, obesity, hypertension, gender, and dyslipidemia, and type 2 diabetes [20] . Chou et al. show that uric acid has a significant association with type 2 diabetes in old and obese individuals [21] . Another study by Meisinger et al. [22] shows that high-level uric acid is associated with incidence of type 2 diabetes in women only with HR 2.5 per 1 mmol/L increase. Carnethon et al. [23] found that the risk of incidence type 2 diabetes increases (OR 1.3, (1.2–1.4)) with every 1.4 mg/dl increase in uric acid level. However, this is in combination with an increase in waist/hip ratio, smoking and obesity. Chien et al. [24] stated that individuals with a uric acid level of 0.486 mmol/L and having metabolic syndrome have a 3.3 times more risk of incidence type 2 diabetes compared to those with a uric acid level of 0.211 mmol/L and not having metabolic syndrome. Nan et al. [25] examined the impact of ethnicity and gender on the association between uric acid and incidence of type 2 diabetes. The authors found that the high serum uric acid is an independent risk factor for type 2 diabetes in Mauritian Indian men compared to Creole men, and there is a no-to-weak association in women of both ethnicity. Similarly, Choi et al. [26] studied the association between uric acid and type 2 diabetes in men having cardiovascular risk profile. The authors concluded that men with cardiovascular profile having high uric acid level are twice likely to develop type 2 diabetes. The authors also stated that this association between uric acid and diabetes is independent of other risk factors such as obesity, age, family history of diabetes, hypertension, and metabolic syndrome. Kramer et al. [27] analyzed the impact of age and impaired fasting glucose (IFG) on the association and found that high uric acid level can independently predict incidence of type 2 diabetes (OR 1.65, 95% CI 1.25–2.18) in older adults having IFG. Lv et al. [28] found that high serum uric acid level is associated to type 2 diabetes in middle-aged or older people (RR 1.56, 95% CI 1.39–1-76).

In summary, the association between high-level serum uric acid remains obscure. It is debatable whether serum uric acid is an independent risk factor for type 2 diabetes or it only emphasizes the association between other independent risk factors and type 2 diabetes. Some studies reported a positive association between high serum uric acid level and incidence of type 2 diabetes [14] , [15] , [16] , [19] , [24] , whereas others [25] , [29] reported no association. On the contrary, some studies reported an inverse association between uric acid and diabetes [30] , [31] , [32] . Furthermore, some studies argue that there is a reverse association, i.e., diabetes leads to high uric acid levels [33] , [34] .

3.2. Sleep quantity/quality

The quality and quantity of sleep are affected by several cultural, social, behavioral, psychological, and environmental factors. The working professionals often experience fatigue, tiredness and daytime napping due to irregular working hours and shifts. Evidence shows that the current average sleep of an individual, i.e., 6.8 h/night, is 1.5 h less than that a century ago [45] . The cause of sleep loss is multi-factorial. For instance 45% of adults report that they sleep fewer hours to get more work done, 43% reported that they watch television or use the Internet, and 22% reported to be suffering from insomnia. The unusual, disturbed and reduced sleep is associated with glucose intolerance [46] .

An individual suffering from sleep disorder, known as obstructive sleep apnea (OSA), experiences: 1) deficiency in the amount of oxygen reaching the tissues by total/partial collapse of upper airways while sleeping (hypoxia) and 2) inflammation. Frequent Hypoxia triggers an increase in sympathetic activity [47] . Increased sympathetic activity and inflammation lead to insulin resistance condition [48] , [49] and eventually to type 2 diabetes. Table 2 shows the characteristics and findings of the work in the literature studying the association between sleep quantity/quality and type 2 diabetes.

Characteristics and findings of the studies examining the association between sleep quantity/quality and type 2 diabetes.

DIS-Difficulty Initiating Sleep, DMS-Difficulty Maintaining Sleep, EPIC-European Prospective Investigation into Cancer and Nutrition, FIN D2D-Finnish type 2 Diabetes, HIPOP-OHP-High risk and Population Strategy for Occupational Health Promotion, IHHP-Isfahan Healthy Heart Program, IRAS-Insulin Resistance Atherosclerosis Study, M-Men, MC-Millennium Cohort, MMAS-Massachusetts Male Aging Study, MONICA-Multinational MONItoring of trends and determinants in CArdiovascular disease, MPP-Malmo Preventive Project, NHANES-National Health and Nutrition Examination Survey, NHIS-National Health Interview Survey, NHS-Nurse Health Study, NHW-Non Hispanic Whites, NIH AARP-National Institutes of Health American Association of Retired Persons Diet and Health Study, QFS-Quebec Family Study, RS-Random Sample, SHHS-Sleep Heart Health Study, W-Women, PCS-Prospective Cohort Study, CSS-Cross-Sectional Study.

The results in the literature show that compared to a reference sleep duration of 7-8 h, an individual having either short sleep duration ( < 6 h) or long sleep duration ( > 8 h) is at high risk of developing type 2 diabetes. However, [50] , [51] concluded that there is no significant association between sleep and incidence of type 2 diabetes. Mallon et al. [52] studied the impact of gender on the association between sleep and diabetes. The authors concluded that short sleep duration increases the risk of incidence diabetes in men, whereas, in women, long sleep duration dominates. The effect of ethnicity on the association is analyzed by [53] , [54] , [55] . Zizi et al. [53] and Jackson et al. [54] showed that the prevalence of type 2 diabetes is more in whites who sleep less than 5 h or more than 8 -9 h compared to blacks. Beihl [55] showed that the association is more in Hispanics/Non-Hispanic Whites compared to that in African-American. Xu et al. examined the association between day-time napping and type 2 diabetes and showed that an individual taking more than 1 h of day-time nap is at 1.5 times more risk to develop diabetes compared to an individual who does not take a nap during the day. In the context of sleep quality, the risk of incidence type 2 diabetes is more in an individual having difficulty initiating sleep (DIS), and the risk increases with increasing DIS frequency [56] , [57] , [58] . Furthermore, the association is more in women having DIS compared to men [59] .

In summary, there is a strong association between sleep quantity/quality and the incidence of type 2 diabetes. The association is stronger in women sleeping for more duration and in men with short sleep duration. Moreover, this association is affected by ethnicity.

3.3. Smoking

Smoking leads to more than 8 million deaths per year [60] . This is from both active and passive uses, i.e, non-smokers exposed to smokers. Smokers are 30–40% more likely to develop type 2 diabetes compared to non-smokers [61] . When an individual smokes, the level of nicotine increases in his/her body. This leads to a reduction in muscle glucose intake, developing insulin resistance and leading to type 2 diabetes [62] . The characteristics and findings of table:smokingtable:smoking/passive smoking and the incidence of type 2 diabetes are presented in Table 3 .

Characteristics and findings of the studies examining the association between smoking and type 2 diabetes.

ZS-Zutphen Study, NHS-Nurse Health Study, NHIS-National Health Interview Survey, HPFS-Health Professionals’ Follow-up Study, RS-Random Sample, SOF-Study of Osteoporotic Fractures, OHS-Osaka Health Survey, PHS-Physicians Health Survey, BRHS-British Regional Health Study, CPS-Cancer Prevention Study, NCDS-National Child Development Study, RIH-Regional Institute for Health, NTHS-Nord Trondelag Health Survey, IRAS-Insulin Resistance Atherosclerosis Study, ARIC-Atherosclerosis Risk in Communities, KCPS-Korean Cancer Prevention Study, JPHC-Japan Public Health Center, WHI-Women Health Initiative, KMIC-Korean Medical Insurance Corporation, M-Men, W-Women, PCS-Prospective Cohort Study, CSS-Cross-Sectional Study.

The results in the literature show that the association between smoking and diabetes increases with an increase in the number of cigarettes smoked/day. Will et al. [63] analyzed the impact of gender on this association and showed that the association between cigarette smoking and type 2 diabetes is more in men compared to women. Similar results are obtained by Jee et al. [64] . Wannamethee et al. [65] revealed that an individual smoking pipe/cigar is 2.15 times more likely to develop type 2 diabetes and an individual smoking cigarette is 1.6 times more likely compared to a non-smoker. Kowall et al. [66] showed that the risk of incidence type 2 diabetes is significantly high in active/passive prediabetic smokers compared to active/passive smokers without prediabetes.

The incidence and prevalence of type 2 diabetes in ex-smokers is examined by [67] , [68] , [69] , [70] , [71] , [72] . and [73] respectively. Results show that ex-smokers are associated with 17–60% increased risk of type 2 diabetes [67] , [68] , [70] , [71] , [72] . However, the results obtained by Simon et al. [73] and Manson et al. [69] showed no association between ex-smokers and type 2 diabetes. This discrepancy in the results can be due to the heterogeneous characteristics (sample size, age range, men/women ratio and ethnicity) of the cohorts used in these studies. Beziaud et al. [74] examined gender-based prevalence of type 2 diabetes in ex-smokers and showed that women are at higher risk compared to men. Furthermore, the duration of smoking cessation also impacts the association in ex-smokers [65] , [75] , [76] , [77] . An individual is at high risk of developing type 2 diabetes during first 5–10 years of smoking cessation. The risk then decreases with an increase in cessation duration. The association between smoking cessation and the incidence of type 2 diabetes is more in women than men [78] .

In summary, both active and passive smoking are strongly associated with the incidence of type 2 diabetes. The association is more in men compared to women. Moreover, the association remains significant in ex-smokers during first the 5–10 years of smoking. After 10 years of smoking cessation, the risk of incidence type 2 diabetes is the same as that in a non-smoker. Women ex-smokers are at a higher risk of developing diabetes compared to men ex-smokers.

3.4. Depression

Depression is a mood disorder that negatively affects the way a person feels, thinks and acts [130] . It can be due to a family history of depression, early childhood trauma, brain structure, medical conditions, drug use or surrounding environment. Depression is associated with multiple health conditions including diabetes [131] . It elevates the sympathetic nervous system activities and hypothalamic–pituitary–adrenal axis activities [132] . Elevated sympathetic nervous system activities lead to an increase in catecholamines and inflammation, and eventually causing insulin resistance [133] . On the other hand, elevated adrenal axis activities lead to an increase in cortisol and eventually blood sugar level [134] . Both insulin resistance and increased blood sugar levels develop type 2 diabetes. The characteristics and findings of the work in the literature examining the association between depression and the incidence of type 2 diabetes are presented in Table 4 .

Characteristics and findings of the studies examining the association between depression and type 2 diabetes.

RS-Random Sample, SDS-Self rating Depression Scale, ECAPS-Epidemiologic Catchment Area Program Survey, NHANES-National Health and Nutrition Examination Survey, ARIC-Atherosclerosis Risk in Communities, RNH-RegistratieNet Huisarts Praktijken, SWAN-Study of Womens’ Health Across the Nation, NTHS-Nord Trondelag Health Study, CHS-Cardiovascular Health Study, CESD-Center for Epidemiological Studies Depression Scale, RBHCDS-Rancho Bernardo Heart and Chronic Disease Study, BDI-Beck Depression Inventory, M-Men, W-Women, PCS-Prospective Cohort Study, CSS-Cross-Sectional Study.

The results show that depression is highly associated with the incidence of type 2 diabetes. In the context of gender, depressed men are at higher risk of incidence type 2 diabetes, whereas depression in women is not associated with type 2 diabetes [135] . Moreover, compared to Caucasian, Hispanic, Japanese-American and Chinese-American, depressed African-Americans are at 2.56 times higher risk of incidence type 2 diabetes [136] . Based on self rating depression scale (SDS) score, an individual having a score of 48–80 is at higher risk of developing diabetes compared to an individual having a score of 20–39 [137] . Similarly, an individual having a score ⩾ 11 using center for epidemiological studies depression scale (CES-D) or a score ⩾ 8 using beck depression inventory (BDI) is at higher risk of incidence type 2 diabetes [138] , [139] .

In summary, depression is associated with type 2 diabetes. However, the association is different in men and women. Moreover, the study by Yu et al. [140] show that depression itself is not a risk factor for diabetes, rather the activities related to depression such as physical inactivity, poor diet, and obesity lead to diabetes. In addition, the medical drugs used to treat depression also have an association with the incidence of type 2 diabetes. Consequently, similar to high-level serum uric acid, depression is not an independent risk factor but it emphasizes the impact of other independent risk factors such as gender, ethnicity, physical inactivity, and obesity.

3.5. Cardiovascular disease

Increased heart rate and cardiovascular disease can elevate the blood pressure in the arteries. As a result, the body’s glucose uptake decreases leading to insulin resistance condition. Consequently, a person suffering from heart disease is at a higher risk of developing type 2 diabetes. However, this association is still obscure. Few studies argue that a history of cardiovascular disease leads to the incidence of type 2 diabetes [141] , while others claim that type 2 diabetes increases the risk of cardiovascular disease [142] , [143] , [144] . Yeung et al. [141] examined the association between family history of coronary heart disease (CHD) and type 2 diabetes ( Table 5 ). The authors concluded that a high family CHD score is associated to the incidence of type 2 diabetes in individuals who have a positive history of family diabetes. For the individuals having a negative family history of diabetes, this association was non-significant. In summary, it is debatable whether cardiovascular disease is a risk factor for type 2 diabetes or not.

Characteristics and findings of the studies examining the association between cardiovascular disease and type 2 diabetes.

ARIC-Atherosclerosis Risk in Communities, CDH-Coronary Heart Disease, M-Men, W-Women, PCS-Prospective Cohort Study.

3.6. Dyslipidemia

Dyslipidemia refers to an abnormal level of lipids, such as triglycerides and cholesterol. It is characterized by high triglyceride levels, increased low-density lipoproteins (LDL) levels and decreased high-density lipoproteins (HDL) levels [145] . Elevated LDL and lowered HDL levels lead to beta-cell dysfunction inhibiting insulin secretion and consequently type 2 diabetes [146] , [147] . Table 6 shows the characteristics and findings of the work in the literature studying the association between dyslipidemia and type 2 diabetes.

Characteristics and findings of the studies examining the association between dyslipidemia and type 2 diabetes.

LWHS-Lowa Women’s Health Study, CCHS-Copenhagen City heart Study, CGPS-Copenhagen General Population Study, MA-Meta Analysis, REACTION-Risk Evaluation of cAncers in Chinese diabeTic Individuals: a lONgitudinal study, RS-Random Sample, M-Men, W-Women, PCS-Prospective Cohort Study.

Dietary fats, that raise the total cholesterol and LDL levels, are considered significant in the development of type 2 diabetes [148] . Substituting saturated fatty acid with polyunsaturated fatty acid and animal fat with vegetable fat can help lower blood cholesterol and eventually type 2 diabetes. This is because both polyunsaturated fatty acid and vegetable fat are inversely related to the risk of incidence type 2 diabetes with RR 0.84 (95% CI 0.71–0.98) and RR 0.78 (95% CI 0.67–0.91) respectively for the highest quintile of intake [148] . Tajima et al. [149] also confirmed the association between high cholesterol diet intake ( > 273 mg/day) and type 2 diabetes (RR 1.25, 95% CI 1.16–1.36) compared to low cholesterol intake ( < 185 mg/day).

In order to reduce elevated LDL level, LDL lowering therapy and drugs are suggested. However, these drugs and therapy are found to be associated with a higher risk of type 2 diabetes [150] . Individuals having familial hypercholesterolemia, a genetic disorder that results in high LDL levels, are less likely to have type 2 diabetes compared to individuals having high LDL levels due to dietary patterns [151] . Zhang et al. [152] in their analysis found that the ratio of non-HDL and HDL levels is an independent risk factor for incidence diabetes. They show that an individual having a ratio of 3.1 is at 40% increased risk of incidence diabetes (OR 1.4, 95% CI 1.1–1.8) compared to an individual having a ratio of 1.4. Elevated non-HDL and lowered HDL levels are significantly associated with incidence diabetes [153] .

On the contrary to studies confirming the association between low-HDL levels and the incidence of type 2 diabetes, Haase et al. [154] in their study concluded that a life-long reduction in HDL levels are not associated with an increased risk of type 2 diabetes. They found that the association is most likely reverse causation, i.e., type 2 diabetes leads to low HDL levels.

3.7. Hypertension

Hypertension, also known as high blood pressure, is a medical condition in which the blood pressure in the arteries is persistently elevated. Hypertension elevates the sympathetic nervous system activity leading to a decrease in the body’s glucose uptake. This causes the condition of insulin resistance and eventually type 2 diabetes. Hypertension elevates sympathetic nervous system activities leading to impaired vasodilation of skeletal muscles. Consequently, muscle glucose uptake decreases with the eventual development of type 2 diabetes. Table 7 shows the characteristics and findings of the work in the literature studying the association between hypertension and type 2 diabetes.

Characteristics and findings of the studies examining the association between hypertension and type 2 diabetes.

RS-Random Sample, ARIC-Atherosclerosis Risk in Communities, WHS, Women’s Health Study, CARDIA-Coronary Artery Risk Development in Young Adults, FHS-Framingham Heart Study, GPPS-Gothenburg Primary Prevention Study, KGES-Korean Genome and Epidemiology Study, M-Men, W-Women, PCS-Prospective Cohort Study.

Hayashi et al. [166] examined the association between high normal blood pressure ( ⩾ 130 and < 140 mmHg/ ⩾ 85 and < 90) and hypertension ( ⩾ 140 mmHg/ ⩾ 90 mmHg), and the incidence of type 2 diabetes in men. The authors concluded that both high normal blood pressure (RR 1.39, 95%1.14–1.69) and hypertension (RR 1.75, 95% CI 1.43–2.16) are associated with an increased risk of type 2 diabetes. This association is dependent on obesity and hypertension medications. Hypertension medications are considered to increase the risk of diabetes depending on the type of medication [167] . For instance, hypertensive individuals taking thiazide diuretics and angiotensin-converting-enzyme medications are at lower risk of diabetes compared to the hypertensive individuals not taking any medication. However, those taking beta-blockers medication are at 28% higher risk of incidence type 2 diabetes (HR 1.28, 95% CI 1.04–1.57) [167] . The association between hypertension and the incidence of type 2 diabetes is significant in women as well [168] . Women having hypertension are at 2 times increased risk of developing diabetes (HR 2.03, 95% CI 1.77–2.32) compared to women having normal blood pressure ( < 120/75) [168] . The association is more in overweight and obese women. Irrespective of gender, prehypertension (HR 1.27, 95%CI 1.09–1.48) and hypertension (HR 1.51, 95% CI 1.29–1.76) are associated with increased risk of incidence type 2 diabetes [169] . In the context of ethnicity, whites individuals having hypertension are at higher risk of developing diabetes (HR 1.25, 95% CI 1.03–1.53), but no such association is seen in African American hypertensive individuals (HR 0.92, 95% CI 0.70–1.21) [170] .

In summary, hypertension is associated with the development of type 2 diabetes in both men and women. However, the association is ethnicity-dependent. The selection of hypertensive medications should be made properly as the medication impacts the strength of the association. Furthermore, an obese individual with hypertension is at higher risk compared to a non-obese.

The number of elderly people (above 60 years) is increasing worldwide. The 900 million global elderly population in 2015 is expected to rise to 2 billion by 2050 [171] . Aging increases the risk of metabolic syndrome and chronic diseases including type 2 diabetes. Aging increases chronic inflammation in an elderly individual leading to insulin resistance [172] . In addition, lipid metabolism disorder due to aging increases the accumulation of body fat leading to elevated free fatty acids concentration in the blood/plasma and eventually insulin resistance [173] . Consequently, an aged individual is at higher risk of developing type 2 diabetes. However, there is not much work concluding that aging is an independent risk factor for type 2 diabetes. Choi et al. [174] concluded that the risk of diabetes increases with aging only in overweight individuals, and the risk decreases with a moderate level of physical activity. Aging can be considered as triggering the association between independent risk factors and risk of diabetes, but more evidence and studies are required to examine the association between aging as an independent factor and diabetes.

3.9. Ethnicity

Ethnicity is associated with a range of health complications including diabetes because of the heterogeneity in the demographic environmental conditions and lifestyle. It is an independent risk factor which tends to be exacerbated by the social disadvantage and the affluent way of living. Table 8 shows the characteristics and findings of the work in the literature studying the association between ethnicity and type 2 diabetes. Compared to white individuals, type 2 diabetes is more prevalent in Pacific Islanders (OR 3.1, 95% CI 1.4–6.8), followed by Blacks (OR 2.3, 95% CI 2.1–2.6), Native Americans (OR 2.2, 95% CI 1.6–2.9), Hispanics (OR 2.0, 95% CI 1.8–2.3), and Multiracial (OR 1.8, 95% CI 1.5–2.9) [175] . In another study by Shai et al. [176] , it was found that compared to whites, Asians (RR 1.94, 95% CI 1.46–2.58), Hispanics (RR 1.70, 95% CI 1.28–2.26), and Blacks (RR 1.36, 95% CI 1.14–1.63) are at higher risk of incidence type 2 diabetes.

Characteristics and findings of the studies examining the association between ethnicity and type 2 diabetes.

RS-Random Sample, BRFSS-Behavioral Risk Factor Surveillance System, NHS-Nurses’ Health Study, PCS-Prospective Cohort Study.

A study by Zimmet et al. [177] showed that type 2 diabetes is 10 times more prevalent in rural Indians compared to rural Melanesians, and 2 times more prevalent in urban Indians compared to urban Melanesians. They also revealed that the prevalence is 5 times more in urban Melanesians compared to rural Melanesians. One of the reason could be that the rural residents have an increased amount of physical activity compared to the urban ones, leading to decreased risk of diabetes [178] . It should thus important to have a moderate amount of physical activity as a therapy for diabetes prevention. Compared to Europeans, type 2 diabetes is 3.8 times more prevalent in Indians, and the prevalence increases to 5 times for 40–64 years old individuals [179] . In another comparison between Asian and non-Asian ethnicity, it is found that the prevalence of type 2 diabetes in Bangladeshis (Asians) is more [180] . Furthermore, the prevalence is high in women (5.75 times) compared to that in men (2.2 times). However, ethnicity can not be considered as an independent risk factor for this association as Bangladeshis had higher smoking rates and a lower ratio of polyunsaturated fatty acids to saturated fatty acids. Consequently, ethnicity, smoking and dyslipidemia all contributed to the risk of incidence type 2 diabetes. Simmons et al. [181] also confirmed in their study that the prevalence is more in Asians compared to Whites. However, in contrast to the results obtained by [180] , Simmons et al. [181] found that the prevalence is more in men compared to women. This inconsistency should be examined further.

In summary, ethnicity is associated with the incidence of type 2 diabetes. However, there is no definite explanation of why individuals of a particular ethnicity are at higher risk of type 2 diabetes compared to the others. One possible explanation can be the ethnicity-dependent relation between BMI and body fat. For instance, Asians have around 3–4 kg/ m 2 lower BMI compared to Caucasians for a given percentage of body fat [182] . Another reason could be ethnicity-based insulin sensitivity. Studies show that Asians, Blacks and Mexican Americans are less insulin sensitive compared to non-Hispanic Whites [183] , [184] .

3.10. Family history of diabetes

Family history information can serve as a useful tool for prognosis/diagnosis and public health. Family history of diabetes reflects both genetic as well as environmental factors and can lead to better prediction of incidence type 2 diabetes than only genetic factors and environmental factors alone [192] . Table 9 shows the characteristics and findings of the work in the literature studying the association between family history of diabetes and type 2 diabetes.

Characteristics and findings of the studies examining the association between family history of diabetes and type 2 diabetes.

RS-Random Sample, SAHS-San Antonio Heart Study, MRFIT-Multiple Risk Factor Intervention Trial, MA-Meta Analysis, THHP-The Honolulu Heart Program, EPIC-European Prospective Investigation into Cancer, FHS-Framingham Heart Study, MONICA-Multinational MONItoring of trends and determinants in CArdiovascular disease, NHANES-National Health and Nutrition Examination Survey, PD-Prediabetes, IFG-Impaired Fasting Glucose, IGT-Impaired Glucose Tolerance, M-Men, W-Women, PCS-Prospective Cohort Study, CSS-Cross-Sectional Study.

A study by Tsenkova et al. [193] revealed that a family history of diabetes is strongly associated with incidence diabetes (OR 2.77, 95% CI 2.03–3.78). Another study also shows that parental history of diabetes is an independent risk factor for diabetes (OR 1.73, 95% CI 1.29–2.33) [194] . However, the association becomes weaker in men free of cardiovascular disease (OR 1.63, 95% CI 1.18-.2.24). Moreover, the association is much higher in 45–54 years old men (OR 1.99, 95% CI 1.38–2.89) compared to 55–68 years old men (OR 1.33, 95% CI 0.70–2.52). Furthermore, the prevalence of type 2 diabetes is stronger in men compared to women [195] . This indicates that parental history of diabetes in combination with other risk factors such as aging, gender and cardiovascular diseases, increases the risk of incidence type 2 diabetes.

Rodríguez-Moran et al. [196] showed that a family history of diabetes in first degree of relative (parents, offspring and siblings) is a strong and independent risk factor for the prevalence of impaired fasting glucose (prediabetes) (OR 11.7, 95% 9.5–21.2) in children and adolescents. This is in the absence of obesity. The results reveal that is it important to consider the parental history of diabetes while screening for diabetes children and adolescents. This is because only obesity-based screening could lead to underestimation. Valdez et al. [197] also showed that the family history of diabetes in at least two first-degree relatives or one first-degree and at least two second-degree relatives is significant for prevalence of type 2 diabetes. However, it can not be denied that the presence of a family history of diabetes can make the association between obesity and diabetes stronger [198] . Given a BMI ⩾ 35, an individual with a family history of diabetes is at a higher risk of incidence diabetes (OR 26.7, 95% CI 14.4–49.4) compared to the one without a family history of diabetes (OR 6.1, 95% CI3.4–11.2). Furthermore, ethnicity is also considered an important factor in an obese individual with a family history of diabetes [199] , [200] .

An individual having a family history of diabetes can have an early onset of diabetes compared to the ones without a family history. However, it is hard to conclude that which among the maternal, paternal and both maternal and paternal family history of diabetes is more significant for incidence/prevalence of type 2 diabetes as the results in the literature are inconsistent [195] , [201] , [202] , [203] , [204] , [205] .

3.11. Obesity

Obesity is a complex health condition that involves an excessive amount of body fat. It is defined by the BMI and further evaluated in terms of fat distribution via the waist-hip ratio. Abdominal fat in the body increases inflammation which decreases insulin sensitivity by disrupting the function of beta-cells. The insulin resistance condition then leads to the prevalence of type 2 diabetes. Table 10 shows the characteristics and findings of the work in the literature studying the association between obesity and type 2 diabetes.

Characteristics and findings of the studies examining the association between obesity and type 2 diabetes.

SWHS-Shanghai Women’s Health Study, BWHS- Black Women’s Health Study, RS-Random Sample, WC-Waist Circumference, TLGS-Tehran Lipid and Glucose Study, MAHES-Massachusetts Hispanic Elderly Study, H-Hispanics, NH-Non Hispanics, M-Men, W-Women, PCS-Prospective Cohort Study, CSS-Cross-Sectional Study.

Ishikawa-Takata et al. [206] found that the risk of diabetes increases significantly for an individual having a BMI greater than 29 kg/ m 2 . The relative risk of diabetes increases up to 38.8 (95% CI 31.9–47.2) for an individual having a BMI greater than 34.9 kg/ m 2 [119] . Furthermore, study shows that the association between obesity and incidence diabetes is gender-dependent [207] . For each 2 kg/ m 2 lower BMI, men are at 23% (15–30%) lower risk of diabetes, whereas women are at 27% (23–32%) lower risk. Further, the association between obesity and diabetes is also dependent on ethnicity [207] . For each 2 kg/ m 2 lower BMI, Asians are at 37% (26–46%) lower risk of diabetes, whereas Australians are at 25% (21–29%) lower risk.

Ohnishi et al. [208] found that compared to overall obesity, central obesity is highly associated with the risk of type 2 diabetes (RR 2.07, 95% CI 1.03–4.16). This association is more in elderly people ( ⩾ 60 years) (OR 3.8, 95% CI 1.8–7.7) [209] . The association between central obesity and the incidence of type 2 diabetes is found significant in both men and women. However, centrally obese women are at higher risk (OR 2.875, 95% CI 1.987–4.160) compared to centrally obese men (OR 2.308, 95% CI 1.473–3.615) [210] . The prevalence of type 2 diabetes in obese individual is ethnicity dependent [211] . Non-Hispanics centrally obese women are at higher risk of developing type 2 diabetes (OR 15.1, 95% CI 1.9–117.6) compared to centrally obese Hispanic women (OR 1.6, 95% CI 1.0–2.8). The centrally Hispanic men are also at risk of developing type 2 diabetes (OR 2.1, 95% CI 1.2–3.9). No such association is found in centrally obese Non-Hispanic men. However, all these studies examining the association between central obesity and the incidence of type 2 diabetes consider different definitions of central obesity. For instance, [208] defines central obesity as waist circumference (WC) ⩾ 85 cm in men and ⩾ 90 cm in women, whereas [211] defines it as WC > 102 cm in men and > 88 cm in women. Consequently, it is difficult to conclude the association between central obesity and the incidence of type 2 diabetes.

In summary, although obesity is a significant predictor, the association between obesity and diabetes is a factor of gender and ethnicity. Women with high BMI are at greater risk of diabetes compared to men. Moreover, the association is stronger in Asians compared to Australians. The association between central obesity is also found to be significant for the prevalence of type 2 diabetes. This association is the strongest in Non-Hispanics women. However, more studies are required to examine the association between central obesity and type 2 diabetes following one standard criterion defining central obesity.

3.12. Physical inactivity

An individual is considered physically inactive if he/she does not get the recommended 30–60 min of exercise three to four times a week. Physical inactivity decreases insulin sensitivity with progressive loss of beta-cells. This leads to impaired glucose tolerance and eventually type 2 diabetes. However, no work examines the association between physical inactivity as an independent factor and the prevalence of diabetes. One of the reasons that physical inactivity leads to type 2 diabetes can be that physical inactivity can cause obesity which in turn is a significant risk factor for type 2 diabetes.

4. Conclusion

Diabetes is a global crisis that is primarily driven by rapid urbanization, changing lifestyles, and uneven dietary patterns [215] , [216] . It is crucial to predict the prevalence of diabetes in an individual to reduce the risk of diabetes development and save lives. Diabetes is thought to prevail due to several risk factors such as high-level serum uric acid, sleep quality/quantity, smoking, depression, cardiovascular disease, dyslipidemia, hypertension, aging, ethnicity, family history of diabetes, physical inactivity, and obesity. Studies in the literature have examined the association between each of these risk factors and the risk of developing type 2 diabetes. In this review, we provide an analysis of the studies in the literature to deduce inferences on the relationship between the risk factors and incidence/prevalence of type 2 diabetes.

In conclusion, it can be observed that sleep quantity/quality, smoking, dyslipidemia, hypertension, ethnicity, family history of diabetes, obesity and physical inactivity are strongly associated with the development of type 2 diabetes. Both sleep quantity and quality are found to be strongly associated with the development of type 2 diabetes. The association is stronger in women sleeping for more hours and in men sleeping for fewer hours. However, the sleeping quantity and quality data in these studies are self-reported by the participants, and therefore, prone to errors. More studies are required that use measurement techniques for data collection to validate the association between sleep quantity/quality and type 2 diabetes. Smoking is also found to be a significant risk factor for type 2 diabetes. Both active and passive smokers are at higher risk of developing type 2 diabetes. Moreover, the risk for developing type 2 diabetes remains high in ex-smokers for the first 5–10 years of smoking cessation. Dyslipidemia is associated with the development of type 2 diabetes. Increased non-HDL and decreased HDL levels are strongly associated with type 2 diabetes. However, in the majority of these studies, the incidence or prevalence of type 2 diabetes is self-reported. Consequently, further studies are needed to validate this association between dyslipidemia and type 2 diabetes using standardized measurement techniques, such as A1C test [217] . Hypertension is a significant risk factor for type 2 diabetes and this is further elevated in obese individuals. Ethnicity strongly associates with the development of type 2 diabetes. This could be due to the fact that insulin sensitivity varies among individuals of different ethnicity. Family history of diabetes in first degree of relatives is strongly associated with the development of type 2 diabetes. In addition, family history of diabetes also signifies the association between obesity and type 2 diabetes. Obesity is found to a significant risk factor for incidence of type 2 diabetes and the association is stronger in women compared to men.

The association between serum uric acid and type 2 diabetes remains obscure. It can not be concluded that serum uric acid is an independent risk factor for type 2 diabetes or it only elevates the association between other independent risk factors such as obesity, hypertension, and dyslipidemia, and type 2 diabetes. Moreover, our analysis shows that there might be no association between serum uric acid and the development of type 2 diabetes, but rather there might be a reverse association, i.e., diabetes leads to elevated serum uric acid level. Similarly, based on the evidence in the literature, aging can not be considered as an independent risk factor for type 2 diabetes. Aging only emphasizes the association between obesity and type 2 diabetes. Depression as well is not found to an independent risk factor contributing to the development of type 2 diabetes. Rather, the activities related to depression such as physical inactivity, poor diet, and obesity leads to diabetes. There is no sufficient evidence to conclude the association between cardiovascular disease and type 2 diabetes. It is debatable whether cardiovascular disease leads to the development of type 2 diabetes. Consequently, more studies are required to study the direct association between these risk factors, i.e., serum uric acid, aging, depression, and cardiovascular disease, and incidence of type 2 diabetes.

Based on this study, we devise recommendations to different stakeholders leading to better patient care. In particular, we provide recommendations for allied healthcare professionals, individuals, and government institutions as follows:

• • Allied healthcare professionals: The hypertensive medications and the LDL lowering therapy and drugs should be carefully prescribed as they are associated with increased risk of type 2 diabetes. In addition, overweight and obese adults should be screened for diabetes.
• • Individuals: A healthy lifestyle, which involves intake of polyunsaturated fatty acids and vegetable fats, regular exercise, a healthy diet and proper sleep, is crucial. Individuals should avoid both active and passive smoking.
• • Government: Physical activity in the nation should be promoted for a healthy nation. Law policies should be implemented to restrict public smoking as passive smoking significantly increases the risk of type 2 diabetes. For instance, designated smoking areas can be established to eliminate the risk of developing passive smokers. It would be beneficial to have periodic surveys that include the demographic and lifestyle features of the citizens and the surveys’ results can be then used to develop a nation-wide diabetes prevention plan, in coordination with the allied health professionals.

CRediT authorship contribution statement

Leila Ismail: Conceptualization, Methodology, Investigation, Writing - original draft, Writing - review & editing. Huned Materwala: Investigation, Writing - original draft. Juma Al Kaabi: Validation, Writing - review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This work is funded by the National Water and Energy Center of the United Arab Emirates University (Grant No. 31R215). We thank the anonymous reviewers for their valuable comments which helped us improve the paper.

Appendix A. 

Search string used to retrieve the studies on the association between risk factor and type 2 diabetes.

Quality assessment of the included studies according to the Quality assessment tool for observational cohort and cross-sectional studies.

Q1. Was the research question or objective in this paper clearly stated?.

Q2. Was the study population clearly specified and defined?.

Q3. Was the participation rate of eligible persons at least 50%?

Q4. Were all the subjects selected or recruited from the same or similar populations (including the same time period)? Were inclusion and exclusion criteria for being in the study prespecified and applied uniformly to all participants?.

Q5. Was a sample size justification, power description, or variance and effect estimates provided?.

Q6. For the analyses in this paper, were the exposure(s) of interest measured prior to the outcome(s) being measured?.

Q7. Was the timeframe sufficient so that one could reasonably expect to see an association between exposure and outcome if it existed?.

Q8. For exposures that can vary in amount or level, did the study examine different levels of the exposure as related to the outcome (e.g., categories of exposure, or exposure measured as continuous variable)?.

Q9. Were the exposure measures (independent variables) clearly defined, valid, reliable, and implemented consistently across all study participants?.

Q10. Was the exposure(s) assessed more than once over time?.

Q11. Were the outcome measures (dependent variables) clearly defined, valid, reliable, and implemented consistently across all study participants?.

Q12. Were the outcome assessors blinded to the exposure status of participants?.

Q13. Was loss to follow-up after baseline 20% or less?.

Q14. Were key potential confounding variables measured and adjusted statistically for their impact on the relationship between exposure(s) and outcome(s)?.

CD-Cannot be Determined; NA-Not Applicable; NR-Not Reported.

• Previous Article
• Next Article

Research Design and Methods

Article information, literature review of type 2 diabetes management and health literacy.

• Split-Screen
• Article contents
• Figures & tables
• Supplementary Data
• Peer Review
• Open the PDF for in another window
• Cite Icon Cite
• Get Permissions

Rulla Alsaedi , Kimberly McKeirnan; Literature Review of Type 2 Diabetes Management and Health Literacy. Diabetes Spectr 1 November 2021; 34 (4): 399–406. https://doi.org/10.2337/ds21-0014

Download citation file:

• Ris (Zotero)
• Reference Manager

The purpose of this literature review was to identify educational approaches addressing low health literacy for people with type 2 diabetes. Low health literacy can lead to poor management of diabetes, low engagement with health care providers, increased hospitalization rates, and higher health care costs. These challenges can be even more profound among minority populations and non-English speakers in the United States.

A literature search and standard data extraction were performed using PubMed, Medline, and EMBASE databases. A total of 1,914 articles were identified, of which 1,858 were excluded based on the inclusion criteria, and 46 were excluded because of a lack of relevance to both diabetes management and health literacy. The remaining 10 articles were reviewed in detail.

Patients, including ethnic minorities and non-English speakers, who are engaged in diabetes education and health literacy improvement initiatives and ongoing follow-up showed significant improvement in A1C, medication adherence, medication knowledge, and treatment satisfaction. Clinicians considering implementing new interventions to address diabetes care for patients with low health literacy can use culturally tailored approaches, consider ways to create materials for different learning styles and in different languages, engage community health workers and pharmacists to help with patient education, use patient-centered medication labels, and engage instructors who share cultural and linguistic similarities with patients to provide educational sessions.

This literature review identified a variety of interventions that had a positive impact on provider-patient communication, medication adherence, and glycemic control by promoting diabetes self-management through educational efforts to address low health literacy.

Diabetes is the seventh leading cause of death in the United States, and 30.3 million Americans, or 9.4% of the U.S. population, are living with diabetes ( 1 , 2 ). For successful management of a complicated condition such as diabetes, health literacy may play an important role. Low health literacy is a well-documented barrier to diabetes management and can lead to poor management of medical conditions, low engagement with health care providers (HCPs), increased hospitalizations, and, consequently, higher health care costs ( 3 – 5 ).

The Healthy People 2010 report ( 6 ) defined health literacy as the “degree to which individuals have the capacity to obtain, process, and understand basic health information and services needed to make appropriate health decisions.” Diabetes health literacy also encompasses a wide range of skills, including basic knowledge of the disease state, self-efficacy, glycemic control, and self-care behaviors, which are all important components of diabetes management ( 3 – 5 , 7 ). According to the Institute of Medicine’s Committee on Health Literacy, patients with poor health literacy are twice as likely to have poor glycemic control and were found to be twice as likely to be hospitalized as those with adequate health literacy ( 8 ). Associations between health literacy and health outcomes have been reported in many studies, the first of which was conducted in 1995 in two public hospitals and found that many patients had inadequate health literacy and could not perform the basic reading tasks necessary to understand their treatments and diagnoses ( 9 ).

Evaluation of health literacy is vital to the management and understanding of diabetes. Several tools for assessing health literacy have been evaluated, and the choice of which to use depends on the length of the patient encounter and the desired depth of the assessment. One widely used literacy assessment tool, the Test of Functional Health Literacy in Adults (TOFHLA), consists of 36 comprehension questions and four numeric calculations ( 10 ). Additional tools that assess patients’ reading ability include the Rapid Estimate of Adult Literacy in Medicine (REALM) and the Literacy Assessment for Diabetes. Tests that assess diabetes numeracy skills include the Diabetes Numeracy Test, the Newest Vital Sign (NVS), and the Single-Item Literacy Screener (SILS) ( 11 ).

Rates of both diabetes and low health literacy are higher in populations from low socioeconomic backgrounds ( 5 , 7 , 12 ). People living in disadvantaged communities face many barriers when seeking health care, including inconsistent housing, lack of transportation, financial difficulties, differing cultural beliefs about health care, and mistrust of the medical professions ( 13 , 14 ). People with high rates of medical mistrust tend to be less engaged in their care and to have poor communication with HCPs, which is another factor HCPs need to address when working with their patients with diabetes ( 15 ).

The cost of medical care for people with diabetes was $327 billion in 2017, a 26% increase since 2012 ( 1 , 16 ). Many of these medical expenditures are related to hospitalization and inpatient care, which accounts for 30% of total medical costs for people with diabetes ( 16 ).

People with diabetes also may neglect self-management tasks for various reasons, including low health literacy, lack of diabetes knowledge, and mistrust between patients and HCPs ( 7 , 15 ).

These challenges can be even more pronounced in vulnerable populations because of language barriers and patient-provider mistrust ( 17 – 19 ). Rates of diabetes are higher among racial and ethnic minority groups; 15.1% of American Indians and Alaskan Natives, 12.7% of Non-Hispanic Blacks, 12.1% of Hispanics, and 8% of Asian Americans have diagnosed diabetes, compared with 7.4% of non-Hispanic Whites ( 1 ). Additionally, patient-provider relationship deficits can be attributed to challenges with communication, including HCPs’ lack of attention to speaking slowly and clearly and checking for patients’ understanding when providing education or gathering information from people who speak English as a second language ( 15 ). White et al. ( 15 ) demonstrated that patients with higher provider mistrust felt that their provider’s communication style was less interpersonal and did not feel welcome as part of the decision-making process.

To the authors’ knowledge, there is no current literature review evaluating interventions focused on health literacy and diabetes management. There is a pressing need for such a comprehensive review to provide a framework for future intervention design. The objective of this literature review was to gather and summarize studies of health literacy–based diabetes management interventions and their effects on overall diabetes management. Medication adherence and glycemic control were considered secondary outcomes.

Search Strategy

A literature review was conducted using the PubMed, Medline, and EMBASE databases. Search criteria included articles published between 2015 and 2020 to identify the most recent studies on this topic. The search included the phrases “diabetes” and “health literacy” to specifically focus on health literacy and diabetes management interventions and was limited to original research conducted in humans and published in English within the defined 5-year period. Search results were exported to Microsoft Excel for evaluation.

Study Selection

Initial screening of the articles’ abstracts was conducted using the selection criteria to determine which articles to include or exclude ( Figure 1 ). The initial search results were reviewed for the following inclusion criteria: original research (clinical trials, cohort studies, and cross-sectional studies) conducted in human subjects with type 2 diabetes in the United States, and published in English between 2015 and 2020. Articles were considered to be relevant if diabetes was included as a medical condition in the study and an intervention was made to assess or improve health literacy. Studies involving type 1 diabetes or gestational diabetes and articles that were viewpoints, population surveys, commentaries, case reports, reviews, or reports of interventions conducted outside of the United States were excluded from further review. The criteria requiring articles to be from the past 5 years and from the United States were used because of the unique and quickly evolving nature of the U.S. health care system. Articles published more than 5 years ago or from other health care systems may have contributed information that was not applicable to or no longer relevant for HCPs in the United States. Articles were screened and reviewed independently by both authors. Disagreements were resolved through discussion to create the final list of articles for inclusion.

PRISMA diagram of the article selection process.

Data Extraction

A standard data extraction was performed for each included article to obtain information including author names, year of publication, journal, study design, type of intervention, primary outcome, tools used to assess health literacy or type 2 diabetes knowledge, and effects of intervention on overall diabetes management, glycemic control, and medication adherence.

A total of 1,914 articles were collected from a search of the PubMed, MEDLINE, and EMBASE databases, of which 1,858 were excluded based on the inclusion and exclusion criteria. Of the 56 articles that met criteria for abstract review, 46 were excluded because of a lack of relevance to both diabetes management and health literacy. The remaining 10 studies identified various diabetes management interventions, including diabetes education tools such as electronic medication instructions and text message–based interventions, technology-based education videos, enhanced prescription labels, learner-based education materials, and culturally tailored interventions ( 15 , 20 – 28 ). Figure 1 shows the PRISMA diagram of the article selection process, and Table 1 summarizes the findings of the article reviews ( 15 , 20 – 28 ).

Findings of the Article Reviews (15,20–28)

SAHLSA, Short Assessment of Health Literacy for Spanish Adults.

Medical mistrust and poor communication are challenging variables in diabetes education. White et al. ( 15 ) examined the association between communication quality and medical mistrust in patients with type 2 diabetes. HCPs at five health department clinics received training in effective health communication and use of the PRIDE (Partnership to Improve Diabetes Education) toolkit in both English and Spanish, whereas control sites were only exposed to National Diabetes Education Program materials without training in effective communication. The study evaluated participant communication using several tools, including the Communication Assessment Tool (CAT), Interpersonal Processes of Care (IPC-18), and the Short Test of Functional Health Literacy in Adults (s-TOFHLA). The authors found that higher levels of mistrust were associated with lower CAT and IPC-18 scores.

Patients with type 2 diabetes are also likely to benefit from personalized education delivery tools such as patient-centered labeling (PCL) of prescription drugs, learning style–based education materials, and tailored text messages ( 24 , 25 , 27 ). Wolf et al. ( 27 ) investigated the use of PCL in patients with type 2 diabetes and found that patients with low health literacy who take medication two or more times per day have higher rates of proper medication use when using PCL (85.9 vs. 77.4%, P = 0.03). The objective of the PCL intervention was to make medication instructions and other information on the labels easier to read to improve medication use and adherence rates. The labels incorporated best-practice strategies introduced by the Institute of Medicine for the Universal Medication Schedule. These strategies prioritize medication information, use of larger font sizes, and increased white space. Of note, the benefits of PCL were largely seen with English speakers. Spanish speakers did not have substantial improvement in medication use or adherence, which could be attributed to language barriers ( 27 ).

Nelson et al. ( 25 ) analyzed patients’ engagement with an automated text message approach to supporting diabetes self-care activities in a 12-month randomized controlled trial (RCT) called REACH (Rapid Education/Encouragement and Communications for Health) ( 25 ). Messages were tailored based on patients’ medication adherence, the Information-Motivation-Behavioral Skills model of health behavior change, and self-care behaviors such as diet, exercise, and self-monitoring of blood glucose. Patients in this trial were native English speakers, so further research to evaluate the impact of the text message intervention in patients with limited English language skills is still needed. However, participants in the intervention group reported higher engagement with the text messages over the 12-month period ( 25 ).

Patients who receive educational materials based on their learning style also show significant improvement in their diabetes knowledge and health literacy. Koonce et al. ( 24 ) developed and evaluated educational materials based on patients’ learning style to improve health literacy in both English and Spanish languages. The materials were made available in multiple formats to target four different learning styles, including materials for visual learners, read/write learners, auditory learners, and kinesthetic learners. Spanish-language versions were also available. Researchers were primarily interested in measuring patients’ health literacy and knowledge of diabetes. The intervention group received materials in their preferred learning style and language, whereas the control group received standard of care education materials. The intervention group showed significant improvement in diabetes knowledge and health literacy, as indicated by Diabetes Knowledge Test (DKT) scores. More participants in the intervention group reported looking up information about their condition during week 2 of the intervention and showed an overall improvement in understanding symptoms of nerve damage and types of food used to treat hypoglycemic events. However, the study had limited enrollment of Spanish speakers, making the applicability of the results to Spanish-speaking patients highly variable.

Additionally, findings by Hofer et al. ( 22 ) suggest that patients with high A1C levels may benefit from interventions led by community health workers (CHWs) to bridge gaps in health literacy and equip patients with the tools to make health decisions. In this study, Hispanic and African American patients with low health literacy and diabetes not controlled by oral therapy benefited from education sessions led by CHWs. The CHWs led culturally tailored support groups to compare the effects of educational materials provided in an electronic format (via iDecide) and printed format on medication adherence and self-efficacy. The study found increased adherence with both formats, and women, specifically, had a significant increase in medication adherence and self-efficacy. One of the important aspects of this study was that the CHWs shared cultural and linguistic characteristics with the patients and HCPs, leading to increased trust and satisfaction with the information presented ( 22 ).

Kim et al. ( 23 ) found that Korean-American participants benefited greatly from group education sessions that provided integrated counseling led by a team of nurses and CHW educators. The intervention also had a health literacy component that focused on enhancing skills such as reading food package labels, understanding medical terminology, and accessing health care services. This intervention led to a significant reduction of 1–1.3% in A1C levels in the intervention group. The intervention established the value of collaboration between CHW educators and nurses to improve health information delivery and disease management.

A collaboration between CHW educators and pharmacists was also shown to reinforce diabetes knowledge and improve health literacy. Sharp et al. ( 26 ) conducted a cross-over study in four primary care ambulatory clinics that provided care for low-income patients. The study found that patients with low health literacy had more visits with pharmacists and CHWs than those with high health literacy. The CHWs provided individualized support to reinforce diabetes self-management education and referrals to resources such as food, shelter, and translation services. The translation services in this study were especially important for building trust with non-English speakers and helping patients understand their therapy. Similar to other studies, the CHWs shared cultural and linguistic characteristics with their populations, which helped to overcome communication-related and cultural barriers ( 23 , 26 ).

The use of electronic tools or educational videos yielded inconclusive results with regard to medication adherence. Graumlich et al. ( 20 ) implemented a new medication planning tool called Medtable within an electronic medical record system in several outpatient clinics serving patients with type 2 diabetes. The tool was designed to organize medication review and patient education. Providers can use this tool to search for medication instructions and actionable language that are appropriate for each patient’s health literacy level. The authors found no changes in medication knowledge or adherence, but the intervention group reported higher satisfaction. On the other hand, Yeung et al. ( 28 ) showed that pharmacist-led online education videos accessed using QR codes affixed to the patients’ medication bottles and health literacy flashcards increased patients’ medication adherence in an academic medical hospital.

Goessl et al. ( 21 ) found that patients with low health literacy had significantly higher retention of information when receiving evidence-based diabetes education through a DVD recording than through an in-person group class. This 18-month RCT randomized participants to either the DVD or in-person group education and assessed their information retention through a teach-back strategy. The curriculum consisted of diabetes prevention topics such as physical exercise, food portions, and food choices. Participants in the DVD group had significantly higher retention of information than those in the control (in-person) group. The authors suggested this may have been because participants in the DVD group have multiple opportunities to review the education material.

Management of type 2 diabetes remains a challenge for HCPs and patients, in part because of the challenges discussed in this review, including communication barriers between patients and HCPs and knowledge deficits about medications and disease states ( 29 ). HCPs can have a positive impact on the health outcomes of their patients with diabetes by improving patients’ disease state and medication knowledge.

One of the common themes identified in this literature review was the prevalence of culturally tailored diabetes education interventions. This is an important strategy that could improve diabetes outcomes and provide an alternative approach to diabetes self-management education when working with patients from culturally diverse backgrounds. HCPs might benefit from using culturally tailored educational approaches to improve communication with patients and overcome the medical mistrust many patients feel. Although such mistrust was not directly correlated with diabetes management, it was noted that patients who feel mistrustful tend to have poor communication with HCPs ( 20 ). Additionally, Latino/Hispanic patients who have language barriers tend to have poor glycemic control ( 19 ). Having CHWs work with HCPs might mitigate some patient-provider communication barriers. As noted earlier, CHWs who share cultural and linguistic characteristics with their patient populations have ongoing interactions and more frequent one-on-one encounters ( 12 ).

Medication adherence and glycemic control are important components of diabetes self-management, and we noted that the integration of CHWs into the diabetes health care team and the use of simplified medication label interventions were both successful in improving medication adherence ( 23 , 24 ). The use of culturally tailored education sessions and the integration of pharmacists and CHWs into the management of diabetes appear to be successful in reducing A1C levels ( 12 , 26 ). Electronic education tools and educational videos alone did not have an impact on medication knowledge or information retention in patients with low health literacy, but a combination of education tools and individualized sessions has the potential to improve diabetes medication knowledge and overall self-management ( 20 , 22 , 30 ).

There were several limitations to our literature review. We restricted our search criteria to articles published in English and studies conducted within the United States to ensure that the results would be relevant to U.S. HCPs. However, these limitations may have excluded important work on this topic. Additional research expanding this search beyond the United States and including articles published in other languages may demonstrate different outcomes. Additionally, this literature review did not focus on A1C as the primary outcome, although A1C is an important indicator of diabetes self-management. A1C was chosen as the method of evaluating the impact of health literacy interventions in patients with diabetes, but other considerations such as medication adherence, impact on comorbid conditions, and quality of life are also important factors.

The results of this work show that implementing health literacy interventions to help patients manage type 2 diabetes can have beneficial results. However, such interventions can have significant time and monetary costs. The potential financial and time costs of diabetes education interventions were not evaluated in this review and should be taken into account when designing interventions. The American Diabetes Association estimated the cost of medical care for people with diabetes to be $327 billion in 2017, with the majority of the expenditure related to hospitalizations and nursing home facilities ( 16 ). Another substantial cost of diabetes that can be difficult to measure is treatment for comorbid conditions and complications such as cardiovascular and renal diseases.

Interventions designed to address low health literacy and provide education about type 2 diabetes could be a valuable asset in preventing complications and reducing medical expenditures. Results of this work show that clinicians who are considering implementing new interventions may benefit from the following strategies: using culturally tailored approaches, creating materials for different learning styles and in patients’ languages, engaging CHWs and pharmacists to help with patient education, using PCLs for medications, and engaging education session instructors who share patients’ cultural and linguistic characteristics.

Diabetes self-management is crucial to improving health outcomes and reducing medical costs. This literature review identified interventions that had a positive impact on provider-patient communication, medication adherence, and glycemic control by promoting diabetes self-management through educational efforts to address low health literacy. Clinicians seeking to implement diabetes care and education interventions for patients with low health literacy may want to consider drawing on the strategies described in this article. Providing culturally sensitive education that is tailored to patients’ individual learning styles, spoken language, and individual needs can improve patient outcomes and build patients’ trust.

Duality of Interest

No potential conflicts of interest relevant to this article were reported.

Author Contributions

Both authors conceptualized the literature review, developed the methodology, analyzed the data, and wrote, reviewed, and edited the manuscript. R.A. collected the data. K.M. supervised the review. K.M. is the guarantor of this work and, as such, has full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Prior Presentation

Portions of this research were presented at the Washington State University College of Pharmacy and Pharmaceutical Sciences Honors Research Day in April 2019.

Email alerts

• Online ISSN 1944-7353
• Print ISSN 1040-9165
• Diabetes Care
• Clinical Diabetes
• Diabetes Spectrum
• Standards of Medical Care in Diabetes
• Scientific Sessions Abstracts
• BMJ Open Diabetes Research & Care
• ShopDiabetes.org
• ADA Professional Books

Clinical Compendia

• Clinical Compendia Home
• Latest News
• DiabetesPro SmartBrief
• Special Collections
• DiabetesPro®
• Diabetes Food Hub™
• Insulin Affordability
• Know Diabetes By Heart™
• About the ADA
• Journal Policies
• For Reviewers
• Advertising in ADA Journals
• Reprints and Permission for Reuse
• Copyright Notice/Public Access Policy
• ADA Professional Membership
• ADA Member Directory
• Diabetes.org
• X (Twitter)
• Cookie Policy
• Accessibility
• Terms & Conditions
• Get Adobe Acrobat Reader
• © Copyright American Diabetes Association

This Feature Is Available To Subscribers Only

Sign In or Create an Account

• Open access
• Published: 13 July 2020

Screening strategies for adults with type 2 diabetes mellitus: a systematic review protocol

• Helen Mearns   ORCID: orcid.org/0000-0002-7971-1256 1 , 2 ,
• Paul Kuodi Otiku 1 , 2 ,
• Mary Shelton 3 ,
• Tamara Kredo 4 , 5 ,
• Benjamin M. Kagina 1 , 2 &
• Bey-Marrié Schmidt 4  

Systematic Reviews volume  9 , Article number:  156 ( 2020 ) Cite this article

6546 Accesses

4 Citations

9 Altmetric

Metrics details

There is limited evidence on whether screening for type 2 diabetes mellitus affects health outcomes. A recent systematic review of randomised clinical trials found only one trial that met their inclusion criteria; therefore, current guidelines for screening interventions for type 2 diabetes mellitus are based on expert opinions and best practice rather than synthesised evidence. This systematic review seeks to collate evidence from non-randomised studies to investigate the effect of screening for adults with type 2 diabetes on outcomes including diabetes-related morbidity, mortality (all-cause and diabetes-related) and harms.

This systematic review will follow Effective Practice and Organisation of Care (EPOC) guidelines for the synthesis of non-randomised studies. We will search PubMed/MEDLINE, Scopus, Web of Science, CINAHL, Academic Search Premier and Health Source Nursing Academic (from inception onwards). We will include non-randomised trials, controlled before-after studies, interrupted time-series studies, repeated measures studies and concurrently controlled prospective cohort studies. The primary outcome will be diabetes-related morbidity (microvascular complications of diabetic retinopathy, nephropathy or neuropathy or macrovascular complications of non-fatal myocardial infarction, peripheral arterial disease or non-fatal stroke). The secondary outcomes will be mortality (all-cause and diabetes-related) and harms of screening strategies to patients (including psychological harms or adverse events following treatments) or to health care system (including resource allocation for false-positives or overdiagnosis). Two reviewers will independently screen all citations and full-text articles. Data will be abstracted by one reviewer and checked by a second. The risk of bias of individual studies will be appraised using the ROBINS-I tool. GRADE will be used to determine the quality of the scientific evidence. If feasible, we will conduct random effects meta-analysis where appropriate. If necessary, analyses will be conducted to explore the potential sources of heterogeneity (e.g. age, sex, socio-economic status, rural versus urban or low-middle income versus high-income country). We will disseminate the findings via publications and through relevant networks.

The protocol outlines the methods for systematically reviewing and synthesising evidence of screening strategies for type 2 diabetes mellitus and their effect on health outcomes associated with the disease. The potential impact of this systematic review is improved evidence-informed decision-making for policies and practice for screening of type-2 diabetes.

Systematic review registration

PROSPERO CRD42020147439

Peer Review reports

Description of the condition

Diabetes mellitus is a disease of increasing global concern. The global prevalence of diabetes was approximately 425 million people in 2017, approximately 8.5% of the adult population, and is expected to double by 2045 [ 1 ]. In high-income countries, type 2 diabetes mellitus accounts for approximately 90% of diabetes cases; there is insufficient data to estimate the ratio of type 2 diabetes mellitus in low- and middle- income countries, but it is assumed to be similar [ 1 , 2 ]. Clinical diabetes is diagnosed through the detection of elevated levels of glucose in the blood (hyperglycemia) [ 3 ]; however, it is estimated that half of the people who have diabetes are not diagnosed [ 1 ].

In addition to those individuals who have clinical diabetes, another 352 million, approximately 7.3% of the adult population, have intermediate blood glucose levels that are considered in between normal and clinically diagnosed diabetes [ 1 , 3 ]. These intermediate blood glucose levels perform as a risk score, where increasing values are associated with an increasing likelihood of progression to diabetes, cardiovascular disease and all-cause mortality [ 2 , 4 ]. Patients who present with intermediate levels of blood glucose are described using a number of terminologies including mild glucose intolerance, non-diabetic hyperglycaemia and prediabetes. The terminology promoted by the World Health Organization (WHO) is impaired glucose tolerance (IGT), impaired fasting glucose (IFG) and intermediate hyperglycaemia [ 3 , 5 ]. The term prediabetes is gaining in popularity even though the WHO has warned its use may lead to disease stigma and detract from the significant cardiovascular risk of this population [ 5 ]. About a third of people with IGT and IFG are young, aged between 20–39 years, meaning they will spend many years at risk of developing diabetes [ 1 ]. Other risk factors, apart from intermediate glucose levels, for the development of diabetes are increasing age of more than 45 years and obesity [ 2 ].

Type 2 diabetes mellitus arises due to defective insulin activity in body tissues, defective insulin secretion from pancreas or a combination of the two [ 2 ]. Type 2 diabetes mellitus usually occurs in older adults, but with a change in lifestyle factors, such as inactivity and obesity, the condition is increasingly being detected in children, adolescents and young adults [ 1 , 2 ]. Current management of type 2 diabetes mellitus involves lifestyle modification: increasing physical activity, improving diet, reaching a healthy body weight and stopping smoking, all monitored by regular screening [ 2 ]. If lifestyle modification does not result in sufficiently decreased blood glucose levels, medication may be prescribed, of which there are a range of treatment options available [ 2 ]. The complication with type 2 diabetes mellitus is the long latency period, often lasting several years, during which time the individual is often asymptomatic and unaware of their condition [ 1 , 2 ]. This prolonged asymptomatic state results in long-term damage to the body’s organs that leads to negative health outcomes including pregnancy complications, oral health problems, disabilities such as blindness, reduced wound healing, foot disease that may require amputation, stroke, heart and kidney disease and death [ 1 , 2 , 3 ].

Description of the intervention

There are many types of screening interventions and strategies that may be used to detect disease in a population often classified as mass, opportunistic and targeted strategies—as presented in Table 1 [ 2 , 6 ]. This systematic review will use these classifications, but if additional strategies are noted, these too will be included.

The biochemical tests commonly used are fasting plasma glucose (FPG), oral glucose tolerance test (OGTT) and detection of glycated haemoglobin A1C (HbA1c) although there are also urine glucose tests available or random blood glucose tests [ 2 , 6 ]. In addition, there are a number of risk scores [ 7 , 8 ], including the Finnish Diabetes Risk Score (FINDRISC) [ 9 ] and the American Diabetes Association’s risk test [ 10 ]; however, these are not commonly used as stand-alone screening tools. Classification of patients post testing can be termed as in the normal range or as having diabetes, impaired glucose tolerance (IGT) or impaired fasting glucose (IFG) (as presented in Table 2 ) [ 1 , 3 ].

How the intervention might work

The theory behind screening for type 2 diabetes mellitus is to identify either disease or associated risk factors to initiate preventative measures that can halt, slow or improve the course of disease [ 11 ]. Therefore, the earlier the disease is detected, especially where there is high risk of disease, theoretically, the better the expected outcomes. The logic model in Fig. 1 describes a complex system in which the intervention interacts with participants, context, implementation and how these affect the outcomes and the impact of this research [ 12 ].

Logic model describing the interactions between screening for diabetes, implementation, context, participants, outcomes and impact

Why it is important to do this review

Guidelines for screening interventions for type 2 diabetes mellitus, such as those released by the UK National Screening Committee [ 13 ], the American Diabetes Association [ 2 ] or the Society for Endocrinology, Metabolism and Diabetes of South Africa [ 14 ], are based on expert opinion and local practice rather than synthesised evidence. This is because there is limited information to provide evidence about best practice for screening interventions for type 2 diabetes mellitus and even less evidence in low- and middle-income countries [ 15 ]. A recently published Cochrane review assessed the effects of any type of screening compared with no screening for type 2 diabetes [ 16 ] and found only one trial, the ADDITION-Cambridge trial [ 17 ], that met their inclusion criteria. The ADDITION-Cambridge trial consisted of 20,184 participants aged 40–69 years from general practices in England who were at risk for diabetes but had no known diabetes. These participants were randomised to screening versus no screening arms, and followed up for a median of 9.6 years (November 2001 to November 2011). The review found moderate certainty evidence that screening for diabetes probably makes little or no difference to all-cause mortality and low certainty evidence that it may make little or no difference to diabetes-related mortality. However, because the review only included one trial, firm conclusions about early diabetes screening on health outcomes cannot be drawn. In consultation with the authors of the unpublished Cochrane review and considering the public health importance of screening and the potential impact on large populations, we propose to assess evidence from non-randomised intervention study designs. The questions for the systematic review will include the following: Does screening for type 2 diabetes mellitus reduce morbidity and/or mortality? Does a particular screening strategy result in a greater reduction of morbidity and/or mortality as compared to another screening strategy? Does screening for type 2 diabetes mellitus result in harms to participants or the health system?

Primary objective

To assess the effectiveness of targeted, opportunistic or mass screening for type 2 diabetes mellitus on reduction of diabetes-associated morbidity in adults

Secondary objectives

To assess the effectiveness of targeted, opportunistic or mass screening for type 2 diabetes mellitus on reduction of mortality (all cause as well as diabetes-associated) in adults

To assess the harms of targeted, opportunistic or mass screening for type 2 diabetes mellitus in adults

The present protocol has been registered within the PROSPERO database (CRD42020147439). This manuscript is being reported in accordance with the reporting guidance provided in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Protocols (PRISMA-P) statement [ 18 ] (see checklist in Additional file 1 ).

Study and source eligibility

Types of studies.

As existing reviews have found limited randomised evidence addressing this question [ 15 , 19 , 20 ], we will focus on non-randomised intervention studies (NRIS). We will employ the Cochrane EPOC criteria [ 21 ], and NRIS of interest will include non-randomised trials, controlled before-after studies, interrupted time-series study, repeated measures study and concurrently controlled prospective cohort study. The difficulty associated with labelling NRIS is well-documented in the literature; several of these designs, for example, have been used interchangeably; we will make use of the EPOC definitions and flow diagram to assist in study design identification ( Appendix 1 ):

Non-randomised trial (NRT) is a study design in which individual participants, or clusters of participants, are allocated to intervention or comparator in a quasi-random or non-random manner. If there is an allocation rule, it is often by, for example, alternation, day of the week, odd/even hospital, or identification number.

Controlled before-after (CBA) is a study design that estimates intervention effectiveness by comparing pre- and post-intervention outcomes in individuals or clusters that receive the intervention and those that do not.

Interrupted time series (ITS) studies design uses multiple observations from individuals or clusters pre-intervention to establish the pre-existing outcome trend; intervention effectiveness is then estimated by measuring post-intervention changes in the expected outcome trend associated with the introduction of an intervention (the ‘interruption’). An ITS study can identify both immediate and long-term changes associated with the intervention. The interrupted time-series studies will be required to have a clearly defined point in time when the intervention occurred and a minimum of 3 time points before and 3 time points after the intervention [ 21 ].

A repeated measures (RM) study is an interrupted time-series study but where the outcomes of interest are measured in the same participants at each point in time.

Concurrently controlled prospective cohort study (PCS) is where subjects are identified prospectively as having received an intervention or comparator and are then followed over time. The allocation rule is often in relation to organizational factors such as ward, clinic, doctor or provider organisation. Control arms should be contemporaneous, we will not include retrospective control arms.

‘PICO’ eligibility

Types of participants.

We will include adults aged 18 years and older without documented diabetes mellitus or pregnancy.

Types of interventions

We will include studies comparing one of the screening strategies, targeted, opportunistic or mass screening interventions for the detection of type 2 diabetes mellitus, against no screening or another of the screening strategies (Table 1 ). There will be a 6-month minimum follow-up time required for the primary clinical outcome of morbidity.

Types of outcome measures

Primary outcomes.

Clinical outcomes

Diabetes-related morbidity defined as study-reported microvascular complications (diabetic retinopathy, diabetic nephropathy, diabetic neuropathy) or macrovascular complications (non-fatal myocardial infarction, peripheral arterial disease, non-fatal stroke) and measured from 6 months after screening

Secondary outcomes

Mortality (all-cause and diabetes-related) defined as death due to any-cause including diabetes or other cardiovascular causes (including acute myocardial infarction, ischemic heart disease, stroke or any cardiovascular disorder that lead to death) and measured at any time after screening

Harms of diabetes screening

Harms to patients is defined as event/s reported in the study at any time after screening.

Psychological harms such as anxiety or stigma that impacts on quality of life due to a false-positive test

Number of days of work lost

Side-effects from treatment

Loss of health insurance benefits

Harms to health care system is defined as event/s reported in the study at any time after screening.

False-positive test resulting in human, physical and financial resource allocation to patients who are not in need

Overdiagnosis may lead to over-extension of human, physical and financial resources for patients who end up in prolonged treatment and engagement with the health system even if they never develop disease

The rationale for prioritisation of outcomes: Primary outcome serves to inform whether screening alters the course of disease as assumed per screening theory [ 11 ] and depicted in Fig. 1 . Secondary outcome of mortality contributes to the current data outlining no reduction in mortality following screening intervention [ 19 ] while also assessing harms that may arise from screening intervention [ 3 ] and therefore contribute to evidence to substantiate policy and practice recommendations.

Search methods for identification of studies

Electronic searches.

The University of Cape Town Health Sciences Reference Librarian (MS) assisted the first author (HM) in developing the search strategy and will provide advice and guidance in conducting the searches for the review.

Electronic Database Search (from inception onwards)

PubMed (MEDLINE)

Scopus (includes majority of EMBASE contents)

Web of Science Platform (Web of Science Core Collection, Biological Abstracts, SciELO Citation Index)

Academic Search Premier (on the EBSCOhost platform)

CINAHL (on the EBSCOhost platform)

Health Source Nursing Academic (on the EBSCOhost platform)

A draft search strategy for PubMed/MEDLINE, based on the original search strategy utilised by the Cochrane Review team and revised by an information specialist, is provided in Appendix 2 (see Appendix 2 ). We will include all studies regardless of publication status; however, we will only include English language studies. We are aware that this decision may lead to language bias [ 22 ], but due to capacity and resource limitation of the systematic review team, we are restricted to English only. We will search all databases from inception to the date of search. The search syntax will first be tested and optimised in PubMed. We will thereafter replicate the searches in the other databases adapting search syntax as necessary for those databases.

Grey literature search

We will conduct a grey literature search to identify studies not indexed in the databases listed above.

OpenGrey (multidisciplinary European database, covering science, technology, biomedical science, economics, social science and humanities)

Conference abstracts from The American Diabetes Association (ADA), the European Association for the Study of Diabetes (EASD) meeting and Diabetologia will be used to track down full text articles.

National Institute for Health Research Economic Evaluation Database (NHS EED)

Cost-Effectiveness Analysis Registry (CEA) (www.healtheconomics)

We will search key references, such as systematic reviews, by cross-checking reference lists for additional potentially eligible primary studies [ 23 ]. We will also contact experts in the field to check if we have missed any relevant studies. We may contact authors of included studies to clarify reported published information and to seek unpublished data.

Methods for screening search results

Screening methods.

We will collate and transfer search results to the Rayyan screening software [ 24 ] and remove duplicate records. At least two review authors will independently screen titles and abstracts of every record retrieved. Outcome measures will not be used to exclude studies during title and abstract screening. The potentially eligible records will be retrieved for full text screening. The two review authors will independently review full text records for compliance of studies with eligibility criteria of the review. A decision tree based on the eligibility criteria will be used to assist in decision making for exclusion of studies (see Appendix 3 ). Two review authors will resolve any disagreements through discussion or, if required, will consult a third review author. A study must meet all inclusion criteria to be included. We will list excluded studies at the full text screening stage in the ‘Characteristics of excluded studies’ table. We will collate multiple reports of the same study so that each study rather than each report is the unit of interest in the review. We will provide any information we can obtain about ongoing studies. We will record the selection process in sufficient detail to complete a PRISMA flow diagram [ 25 ].

Data collection and analysis

Data extraction.

We will use a standard data extraction form in Microsoft Excel to capture study characteristics and outcome data [ 22 , 26 ]; we will pilot the form on at least one eligible study. One review author will extract the following study characteristics from the included studies, and an independent review author will check the extraction:

Source: study ID (created by review author), review author ID (created by review author), citation and contact details

Eligibility: confirm eligibility for review, reason for exclusion

Methods: study design, number of study centres and location, study setting, withdrawals, date of study, follow-up, confounding factors considered, and the methods used to control for confounding, aspects of risk of bias specific for NRIS (see “Assessment of risk of bias in included studies” below), how missing data was handled

Participants: number, mean/median age, age range, gender, severity of condition, diagnostic criteria, inclusion criteria, exclusion criteria, screening criteria, diagnostic criteria, presence of known risk factors for type 2 diabetes mellitus (obesity, family history), co-morbidity (hypertension, dyslipidaemia), socio-demographics

Interventions: intervention components, comparison, fidelity assessment using the Template for Intervention Description and Replication (TIDieR) as a guide [ 27 ]

Outcomes: primary and secondary outcomes specified above in the section “Types of outcome measures”.

Miscellaneous: funding source, notable conflicts of interest of study authors, ethical approval, key conclusions of the study authors, miscellaneous comments from the study authors, references to other relevant studies, correspondence required, miscellaneous comments by the review authors.

One review author will extract outcome data from included studies, and an independent review author will check extracted data. We will note in the ‘Characteristics of included studies’ table if outcome data were reported in an unusable way. We will resolve disagreements by consensus or by involving a third review author.

Assessment of risk of bias in included studies

Two review authors will independently assess risk of bias for each study using the ROBINS-I tool [ 28 ]. Any disagreement will be resolved by discussion or by involving a third review author.

We will assess the risk of bias according to the following domains:

Pre-intervention: bias due to confounding

Pre-intervention: bias in selection of participants into the study

At intervention: bias in classification of interventions

Post-intervention: bias due to deviations from intended interventions

Post-intervention: bias due to missing data

Post-intervention: bias in measurement of outcomes

Post-intervention: bias in selection of the reported result

We will judge each potential source of bias as low risk, moderate risk, serious risk, critical risk of bias or no information. We will summarise the ‘Risk of bias’ judgements across different studies for each of the domains listed. We will consider blinding separately for different key outcomes where necessary (e.g. for unblinded outcome assessment, risk of bias for all-cause mortality may be very different than for a patient reported pain scale). Where information on risk of bias relates to unpublished data or correspondence with a trialist, we will note this in the ‘Risk of bias’ table. We will not exclude studies on the grounds of their risk of bias but will clearly report the risk of bias when presenting the results of the studies. When considering treatment effects, we will take into account the risk of bias for the studies that contribute to that outcome. We will conduct the review according to this published protocol and report any deviations from it in the ‘Differences between protocol and review’ section of the systematic review.

Dealing with missing data

Authors will be contacted, and missing data will be requested. If only returned in part and data can be logically imputed, such as standard errors, this will occur. All missing data will be clearly reported in the data extraction forms and risk of bias table and as such be assessed in the sensitivity analysis.

Data management

EndNote X9 and Microsoft Excel will be used for data management. If there is a conflict between data reported across multiple sources for a single study (e.g. between a published article and a trial registry record), we will report the data from the first peer-reviewed published article.

Data synthesis

Preparation for data synthesis.

In preparation for synthesis (either meta-analyses or synthesis without meta-analysis), we will assess how much data are available for each of our objectives by creating a table to compare the PICO elements and the study design features as well as the extracted numerical data for the compilation of a meta-analysis.

Measures of treatment effect

We will estimate the effect of the intervention using risk ratio for dichotomous data, and mean difference or standardised mean difference for continuous data. Time to event outcomes will be reported as hazard ratios. If other effect estimates are provided, we will convert between estimates where possible. Measures of precision will be 95% confidence intervals. We will ensure that an increase in scores for continuous outcomes can be interpreted in the same way for each outcome, explain the direction to the reader, and report where the directions were reversed if this was necessary. Interrupted time series data will be analysed and, if required, a statistical comparison of time trends before and after the intervention will be performed. For ITS studies, the guideline as outlined in Analysis in EPOC reviews will be followed with assistance of a statistician to ensure integrity of analysis [ 29 ].

Unit of analysis issues

To avoid unit of analysis errors we will consider the unit used to cluster the intervention (such as a ward, clinic, doctor or provider organisation) or in the case of repeated measures that there will be multiple observations for the same outcome. For instance, multiple screening intervention events per participant may occur over time that may cause a unit-of-analysis error. In order to calculate the confidence intervals, the participants per treatment group rather than the number of intervention attempts will be used [ 22 ]. Multiple intervention groups could create unit-of-analysis issues especially if different screening interventions are compared against no screening intervention and use the same participants with no screening intervention in both comparisons [ 22 ]. If there is more than one comparison in the study design, we will combine groups into a single pairwise comparison. If there is a unit of analysis error in the reported analysis for a study and there is insufficient information to reanalyse the results, the study authors will be contacted to obtain necessary data. If these data are not available, we will not report confidence intervals or p values for which there is a unit of analysis error [ 30 ].

Quantitative synthesis

We will undertake meta-analyses only where this is meaningful, i.e. if the interventions, participants and the underlying clinical question are similar enough for pooling to make sense. If feasible and appropriate, outcome data from primary studies will be used to perform random effects meta-analyses. Since heterogeneity is expected a priori, we will estimate the pooled treatment effect estimates and its 95% confidence interval using the random effects model. The random effects model assumes that the effect estimates follow a normal distribution, considering both within-study and between-study variation.

Assessment of heterogeneity

Forest plots will be used to visualise the extent of heterogeneity among studies. We will quantify statistical heterogeneity by estimating the variance between studies using I 2 statistic. The I 2 is the proportion of variation in effect estimates that is due to genuine variation rather than sampling (random) error. I 2 ranges between 0 and 100% (with values of 0–25% and 75–100% taken to indicate low and considerable heterogeneity, respectively) [ 22 ]. We will also calculate the chi-squared test where a p value < 0.1 indicates statistically significant heterogeneity.

Assessment of publication bias

If we include more than 10 studies investigating a particular outcome, we will use a funnel plot to explore possible publication bias, interpreting the results with caution [ 31 ].

Subgroup analysis and investigation of heterogeneity

We expect the following population characteristics may introduce clinical heterogeneity: age, sex, socio-economic status [ 6 ].

We expect the following contexts may introduce health system heterogeneity: study setting of rural or urban or in a low-middle income country or a high-income country (as defined by the World Bank) [ 6 ].

We will use the following outcomes in subgroup analysis:

Diabetes-associated morbidity

Mortality (all-cause and diabetes-associated)

Sensitivity analysis

We may conduct a sensitivity analysis to explore the influence of various factors on the effect size of the primary outcomes of the review only. We will stratify studies according to:

Restricting the analysis to published studies.

Restricting the analysis to studies with a low risk of bias, as specified in “Assessment of risk of bias in included studies”

Imputing missing data.

Any post hoc sensitivity analyses that may arise during the review process will be justified in the final report.

Assessment of certainty of evidence using the GRADE approach

Two review authors will independently assess the certainty of the evidence (high, moderate, low and very low) for each outcome using the five GRADE considerations for downgrading the certainty of evidence (risk of bias, consistency of effect, imprecision, indirectness, and publication bias) and the three criteria for upgrading the certainty of evidence (large effect, dose response and residual confounding opposing the observed effect) [ 32 ]. We will use the GRADEpro software GDT [ 33 ] to create the ‘Summary of findings’ tables for the main intervention comparisons and include the following outcomes: diabetes-associated morbidity, mortality (all-cause and diabetes-associated), harms (see Appendix 4 for SoF). We will resolve disagreements on certainty ratings by discussion and provide justification for decisions to down- or upgrade the ratings using footnotes in the SoF table and make comments to aid readers’ understanding of the review where necessary. We will use plain language statements to report these findings in the review [ 34 ]. The SoF tables will be used to draw conclusions about the certainty of the evidence within the text of the review. If during the review process, we become aware of an important outcome that we failed to list in our planned ‘SoF’ tables, we will include the relevant outcome and explain the reasons for this is the section ‘Differences between protocol and review’.

Systematic reviews of screening for type 2 diabetes have found no evidence that this intervention saves lives [ 15 , 19 , 20 ]; therefore, this review will primarily focus on the impact of screening on the reduction of diabetes-associated morbidities. The impact of this review is synthesised data for the provision of evidence-based decision-making for informing policy and practice around screening strategies for type 2 diabetes mellitus. Important protocol amendments will be documented and noted in the discussion.

Limitations

The potential limitations of this review at a study (outcome) level include the following: the potential finding of insufficient studies of similar study design and clinical question to synthesise abstracted study data; the overall completeness and applicability of evidence and quality of evidence especially due to the limitation to non-randomised studies due to the lack of randomised studies and therefore the lower quality of evidence; the limitation to English studies only and therefore the potential to miss published research; the limitation of not being able to discern between all-cause mortality and diabetes-related mortality and therefore combining this outcome under one mortality outcome. The potential limitation of this review at a systematic review process level includes the potential biases in the review process such as post hoc analysis and focus of outcome objectives.

Availability of data and materials

Not applicable.

Abbreviations

American Diabetes Association

Controlled before-after

Cost-Effectiveness Analysis Registry

Collaboration for Evidence-Based Healthcare and Public Health in Africa

European Association for the Study of Diabetes

Effective Practice and Organisation of Care

Fasting plasma glucose

Grading of Recommendations Assessment, Development and Evaluation

Detection of glycated haemoglobin A1C

Interrupted time series

NHS Economic Evaluation Database

Non-randomised trial

Non-randomised intervention studies

Oral glucose tolerance test

Prospective cohort study

Preferred Reporting Items for Systematic Reviews and Meta-Analyses

Repeated measures

Summary of findings

Template for Intervention Description and Replication

Ogurtsova K, da Rocha Fernandes JD, Huang Y, Linnenkamp U, Guariguata L, Cho NH, et al. IDF Diabetes Atlas: global estimates for the prevalence of diabetes for 2015 and 2040. Diabetes Res Clin Pract. 2017;128:40–50.

Article   CAS   Google Scholar  

American Diabetes Association. 2. Classification and diagnosis of diabetes: standards of medical care in diabetes-2018. Diabetes Care. 2018;41(Suppl 1):S13–27.

Article   Google Scholar  

World Health Organization. Global report on diabetes. 2016.

Sorkin JD, Muller DC, Fleg JL, Andres R. The relation of fasting and 2-h postchallenge plasma glucose concentrations to mortality: data from the Baltimore Longitudinal Study of Aging with a critical review of the literature. Diabetes Care. 2005;28(11):2626–32.

World Health Organization, International Diabetes Federation. Definition and diagnosis of diabetes mellitus and intermediate hyperglycaemia : report of a WHO/IDF consultation. World Health Organization. 2006.

World Health Organization. Screening for type 2 diabetes. Geneva, Switzerland: Department of Noncommunicable Disease Management; 2003.

Google Scholar  

Schulze MB, Hoffmann K, Boeing H, Linseisen J, Rohrmann S, Mohlig M, et al. An accurate risk score based on anthropometric, dietary, and lifestyle factors to predict the development of type 2 diabetes. Diabetes Care. 2007;30(3):510–5.

Muhlenbruch K, Ludwig T, Jeppesen C, Joost HG, Rathmann W, Meisinger C, et al. Update of the German Diabetes Risk Score and external validation in the German MONICA/KORA study. Diabetes Res Clin Pract. 2014;104(3):459–66.

Jolle A, Midthjell K, Holmen J, Carlsen SM, Tuomilehto J, Bjorngaard JH, et al. Validity of the FINDRISC as a prediction tool for diabetes in a contemporary Norwegian population: a 10-year follow-up of the HUNT study. BMJ Open Diabetes Res Care. 2019;7(1):e000769.

American Diabetes Association. Screening for diabetes. Diabetes Care. 2002;25(suppl 1):s21–s4.

Speechley M, Kunnilathu A, Aluckal E, Balakrishna MS, Mathew B, George EK. Screening in public health and clinical care: similarities and differences in definitions, types, and aims - a systematic review. J Clin Diagn Res. 2017;11(3):LE01–LE4.

PubMed   PubMed Central   Google Scholar  

Rohwer A, Pfadenhauer L, Burns J, Brereton L, Gerhardus A, Booth A, et al. Series: clinical epidemiology in South Africa. Paper 3: logic models help make sense of complexity in systematic reviews and health technology assessments. J Clin Epidemiol. 2017;83:37–47.

Waugh NR, Shyangdan D, Taylor-Phillips S, Suri G, Hall B. Screening for type 2 diabetes: a short report for the National Screening Committee. Health Technol Assess. 2013;17(35):1–90.

Amod A. The 2017 SEMDSA guideline for the management of type 2 diabetes. JEMDSA. 2017;22(1):S1–S196.

Selph S, Dana T, Blazina I, Bougatsos C, Patel H, Chou R. Screening for type 2 diabetes mellitus: a systematic review for the U.S. Preventive Services Task Force. Ann Intern Med. 2015;162(11):765–76.

Peer N, Balakrishna Y, Durao S. Screening for type 2 diabetes mellitus. Cochrane Database Syst Rev. 2020;5:CD005266.

PubMed   Google Scholar  

Simmons RK, Echouffo-Tcheugui JB, Sharp SJ, Sargeant LA, Williams KM, Prevost AT, et al. Screening for type 2 diabetes and population mortality over 10 years (ADDITION-Cambridge): a cluster-randomised controlled trial. Lancet. 2012;380(9855):1741–8.

Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, Petticrew M, et al. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev. 2015;4:1.

Saquib N, Saquib J, Ioannidis JP. Does screening for disease save lives in asymptomatic adults? Systematic review of meta-analyses and randomized trials. Int J Epidemiol. 2015;44(1):264–77.

Durao S, Ajumobi O, Kredo T, Naude C, Levitt NS, Steyn K, et al. Evidence insufficient to confirm the value of population screening for diabetes and hypertension in low- and-middle-income settings. S Afr Med J. 2015;105(2):98–102.

Cochrane Effective Practice and Organisation of Care (EPOC). EPOC review and what should they be called? EPOC resources for review authors. 2017.

Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, et al. Cochrane handbook for systematic reviews of interventions 2019 July 2019.

Horsley T, Dingwall O, Sampson M. Checking reference lists to find additional studies for systematic reviews. Cochrane Database Syst Rev. 2011;8:MR000026.

Ouzzani M, Hammady H, Fedorowicz Z, Elmagarmid A. Rayyan-a web and mobile app for systematic reviews. Syst Rev. 2016;5(1):210.

Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, Ioannidis JP, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Med. 2009;6(7):e1000100.

Cochrane Effective Practice and Organisation of Care (EPOC). Data collection form. 2013.

Hoffmann TC, Glasziou PP, Boutron I, Milne R, Perera R, Moher D, et al. Better reporting of interventions: template for intervention description and replication (TIDieR) checklist and guide. BMJ : British Medical Journal. 2014;348:g1687.

Sterne JA, Hernan MA, Reeves BC, Savovic J, Berkman ND, Viswanathan M, et al. ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions. BMJ. 2016;355:i4919.

Cochrane Effective Practice and Organisation of Care (EPOC). Interrupted time series (ITS) analyses. 2017.

Cochrane Effective Practice and Organisation of Care (EPOC). Analysis in EPOC reviews. 2017.

Sterne JA, Sutton AJ, Ioannidis JP, Terrin N, Jones DR, Lau J, et al. Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials. BMJ. 2011;343:d4002.

Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Alonso-Coello P, et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ. 2008;336(7650):924–6.

McMaster University. GRADEpro GDT. In: Group HOGW, editor. 2014.

Cochrane Effective Practice and Organisation of Care (EPOC). EPOC worksheets for preparing a 'Summary of findings' table using GRADE. 2013.

Download references

Acknowledgements

The authors would like to thank the following people who have supported this protocol:

• The authors of an unpublished Cochrane systematic review titled “Screening for type 2 diabetes mellitus”: Nasheeta Peer and Solange Durao, South African Medical Research Council, for providing their protocol to form the basis of this protocol, as well as a summary of their findings to form the rationale of the protocol and a quote to summarise the findings of their unpublished review.

• Information specialist Maria-Inti Metzendorf, Cochrane Metabolic and Endocrine Disorders Group, provided search strategy for the unpublished Cochrane review by Peer and Durao upon which the current search strategy is based.

• CEBHA+ Methodological Support Group Jake Burns and Peter Philipsborn, Ludwig-Maximillians Universitat Munchen, are internal reviewers for CEBHA+ which funds this review and assisted by providing critical review of the draft protocol.

• Prof Naomi Levitt, Chronic Disease for Africa, UCT, is a clinical Endocrinologist/Diabetologist who contributed to the PICO of this review when conceptualising the project and gave critical feedback to assist with responding to reviewers comments on the protocol before resubmission.

The work reported herein was made possible through Cochrane South Africa, South African Medical Research Council under the Collaboration for Evidence-Based Healthcare and Public Health in Africa (CEBHA+) Scholarship Programme. CEBHA+ receives funding from the Federal Ministry for Education and Research (Bundesministerium fur Bildung und Forschung, BMBF), Germany, through the BMBF funding of Research Networks for Health Innovation in Sub-Saharan Africa. The content hereof is the sole responsibility of the authors and does not necessarily represent the official views of SAMRC or the funders. The funding body, the Federal Ministry for Education and Research (Bundesministerium fur Bildung und Forschung, BMBF), Germany, played no role in the design and writing of the protocol; however, members of the Collaboration for Evidence-Based Healthcare and Public Health in Africa (CEBHA+) provided critical review and expert advice on the penultimate version of the protocol.

Author information

Authors and affiliations.

Vaccines for Africa Initiative, University of Cape Town, Cape Town, South Africa

Helen Mearns, Paul Kuodi Otiku & Benjamin M. Kagina

School of Public Health & Family Medicine, University of Cape Town, Cape Town, South Africa

Health Sciences Library, University of Cape Town, Cape Town, South Africa

Mary Shelton

Cochrane South Africa, South African Medical Research Council, Cape Town, South Africa

Tamara Kredo & Bey-Marrié Schmidt

Division of Clinical Pharmacology, Department of Medicine, Faculty of Medicine and Health Sciences, University of Stellenbosch, Cape Town, South Africa

Tamara Kredo

You can also search for this author in PubMed   Google Scholar

Contributions

Conceiving the protocol: BS, TK. Designing the protocol: HM, BS. Co-ordinating the protocol: HM. Designing search strategies: MS, HM. Writing the protocol: HM. Providing general advice on the protocol and approving the final version: PKO, MS, BK, TK, BS. Securing funding for the protocol: BS, TK. Performing previous work that was the foundation of the current study: NP, SD. Guarantor of the review: BS. The authors read and approved the final manuscript.

Corresponding author

Correspondence to Helen Mearns .

Ethics declarations

Ethics approval and consent to participate.

Ethics approval is not required for a systematic review of secondary data.

Consent for publication

Competing interests.

The authors declare that they have no competing interests.

HM: None known

PKO: None known

MS: None known

BMK: None known

TK: None known

BS: None known

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Additional file 1..

PRISMA Checklist.

Flow diagram to assist with identifying the type of study (modified from [ 34 ])

Search Strategy for PubMed:

Set 1: Diabetes Mellitus, Type 2 [MeSH] OR [Text Word field:] Adult onset diabetes OR late onset diabetes OR latent diabetes OR mature onset diabetes OR MODY OR NIDDM OR noninsulin-dependent diabetes OR slow onset diabetes OR stable onset diabetes OR type 2 diabetes OR type II diabetes OR T2DM OR T2D

Set 2: Diabetes Insipidus [MeSH] OR [Text Word field:] diabetes insipidus

Set 3 : 1 NOT 2

Set 4: Mass screening [MeSH] OR [Text Word field:] screening

Set 5: 3 AND 4

Set 6: Animals [MeSH] NOT Humans [MeSH]

Set 7: 5 NOT 6

Set 8: [All fields:] Trial OR trials OR before-and-after study OR before-and-after studies OR cohort OR comparative study OR comparative studies OR Controlled OR evaluation study OR evaluation studies OR follow-up study OR follow-up studies OR interrupted time series OR longitudinal study OR longitudinal studies OR non-randomised OR non-randomized OR nonrandomised OR nonrandomized OR non randomised OR non randomized OR program evaluation OR programme evaluation OR prospective study OR prospective studies OR quantitative study OR quantitative studies OR quasi experimental OR repeated measures

Set 9: 7 AND 8

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Mearns, H., Otiku, P.K., Shelton, M. et al. Screening strategies for adults with type 2 diabetes mellitus: a systematic review protocol. Syst Rev 9 , 156 (2020). https://doi.org/10.1186/s13643-020-01417-3

Download citation

Received : 21 December 2019

Accepted : 30 June 2020

Published : 13 July 2020

DOI : https://doi.org/10.1186/s13643-020-01417-3

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Mass screening
• Opportunistic
• Type 2 diabetes mellitus

Systematic Reviews

ISSN: 2046-4053

• Submission enquiries: Access here and click Contact Us
• General enquiries: [email protected]

• - Google Chrome

Intended for healthcare professionals

• Access provided by Google Indexer
• My email alerts
• BMA member login
• Username * Password * Forgot your log in details? Need to activate BMA Member Log In Log in via OpenAthens Log in via your institution

Search form

• Advanced search
• Search responses
• Search blogs
• Trends in incidence of...

Trends in incidence of total or type 2 diabetes: systematic review

Visual summary available

Showing the turning point in diabetes incidence in 61 populations

Linked editorial

Trends in type 2 diabetes

• Related content
• Peer review
• Dianna J Magliano , laboratory head of diabetes and population health 1 2 ,
• Rakibul M Islam , postdoctoral research fellow 1 2 ,
• Elizabeth L M Barr , postdoctoral research fellow 1 ,
• Edward W Gregg , chair in diabetes and cardiovascular disease epidemiology 3 4 ,
• Meda E Pavkov , physician scientist 3 ,
• Jessica L Harding , research fellow 3 ,
• Maryam Tabesh , research study coordinator 1 2 ,
• Digsu N Koye , postdoctoral research fellow 1 2 ,
• Jonathan E Shaw , deputy director of Baker Heart and Diabetes Institute 1 2
• 1 Baker Heart and Diabetes Institute, Melbourne, VIC 3004, Australia
• 2 School of Public Health and Preventive Medicine, Monash University, Melbourne, VIC 3004, Australia
• 3 Centres for Diseases Control and Prevention, Division of Diabetes Translation, Atlanta, GA, USA
• 4 School of Public Health, Epidemiology and Biostatistics, Imperial College London, London, UK
• Correspondence to: D J Magliano dianna.magliano{at}baker.edu.au
• Accepted 16 July 2019

Objective To assess what proportions of studies reported increasing, stable, or declining trends in the incidence of diagnosed diabetes.

Design Systematic review of studies reporting trends of diabetes incidence in adults from 1980 to 2017 according to PRISMA guidelines.

Data sources Medline, Embase, CINAHL, and reference lists of relevant publications.

Eligibility criteria Studies of open population based cohorts, diabetes registries, and administrative and health insurance databases on secular trends in the incidence of total diabetes or type 2 diabetes in adults were included. Poisson regression was used to model data by age group and year.

Results Among the 22 833 screened abstracts, 47 studies were included, providing data on 121 separate sex specific or ethnicity specific populations; 42 (89%) of the included studies reported on diagnosed diabetes. In 1960-89, 36% (8/22) of the populations studied had increasing trends in incidence of diabetes, 55% (12/22) had stable trends, and 9% (2/22) had decreasing trends. In 1990-2005, diabetes incidence increased in 66% (33/50) of populations, was stable in 32% (16/50), and decreased in 2% (1/50). In 2006-14, increasing trends were reported in only 33% (11/33) of populations, whereas 30% (10/33) and 36% (12/33) had stable or declining incidence, respectively.

Conclusions The incidence of clinically diagnosed diabetes has continued to rise in only a minority of populations studied since 2006, with over a third of populations having a fall in incidence in this time period. Preventive strategies could have contributed to the fall in diabetes incidence in recent years. Data are limited in low and middle income countries, where trends in diabetes incidence could be different.

Systematic review registration Prospero CRD42018092287.

Introduction

Over the past few decades, the prevalence of diabetes in developed and developing countries has risen substantially, making diabetes a key health priority globally. 1 Examination of trends in total burden of diabetes is an essential part of the monitoring of this health priority area, but, to date, it has consisted primarily of studies looking at diabetes prevalence. 1 2 3 4 5 Prevalence estimates suggest that the diabetes burden is still rising in most countries, and this is often interpreted as evidence of increasing risk in the population. However, selective incidence studies 6 7 and some accompanying risk factor data 8 suggest otherwise. Prevalence can be a crude and misleading metric of the trajectory of an epidemic, because increasing prevalence of a disease might be due to either increasing incidence or to improved survival. Furthermore, prevalence cannot be reliably used to study the effects of changes in population risk factors, because their effects are detected earlier with incidence trends than with prevalence trends, and incidence is not affected by changes in survival.

Incidence measures the proportion of people who develop diabetes over a period of time among the population at risk. It is the appropriate measure of population risk, and a valuable way of assessing whether public health campaigns for diabetes prevention are succeeding. While prevalence can rise simply because mortality falls, incidence of diagnosed diabetes is affected only by the risk of the population and the amount of screening undertaken. Changes in prevalence might be an inadequate guide to the effects of prevention activities, and could lead to the inappropriate rejection of effective interventions. It is only by measuring both incidence and prevalence that a better understanding of the extent of diabetes can be achieved.

Among existing diabetes incidence data, a few studies suggest that diabetes incidence could be falling despite rising or stable prevalence, 6 7 9 but not all data are consistently showing the same trends. For example, studies from England and Wales (1994-98), 10 Portugal (1992-2015), 11 and Canada (1995-2007) 12 are reporting increases in diabetes incidence. To understand what is happening at a global level over time, a systematic approach to review all incidence trend data should be undertaken to study patterns and distributions of incidence trends by time, age, and sex. So far, no systematic reviews have reported on trends in the incidence of diabetes. Therefore, we conducted a systematic review of the literature reporting diabetes incidence trends.

Data sources and searches

We conducted a systematic review in accordance with PRISMA guidelines. 13 We searched Medline, Embase, and CINAHL from January 1980 to December 2017 without language restrictions. The full search strategy is available in supplementary table 1.

Study selection

Inclusion and exclusion criteria.

Eligible studies needed to report diabetes incidence in two or more time periods. Study populations derived from open, population based cohort studies (that is, with ongoing recruitment over time), diabetes registries, or administrative or health insurance databases based mainly or wholly in primary care (electronic medical records, health insurance databases, or health maintenance organisations). We also included serial, cross sectional, population based studies where incidence was defined as a person reporting the development of diabetes in the 12 months before the survey. Studies were required to report on the incidence of either total diabetes or type 2 diabetes. We excluded studies reporting incidence restricted to select groups (eg, people with heart failure) and studies reporting only on children or youth.

Each title and abstract was screened by at least two authors (DJM, JES, DNK, JLH, and MT) and discrepancies were resolved by discussion. We aimed to avoid overlap of populations between studies. Therefore, if national data and regional data were available from the same country over the same time period, we only included the national data. If multiple publications used the same data source, over the same time period, we chose the publication that covered the longest time period.

Outcome measure

Our outcome was diabetes incidence using various methods of diabetes ascertainment including: blood glucose, glycated haemoglobin (HbA1c), linkage to drug treatment or reimbursement registries, clinical diagnosis by physicians, administrative data (ICD codes (international classification of diseases)), or self report. Several studies developed algorithms based on several of these elements to define diabetes. We categorised the definition of diabetes into one of five groups: clinical diagnosis, diabetes treatment, algorithm derived, glycaemia defined (blood glucose or HbA1c, with or without treatment), and self report.

Data extraction and quality of studies

We extracted crude and standardised incidence by year (including counts and denominators) and the reported pattern of the trends (increasing, decreasing, or stable, (that is, no statistically significant change)) in each time period as well as study and population characteristics. Age specific data were also extracted if available. Data reported only in graphs were extracted by DigitizeIt software (European Organisation for Nuclear Research, Germany). We assessed study quality using a modified Newcastle-Ottawa scale for assessing the risk of bias of cohort studies 14 (supplementary material).

Statistical methods

Data were reported as incidence density (per person year) or yearly rates (percentage per year). From every study, we extracted data from every subpopulation reported, such that a study reporting incidence in men and women separately contributed two populations to this analysis. If studies reported two different trends over different time periods, we considered these as two populations. Further, if the study was over 10 years in duration, we treated these as two separate time periods. To avoid double counting, when the data were reported in the total population as well as by sex and ethnic groups, we only included data once and prioritised ethnicity specific data over sex specific data.

We extracted the age specific incidence data reported for every individual calendar year. These data were then categorised into four age bands (<40, 40-54, 55-69, and ≥70), and were plotted against calendar year. In studies where counts and denominators were reported by smaller age groups than we used, we recalculated incidence across our specified larger age groups. If we found multiple age groups within any of our broader age groups, but with insufficient information to combine the data into a new category, only data from one age group were used. To limit overcrowding on plots, if data were available for men, women, and the total population, only total population data were plotted. Data from populations with high diabetes incidence such as Mauritians 15 and First Nation populations from Canada 16 were plotted separately to allow the examination of most of the data more easily on a common scale (supplementary material). Furthermore, studies reporting data before 1991 or populations with fewer than three data points were not plotted. We also categorised studies into European and non-European populations on the basis of the predominant ethnicity of the population in which they were conducted. Studies conducted in Israel, Canada, and the United States were assigned to the European category.

We took two approaches to analyse trends of diabetes incidence over time. Firstly, we allocated the reported trend (increasing, decreasing, or stable (that is, no statistically significant change)) of each population to the mid-point of each study’s observational period, and then assigned this trend into one of five time periods (1960-79, 1980-89, 1990-99, 2000-05, and 2006-14). Where a test of significance of trends was not reported or when a time period was longer than 10 years, we performed Joinpoint trend analyses 17 18 to observe any significant trends in the data (assuming a constant standard deviation). Joinpoint Trend Analysis Software (version 4.5.0.1) uses permutation tests to identify points where linear trends change significantly in direction or in magnitude, and calculates an annual percentage change for each time period identified. In sensitivity analyses we also tested different cut points in the last two time periods.

The second approach was used to more accurately allocate trends to the prespecified time periods. Among the studies that reported raw counts of diabetes cases and denominators, we examined the association between calendar year and incidence, using Poisson models with the log person years as offset. The midpoints of age and calendar period were used as continuous covariates, and the effects of these were taken as linear functions. We analysed each study separately by prespecified time periods, and reported annual percentage change when the number of data points in the time period was at least four. For studies that did not provide raw data but did report a sufficient number of points, we analysed the relation between year and incidence using Joinpoint regression across the time periods specified above and reported annual percentage change. Analyses were conducted with Stata software version 14.0 (Stata Corporation, College Station, TX, USA), and Joinpoint (Joinpoint Desktop Software Version 4.5.0.1). 17 18

Patient and public involvement

No patients or members of the public were involved in setting the research question or the outcome measures for this study. No patients were asked to advise on interpretation or writing up of results. We intend to disseminate this research through press releases and at research meetings.

We found 22 833 unique abstracts from 1 January 1980 to the end of 2017. Among these, 80 described trends of diabetes incidence, of which 47 met all inclusion criteria. Articles describing trends were excluded for the following reasons: duplicated data (n=21), closed cohorts (n=5), populations included youth only (n=1), occupational cohorts (n=2), or no usable data presented (n=4; fig 1 ).

Flowchart of study selection

• Download figure
• Open in new tab
• Download powerpoint

Table 1 and supplementary material table 2 describe the characteristics of the included studies. Only 19% (9/47) of studies were from predominantly non-Europid populations and 4% (2/47) of studies were from low or middle income countries (China 25 and Mauritius 15 ). Administrative datasets, health insurance data, registry data, survey data, and cohort studies accounted for 38% (n=18), 21% (n=10), 19% (n=9), 11% (n=5), and 11% (n=5) of the 47 data sources, respectively. Among the 47 studies, diabetes was defined by a clinical diagnosis, diabetes treatment (via linkage to drug treatment registers), an algorithm, blood glucose, and self report in 28% (n=13), 9% (n=4), 47% (n=22), 11% (n=5), and 6% (n=3) of studies, respectively. Sample sizes of the populations were greater than 10 000 in every year in 85% (n=40) of the studies, and greater than 130 000 per year in 70% (n=33) of the studies. A total of 62% (n=29) of the 47 included studies exclusively reported on type 2 diabetes, and 38% (n=18) reported on total diabetes.

Characteristics of 47 included studies reporting on diabetes incidence trends, by country

• View inline

Summary of patterns of diabetes incidence trends based on analyses reported in publications in 1960-99

Trends of diabetes incidence

Among the 47 studies, 16 provided information on incidence by age group. Of these 16 studies, 14 were plotted in figure 2 , with those from high incidence countries plotted in supplementary figure 1. In these figures, incidence in most studies increased progressively until the mid-2000s in all age groups. Thereafter, most studies showed a stable or decreasing trend, apart from studies in Denmark 26 27 and Germany 31 and in a US health insurance population 9 where the incidence inflected upwards in the later years for some age groups.

Incidence of diabetes over time for populations aged under 40, 40-54, 55-69, and 70 or more, among studies reporting age specific data. Only populations with at least three points were plotted. NHIS=National Health Interview Survey

Using the first approach to analyse trends of diabetes incidence over time, we separated the data into populations based on sex and ethnicity, and allocated a time period to each population, generating 105 populations for analysis. Seventy four and 31 populations were predominantly Europid and non-Europid, respectively. Table 2 and table 3 show the reported trend for each population. Table 4 summarises the findings in table 2 and table 3 , and shows that the proportion of populations reporting increasing trends peaked in 1990-99 and fell progressively in the two later time periods. Between 1960 and 1989, 36% (8/22) of the populations studied had increasing trends in incidence of diabetes, 55% (12/22) had stable trends, and 9% (2/22) had decreasing trends. In 1990-2005, diabetes incidence increased in 66% (33/50) of populations, was stable in 32% (16/50), and decreased in 2% (1/50). In 2006-14, increasing trends were reported in 33% (11/33) of populations, whereas 30% (10/33) and 36% (12/33) had stable or declining incidence, respectively.

Summary of patterns of diabetes incidence trends based on analyses reported in publications in 2000-14

Summary of incidence trends over time of total or type 2 diabetes

Populations that reported a decrease in incidence after 2005 came from the US, 6 9 Israel, 34 Switzerland, 46 Hong Kong, 32 Sweden, 43 and Korea. 36 Populations reporting increasing incidence after 2005 included Portugal, 11 Denmark, 26 27 and Germany, 31 while populations from Canada, 19 Italy, 35 Scotland, 40 Norway, 39 US (non-Hispanic white), 56 and the United Kingdom 50 showed stable incidence. For two studies (16 populations), 16 29 we could not determine a direction of a trend (increasing, decreasing, or stable), because they showed three phases of change with the trend of the middle phase differing from the trend of the first and last phase. Across the total time period, we observed a higher proportion of populations reporting stable or decreasing trends in predominantly Europid than in non-Europid populations (52% v 41%).

Using the second approach to analyse trends of diabetes incidence over time, we modelled 21 studies (62 populations) that reported diabetes counts and denominators specifically within each time period ( table 5 ). The percentage of populations with a decreased or stable incidence was highest in 1980-89 (88%; 7/8), but this proportion was based on only eight populations in three studies. From 1990 onwards, the percentage with decreasing or stable incidence increased progressively, reaching 83% (19/23) of populations in 2006-14. Eight studies (21 populations) that were analysed by Joinpoint had no data on counts or denominators (supplementary table 3). When these data were considered with the data in table 5 , the percentage of populations in 2006-14 with decreasing or stable incidence fell to 70% (19/27), but this proportion was still the highest of all the time periods, whereas the percentage for 1990-99 remained the lowest at 31% (5/16).

Annual percentage change in diabetes incidence in men (M), women (W), or total population (T) among studies that provided counts and denominators, by time period

In a sensitivity analysis, we tested whether our selection of time periods was driving our results. When we defined the final time periods to be 2000-07 and 2008-14, our results were not altered, with 66% (21/32) of the populations in the last time period showing decreasing or stable trends. We also repeated the analysis in table 4 and excluded cohort studies and surveys, and found that the results were not materially altered, with 65% (20/31) of populations in the last time period (from 2006 onwards) showing decreasing or stable incidence of diabetes.

Quality of studies

The median score for study quality was 10 (interquartile range 8-11; supplementary table 4). We repeated the analyses reported in table 4 after excluding studies that had quality scores in the lowest quarter, and observed similar results to the main findings. For example, in 1960-89, 67% (10/15) of populations reported stable or decreasing incidence, while in the final time period, 67% (18/27) of populations reported stable or decreasing incidence of diagnosed diabetes.

Principal findings

In this systematic review of population based studies on diabetes incidence, we show evidence that the incidence of diagnosed diabetes increased in most populations from the 1960s to the early 2000s, after which a pattern emerged of levelling trends in 30% and declining trends in 36% of the reported populations. Although the lack of data for non-Europid populations leaves global trends in incidence unclear, these findings suggest that trends in the diabetes epidemic in some high income countries have turned in a more encouraging direction compared with previous decades. It is important to note that these results apply predominantly to type 2 diabetes, as even though many studies did not accurately define diabetes type, the incidence of type 2 diabetes in adults is an order of magnitude greater than that of type 1 diabetes.

The countries that showed stable or decreasing trends in the last time period were from Europe and east Asia, with no obvious clustering or commonalities. For the countries showing decreasing or stable diabetes trends, if the prevalence data were used to understand the diabetes epidemic in that country, a different message would be obtained. For example, national data from Korea showed that the prevalence of diabetes increased from 2000 to 2010. 59 Similarly in Sweden, the prevalence of pharmacologically treated diabetes increased moderately from 2006 to 2014. 43 In the US, the prevalence of diabetes reached a plateau when incidence began to decrease. However, we lacked incidence data from many areas of the world where the most steady and substantial increases in prevalence have been reported, including the Pacific Islands, Middle East, and south Asia. Large increases in incidence could still be occurring in these areas. The lack of incidence data for much of the world, combined with the common observation of discordance between incidence and prevalence rates where such data exist, both underscore the importance of using incidence data to understand the direction of the diabetes epidemic.

Incidence could be starting to fall for several reasons. Firstly, we might be starting to benefit from prevention activities of type 2 diabetes, including increased awareness, education, and risk factor modification. These activities have involved both targeted prevention among high risk individuals, similar to that conducted in the Diabetes Prevention study 60 and Diabetes Prevention Programme 61 62 in many countries, 63 and less intensive interventions with broader reach such as telephone counselling in the general community. 64 65 67 Secondly, health awareness and education programmes have also been implemented in schools and work places, and many changes to the physical environment, such as the introduction of bike tracks and exercise parks, have occurred. 68 Thirdly, favourable trends in selected risk factors of type 2 diabetes in some countries provide indirect evidence of positive changes to reduce diabetes incidence. Finally, in the US, there is some evidence in recent years of improved diets and related behaviours, which include reductions in intake of sugar sweetened beverages 69 and fat, 70 small declines in overall energy intake, and declines in some food purchases. 8 71

Similar reduction in consumptions of sugar sweetened beverages have occurred in Norway 72 and Australia 73 and fast food intake has decreased in Korea. 74 Some of these changes could be linked to a fall in diabetes incidence. Some places such as Scotland 75 have also had a plateauing of obesity prevalence, but this is not universal. In the US, despite earlier studies suggesting that the rate of increase in obesity might be slowing down, 76 77 more recent data show a small increase. 78 79 While some evidence supports the hypothesis that these prevention activities for type 2 diabetes and an improved environment could trigger sufficient behaviour change to have an effect on diabetes incidence, other data, such as the continuing rising obesity prevalence in the US, 79 casts some doubt over the explanations underpinning our findings on diabetes incidence trends.

Other factors might have also influenced reported diabetes incidence. Only 11% (n=5) of the studies reported here screened for undiagnosed diabetes, and therefore trends could have been influenced by secular changes in diagnostic behaviour. In 1997, the threshold for fasting plasma glucose for diagnosis of diabetes was reduced from 7.8 to 7.0 mmol/L, which could increase diagnosis of new cases of type 2 diabetes. In 2009-10, HbA1c was then introduced as an alternative way to diagnose diabetes. 80 Evidence from some studies suggests that the HbA1c diagnostic threshold detects fewer people with diabetes than do the thresholds for fasting plasma blood glucose, 80 81 potentially leading to a lowering of incidence estimates. However, across multiple studies, prevalence estimates based on fasting plasma glucose only versus HbA1c definitions are similar. 82 Furthermore, because HbA1c can be measured in the non-fasting state (unlike the fasting blood glucose or oral glucose tolerance test), the number of people who actually undergo diagnostic testing could be higher with HbA1c. Nichols and colleagues 56 reported that among seven million insured US adults, despite a shift towards HbA1c as the diagnostic test in 2010, the incidence of diabetes did not change from 2010 to 2011.

Another potential explanation for declining or stable diabetes incidence after the mid-2000s is a reduction in the pool of undiagnosed diabetes 83 through the intensification of diagnostic and screening activities 83 84 and changing diagnostic criteria during the previous decade. 80 Data from Read and colleagues provide some evidence to support this notion. 41

Among the included studies, two studies specifically examined clinical screening patterns in parallel with incidence trends. These studies reported that the proportion of the population screened for diabetes increased over time, and the incidence of diabetes remained stable 56 or fell. 34 While the Karpati study 34 combined data for glucose testing with HbA1c testing, the study by Nichols and colleagues 56 separated the two, and showed that both glucose testing and HbA1c testing increased over time. A third study, in Korea, 36 also noted that the incidence of diabetes decreased in the setting of an increase in the uptake of the national health screening programme. Despite the introduction of HbA1c for diagnosis of diabetes by the World Health Organization, this practice has not been adopted everywhere. For example, neither Scotland nor Hong Kong have introduced the use of HbA1c for screening or diagnosis of diabetes, and studies in these areas showed a levelling of diabetes incidence trends and decreasing trends, respectively.

Our findings appear to contrast with data showing increasing global prevalence of diabetes. 1 3 However, increasing prevalence could be influenced by improved survival of people with diabetes, because this increases the length of time that each individual remains within the diabetes population. As is shown in several studies in this review, 23 41 mortality from diabetes and incidence of diabetes might both be falling but as long as mortality is lower than incidence, prevalence will rise. Therefore, we argue that prevalence alone is an insufficient measure to track the epidemic of diabetes and other non-communicable diseases.

Strengths and weaknesses of this study

A key strength of this work was the systematic approach and robust methodology to describe trends in diagnosed diabetes incidence. We also presented the reported trends allocated to approximate time periods, as well as conducting our own regression within exact time periods. The following limitations should also be considered. Firstly, we did not formally search the grey literature, because a preliminary grey literature search revealed only low quality studies, with inadequate methodological detail to provide confidence in any observed incidence trends, and thus review could be subject to publication bias. Secondly, we were not able to source age or sex specific data on all populations. Thirdly, it was not possible to adjust for different methods of diabetes diagnosis or ascertain trends by different definitions of diabetes. Fourthly, most data sources reported only on clinically diagnosed diabetes and so were subject to influence from diagnostic behaviour and coding practices. Fifthly, study type changed over time, with large administrative datasets becoming more common and cohort studies becoming less common over time. Nevertheless, the size and absence of volunteer bias in administrative datasets likely make them less biased. Finally, data were limited in low and middle income countries.

Conclusions and unanswered questions

This systematic review shows that in most countries for which data are available, the incidence of diagnosed diabetes was rising from the 1990s to the mid-2000s, but has been stable or falling since. Preventive strategies and public health education and awareness campaigns could have contributed to this recent trend. Data are limited in low and middle income countries where trends in diabetes incidence might be different. Improvement of the collection, availability, and analysis of incidence data will be important to effectively monitor the epidemic and guide prevention efforts into the future.

What is already known on this topic

Monitoring of the diabetes epidemic has mainly focused on reporting diabetes prevalence, which continues to rise; however, increasing prevalence is partly driven by improved medical treatment and declining mortality

Studies on diabetes incidence are scarce, but among those that exist, some report a fall or stabilisation of diabetes incidence;

Whether the proportion of studies reporting falling incidence has changed over time is not known

What this study adds

This systematic review of published data reporting diabetes incidence trends over time shows that in most countries with available data, incidence of diabetes (mainly diagnosed diabetes) increased from the 1990s to the mid-2000s, and has been stable or falling since

Preventive strategies and public health education and awareness campaigns could have contributed to this flattening of rates, suggesting that worldwide efforts to curb the diabetes epidemic over the past decade might have been effective

Published data were very limited in low and middle income countries, where trends in diabetes incidence might be different

Acknowledgments

The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the US Centers for Disease Control and Prevention (CDC).

Contributors: MT, DNK, JLH, and RMI are postdoctoral fellows who screened abstracts for selection into the systematic review. JES and DJM also screened abstracts. ELMB applied the quality criteria to the selected articles. RMI extracted data, applied quality criteria to selected articles, and contributed to preparing the manuscript. DJM conceived the project, screened abstracts, extracted the data, analysed the data, and wrote the manuscript. JES, MEP, and EWG conceived the project, edited the manuscript, and provided intellectual input throughout the process. The funder of the study (CDC) was part of the study group and contributed to data collection, data analysis, data interpretation, and writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication. DJM is guarantor. The corresponding author attests that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted.

Funding: Funded by the CDC. The researchers were independent from the funders.

Competing interests: All authors have completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf and declare: support from the CDC for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work.

Ethical approval: Not required because this work was a systematic review.

Data sharing: Data are available from the corresponding author ([email protected]).

The lead author affirms that the manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained.

This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/ .

• Finucane MM ,
• Global Burden of Metabolic Risk Factors of Chronic Diseases Collaborating Group (Blood Glucose)
• Whiting DR ,
• Guariguata L ,
• International Diabetes Federation
• NCD Risk Factor Collaboration (NCD-RisC)
• Karuranga S ,
• Abraham TM ,
• Pencina KM ,
• Pencina MJ ,
• Slining MM ,
• Kimball ES ,
• Newnham A ,
• de Sousa-Uva M ,
• Antunes L ,
• Johnson JA ,
• Hemmelgarn BR ,
• Liberati A ,
• Tetzlaff J ,
• Altman DG ,
• PRISMA Group
• ↵ Wells G, Shea B, O’connell D, et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. Ottawa Hospital Research Institute, 2014. www.ohri.ca/programs/clinical_epidemiology/oxford.asp Last accessed 14 December 2018.
• Söderberg S ,
• Tuomilehto J ,
• ↵ Joinpoint Regression Program. 4.6.0.0 version. Statistical Methodology and Applications Branch, Surveillance Research Program: National Cancer Institute, 2018. https://surveillance.cancer.gov/joinpoint/ .
• Midthune DN
• ↵ Canadian Chronic Disease Surveillance System. Canadian Chronic Disease Surveillance System 2017. https://www.canada.ca/en/public-health.html
• Blanchard JF ,
• Lipscombe LL ,
• Jacobs-Whyte H ,
• Paradis G ,
• Macaulay AC
• Carstensen B ,
• Kristensen JK ,
• Ottosen P ,
• Borch-Johnsen K ,
• Steering Group of the National Diabetes Register
• Jensen PB ,
• Abouzeid M ,
• Wikström K ,
• Peltonen M ,
• Reunanen A ,
• Klaukka T ,
• Maatela J ,
• Michaelis D ,
• Boehme MW ,
• Buechele G ,
• Frankenhauser-Mannuss J ,
• Vilbergsson S ,
• Sigurdsson G ,
• Sigvaldason H ,
• Hreidarsson AB ,
• Sigfusson N
• Karpati T ,
• Cohen-Stavi CJ ,
• Leibowitz M ,
• Feldman BS ,
• Baviera M ,
• Marzona I ,
• Zimmet PZ ,
• Ruwaard D ,
• Bartelds AI ,
• Hirasing RA ,
• Verkleij H ,
• Birkeland KI ,
• Barnett KN ,
• Ogston SA ,
• Kerssens JJ ,
• McAllister DA ,
• Scottish Diabetes Research Network Epidemiology Group
• Stenström G ,
• Sundkvist G
• Jansson SP ,
• Andersson DK ,
• Svärdsudd K
• Ringborg A ,
• Lindgren P ,
• Martinell M ,
• Stålhammar J
• Schwenkglenks M ,
• Holden SH ,
• Barnett AH ,
• Peters JR ,
• Zghebi SS ,
• Steinke DT ,
• Rutter MK ,
• Emsley RA ,
• Ashcroft DM
• Akushevich I ,
• Kravchenko J ,
• Ukraintseva S ,
• O’Brien P ,
• Centers for Disease Control and Prevention (CDC)
• McBean AM ,
• Gilbertson DT ,
• Narayanan ML ,
• Schraer CD ,
• Bulkow LR ,
• Nichols GA ,
• Schroeder EB ,
• Karter AJ ,
• SUPREME-DM Study Group
• Tabaei BP ,
• Chamany S ,
• Driver CR ,
• Pavkov ME ,
• Hanson RL ,
• Knowler WC ,
• Bennett PH ,
• Krakoff J ,
• Lindström J ,
• Eriksson JG ,
• Finnish Diabetes Prevention Study Group
• Barrett-Connor E ,
• Fowler SE ,
• Diabetes Prevention Program Research Group
• Saaristo T ,
• Moilanen L ,
• Korpi-Hyövälti E ,
• Troughton J ,
• Chatterjee S ,
• Schmittdiel JA ,
• Neugebauer R ,
• Solomon LS ,
• Giles-Corti B ,
• Vernez-Moudon A ,
• Bolt-Evensen K ,
• Brand-Miller JC ,
• ↵ Bromley C, Dowling S, L G. The Scottish Health Survey. Scotland: A National Statistics Publication for Scotland, 2013.
• Carroll MD ,
• Flegal KM ,
• Kruszon-Moran D ,
• Freedman DS ,
• American Diabetes Association
• Lorenzo C ,
• Rasmussen SS ,
• Johansen NB ,

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here .

Loading metrics

Open Access

Peer-reviewed

Registered Report Protocol

Registered Report Protocols describe a study’s rationale and methods for which the planned work was peer-reviewed prior to data collection.

See all article types »

Type 1 and type 2 diabetes mellitus: Clinical outcomes due to COVID-19. Protocol of a systematic literature review

Contributed equally to this work with: Juan Pablo Pérez Bedoya, Alejandro Mejía Muñoz

Roles Conceptualization, Investigation, Methodology, Project administration, Writing – original draft

* E-mail: [email protected]

Current address: National Faculty of Public Health, University of Antioquia, Medellin, Antioquia, Colombia

Affiliation Epidemiology Group, National Faculty of Public Health, University of Antioquia, Medellín, Colombia

Affiliation Biology and Control of Infectious Diseases Group, Faculty of Exact and Natural Sciences, University of Antioquia, Medellín, Colombia

Roles Supervision, Validation, Writing – review & editing

¶ ‡ NCB and PADV also contributed equally to this work.

Affiliation Department of Translational Medicine, Herbert Wertheim College of Medicine & Department of Global Health, Robert Stempel College of Public Health and Social Work, Florida International University, Miami, FL, United States of America

• Juan Pablo Pérez Bedoya, 
• Alejandro Mejía Muñoz, 
• Noël Christopher Barengo, 
• Paula Andrea Diaz Valencia

• Published: September 9, 2022
• https://doi.org/10.1371/journal.pone.0271851
• See the preprint
• Peer Review
• Reader Comments

Introduction

Diabetes has been associated with an increased risk of complications in patients with COVID-19. Most studies do not differentiate between patients with type 1 and type 2 diabetes, which correspond to two pathophysiological distinct diseases that could represent different degrees of clinical compromise.

To identify if there are differences in the clinical outcomes of patients with COVID-19 and diabetes (type 1 and type 2) compared to patients with COVID-19 without diabetes.

Observational studies of patients with COVID-19 and diabetes (both type 1 and type 2) will be included without restriction of geographic region, gender or age, whose outcome is hospitalization, admission to intensive care unit or mortality compared to patients without diabetes. Two authors will independently perform selection, data extraction, and quality assessment, and a third reviewer will resolve discrepancies. The data will be synthesized regarding the sociodemographic and clinical characteristics of patients with diabetes and without diabetes accompanied by the measure of association for the outcomes. The data will be synthesized regarding the sociodemographic and clinical characteristics of patients with diabetes and without diabetes accompanied by the measure of association for the outcomes.

Expected results

Update the evidence regarding the risk of complications in diabetic patients with COVID-19 and in turn synthesize the information available regarding type 1 and type 2 diabetes mellitus, to provide keys to a better understanding of the pathophysiology of diabetics.

Systematic review registry

This study was registered at the International Prospective Registry for Systematic Reviews (PROSPERO)— CRD42021231942 .

Citation: Pérez Bedoya JP, Mejía Muñoz A, Barengo NC, Diaz Valencia PA (2022) Type 1 and type 2 diabetes mellitus: Clinical outcomes due to COVID-19. Protocol of a systematic literature review. PLoS ONE 17(9): e0271851. https://doi.org/10.1371/journal.pone.0271851

Editor: Alok Raghav, GSVM Medical College, INDIA

Received: July 7, 2022; Accepted: August 23, 2022; Published: September 9, 2022

Copyright: © 2022 Pérez Bedoya et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: No datasets were generated or analysed during the current study. All relevant data from this study will be made available upon study completion.

Funding: This research was developed within the framework of the project "Repository for the surveillance of risk factors for chronic diseases in Colombia, the Caribbean and the Americas" and has the financial support of the Ministry of Science, Technology and Innovation of Colombia—Minciencias 844 (grant number 111584467754). The opinions expressed are those of the authors and not necessarily of Minciencias.

Competing interests: The authors have declared that no competing interests exist.

The Severe Acute Respiratory Syndrome Coronavirus type 2 (SARS-CoV-2), the causal viral agent of coronavirus disease 2019 (COVID-19), currently has the world in one of the greatest public health crises of recent times since its appearance at the end of 2019 in the city of Wuhan, China [ 1 ]. The infection has a mild or even asymptomatic course in most cases, but in elderly patients (over 60 years-of-age) and in those with pre-existing chronic comorbidities, it can progress severe complications such as pneumonia, acute respiratory distress (ARDS) with hyperinflammatory involvement and multi-organ failure, leading in some cases to death [ 2 ].

Different studies have reported that patients diagnosed with diabetes who suffer from COVID-19 disease have higher morbidity and mortality compared with people without diabetes [ 3 ]. An analysis by Gude Sampedro et al. using prognostic models found that diabetic patients had greater odds of being hospitalized (OR 1.43; 95% CI: 1.18 to 1.73), admitted to the intensive care unit (OR 1.61; 95% CI: 1.12 to 2.31) and dying from COVID-19 (OR 1.79; 95% CI %: 1.38 to 2.32) compared with patients without diabetes [ 4 ]. However, it is difficult to establish whether diabetes alone directly contributed to the increase likelihood of complications.

Several studies using secondary data have emerged during the course of the pandemic that seek to determine the association of diabetes with mortality and other clinical outcomes in patients with COVID-19, such as, for example, a meta-analysis carried out by Shang et al. of severe infection and mortality from COVID-19 in diabetic patients compared with those without diabetes. They reported that patients with COVID-19 and diabetes had higher odds of serious infection (OR = 2.38, 95% CI: 2.05 to 2.78) and mortality (OR = 2, 21, 95% CI: 1.83 to 2.66) than patients without diabetes [ 5 ]. Despite the fact that there are several primary studies that attempt to explain the association between diabetes and COVID-19, most studies lack epidemiological rigor in the design and methodology used [ 6 ]. In addition, many of them did not distinguish between type 1 and type 2 diabetes, which are two very different conditions with different clinical development and pathophysiological mechanisms [ 7 ]. This may lead to different degrees of clinical complications from COVID-19. Currently, there is a gap in knowledge about the complications in patients with COVID-19 according to the type of diabetes. Moreover, only limited information exist how COVID-19 affects type 1 patients [ 8 , 9 ].

The objective of this systematic literature review will be to identify whether there are differences in the clinical outcomes of both type 1 and type 2 diabetes patients diagnosed with COVID-19 compared with patients with COVID-19 without a diagnosis of diabetes. This study will provide scientific evidence regarding the risk of complications in diabetic patients with COVID-19 and, in turn, synthesize the available information regarding to type 1 and type 2 diabetes.

Study design

This systematic literature review protocol was prepared according to the Preferred Reporting Elements for Systematic Review and Meta-Analysis Protocols (PRISMA-P) [ 10 ] ( S1 Appendix ). The results of the final systematic review will be reported according to the preferred reporting items for systematic reviews and meta-analyses (PRISMA 2020) [ 11 , 12 ]. In the event of significant deviations from this protocol, they will be reported and published with the results of the review.

Eligibility criteria

Participants (population)..

Patients with a confirmed diagnosis of COVID-19 without restriction of geographic region, sex, or age. For the diagnosis of COVID-19, the operational definition of confirmed case of the World Health Organization in its latest update will be used as a reference. Confirmed case of SARS-CoV-2 infection: a person with a positive Nucleic Acid Amplification Test (NAAT), regardless of clinical criteria OR epidemiological criteria or a person meeting clinical criteria AND/OR epidemiological criteria (suspect case A) with a positive professional- use or self-test SARS-CoV-2 Antigen RDT [ 13 ].

Patients with COVID-19 and concomitant diagnosis of unspecified diabetes mellitus, differentiated into type 1 diabetes mellitus or type 2 diabetes mellitus, without restriction of geographic region, gender, or age of the patients, who present definition of clinical criteria and /or paraclinical tests used by researchers to classify patients according to their diabetes status.

The operational definition of a confirmed case of diabetes mellitus provided by the American Diabetes Association will be used as a guide. The reference diagnostic criteria for diabetes are fasting plasma glucose ≥126 mg/dL (7.0 mmol/L). Fasting is defined as no caloric intake for at least 8 h or 2-h plasma glucose ≥ 200 mg/dL (11.1 mmol/L) during OGTT or hemoglobin A1C ≥6.5% (48 mmol/mol) or in a patient with classic symptoms of hyperglycemia or hyperglycemic crisis, at random plasma glucose ≥200 mg/dL [ 14 ].

In selected primary studies, identification of diabetes status may be based on medical history and International Classification of Diseases codes for type 1 or type 2 diabetes, use of antidiabetic medications, or previously defined diagnostic criteria.

Comparator.

Patients with COVID-19 who do not have a concomitant diagnosis of diabetes mellitus.

The main endpoint is all-cause mortality (according to the definitions of each primary study) and the secondary outcomes are hospitalization and admission to the ICU, where the authors specify a clear definition based on clinical practice guidelines and provide a well-defined criteria for patient outcomes.

Type of study.

Primary observational original research studies (prospective or retrospective cohort, case-control design, and cross-sectional studies) will be included in this systematic review.

Exclusion criteria

Clinical trials, editorials, letters to the editor, reviews, case reports, case series, narrative reviews or systematic reviews and meta-analyses, as well as research in the field of basic sciences based on experimental laboratory models, will be excluded. Original research articles that only include other types of diabetes, such as monogenic diabetes, gestational diabetes, latent autoimmune diabetes in adults, ketosis-prone diabetes, among others, or articles with publication status prior to publication will not be considered. In addition, articles whose main hypothesis is not diabetes and do not have the established outcomes will be excluded.

Information sources and search strategy

Electronic bibliographic databases..

For the preparation of the search strategy, the recommendations of the PRISMA-S guide [ 15 ] will be adopted. Relevant articles will be identified by electronic search applying the equation previously developed by the researchers and validated by an expert librarian ( S2 Appendix ). The following electronic bibliographic databases will be used: MEDLINE, EMBASE, LILACS, OVID MEDLINE, WHO (COVID-19 Global literature on coronavirus disease) and SCOPUS with a publication date from December 2019 to August 15, 2022, without language restriction.

The search for potential primary studies published in gray literature will be performed through the World Health Organization database for COVID-19 (WHO COVID-19 Global literature on coronavirus disease). This database contains different electronic bibliographic databases incorporated into its browser, including Web of Science, EuropePMC and Gray literature, among others.

Unlike electronic bibliographic databases.

To identify other potentially eligible studies, the references of relevant publications will be reviewed to perform a snowball manual search. This technique consists of searching for new articles from the primary studies already selected in order to guarantee exhaustiveness in the search.

Study selection process

Two researchers will independently evaluate all the titles and abstracts of the retrieved articles, using the free access Rayyan® software [ 16 ] with previously established selection criteria. Disagreements will be resolved in first instance through discussion and in the second instance through a third reviewer. Subsequently, the full text of the articles selected in the eligibility phase will be read independently by two researchers, both using the same instrument previously validated in Excel according to predefined criteria. Discrepancies will be resolved by discussion or a third reviewer. The process of identification, selection and inclusion of primary studies will be described and presented using the flowchart recommended by the PRISMA statement in its latest version 2020 [ 11 , 12 ].

Data collection and extraction

Standardized and validated forms will be used to collect the data extracted from the primary studies, accompanied by a detailed instruction manual to specify the guiding questions, and avoid the introduction of bias. Data will be extracted from those articles in full text format. If the full text is not available, contact the author or search for the manuscript with the help of the library system. This process will be carried out by two researchers independently. A third investigator will verify the extracted data to ensure the accuracy of the records. The authors of the primary studies will be contacted to resolve any questions that may arise. The reviewers will resolve the disagreements through discussion and one of the two referees will adjudicate the discrepancies presented through discussion and consensus.

In specific terms, the following data will be collected both for the primary studies that report diabetes and COVID-19 and for those that differentiate between DMT1 and DMT2: author, year and country where the study was carried out; study design; general characteristics of the population, sample size, demographic data of the participants (sex, age, ethnicity), percentage of patients with diabetes, percentage of patients with type 1 and/or type 2 diabetes, percentage of patients without diabetes, frequency of comorbidities in diabetics and non-diabetics, percentage of diabetic and non-diabetic patients who presented the outcomes (hospitalization, ICU admission and mortality) and association measures reported for the outcomes. Data extraction will be done using a Microsoft Excel 365 ® spreadsheets.

Quality evaluation

The study quality assessment tool provided by the National Institutes of Health (NIH) [ 17 ] will be used for observational studies such as cohort, case-control, and cross-sectional. Two tools will be sued: one for cohort and cross-sectional studies (14 questions/domains) and one for case-control studies (12 questions/domains). These tools are aimed at detecting elements that allow evaluation of possible methodological problems, including sources of bias (for example, patient selection, performance, attrition and detection), confounding, study power, the strength of causality in the association between interventions and outcomes, among other factors. The different tools that will be used reflect a score of "1" or "0" depending on the answer "yes" or "no", respectively for each question or domain evaluated, or failing that, the indeterminate criterion option. For observational cohort studies, which consist of 14 risk of bias assessment domains, the studies will be classified as having good quality if they obtain ≥10 points, of fair quality if they obtain 8 to 9 points, and of poor quality if they obtain less than 8 points. On the other hand, in the case of case-control studies that consist of 12 bias risk assessment domains, the studies will be classified as good quality if they obtained ≥8 points, regular quality if they obtained 6 to 7 points and of poor quality if they obtained less than 6 points. However, the internal discussion between the research team will always be considered as the primary quality criterion.

Data synthesis

A narrative synthesis with summary tables will be carried out according to the recommendations adapted from the Synthesis Without Meta-analysis (SWiM) guide to describe in a structured way the methods used, and the findings found in the primary studies, as well as the criteria for grouping of the studies [ 18 ]. A narrative synthesis will be presented in two sections, one for patients with COVID-19 and diabetes and another for patients with COVID-19 and type 1 or type 2 diabetes.

Assessment of clinical and methodological heterogeneity will determine the feasibility of the meta-analysis. Possible sources of heterogeneity identified are the clinical characteristics of the study population, the criteria used to define the outcomes in the groups of patients, the time period of the pandemic in which the study was carried out, and the availability of measurement and control for potential confounding factors. For this reason, it is established a priori that this diversity of findings will make it difficult to carry out an adequate meta-analysis [ 19 ]. However, if meta-analysis is considered feasible, the random effects model will be used due to the high probability of heterogeneity between studies. Statistical heterogeneity will be assessed using the X 2 test and the I 2 statistic, and publication bias assessed using funnel plots if there are sufficient (>10) studies [ 20 ].

Exploratory ecological analysis

An exploratory ecological analysis of the association between the frequency of clinical outcomes of diabetic patients with COVID-19 and the indicators related to the health care dimension, reported for the different countries analyzed by means of the correlation coefficient, will be carried out. The open public databases of the World Bank (WB) [ 21 ], the World Health Organization (WHO) [ 22 ] and Our World In Data [ 23 ] will be used to extract population indicators related to health care, among those prioritized, universal health coverage, hospital beds per 1,000 people, doctors per 1,000 people, current health spending as a percentage of gross domestic product (GDP), percentage of complete vaccination coverage for COVID-19.

Since the first epidemiological and clinical reports were released from the city of Wuhan regarding the clinical characteristics of patients with COVID-19, a high incidence of chronic non-communicable diseases has been observed in Covid-19 patients. Current scientific evidence has shown that certain comorbidities increase the risk for hospitalization, severity of illness or death from COVID-19, such as hypertension, cardiovascular disease, chronic kidney disease, chronic respiratory disease, diabetes, among others [ 24 ].

One of the main chronic comorbidities affected by the COVID-19 pandemic is diabetes. Multivariate analysis of several observational epidemiological studies have revealed that COVID-19 patients with diabetes were at increased risk of hospitalization, ICU admission, and mortality compared with patients without diabetes [ 4 ].

For this reason, it is expected that this systematic literature review will provide scientific support regarding the outcomes and complications that patients diagnosed with COVID-19 with type 1 or type 2 diabetes present compared with patients without diabetes. This information will be useful for healthcare personnel, public health professionals and epidemiologists involved in patient care or decision making, generating epidemiological evidence. Thus, highlighting the decisive role of epidemiological research in the context of the pandemic, especially in the field of diabetes epidemiology may improve comprehensive management and care of diabetic patients. This study may also provide important information that can be used to update of clinical practice guidelines.

Limitations

There are some potential limitations to the proposed systematic review. Firstly, both type 1 and type 2 diabetes may have different key search terms and some studies may be missed. To minimize this limitation, different search equations have been designed for each database in an exhaustive and sensitive manner. In addition to reading references and level ball as an additional strategy. Another limitation is that observational studies evaluating the effect of an intervention may be susceptible to significant confounding bias and may present high heterogeneity in the findings. To report these possible biases, an adequate quality assessment will be carried out, with highly sensitive and previously validated tools, exclusive for each type of observational design. The review is intended for publication in a peer-reviewed journal.

The status of the study

The study is in the selection phase of the records by applying the eligibility criteria to the titles and abstracts. Completion of the project is expected in September 2022 with the publication of the results.

Conclusions

This report describes the systematic review protocol that will be utilized to update the evidence regarding the risk of complications in diabetic patients with COVID-19 and in turn synthesize the information available regarding DM1 and DM2, to provide keys to a better understanding of the pathophysiology of diabetics.

Supporting information

S1 appendix. prisma-p (preferred reporting items for systematic review and meta-analysis protocols) 2015 checklist: recommended items to address in a systematic review protocol..

https://doi.org/10.1371/journal.pone.0271851.s001

S2 Appendix. Search string details for each database.

https://doi.org/10.1371/journal.pone.0271851.s002

• View Article
• PubMed/NCBI
• Google Scholar
• 13. World Health Organization. WHO COVID-19 Case definition Updated in Public health surveillance for COVID-19. 2022 July 22. [cited 18 August 2022]. In: World Health Organization [Internet]. Available from: https://www.who.int/publications/i/item/WHO-2019-nCoV-Surveillance_Case_Definition-2022.1
• 17. National Institutes of Health (NIH) [Internet]. Study Quality Assessment Tools; 2021. [cited 16 June 2022]. Available at: https://www.nhlbi.nih.gov/health-topics/study-quality-assessment-tools
• 19. Cochrane Handbook for Systematic Reviews of Interventions; 2022. [cited 16 June 2022]. Available at: https://training.cochrane.org/handbook/current
• 21. World Bank Group [Internet]. World Bank Indicators; 2022. [cited 16 June 2022]. Available at: https://datos.bancomundial.org/indicador
• 22. World Health Organization (WHO) [Internet]. Global Health Observatory Data Repository; 2021. [cited 16 June 2022]. Available at: https://apps.who.int/gho/data/node.home
• 23. Our World In Data [Internet]. Statistics and Research Coronavirus Pandemic (COVID-19); 2022. [cited 16 June 2022]. Available at: https://ourworldindata.org/coronavirus
• 24. Centers for Disease Control and Prevention [Internet]. COVID-19. People with Certain Medical Conditions; 2021. [cited 16 June 2022]. Available at: https://www.cdc.gov/coronavirus/2019-ncov/need-extra-precautions/people-with-medical-conditions.html?CDC_AA_refVal=ht

Utilities for Complications Associated with Type 2 Diabetes: A Review of the Literature

• Open access
• Published: 21 May 2024

Cite this article

You have full access to this open access article

• William J. Valentine   ORCID: orcid.org/0000-0003-4844-6813 1 ,
• Kirsi Norrbacka 2 &
• Kristina S. Boye 3  

201 Accesses

1 Altmetric

Explore all metrics

Introduction

Utility values are used in health economic modeling analyses of type 2 diabetes (T2D) to quantify the effect of acute and long-term complications on quality of life (QoL). For accurate modeling projections, it is important that the utility values used are up to date, accurate and representative of the simulated model cohort.

A literature review was performed to identify utility values for health states representing acute and chronic T2D-related complications including cardiovascular complications, stroke, renal disease, ophthalmic complications, neuropathy, diabetic foot, amputation and hypoglycemia. Searches were performed using the PubMed, Embase and Cochrane Library databases and limited to articles published since 2010. Supplementary searches were performed to identify data published at congresses in 2019–2023.

A total of 54 articles were identified that reported utility values for T2D-related complications. The most frequently used elicitation method/instrument was the EQ-5D ( n  = 42 studies) followed by the Short Form-6 dimensions ( n  = 6), time tradeoff ( n  = 5), the Health Utilities Index Mark 2 or Mark 3 ( n  = 2), 15D ( n  = 1), visual analog scale ( n  = 1) and standard gamble ( n  = 1). Stroke and amputation were consistently associated with the largest decrements in QoL. There is a lack of published data that distinguishes between severity of several complications including renal disease, retinopathy and neuropathy.

Conclusions

Diabetes-related complications can have a profound impact on QoL; therefore, it is important that these are captured accurately and appropriately in health economic models. Recently published utility values for diabetes-related complications that can be used to inform health economic models are summarized here.

Similar content being viewed by others

Estimating Quality of Life Decrements Due to Diabetes Complications in the United States: The Health Utility Index (HUI) Diabetes Complication Equation

Evaluation of health utility values for diabetic complications, treatment regimens, glycemic control and other subjective symptoms in diabetic patients using the EQ-5D-5L

Does the choice of eq-5d tariff matter a comparison of the swedish eq-5d-3l index score with uk, us, germany and denmark among type 2 diabetes patients.

Avoid common mistakes on your manuscript.

Health state utility values provide a quantitative indication of the extent to which a particular disease, complication or treatment-related side effect influences the quality of life (QoL) of an individual. Utility values quantify QoL on a scale of 0–1, where 1 represents perfect health and 0 represents death. Some utility elicitation methods, such as the EuroQoL 5 Dimensions (EQ-5D), allow negative values, which correspond to states considered to be worse than death [ 1 , 2 ]. Utility values are key inputs in health economic models that are used to project the long-term clinical and economic outcomes associated with new treatments and upon which payer and reimbursement decisions are often based. It is therefore important that these values accurately represent the QoL decrement associated with specific health states.

Individuals with type 2 diabetes (T2D) are at an elevated risk for a number of acute and long-term complications including cardiovascular disease, stroke, ophthalmic complications, renal disease, diabetic foot, neuropathy and hypoglycemic events, many of which can have a considerable impact on QoL. Poor glycemic control exacerbates the risk for long-term complications, and pharmacologic treatments aimed at improving glycemic control can in turn reduce the risk for long-term complications. However, the duration of clinical trials is not sufficient to capture the effect of new treatments on the incidence of long-term complications. Consequently, long-term health economic modeling analyses represent an important component of the reimbursement decision-making process for payers and policy makers.

The International Society for Pharmacoeconomics and Outcomes Research (ISPOR) good practice recommendations state that the use of up-to-date QoL data is an important aspect of long-term health economic analyses but also note that, in practice, many economic models use outdated utility values that may not adequately capture recent advances in treatment [ 3 ]. A previous review published by Beaudet et al. collated utility values for T2D-related complications, published over the period 1995–2012 [ 4 ]. The aim of the current review was to update and expand the previously published review by Beaudet et al. to provide a synopsis of utility values that could be utilized in future health economic models of T2D. The scope of the review was expanded to capture utility values elicited by the EQ-5D and other methods including (but not limited to) the Short Form 6 Dimensions (SF-6D), 15D, Health Utilities Index-2 or -3 (HUI-2/3), time-trade-off (TTO) and standard gamble (SG) and also to include utility/disutility values related to treatment-related attributes such as dosing frequency and timing flexibility, injection device-related attributes and unpleasant treatment-related side effects such as nausea or weight gain. The time frame of the review was also updated such that the review was limited to studies published since 2010, thereby capturing contemporary clinical practice relating to the management of people with T2D. Only findings related to T2D-related complications are presented here; summary findings of utilities for treatment-related attributes are presented separately in Part 2 of this review.

The literature review was performed using the PubMed, Embase and Cochrane Library databases. Search strategies were designed in alignment with recommendations outlined in the UK-based National Institute for Health and Care Excellence (NICE) Decision Support Unit (DSU) Technical Support Document 9 [ 5 ]. Search strategies utilized high level Medical Subject Heading (MeSH) terms supplemented with free-text terms, and search syntax was adjusted as required for use across the different databases (full details of the search strategies used are provided in Supplementary Material, Tables 1–3). Supplementary hand searches were also performed to identify pertinent studies presented at major congresses between late 2019 and 2023( specifically the virtual meeting of the American Diabetes Association [ADA], the ISPOR Annual Congress and the 55th annual meeting of the European Association for the Study of Diabetes [EASD] in 2019). Relevant abstracts presented at the 2019 ISPOR meeting have been published; therefore, relevant publications should have been captured within the literature database searches. Studies published only in abstract form prior to 2019 were excluded on the basis that study results were likely to have been subsequently published in full-text form.

The time horizon of the searches was limited to articles published since 2010, and all searches were performed in March 2020. For inclusion in the review, studies were required to be published in full-text form (except for recent abstracts as outlined above) in English and present utility or disutility values for health states related to acute or long-term T2D-related complications or treatment-related attributes or process characteristics. Complications captured in the review included cardiovascular disease (angina, myocardial infarction [MI], and congestive heart failure [CHF]), stroke, renal disease (albuminuria/proteinuria, end-stage renal disease [ESRD] and renal transplant), ophthalmic complications (retinopathy, macular edema, cataract and severe vision loss/blindness), neuropathy, diabetic foot, amputation, peripheral vascular disease, overweight/obesity, hypoglycemia and fear of hypoglycemia (FoH). Studies that were conducted in mixed populations of patients with type 1 and type 2 diabetes were excluded if results were not presented according to diabetes type. Secondary studies (i.e., studies listing previously published utility values), and discrete choice experiments were also excluded.

Here, reporting of results is limited to acute and long-term complications; utility values for treatment-related attributes and process characteristics are reported in a separate review [ 6 ].

This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.

Literature Searches

Literature searches across the three databases yielded a total of 8566 hits, which included 1383 duplicates, therefore resulting in a total of 7183 unique hits. First-round screening of titles and abstracts was performed by one investigator and identified a total of 241 hits for full-text review (Fig.  1 ). During second-round screening, a further 176 articles were excluded, leaving a total of 65 articles detailing utilities/disutilities associated with either T2D-related complications or treatment-related attributes for inclusion. A further three articles were identified via bibliographies of included articles. Searches of meeting abstract databases identified one relevant abstract for inclusion. The final review therefore included a total of 69 studies. Of these, a total of 39 presented findings exclusively related to acute or long-term diabetes-related complications, 15 presented findings exclusively related to the influence of treatment-related attributes on QoL, and 15 captured findings on both complications and treatment attributes. In total, 54 articles presented utility/disutility values for T2D-related complications (Table  1 ); these included a total of 18 studies conducted in Asia [ 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 ], 13 conducted in Europe [ 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 ], 13 in North America [ 38 , 39 , 40 , 41 , 42 , 43 , 44 , 45 , 46 , 47 , 48 , 49 , 50 ], 2 in Latin America [ 51 , 52 ], one in the Middle East [ 53 ] and 1 in Africa [ 54 ], and 6 were multinational (or setting not stated) [ 55 , 56 , 57 , 58 , 59 , 60 ]. Nearly all identified studies were conducted in individuals with T2D, although two presented data gathered from general population samples [ 34 , 56 ].

Summary of literature searches. T2D, type 2 diabetes. Publication/study type refers to articles that were reviews, editorials, letters, case reports, secondary sources of utility values or discrete choice experiments

T2D with No Complications

Eleven studies were identified that provided baseline utility values for individuals with T2D with no complications, all of which used the EQ-5D-3L or EQ-5D-5L (where studies did not state which version of the EQ-5D was used it was assumed that the EQ-5D-3L was used) (Table  2 ) [ 7 , 14 , 16 , 22 , 24 , 28 , 35 , 36 , 38 , 44 , 49 ] with one additionally using the SF-6D [ 38 ]. Five of the identified studies were conducted in Asia [ 7 , 14 , 16 , 22 , 24 ], three in North America [ 38 , 44 , 49 ] and three in European populations [ 28 , 35 , 36 ].

Using the EQ-5D, reported mean baseline values for individuals with T2D with no complications ranged from a low of 0.76 in a community-dwelling sample in Canada aged ≥ 40 years and with HbA1c ≥ 7% [ 44 ] to a high of 0.956 in a sample in China with mean (SD) age of 64.9 (9.1) years [ 14 ]. One study conducted in Vietnam reported a median baseline value of 1.0, which corresponds to perfect health [ 16 ], and is also slightly higher than that previously reported in a general population sample in Vietnam [ 61 ]. Differences in baseline characteristics such as age, BMI and duration of disease as well as the source of utility weights may have contributed towards the differences in baseline values between the different study populations.

For economic modeling analyses, the optimal baseline utility values should be as closely representative of the simulated patient population under investigation as possible. In particular, factors such as setting, baseline age and duration of diabetes should be taken into account and, if necessary, adjustments applied to the reference population. The description of the reference population of patients with no complications varied between studies; however, one US-based study provided a comprehensive description. Zhang et al. [ 49 ] reported a baseline utility value of 0.92 for no complications with this value referring specifically to individuals that were male, non-Hispanic white, with BMI < 30 kg/m 2 and with no risk factors for cardiovascular disease, with income > USD 40,000 per annum and not treated with insulin.

Cardiovascular and Cerebrovascular Complications

A total of ten studies that presented disutility values for MI [ 10 , 26 , 28 , 33 , 37 , 44 , 46 , 55 , 57 , 58 ], were identified (Table  3 ) (three studies presented utility values for MI; data not shown [ 11 , 36 , 46 ]). Two further studies presented disutilities for angina [ 10 , 46 ], and nine studies were identified that presented disutility values for CHF [ 10 , 27 , 28 , 37 , 46 , 49 , 55 , 57 , 58 ]. In studies that utilized the EQ-5D, the reported disutility associated with an MI ranged from − 0.073 in a sample of individuals with T2D attending outpatient clinics in South Korea [ 10 ] to − 0.0119 in a sample of patients in Sweden [ 28 ]. One study also reported a positive value of + 0.004 in a post hoc analysis of the multinational LEADER trial [ 58 ]. However, this value was not an event-related disutility per se but specifically referred to the change in utility value reported in patients who experienced an MI at any point during the 36-month follow-up period rather than a recent MI and may therefore potentially have captured improvements in QoL in patients recovering from an MI. Indeed, in a post hoc analysis of ACCORD trial data, Shao et al. noted that the timing of the event relative to the timing of measurement of QoL was an important determinant of QoL and although the effect of an MI waned over time a degree of long-term impairment persisted [ 46 ]. More specifically, using the HUI-3, Shao et al. reported that the mean (SE) disutility associated with MI was − 0.042 (0.016) in the year of the event but − 0.011 (0.006) in subsequent years after the event.

Two studies reported disutility values associated with angina (two studies also reported utility values for patients with angina; data not shown [ 36 , 46 ]). In a South Korean study that utilized the EQ-5D, Lee et al. reported a mean (SE) disutility of − 0.0266 (0.0114) [ 10 ], while for US-based patients, Shao et al., who utilized the HUI-3, reported a disutility of − 0.032 (0.006) in the year of event but − 0.010 (0.021) in subsequent years [ 46 ].

A total of nine studies reported disutility values for CHF [ 10 , 27 , 28 , 37 , 46 , 49 , 55 , 57 , 58 ], eight of which used the EQ-5D [ 10 , 27 , 28 , 37 , 49 , 55 , 57 , 58 ], with the remaining study using the HUI-3 [ 46 ] (four studies presented utility values for patients with CHF; data not shown [ 11 , 27 , 36 , 46 ]). In studies using the EQ-5D, the mean disutility for CHF ranged from − 0.0821 in Sweden [ 28 ] to − 0.037 in a study conducted in Greece [ 37 ]. Additionally, a US-based study that used the HUI-3 reported a disutility of − 0.089 in the year of diagnosis and − 0.041 in subsequent years [ 46 ].

The interpretation of the QoL impact of stroke was complicated by the heterogeneity in the definition of stroke used between different studies. In two studies, separate disutility values were presented for stroke and transient ischemic attack (TIA) [ 10 , 46 ] and one study further separated stroke into events with and without hemiplegia [ 46 ]. In comparison, other studies either did not state how stroke was defined or grouped TIA and stroke together. A total of 14 studies reported disutility values for stroke [ 10 , 24 , 26 , 28 , 33 , 35 , 37 , 44 , 46 , 49 , 54 , 55 , 57 , 58 ] (Table  4 ), and 8 studies presented utility values for stroke [ 11 , 18 , 24 , 29 , 36 , 38 , 46 , 54 ] (data not shown). Overall, stroke was typically associated with profound deficits in QoL relative to other complications. In studies that used the term stroke, disutility values ranged from − 0.59 to − 0.035 (Table  4 ). Notably, the two studies that reported the largest disutilities for stroke used the HUI-2 or HUI-3 [ 46 , 54 ], and another study reporting one of the smallest decrements of − 0.04 utilized the SF-6D [ 33 ]. When limited to studies that elicited utility values using the EQ-5D, the disutility associated with stroke ranged from − 0.135 [ 35 ] to − 0.035 [ 49 ]. Notably, the − 0.035 value reported in a US-based study by Zhang et al. did not include strokes resulting in hemiplegia; for events resulting in hemiplegia, the disutility was − 0.094 [ 49 ].

Renal Disease

Heterogeneity regarding definitions applied was also evident for studies presenting disutility values for renal complications. The literature review was initially designed to capture studies that reported utility/disutility values for the health states of microalbuminuria, macroalbuminuria, end-stage renal disease (ESRD), dialysis and renal transplant. However, few studies were identified that delineated renal disease according to this terminology, with many studies instead utilizing the overarching term of nephropathy. However, the definition of the nephropathy health state was in general poorly defined, and in instances where definitions were provided, these were inconsistent between studies. For example, Grandy et al. defined nephropathy as “self-reported diagnosis of chronic kidney disease, dialysis, ESRD, kidney transplant or proteinuria” [ 39 ]. In contrast, Luk et al. applied a more specific definition of nephropathy as “either proteinuria or chronic kidney disease (defined by the Renal Association as estimated glomerular filtration rate < 60 ml/min/1.73 m [ 2 ] on at least two occasions 90 days apart with or without markers of kidney damage, which include albuminuria, hematuria, electrolyte disorders due to tubular disorders, renal histologic abnormalities, structural abnormalities evident on imaging or a history of renal transplantation” [ 62 ]) [ 12 ]. The broad definition applied in analyses such as that presented by Grandy et al. means that the term nephropathy captures both patients with microalbuminuria and ESRD, despite these states representing very different severities of renal disease.

A total of seven studies presented disutility values for the broad health state of nephropathy [ 10 , 12 , 16 , 37 , 39 , 50 , 58 ] (all of which used the EQ-5D), and a further four studies presented values for ESRD [ 44 , 46 , 49 , 57 ], three of which used the EQ-5D [ 44 , 49 , 57 ] and one used the HUI-3 (Table  5 ) [ 46 ]. The mean disutility for nephropathy ranged from − 0.08 in a sample of patients with good glycemic control (median HbA1c 6.8%) and median duration of diabetes of 7 years based in Vietnam [ 16 ] to –0.0044 in patients with a mean age of 57 years based in South Korea [ 10 ]. For individuals with ESRD, the mean disutility value in EQ-5D studies ranged from –0.049 [ 57 ] to − 0.1018 [ 44 ], while the one study that utilized the HUI-3 reported a disutility of − 0.024 [ 46 ].

Ophthalmic Complications

The review was designed to capture the QoL impact of ophthalmic complications including retinopathy (background and proliferative), macular edema, cataract and severe vision loss/blindness. However, no studies were identified that examined the effect of macular edema on QoL, and only one study that presented a disutility value for cataract was identified (Table  6 ) [ 10 ]. Additionally, none of the retinopathy studies identified distinguished in severity between background and proliferative retinopathy; however, one Chinese study did distinguish between unilateral and bilateral retinopathy [ 15 ]. Here, using the EQ-5D-5L, the mean disutilities for unilateral and bilateral retinopathy were –0.013 and − 0.019, respectively [ 15 ]. A total of nine further studies reported disutility values for retinopathy elicited using the EQ-5D [ 10 , 14 , 16 , 24 , 26 , 28 , 37 , 39 , 58 ]. These included one US-based longitudinal study that captured the decline in QoL in individuals with retinopathy over a 5-year period. Using the EQ-5D, Grandy et al. reported a decline in EQ-5D of − 0.058 over 5 years, assuming a linear rate of decline over time corresponding to an annual decline of − 0.0116 per year [ 39 ]. Excluding the longitudinal study by Grandy et al., the mean disutility associated with retinopathy, elicited using the EQ-5D, ranged from − 0.0578 in a sample of people with T2D aged > 65 years (mean [SD] age 70.3 [9.3] years) based in Germany [ 26 ] to − 0.001 in a multinational study by Nauck et al. [ 58 ]. A further study utilized the 15D questionnaire and reported a mean disutility for retinopathy of − 0.036 [ 29 ]. The same authors also noted that in their analysis both the EQ-5D and SF-6D were relatively insensitive to retinopathy.

Three further studies presented disutility values for health states defined as “impaired vision,” “severe vision loss” and “blindness” (Table  6 ), findings that collectively suggested a deterioration in QoL with increasing severity of vision loss. In a Norwegian study, visual impairment (severity not defined) was associated with a mean (EQ-5D) disutility of − 0.012 [ 35 ]. In a US-based study, severe vision loss (defined as visual acuity of < 20/200 on a Snellen chart) was associated with a mean (HUI-3) disutility of − 0.057 [ 46 ], and in a multinational study blindness was associated with a mean (EQ-5D) disutility of − 0.083 [ 57 ].

Neuropathy, Foot Ulcer and Amputation

A total of 11 studies reported disutility values for people experiencing neuropathy (Table  7 ) [ 12 , 14 , 19 , 26 , 29 , 35 , 37 , 39 , 46 , 49 , 51 ]; however, only 1 disutility study distinguished between painful and non-painful neuropathy [ 49 ] (8 studies also reported utility values for the health state of neuropathy, 1 of which distinguished between painful and non-painful neuropathy; data not shown [ 9 , 12 , 13 , 14 , 18 , 24 , 29 , 46 ]). For US-based patients, using the EQ-5D, the mean disutility for patients with non-painful neuropathy was –0.039, but for painful neuropathy the mean decrement was as expected, notably more pronounced at –0.105 [ 49 ]. Another study identified in the review compared utility (rather than disutility values) and patient characteristics for those with painful versus non-painful neuropathy [ 9 ]. Patients with painful neuropathy were slightly older and had longer duration of disease than those with non-painful neuropathy, and the mean (SD) EQ-5D score was 0.8 (0.1) for those with non-painful neuropathy compared with 0.6 (0.3) for those with painful neuropathy [ 9 ]. Allied to this, a further US-based study captured the decline in QoL over time in patients with neuropathy, which may at least partially address the issue of progression/increased severity over time. Using the EQ-5D, Grandy et al. reported a decline in QoL of − 0.061 over a period of 5 years, which corresponds to an annual decline of − 0.0122 per year assuming a linear rate of decline [ 39 ].

The magnitude of the disutility associated with neuropathy was also influenced by the elicitation method used. For patients in Greece, Kontodimopoulos et al. report a disutility for neuropathy of − 0.117 using the SF-6D, but using the EQ-5D the corresponding value was − 0.247, suggesting different sensitivities to the impact of neuropathy between the two methods used [ 29 ].

Diabetic Foot and Amputation

The definitions and terminology used for diabetic foot/foot ulcer health states were heterogenous between studies and in some instances poorly defined, thereby making it challenging to differentiate between the impact of events such as the occurrence of superficial uninfected ulcers that healed without complications and infected/gangrenous ulcer. Additionally, two studies included amputation within the overarching terms of diabetic foot [ 50 , 58 ]. Reported mean disutility values for diabetic foot/foot ulcer ranged from − 0.016 in a Norwegian study by Solli et al. [ 35 ] to − 0.206 in a Greek study by Kontodimopoulos et al. [ 29 ], with both of these values elicited using the EQ-5D (Table  7 ). The broad range of disutility values may be partly attributable to the heterogeneity and differences in severity in terms of the terminology and categorizations used between different studies. Furthermore, as with neuropathy, reported disutility values were influenced by the elicitation method used. Using both the 15D and SF-6D Kontodimopoulos et al. reported a mean disutility of − 0.093, but when the EQ-5D-3L was used the mean utility decrement was notably greater at − 0.206 [ 29 ]. Additionally, for amputation specifically, reported disutility values (elicited using the EQ-5D) ranged from − 0.122 [ 57 ] to − 0.0631 [ 44 ] (Table  7 ).

Hypoglycemia

Literature searches identified a total of 24 studies that presented either utility or disutility values for patients experiencing hypoglycemia of various severities [ 11 , 12 , 17 , 20 , 21 , 24 , 25 , 30 , 31 , 32 , 34 , 37 , 40 , 41 , 42 , 43 , 45 , 47 , 48 , 55 , 56 , 58 , 59 , 60 ] as well as three studies that examined the effect of FoH on QoL [ 34 , 35 , 47 ]. As with other complications, the terminology and categorization of hypoglycemia varied widely between studies. Some studies categorized hypoglycemic events as either non-severe or severe, with non-severe events typically defined as events that could be resolved by the individual and severe events as those requiring third-party assistance. Other investigators used the categorization of mild, moderate, severe and very severe, with mild events typically defined as events causing no interruption of activities, moderate events as those resulting in some interruption of activities, severe events as requiring the assistance of others and very severe events as events that required medical assistance [ 41 , 60 ]. Some studies employed broader terminology, presenting utility/disutility values for any symptomatic event, and several focused on historical events describing utility values according to hypoglycemia experienced in the previous 1-, 3- or 6-month period. Additionally, a total of four studies distinguished between daytime and nocturnal hypoglycemic events [ 11 , 20 , 40 , 56 ]. As with long-term complications, the EQ-5D was the most frequently used utility elicitation method employed to determine the influence of hypoglycemic events on QoL; however, a total of five studies that specifically focused on hypoglycemic events used TTO methodology to determine disutility values [ 20 , 34 , 40 , 56 , 59 ].

Three studies reported disutility values for hypoglycemic events categorized as non-severe, all three of which included values elicited from both people with T2D and general population samples using TTO methodology [ 20 , 40 , 56 ]. For people with T2D, the disutility for a daytime non-severe event ranged from − 0.0028 in a Canadian study [ 40 ] to − 0.0283 in Malaysia (Table  8 ) [ 20 ]. Similarly, for general population samples, the corresponding range was from − 0.004 in a multinational study [ 56 ] to − 0.0354 in Malaysia [ 20 ]. For people with T2D, two of the three studies reported that a nocturnal non-severe event was associated with a greater decrement compared with an event occurring during the daytime. For example, in Canada, the decrement associated with a daytime non-severe event was − 0.0028, whereas a nocturnal non-severe event was associated with a utility decrement of − 0.0076 [ 40 ].

A further two studies (both of which used the EQ-5D) categorized events as either mild or moderate and reported disutility values of − 0.009 [ 41 ] and − 0.018 [ 31 ] for mild events and − 0.055 [ 41 ] and − 0.043 [ 31 ] for moderate events, respectively. A total of 13 studies examined the effect of severe hypoglycemic events on QoL [ 11 , 20 , 24 , 31 , 34 , 37 , 40 , 41 , 43 , 46 , 56 , 58 , 60 ], including 8 that reported disutility values for severe events [ 20 , 24 , 31 , 37 , 40 , 41 , 56 , 58 ]. In people with T2D, reported mean disutility values for a severe hypoglycemic event ranged from − 0.008 in China [ 24 ] (determined using the EQ-5D-3L) to − 0.1938 (determined using TTO methods) in a cross-sectional study conducted in Malaysia [ 20 ]. Three studies distinguished between severe events occurring during the day and nocturnal severe events [ 20 , 40 , 56 ], and in people with T2D nocturnal events were consistently associated with a greater disutility than daytime events.

Additionally, in a US-based study presenting disutility values elicited using the EQ-5D, Marrett et al. [ 41 ] distinguished between severe events requiring third-party assistance and very severe events requiring medical assistance; the mean disutilities associated with severe and very severe events were − 0.131 and − 0.208, respectively.

Three further studies quantified the influence of FoH on QoL; of these, two were conducted in people with T2D in Norway [ 35 ] and the USA [ 47 ], respectively, and the third was conducted in a UK-based general population sample [ 34 ]. Both studies conducted in people with T2D used the EQ-5D, and in the US, Shi et al. [ 47 ] reported that any FoH was associated with a disutility of − 0.003. In Norway, Solli et al. [ 35 ] reported that a large FoH was associated with a disutility of − 0.078 compared with a small FoH.

It is well established that diabetes-related complications can have a profound effect on QoL. Here, a synopsis of recently published disutility values that reflect current management/treatment practices is provided, which can be used in turn to inform future health economic models and analyses of novel interventions. New treatments for T2D generally provide incremental benefits in terms of glycemic control and/or adverse event rates, in particular hypoglycemia, relative to the standard of care. Glycemic control is a key determinant of the risk for long-term complications [ 63 ]. As such, for economic models to project valid outcomes such as quality-adjusted life expectancy for diabetes interventions, it is important that the most appropriate disutility values are used to best reflect the impact of individual complications on QoL.

When selecting utility/disutility values to inform economic models, there are a number of issues that warrant consideration by model developers. These include the elicitation method used and whether utility/disutility values selected are from a population that aligns with the simulated patient cohort under investigation. The choice of elicitation method is in some instances influenced by national guidelines. For example, for economic analyses performed in Sweden or Denmark, direct methods (e.g., TTO or SG) are preferred, whereas in other jurisdictions (e.g., England, Scotland) guidelines advocate the use of utility values elicited using the EQ-5D [ 64 ]. Indeed, in national guidelines where a specific generic multi-attribute utility instrument is recommended, this is most commonly the EQ-5D, although several countries do not stipulate which instrument should be used, instead providing examples of acceptable methods [ 65 ]. In terms of population, it may also be desirable to source disutility values that are generalizable to the simulated model cohort in terms of location, baseline demographics and disease characteristics. For example, values derived from individuals with newly diagnosed T2D based in Europe may not be appropriate when modeling new interventions in a population with long-standing disease in a country in Asia such as China or Singapore. The statistical approach used and the parameters adjusted for should also be considered when selecting utility values. In some instances, disutilities were calculated simply by subtracting the mean utility value of patients with a particular complication from the mean value for those without, while in other analyses disutilities were calculated using multivariate regression models and adjusted for baseline demographics and disease characteristics. The factors controlled for should be considered when selecting sources of disutility values. For example, women were consistently shown to have lower QoL relative to men [ 14 , 24 , 28 , 33 , 46 , 49 , 53 ], so it is important to consider whether values have been adjusted according to gender and other baseline demographics and disease characteristics.

Model developers may also need to consider the best way in which to address data gaps in terms of the need to utilize disutility values from multiple different sources. Research across different therapy areas has consistently shown that, for particular health states, utility values elicited using direct methods are consistently higher than values for the same state elicited using indirect methods [ 66 , 67 ], which may potentially lead to the introduction of bias in an analysis. Allied to this, for indirect methods consistency in terms of the source of preference weights used should also be considered owing to potential differences in cultural norms and healthcare provision between different settings [ 46 ]. Two studies identified in the current review showed that the source of preference weights can have a considerable influence on baseline utility values [ 28 , 57 ]. In particular, in a Swedish analysis, Kiadaliri et al. [ 28 ] noted that when using a Swedish tariff, the mean utility value for individuals with no complications was 0.89, but when using a UK tariff the corresponding value was considerably lower at 0.79. Similarly, when using the UK tariff, the disutilities associated with MI, heart failure and stroke were approximately two-fold greater than with the Swedish tariff [ 28 ].

A few studies identified in the review documented the phenomenon of diminishing marginal disutility specifically relating to cardiovascular events. Diminishing marginal disutility has previously been demonstrated with hypoglycemic events [ 68 ] and refers to instances where subsequent events are judged to have a lesser effect on QoL than first events. Briggs et al. observed diminishing marginal disutility with a composite endpoint of major cardiovascular event [ 55 ]. Here, a first major cardiovascular event was associated with a disutility of − 0.05, but the decrement associated with a subsequent event was less at − 0.038. Similar findings were reported by Kiadaliri et al. in terms of the effect of first and subsequent MIs [ 28 ]. Shao et al. also demonstrated that while the QoL impact of events such as MIs wanes over time, an MI within the previous year was associated with a disutility of − 0.042 but the disutility associated with a history of MI was − 0.011 [ 46 ]. However, with any retrospective analysis of events, it is possible that the interval between the event and utility elicitation may influence the QoL finding owing to response shift or recall bias, the potential impact of which should be considered when interpreting results [ 69 ].

Several diabetes-related complications such as retinopathy or renal disease may be classified according to severity. However, few studies identified in the review captured differing levels of severity for complications, although the notable exception to this was hypoglycemia, where events were frequently classified as non-severe and severe, or mild, moderate, severe and very severe. For retinopathy, no studies were identified that distinguished between background retinopathy and proliferative retinopathy. The decrement associated with vision loss/blindness, which may occur as a result of proliferative retinopathy, was however captured by several investigators. Additionally, only two studies distinguished between painful and non-painful neuropathy [ 9 , 49 ]. However, a longitudinal US-based study by Grandy et al. [ 39 ] captured the decline in EQ-5D index score over a time period of 5 years for patients with various complications including retinopathy and neuropathy. Annualizing this decrement may therefore represent an alternative method of modeling progression/increasing severity of complications over time.

For most complications, the EQ-5D was the most frequently used elicitation method, with more recently published studies tending to use the EQ-5D-5L rather than EQ-5D-3L. Two studies also directly compared values derived from the EQ-5D-3L and EQ-5D-5L in people with T2D, with both concluding that the 5L version showed greater sensitivity and discriminative ability as well as a reduced ceiling effect relative to the 3L [ 13 , 22 ]. Although the EQ-5D was frequently used for long-term complications, three studies that focused exclusively on hypoglycemic events used TTO methodology [ 20 , 40 , 56 ], with the rationale for this being the greater sensitivity of TTO methods relative to generic instruments such as the EQ-5D. This increased sensitivity may be particularly beneficial for complications such as hypoglycemia where events are distinguished by severity as well as the time of day/night at which events occur.

As with all literature reviews, there are limitations associated with the present review. No formal assessment of study quality or critical evaluation was performed; as a result, no ranking of the data is identified. This was a deliberate approach as different modeling and country-specific approaches may lead to differential priorities for the selection of utilities in a health economic analysis. Little was reported on the management of diabetes-related complications in the studies identified in this review. Different approaches to managing these conditions could lead to different utility scores being reported by patients across studies, likely contributing to between-country variation in outcomes reported. Demographic and socio-economic factors regarding utility scores were not reported in this review, which may also influence the selection of the most appropriate utility scores for health economic analysis. The application of inclusion and exclusion criteria was limited in the present review in an effort to align the work with that of Beaudet et al. and to minimize the risk of introducing bias through study selection [ 4 ]. However, this approach also has the potential to capture studies with smaller populations and/or limited quality and imply an equal weight to their results.

The ISPOR good practice recommendations note that economic modeling analyses should utilize up-to-date utility values as these reflect contemporary clinical practices and recent advances in treatment that may influence patients’ QoL [ 3 ]. The findings presented here provide a synopsis of recently published disutility values for major complications in T2D diabetes than can be used to inform future economic modeling analyses.

EuroQol 5 dimension questionnaire. Available at: https://euroqol.org/eq-5d-instruments/ . Last accessed September 14, 2020.

Brooks R, Boye KS, Slaap B. EQ-5D: a plea for accurate nomenclature. J Patient Rep Outcomes. 2020;4(1):52.

Article   PubMed   PubMed Central   Google Scholar  

Brazier J, Ara R, Azzabi I, Busschbach J, Chevrou-Séverac H, Crawford B, Cruz L, Karnon J, Lloyd A, Paisley S, Pickard AS. Identification, review, and use of health state utilities in cost-effectiveness models: an ISPOR good practices for outcomes research task force report. Value Health. 2019;22(3):267–75.

Article   PubMed   Google Scholar  

Beaudet A, Clegg J, Thuresson PO, Lloyd A, McEwan P. Review of utility values for economic modeling in type 2 diabetes. Value Health. 2014;17(4):462–70.

Papaioannou D, Brazier J, Paisley S. NICE DSU Technical Support Document 9: the identification, review and synthesis of health state utility values from the literature. October 201. http://www.nicedsu.org.uk .

Valentine W, Norrbacka K, Boye KS. Evaluating the impact of therapy on quality of life in type 2 diabetes: a literature review of utilities associated with treatment-related attributes. Patient Relat Outcome Meas. 2022;12(13):97–111.

Article   Google Scholar  

Butt M, Ali AM, Bakry MM. Health-related quality of life in poorly controlled type 2 diabetes patients- association of patients’ characteristics with EQ-5D domains, mean EQ-5D scores, and visaul analog scale score. Asian J Pharm Clin Res. 2018;11(1):93–8.

Ji L, Zou D, Liu L, Qian L, Kadziola Z, Babineaux S, Zhang HN, Wood R. Increasing body mass index identifies Chinese patients with type 2 diabetes mellitus at risk of poor outcomes. J Diabetes Complicat. 2015;29(4):488–96.

Kim SS, Won JC, Kwon HS, Kim CH, Lee JH, Park TS, Ko KS, Cha BY. Prevalence and clinical implications of painful diabetic peripheral neuropathy in type 2 diabetes: results from a nationwide hospital-based study of diabetic neuropathy in Korea. Diabetes Res Clin Pract. 2014;103(3):522–9.

Lee WJ, Song KH, Noh JH, Choi YJ, Jo MW. Health-related quality of life using the EuroQol 5D questionnaire in Korean patients with type 2 diabetes. J Korean Med Sci. 2012;27(3):255–60.

Lin YJ, Wang CY, Cheng SW, Ko Y. Patient preferences for diabetes-related complications in Taiwan. Curr Med Res Opin. 2019;35(1):7–13.

Article   CAS   PubMed   Google Scholar  

Luk AOY, Zhang Y, Ko GTC, Brown N, Ozaki R, et al. Health-related quality of life in chinese patients with type 2 diabetes: an analysis of the Joint Asia Diabetes Evaluation (JADE) program. J Diabetes Metab. 2014;5:333.

Google Scholar  

Pan CW, Sun HP, Wang X, Ma Q, Xu Y, Luo N, Wang P. The EQ-5D-5L index score is more discriminative than the EQ-5D-3L index score in diabetes patients. Qual Life Res. 2015;24(7):1767–74.

Pan CW, Sun HP, Zhou HJ, Ma Q, Xu Y, Luo N, Wang P. Valuing health-related quality of life in type 2 diabetes patients in China. Med Decis Making. 2016;36(2):234–41.

Pan CW, Wang S, Wang P, Xu CL, Song E. Diabetic retinopathy and health-related quality of life among Chinese with known type 2 diabetes mellitus. Qual Life Res. 2018;27(8):2087–93.

Pham TB, Nguyen TT, Truong HT, Trinh CH, Du HNT, Ngo TT, Nguyen LH. Effects of diabetic complications on health-related quality of life impairment in Vietnamese patients with type 2 diabetes. J Diabetes Res. 2020;23(2020):4360804.

Pratipanawatr T, Satirapoj B, Ongphiphadhanakul B, Suwanwalaikorn S, Nitiyanant W. Impact of hypoglycemia on health-related quality of life among type 2 diabetes: a cross-sectional study in Thailand. J Diabetes Res. 2019;23(2019):5903820.

Quah JH, Luo N, Ng WY, How CH, Tay EG. Health-related quality of life is associated with diabetic complications, but not with short-term diabetic control in primary care. Ann Acad Med Singap. 2011;40(6):276–86.

Riandini T, Wee HL, Khoo EYH, Tai BC, Wang W, Koh GCH, Tai ES, Tavintharan S, Chandran K, Hwang SW, Venkataraman K. Functional status mediates the association between peripheral neuropathy and health-related quality of life in individuals with diabetes. Acta Diabetol. 2018;55(2):155–64.

Shafie AA, Ng CH, Thanimalai S, Haron N, Manocha AB. Estimating the utility value of hypoglycaemia according to severity and frequency using the visual analogue scale (VAS) and time trade-off (TTO) survey. J Diabetes Metab Disord. 2018;17(2):269–75.

Terauchi Y, Ozaki A, Zhao X, Teoh C, Jaffe D, Tajima Y, Shuto Y. Humanistic and economic burden of cardiovascular disease related comorbidities and hypoglycaemia among patients with type 2 diabetes in Japan. Diabetes Res Clin Pract. 2019;149:115–25.

Wang P, Luo N, Tai ES, Thumboo J. The EQ-5D-5L is more discriminative than the EQ-5D-3L in patients with diabetes in Singapore. Value Health Reg Issues. 2016;9:57–62.

Yan BP, Zhang Y, Kong AP, Luk AO, Ozaki R, Yeung R, Tong PC, Chan WB, Tsang CC, Lau KP, Cheung Y, Wolthers T, Lyubomirsky G, So WY, Ma RC, Chow FC, Chan JC, Hong Kong JADE Study Group. Borderline ankle-brachial index is associated with increased prevalence of micro- and macrovascular complications in type 2 diabetes: a cross-sectional analysis of 12,772 patients from the Joint Asia Diabetes Evaluation Program. Diab Vasc Dis Res. 2015;12(5):334–44.

Zhang Y, Wu J, Chen Y, Shi L. EQ-5D-3L decrements by diabetes complications and comorbidities in China. Diabetes Ther. 2020;11(4):939–50.

Cvetanović G, Stojiljković M. Miljković M Estimation of the influence of hypoglycemia and body mass index on health-related quality of life, in patients with type 2 diabetes mellitus. Vojnosanit Pregl. 2017;74(9):831–9.

Hunger M, Schunk M, Meisinger C, Peters A, Holle R. Estimation of the relationship between body mass index and EQ-5D health utilities in individuals with type 2 diabetes: evidence from the population-based KORA studies. J Diabetes Complicat. 2012;26(5):413–8.

Kamradt M, Krisam J, Kiel M, Qreini M, Besier W, Szecsenyi J, Ose D. Health-related quality of life in primary care: which aspects matter in multimorbid patients with type 2 diabetes mellitus in a community setting? PLoS ONE. 2017;12(1): e0170883.

Kiadaliri AA, Gerdtham UG, Eliasson B, Gudbjörnsdottir S, Svensson AM, Carlsson KS. Health utilities of type 2 diabetes-related complications: a cross-sectional study in Sweden. Int J Environ Res Public Health. 2014;11(5):4939–52.

Kontodimopoulos N, Pappa E, Chadjiapostolou Z, Arvanitaki E, Papadopoulos AA, Niakas D. Comparing the sensitivity of EQ-5D, SF-6D and 15D utilities to the specific effect of diabetic complications. Eur J Health Econ. 2012;13(1):111–2.

Mitchell BD, Vietri J, Zagar A, Curtis B, Reaney M. Hypoglycaemic events in patients with type 2 diabetes in the United Kingdom: associations with patient-reported outcomes and self-reported HbA1c. BMC Endocr Disord. 2013;19(13):59.

Pagkalos E, Thanopoulou A, Sampanis C, Bousboulas S, Melidonis A, Tentolouris N, Alexandrides T, Migdalis I, Karamousouli E, Papanas N. The real-life effectiveness and care patterns of type 2 diabetes management in Greece. Exp Clin Endocrinol Diabetes. 2018;126(1):53–60.

Pettersson B, Rosenqvist U, Deleskog A, Journath G, Wändell P. Self-reported experience of hypoglycemia among adults with type 2 diabetes mellitus (Exhype). Diabetes Res Clin Pract. 2011;92(1):19–25.

Schunk M, Reitmeir P, Schipf S, Völzke H, Meisinger C, Ladwig KH, Kluttig A, Greiser KH, Berger K, Müller G, Ellert U, Neuhauser H, Tamayo T, Rathmann W, Holle R. Health-related quality of life in women and men with type 2 diabetes: a comparison across treatment groups. J Diabetes Complicat. 2015;29(2):203–11.

Article   CAS   Google Scholar  

Shingler S, Fordham B, Evans M, Schroeder M, Thompson G, Dewilde S, Lloyd AJ. Utilities for treatment-related adverse events in type 2 diabetes. J Med Econ. 2015;18(1):45–55.

Solli O, Stavem K, Kristiansen IS. Health-related quality of life in diabetes: the associations of complications with EQ-5D scores. Health Qual Life Outcomes. 2010;4(8):18.

Wermeling PR, Gorter KJ, van Stel HF, Rutten GE. Both cardiovascular and non-cardiovascular comorbidity are related to health status in well-controlled type 2 diabetes patients: a cross-sectional analysis. Cardiovasc Diabetol. 2012;5(11):121.

Yfantopoulos J, Chantzaras A. Health-related quality of life and health utilities in insulin-treated type 2 diabetes: the impact of related comorbidities/complications. Eur J Health Econ. 2020. https://doi.org/10.1007/s10198-020-01167-y . ( Epub ahead of print ).

Sayah FA, Qiu W, Xie F, Johnson JA. Comparative performance of the EQ-5D-5L and SF-6D index scores in adults with type 2 diabetes. Qual Life Res. 2017;26(8):2057–66.

Grandy S, Fox KM, SHIELD Study Group. Change in health status (EQ-5D) over 5 years among individuals with and without type 2 diabetes mellitus in the SHIELD longitudinal study. Health Qual Life Outcomes. 2012;10:99.

Harris S, Mamdani M, Galbo-Jørgensen CB, Bøgelund M, Gundgaard J, Groleau D. The effect of hypoglycemia on health-related quality of life: Canadian results from a multinational time trade-off survey. Can J Diabetes. 2014;38(1):45–52.

Marrett E, Radican L, Davies MJ, Zhang Q. Assessment of severity and frequency of self-reported hypoglycemia on quality of life in patients with type 2 diabetes treated with oral antihyperglycemic agents: a survey study. BMC Res Notes. 2011;21(4):251.

McCoy RG, Van Houten HK, Ziegenfuss JY, Shah ND, Wermers RA, Smith SA. Self-report of hypoglycemia and health-related quality of life in patients with type 1 and type 2 diabetes. Endocr Pract. 2013;19(5):792–9.

Meneghini LF, Lee LK, Gupta S, Preblick R. Association of hypoglycaemia severity with clinical, patient-reported and economic outcomes in US patients with type 2 diabetes using basal insulin. Diabetes Obes Metab. 2018;20(5):1156–65.

Article   CAS   PubMed   PubMed Central   Google Scholar  

O’Reilly DJ, Xie F, Pullenayegum E, Gerstein HC, Greb J, Blackhouse GK, Tarride JE, Bowen J, Goeree RA. Estimation of the impact of diabetes-related complications on health utilities for patients with type 2 diabetes in Ontario, Canada. Qual Life Res. 2011;20(6):939–43.

Pawaskar M, Iglay K, Witt EA, Engel SS, Rajpathak S. Impact of the severity of hypoglycemia on health—related quality of life, productivity, resource use, and costs among US patients with type 2 diabetes. J Diabetes Complicat. 2018;32(5):451–7.

Shao H, Yang S, Fonseca V, Stoecker C, Shi L. Estimating quality of life decrements due to diabetes complications in the United States: the Health Utility Index (HUI) diabetes complication equation. Pharmacoeconomics. 2019;37(7):921–9.

Shi L, Shao H, Zhao Y, Thomas NA. Is hypoglycemia fear independently associated with health-related quality of life? Health Qual Life Outcomes. 2014;30(12):167.

Williams SA, Pollack MF, Dibonaventura M. Effects of hypoglycemia on health-related quality of life, treatment satisfaction and healthcare resource utilization in patients with type 2 diabetes mellitus. Diabetes Res Clin Pract. 2011;91(3):363–70.

Zhang P, Brown MB, Bilik D, Ackermann RT, Li R, Herman WH. Health utility scores for people with type 2 diabetes in US managed care health plans: results from Translating Research Into Action for Diabetes (TRIAD). Diabetes Care. 2012;35(11):2250–6.

Zhao H, McClure NS, Johnson JA, Soprovich A, Al Sayah F, Eurich DT. A longitudinal study on the association between diabetic foot disease and health-related quality of life in adults with type 2 diabetes. Can J Diabetes. 2020;44(3):280-286.e1.

da Mata AR, Álvares J, Diniz LM, da Silva MR, Alvernaz dos Santos BR, Guerra Júnior AA, Cherchiglia ML, Andrade EI, Godman B, Acurcio Fde A. Quality of life of patients with diabetes mellitus types 1 and 2 from a referal health centre in Minas Gerais, Brazil. Expert Rev Clin Pharmacol. 2016;9(5):739–46.

Romero-Naranjo F, Espinosa-Uquillas C, Gordillo-Altamirano F, Barrera-Guarderas F. Which factors may reduce the health-related quality of life of ecuadorian patients with diabetes? P R Health Sci J. 2019;38(2):102–8.

PubMed   Google Scholar  

Javanbakht M, Abolhasani F, Mashayekhi A, Baradaran HR, Jahangiri Noudeh Y. Health related quality of life in patients with type 2 diabetes mellitus in Iran: a national survey. PLoS ONE. 2012;7(8):e44526.

Adibe MO, Aguwa CN. Sensitivity and responsiveness of health utility indices (HUI2 and HUI3) among type 2 diabetes patients. Trop J Pharm Res. 2013;12(5):835–42.

Briggs AH, Bhatt DL, Scirica BM, Raz I, Johnston KM, Szabo SM, Bergenheim K, Mukherjee J, Hirshberg B, Mosenzon O. Health-related quality-of-life implications of cardiovascular events in individuals with type 2 diabetes mellitus: a subanalysis from the Saxagliptin Assessment of Vascular Outcomes Recorded in Patients with Diabetes Mellitus (SAVOR)-TIMI 53 trial. Diabetes Res Clin Pract. 2017;130:24–33.

Evans M, Khunti K, Mamdani M, Galbo-Jørgensen CB, Gundgaard J, Bøgelund M, Harris S. Health-related quality of life associated with daytime and nocturnal hypoglycaemic events: a time trade-off survey in five countries. Health Qual Life Outcomes. 2013;3(11):90.

Hayes A, Arima H, Woodward M, Chalmers J, Poulter N, Hamet P, Clarke P. Changes in quality of life associated with complications of diabetes: results from the ADVANCE study. Value Health. 2016;19(1):36–41.

Nauck MA, Buse JB, Mann JFE, Pocock S, Bosch-Traberg H, Frimer-Larsen H, Ye Q, Gray A, LEADER Publication Committee for the LEADER Trial Investigators. Health-related quality of life in people with type 2 diabetes participating in the LEADER trial. Diabetes Obes Metab. 2019;21(3):525–32.

Polster M, Zanutto E, McDonald S, Conner C, Hammer M. A comparison of preferences for two GLP-1 products—liraglutide and exenatide—for the treatment of type 2 diabetes. J Med Econ. 2010;13(4):655–61.

Sheu WH, Ji LN, Nitiyanant W, Baik SH, Yin D, Mavros P, Chan SP. Hypoglycemia is associated with increased worry and lower quality of life among patients with type 2 diabetes treated with oral antihyperglycemic agents in the Asia-Pacific region. Diabetes Res Clin Pract. 2012.

Nguyen LH, Tran BX, Le Hoang QN, Tran TT, Latkin CA. Quality of life profile of general Vietnamese population using EQ-5D-5L. Health Qual Life Outcomes. 2017;15(1):199.

The Renal Association. Chronic kidney disease stages. Available at: https://renal.org/information-resources/the-uk-eckd-guide/ckd-stages/ . Last accessed July 29, 2020.

Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). UK Prospective Diabetes Study (UKPDS) Group. Lancet. 1998;352(9131):837–53. ( Erratum in: Lancet 1999 Aug 14;354(9178):602 ).

European Network for Health Technology Assessment. 2015. Methods for health economic evaluations. Available at: https://www.eunethta.eu/wpcontent/uploads/2018/03/Methods_for_health_economic_evaluations.pdf . Last accessed September 23, 2020.

Kennedy-Martin M, Slaap B, Herdman M, van Reenen M, Kennedy-Martin T, Greiner W, Busschbach J, Boye KS. Which multi-attribute utility instruments are recommended for use in cost-utility analysis? A review of national health technology assessment (HTA) guidelines. Eur J Health Econ. 2020;21(8):1245–57.

Blieden Betts M, Gandra SR, Cheng LI, Szatkowski A, Toth PP. Differences in utility elicitation methods in cardiovascular disease: a systematic review. J Med Econ. 2018;21(1):74–84.

Arnold D, Girling A, Stevens A, Lilford R. Comparison of direct and indirect methods of estimating health state utilities for resource allocation: review and empirical analysis. BMJ. 2009;22(339): b2688.

Lauridsen JT, Lønborg J, Gundgaard J, Jensen HH. Diminishing marginal disutility of hypoglycaemic events: results from a time trade-off survey in five countries. Qual Life Res. 2014;23(9):2645–50.

McPhail S, Haines T. Response shift, recall bias and their effect on measuring change in health-related quality of life amongst older hospital patients. Health Qual Life Outcomes. 2010;10(8):65.

Download references

Medical Writing and Editorial Assistance.

The authors are grateful to Jayne Smith-Palmer for medical writing support (when Dr. Palmer was an employee of Ossian). Eli Lilly and Company provided financial support to Ossian Health Economics and Communications to perform the medical writing work.

Eli Lilly and Company provided financial support to Ossian Health Economics and Communications to perform the literature review described in this paper. Eli Lilly and Company also funded the journal’s Rapid Service Fee and the Open Access Fee.

Author information

Authors and affiliations.

Ossian Health Economics and Communications GmbH, Bäumleingasse 20, 4051, Basel, Switzerland

William J. Valentine

Eli Lilly Finland, Helsinki, Finland

Kirsi Norrbacka

Eli Lilly and Company, Indianapolis, IN, USA

Kristina S. Boye

You can also search for this author in PubMed   Google Scholar

Contributions

William J Valentine, Kristina Secnik Boye and Kirsi Norrbacka were involved in the concept and design of the literature review, review of the data for inclusion in the publication and review and/or drafting of the manuscript for publication.

Corresponding author

Correspondence to William J. Valentine .

Ethics declarations

Conflict of interest.

William J Valentine is an employee of Ossian Health Economics and Communications. Krisi Norrbacka and Kristina Secnik Boye were both employees of Eli Lilly and Company at the time of the literature review.

Ethical Approval

Supplementary information.

Below is the link to the electronic supplementary material.

Supplementary file1 (PDF 213 KB)

Rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License, which permits any non-commercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc/4.0/ .

Reprints and permissions

About this article

Valentine, W.J., Norrbacka, K. & Boye, K.S. Utilities for Complications Associated with Type 2 Diabetes: A Review of the Literature. Adv Ther (2024). https://doi.org/10.1007/s12325-024-02878-x

Download citation

Received : 22 May 2023

Accepted : 17 April 2024

Published : 21 May 2024

DOI : https://doi.org/10.1007/s12325-024-02878-x

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Diabetes-related complications
• Health state utility
• Quality of life
• Type 2 diabetes
• Find a journal
• Publish with us
• Track your research

• Reference Manager
• Simple TEXT file

People also looked at

Review article, a systematic review of type 2 diabetes mellitus and hypertension in imaging studies of cognitive aging: time to establish new norms.

• 1 Baycrest Centre, Rotman Research Institute, Toronto, ON, Canada
• 2 Sunnybrook Research Institute, Heart and Stroke Foundation Canadian Partnership for Stroke Recovery, Toronto, ON, Canada
• 3 Department of Medical Biophysics, Faculty of Medicine, University of Toronto, Toronto, ON, Canada
• 4 Department of Nutritional Sciences, Faculty of Medicine, University of Toronto, Toronto, ON, Canada
• 5 Departments of Psychology and Psychiatry, University of Toronto, Toronto, ON, Canada

The rising prevalence of type 2 diabetes (T2DM) and hypertension in older adults, and the deleterious effect of these conditions on cerebrovascular and brain health, is creating a growing discrepancy between the “typical” cognitive aging trajectory and a “healthy” cognitive aging trajectory. These changing health demographics make T2DM and hypertension important topics of study in their own right, and warrant attention from the perspective of cognitive aging neuroimaging research. Specifically, interpretation of individual or group differences in blood oxygenation level dependent magnetic resonance imaging (BOLD MRI) or positron emission tomography (PET H 2 O 15 ) signals as reflective of differences in neural activation underlying a cognitive operation of interest requires assumptions of intact vascular health amongst the study participants. Without adequate screening, inclusion of individuals with T2DM or hypertension in “healthy” samples may introduce unwanted variability and bias to brain and/or cognitive measures, and increase potential for error. We conducted a systematic review of the cognitive aging neuroimaging literature to document the extent to which researchers account for these conditions. Of the 232 studies selected for review, few explicitly excluded individuals with T2DM (9%) or hypertension (13%). A large portion had exclusion criteria that made it difficult to determine whether T2DM or hypertension were excluded (44 and 37%), and many did not mention any selection criteria related to T2DM or hypertension (34 and 22%). Of all the surveyed studies, only 29% acknowledged or addressed the potential influence of intersubject vascular variability on the measured BOLD or PET signals. To reinforce the notion that individuals with T2DM and hypertension should not be overlooked as a potential source of bias, we also provide an overview of metabolic and vascular changes associated with T2DM and hypertension, as they relate to cerebrovascular and brain health.

Introduction

Amongst middle-aged and older adults, the rising prevalence of T2DM, hypertension, and other conditions that comprise the metabolic syndrome is a global health epidemic, attributed largely to sedentary lifestyles, poor diet, and lack of exercise. In 2008, it was estimated that 347 million adults worldwide had T2DM, up from 153 million in 1980 ( Danaei et al., 2011 ). Over the next two decades, it is expected that these numbers will continue to rise, by as much as 38% by 2030 ( Shaw et al., 2010 ). Prevalence rates of hypertension are even higher. In 2000, the global prevalence of hypertension was 26.4%, affecting an estimated 972 million people worldwide. Again, these numbers are expected to increase by approximately 60% by 2025, to a total of 1.56 billion people ( Kearney et al., 2005 ). Critically, hypertension is present in up to 75% of individuals with T2DM ( Colosia et al., 2013 ). The growing number of middle-aged and older adults living with T2DM and/or hypertension makes these conditions important topics of study in their own right.

Better long-term health care and disease management allow middle-aged and older adults to live with T2DM and hypertension for many years; however, both of these conditions have long-term deleterious effects on cerebrovascular and brain health, and contribute to cognitive impairment and decline ( Gorelick et al., 2011 ). T2DM and midlife hypertension confer a high risk for development of mild cognitive impairment (MCI) and dementia ( Launer et al., 2000 ; Kloppenborg et al., 2008 ; Creavin et al., 2012 ; Crane et al., 2013 ; Roberts et al., 2014 ), and older individuals with T2DM progress to dementia at faster rates ( Xu et al., 2010 ; Morris et al., 2014 ). These changing health demographics have created a discrepancy: what we define as “normal” or “typical” cognitive aging is becoming farther and farther removed from what would be considered optimal, or “healthy” cognitive aging.

This trend warrants attention from the perspective of cognitive aging research. Without adequate screening procedures in place, inclusion of individuals with T2DM and hypertension in otherwise healthy study samples may introduce unwanted variability and bias to brain and/or cognitive measures, and increase the potential for type 1 and type 2 errors. Functional neuroimaging studies may be particularly vulnerable in this regard. Blood oxygenation level dependent magnetic resonance imaging (BOLD MRI) and positron emission tomography (PET H 2 O 15 ) measure hemodynamic changes associated with neural activity, and thus provide an indirect measure of neural function ( Logothetis et al., 2001 ). To interpret individual or group differences in BOLD or PET signaling as reflective of individual or group differences in neural activation underlying a cognitive operation of interest, we rely on assumptions of intact neurovascular signaling, cerebrovascular reactivity, and vascular health amongst the study participants. These assumptions may be true in young and healthy individuals, but do not hold in older adults with conditions that affect vascular health ( D'Esposito et al., 2003 ). Even normal, age-related changes in the integrity of the cerebrovascular system can undermine these assumptions ( D'Esposito et al., 1999 ).

Yet, it was our impression that relatively few studies in the cognitive aging neuroimaging literature consider T2DM or hypertension during recruitment, or control for potential confounds associated with these conditions during analysis. To clarify the extent to which current research practices consider T2DM and hypertension in study design, we present the results of a systematic review of the cognitive aging neuroimaging literature, looking at study inclusion/exclusion criteria and methodology related to T2DM and hypertension. Then, to reinforce the notion that individuals with T2DM and hypertension should not be overlooked as a potential source of bias, we provide an overview of metabolic and vascular changes associated with T2DM and hypertension, as they relate to vascular health, structural brain atrophy, and functional integrity. The final section discusses best practices moving forward.

Systematic Review

This review focuses on the cognitive aging neuroimaging literature, however the issues associated with inclusion of individuals with T2DM and hypertension in study samples are by no means limited to this area of research. Any research study whose population of interest has high prevalence rates of T2DM or hypertension should be cognizant of these issues. For example, psychiatric populations have a higher incidence of metabolic disruption and T2DM that is mediated, at least partially, by the use of mood stabilizers, anticonvulsants, and antipsychotic medications ( Regenold et al., 2002 ; Newcomer and Haupt, 2006 ).

It should also be noted that the purpose of this review is not to quantitatively compare the results of studies that have excluded T2DM and/or hypertension with those that have not. This type of comparison is not feasible for numerous reasons, the primary one being that the extent to which individuals with T2DM or hypertension were present in study samples that did not screen for either condition is unknown. Rather, the aim of this review is to highlight the proportion of studies in the cognitive aging neuroimaging literature that consider T2DM and/or hypertension in their inclusion/exclusion criteria, or attempt to account for the potential bias introduced by inclusion of these individuals in their study groups.

We searched PsychInfo, MedLine, and PubMed between 1995 and February, 2013 using the search terms [“functional magnetic resonance imaging” or “positron emission tomography”], [“geriatrics” or “aging” or “age differences”], and [“cognit*” or “neuropsych*” or “memory” or “attention”]. Across the three databases, these search terms produced 704 unique empirical studies. From these results, we excluded studies that did not include a “healthy” or “normal” older adult sample ( n = 125), included a clinical sample other than MCI or Alzheimer disease (AD)/dementia (e.g., psychiatric; n = 46), did not use BOLD or PET H 2 O 15 imaging ( n = 227), and did not scan during a cognitive or resting state task ( n = 74; Figure 1 ).

Figure 1. Literature search terms and exclusion criteria . Based on these criteria, 232 studies were selected for review.

Based on these criteria, 232 studies were selected for review. These studies are identified with an asterisk (*) in the reference section. Two hundred and nineteen of these used BOLD imaging, one used both BOLD and PET H 2 O 15 , and 12 used PET H 2 O 15 only. One hundred and sixty five of these studies compared a “healthy” older group with a group of young participants, 34 studies compared a “healthy” older sample to an MCI and/or AD group (two of which also included a young adult comparison group), and the remaining 33 studies looked only at a “healthy” older sample. The majority of surveyed studies employed a memory paradigm during imaging (e.g., encoding/recognition of words, pictures, scenes, faces, autobiographical memory, spatial memory, associative memory, implicit learning). Working memory and executive processes were also well-studied (e.g., cognitive control, inhibition, decision making, mental rotation, task-switching, attention, judgment, processing speed, naming, imagery, verb generation, fluency). We also included resting-state studies in the sample.

Our primary concern was how sample selection was reported to have occurred. In particular, we were interested to learn how many studies specifically screened for T2DM and/or hypertension in their healthy older adult samples. For each of the 232 identified studies, the inclusion/exclusion criteria were examined according to the following criteria: (i) explicit exclusion of T2DM and/or hypertension, or exclusion of medical disorders/physical illnesses/systemic illnesses (implying that all medical conditions, including T2DM and hypertension, were excluded); (ii) exclusion of “significant,” “major,” or “severe” medical/physical/systemic disorders; or (iii) no screening criteria related to T2DM and/or hypertension provided. We also surveyed each of the 232 studies to determine how subjects were screened (e.g., self-report questionnaire, clinical assessment with a medical doctor, laboratory testing), and how—if at all—the potential influence of intersubject vascular variability on the measured BOLD or PET signals was addressed.

In each section below, superscript numbers, letters, and symbols are used to represent the extent to which studies screened for T2DM and hypertension, the screening method, and the degree to which studies attempted to account for intersubject vascular variability, respectively. The identified studies are denoted in the reference section according to these superscript classifiers.

Inclusion/exclusion of T2DM and hypertension

Of the 232 studies surveyed, only 22 (9.5%) explicitly excluded individuals with T2DM( 1 ), and only 29 (12.5%) explicitly excluded individuals with hypertension( 2 ). Thirteen studies—approximately 6%—excluded both T2DM and hypertension. Fourteen studies (6.0%) excluded individuals on antihypertensive medication( 3 ), however few of these studies also clarified whether individuals were assessed for untreated hypertension and excluded, if present. Nineteen studies (8.2%) excluded medical illnesses, systemic illnesses, medical disorders or physical illnesses( 4 ). This criterion implies that all medical conditions, including T2DM and hypertension, were excluded.

In contrast, almost half of the included studies (102; 44.0%) had exclusion criteria that made it difficult to determine whether T2DM was excluded( 5 ), and 85 studies (36.6%) had exclusion criteria that made it difficult to determine whether hypertension and/or antihypertensive medications were excluded( 6 ). These studies listed “major medical illnesses,” “significant medical conditions,” “serious systemic illnesses,” “conditions/medications interfering with cognitive and/or brain function,” “vascular disease,” “cardiovascular disease,” and/or “conditions/medications interfering with the fMRI signal” as exclusion criteria, or simply described their sample as “healthy.” There were also many studies that did not mention any selection criteria related to T2DM (80; 34.5%)( 7 ) or hypertension (51; 22.0%)( 8 ).

In addition, 26 studies (11.2%) included individuals with controlled hypertension( 9 ), 8 studies (3.5%) included controlled and uncontrolled hypertension( 10 ), 3 studies (1.3%) included individuals with controlled T2DM( 11 ), and 6 studies (2.5%) included individuals with controlled and uncontrolled T2DM in their healthy cohort( 12 ). Figure 2 provides a visual depiction of these results.

Figure 2. The extent to which T2DM and hypertension were accounted for in the inclusion/exclusion criteria of the healthy samples that were surveyed .

Screening method

The majority of studies (173; 75%) did not report how they conducted their medical screening( a ). Only 28 studies (12%) reported having screened subjects with physician-conducted medical examinations and/or laboratory testing( b ). Sixteen studies (7%) screened participants with telephone interviews, in-person clinical interviews, medical history, chart reviews, or a combination of these methods( c ). The remaining 15 studies (6%) used a self-report questionnaire to assess medical status( d ).

Accounting for intersubject vascular variability

A survey of the 232 included studies found that just under one third (29%) acknowledged and/or addressed the potential influence of intersubject vascular variability on the reported results. Many excluded subjects with a high vascular burden by screening for white matter hyperintensities in the imaging data( ■ ). Others compared groups on vascular risk factors( + ), compared outcome measures on hypertension status or antihypertensive treatment status( ♦ ), or attempted to control for health, blood pressure, and/or white matter hyperintensities in the reported associations ( ❖ ). Several studies noted in their discussion the possibility that the reported results were influenced by vascular factors, or explained why they did not think this was an issue( • ). A few studies used the measured BOLD or PET signals to examine and account for individual differences in vascular health( □ ); for example, by ensuring that groups were equated on BOLD signal variability, by comparing the temporal characteristics of the hemodynamic response curve across groups, with proportional scaling of the BOLD or PET signal, or by focusing on group by task interactions (instead of group main effects) or comparing within-subject task contrasts across individuals or groups to minimize any individual or group differences in vascular integrity.

There are rigorous ways to account for intersubject vascular variability, such as additional task data or an additional imaging contrast. Several studies included in the present review used arterial spin labeling (ASL) MRI ( ▴ ) or PET ( ▾ ) to measure resting cerebral blood flow and control for individual differences in perfusion. Three studies used a breath-hold task to index individual differences in cerebrovascular reactivity ( ❍ ), and two studies included a low-level motor or baseline task to ensure that participants demonstrated an adequate hemodynamic response ( × ).

Our results found that fewer than 10% of the selected functional imaging studies on cognitive aging explicitly excluded individuals with T2DM from their normative samples, and fewer than 15% explicitly excluded individuals with hypertension. A number of studies reported selection criteria that were insufficient to determine whether T2DM or hypertension were screened. Critically, one third of included studies had no reported inclusion or exclusion criteria related to T2DM, while almost a quarter had no reported inclusion or exclusion criteria related to hypertension. Only 67 of the 232 selected studies (29%) acknowledged or addressed the potential influence of intersubject vascular variability on the measured BOLD or PET signals.

Moreover, the large majority of studies did not include information about the medical screening process itself (e.g., laboratory testing vs. clinical interview vs. self-report questionnaire). This is not ideal when established tests for T2DM and hypertension are available (for example, 24-h ambulatory blood pressure monitoring would be the gold-standard for determining hypertension status, and an oral glucose tolerance test for determining T2DM status). Furthermore, we posit that participants may be less likely to volunteer T2DM or hypertension status as a “significant” medical illness without specific probing (i.e., compared to cancer, HIV, multiple sclerosis, or heart disease), because when these conditions are well-controlled they can have a minimal impact on day-to-day functioning, and, in the case of T2DM, can be controlled by diet alone. Collectively, these observations point to a lack of awareness that T2DM and hypertension are major medical illnesses that interfere significantly with cognitive and brain function in older adults.

Overview: Metabolic and Vascular Complications of Type 2 Diabetes Mellitus and Hypertension

To reinforce the position that T2DM and hypertension are conditions that can have a major effect on brain health and cognitive aging, this next section reviews evidence on the cognitive deficits, structural changes, and functional consequences associated with T2DM and hypertension, and describes some of the mechanisms that mediate these changes.

Type 2 Diabetes Mellitus

T2DM is the result of peripheral insulin resistance, which leads to insulin dysregulation and hyperglycemia. These metabolic changes affect cerebrovascular health, structural integrity, and brain function, and underlie the associations between T2DM, cognitive decline, and dementia risk.

Insulin dysregulation

Insulin is a peptide hormone that is critical for regulation of blood glucose levels. Binding of insulin to its receptors, found on nearly all cells throughout the body, facilitates the cellular uptake of glucose from the blood. When bound, insulin and insulin-like growth factor also activate complex intracellular signaling pathways that promote cell growth and survival, regulate glucose metabolism, and inhibit oxidative stress and apoptosis (for a review, see Nakae et al., 2001 ).

The defining characteristic of T2DM is peripheral insulin resistance, which occurs when cells in the body decrease their response to insulin stimulation. In the developing stage of this disease, the pancreas is able to produce enough insulin to overcome this resistance. This results in peripheral hyperinsulinemia, and blood glucose levels remain within the normal range. As the disease progresses, however, the pancreas can no longer keep up, and blood glucose levels begin to rise. When blood glucose levels are high even in the fasting state, T2DM is diagnosed.

Peripheral insulin resistance and hyperinsulinemia have a counterintuitive impact on insulin levels within the central nervous system. In the face of peripheral hyperinsulinemia, insulin transport across the blood brain barrier is effectively reduced, resulting in a brain hypo -insulinemic state (e.g., Heni et al., 2013 ). Low brain insulin levels and disrupted insulin signaling contribute to cognitive impairments directly, particularly in medial temporal lobe regions where insulin receptors are abundant ( Convit, 2005 ; Craft, 2006 ). Indirectly, low brain insulin levels exacerbate amyloid beta (Aβ) and tau pathology, hallmarks of Alzheimer disease (AD). It is here that we see the link between T2DM and Alzheimer disease pathology: brain insulin deficiency results in the down-regulation of insulin degrading enzyme (IDE; Luchsinger, 2008 ), which also has a role in degrading Aβ ( Carlsson, 2010 ). As a result, Aβ degradation is effectively reduced, contributing to its aggregation and amyloid plaque formation. Decreased brain insulin levels also suppress the enzymes involved in tau phosphorylation, contributing to the formation of neurofibrillary tangles ( Akter et al., 2011 ). While the downstream impact of T2DM-mediated brain insulin deficiency and insulin resistance is more moderate than that associated with AD, the underlying pathogenic mechanisms are similar ( Steen et al., 2005 ), and it has been proposed that AD is a form of diabetes mellitus that selectively affects the brain (T3DM; for discussion, see de la Monte and Wands, 2008 ). Given this, is not surprising that individuals with T2DM show a pattern of memory impairment, medial temporal lobe atrophy, and reduced hippocampal connectivity that is similar to the classic pattern of memory deficits, neurodegeneration, and network disruption in AD (e.g., Gold et al., 2007 ; Zhou et al., 2010 ; Baker et al., 2011 ; Cui et al., 2014 ).

Hyperglycemia

When cells in the body become resistant to the effects of insulin, blood glucose levels rise, resulting in hyperglycemia. Endothelial cells are particularly vulnerable to the effects of hyperglycemia, because they are less efficient at reducing glucose uptake in the face of high blood glucose levels ( Kaiser et al., 1993 ). Under such conditions, the resultant intracellular hyperglycemia induces an overproduction of reactive oxygen species in the mitochondria, which increases oxidative stress within the cell. This initiates a cascade of biochemical events that mediate much of the microvascular and macrovascular damage associated with T2DM including, but not limited to, increased intracellular formation of advanced glycation end-products (AGEs) and protein kinase C activation ( Du et al., 2000 ; Nishikawa et al., 2000 ; Brownlee, 2005 ; Giacco and Brownlee, 2010 ; Johnson, 2012 ).

AGEs are formed during normal metabolism on proteins with slower rates of turnover, in almost all cells throughout the body. AGE accumulation over time is a major factor in normal aging; however, under hyperglycemic conditions, AGE production is exacerbated beyond normal levels. AGEs cause intracellular damage and induce apoptosis through a process called cross-linking ( Shaikh and Nicholson, 2008 ). AGEs also contribute to oxidative stress, and themselves activate inflammatory signaling cascades (for a review, see Yan et al., 2008 ). Critically, under hyperglycemic conditions, the Aβ protein itself can act as an AGE ( Granic et al., 2009 ), which enhances its own aggregation and further increases amyloid plaque formation.

Protein kinase C activation, on the other hand, affects a variety of changes in gene expression that culminate in vascular dysfunction. Production of nitric oxide (NO), a vasodilator, is decreased, and production of endothelin-1, a vasoconstrictor, is increased. As a result, blood vessels are less able to dilate to accommodate increased blood flow demand. Over time, chronic exposure to high concentrations of endothelin-1 and decreased concentrations of NO contribute to diminished vessel elasticity, and structural changes in the vessel wall that result in atherosclerotic plaque formation ( Kalani, 2008 ).

In the brain, hyperglycemia-mediated macro- and microvascular damage reduces the delivery of nutrients and oxygen required to meet metabolic demands. Altered cerebral autoregulation has been observed in middle-aged adults with T2DM ( Brown et al., 2008 ), and may be an early manifestation of microvascular disease ( Kim et al., 2008 ). Older adults with T2DM show decreased blood flow velocity, increased cerebrovascular resistance, and impaired vasoreactivity ( Novak et al., 2006 ). Over time, declines in cerebrovascular health and reduced perfusion of brain tissue lead to structural atrophy and altered brain function.

Cognitive effects

The cognitive profile of individuals with T2DM includes deficits in attention, processing speed, learning and memory, and executive function (e.g., Reaven et al., 1990 ; Brands et al., 2007 ; Yeung et al., 2009 ; Whitehead et al., 2011 ). Moreover, these individuals, and individuals with pre-diabetes (impaired glucose tolerance), show an accelerated trajectory of cognitive decline relative to that associated with healthy aging ( Gregg et al., 2000 ; Fontbonne et al., 2001 ; Arvanitakis et al., 2004 ; Yaffe et al., 2004 ; Fischer et al., 2009 ; Nooyens et al., 2010 ; Espeland et al., 2011 ; for conflicting results, see van den Berg et al., 2010 ).

Cognitive deficits in T2DM have been linked to multiple disease-related processes, including: (i) poor glucose control (i.e., hemoglobin A1c [HbA1c]; Ryan and Geckle, 2000 ; Kanaya et al., 2004 ; Cukierman-Yaffe et al., 2009 ; Maggi et al., 2009 ; Luchsinger et al., 2011 ; Tuligenga et al., 2014 ; for conflicting results, see Christman et al., 2011 ), (ii) glucose intolerance ( Rizzo et al., 2010 ; Zhong et al., 2012b ), (iii) high peripheral AGE levels ( Yaffe et al., 2011 ), (iv) high levels of inflammatory cytokines ( Marioni et al., 2010 ), and (v) peripheral hyperinsulinemia and insulin resistance ( Bruehl et al., 2010 ; Zhong et al., 2012a ). Even in non-diabetic adults, poorer glucoregulation has been associated with deficits and/or declines in verbal memory, working memory, processing speed, and executive function ( Dahle et al., 2009 ; Bruehl et al., 2010 ; Messier et al., 2010 , 2011 ; Ravona-Springer et al., 2012 ).

The link between cognitive impairment and poor metabolic control may be largely mediated by the structural and functional brain changes that occur in the presence of chronic insulin dysregulation and hyperglycemia. Associations between glucoregulation, hypoperfusion in temporal regions, hippocampal atrophy, and memory impairment have been observed in T2DM ( Gold et al., 2007 ; Last et al., 2007 ), and in non-diabetic adults with decreased peripheral glucose regulation ( Convit et al., 2003 ), or high fasting plasma glucose levels within the normal range ( Cherbuin et al., 2012 ; Kerti et al., 2013 ). In other studies of T2DM, cognitive deficits and structural brain atrophy were linked to cerebral hypoperfusion and altered vascular reactivity ( Last et al., 2007 ; Brundel et al., 2012 ), and disrupted default-mode network connectivity was associated with peripheral hyperinsulinemia, insulin resistance, and white matter integrity ( Musen et al., 2012 ; Hoogenboom et al., 2014 ). Regardless of the underlying cause, brain atrophy in T2DM is associated with poor cognition ( Moran et al., 2013 ), and cognitive declines have been associated with progression of brain atrophy over time ( van Elderen et al., 2010 ; Reijmer et al., 2011 ). Some studies suggest that structural changes may occur early in the course of T2DM; enlarged lateral ventricles, particularly within the frontal horns, have been observed less than a year after diagnosis ( Lee et al., 2013 ), and middle-aged, as well as older adults with T2DM, show reduced prefrontal volumes ( Bruehl et al., 2009 ) and generalized global atrophy ( de Bresser et al., 2010 ; Kamiyama et al., 2010 ; Espeland et al., 2013 ).

Hypertension

The brain is one of the most highly perfused organs. The cerebral hemispheres are supplied by capillary beds connected to the pial vasculature by penetrating arterioles, and the pial vasculature stems from a system of arteries branching off the anterior, middle, and posterior cerebral arteries. Maintenance of brain function depends on a constant blood supply through this network. Hypertension causes changes to the structure and function of these blood vessels, which impacts perfusion in affected areas. Hypoperfusion, for example, can interfere with the delivery of oxygen and nutrients required to meet metabolic demands, and makes hypertension a major risk factor for vascular cognitive impairment, stroke, and dementia.

Cerebrovascular changes

Hypertension places enormous stress on the cerebral circulation (for a comprehensive review, see Pires et al., 2013 ). A hallmark of chronic hypertension is increased vascular resistance, particularly in the small blood vessels that perfuse the brain. Vascular resistance increases as vessel walls thicken. This remodeling is an adaptive response required to maintain chronically increased blood pressure, but it decreases the interior space of the blood vessels (the lumen). Vascular resistance also increases as the number of blood vessels decrease. Rat models of hypertension have shown both of these effects: reductions in lumen diameter and in the number of capillaries making up capillary beds in the cerebral vasculature ( Sokolova et al., 1985 ).

Blood flow is reduced when vascular resistance is high, and chronic hypertension-mediated hypoperfusion has been linked to white matter degradation, gray matter atrophy, and cognitive deficits. Studies of older adults with hypertension show reduced blood flow, particularly in occipito-temporal, prefrontal, and medial temporal lobe regions ( Beason-Held et al., 2007 ), positive correlations between blood pressure and white matter burden ( White et al., 2011 ; Raji et al., 2012 ), and negative correlations between blood pressure and total brain volume ( Nagai et al., 2008 ). Blood vessel function is also impacted by hypertension. Cerebral autoregulation (i.e., the ability to maintain a constant perfusion rate over a range of arterial pressures) is impaired, as is cerebrovascular reactivity, the ability of blood vessels to dilate to accommodate increased blood flow demand ( Last et al., 2007 ; Hajjar et al., 2010 ).

The cognitive profile of older adults with hypertension includes poorer performance on tests of executive function, including verbal fluency, Trails B-A switching score, Stroop interference scores ( Bucur and Madden, 2010 ), slowed processing speed ( Dahle et al., 2009 ), and deficits in attention and memory (see Gifford et al., 2013 for a meta-analysis). Prospective cohort studies show that midlife cardiovascular risk factors like hypertension predict cognitive impairment in later life (e.g., Virta et al., 2013 ), and, similarly, cross-sectional studies show a relation between higher systolic blood pressure and poorer cognitive performance, even within the normotensive range, a relation that is particularly strong in midlife (e.g., Knecht et al., 2008 , 2009 ). Hypertension is associated with decreases in cognitive reserve ( Giordano et al., 2012 ), and older adults with MCI and cardiovascular risk factors like hypertension are twice as likely to develop dementia compared to those without such risk factors ( Johnson et al., 2010 ; Ettorre et al., 2012 ). Moreover, cognitive declines may be faster in those with MCI and hypertension, compared to those without hypertension ( Li et al., 2011 ; Goldstein et al., 2013 ).

The association between hypertension and cognitive decline appears to be strongest in executive and processing speed domains, and weakest in memory and language domains. Hypertension increased the risk of non-amnestic MCI, but not amnestic MCI, regardless of APOEε 4 genotype or hypertensive medication status ( Reitz et al., 2007 ), and predicted progression to dementia in non-amnestic MCI, but not amnestic or multi-domain MCI ( Oveisgharan and Hachinski, 2010 ). The impact of hypertension on executive and processing speed domains is consistent with studies that show a positive relation between hypertension and white matter changes ( Kennedy and Raz, 2009 ; Raz et al., 2012 ), and between white matter changes and deficits in processing speed, executive function, and attention, but not memory (e.g., Debette et al., 2011 ).

Cognitive deficits in hypertensive adults are linked to various indicators of vascular and brain health. There are correlations between white matter integrity and performance on tests of executive function and attention ( Hannesdottir et al., 2009 ), and between decreased flow-mediated dilation and poorer executive function ( Smith et al., 2011 ). Deficits in attention and psychomotor speed in late middle-aged adults with hypertension are associated with reductions in global brain perfusion, reductions that were not fully ameliorated following 6-months of antihypertensive treatment ( Efimova et al., 2008 ). Global cognitive decline has been linked to reduced cerebral blood flow in the face of white matter lesions and lacunar infarcts ( Kitagawa et al., 2009 ), to higher pulse pressure and arterial stiffness ( Scuteri et al., 2007 ; Waldstein et al., 2008 ; Triantafyllidi et al., 2009 ), and to hypertension-mediated deep-brain vascular pathology ( Yakushiji et al., 2012 ). In another large study of patients with MCI, those with hypertension and deep white matter lesions were at higher risk of dementia ( Clerici et al., 2012 ).

Conclusions

Taken together, these studies provide abundant evidence that middle-aged and older adults with T2DM and hypertension, relative to healthy older adults, are more likely to show signs of cognitive dysfunction, widespread structural atrophy, vascular damage, and functional changes. In light of their rising prevalence amongst older adults, there is an increasing likelihood that, without adequate screening at recruitment, individuals with T2DM and/or hypertension will be included in healthy older adult samples. This may introduce unwanted variability and bias to brain and/or cognitive measures, and increase the potential for type 1 and type 2 errors. Given the state of the neuroimaging literature on this topic and the need to advance our understanding, we view T2DM and hypertension as important new frontiers in cognitive neuroscience.

Moving forward, there is an opportunity to develop best practices when it comes to cognitive neuroscience research in older adult populations. Reconciling the vascular risk component in T2DM and hypertension may be the most tractable option since there are myriad approaches one can take to do this. The most rigorous approach in this respect may be inclusion of a breath-hold task, or a measure of cerebral blood flow (e.g., ASL) in the functional imaging protocol, as this allows for a direct estimate of each subject's vascular health. Breath-hold tasks can be used to index cerebrovascular reactivity in response to non-neuronal signals. The breath-hold period induces hypercapnia, which stimulates vasodilation and increases blood flow and blood volume in the brain, a signal change that occurs independently of neuronal activation. ASL or resting-state PET scans provide a direct measure of blood flow, and can be used to account for individual differences in perfusion. As noted above, these methods have already been used in some studies of cognitive aging to account for individual differences in cerebrovascular health. Whether other means of equating vascular risk across participants or across groups (e.g., screening participants for excessive white matter hyperintensities, post-hoc comparison of outcome measures or study groups on vascular risk factors, or statistical analyses aimed at controlling for the effects of vascular variability in the reported results) are similarly effective requires further study.

It may also be important for investigators to acknowledge a distinction between “healthy” and “typical” brain aging. Studies characterizing healthy aging should adopt T2DM and hypertension as exclusion criteria. Conversely, given the high prevalence of T2DM and hypertension in older adults, community- or population-based studies characterizing the typical trajectory of cognitive aging would benefit by including these participants to maximize the generalizability of results, and reconciling the heterogeneity through study design groups (e.g., stratifying based on diagnosis of T2DM and hypertension) or covariates in their analysis.

As the proportion of older adults living with T2DM and hypertension increase, it is imperative that functional imaging studies are designed to account for these population trends. The current state of the cognitive aging neuroimaging literature suggests that there is limited appreciation and/or awareness that T2DM and hypertension are significant medical illnesses that disrupt brain vasculature, brain structure, and brain function. By adopting best practices that take T2DM and hypertension into account, we can advance our understanding of these conditions, and of cognitive aging in general.

Author Contributions

Liesel-Ann C. Meusel selecting, indexing, and reviewing articles, writing of drafts; Nisha Kansal selecting articles, editing of drafts; Ekaterina Tchistiakova contributing to the first draft, editing of drafts; William Yuen selecting articles, contributing to the first draft, editing of drafts; Bradley J. MacIntosh provided conceptual foundation for paper, editing of drafts; Carol E. Greenwood provided conceptual foundation for paper, editing of drafts; Nicole D. Anderson provided conceptual foundation for paper, editing of drafts.

Conflict of Interest Statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Acknowledgments

This research was supported in part by postdoctoral fellowships from the Centre for Stroke Recovery and the Alzheimer Society of Canada awarded to Liesel-Ann C. Meusel, and grant funds from CIHR (MOP111244).

Akter, K., Lanza, E. A., Martin, S. A., Myronyuk, N., Rua, M., and Raffa, R. B. (2011). Diabetes mellitus and Alzheimer's disease: shared pathology and treatment? Br. J. Clin. Pharmacol . 71, 365–376. doi: 10.1111/j.1365-2125.2010.03830.x

Pubmed Abstract | Pubmed Full Text | CrossRef Full Text

* 5,6,a Anderson, K. E., Lynch, K., Zarahn, E., Scarmeas, N., Van Heertum, R., Sackeim, H. et al. (2005). H215O PET study of impairment of nonverbal recognition with normal aging. J. Neuropsychiatry Clin. Neurosci . 17, 192–200. doi: 10.1176/appi.neuropsych.17.2.192

* 7,8,d Anguera, J. A., Reuter-Lorenz, P. A., Willingham, D. T., and Seidler, R. D. (2011). Failure to engage spatial working memory contributes to age-related declines in visuomotor learning. J. Cogn. Neurosci . 23, 11–25. doi: 10.1162/jocn.2010.21451

* 5,6,a Ansado, J., Monchi, O., Ennabil, N., Faure, S., and Joanette, Y. (2012). Load-dependent posterior-anterior shift in aging in complex visual selective attention situations. Brain Res . 1454, 14–22. doi: 10.1016/j.brainres.2012.02.061

* 5,6,a,□ Antonova, E., Parslow, D., Brammer, M., Dawson, G. R., Jackson, S. H. D., and Morris, R. G. (2009). Age-related neural activity during allocentric spatial memory. Memory 17, 125–143. doi: 10.1080/09658210802077348

Arvanitakis, Z., Wilson, R. S., Bienias, J. L., Evans, D. A., and Bennett, D. A. (2004). Diabetes mellitus and risk of Alzheimer disease and decline in cognitive function. Arch. Neurol . 61, 661–666. doi: 10.1001/archneur.61.5.661

* 5,6,b Bäckman, L., Karlsson, S., Fischer, H., Karlsson, P., Brehmer, Y., Rieckmann, A. et al. (2011). Dopamine D(1) receptors and age differences in brain activation during working memory. Neurobiol. Aging 32, 1849–1856. doi: 10.1016/j.neurobiolaging.2009.10.018

* 5,6,a Bagurdes, L. A., Mesulam, M. M., Gitelman, D. R., Weintraub, S., and Small, D. M. (2008). Modulation of the spatial attention network by incentives in healthy aging and mild cognitive impairment. Neuropsychologia 46, 2943–2948. doi: 10.1016/j.neuropsychologia.2008.06.005

* 5,6,c Bai, F., Liao, W., Watson, D. R., Shi, Y., Wang, Y., Yue, C. et al. (2011). Abnormal whole-brain functional connection in amnestic mild cognitive impairment patients. Behav. Brain Res . 216, 666–672. doi: 10.1016/j.bbr.2010.09.010

* 5,6,a Bai, F., Zhang, Z., Yu, H., Shi, Y., Yuan, Y., Zhu, W. et al. (2008). Default-mode network activity distinguishes amnestic type mild cognitive impairment from healthy aging: a combined structural and resting-state functional MRI study. Neurosci. Lett . 438, 111–115. doi: 10.1016/j.neulet.2008.04.021

Baker, L. D., Cross, D. J., Minoshima, S., Belongia, D., Watson, G. S., and Craft, S. (2011). Insulin resistance and Alzheimer-like reductions in regional cerebral glucose metabolism for cognitively normal adults with prediabetes or early type 2 diabetes. Arch. Neurol . 68, 51–57. doi: 10.1001/archneurol.2010.225

* 5,6,c,▴ Bangen, K. J., Kaup, A. R., Mirzakhanian, H., Wierenga, C. E., Jeste, D. V., and Eyler, L. T. (2012). Compensatory brain activity during encoding among older adults with better recognition memory for face-name pairs: an integrative functional, structural, and perfusion imaging study. J. Int. Neuropsychol. Soc . 18, 402–413. doi: 10.1017/S1355617712000197

* 9,12,b,▴ Bangen, K. J., Restom, K., Liu, T. T., Jak, A. J., Wierenga, C. E., Salmon, D. P. et al. (2009). Differential age effects on cerebral blood flow and BOLD response to encoding: associations with cognition and stroke risk. Neurobiol. Aging 30, 1276–1287. doi: 10.1016/j.neurobiolaging.2007.11.012

* 7,8,a,▾ Beason-Held, L. L., Kraut, M. A., and Resnick, S. M. (2008). I. Longitudinal changes in aging brain function. Neurobiol. Aging 29, 483–496. doi: 10.1016/j.neurobiolaging.2006.10.031

Beason-Held, L. L., Moghekar, A., Zonderman, A. B., Kraut, M. A., and Resnick, S. M. (2007). Longitudinal changes in cerebral blood flow in the older hypertensive brain. Stroke 38, 1766–1773. doi: 10.1161/STROKEAHA.106.477109

* 9,12,c,+ Beeri, M. S., Lee, H., Cheng, H., Wollman, D., Silverman, J. M., and Prohovnik, I. (2011). Memory activation in healthy nonagenarians. Neurobiol. Aging 32, 515–523. doi: 10.1016/j.neurobiolaging.2009.02.022

* 1,2,a Berlingeri, M., Bottini, G., Danelli, L., Ferri, F., Traficante, D., Sacheli, L. et al. (2010). With time on our side? Task-dependent compensatory processes in graceful aging. Exp. Brain Res . 205, 307–324. doi: 10.1007/s00221-010-2363-7

* 4,a Bernard, F. A., Desgranges, B., Eustache, F., and Baron, J.-C. (2007). Neural correlates of age-related verbal episodic memory decline: a PET study with combined subtraction/correlation analysis. Neurobiol. Aging 28, 1568–1576. doi: 10.1016/j.neurobiolaging.2006.07.004

* 5,6,a Bollinger, J., Rubens, M. T., Masangkay, E., Kalkstein, J., and Gazzaley, A. (2011). An expectation-based memory deficit in aging. Neuropsychologia 49, 1466–1475. doi: 10.1016/j.neuropsychologia.2010.12.021

Brands, A. M. A., Van den Berg, E., Manschot, S. M., Biessels, G. J., Kappelle, L. J., De Haan, E. H. F. et al. (2007). A detailed profile of cognitive dysfunction and its relation to psychological distress in patients with type 2 diabetes mellitus. J. Int. Neuropsychol. Soc . 13, 288–297. doi: 10.1017/S1355617707070312

* 2,5,a Braskie, M. N., Landau, S. M., Wilcox, C. E., Taylor, S. D., O'Neil, J. P., Baker, S. L. et al. (2011). Correlations of striatal dopamine synthesis with default network deactivations during working memory in younger adults. Hum. Brain Mapp . 32, 947–961. doi: 10.1002/hbm.21081

* 5,6,a Braskie, M. N., Small, G. W., and Bookheimer, S. Y. (2009). Entorhinal cortex structure and functional MRI response during an associative verbal memory task. Hum. Brain Mapp . 30, 3981–3992. doi: 10.1002/hbm.20823

* 1,10,b,❖ Braskie, M. N., Small, G. W., and Bookheimer, S. Y. (2010). Vascular health risks and fMRI activation during a memory task in older adults. Neurobiol. Aging 31, 1532–1542. doi: 10.1016/j.neurobiolaging.2008.08.016

Brown, C. M., Marthol, H., Zikeli, U., Ziegler, D., and Hilz, M. J. (2008). A simple deep breathing test reveals altered cerebral autoregulation in type 2 diabetic patients. Diabetologia 51, 756–761. doi: 10.1007/s00125-008-0958-3

Brownlee, M. (2005). The pathobiology of diabetic complications: a unifying mechanism. Diabetes 54, 1615–1625. doi: 10.2337/diabetes.54.6.1615

Bruehl, H., Sweat, V., Hassenstab, J., Polyakov, V., and Convit, A. (2010). Cognitive impairment in nondiabetic middle-aged and older adults is associated with insulin resistance. J. Clin. Exp. Neuropsychol . 32, 487–493. doi: 10.1080/13803390903224928

Bruehl, H., Wolf, O. T., Sweat, V., Tirsi, A., Richardson, S., and Convit, A. (2009). Modifiers of cognitive function and brain structure in middle-aged and elderly individuals with type 2 diabetes mellitus. Brain Res . 1280, 186–194. doi: 10.1016/j.brainres.2009.05.032

Brundel, M., van den Berg, E., Reijmer, Y. D., de Bresser, J., Kappelle, L. J., and Biessels, G. J. (2012). Cerebral haemodynamics, cognition and brain volumes in patients with type 2 diabetes. J. Diabetes Complicat . 26, 205–209. doi: 10.1016/j.jdiacomp.2012.03.021

Bucur, B., and Madden, D. J. (2010). Effects of adult age and blood pressure on executive function and speed of processing. Exp. Aging Res . 36, 153–168. doi: 10.1080/03610731003613482

* 7,9,a Burgmans, S., van Boxtel, M. P. J., Vuurman, E. F. P. M., Evers, E. A. T., and Jolles, J. (2010). Increased neural activation during picture encoding and retrieval in 60-year-olds compared to 20-year-olds. Neuropsychologia 48, 2188–2197. doi: 10.1016/j.neuropsychologia.2010.04.011

* 2,5,a Cabeza, R., Anderson, N. D., Locantore, J. K., and McIntosh, A. R. (2002). Aging gracefully: compensatory brain activity in high-performing older adults. Neuroimage 17, 1394–1402. doi: 10.1006/nimg.2002.1280

* 2,5,d Cabeza, R., Daselaar, S. M., Dolcos, F., Prince, S. E., Budde, M., and Nyberg, L. (2004). Task-independent and task-specific age effects on brain activity during working memory, visual attention and episodic retrieval. Cereb. Cortex 14, 364–375. doi: 10.1093/cercor/bhg133

* 2,5,a Cabeza, R., Grady, C. L., Nyberg, L., McIntosh, A. R., Tulving, E., Kapur, S. et al. (1997). Age-related differences in neural activity during memory encoding and retrieval: a positron emission tomography study. J. Neurosci . 17, 391–400.

Pubmed Abstract | Pubmed Full Text

* 7,8,a Campbell, K. L., Grady, C. L., Ng, C., and Hasher, L. (2012). Age differences in the frontoparietal cognitive control network: implications for distractibility. Neuropsychologia 50, 2212–2223. doi: 10.1016/j.neuropsychologia.2012.05.025

* 7,8,a Cappell, K. A., Gmeindl, L., and Reuter-Lorenz, P. A. (2010). Age differences in prefontal recruitment during verbal working memory maintenance depend on memory load. Cortex 46, 462–473. doi: 10.1016/j.cortex.2009.11.009

* 7,8,a Carlson, M. C., Erickson, K. I., Kramer, A. F., Voss, M. W., Bolea, N., Mielke, M. et al. (2009). Evidence for neurocognitive plasticity in at-risk older adults: the experience corps program. J. Gerontol. A Biol. Sci. Med. Sci . 64, 1275–1282. doi: 10.1093/gerona/glp117

Carlsson, C. M. (2010). Type 2 diabetes mellitus, dyslipidemia, and Alzheimer's disease. J. Alzheimers. Dis . 20, 711–722. doi: 10.3233/JAD-2010-100012

* 7,8,a Celone, K. A., Calhoun, V. D., Dickerson, B. C., Atri, A., Chua, E. F., Miller, S. L. et al. (2006). Alterations in memory networks in mild cognitive impairment and Alzheimer's disease: an independent component analysis. J. Neurosci . 26, 10222–10231. doi: 10.1523/JNEUROSCI.2250-06.2006

* 2,5,a Chen, N., Chou, Y., Song, A. W., and Madden, D. J. (2009). Measurement of spontaneous signal fluctuations in fMRI: adult age differences in intrinsic functional connectivity. Brain Struct. Funct . 213, 571–585. doi: 10.1007/s00429-009-0218-4

Cherbuin, N., Sachdev, P., and Anstey, K. J. (2012). Higher normal fasting plasma glucose is associated with hippocampal atrophy: the PATH Study. Neurology 79, 1019–1026. doi: 10.1212/WNL.0b013e31826846de

Christman, A. L., Matsushita, K., Gottesman, R. F., Mosley, T., Alonso, A., Coresh, J. et al. (2011). Glycated haemoglobin and cognitive decline: the Atherosclerosis Risk in Communities (ARIC) study. Diabetologia 54, 1645–1652. doi: 10.1007/s00125-011-2095-7

* 5,6,a Chua, E. F., Schacter, D. L., and Sperling, R. A. (2009). Neural basis for recognition confidence in younger and older adults. Psychol. Aging 24, 139–153. doi: 10.1037/a0014029

* 5,6,b Clément, F., and Belleville, S. (2009). Test-retest reliability of fMRI verbal episodic memory paradigms in healthy older adults and in persons with mild cognitive impairment. Hum. Brain Mapp . 30, 4033–4047. doi: 10.1002/hbm.20827

Clerici, F., Caracciolo, B., Cova, I., Fusari Imperatori, S., Maggiore, L., Galimberti, D. et al. (2012). Does vascular burden contribute to the progression of mild cognitive impairment to dementia? Dement. Geriatr. Cogn. Disord . 34, 235–243. doi: 10.1159/000343776

Colosia, A. D., Palencia, R., and Khan, S. (2013). Prevalence of hypertension and obesity in patients with type 2 diabetes mellitus in observational studies: a systematic literature review. Diabetes Metab. Syndr. Obes . 6, 327–338. doi: 10.2147/DMSO.S51325

Convit, A. (2005). Links between cognitive impairment in insulin resistance: an explanatory model. Neurobiol. Aging 26(Suppl. 1), 31–35. doi: 10.1016/j.neurobiolaging.2005.09.018

Convit, A., Wolf, O. T., Tarshish, C., and de Leon, M. J. (2003). Reduced glucose tolerance is associated with poor memory performance and hippocampal atrophy among normal elderly. Proc. Natl. Acad. Sci. U.S.A . 100, 2019–2022. doi: 10.1073/pnas.0336073100

* 7,8,b Cook, I. A., Bookheimer, S. Y., Mickes, L., Leuchter, A. F., and Kumar, A. (2007). Aging and brain activation with working memory tasks: an fMRI study of connectivity. Int. J. Geriatr. Psychiatry 22, 332–342. doi: 10.1002/gps.1678

Craft, S. (2006). Insulin resistance syndrome and Alzheimer disease: pathophysiologic mechanisms and therapeutic implications. Alzheimer Dis. Assoc. Disord . 20, 298–301. doi: 10.1097/01.wad.0000213866.86934.7e

Crane, P. K., Walker, R., Hubbard, R. A., Li, G., Nathan, D. M., Zheng, H. et al. (2013). Glucose levels and risk of dementia. N. Engl. J. Med . 369, 540–548. doi: 10.1056/NEJMoa1215740

Creavin, S. T., Gallacher, J., Bayer, A., Fish, M., Ebrahim, S., and Ben-Shlomo, Y. (2012). Metabolic syndrome, diabetes, poor cognition, and dementia in the Caerphilly prospective study. J. Alzheimers Dis . 28, 931–939. doi: 10.3233/JAD-2011-111550

Cui, Y., Jiao, Y., Chen, Y.-C., Wang, K., Gao, B., Wen, S. et al. (2014). Altered spontaneous brain activity in type 2 diabetes: a resting-state functional MRI study. Diabetes 63, 749–760. doi: 10.2337/db13-0519

Cukierman-Yaffe, T., Gerstein, H. C., Williamson, J. D., Lazar, R. M., Lovato, L., Miller, M. E. et al. (2009). Relationship between baseline glycemic control and cognitive function in individuals with type 2 diabetes and other cardiovascular risk factors: the action to control cardiovascular risk in diabetes-memory in diabetes (ACCORD-MIND) trial. Diabetes Care 32, 221–226. doi: 10.2337/dc08-1153

Dahle, C. L., Jacobs, B. S., and Raz, N. (2009). Aging, vascular risk, and cognition: blood glucose, pulse pressure, and cognitive performance in healthy adults. Psychol. Aging 24, 154–162. doi: 10.1037/a0014283

Danaei, G., Finucane, M. M., Lu, Y., Singh, G. M., Cowan, M. J., Paciorek, C. J. et al. (2011). National, regional, and global trends in fasting plasma glucose and diabetes prevalence since 1980: systematic analysis of health examination surveys and epidemiological studies with 370 country-years and 2·7 million participants. Lancet 378, 31–40. doi: 10.1016/S0140-6736(11)60679-X

* 2,5,d Daselaar, S. M., Fleck, M. S., Dobbins, I. G., Madden, D. J., and Cabeza, R. (2006). Effects of healthy aging on hippocampal and rhinal memory functions: an event-related fMRI study. Cereb. Cortex 16, 1771–1782. doi: 10.1093/cercor/bhj112

* 7,8,d Daselaar, S. M., Veltman, D. J., Rombouts, S. A. R. B., Raaijmakers, J. G. W., and Jonker, C. (2003). Neuroanatomical correlates of episodic encoding and retrieval in young and elderly subjects. Brain 126, 43–56. doi: 10.1093/brain/awg005

* 2,5,a Davis, S. W., Dennis, N. A., Daselaar, S. M., Fleck, M. S., and Cabeza, R. (2008). Que PASA? The posterior-anterior shift in aging. Cereb. Cortex 18, 1201–1209. doi: 10.1093/cercor/bhm155

Debette, S., Seshadri, S., Beiser, A., Au, R., Himali, J. J., Palumbo, C. et al. (2011). Midlife vascular risk factor exposure accelerates structural brain aging and cognitive decline. Neurology 77, 461–468. doi: 10.1212/WNL.0b013e318227b227

de Bresser, J., Tiehuis, A. M., van den Berg, E., Reijmer, Y. D., Jongen, C., Kappelle, L. J. et al. (2010). Progression of cerebral atrophy and white matter hyperintensities in patients with type 2 diabetes. Diabetes Care 33, 1309–1314. doi: 10.2337/dc09-1923

* 7,9,a de Chastelaine, M., Wang, T. H., Minton, B., Muftuler, L. T., and Rugg, M. D. (2011). The effects of age, memory performance, and callosal integrity on the neural correlates of successful associative encoding. Cereb. Cortex 21, 2166–2176. doi: 10.1093/cercor/bhq294

de la Monte, S. M., and Wands, J. R. (2008). Alzheimer's disease is type 3 diabetes-evidence reviewed. J. Diabetes Sci. Technol . 2, 1101–1113. doi: 10.1177/193229680800200619

* 5,6,a Dennis, N. A., and Cabeza, R. (2011). Age-related dedifferentiation of learning systems: an fMRI study of implicit and explicit learning. Neurobiol. Aging 32, 2318.e17–30. doi: 10.1016/j.neurobiolaging.2010.04.004

* 5,6,a Dennis, N. A., Daselaar, S., and Cabeza, R. (2007a). Effects of aging on transient and sustained successful memory encoding activity. Neurobiol. Aging 28, 1749–1758. doi: 10.1016/j.neurobiolaging.2006.07.006

* 5,6,a Dennis, N. A., Hayes, S. M., Prince, S. E., Madden, D. J., Huettel, S. A., and Cabeza, R. (2008a). Effects of aging on the neural correlates of successful item and source memory encoding. J. Exp. Psychol. Learn. Mem. Cogn . 34, 791–808. doi: 10.1037/0278-7393.34.4.791

* 5,6,a Dennis, N. A., Kim, H., and Cabeza, R. (2008b). Age-related differences in brain activity during true and false memory retrieval. J. Cogn. Neurosci . 20, 1390–1402. doi: 10.1162/jocn.2008.20096

* 5,6,a Dennis, N. A., Kim, H., and Cabeza, R. (2007b). Effects of aging on true and false memory formation: an fMRI study. Neuropsychologia 45, 3157–3166. doi: 10.1016/j.neuropsychologia.2007.07.003.

D'Esposito, M., Deouell, L. Y., and Gazzaley, A. (2003). Alterations in the BOLD fMRI signal with ageing and disease: a challenge for neuroimaging. Nat. Rev. Neurosci . 4, 863–872. doi: 10.1038/nrn1246

D'Esposito, M., Zarahn, E., Aguirre, G. K., and Rypma, B. (1999). The effect of normal aging on the coupling of neural activity to the bold hemodynamic response. Neuroimage 10, 6–14. doi: 10.1006/nimg.1999.0444

* 5,6,b,• Dickerson, B. C., Salat, D. H., Greve, D. N., Chua, E. F., Rand-Giovannetti, E., Rentz, D. M. et al. (2005). Increased hippocampal activation in mild cognitive impairment compared to normal aging and AD. Neurology 65, 404–411. doi: 10.1212/01.wnl.0000171450.97464.49

* 5,6,a,× DiGirolamo, G. J., Kramer, A. F., Barad, V., Cepeda, N. J., Weissman, D. H., Milham, M. P. et al. (2001). General and task-specific frontal lobe recruitment in older adults during executive processes: a fMRI investigation of task-switching. Neuroreport 12, 2065–2071. doi: 10.1097/00001756-200107030-00054

* 1,6,b,■ Donix, M., Poettrich, K., Weiss, P. H., Werner, A., von Kummer, R., Fink, G. R. et al. (2010). Age-dependent differences in the neural mechanisms supporting long-term declarative memories. Arch. Clin. Neuropsychol . 25, 383–395. doi: 10.1093/arclin/acq037

* 5,6,a Drobyshevsky, A., Baumann, S. B., and Schneider, W. (2006). A rapid fMRI task battery for mapping of visual, motor, cognitive, and emotional function. Neuroimage 31, 732–744. doi: 10.1016/j.neuroimage.2005.12.016

Du, X. L., Edelstein, D., Rossetti, L., Fantus, I. G., Goldberg, H., Ziyadeh, F. et al. (2000). Hyperglycemia-induced mitochondrial superoxide overproduction activates the hexosamine pathway and induces plasminogen activator inhibitor-1 expression by increasing Sp1 glycosylation. Proc. Natl. Acad. Sci. U.S.A . 97, 12222–12226. doi: 10.1073/pnas.97.22.12222

* 1,2,a,■ Duarte, A., Graham, K. S., and Henson, R. N. (2010). Age-related changes in neural activity associated with familiarity, recollection and false recognition. Neurobiol. Aging 31, 1814–1830. doi: 10.1016/j.neurobiolaging.2008.09.014

* 1,2,a,■ Duarte, A., Henson, R. N., and Graham, K. S. (2008). The effects of aging on the neural correlates of subjective and objective recollection. Cereb. Cortex 18, 2169–2180. doi: 10.1093/cercor/bhm243

* 3,5,a Dulas, M. R., and Duarte, A. (2011). The effects of aging on material-independent and material-dependent neural correlates of contextual binding. Neuroimage 57, 1192–1204. doi: 10.1016/j.neuroimage.2011.05.036

* 3,5,a Dulas, M. R., and Duarte, A. (2012). The effects of aging on material-independent and material-dependent neural correlates of source memory retrieval. Cereb. Cortex 22, 37–50. doi: 10.1093/cercor/bhr056

* 7,9,a Duverne, S., Motamedinia, S., and Rugg, M. D. (2009). The relationship between aging, performance, and the neural correlates of successful memory encoding. Cereb. Cortex 19, 733–744. doi: 10.1093/cercor/bhn122

Efimova, I. Y., Efimova, N. Y., Triss, S. V., and Lishmanov, Y. B. (2008). Brain perfusion and cognitive function changes in hypertensive patients. Hypertens. Res . 31, 673–678. doi: 10.1291/hypres.31.673

* 1,2,a Emery, L., Heaven, T. J., Paxton, J. L., and Braver, T. S. (2008). Age-related changes in neural activity during performance matched working memory manipulation. Neuroimage 42, 1577–1586. doi: 10.1016/j.neuroimage.2008.06.021

* 7,8,a Erickson, K. I., Colcombe, S. J., Wadhwa, R., Bherer, L., Peterson, M. S., Scalf, P. E. et al. (2007). Training-induced plasticity in older adults: effects of training on hemispheric asymmetry. Neurobiol. Aging 28, 272–283. doi: 10.1016/j.neurobiolaging.2005.12.012

Espeland, M. A., Bryan, R. N., Goveas, J. S., Robinson, J. G., Siddiqui, M. S., Liu, S. et al. (2013). Influence of type 2 diabetes on brain volumes and changes in brain volumes: results from the Women's Health Initiative Magnetic Resonance Imaging studies. Diabetes Care 36, 90–97. doi: 10.2337/dc12-0555

Espeland, M. A., Miller, M. E., Goveas, J. S., Hogan, P. E., Coker, L. H., Williamson, J. et al. (2011). Cognitive function and fine motor speed in older women with diabetes mellitus: results from the women's health initiative study of cognitive aging. J. Womens. Health (Larchmt) . 20, 1435–1443. doi: 10.1089/jwh.2011.2812

Ettorre, E., Cerra, E., Marigliano, B., Vigliotta, M., Vulcano, A., Fossati, C. et al. (2012). Role of cardiovascular risk factors (CRF) in the patients with mild cognitive impairment (MCI). Arch. Gerontol. Geriatr . 54, 330–332. doi: 10.1016/j.archger.2011.04.025

* 5,6,a Fakhri, M., Sikaroodi, H., Maleki, F., Ali Oghabian, M., and Ghanaati, H. (2012). Age-related frontal hyperactivation observed across different working memory tasks: an fMRI study. Behav. Neurol . 25, 351–361. doi: 10.3233/BEN-2012-120280

* 4,a,• Fera, F., Weickert, T. W., Goldberg, T. E., Tessitore, A., Hariri, A., Das, S. et al. (2005). Neural mechanisms underlying probabilistic category learning in normal aging. J. Neurosci . 25, 11340–11348. doi: 10.1523/JNEUROSCI.2736-05.2005

* 7,9,a Fernandes, M. A., Pacurar, A., Moscovitch, M., and Grady, C. (2006). Neural correlates of auditory recognition under full and divided attention in younger and older adults. Neuropsychologia 44, 2452–2464. doi: 10.1016/j.neuropsychologia.2006.04.020

* 2,11,a,■,▴ Filippini, N., Ebmeier, K. P., MacIntosh, B. J., Trachtenberg, A. J., Frisoni, G. B., Wilcock, G. K. et al. (2011). Differential effects of the APOE genotype on brain function across the lifespan. Neuroimage 54, 602–610. doi: 10.1016/j.neuroimage.2010.08.009

* 2,12,a,■,• Filippini, N., Nickerson, L. D., Beckmann, C. F., Ebmeier, K. P., Frisoni, G. B., Matthews, P. M. et al. (2012). Age-related adaptations of brain function during a memory task are also present at rest. Neuroimage 59, 3821–3828. doi: 10.1016/j.neuroimage.2011.11.063

Fischer, A. L., de Frias, C. M., Yeung, S. E., and Dixon, R. A. (2009). Short-term longitudinal trends in cognitive performance in older adults with type 2 diabetes. J. Clin. Exp. Neuropsychol . 31, 809–822. doi: 10.1080/13803390802537636

Fontbonne, A., Berr, C., Ducimetière, P., and Alpérovitch, A. (2001). Changes in cognitive abilities over a 4-year period are unfavorably affected in elderly diabetic subjects: results of the Epidemiology of Vascular Aging Study. Diabetes Care 24, 366–370. doi: 10.2337/diacare.24.2.366

* 3,5,a Gandini, D., Lemaire, P., Anton, J.-L., and Nazarian, B. (2008). Neural correlates of approximate quantification strategies in young and older adults: an fMRI study. Brain Res . 1246, 144–157. doi: 10.1016/j.brainres.2008.09.096

* 5,6,d,■ Garrett, D. D., Kovacevic, N., McIntosh, A. R., and Grady, C. L. (2011). The importance of being variable. J. Neurosci . 31, 4496–4503. doi: 10.1523/JNEUROSCI.5641-10.2011

* 3,5,a,□ Gazzaley, A., Cooney, J. W., Rissman, J., and D'Esposito, M. (2005). Top-down suppression deficit underlies working memory impairment in normal aging. Nat. Neurosci . 8, 1298–1300. doi: 10.1038/nn1543

Giacco, F., and Brownlee, M. (2010). Oxidative stress and diabetic complications. Circ. Res . 107, 1058–1070. doi: 10.1161/CIRCRESAHA.110.223545

Gifford, K. A., Badaracco, M., Liu, D., Tripodis, Y., Gentile, A., Lu, Z. et al. (2013). Blood pressure and cognition among older adults: a meta-analysis. Arch. Clin. Neuropsychol . 28, 649–664. doi: 10.1093/arclin/act046

* 7,8,a,■ Gigi, A., Babai, R., Penker, A., Hendler, T., and Korczyn, A. D. (2010). Prefrontal compensatory mechanism may enable normal semantic memory performance in mild cognitive impairment (MCI). J. Neuroimaging 20, 163–168. doi: 10.1111/j.1552-6569.2009.00386.x

Giordano, N., Tikhonoff, V., Palatini, P., Bascelli, A., Boschetti, G., De Lazzari, F. et al. (2012). Cognitive functions and cognitive reserve in relation to blood pressure components in a population-based cohort aged 53 to 94 years. Int. J. Hypertens . 2012:274851. doi: 10.1155/2012/274851

* 5,6,a,+ Giovanello, K. S., De Brigard, F., Hennessey Ford, J., Kaufer, D. I., Burke, J. R., Browndyke, J. N. et al. (2012). Event-related functional magnetic resonance imaging changes during relational retrieval in normal aging and amnestic mild cognitive impairment. J. Int. Neuropsychol. Soc . 18, 886–897. doi: 10.1017/S1355617712000689

* 5,6,a Giovanello, K. S., Kensinger, E. A., Wong, A. T., and Schacter, D. L. (2010). Age-related neural changes during memory conjunction errors. J. Cogn. Neurosci . 22, 1348–1361. doi: 10.1162/jocn.2009.21274

Gold, S. M., Dziobek, I., Sweat, V., Tirsi, A., Rogers, K., Bruehl, H. et al. (2007). Hippocampal damage and memory impairments as possible early brain complications of type 2 diabetes. Diabetologia 50, 711–719. doi: 10.1007/s00125-007-0602-7

* 5,6,b Gold, B. T., Jiang, Y., Jicha, G. A., and Smith, C. D. (2010a). Functional response in ventral temporal cortex differentiates mild cognitive impairment from normal aging. Hum. Brain Mapp . 31, 1249–1259. doi: 10.1002/hbm.20932

* 2,7,d,■ Gold, B. T., Powell, D. K., Xuan, L., Jicha, G. A., and Smith, C. D. (2010b). Age-related slowing of task switching is associated with decreased integrity of frontoparietal white matter. Neurobiol. Aging 31, 512–522. doi: 10.1016/j.neurobiolaging.2008.04.005

Goldstein, F. C., Levey, A. I., and Steenland, N. K. (2013). High blood pressure and cognitive decline in mild cognitive impairment. J. Am. Geriatr. Soc . 61, 67–73. doi: 10.1111/jgs.12067

Gorelick, P. B., Scuteri, A., Black, S. E., DeCarli, C., Greenberg, S. M., Iadecola, C. et al. (2011). Vascular contributions to cognitive impairment and dementia: a statement for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 42, 2672–2713. doi: 10.1161/STR.0b013e3182299496

* 4, a,• Gould, R. L., Brown, R. G., Owen, A. M., Ffytche, D. H., and Howard, R. J. (2003). fMRI BOLD response to increasing task difficulty during successful paired associates learning. Neuroimage 20, 1006–1019. doi: 10.1016/S1053-8119(03)00365-3

* 1,9,a Grady, C. L., Grigg, O., and Ng, C. (2012). Age differences in default and reward networks during processing of personally relevant information. Neuropsychologia 50, 1682–1697. doi: 10.1016/j.neuropsychologia.2012.03.024

* 5,6,d,■,• Grady, C. L., Protzner, A. B., Kovacevic, N., Strother, S. C., Afshin-Pour, B., Wojtowicz, M. et al. (2010). A multivariate analysis of age-related differences in default mode and task-positive networks across multiple cognitive domains. Cereb. Cortex 20, 1432–1447. doi: 10.1093/cercor/bhp207

* 7,9,a,■,• Grady, C. L., Springer, M. V., Hongwanishkul, D., McIntosh, A. R., and Winocur, G. (2006). Age-related changes in brain activity across the adult lifespan. J. Cogn. Neurosci . 18, 227–241. doi: 10.1162/089892906775783705

Granic, I., Dolga, A. M., Nijholt, I. M., van Dijk, G., and Eisel, U. L. M. (2009). Inflammation and NF-kappaB in Alzheimer's disease and diabetes. J. Alzheimers Dis . 16, 809–821. doi: 10.3233/JAD-2009-0976

Gregg, E. W., Yaffe, K., Cauley, J. A., Rolka, D. B., Blackwell, T. L., Narayan, K. M. et al. (2000). Is diabetes associated with cognitive impairment and cognitive decline among older women? Study of Osteoporotic Fractures Research Group. Arch. Intern. Med . 160, 174–180. doi: 10.1001/archinte.160.2.174

* 5,6,c Grön, G., Bittner, D., Schmitz, B., Wunderlich, A. P., Tomczak, R., and Riepe, M. W. (2003). Variability in memory performance in aged healthy individuals: an fMRI study. Neurobiol. Aging 24, 453–462. doi: 10.1016/S0197-4580(02)00128-8

* 6,7,a,□ Grossman, M., Cooke, A., DeVita, C., Alsop, D., Detre, J., Chen, W. et al. (2002). Age-related changes in working memory during sentence comprehension: an fMRI study. Neuroimage 15, 302–317. doi: 10.1006/nimg.2001.0971

Hajjar, I., Zhao, P., Alsop, D., and Novak, V. (2010). Hypertension and cerebral vasoreactivity: a continuous arterial spin labeling magnetic resonance imaging study. Hypertension 56, 859–864. doi: 10.1161/HYPERTENSIONAHA.110.160002

* 5,6,a Hampstead, B. M., Stringer, A. Y., Stilla, R. F., Amaraneni, A., and Sathian, K. (2011). Where did I put that? Patients with amnestic mild cognitive impairment demonstrate widespread reductions in activity during the encoding of ecologically relevant object-location associations. Neuropsychologia 49, 2349–2361. doi: 10.1016/j.neuropsychologia.2011.04.008

* 5,6,a Hampstead, B. M., Stringer, A. Y., Stilla, R. F., Giddens, M., and Sathian, K. (2012). Mnemonic strategy training partially restores hippocampal activity in patients with mild cognitive impairment. Hippocampus 22, 1652–1658. doi: 10.1002/hipo.22006

* 7,8,a Han, S. D., Arfanakis, K., Fleischman, D. A., Leurgans, S. E., Tuminello, E. R., Edmonds, E. C. et al. (2012a). Functional connectivity variations in mild cognitive impairment: associations with cognitive function. J. Int. Neuropsychol. Soc . 18, 39–48. doi: 10.1017/S1355617711001299

* 4, b Han, Y., Lui, S., Kuang, W., Lang, Q., Zou, L., and Jia, J. (2012b). Anatomical and functional deficits in patients with amnestic mild cognitive impairment. PLoS ONE 7:e28664. doi: 10.1371/journal.pone.0028664

Hannesdottir, K., Nitkunan, A., Charlton, R. A., Barrick, T. R., MacGregor, G. A., and Markus, H. S. (2009). Cognitive impairment and white matter damage in hypertension: a pilot study. Acta Neurol. Scand . 119, 261–268. doi: 10.1111/j.1600-0404.2008.01098.x

* 7,8,a Hartley, A. A., Jonides, J., and Sylvester, C.-Y. C. (2011). Dual-task processing in younger and older adults: similarities and differences revealed by fMRI. Brain Cogn . 75, 281–291. doi: 10.1016/j.bandc.2011.01.004

* 3,7,a,• Hedden, T., Van Dijk, K. R. A., Shire, E. H., Sperling, R. A., Johnson, K. A., and Buckner, R. L. (2012). Failure to modulate attentional control in advanced aging linked to white matter pathology. Cereb. Cortex 22, 1038–1051. doi: 10.1093/cercor/bhr172

Heni, M., Schöpfer, P., Peter, A., Sartorius, T., Fritsche, A., Synofzik, M. et al. (2013). Evidence for altered transport of insulin across the blood–brain barrier in insulin-resistant humans. Acta Diabetol . doi: 10.1007/s00592-013-0546-y. [Epub ahead of print].

* 5,9,b Holtzer, R., Rakitin, B. C., Steffener, J., Flynn, J., Kumar, A., and Stern, Y. (2009). Age effects on load-dependent brain activations in working memory for novel material. Brain Res . 1249, 148–161. doi: 10.1016/j.brainres.2008.10.009

Hoogenboom, W. S., Marder, T. J., Flores, V. L., Huisman, S., Eaton, H. P., Schneiderman, J. S. et al. (2014). Cerebral white matter integrity and resting-state functional connectivity in middle-aged patients with type 2 diabetes. Diabetes 63, 728–738. doi: 10.2337/db13-1219

* 5,6,a Hosseini, S. M. H., Rostami, M., Yomogida, Y., Takahashi, M., Tsukiura, T., and Kawashima, R. (2010). Aging and decision making under uncertainty: behavioral and neural evidence for the preservation of decision making in the absence of learning in old age. Neuroimage 52, 1514–1520. doi: 10.1016/j.neuroimage.2010.05.008

* 5,6,a Huang, C.-M., Polk, T. A., Goh, J. O., and Park, D. C. (2012). Both left and right posterior parietal activations contribute to compensatory processes in normal aging. Neuropsychologia 50, 55–66. doi: 10.1016/j.neuropsychologia.2011.10.022

* 5,6,a Hubert, V., Beaunieux, H., Chételat, G., Platel, H., Landeau, B., Viader, F. et al. (2009). Age-related changes in the cerebral substrates of cognitive procedural learning. Hum. Brain Mapp . 30, 1374–1386. doi: 10.1002/hbm.20605

* 4, c, • Iidaka, T., Sadato, N., Yamada, H., Murata, T., Omori, M., and Yonekura, Y. (2001). An fMRI study of the functional neuroanatomy of picture encoding in younger and older adults. Brain Res. Cogn. Brain Res . 11, 1–11. doi: 10.1016/S0926-6410(00)00058-6

* 1,2,b,♦ Jennings, J. R., van der Veen, F. M., and Meltzer, C. C. (2006). Verbal and spatial working memory in older individuals: a positron emission tomography study. Brain Res . 1092, 177–189. doi: 10.1016/j.brainres.2006.03.077

* 1,2,a Jimura, K., and Braver, T. S. (2010). Age-related shifts in brain activity dynamics during task switching. Cereb. Cortex 20, 1420–1431. doi: 10.1093/cercor/bhp206

* 5,6,a Jin, M., Pelak, V. S., and Cordes, D. (2012). Aberrant default mode network in subjects with amnestic mild cognitive impairment using resting-state functional MRI. Magn. Reson. Imaging 30, 48–61. doi: 10.1016/j.mri.2011.07.007

Johnson, E. L. (2012). Glycemic variability in type 2 diabetes mellitus: oxidative stress and macrovascular complications. Adv. Exp. Med. Biol . 771, 139–154.

* 5,6,a Johnson, M. K., Mitchell, K. J., Raye, C. L., and Greene, E. J. (2004). An age-related deficit in prefrontal cortical function associated with refreshing information. Psychol. Sci . 15, 127–132. doi: 10.1111/j.0963-7214.2004.01502009.x

Johnson, J. K., Pa, J., Boxer, A. L., Kramer, J. H., Freeman, K., and Yaffe, K. (2010). Baseline predictors of clinical progression among patients with dysexecutive mild cognitive impairment. Dement. Geriatr. Cogn. Disord . 30, 344–351. doi: 10.1159/000318836

* 9,11,a,■ Johnson, S. C., Schmitz, T. W., Asthana, S., Gluck, M. A., and Myers, C. (2008). Associative learning over trials activates the hippocampus in healthy elderly but not mild cognitive impairment. Neuropsychol. Dev. Cogn. B Aging Neuropsychol. Cogn . 15, 129–145. doi: 10.1080/13825580601139444

* 10,12,c Jones, D. T., Machulda, M. M., Vemuri, P., McDade, E. M., Zeng, G., Senjem, M. L. et al. (2011). Age-related changes in the default mode network are more advanced in Alzheimer disease. Neurology 77, 1524–1531. doi: 10.1212/WNL.0b013e318233b33d

Kaiser, N., Sasson, S., Feener, E. P., Boukobza-Vardi, N., Higashi, S., Moller, D. E. et al. (1993). Differential regulation of glucose transport and transporters by glucose in vascular endothelial and smooth muscle cells. Diabetes 42, 80–89. doi: 10.2337/diab.42.1.80

Kalani, M. (2008). The importance of endothelin-1 for microvascular dysfunction in diabetes. Vasc. Health Risk Manag . 4, 1061–1068. doi: 10.2147/VHRM.S3920

* 5,6,a Kalkstein, J., Checksfield, K., Bollinger, J., and Gazzaley, A. (2011). Diminished top-down control underlies a visual imagery deficit in normal aging. J. Neurosci . 31, 15768–15774. doi: 10.1523/JNEUROSCI.3209-11.2011

Kamiyama, K., Wada, A., Sugihara, M., Kurioka, S., Hayashi, K., Hayashi, T. et al. (2010). Potential hippocampal region atrophy in diabetes mellitus type 2: a voxel-based morphometry VSRAD study. Jpn. J. Radiol . 28, 266–272. doi: 10.1007/s11604-009-0416-2

Kanaya, A. M., Barrett-Connor, E., Gildengorin, G., and Yaffe, K. (2004). Change in cognitive function by glucose tolerance status in older adults: a 4-year prospective study of the Rancho Bernardo study cohort. Arch. Intern. Med . 164, 1327–1333. doi: 10.1001/archinte.164.12.1327

* 5,6,a,° Kannurpatti, S. S., Motes, M. A., Rypma, B., and Biswal, B. B. (2010). Neural and vascular variability and the fMRI-BOLD response in normal aging. Magn. Reson. Imaging 28, 466–476. doi: 10.1016/j.mri.2009.12.007

* 5,6,a,° Kannurpatti, S. S., Motes, M. A., Rypma, B., and Biswal, B. B. (2011). Increasing measurement accuracy of age-related BOLD signal change: minimizing vascular contributions by resting-state-fluctuation-of-amplitude scaling. Hum. Brain Mapp . 32, 1125–1140. doi: 10.1002/hbm.21097

* 5,6,a Kaufmann, L., Ischebeck, A., Weiss, E., Koppelstaetter, F., Siedentopf, C., Vogel, S. E. et al. (2008). An fMRI study of the numerical Stroop task in individuals with and without minimal cognitive impairment. Cortex 44, 1248–1255. doi: 10.1016/j.cortex.2007.11.009

Kearney, P. M., Whelton, M., Reynolds, K., Muntner, P., Whelton, P. K., and He, J. (2005). Global burden of hypertension: analysis of worldwide data. Lancet 365, 217–223. doi: 10.1016/S0140-6736(05)17741-1

Kennedy, K. M., and Raz, N. (2009). Pattern of normal age-related regional differences in white matter microstructure is modified by vascular risk. Brain Res . 1297, 41–56. doi: 10.1016/j.brainres.2009.08.058

* 7,8,a Kennedy, K. M., Rodrigue, K. M., Devous, M. D. Sr., Hebrank, A. C., Bischof, G. N., and Park, D. C. (2012). Effects of beta-amyloid accumulation on neural function during encoding across the adult lifespan. Neuroimage 62, 1–8. doi: 10.1016/j.neuroimage.2012.03.077

Kerti, L., Witte, A. V., Winkler, A., Grittner, U., Rujescu, D., and Floel, A. (2013). Higher glucose levels associated with lower memory and reduced hippocampal microstructure. Neurology 81, 1746–1752. doi: 10.1212/01.wnl.0000435561.00234.ee

* 9,11,a,■,× Kikuchi, M., Hirosawa, T., Yokokura, M., Yagi, S., Mori, N., Yoshikawa, E. et al. (2011). Effects of brain amyloid deposition and reduced glucose metabolism on the default mode of brain function in normal aging. J. Neurosci . 31, 11193–11199. doi: 10.1523/JNEUROSCI.2535-11.2011

Kim, E., Cho, M. H., Cha, K. R., Park, J. S., Ahn, C.-W., Oh, B. H. et al. (2008). Interactive effect of central obesity and hypertension on cognitive function in older out-patients with Type 2 diabetes. Diabet. Med . 25, 1440–1446. doi: 10.1111/j.1464-5491.2008.02612.x

* 6,7,a Kim, S.-Y., and Giovanello, K. S. (2011). The effects of attention on age-related relational memory deficits: fMRI evidence from a novel attentional manipulation. J. Cogn. Neurosci . 23, 3637–3656. doi: 10.1162/jocn_a_00058

* 4,b Kircher, T., Weis, S., Leube, D., Freymann, K., Erb, M., Jessen, F. et al. (2008). Anterior hippocampus orchestrates successful encoding and retrieval of non-relational memory: an event-related fMRI study. Eur. Arch. Psychiatry Clin. Neurosci . 258, 363–372. doi: 10.1007/s00406-008-0805-z

* 1,9,a Kirchhoff, B. A., Anderson, B. A., Barch, D. M., and Jacoby, L. L. (2012). Cognitive and neural effects of semantic encoding strategy training in older adults. Cereb. Cortex 22, 788–799. doi: 10.1093/cercor/bhr129

Kitagawa, K., Oku, N., Kimura, Y., Yagita, Y., Sakaguchi, M., Hatazawa, J. et al. (2009). Relationship between cerebral blood flow and later cognitive decline in hypertensive patients with cerebral small vessel disease. Hypertens. Res . 32, 816–820. doi: 10.1038/hr.2009.100

Kloppenborg, R. P., van den Berg, E., Kappelle, L. J., and Biessels, G. J. (2008). Diabetes and other vascular risk factors for dementia: which factor matters most? A systematic review. Eur. J. Pharmacol . 585, 97–108. doi: 10.1016/j.ejphar.2008.02.049

* 2,5,a Klostermann, E. C., Braskie, M. N., Landau, S. M., O'Neil, J. P., and Jagust, W. J. (2012). Dopamine and frontostriatal networks in cognitive aging. Neurobiol. Aging 33, 623.e15–24. doi: 10.1016/j.neurobiolaging.2011.03.002

Knecht, S., Wersching, H., Lohmann, H., Berger, K., and Ringelstein, E. B. (2009). How much does hypertension affect cognition? Explained variance in cross-sectional analysis of non-demented community-dwelling individuals in the SEARCH study. J. Neurol. Sci . 283, 149–152. doi: 10.1016/j.jns.2009.02.362

Knecht, S., Wersching, H., Lohmann, H., Bruchmann, M., Duning, T., Dziewas, R. et al. (2008). High-normal blood pressure is associated with poor cognitive performance. Hypertension 51, 663–668. doi: 10.1161/HYPERTENSIONAHA.107.105577

* 7,8,b,• Koch, W., Teipel, S., Mueller, S., Buerger, K., Bokde, A. L. W., Hampel, H. et al. (2010). Effects of aging on default mode network activity in resting state fMRI: does the method of analysis matter? Neuroimage 51, 280–287. doi: 10.1016/j.neuroimage.2009.12.008

* 7,8,a Krause, J. B., Taylor, J. G., Schmidt, D., Hautzel, H., Mottaghy, F. M., and Müller-Gärtner, H. W. (2000). Imaging and neural modelling in episodic and working memory processes. Neural Netw . 13, 847–859

* 1,9,a Kühn, S., Schmiedek, F., Schott, B., Ratcliff, R., Heinze, H.-J., Düzel, E. et al. (2011). Brain areas consistently linked to individual differences in perceptual decision-making in younger as well as older adults before and after training. J. Cogn. Neurosci . 23, 2147–2158. doi: 10.1162/jocn.2010.21564

* 7,8,a Kukolja, J., Thiel, C. M., Wilms, M., Mirzazade, S., and Fink, G. R. (2009). Ageing-related changes of neural activity associated with spatial contextual memory. Neurobiol. Aging 30, 630–645. doi: 10.1016/j.neurobiolaging.2007.08.015

* 6,7,a Kukolja, J., Thiel, C. M., Wolf, O. T., and Fink, G. R. (2008). Increased cortisol levels in cognitively challenging situations are beneficial in young but not older subjects. Psychopharmacology (Berl.) 201, 293–304. doi: 10.1007/s00213-008-1275-8

* 1,2,a,• Lamar, M., Yousem, D. M., and Resnick, S. M. (2004). Age differences in orbitofrontal activation: an fMRI investigation of delayed match and nonmatch to sample. Neuroimage 21, 1368–1376. doi: 10.1016/j.neuroimage.2003.11.018

* 7,8,a Langenecker, S. A., Briceno, E. M., Hamid, N. M., and Nielson, K. A. (2007). An evaluation of distinct volumetric and functional MRI contributions toward understanding age and task performance: a study in the basal ganglia. Brain Res . 1135, 58–68. doi: 10.1016/j.brainres.2006.11.068

* 5,6,a Langenecker, S. A., and Nielson, K. A. (2003). Frontal recruitment during response inhibition in older adults replicated with fMRI. Neuroimage 20, 1384–1392. doi: 10.1016/S1053-8119(03)00372-0

* 5,6,a Langenecker, S. A., Nielson, K. A., and Rao, S. M. (2004). fMRI of healthy older adults during Stroop interference. Neuroimage 21, 192–200. doi: 10.1016/j.neuroimage.2003.08.027

Last, D., Alsop, D. C., Abduljalil, A. M., Marquis, R. P., de Bazelaire, C., Hu, K. et al. (2007). Global and regional effects of type 2 diabetes on brain tissue volumes and cerebral vasoreactivity. Diabetes Care 30, 1193–1199. doi: 10.2337/dc06-2052

Launer, L. J., Ross, G. W., Petrovitch, H., Masaki, K., Foley, D., White, L. R. et al. (2000). Midlife blood pressure and dementia: the Honolulu-Asia aging study. Neurobiol. Aging 21, 49–55. doi: 10.1016/S0197-4580(00)00096-8

Lee, J. H., Yoon, S., Renshaw, P. F., Kim, T.-S., Jung, J. J., Choi, Y. et al. (2013). Morphometric changes in lateral ventricles of patients with recent-onset type 2 diabetes mellitus. PLoS ONE 8:e60515. doi: 10.1371/journal.pone.0060515

* 5,6,a Leshikar, E. D., Gutchess, A. H., Hebrank, A. C., Sutton, B. P., and Park, D. C. (2010). The impact of increased relational encoding demands on frontal and hippocampal function in older adults. Cortex 46, 507–521. doi: 10.1016/j.cortex.2009.07.011

* 5,6,a Li, C., Zheng, J., Wang, J., Gui, L., and Li, C. (2009a). An fMRI stroop task study of prefrontal cortical function in normal aging, mild cognitive impairment, and Alzheimer's disease. Curr. Alzheimer Res . 6, 525–530. doi: 10.2174/156720509790147142

Li, J., Wang, Y. J., Zhang, M., Xu, Z. Q., Gao, C. Y., Fang, C. Q. et al. (2011). Vascular risk factors promote conversion from mild cognitive impairment to Alzheimer disease. Neurology 76, 1485–1491. doi: 10.1212/WNL.0b013e318217e7a4

* 7,8,a Li, Z., Moore, A. B., Tyner, C., and Hu, X. (2009b). Asymmetric connectivity reduction and its relationship to “HAROLD” in aging brain. Brain Res . 1295, 149–158. doi: 10.1016/j.brainres.2009.08.004

Logothetis, N. K., Pauls, J., Augath, M., Trinath, T., and Oeltermann, A. (2001). Neurophysiological investigation of the basis of the fMRI signal. Nature 412, 150–157. doi: 10.1038/35084005

Luchsinger, J. A. (2008). Adiposity, hyperinsulinemia, diabetes and Alzheimer's disease: an epidemiological perspective. Eur. J. Pharmacol . 585, 119–129. doi: 10.1016/j.ejphar.2008.02.048

Luchsinger, J. A., Palmas, W., Teresi, J. A., Silver, S., Kong, J., Eimicke, J. P. et al. (2011). Improved diabetes control in the elderly delays global cognitive decline. J. Nutr. Health Aging 15, 445–449. doi: 10.1007/s12603-011-0057-x

* 3,5,a MacDonald, S. W. S., Nyberg, L., Sandblom, J., Fischer, H., and Bäckman, L. (2008). Increased response-time variability is associated with reduced inferior parietal activation during episodic recognition in aging. J. Cogn. Neurosci . 20, 779–786. doi: 10.1162/jocn.2008.20502

* 2,5,a,■ Madden, D. J., Costello, M. C., Dennis, N. A., Davis, S. W., Shepler, A. M., Spaniol, J. et al. (2010). Adult age differences in functional connectivity during executive control. Neuroimage 52, 643–657. doi: 10.1016/j.neuroimage.2010.04.249

* 1,2,d,■ Madden, D. J., Langley, L. K., Denny, L. L., Turkington, T. G., Provenzale, J. M., Hawk, T. C. et al. (2002a). Adult age differences in visual word identification: functional neuroanatomy by positron emission tomography. Brain Cogn . 49, 297–321. doi: 10.1006/brcg.2001.1502

* 2,5,d,■ Madden, D. J., Spaniol, J., Whiting, W. L., Bucur, B., Provenzale, J. M., Cabeza, R. et al. (2007). Adult age differences in the functional neuroanatomy of visual attention: a combined fMRI and DTI study. Neurobiol. Aging 28, 459–476. doi: 10.1016/j.neurobiolaging.2006.01.005

* 1,2,d,■ Madden, D. J., Turkington, T. G., Provenzale, J. M., Denny, L. L., Langley, L. K., Hawk, T. C. et al. (2002b). Aging and attentional guidance during visual search: functional neuroanatomy by positron emission tomography. Psychol. Aging 17, 24–43. doi: 10.1037/0882-7974.17.1.24

Maggi, S., Limongi, F., Noale, M., Romanato, G., Tonin, P., Rozzini, R. et al. (2009). Diabetes as a risk factor for cognitive decline in older patients. Dement. Geriatr. Cogn. Disord . 27, 24–33. doi: 10.1159/000183842

* 1,9,a Maillet, D., and Rajah, M. N. (2011). Age-related changes in the three-way correlation between anterior hippocampus volume, whole-brain patterns of encoding activity and subsequent context retrieval. Brain Res . 1420, 68–79. doi: 10.1016/j.brainres.2011.08.071

Marioni, R. E., Strachan, M. W. J., Reynolds, R. M., Lowe, G. D. O., Mitchell, R. J., Fowkes, F. G. R. et al. (2010). Association between raised inflammatory markers and cognitive decline in elderly people with type 2 diabetes: the Edinburgh Type 2 Diabetes Study. Diabetes 59, 710–713. doi: 10.2337/db09-1163

* 3,5,b Mathis, A., Schunck, T., Erb, G., Namer, I. J., and Luthringer, R. (2009). The effect of aging on the inhibitory function in middle-aged subjects: a functional MRI study coupled with a color-matched Stroop task. Int. J. Geriatr. Psychiatry 24, 1062–1071. doi: 10.1002/gps.2222

* 4,a Mattay, V. S., Fera, F., Tessitore, A., Hariri, A. R., Berman, K. F., Das, S. et al. (2006). Neurophysiological correlates of age-related changes in working memory capacity. Neurosci. Lett . 392, 32–37. doi: 10.1016/j.neulet.2005.09.025

* 1,2,a,• Matthäus, F., Schmidt, J.-P., Banerjee, A., Schulze, T. G., Demirakca, T., and Diener, C. (2012). Effects of age on the structure of functional connectivity networks during episodic and working memory demand. Brain Connect . 2, 113–124. doi: 10.1089/brain.2012.0077

* 7,8,a,° Mayhew, S. D., Li, S., Storrar, J. K., Tsvetanov, K. A., and Kourtzi, Z. (2010). Learning shapes the representation of visual categories in the aging human brain. J. Cogn. Neurosci . 22, 2899–2912. doi: 10.1162/jocn.2010.21415

* 7,8,a,□ McGeown, W. J., Shanks, M. F., Forbes-McKay, K. E., and Venneri, A. (2009). Patterns of brain activity during a semantic task differentiate normal aging from early Alzheimer's disease. Psychiatry Res . 173, 218–227. doi: 10.1016/j.pscychresns.2008.10.005

* 5,6,a Meier, T. B., Desphande, A. S., Vergun, S., Nair, V. A., Song, J., Biswal, B. B. et al. (2012). Support vector machine classification and characterization of age-related reorganization of functional brain networks. Neuroimage 60, 601–613. doi: 10.1016/j.neuroimage.2011.12.052

* 7,9,c Meinzer, M., Flaisch, T., Seeds, L., Harnish, S., Antonenko, D., Witte, V. et al. (2012a). Same modulation but different starting points: performance modulates age differences in inferior frontal cortex activity during word-retrieval. PLoS ONE 7:e33631. doi: 10.1371/journal.pone.0033631

* 5,6,d Meinzer, M., Flaisch, T., Wilser, L., Eulitz, C., Rockstroh, B., Conway, T. et al. (2009). Neural signatures of semantic and phonemic fluency in young and old adults. J. Cogn. Neurosci . 21, 2007–2018. doi: 10.1162/jocn.2009.21219

* 7,9,c Meinzer, M., Seeds, L., Flaisch, T., Harnish, S., Cohen, M. L., McGregor, K. et al. (2012b). Impact of changed positive and negative task-related brain activity on word-retrieval in aging. Neurobiol. Aging 33, 656–669. doi: 10.1016/j.neurobiolaging.2010.06.020

Messier, C., Awad-Shimoon, N., Gagnon, M., Desrochers, A., and Tsiakas, M. (2011). Glucose regulation is associated with cognitive performance in young nondiabetic adults. Behav. Brain Res . 222, 81–88. doi: 10.1016/j.bbr.2011.03.023

Messier, C., Tsiakas, M., Gagnon, M., and Desrochers, A. (2010). Effect of age and glucoregulation on cognitive performance. J. Clin. Exp. Neuropsychol . 32, 809–821. doi: 10.1080/13803390903540323

* 7,9,a Meulenbroek, O., Kessels, R. P. C., de Rover, M., Petersson, K. M., Rikkert, M. G. M. O., Rijpkema, M. et al. (2010a). Age-effects on associative object-location memory. Brain Res . 1315, 100–110. doi: 10.1016/j.brainres.2009.12.011

* 1,2,a Meulenbroek, O., Petersson, K. M., Voermans, N., Weber, B., and Fernández, G. (2004). Age differences in neural correlates of route encoding and route recognition. Neuroimage 22, 1503–1514. doi: 10.1016/j.neuroimage.2004.04.007

* 7,8,a Meulenbroek, O., Rijpkema, M., Kessels, R. P. C., Rikkert, M. G. M. O., and Fernández, G. (2010b). Autobiographical memory retrieval in patients with Alzheimer's disease. Neuroimage 53, 331–340. doi: 10.1016/j.neuroimage.2010.05.082

* 7,8,b,■ Miettinen, P. S., Pihlajamäki, M., Jauhiainen, A. M., Niskanen, E., Hänninen, T., Vanninen, R. et al. (2011). Structure and function of medial temporal and posteromedial cortices in early Alzheimer's disease. Eur. J. Neurosci . 34, 320–330. doi: 10.1111/j.1460-9568.2011.07745.x

* 3,7,a,• Milham, M. P., Erickson, K. I., Banich, M. T., Kramer, A. F., Webb, A., Wszalek, T. et al. (2002). Attentional control in the aging brain: insights from an fMRI study of the stroop task. Brain Cogn . 49, 277–296. doi: 10.1006/brcg.2001.1501

* 5,6,a Mitchell, K. J., Johnson, M. K., Raye, C. L., and D'Esposito, M. (2000). fMRI evidence of age-related hippocampal dysfunction in feature binding in working memory. Brain Res. Cogn. Brain Res . 10, 197–206. doi: 10.1016/S0926-6410(00)00029-X

* 7,8,a Mitchell, K. J., Johnson, M. R., Higgins, J. A., and Johnson, M. K. (2010). Age differences in brain activity during perceptual versus reflective attention. Neuroreport 21, 293–297. doi: 10.1097/WNR.0b013e32833730d6

* 7,8,a Mitchell, K. J., Raye, C. L., Johnson, M. K., and Greene, E. J. (2006). An fMRI investigation of short-term source memory in young and older adults. Neuroimage 30, 627–633. doi: 10.1016/j.neuroimage.2005.09.039

* 1,2,a,▴ Mohtasib, R. S., Lumley, G., Goodwin, J. A., Emsley, H. C. A., Sluming, V., and Parkes, L. M. (2012). Calibrated fMRI during a cognitive Stroop task reveals reduced metabolic response with increasing age. Neuroimage 59, 1143–1151. doi: 10.1016/j.neuroimage.2011.07.092

Moran, C., Phan, T. G., Chen, J., Blizzard, L., Beare, R., Venn, A. et al. (2013). Brain atrophy in type 2 diabetes: regional distribution and influence on cognition. Diabetes Care 36, 4036–4042. doi: 10.2337/dc13-0143

* 3,5,a Morcom, A. M., and Friston, K. J. (2012). Decoding episodic memory in ageing: a Bayesian analysis of activity patterns predicting memory. Neuroimage 59, 1772–1782. doi: 10.1016/j.neuroimage.2011.08.071

* 3,5,a,□ Morcom, A. M., Good, C. D., Frackowiak, R. S. J., and Rugg, M. D. (2003). Age effects on the neural correlates of successful memory encoding. Brain 126, 213–229. doi: 10.1093/brain/awg020

* 5,6,a Mormino, E. C., Brandel, M. G., Madison, C. M., Marks, S., Baker, S. L., and Jagust, W. J. (2012). Aβ Deposition in aging is associated with increases in brain activation during successful memory encoding. Cereb. Cortex 22, 1813–1823. doi: 10.1093/cercor/bhr255

* 5,6,a Mormino, E. C., Smiljic, A., Hayenga, A. O., Onami, S. H., Greicius, M. D., Rabinovici, G. D. et al. (2011). Relationships between β-amyloid and functional connectivity in different components of the default mode network in aging. Cereb. Cortex 21, 2399–2407. doi: 10.1093/cercor/bhr025

Morris, J. K., Vidoni, E. D., Honea, R. A., and Burns, J. M. (2014). Impaired glycemia increases disease progression in mild cognitive impairment. Neurobiol. Aging 35, 585–589. doi: 10.1016/j.neurobiolaging.2013.09.033

* 5,6,a Mowinckel, A. M., Espeseth, T., and Westlye, L. T. (2012). Network-specific effects of age and in-scanner subject motion: a resting-state fMRI study of 238 healthy adults. Neuroimage 63, 1364–1373. doi: 10.1016/j.neuroimage.2012.08.004

* 5,6,a Murphy, K., and Garavan, H. (2004). Artifactual fMRI group and condition differences driven by performance confounds. Neuroimage 21, 219–228. doi: 10.1016/j.neuroimage.2003.09.016

* 3,7,b Murty, V. P., Sambataro, F., Das, S., Tan, H.-Y., Callicott, J. H., Goldberg, T. E. et al. (2009). Age-related alterations in simple declarative memory and the effect of negative stimulus valence. J. Cogn. Neurosci . 21, 1920–1933. doi: 10.1162/jocn.2009.21130

Musen, G., Jacobson, A. M., Bolo, N. R., Simonson, D. C., Shenton, M. E., McCartney, R. L. et al. (2012). Resting-state brain functional connectivity is altered in type 2 diabetes. Diabetes 61, 2375–2379. doi: 10.2337/db11-1669

Nagai, M., Hoshide, S., Ishikawa, J., Shimada, K., and Kario, K. (2008). Ambulatory blood pressure as an independent determinant of brain atrophy and cognitive function in elderly hypertension. J. Hypertens . 26, 1636–1641. doi: 10.1097/HJH.0b013e3283018333

* 7,8,a Nagel, I. E., Preuschhof, C., Li, S.-C., Nyberg, L., Bäckman, L., Lindenberger, U. et al. (2009). Performance level modulates adult age differences in brain activation during spatial working memory. Proc. Natl. Acad. Sci. U.S.A . 106, 22552–22557. doi: 10.1073/pnas.0908238106

* 10,12,a Nagel, I. E., Preuschhof, C., Li, S.-C., Nyberg, L., Bäckman, L., Lindenberger, U. et al. (2011). Load modulation of BOLD response and connectivity predicts working memory performance in younger and older adults. J. Cogn. Neurosci . 23, 2030–2045. doi: 10.1162/jocn.2010.21560

Nakae, J., Kido, Y., and Accili, D. (2001). Distinct and overlapping functions of insulin and IGF-I receptors. Endocr. Rev . 22, 818–835. doi: 10.1210/edrv.22.6.0452

Newcomer, J. W., and Haupt, D. W. (2006). The metabolic effects of antipsychotic medications. Can. J. Psychiatry 51, 480–491.

* 4, b, ■ Nichols, L. M., Masdeu, J. C., Mattay, V. S., Kohn, P., Emery, M., Sambataro, F. et al. (2012). Interactive effect of apolipoprotein e genotype and age on hippocampal activation during memory processing in healthy adults. Arch. Gen. Psychiatry 69, 804–813. doi: 10.1001/archgenpsychiatry.2011.1893

* 7,8,a,• Nielson, K. A., Douville, K. L., Seidenberg, M., Woodard, J. L., Miller, S. K., Franczak, M. et al. (2006). Age-related functional recruitment for famous name recognition: an event-related fMRI study. Neurobiol. Aging 27, 1494–1504. doi: 10.1016/j.neurobiolaging.2005.08.022

* 5,6,a Nielson, K. A., Langenecker, S. A., and Garavan, H. (2002). Differences in the functional neuroanatomy of inhibitory control across the adult life span. Psychol. Aging 17, 56–71. doi: 10.1037/0882-7974.17.1.56

* 3,5,a,□ Nielson, K. A., Langenecker, S. A., Ross, T. J., Garavan, H., Rao, S. M., and Stein, E. A. (2004). Comparability of functional MRI response in young and old during inhibition. Neuroreport 15, 129–133. doi: 10.1097/00001756-200401190-00025

Nishikawa, T., Edelstein, D., Du, X. L., Yamagishi, S., Matsumura, T., Kaneda, Y. et al. (2000). Normalizing mitochondrial superoxide production blocks three pathways of hyperglycaemic damage. Nature 404, 787–790. doi: 10.1038/35008121

Nooyens, A. C. J., Baan, C. A., Spijkerman, A. M. W., and Verschuren, W. M. M. (2010). Type 2 diabetes and cognitive decline in middle-aged men and women: the Doetinchem Cohort Study. Diabetes Care 33, 1964–1969. doi: 10.2337/dc09-2038

* 7,10,c,❖ Nordahl, C. W., Ranganath, C., Yonelinas, A. P., Decarli, C., Fletcher, E., and Jagust, W. J. (2006). White matter changes compromise prefrontal cortex function in healthy elderly individuals. J. Cogn. Neurosci . 18, 418–429. doi: 10.1162/089892906775990552

Novak, V., Last, D., Alsop, D. C., Abduljalil, A. M., Hu, K., Lepicovsky, L. et al. (2006). Cerebral blood flow velocity and periventricular white matter hyperintensities in type 2 diabetes. Diabetes Care 29, 1529–1534. doi: 10.2337/dc06-0261

* 5,6,a Nyberg, L., Dahlin, E., Stigsdotter Neely, A., and Bäckman, L. (2009). Neural correlates of variable working memory load across adult age and skill: dissociative patterns within the fronto-parietal network. Scand. J. Psychol . 50, 41–46. doi: 10.1111/j.1467-9450.2008.00678.x

* 7,8,a,• O'Brien, J. L., O'Keefe, K. M., LaViolette, P. S., DeLuca, A. N., Blacker, D., Dickerson, B. C. et al. (2010). Longitudinal fMRI in elderly reveals loss of hippocampal activation with clinical decline. Neurology 74, 1969–1976. doi: 10.1212/WNL.0b013e3181e3966e

* 7,8,a Osaka, M., Otsuka, Y., and Osaka, N. (2012a). Verbal to visual code switching improves working memory in older adults: an fMRI study. Front. Hum. Neurosci . 6:24. doi: 10.3389/fnhum.2012.00024

* 7,8,a Osaka, M., Yaoi, K., Otsuka, Y., Katsuhara, M., and Osaka, N. (2012b). Practice on conflict tasks promotes executive function of working memory in the elderly. Behav. Brain Res . 233, 90–98. doi: 10.1016/j.bbr.2012.04.044

* 7,8,a Otsuka, Y., Osaka, N., Morishita, M., Kondo, H., and Osaka, M. (2006). Decreased activation of anterior cingulate cortex in the working memory of the elderly. Neuroreport 17, 1479–1482. doi: 10.1097/01.wnr.0000236852.63092.9f

Oveisgharan, S., and Hachinski, V. (2010). Hypertension, executive dysfunction, and progression to dementia: the canadian study of health and aging. Arch. Neurol . 67, 187–192. doi: 10.1001/archneurol.2009.312

* 6,7,a Pacheco, J., Beevers, C. G., McGeary, J. E., and Schnyer, D. M. (2012). Memory monitoring performance and PFC activity are associated with 5-HTTLPR genotype in older adults. Neuropsychologia 50, 2257–2270. doi: 10.1016/j.neuropsychologia.2012.05.030

* 7,8,a Park, D. C., Welsh, R. C., Marshuetz, C., Gutchess, A. H., Mikels, J., Polk, T. A. et al. (2003). Working memory for complex scenes: age differences in frontal and hippocampal activations. J. Cogn. Neurosci . 15, 1122–1134. doi: 10.1162/089892903322598094

* 4, a, □ Park, J., Carp, J., Hebrank, A., Park, D. C., and Polk, T. A. (2010). Neural specificity predicts fluid processing ability in older adults. J. Neurosci . 30, 9253–9259. doi: 10.1523/JNEUROSCI.0853-10.2010

* 1,2,a Paxton, J. L., Barch, D. M., Racine, C. A., and Braver, T. S. (2008). Cognitive control, goal maintenance, and prefrontal function in healthy aging. Cereb. Cortex 18, 1010–1028. doi: 10.1093/cercor/bhm135

* 1,9,b Persson, J., Kalpouzos, G., Nilsson, L.-G., Ryberg, M., and Nyberg, L. (2011). Preserved hippocampus activation in normal aging as revealed by fMRI. Hippocampus 21, 753–766. doi: 10.1002/hipo.20794

* 7,10,a,+ Persson, J., Nyberg, L., Lind, J., Larsson, A., Nilsson, L.-G., Ingvar, M. et al. (2006). Structure-function correlates of cognitive decline in aging. Cereb. Cortex 16, 907–915. doi: 10.1093/cercor/bhj036

* 7,10,c,♦ Persson, J., Pudas, S., Lind, J., Kauppi, K., Nilsson, L.-G., and Nyberg, L. (2012). Longitudinal structure-function correlates in elderly reveal MTL dysfunction with cognitive decline. Cereb. Cortex 22, 2297–2304. doi: 10.1093/cercor/bhr306

* 2,5,a Persson, J., Sylvester, C.-Y. C., Nelson, J. K., Welsh, K. M., Jonides, J., and Reuter-Lorenz, P. A. (2004). Selection requirements during verb generation: differential recruitment in older and younger adults. Neuroimage 23, 1382–1390. doi: 10.1016/j.neuroimage.2004.08.004

* 7,8,a Petrella, J. R., Townsend, B. A., Jha, A. P., Ziajko, L. A., Slavin, M. J., Lustig, C. et al. (2005). Increasing memory load modulates regional brain activity in older adults as measured by fMRI. J. Neuropsychiatry Clin. Neurosci . 17, 75–83. doi: 10.1176/appi.neuropsych.17.1.75

* 5,6,a Piefke, M., Onur, Ö. A., and Fink, G. R. (2012). Aging-related changes of neural mechanisms underlying visual-spatial working memory. Neurobiol. Aging 33, 1284–1297. doi: 10.1016/j.neurobiolaging.2010.10.014

* 7,8,a Pihlajamäki, M., and Sperling, R. A. (2009). Functional MRI assessment of task-induced deactivation of the default mode network in Alzheimer's disease and at-risk older individuals. Behav. Neurol . 21, 77–91. doi: 10.3233/BEN-2009-0231

Pires, P. W., Dams Ramos, C. M., Matin, N., and Dorrance, A. M. (2013). The effects of hypertension on the cerebral circulation. Am. J. Physiol. Heart Circ. Physiol . 304, H1598–H1614. doi: 10.1152/ajpheart.00490.2012

* 4, c Podell, J. E., Sambataro, F., Murty, V. P., Emery, M. R., Tong, Y., Das, S. et al. (2012). Neurophysiological correlates of age-related changes in working memory updating. Neuroimage 62, 2151–2160. doi: 10.1016/j.neuroimage.2012.05.066

* 7,8,a Prakash, R. S., Erickson, K. I., Colcombe, S. J., Kim, J. S., Voss, M. W., and Kramer, A. F. (2009). Age-related differences in the involvement of the prefrontal cortex in attentional control. Brain Cogn . 71, 328–335. doi: 10.1016/j.bandc.2009.07.005

* 7,8,a Protzner, A. B., Mandzia, J. L., Black, S. E., and McAndrews, M. P. (2011). Network interactions explain effective encoding in the context of medial temporal damage in MCI. Hum. Brain Mapp . 32, 1277–1289. doi: 10.1002/hbm.21107

* 7,8,a,• Putcha, D., O'Keefe, K., LaViolette, P., O'Brien, J., Greve, D., Rentz, D. M. et al. (2011). Reliability of functional magnetic resonance imaging associative encoding memory paradigms in non-demented elderly adults. Hum. Brain Mapp . 32, 2027–2044. doi: 10.1002/hbm.21166

* 1,9,d Rajah, M. N., Languay, R., and Grady, C. L. (2011). Age-related changes in right middle frontal gyrus volume correlate with altered episodic retrieval activity. J. Neurosci . 31, 17941–17954. doi: 10.1523/JNEUROSCI.1690-11.2011

* 7,8,a,• Rajah, M. N., Languay, R., and Valiquette, L. (2010). Age-related changes in prefrontal cortex activity are associated with behavioural deficits in both temporal and spatial context memory retrieval in older adults. Cortex 46, 535–549. doi: 10.1016/j.cortex.2009.07.006

* 5,6,a,□ Rajah, M. N., and McIntosh, A. R. (2008). Age-related differences in brain activity during verbal recency memory. Brain Res . 1199, 111–125. doi: 10.1016/j.brainres.2007.12.051

Raji, C. A., Lopez, O. L., Kuller, L. H., Carmichael, O. T., Longstreth, W. T. Jr., Gach, H. M. et al. (2012). White matter lesions and brain gray matter volume in cognitively normal elders. Neurobiol. Aging 33, 834.e7–16. doi: 10.1016/j.neurobiolaging.2011.08.010

* 2,7,d,▴ Ramsøy, T. Z., Liptrot, M. G., Skimminge, A., Lund, T. E., Sidaros, K., Christensen, M. S. et al. (2012). Healthy aging attenuates task-related specialization in the human medial temporal lobe. Neurobiol. Aging 33, 1874–1889. doi: 10.1016/j.neurobiolaging.2011.09.032

* 6,7,a Rand-Giovannetti, E., Chua, E. F., Driscoll, A. E., Schacter, D. L., Albert, M. S., and Sperling, R. A. (2006). Hippocampal and neocortical activation during repetitive encoding in older persons. Neurobiol. Aging 27, 173–182. doi: 10.1016/j.neurobiolaging.2004.12.013

Ravona-Springer, R., Moshier, E., Schmeidler, J., Godbold, J., Akrivos, J., Rapp, M. et al. (2012). Changes in glycemic control are associated with changes in cognition in non-diabetic elderly. J. Alzheimers Dis . 30, 299–309. doi: 10.3233/JAD-2012-120106

* 5,6,a Raye, C. L., Mitchell, K. J., Reeder, J. A., Greene, E. J., and Johnson, M. K. (2008). Refreshing one of several active representations: behavioral and functional magnetic resonance imaging differences between young and older adults. J. Cogn. Neurosci . 20, 852–862. doi: 10.1162/jocn.2008.20508

Raz, N., Yang, Y., Dahle, C. L., and Land, S. (2012). Volume of white matter hyperintensities in healthy adults: contribution of age, vascular risk factors, and inflammation-related genetic variants. Biochim. Biophys. Acta 1822, 361–369. doi: 10.1016/j.bbadis.2011.08.007

Reaven, G. M., Thompson, L. W., Nahum, D., and Haskins, E. (1990). Relationship between hyperglycemia and cognitive function in older NIDDM patients. Diabetes Care 13, 16–21. doi: 10.2337/diacare.13.1.16

Regenold, W. T., Thapar, R. K., Marano, C., Gavirneni, S., and Kondapavuluru, P. V. (2002). Increased prevalence of type 2 diabetes mellitus among psychiatric inpatients with bipolar I affective and schizoaffective disorders independent of psychotropic drug use. J. Affect. Disord . 70, 19–26. doi: 10.1016/S0165-0327(01)00456-6

Reijmer, Y. D., van den Berg, E., de Bresser, J., Kessels, R. P. C., Kappelle, L. J., Algra, A. et al. (2011). Accelerated cognitive decline in patients with type 2 diabetes: MRI correlates and risk factors. Diabetes Metab. Res. Rev . 27, 195–202. doi: 10.1002/dmrr.1163

Reitz, C., Tang, M.-X., Manly, J., Mayeux, R., and Luchsinger, J. A. (2007). Hypertension and the risk of mild cognitive impairment. Arch. Neurol . 64, 1734–1740. doi: 10.1001/archneur.64.12.1734

* 5,6,a,▴ Restom, K., Bangen, K. J., Bondi, M. W., Perthen, J. E., and Liu, T. T. (2007). Cerebral blood flow and BOLD responses to a memory encoding task: a comparison between healthy young and elderly adults. Neuroimage 37, 430–439. doi: 10.1016/j.neuroimage.2007.05.024

* 7,8,a Rieckmann, A., Fischer, H., and Bäckman, L. (2010). Activation in striatum and medial temporal lobe during sequence learning in younger and older adults: relations to performance. Neuroimage 50, 1303–1312. doi: 10.1016/j.neuroimage.2010.01.015

* 7,8,a Rieckmann, A., Karlsson, S., Fischer, H., and Bäckman, L. (2011). Caudate dopamine D1 receptor density is associated with individual differences in frontoparietal connectivity during working memory. J. Neurosci . 31, 14284–14290. doi: 10.1523/JNEUROSCI.3114-11.2011

Rizzo, M. R., Marfella, R., Barbieri, M., Boccardi, V., Vestini, F., Lettieri, B. et al. (2010). Relationships between daily acute glucose fluctuations and cognitive performance among aged type 2 diabetic patients. Diabetes Care 33, 2169–2174. doi: 10.2337/dc10-0389

Roberts, R. O., Knopman, D. S., Geda, Y. E., Cha, R. H., Pankratz, V. S., Baertlein, L. et al. (2014). Association of diabetes with amnestic and nonamnestic mild cognitive impairment. Alzheimers. Dement . 10, 18–26. doi: 10.1016/j.jalz.2013.01.001

* 5,6,a Rombouts, S. A. R. B., Barkhof, F., Goekoop, R., Stam, C. J., and Scheltens, P. (2005a). Altered resting state networks in mild cognitive impairment and mild Alzheimer's disease: an fMRI study. Hum. Brain Mapp . 26, 231–239. doi: 10.1002/hbm.20160

* 5,6,a,• Rombouts, S. A. R. B., Goekoop, R., Stam, C. J., Barkhof, F., and Scheltens, P. (2005b). Delayed rather than decreased BOLD response as a marker for early Alzheimer's disease. Neuroimage 26, 1078–1085. doi: 10.1016/j.neuroimage.2005.03.022

* 7,10,b,❖ Rosano, C., Venkatraman, V. K., Guralnik, J., Newman, A. B., Glynn, N. W., Launer, L. et al. (2010). Psychomotor speed and functional brain MRI 2 years after completing a physical activity treatment. J. Gerontol. A Biol. Sci. Med. Sci . 65, 639–647. doi: 10.1093/gerona/glq038

* 3,5,a Rosen, A. C., Gabrieli, J. D. E., Stoub, T., Prull, M. W., O'Hara, R., Yesavage, J. et al. (2005). Relating medial temporal lobe volume to frontal fMRI activation for memory encoding in older adults. Cortex 41, 595–602. doi: 10.1016/S0010-9452(08)70199-0

Ryan, C. M., and Geckle, M. O. (2000). Circumscribed cognitive dysfunction in middle-aged adults with type 2 diabetes. Diabetes Care 23, 1486–1493. doi: 10.2337/diacare.23.10.1486

* 4, b Rypma, B., Berger, J. S., Genova, H. M., Rebbechi, D., and D'Esposito, M. (2005). Dissociating age-related changes in cognitive strategy and neural efficiency using event-related fMRI. Cortex 41, 582–594. doi: 10.1016/S0010-9452(08)70198-9

* 4, b Rypma, B., Eldreth, D. A., and Rebbechi, D. (2007). Age-related differences in activation-performance relations in delayed-response tasks: a multiple component analysis. Cortex 43, 65–76. doi: 10.1016/S0010-9452(08)70446-5

* 5,6,a Sala-Llonch, R., Arenaza-Urquijo, E. M., Valls-Pedret, C., Vidal-Piñeiro, D., Bargall,ó, N., Junqu,é, C. et al. (2012). Dynamic functional reorganizations and relationship with working memory performance in healthy aging. Front. Hum. Neurosci . 6, 152. doi: 10.3389/fnhum.2012.00152

* 5,6,b Salami, A., Eriksson, J., and Nyberg, L. (2012). Opposing effects of aging on large-scale brain systems for memory encoding and cognitive control. J. Neurosci . 32, 10749–10757. doi: 10.1523/JNEUROSCI.0278-12.2012

* 4, a, • Sambataro, F., Murty, V. P., Callicott, J. H., Tan, H.-Y., Das, S., Weinberger, D. R. et al. (2010). Age-related alterations in default mode network: impact on working memory performance. Neurobiol. Aging 31, 839–852. doi: 10.1016/j.neurobiolaging.2008.05.022

* 3,5,a Schneider-Garces, N. J., Gordon, B. A., Brumback-Peltz, C. R., Shin, E., Lee, Y., Sutton, B. P. et al. (2010). Span, CRUNCH, and beyond: working memory capacity and the aging brain. J. Cogn. Neurosci . 22, 655–669. doi: 10.1162/jocn.2009.21230

* 7,9,c Schulte, T., Müller-Oehring, E. M., Chanraud, S., Rosenbloom, M. J., Pfefferbaum, A., and Sullivan, E. V. (2011). Age-related reorganization of functional networks for successful conflict resolution: a combined functional and structural MRI study. Neurobiol. Aging 32, 2075–2090. doi: 10.1016/j.neurobiolaging.2009.12.002

Scuteri, A., Tesauro, M., Appolloni, S., Preziosi, F., Brancati, A. M., and Volpe, M. (2007). Arterial stiffness as an independent predictor of longitudinal changes in cognitive function in the older individual. J. Hypertens . 25, 1035–1040. doi: 10.1097/01.hjh.0000170384.38708.b7

* 7,8,a Shafto, M. A., Stamatakis, E. A., Tam, P. P., and Tyler, L. K. (2010). Word retrieval failures in old age: the relationship between structure and function. J. Cogn. Neurosci . 22, 1530–1540. doi: 10.1162/jocn.2009.21321

Shaikh, S., and Nicholson, L. F. B. (2008). Advanced glycation end products induce in vitro cross-linking of alpha-synuclein and accelerate the process of intracellular inclusion body formation. J. Neurosci. Res . 86, 2071–2082. doi: 10.1002/jnr.21644

Shaw, J. E., Sicree, R. A., and Zimmet, P. Z. (2010). Global estimates of the prevalence of diabetes for 2010 and 2030. Diabetes Res. Clin. Pract . 87, 4–14. doi: 10.1016/j.diabres.2009.10.007

* 5,6,b,• Siedlecki, K. L., Habeck, C. G., Brickman, A. M., Gazes, Y., and Stern, Y. (2009). Examining the multifactorial nature of cognitive aging with covariance analysis of positron emission tomography data. J. Int. Neuropsychol. Soc . 15, 973–981. doi: 10.1017/S1355617709990592

* 7,8,a Simon, J. R., Vaidya, C. J., Howard, J. H., and Howard, D. V. (2012). The effects of aging on the neural basis of implicit associative learning in a probabilistic triplets learning task. J. Cogn. Neurosci . 24, 451–463. doi: 10.1162/jocn_a_00116

Smith, P. J., Blumenthal, J. A., Babyak, M. A., Hinderliter, A., and Sherwood, A. (2011). Association of vascular health and neurocognitive performance in overweight adults with high blood pressure. J. Clin. Exp. Neuropsychol . 33, 559–566. doi: 10.1080/13803395.2010.537648

Sokolova, I. A., Manukhina, E. B., Blinkov, S. M., Koshelev, V. B., Pinelis, V. G., and Rodionov, I. M. (1985). Rarefication of the arterioles and capillary network in the brain of rats with different forms of hypertension. Microvasc. Res . 30, 1–9. doi: 10.1016/0026-2862(85)90032-9

* 2,5,d Solbakk, A.-K., Fuhrmann Alpert, G., Furst, A. J., Hale, L. A., Oga, T., Chetty, S. et al. (2008). Altered prefrontal function with aging: insights into age-associated performance decline. Brain Res . 1232, 30–47. doi: 10.1016/j.brainres.2008.07.060

* 5,6,a Solé-Padullés, C., Bartrés-Faz, D., Junqué, C., Vendrell, P., Rami, L., Clemente, I. C. et al. (2009). Brain structure and function related to cognitive reserve variables in normal aging, mild cognitive impairment and Alzheimer's disease. Neurobiol. Aging 30, 1114–1124. doi: 10.1016/j.neurobiolaging.2007.10.008

* 5,6,d,■ Spaniol, J., and Grady, C. (2012). Aging and the neural correlates of source memory: over-recruitment and functional reorganization. Neurobiol. Aging 33, 425.e3–18. doi: 10.1016/j.neurobiolaging.2010.10.005

* 6,7,a Sperling, R. A., Bates, J. F., Chua, E. F., Cocchiarella, A. J., Rentz, D. M., Rosen, B. R. et al. (2003). fMRI studies of associative encoding in young and elderly controls and mild Alzheimer's disease. J. Neurol. Neurosurg. Psychiatr . 74, 44–50. doi: 10.1136/jnnp.74.1.44

* 10,12,b Stebbins, G. T., Carrillo, M. C., Dorfman, J., Dirksen, C., Desmond, J. E., Turner, D. A. et al. (2002). Aging effects on memory encoding in the frontal lobes. Psychol. Aging 17, 44–55. doi: 10.1037/0882-7974.17.1.44

Steen, E., Terry, B. M., Rivera, E. J., Cannon, J. L., Neely, T. R., Tavares, R. et al. (2005). Impaired insulin and insulin-like growth factor expression and signaling mechanisms in Alzheimer's disease—is this type 3 diabetes? J. Alzheimers Dis . 7, 63–80.

* 5,6,b Steffener, J., Brickman, A. M., Rakitin, B. C., Gazes, Y., and Stern, Y. (2009). The impact of age-related changes on working memory functional activity. Brain Imaging Behav . 3, 142–153. doi: 10.1007/s11682-008-9056-x

* 5,6,a Stern, Y., Habeck, C., Moeller, J., Scarmeas, N., Anderson, K. E., Hilton, H. J. et al. (2005). Brain networks associated with cognitive reserve in healthy young and old adults. Cereb. Cortex 15, 394–402. doi: 10.1093/cercor/bhh142

* 4, a Stern, Y., Rakitin, B. C., Habeck, C., Gazes, Y., Steffener, J., Kumar, A. et al. (2012). Task difficulty modulates young-old differences in network expression. Brain Res . 1435, 130–145. doi: 10.1016/j.brainres.2011.11.061

* 7,8,a Stern, Y., Zarahn, E., Habeck, C., Holtzer, R., Rakitin, B. C., Kumar, A. et al. (2008). A common neural network for cognitive reserve in verbal and object working memory in young but not old. Cereb. Cortex 18, 959–967. doi: 10.1093/cercor/bhm134

* 4, a Stevens, W. D., Hasher, L., Chiew, K. S., and Grady, C. L. (2008). A neural mechanism underlying memory failure in older adults. J. Neurosci . 28, 12820–12824. doi: 10.1523/JNEUROSCI.2622-08.2008

* 7,9,a St Jacques, P. L., Rubin, D. C., and Cabeza, R. (2012). Age-related effects on the neural correlates of autobiographical memory retrieval. Neurobiol. Aging 33, 1298–1310. doi: 10.1016/j.neurobiolaging.2010.11.007

* 5,6,a Thiyagesh, S. N., Farrow, T. F. D., Parks, R. W., Accosta-Mesa, H., Young, C., Wilkinson, I. D. et al. (2009). The neural basis of visuospatial perception in Alzheimer's disease and healthy elderly comparison subjects: an fMRI study. Psychiatry Res . 172, 109–116. doi: 10.1016/j.pscychresns.2008.11.002

* 7,8,a Thomsen, T., Specht, K., Rimol, L. M., Hammar, A., Nyttingnes, J., Ersland, L. et al. (2004). Brain localization of attentional control in different age groups by combining functional and structural MRI. Neuroimage 22, 912–919. doi: 10.1016/j.neuroimage.2004.02.015

* 5,9,a Townsend, J., Adamo, M., and Haist, F. (2006). Changing channels: an fMRI study of aging and cross-modal attention shifts. Neuroimage 31, 1682–1692. doi: 10.1016/j.neuroimage.2006.01.045

Triantafyllidi, H., Arvaniti, C., Lekakis, J., Ikonomidis, I., Siafakas, N., Tzortzis, S. et al. (2009). Cognitive impairment is related to increased arterial stiffness and microvascular damage in patients with never-treated essential hypertension. Am. J. Hypertens . 22, 525–530. doi: 10.1038/ajh.2009.35

* 5,6,b Trivedi, M. A., Murphy, C. M., Goetz, C., Shah, R. C., Gabrieli, J. D. E., Whitfield-Gabrieli, S. et al. (2008a). fMRI activation changes during successful episodic memory encoding and recognition in amnestic mild cognitive impairment relative to cognitively healthy older adults. Dement. Geriatr. Cogn. Disord . 26, 123–137. doi: 10.1159/000148190

* 1,6,a Trivedi, M. A., Schmitz, T. W., Ries, M. L., Hess, T. M., Fitzgerald, M. E., Atwood, C. S. et al. (2008b). fMRI activation during episodic encoding and metacognitive appraisal across the lifespan: risk factors for Alzheimer's disease. Neuropsychologia 46, 1667–1678. doi: 10.1016/j.neuropsychologia.2007.11.035

* 4, b Trivedi, M. A., Stoub, T. R., Murphy, C. M., George, S., deToledo-Morrell, L., Shah, R. C. et al. (2011). Entorhinal cortex volume is associated with episodic memory related brain activation in normal aging and amnesic mild cognitive impairment. Brain Imaging Behav . 5, 126–136. doi: 10.1007/s11682-011-9117-4

* 5,6,a Tsukiura, T., Sekiguchi, A., Yomogida, Y., Nakagawa, S., Shigemune, Y., Kambara, T. et al. (2011). Effects of aging on hippocampal and anterior temporal activations during successful retrieval of memory for face-name associations. J. Cogn. Neurosci . 23, 200–213. doi: 10.1162/jocn.2010.21476

Tuligenga, R. H., Dugravot, A., Tabák, A. G., Elbaz, A., Brunner, E. J., Kivimäki, M. et al. (2014). Midlife type 2 diabetes and poor glycaemic control as risk factors for cognitive decline in early old age: a post-hoc analysis of the Whitehall II cohort study. Lancet Diabetes Endocrinol . 2, 228–235. doi: 10.1016/S2213-8587(13)70192-X.

* 7,8,a Tyler, L. K., Shafto, M. A., Randall, B., Wright, P., Marslen-Wilson, W. D., and Stamatakis, E. A. (2010). Preserving syntactic processing across the adult life span: the modulation of the frontotemporal language system in the context of age-related atrophy. Cereb. Cortex 20, 352–364. doi: 10.1093/cercor/bhp105

* 7,8,a Vallesi, A., McIntosh, A. R., and Stuss, D. T. (2009). Temporal preparation in aging: a functional MRI study. Neuropsychologia 47, 2876–2881. doi: 10.1016/j.neuropsychologia.2009.06.013

* 5,6,a,■ Vallesi, A., McIntosh, A. R., and Stuss, D. T. (2011). Overrecruitment in the aging brain as a function of task demands: evidence for a compensatory view. J. Cogn. Neurosci . 23, 801–815. doi: 10.1162/jocn.2010.21490

van den Berg, E., Reijmer, Y. D., de Bresser, J., Kessels, R. P. C., Kappelle, L. J., and Biessels, G. J. (2010). A 4 year follow-up study of cognitive functioning in patients with type 2 diabetes mellitus. Diabetologia 53, 58–65. doi: 10.1007/s00125-009-1571-9

* 4, a Vandenbroucke, M. W. G., Goekoop, R., Duschek, E. J. J., Netelenbos, J. C., Kuijer, J. P. A., Barkhof, F. et al. (2004). Interindividual differences of medial temporal lobe activation during encoding in an elderly population studied by fMRI. Neuroimage 21, 173–180. doi: 10.1016/j.neuroimage.2003.09.043

* 4, c van der Veen, F. M., Nijhuis, F. A. P., Tisserand, D. J., Backes, W. H., and Jolles, J. (2006). Effects of aging on recognition of intentionally and incidentally stored words: an fMRI study. Neuropsychologia 44, 2477–2486. doi: 10.1016/j.neuropsychologia.2006.04.023

van Elderen, S. G. C., de Roos, A., de Craen, A. J. M., Westendorp, R. G. J., Blauw, G. J., Jukema, J. W. et al. (2010). Progression of brain atrophy and cognitive decline in diabetes mellitus: a 3-year follow-up. Neurology 75, 997–1002. doi: 10.1212/WNL.0b013e3181f25f06

* 7,8,a van Impe, A., Coxon, J. P., Goble, D. J., Wenderoth, N., and Swinnen, S. P. (2011). Age-related changes in brain activation underlying single- and dual-task performance: visuomanual drawing and mental arithmetic. Neuropsychologia 49, 2400–2409. doi: 10.1016/j.neuropsychologia.2011.04.016

* 5,6,a,• Velanova, K., Lustig, C., Jacoby, L. L., and Buckner, R. L. (2007). Evidence for frontally mediated controlled processing differences in older adults. Cereb. Cortex 17, 1033–1046. doi: 10.1093/cercor/bhl013

* 7,8,a,❖ Venkatraman, V. K., Aizenstein, H., Guralnik, J., Newman, A. B., Glynn, N. W., Taylor, C. et al. (2010). Executive control function, brain activation and white matter hyperintensities in older adults. Neuroimage 49, 3436–3442. doi: 10.1016/j.neuroimage.2009.11.019

Virta, J. J., Heikkilä, K., Perola, M., Koskenvuo, M., Räihä, I., Rinne, J. O. et al. (2013). Midlife cardiovascular risk factors and late cognitive impairment. Eur. J. Epidemiol . 28, 405–416. doi: 10.1007/s10654-013-9794-y

* 7,8,a Voelcker-Rehage, C., Godde, B., and Staudinger, U. M. (2010). Physical and motor fitness are both related to cognition in old age. Eur. J. Neurosci . 31, 167–176. doi: 10.1111/j.1460-9568.2009.07014.x

* 7,8,a Waiter, G. D., Fox, H. C., Murray, A. D., Starr, J. M., Staff, R. T., Bourne, V. J. et al. (2008). Is retaining the youthful functional anatomy underlying speed of information processing a signature of successful cognitive ageing? An event-related fMRI study of inspection time performance. Neuroimage 41, 581–595. doi: 10.1016/j.neuroimage.2008.02.045

Waldstein, S. R., Carrington, S., Thayer, J. F., Najjar, S. S., Scuteri, A., and Zonderman, A. B. (2008). Pulse pressure and pulse wave velocity are related to cognitive decline in the Baltimore Longitudinal Study of Aging. Hypertension 51, 99–104. doi: 10.1161/HYPERTENSIONAHA.107.093674

* 5,6,a Wang, L., Laviolette, P., O'Keefe, K., Putcha, D., Bakkour, A., Van Dijk, K. R. A. et al. (2010a). Intrinsic connectivity between the hippocampus and posteromedial cortex predicts memory performance in cognitively intact older individuals. Neuroimage 51, 910–917. doi: 10.1016/j.neuroimage.2010.02.046

* 7,9,a Wang, L., Li, Y., Metzak, P., He, Y., and Woodward, T. S. (2010b). Age-related changes in topological patterns of large-scale brain functional networks during memory encoding and recognition. Neuroimage 50, 862–872. doi: 10.1016/j.neuroimage.2010.01.044

* 7,9,a Wang, T. H., Kruggel, F., and Rugg, M. D. (2009). Effects of advanced aging on the neural correlates of successful recognition memory. Neuropsychologia 47, 1352–1361. doi: 10.1016/j.neuropsychologia.2009.01.030

Weis, S., Leube, D., Erb, M., Heun, R., Grodd, W., and Kircher, T. (2011). Functional neuroanatomy of sustained memory encoding performance in healthy aging and in Alzheimer's disease. Int. J. Neurosci . 121, 384–392. doi: 10.3109/00207454.2011.565892

White, W. B., Wolfson, L., Wakefield, D. B., Hall, C. B., Campbell, P., Moscufo, N. et al. (2011). Average daily blood pressure, not office blood pressure, is associated with progression of cerebrovascular disease and cognitive decline in older people. Circulation 124, 2312–2319. doi: 10.1161/CIRCULATIONAHA.111.037036

Whitehead, B. P., Dixon, R. A., Hultsch, D. F., and MacDonald, S. W. S. (2011). Are neurocognitive speed and inconsistency similarly affected in type 2 diabetes? J. Clin. Exp. Neuropsychol . 33, 647–657. doi: 10.1080/13803395.2010.547845

* 7,9,a,× Wierenga, C. E., Benjamin, M., Gopinath, K., Perlstein, W. M., Leonard, C. M., Rothi, L. J. G. et al. (2008). Age-related changes in word retrieval: role of bilateral frontal and subcortical networks. Neurobiol. Aging 29, 436–451. doi: 10.1016/j.neurobiolaging.2006.10.024

* 7,9,b,■,□ Wierenga, C. E., Stricker, N. H., McCauley, A., Simmons, A., Jak, A. J., Chang, Y.-L. et al. (2010). Increased functional brain response during word retrieval in cognitively intact older adults at genetic risk for Alzheimer's disease. Neuroimage 51, 1222–1233. doi: 10.1016/j.neuroimage.2010.03.021

* 5,6,a Wood, G., Ischebeck, A., Koppelstaetter, F., Gotwald, T., and Kaufmann, L. (2009). Developmental trajectories of magnitude processing and interference control: an FMRI study. Cereb. Cortex 19, 2755–2765. doi: 10.1093/cercor/bhp056

* 7,8,a Woodard, J. L., Seidenberg, M., Nielson, K. A., Antuono, P., Guidotti, L., Durgerian, S. et al. (2009). Semantic memory activation in amnestic mild cognitive impairment. Brain 132, 2068–2078. doi: 10.1093/brain/awp157

* 5,6,a Woodard, J. L., Seidenberg, M., Nielson, K. A., Miller, S. K., Franczak, M., Antuono, P. et al. (2007). Temporally graded activation of neocortical regions in response to memories of different ages. J. Cogn. Neurosci . 19, 1113–1124. doi: 10.1162/jocn.2007.19.7.1113

* 5,6,c Woodard, J. L., Seidenberg, M., Nielson, K. A., Smith, J. C., Antuono, P., Durgerian, S. et al. (2010). Prediction of cognitive decline in healthy older adults using fMRI. J. Alzheimers Dis . 21, 871–885. doi: 10.3233/JAD-2010-091693

* 5,6,c Woodard, J. L., Sugarman, M. A., Nielson, K. A., Smith, J. C., Seidenberg, M., Durgerian, S. et al. (2012). Lifestyle and genetic contributions to cognitive decline and hippocampal structure and function in healthy aging. Curr. Alzheimer Res . 9, 436–446. doi: 10.2174/156720512800492477

* 5,6,a,• Wu, J.-T., Wu, H.-Z., Yan, C.-G., Chen, W.-X., Zhang, H.-Y., He, Y. et al. (2011). Aging-related changes in the default mode network and its anti-correlated networks: a resting-state fMRI study. Neurosci. Lett . 504, 62–67. doi: 10.1016/j.neulet.2011.08.059

Xu, W., Caracciolo, B., Wang, H.-X., Winblad, B., Bäckman, L., Qiu, C. et al. (2010). Accelerated progression from mild cognitive impairment to dementia in people with diabetes. Diabetes 59, 2928–2935. doi: 10.2337/db10-0539

Yaffe, K., Blackwell, T., Kanaya, A. M., Davidowitz, N., Barrett-Connor, E., and Krueger, K. (2004). Diabetes, impaired fasting glucose, and development of cognitive impairment in older women. Neurology 63, 658–663. doi: 10.1212/01.WNL.0000134666.64593.BA

Yaffe, K., Lindquist, K., Schwartz, A. V., Vitartas, C., Vittinghoff, E., Satterfield, S. et al. (2011). Advanced glycation end product level, diabetes, and accelerated cognitive aging. Neurology 77, 1351–1356. doi: 10.1212/WNL.0b013e3182315a56

Yakushiji, Y., Noguchi, T., Hara, M., Nishihara, M., Eriguchi, M., Nanri, Y. et al. (2012). Distributional impact of brain microbleeds on global cognitive function in adults without neurological disorder. Stroke 43, 1800–1805. doi: 10.1161/STROKEAHA.111.647065

Yan, S. F., Ramasamy, R., and Schmidt, A. M. (2008). Mechanisms of disease: advanced glycation end-products and their receptor in inflammation and diabetes complications. Nat. Clin. Pract. Endocrinol. Metab . 4, 285–293. doi: 10.1038/ncpendmet0786

Yeung, S. E., Fischer, A. L., and Dixon, R. A. (2009). Exploring effects of type 2 diabetes on cognitive functioning in older adults. Neuropsychology 23, 1–9. doi: 10.1037/a0013849

* 5,6,c Ystad, M., Eichele, T., Lundervold, A. J., and Lundervold, A. (2010). Subcortical functional connectivity and verbal episodic memory in healthy elderly—a resting state fMRI study. Neuroimage 52, 379–388. doi: 10.1016/j.neuroimage.2010.03.062

* 5,6,c Ystad, M., Hodneland, E., Adolfsdottir, S., Haász, J., Lundervold, A. J., Eichele, T. et al. (2011). Cortico-striatal connectivity and cognition in normal aging: a combined DTI and resting state fMRI study. Neuroimage 55, 24–31. doi: 10.1016/j.neuroimage.2010.11.016

Zhong, Y., Miao, Y., Jia, W. P., Yan, H., Wang, B. Y., and Jin, J. (2012a). Hyperinsulinemia, insulin resistance and cognitive decline in older cohort. Biomed. Environ. Sci . 25, 8–14. doi: 10.3967/0895-3988.2012.01.002

Zhong, Y., Zhang, X. Y., Miao, Y., Zhu, J. H., Yan, H., Wang, B. Y. et al. (2012b). The relationship between glucose excursion and cognitive function in aged type 2 diabetes patients. Biomed. Environ. Sci . 25, 1–7. doi: 10.3967/0895-3988.2012.01.001

Zhou, H., Lu, W., Shi, Y., Bai, F., Chang, J., Yuan, Y. et al. (2010). Impairments in cognition and resting-state connectivity of the hippocampus in elderly subjects with type 2 diabetes. Neurosci. Lett . 473, 5–10. doi: 10.1016/j.neulet.2009.12.057

* 7,8,a,• Zhu, D. C., Zacks, R. T., and Slade, J. M. (2010). Brain activation during interference resolution in young and older adults: an fMRI study. Neuroimage 50, 810–817. doi: 10.1016/j.neuroimage.2009.12.087

Keywords: type 2 diabetes mellitus, hypertension, cognition, aging, imaging

Citation: Meusel L-AC, Kansal N, Tchistiakova E, Yuen W, MacIntosh BJ, Greenwood CE and Anderson ND (2014) A systematic review of type 2 diabetes mellitus and hypertension in imaging studies of cognitive aging: time to establish new norms. Front. Aging Neurosci . 6 :148. doi: 10.3389/fnagi.2014.00148

Received: 25 January 2014; Accepted: 17 June 2014; Published online: 08 July 2014.

Reviewed by:

Copyright © 2014 Meusel, Kansal, Tchistiakova, Yuen, MacIntosh, Greenwood and Anderson. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) . The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

*Correspondence: Nicole D. Anderson, Rotman Research Institute, Baycrest, 3560 Bathurst Street, Toronto, ON M6A 2E1, Canada e-mail: [email protected]

• Open access
• Published: 16 May 2024

DNA methylation and type 2 diabetes: a systematic review

• Nikhil Nadiger 1 , 2 ,
• Jyothisha Kana Veed 2 ,
• Priyanka Chinya Nataraj 2   nAff3 &
• Arpita Mukhopadhyay 2  

Clinical Epigenetics volume  16 , Article number:  67 ( 2024 ) Cite this article

516 Accesses

Metrics details

DNA methylation influences gene expression and function in the pathophysiology of type 2 diabetes mellitus (T2DM). Mapping of T2DM-associated DNA methylation could aid early detection and/or therapeutic treatment options for diabetics.

A systematic literature search for associations between T2DM and DNA methylation was performed. Prospero registration ID: CRD42020140436.

PubMed and ScienceDirect databases were searched (till October 19, 2023). Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and New Castle Ottawa scale were used for reporting the selection and quality of the studies, respectively.

Thirty-two articles were selected. Four of 130 differentially methylated genes in blood, adipose, liver or pancreatic islets ( TXNIP , ABCG1 , PPARGC1A , PTPRN2 ) were reported in > 1 study. TXNIP was hypomethylated in diabetic blood across ethnicities. Gene enrichment analysis of the differentially methylated genes highlighted relevant disease pathways (T2DM, type 1 diabetes and adipocytokine signaling). Three prospective studies reported association of methylation in IGFBP2 , MSI2 , FTO , TXNIP , SREBF1 , PHOSPHO1 , SOCS3 and ABCG1 in blood at baseline with incident T2DM/hyperglycemia. Sex-specific differential methylation was reported only for HOOK2 in visceral adipose tissue (female diabetics: hypermethylated, male diabetics: hypomethylated). Gene expression was inversely associated with methylation status in 8 studies, in genes including ABCG1 (blood), S100A4 (adipose tissue), PER2 (pancreatic islets), PDGFA (liver) and PPARGC1A (skeletal muscle).

This review summarizes available evidence for using DNA methylation patterns to unravel T2DM pathophysiology. Further validation studies in diverse populations will set the stage for utilizing this knowledge for identifying early diagnostic markers and novel druggable pathways.

Introduction

Type 2 diabetes mellitus (T2DM) is a disorder of genetic and environmental factors. It is projected to affect 693 million people worldwide by 2045 [ 1 ]. DNA methylation had been proposed as one of the epigenetic phenomena for explaining the missing heritability of T2DM, as multiple, large genome-wide association studies have been able to account for only < 20% of the estimated T2DM heritability [ 2 ]. DNA methylation is an epigenetic phenomenon in which the C5 carbon of the cytosine residue is attached to a methyl group, predominantly in cytosine-phosphate-guanine (CpG) sites [ 3 , 4 , 5 ]. This epigenetic alteration influences gene expression, and thereby, gene function [ 6 , 7 ].

DNA methylation has been studied extensively in relation to T2DM, and 3 systematic reviews have summarized the findings a few years back [ 8 , 9 , 10 ]. From systematic literature done till August 2015, Muka et al. [ 10 ] could not find any consistent association between global DNA methylation with T2DM, glucose, insulin and insulin resistance and reported epigenetic regulation of few candidate genes in blood cells, muscle, adipose tissue and placenta without any overlap between them. Walaszczyk et al . [ 9 ] could replicate association of methylation with T2DM in blood samples from the Lifelines study at 5 CpGs (in ABCG1 , LOXL2 , TXNIP , SLC1A5 and SREBF1 ) out of the 52 CpGs they identified as reported to be differentially methylated in T2DM through a systematic review of the literature done till April 2017. Willmer et al . [ 8 ] also focused on differential methylation signatures in blood samples and reported TCF7L2 , KCNQ1 , ABCG1 , TXNIP , PHOSPHO1 , SREBF1 , SLC30A8 and FTO genes to be reproducibly associated with T2DM across multiple population groups in the literature reviewed between January 2002 and July 2018.

DNA methylation has been touted as a strong candidate biological process for identification of diagnostic and therapeutics for T2DM [ 11 ]. While the available systematic reviews have looked at DNA methylation associated with T2DM [ 8 , 9 , 10 ], they have not evaluated T2DM-associated DNA methylation comprehensively in all available human tissue and cell types. We set out to fill this research gap with the no time period cutoff until October 19, 2023, and including all available human tissue and cell types. We also report associated gene expression data, role of sex and ethnicity, in relation to DNA methylation in our review.

PubMed and Science Direct databases were independently searched by authors (NN, PN and JKV) using the key terms “type 2 diabetes mellitus” and “DNA methylation,” and their associated terms for all studies published up to October 19, 2023. All articles from the time of publication listing were considered, and as such no start date was set. No filters were applied during the search using the keywords, so as to not exclude any mislabeled/mis-annotated article type. The detailed search strategy is given in Additional file 1 : Table S1.

Study inclusion and exclusion criteria

The inclusion criteria were full-text English language articles on DNA methylation associated with T2DM in human subjects. Case–control and prospective studies investigating genome-wide methylation were included. Reviews, animal model studies, in vitro studies, irrelevant articles and articles published in other languages were excluded.

All participants, regardless of gender and ethnicity, classified as adults aged 18 years and above were included. All individuals who did not satisfy these criteria—children and adolescents under 18 years of age; as well as subjects with type 1 diabetes (T1DM) or gestational diabetes were excluded. As the association of DNA methylation with T2DM was the focus of this systematic review, intervention studies and clinical trials were excluded. Studies reporting association of DNA methylation with diabetes-related traits (hyperglycemia and insulin resistance) were retained.

All the articles were assessed for their eligibility based on their abstract or full text.

Disagreements between the authors, such as categorization and selection of articles, and data extraction, were resolved through discussion with AM. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklist was followed to represent the method used [ 12 ]. A total of 32 full-text articles are included in this systematic review.

The assessment of quality of the studies was done by adapting the New Castle Ottawa scale (NOS) [ 13 ]. The parameters used for the assessment are adequacy of case definition, representativeness of cases, selection of controls, definition of controls, comparability of cases and controls, ascertainment of exposure and method used for ascertainment of cases and controls. Scores were given to each of the included studies, and the total score was calculated according to the score sheet (NOS).

This review protocol was registered with the International Prospective Register of Systematic Reviews (PROSPERO) database ( https://www.crd.york.ac.uk/prospero/ ) [ 14 ] (accessed April 18, 2023) (registration ID: CRD42020140436).

Pathologically connected pathways with differentially methylated genes in T2DM were analyzed using Kyoto Encyclopedia of Genes and Genomes (KEGG) and Jensen Disease database via Enrichr-KG [ 15 ].

We identified a total of 5819 articles during the initial search. Duplicates, irrelevant articles based on the study design, publication language, article type, and other articles not within our scope of review were removed. Thirty-two full-text articles were finally selected (Fig.  1 ).

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [ 12 ] flowchart for the literature search process, performed up to October 19, 2023

NOS was used to access the quality of the articles. Of the 32 studies, 16 were assigned a score of more than 5, indicating high quality (Additional file 2 : Table S2). As all the studies have used the same method of ascertainment for cases and controls, and the authors are not blinded to case–control status, these redundant scores are not presented. As the nonresponse rate was not available for any of the studies, this also has been omitted from the quality assessment table.

Case–control studies that reported differential DNA methylation between T2DM (case) and normoglycemic (control) subjects or reported associations between DNA methylation and clinical parameters related to glycemic control of the subjects (HbA1c, fasting blood glucose) and prospective nested case–control studies that reported differential DNA methylation measured at baseline/recruitment between subjects who developed T2DM (incident cases) and those that remained normoglycemic (control) during the follow-up period were finally included.

Participant details such as number of cases and controls and location of the study are also included. Details of the study participants who do not explicitly belong to either case or control group are also presented. The tissue source of the gene/loci identified in; method used for determining methylation status; and the validation method used for confirming the methylation status are tabulated in Table  1 .

The loci/genes reported to be differentially methylated are tabulated in Table  2 , where their methylation status is represented as downward arrow (hypomethylation) or upward arrow (hypermethylation). Wherever reported, the statistical significance of methylation ( P value) is also mentioned. For studies reporting more than 10 differentially methylated genes, the top 5 hypo- and hypermethylated genes are listed.

Methods of DNA methylation analysis

Majority of the evaluated studies had employed array-based techniques to assess DNA methylation levels. Eighteen of 32 studies used Illumina 450 k array. Other array-based studies used Illumina 27 k array (2 studies), Illumina EPIC BeadChip array (4 studies; of which 2 studies specifically mentioned using the 850 k array—EPIC v1 array targeting 850 k probes), Affymetrix SNP6 microarray (1 study), Affymetrix GeneChip promoter 1.0R array (1 study) or Affymetrix axiom genome-wide Taiwan BioBank (TWB) array (1 studies). Rest of the studies used techniques such as methylated DNA immunoprecipitation (MEDIP) (2 studies), MEDIP-chromatin immune precipitation (1 study), reduced representation bisulfite sequencing (1 study) or next-generation sequencing (1 study) to measure DNA methylation levels.

Tissues used in DNA methylation analyses

Of the 32 articles retrieved, 17 (53%) studies used blood samples, 4 (13%) studies used pancreatic islet samples, 5 (16%) studies used adipose tissue samples, 2 (6%) studies used liver samples, 1 (3%) study used spermatozoa samples and 3 (9%) used skeletal muscle samples for their DNA methylation analyses. None of the 32 studies reviewed here utilized more than one tissue from the same subjects for DNA methylation analyses.

Genome-wide methylation analysis for T2DM

Of the 32 genome-wide methylation studies reviewed here, we identified a total of 130 loci that were differentially methylated between T2DM cases and controls across. In an instance where a study reports < 10 differentially methylated genes/loci, they are presented individually. However, in the case of a study which reports > 10 genes/loci, only the top 5 hypo- and 5 hypermethylated genes are highlighted for brevity and reported in Table  2 . The direction of methylation (hyper- or hypomethylated in T2DM, compared to controls) and the reported P values (both unadjusted, and after multiple testing correction) have been included.

We identified genes such as ABCG1, PPARGC1A , PTPRN2 and TXNIP with well-known T2DM genetic risk variants, which were consistently reported to be differentially methylated in more than one study (Fig.  2 ). Tissues used in identification of these gene were blood cells, liver, pancreatic islets and adipose tissue. TXNIP (cg19693031) was the most common gene identified consistently as hypomethylated in diabetic blood (9 studies). TXNIP also harbors established T2DM genetic risk variants [ 16 , 17 ].

A pie chart depicting the genes that were consistently reported to be differentially methylated in ≥ 2 studies in various tissues from T2DM subjects. ↑: Hypomethylation, ↓: Hypermethylation in T2DM individuals compared to normoglycemics. PPARGC1A (chr4: 24,024,251–500) hypomethylated, (chr4: 24,111,501–750) hypermethylated in spermatozoa [ 57 ]

Although blood is not an insulin-responsive tissue, it is the prime minimally invasive tissue available for investigating T2DM-associated epigenetic markers. With the bulk (50%) of the studies coming from Europe, ABCG1 [ 18 , 19 ] and TXNIP [ 16 , 17 , 19 , 20 , 21 , 22 , 23 , 24 , 25 ] were some of the blood-based epigenetic markers which were reported to be associated with T2DM in more than one study. We were unable to find any study where differential methylation was investigated simultaneously in blood and other tissues from the same subjects.

Pancreatic islets

Insufficient secretion of insulin from pancreatic beta cells and increased secretion of glucagon from pancreatic alpha cells leads to development of T2DM and is known to be regulated by DNA methylation [ 26 ]. Three of the 32 studies, from Italy, South Korea and Sweden, included in this review have interrogated DNA methylation in pancreatic islets from T2DM individuals, donated after their death. Regions in SFRS2IP [ 3 ], MSI2 [ 27 ], which are known to be associated with critical roles in nucleic acid binding, and B3GNT7 [ 28 ] that is involved in synthesis of glycoprotein, were reported to be hypomethylated in pancreatic islets from T2DM individuals. Considering that DNA methylation can change based on the time of collection of tissue after death [ 29 , 30 ], findings from these studies need to be interpreted in cognizance of the lack of details available in these studies about the cause of death or collection and storage of pancreatic islet tissue after death.

Adipose tissue

Adipose tissue is known to play a critical role in regulating body metabolism and energy homeostasis [ 31 ]. Dysregulation in adipose biology imposes serious health complications such as obesity and development of T2DM [ 31 ]. DNA methylation is an important regulator factor in development [ 32 , 33 ] and dysfunction [ 34 , 35 ] of adipose tissue. Five studies—4 of these representing the European population—included in this review have dissected whether T2DM, and related risk factors are associated with epigenetic modifications in human adipose tissue [ 36 , 37 , 38 , 39 , 40 ]. It is possible that DNA methylation alterations in these reported genes including C1orf52 [ 36 ], HOOK2 [ 37 ], MFSD1 [ 38 ], HNF4A [ 39 ] and L1TD1 [ 40 ] contribute to or are caused by T2DM.

C1orf52 is involved in RNA binding in adipose tissue [ 41 ], and HOOK2 is responsible for cytoskeleton maintenance via regulation of microtubules [ 42 ], while MSFD1 regulates lysosome transport [ 43 ]. Epigenetic alterations in such genes involved in cell structure and function can cause dysfunction in adipose tissue, thereby leading to insulin resistance. While HNF4A mainly regulates transcription in hepatocytes and is associated with Fanconi renotubular syndrome 4 with maturity-onset diabetes of the young [ 44 ] and maturity-onset diabetes of the young, type 1 [ 45 ], it is also known to play a role in lipid and glucose metabolism [ 46 , 47 ]. L1TD1 is predicted to be involved in single-stranded RNA-binding activity [ 48 ].

Liver is known to be involved in regulating glucose level by storing and releasing glycogen in response to insulin and glucagon [ 49 ]. Impaired hepatic gluconeogenesis, glycogenolysis and insulin sensitivity are known to play an important role in T2DM development and other risk factors. Altered hepatic metabolism could be the cause or consequence of DNA methylation modification. Genes involved in intracellular tyrosine kinase activity— PDGFA [ 50 ], transferring phosphorus-containing groups and protein tyrosine kinase activity— RIPK4 [ 51 ], heme binding and oxidoreductase activity— CYB561D1 [ 51 ], were found to be hypomethylated in the diabetic groups. However, the gene involved in inflammation— IL23Ap19 [ 51 ] was identified to be hypermethylated in the diabetic group. Of the two studies reported here, one was from France and the other from Finland.

Gene expression studies

Out of the 32 studies reviewed, 8 had also examined differences in gene expression between T2DM and normoglycemic individuals. To examine if increase in methylation of a gene causes decrease in expression of that gene, we analyzed the studies that report both differentially methylated genes and gene expression, in the same population and study setting, using tissues from the same study participants (Table  3 ). For most of the loci with both DNA methylation and gene expression data available, we found that increase in methylation was associated with decrease in expression, concurrent to the current understanding [ 6 ]. Hypermethylation of PPARGC1A in skeletal muscles [ 52 ], ABCG1 in blood [ 18 ] and PER2 in pancreatic islets [ 3 ] was associated with lower expression of the corresponding genes.

Twin studies

Five of the 32 studies reviewed here have investigated DNA methylation in monozygotic twin cohorts [ 17 , 21 , 28 , 36 , 53 ] (Table  4 ). MALTI [ 53 ] which is known to be involved in energy and insulin signaling pathways [ 54 ], PTBP1 [ 36 ] that is involved in nucleic acid binding, and ANO8 [ 28 ] that is involved in calcium transport, were hypermethylated in diabetic twins in peripheral blood, adipose tissue and pancreatic islets, respectively. TXNIP [ 17 , 21 ], COL21A1 [ 36 ] and B3GNT7 [ 28 ] were hypomethylated in blood cells, adipose tissue and pancreatic islets, respectively, from the diabetic twins. Dayeh et al . reported differential methylation of ABCG1 (hypermethylated in blood and adipose tissue) and PHOSPHO1 (hypomethylated in skeletal muscle) in monozygotic twins discordant for T2DM [ 55 ].

Association between diabetes related traits and DNA methylation

Only 4 of the 32 studies reported association between diabetes-related traits (hyperglycemia and insulin resistance) and DNA methylation [ 17 , 18 , 19 , 22 ]. Kriebel et al . reported significant association between measures of glucose metabolism phenotypic traits and methylation levels of 31 CpG sites in PBMCs [ 18 ]. Five CpGs were found to be associated with fasting glucose, 1 CpG with 2-h glucose, 8 with fasting insulin and 26 with Homeostatic Model Assessment of Insulin Resistance (HOMA-IR) in model 1 (Table  2 ) [ 18 ]. There was no significant association between HbA1c and DNA methylation levels in model 1; in model 2, after adjustment for body mass index (BMI), the effect strength was reduced by 30% for DNA methylation associations with fasting glucose suggesting that the associations between DNA methylation and diabetes-related traits are partially mediated by BMI [ 18 ].

Kulkarni et al . investigated association between 446,356 autosomal CpG sites and phenotypic traits in PBMCs, of which a total of 51 CpG sites were significantly associated with T2DM, 19 with FBG and 24 with HOMA-IR (Table  2 ) [ 19 ].

Wang et al . report association between 63 differential methylated loci and fasting blood glucose and association between 6 differentially methylated loci with HbA1c in blood samples from twins discordant for diabetes [ 17 ]. Among these, hypomethylation of TXNIP [ 17 , 19 ] and hypermethylation of ABCG1 [ 18 , 19 ] were positively associated with fasting blood glucose (FBG), and hypermethylation of SAMD12 was negatively associated with FBG [ 19 ]. TXNIP hypomethylation in blood cells was found to be associated with hyperglycemia in individuals from Taiwan [ 23 ], France [ 24 ], the USA [ 21 ] and China [ 17 ].

Dawes et al . performed genome-wide DNA methylation on blood samples from normoglycemic (n = 142), pre-diabetic (n = 274) and diabetic (n = 90) individuals [ 22 ]. They identified HbA1c-associated DNA methylation loci by regressing the probes against HbA1c values, while controlling for age, sex and BMI [ 22 ]. They report cg19693031 ( TXNIP ) as the locus most highly associated with HbA1c [ 22 ].

Enrichment analysis of genes differentially methylated in T2DM

Enrichment analysis of signaling pathways relevant to the pathophysiology of T2DM using Enrichr-KG [ 15 ] was done in two steps. Initially, all 130 genes differentially methylated in T2DM in all 32 studies reviewed were included (Fig.  3 ). To take into account reproducibility of these findings, enrichment analysis was separately done specifically for the genes ( ABCG1 , TXNIP , PTPRN2 , PPARGC1A ) that were reported to be differentially methylated in T2DM in more than one study (Fig.  4 ). TXNIP hypomethylation in blood was linked to hyperglycemia. PPARGC1A hypermethylation in skeletal muscles, and two CpG sites that were hyper- and hypomethylated, respectively, in spermatozoa, was linked to hyperglycemia and adipocytokine signaling pathway. PTPRN2 that was reported to be hypermethylated in blood and hypomethylated in adipose tissue was associated with T2DM and T1DM.

Gene enrichment analysis of 17 of the 130 genes reported to be differentially methylated in T2DM subjects in the 32 studies included for review using Enrichr-KG. These genes were mapped to diabetes and related disorders. Insulin resistance, glucagon signaling pathway, glaucoma, AMPK signaling pathway, cholinergic synapse, ovarian cancer, amphetamine addiction and Huntington’s disease were found to be associated with KCNQ1 , FTO , PPARGC1A , PTPRN2 , ELOVL5 , HNF1B , HNF4A , VPS13A , MAEA , CREB1 , CPT1A , PRKCZ , PRKCB , CREB3L2 , CDKN2A and TGFBR3

Gene enrichment analysis of 4 genes reported to be differentially methylated in T2DM subjects in > 1 study from among the 32 studies included for review using Enrichr-KG. Hyperglycemia, type 1 diabetes, adipocytokine signaling pathway, glucagon signaling pathway, longevity regulating pathway and ABC transporters were found to be associated with PPARGC1A , TXNIP , PTPRN2 and ABCG1

Subgroup analysis based on ethnicity

Out of the 32 studies, 16 (50%) were from Europe, 4 (13%) were from North America, 8 (25%) were from Asia and 1 (3%) from Africa. Three studies (9%) did not report their subjects’ ethnicity/demography.

TXNIP was the most commonly reported hypomethylated gene in blood cells of T2DM individuals from all the geographic locations [ 16 , 17 , 19 , 20 , 21 , 22 , 23 , 24 ]. ABCG1 was found be to hypermethylated in blood cells of type 2 diabetics in studies from Europe [ 18 ] and the USA [ 19 ]. PTPRN2 was reported to be hypermethylated in peripheral blood in studies from China [ 56 ] and France [ 24 ]. Conversely, PTPRN2 was reported to be hypomethylated in adipose tissue from a Spanish study [ 37 ].

Subgroup analysis based on sex

PPARGC1A was assessed for differential methylation in two studies which had only male participants [ 52 , 57 ]. PPARGC1A was hypermethylated in skeletal muscle of T2DM men [ 52 ]. Of the two differentially methylated regions in PPARGC1A identified in sperm, chr4: 24,111,501–750 was reported to be hypermethylated, and chr4: 24,024,251–500 was reported to be hypomethylated [ 57 ]. We did not find other epigenome-wide studies that report differential methylation of PPARGC1A in female-only or mixed-sex populations.

PDGFA was found to be hypomethylated in hepatocytes from liver biopsies of female T2DM participants of the discovery group and was later confirmed in both men and women by Abderrahmani et al . [ 50 ]. Similarly, hypomethylation of MSI2 in blood cells was first observed in a discovery group comprised of only men, and then in a replication group of both men and women by Jeon et al . [ 27 ].

In the cg 11,738,485-region (5 CpG nucleotides) of HOOK2 , female T2DM visceral adipose tissue samples were hypermethylated, while male T2DM samples were hypomethylated, compared to non-diabetic sex-matched control samples [ 37 ]. None of the other loci/genes were reported to be differentially methylated in a sex-specific manner.

Internal and/or external validation

Only 22% of the studies reviewed (7 out of 32) validated their findings in an independent set of subjects using the same DNA methylation measurement method that they had used for the discovery set of samples [ 17 , 25 , 27 , 36 , 37 , 50 , 53 ]. Others used either bisulfite pyrosequencing/sequencing (10 studies) [ 3 , 19 , 27 , 28 , 37 , 39 , 52 , 58 , 59 , 60 ], qPCR (1 study) [ 51 ], EpiTYPER (1 study) [ 16 ], Illumina 450 k (3 studies) [ 36 , 50 , 53 ] or MEDIP (1 study) [ 61 ] for their internal validation. Sixteen studies (50%) did not perform any validation for their findings.

Replication for case–control studies

We later looked for candidate-gene DNA methylation studies to see if the differentially methylated genes found in genome-wide studies have been confirmed in them. The following genes were reported to be differentially methylated in T2DM compared to normoglycemic controls in independent candidate-gene DNA methylation studies in the same tissue as the initial discovery group— ABCG1 [ 62 , 63 ], FTO [ 64 , 65 , 66 ], TXNIP [ 67 ] and KCNQ1 [ 64 , 68 ] in PBMCs, and PPARGC1A in pancreatic islets [ 69 ].

Prospective studies

As prospective studies observe the disease condition over a long period, they help in better understanding the role of a gene/set of genes toward pathogenesis. In our review, we came across three such studies that looked at incidence of T2DM and epigenetic modifications in genes associated with this incidence (Table  5 ).

In a 1:1 matched nested case–control study of 290 incident diabetics, who developed T2DM and 290 controls, who remained normoglycemic during the 4-year follow-up, baseline methylation at 7 CpG sites of IGFBP2 in blood cells (4 hypermethylated and 3 hypomethylated in cases) was associated with increased risk of incident T2DM [ 70 ].

Jeon et al . reported that differential methylation of three CpG sites in blood cells at baseline was associated with T2DM/hyperglycemia after a 10-year follow-up [ 27 ]. These CpG sites were cg23586172 (annotated to MSI2 , hypomethylated), cg22604213 (annotated to CXXC4, hypomethylated) and cg25290098 (hypomethylated) in T2DM [ 27 ]. They further reported MSI2 hypomethylation in a replication group of 220 normoglycemic and 220 T2DM individuals [ 27 ]. Furthermore, whole-genome bisulfite sequencing of pancreatic islets of 2 T2DM and 16 normoglycemic individuals revealed that chr17:55,484,635 in MSI2 was hypomethylated in T2DM [ 27 ]. While MSI2 hypomethylation was seen in both pancreatic islets and PBMCs, pancreatic islets showed increased difference of 16% methylation versus 3% in PBMCs of MSI2 in T2DM when compared to normoglycemics [ 27 ]. MSI2 differential methylation was not found to be replicated in locus-specific case–control studies.

From the Jerusalem LRC longitudinal study, Toperoff et al . selected 58 individuals who developed impaired glucose metabolism over a 13-year follow-up and reported hypomethylation of a single CpG site in the first intron of FTO in peripheral blood samples collected at baseline [ 58 ]. Chen et al . similarly reported hypomethylation of FTO in their case–control study [ 57 ].

In a longitudinal study of Indian Asians living in London, UK (1074 incident T2DM and 1590 normoglycemic controls), over 8 years of follow-up, Chambers et al . reported that DNA methylation levels of TXNIP , PROC , C7orf29 , SREBF1 , PHOSPHO1 , SOCS3 and ABCG1 in blood cells were positively associated with future T2DM incidence [ 71 ]. Of these, higher baseline methylation levels in TXNIP , SREBF1 , PHOSPHO1 , SOCS3 and ABCG1 were also associated with incident T2DM in an European cohort of 377 incident T2DM and 764 normoglycemic individuals [ 71 ].

Differential methylation in animal models

To check if animal model studies exist that have investigated or reported differential methylation in the genes identified as differentially methylated in the human case–control studies as playing causal or mechanistic role in the development of T2DM, a simple literature search was done using PubMed and bibliography search. A study in rat pancreatic islets reported Kcnq1 was hypomethylated in older rats (15 months of age) when compared to younger rats (3 months of age), but this difference was not statistically significant, while there was no comparison done with a rat T2DM model [ 72 ]. Though Toperoff et al . reported hypomethylation of KCNQ1 in blood cells [ 58 ], there are no human pancreatic islet studies reporting hypomethylation of KCNQ1 . Identification of multiple variants in genome-wide association studies [ 73 , 74 , 75 , 76 , 77 , 78 , 79 , 80 , 81 ] points toward the likely importance of KCNQ1 in T2DM pathophysiology.

High-fat diet was shown to induce hypermethylation of Tcf7l2, and subsequently, gene expression was decreased in mouse islets [ 82 ]. This is in contrast to the findings where TCF7L2 is hypomethylated in T2DM human blood cells [ 58 ] and pancreatic islets [ 59 ]. It is to be noted that the mice used were non-diabetic adult males aged 8 weeks (equivalent to middle-aged humans [ 83 ]) [ 82 ], while the human study group were a mix of men and women aged about 58–65 years, and for the human pancreatic islet study, the samples had been collected post-mortem [ 58 , 59 ]. Although there is an inverse differential methylation status among mice and humans, it is important to note that a high-fat diet caused suppression of Tcf7l2 gene expression and thus decreases pancreatic beta-cell survival (mediated via the transcription of Wnt/Beta-catenin signaling pathway [ 84 ]) [ 82 ].

From the 32 studies finally included for this systematic review, we identified 130 genes with T2DM-associated differential methylation across all tissues analyzed. These comprise of the top 5 hypo- and hypermethylated genes for studies reporting more than 10 differentially methylated genes/loci. Of these 130 genes, 4 (3%; ABCG1, PPARGC1A , PTPRN2 and TXNIP ) were reported in > 1 studies. The genes and associated pathways with altered DNA methylation in T2DM are conceptually summarized in Fig.  3 (for 16 of the 130 genes, for which pathway analysis could be conducted) and Fig.  4 (for the 4 genes reported to be differentially methylated in > 1 studies).

Previous systematic reviews [ 8 , 9 ] have reported differentially methylated loci in genes in T2DM blood cells including ABCG1 , TXNIP , KCNQ1 . While another such review by Muka et al . reported several epigenetically regulated genes from blood cells, adipose tissue, muscle and placenta, there was no overlap between them, and no association was found between global DNA methylation and T2DM/hyperglycemic markers [ 10 ].

We did not limit our search to a particular method used to identify DNA methylation, and several studies included have used Illumina’s 450 k array. The common method of validation/replication in the studies reviewed here was bisulfite pyrosequencing. We also looked at candidate-gene DNA methylation studies which aimed to replicate/validate the epigenome-wide studies reviewed here and found that in blood cells, ABCG1 [ 62 ], FTO [ 64 ] and KCNQ1 [ 64 ] were hypermethylated, while TXNIP was hypomethylated [ 67 ]. TXNIP codes for thioredoxin-interacting protein, and this protein plays a major role in pathways generating reactive oxygen species [ 85 ], regulating redox-dependent signaling pathways, mediating oxidative stress, suppressing cell growth and inducing pancreatic beta-cell apoptosis [ 86 ]. ABCG1 codes for the protein responsible for intracellular sterol transport [ 87 ], and it regulates cholesterol efflux from macrophages to high-density lipoprotein in diabetics [ 88 ], indicated by altered lipid levels [ 89 ]. While genetic variants and epigenetic modification of KCNQ1 have been linked with T2DM via whole body insulin sensitivity [ 90 ], there is no clear evidence for the mechanistic link. Likewise, there has been no clear evidence of FTO link with T2DM.

As gene expression is known to be regulated by DNA methylation, it is important to validate this claim in the epigenome-wide association studies. We were able to report the relation between DNA methylation in the promoter region and expression of the corresponding gene, as none of the studies had mentioned methylation status of other regions of the genes. Of the studies reviewed here, we found that DNA methylation of genes was inversely related to gene expression. For example, hypomethylation of S100A4 in adipose tissue [ 36 ] and PDGFA in hepatocytes [ 50 ] was associated with increased expression of these genes, and hypermethylation of PPARGC1A in skeletal muscles [ 52 ], ABCG1 in blood [ 18 ] and PER2 in pancreatic islets [ 3 ] was associated with lower expression of the corresponding genes. Even though we observed DNA methylation being related inversely with expression of the corresponding gene in the studies reviewed, this is not a rule as has been reported repeatedly [ 91 ]. It is also important to note that there have been reports of methylation levels differing between different regions of the gene that influence gene expression; for instance, Anastasiadi et al . recently reported that gene expression is dependent on methylation of the first exon, more than methylation of the promoter region [ 92 ]. Moreover, in other studies such as one by Ball and colleagues, highly expressed genes have been reported to have low methylation levels in the promoter region and high methylation levels in rest of the gene body [ 93 ]. We could not, however, evaluate the relations between DNA methylation in various regions of a gene and its corresponding expression in this study since the studies reviewed by us have reported DNA methylation specifically in the promoter region.

Epigenetic studies on twins discordant for disease status are crucial in understanding the genetic basis of epigenetic differences observed in cross-sectional studies. Of the 5 studies included in our search, 3 did not have any common differentially methylated genes among them, while the other two studies that used blood cells as the source tissue had TXNIP as the common differentially methylated gene between them, with hypomethylation of TXNIP in diabetic blood samples observed in both these studies [ 17 , 21 ]. TXNIP is the only gene reported to be hypomethylated in diabetic blood in both case–control studies [ 55 ] and in twin studies [ 17 , 21 ] where the influence of underlying genetic factors is not masked. TXNIP has also been reported to be hypomethylated in diabetic pancreatic islets [ 55 ] and skeletal muscle [ 55 ], making it a potentially important causal gene in the pathophysiology of T2DM.

T2DM is known to be associated with other comorbidities such as obesity and cardiovascular complication. These comorbidities share some common risk factors like age, BMI and cholesterol content in blood. These risk factors are influenced by genes such as KCNQ1 , TCF7L2 and FTO [ 94 ]. Other systematic reviews have looked at epigenetic changes in obesity [ 95 ], aging [ 96 , 97 ] and cardiovascular complications [ 98 ]. Andrade et al . aimed to identify epigenetic changes in human adipose tissue from obese/overweight individuals with and without metabolic disorders like T2DM [ 95 ]. They also report differentially methylated genes that we have been reported in this review, such as KCNQ1 , FASN , MFSD1 , TXNIP , PPARG , IRS1 and TCF7L2 , from the same studies [ 95 ]. Krolevets et al . report that in addition to about 75,000 CpG sites and 19,000 genes, PTPRN2 was among the most frequently reported gene that was associated with cardiac disorders, although the direction of methylation is not specified [ 98 ]. Of the two studies that investigated DNA methylation in aging [ 96 , 97 ], no genes/CpG sites/studies were common with the ones mentioned in our review.

One of the most important factors in looking at T2DM as an epidemic is the geographic location of the site of reported data. With a large amount of data coming in from Europe alone, it is important to perform similar studies in other parts of the world and including various other ethnic groups to validate these reports and also help in mapping the genetic diversity to be able to tackle T2DM. India being the most populous country [ 99 ] with about 11% of Indians suffering from T2DM (in 2020) [ 100 ], it is imperative to study this population to uncover T2DM susceptible loci/genes. Of note, Chambers et al . have followed up London resident Indian Asians, for 8 years, and found that DNA methylation levels of TXNIP , PROC , C7orf29 , SREBF1 , PHOSPHO1 , SOCS3 and ABCG1 were positively associated with future T2DM incidence [ 71 ], but similar studies are lacking in Indians living in India, where exposure to pollution and availability and consumption of healthy diet are vastly different.

As for sex-specific methylation signatures of T2DM, differences were not seen between men and women except in genes HOOK2 [ 37 ] and MSI2 [ 27 ], which were hypermethylated in adipose tissue, and hypomethylated in blood, respectively . Finally, we searched if the genes which we found to be highly reported to be differentially methylated in human were also reported to be differentially methylated in animal models. KCNQ1 was reported to be hypomethylated in both T2DM human [ 58 ], and older mice model compared with younger mice [ 72 ] suggesting age-related methylation changes across species. In both humans [ 58 ], and mice fed with a high-fat diet, TCF7L2 was hypomethylated, and this DNA methylation change in mice was induced because of their diet [ 82 ], suggesting that nutrient consumption plays a role in epigenetic modification of genes involved in beta-cell function, and a healthy diet can have a protective role in maintaining homeostasis.

Although we did not look at clinical trials and candidate-gene studies that report differential DNA methylation, our review is an up-to-date report of epigenome-wide studies that includes prospective studies. We also report gene expression data in comparison with DNA methylation. Furthermore, a systematic report of differentially methylated gene/loci in tissues including blood cells, adipose tissue, pancreatic islet, skeletal muscles, liver and spermatozoa is included. While sex and ethnicity play a major role in pathology, we have tried to highlight these effects.

As with previous reviews, we emphasize the need for more prospective studies and replication of genome-wide association studies in different tissue types and populations.

From the 32 studies that report differentially methylated genes/loci between T2DM and normoglycemic individuals, ABCG1 (hypermethylated in blood), FTO (hypermethylated in blood and spermatozoa), KCNQ1 (hypermethylated in blood and hypomethylated in spermatozoa), TXNIP (hypomethylated in blood), PPARGC1A loci at chr4: 24,111,501–750 (hypermethylated in skeletal muscle and spermatozoa) and loci at chr4: 24,024,251–500 (hypomethylated in spermatozoa), PTPRN2 (hypermethylated in blood, hypomethylated in adipose tissue) were reported in more than one study. We found reports of hypermethylation of these genes that were associated with decreased gene expression, and vice versa. We also report findings from studies done on monozygotic twins. Various traits that can affect T2DM such as sex, glucose levels, BMI and ethnicity were also taken into consideration.

As there were multiple methods that were used to measure DNA methylation, internal and external validation of these results is also reported. Finally, animal model studies that have reported differential DNA methylation of the genes that were found to be differentially methylated in human studies were looked at to get an understanding of the likely mechanisms linking epigenetic dysregulation of these genes in T2DM to its pathophysiology.

Although the majority of the top differentially methylated genes are well known, other more recent genes reported here should be investigated further to understand their role in pathogenesis of T2DM.

Data availability statement

All relevant data are presented as tables and/or figures.

Abbreviations

ATP-Binding Cassette Subfamily G Member 1

Anoctamin 8

Beta 1,3-N-Acetylglucosaminyltransferase 7

Chromosome 1 Open Reading Frame 52

Chromosome 7 Open Reading Frame 29

Collagen Type XXI Alpha 1

Cytochrome B561 Family Member D1

CXXC Finger Protein 4

Alpha-Ketoglutarate Dependent Dioxygenase

Glucagon Like Peptide 1 Receptor

Glutathione Peroxidase 6

Hepatocyte Nuclear Factor 4 Alpha

Hook Microtubule Tethering Protein 2

Insulin-Like Growth Factor-Binding Protein 2

Interleukin-23 Subunit Alpha

Potassium Voltage-Gated Channel Subfamily Q Member 1

LINE1 Type Transposase Domain Containing 1

Lysyl Oxidase Homolog 2

Mucosa-Associated Lymphoid Tissue Lymphoma Translocation Protein 1

Major Facilitator Superfamily Domain Containing 1

Musashi RNA-Binding Protein 2

Platelet Derived Growth Factor Subunit A

Pancreatic and Duodenal Homeobox 1

Period Circadian Regulator 2

Phosphoethanolamine/Phosphocholine Phosphatase 1

Peroxisome Proliferator-Activated Receptor Gamma Coactivator 1-Alpha

Protein C, Inactivator Of Coagulation Factors Va And VIIIa

Polypyrimidine Tract-Binding Protein 1

Protein Tyrosine Phosphatase Receptor Type N2

Receptor Interacting Serine/Threonine Kinase 4

S100 Calcium-Binding Protein A4

Sterile Alpha Motif Domain Containing 12

Solute Carrier Family 1 Member 5

Solute Carrier Family 22 Member 1

Solute Carrier Family 22 Member 3

Solute Carrier Family 30 Member 8

Sterol Regulatory Element-Binding Transcription Factor 1

Suppressor Of Cytokine Signaling 3

Transcription Factor 7-Like 2

Thioredoxin-Interacting Protein

Cho NH, Shaw JE, Karuranga S, Huang Y, da Rocha Fernandes JD, Ohlrogge AW, et al. IDF diabetes atlas: global estimates of diabetes prevalence for 2017 and projections for 2045. Diabetes Res Clin Pract. 2018;138:271–81.

Article   CAS   PubMed   Google Scholar  

Prasad R, Groop L. Genetics of type 2 diabetes—pitfalls and possibilities. Genes. 2015;6(1):87–123.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Volkmar M, Dedeurwaerder S, Cunha DA, Ndlovu MN, Defrance M, Deplus R, et al. DNA methylation profiling identifies epigenetic dysregulation in pancreatic islets from type 2 diabetic patients. EMBO J. 2012;31(6):1405–26.

Jin B, Li Y, Robertson KD. DNA methylation: superior or subordinate in the epigenetic hierarchy? Genes Cancer. 2011;2(6):607–17.

Robertson KD. DNA methylation and human disease. Nat Rev Genet. 2005;6(8):597–610.

Moore LD, Le T, Fan G. DNA methylation and its basic function. Neuropsychopharmacology. 2013;38(1):23–38.

Hall E, Dayeh T, Kirkpatrick CL, Wollheim CB, Dekker Nitert M, Ling C. DNA methylation of the glucagon-like peptide 1 receptor (GLP1R) in human pancreatic islets. BMC Med Genet. 2013;14:76–76.

Willmer T, Johnson R, Louw J, Pheiffer C. Blood-based DNA methylation biomarkers for type 2 diabetes: potential for clinical applications. Front Endocrinol. 2018;4(9):744.

Article   Google Scholar  

Walaszczyk E, Luijten M, Spijkerman AMW, Bonder MJ, Lutgers HL, Snieder H, et al. DNA methylation markers associated with type 2 diabetes, fasting glucose and HbA1c levels: a systematic review and replication in a case–control sample of the Lifelines study. Diabetologia. 2018;61(2):354–68.

Muka T, Nano J, Voortman T, Braun KVE, Ligthart S, Stranges S, et al. The role of global and regional DNA methylation and histone modifications in glycemic traits and type 2 diabetes: a systematic review. Nutr Metab Cardiovasc Dis. 2016;26(7):553–66.

Rönn T, Ling C. DNA methylation as a diagnostic and therapeutic target in the battle against Type 2 diabetes. Epigenomics. 2015;7(3):451–60.

Article   PubMed   Google Scholar  

Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, Petticrew M, et al. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev. 2015;1(4):1.

Wells G, Shea B, O’Connell D, Peterson J, Welch, Losos M, et al. The Newcastle-Ottawa Scale (NOS) for Assessing the Quality of Nonrandomised Studies in Meta-Analyses. 2014. http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp

PROSPERO. [cited 2023 Apr 18]. https://www.crd.york.ac.uk/prospero/

Evangelista JE, Xie Z, Marino GB, Nguyen N, Clarke DJB, Ma’ayan A. Enrichr-KG: bridging enrichment analysis across multiple libraries. Nucl Acids Res. 2023;gkad393.

Soriano-Tárraga C, Jiménez-Conde J, Giralt-Steinhauer E, Mola-Caminal M, Vivanco-Hidalgo RM, Ois A, et al. Epigenome-wide association study identifies TXNIP gene associated with type 2 diabetes mellitus and sustained hyperglycemia. Hum Mol Genet. 2016;25(3):609–19.

Wang Z, Peng H, Gao W, Cao W, Lv J, Yu C, et al. Blood DNA methylation markers associated with type 2 diabetes, fasting glucose, and HbA1c levels: an epigenome-wide association study in 316 adult twin pairs. Genomics. 2021;113(6):4206–13.

Kriebel J, Herder C, Rathmann W, Wahl S, Kunze S, Molnos S, et al. Association between DNA methylation in whole blood and measures of glucose metabolism: KORA F4 study. PLoS ONE. 2016;11(3):e0152314–e0152314.

Article   PubMed   PubMed Central   Google Scholar  

Kulkarni H, Kos MZ, Neary J, Dyer TD, Kent JWJ, Göring HHH, et al. Novel epigenetic determinants of type 2 diabetes in Mexican-American families. Hum Mol Genet. 2015;24(18):5330–44.

Meeks KAC, Henneman P, Venema A, Addo J, Bahendeka S, Burr T, et al. Epigenome-wide association study in whole blood on type 2 diabetes among sub-Saharan African individuals: findings from the RODAM study. Int J Epidemiol. 2019;48(1):58–70.

Xiang Y, Wang Z, Hui Q, Gwinn M, Vaccarino V, Sun YV. DNA Methylation of TXNIP independently associated with inflammation and diabetes mellitus in twins. Twin Res Hum Genet. 2021;24(5):273–80.

Dawes K, Philibert W, Darbro B, Simons RL, Philibert R. Additive and interactive genetically contextual effects of HbA1c on cg19693031 methylation in type 2 diabetes. Genes (Basel). 2022;13(4):683.

Tsai HH, Shen CY, Ho CC, Hsu SY, Tantoh DM, Nfor ON, et al. Interaction between a diabetes-related methylation site (TXNIP cg19693031) and variant (GLUT1 rs841853) on fasting blood glucose levels among non-diabetics. J Transl Med. 2022;20(1):87.

Khamis A, Ning L, Balkau B, Bonnefond A, Canouil M, Roussel R, et al. Epigenetic changes associated with hyperglycaemia exposure in the longitudinal DESIR cohort. Diabetes Metab. 2022;48(4):101347.

Florath I, Butterbach K, Heiss J, Bewerunge-Hudler M, Zhang Y, Schöttker B, et al. Type 2 diabetes and leucocyte DNA methylation: an epigenome-wide association study in over 1,500 older adults. Diabetologia. 2016;59(1):130–8.

Kuroda A, Rauch TA, Todorov I, Ku HT, Al-Abdullah IH, Kandeel F, et al. Insulin gene expression is regulated by DNA methylation. PLoS ONE. 2009;4(9):6953.

Jeon JP, Koh IU, Choi NH, Kim BJ, Han BG, Lee S. Differential DNA methylation of MSI2 and its correlation with diabetic traits. PLoS ONE. 2017;12(5):e0177406–e0177406.

Dayeh T, Volkov P, Salö S, Hall E, Nilsson E, Olsson AH, et al. Genome-wide DNA methylation analysis of human pancreatic islets from type 2 diabetic and non-diabetic donors identifies candidate genes that influence insulin secretion. PLoS Genet. 2014;10(3):e1004160–e1004160.

Sjöholm LK, Ransome Y, Ekström TJ, Karlsson O. Evaluation of post-mortem effects on global brain DNA methylation and hydroxymethylation. Basic Clin Pharmacol Toxicol. 2018;122(2):208–13.

Vilahur N, Baccarelli AA, Bustamante M, Agramunt S, Byun HM, Fernandez MF, et al. Storage conditions and stability of global DNA methylation in placental tissue. Epigenomics. 2013;5(3):341–8.

Makki K, Froguel P, Wolowczuk I. Adipose tissue in obesity-related inflammation and insulin resistance: cells, cytokines, and chemokines. ISRN Inflammation. 2013;22(2013):1–12.

Dahlman I, Sinha I, Gao H, Brodin D, Thorell A, Rydén M, et al. The fat cell epigenetic signature in post-obese women is characterized by global hypomethylation and differential DNA methylation of adipogenesis genes. Int J Obes. 2015;39(6):910–9.

Article   CAS   Google Scholar  

Fujiki K, Shinoda A, Kano F, Sato R, Shirahige K, Murata M. PPARγ-induced PARylation promotes local DNA demethylation by production of 5-hydroxymethylcytosine. Nat Commun. 2013;4(1):2262.

Pfeiffer S, Krüger J, Maierhofer A, Böttcher Y, Klöting N, El Hajj N, et al. Hypoxia-inducible factor 3A gene expression and methylation in adipose tissue is related to adipose tissue dysfunction. Sci Rep. 2016;6(1):27969.

Wang X, Cao Q, Yu L, Shi H, Xue B, Shi H. Epigenetic regulation of macrophage polarization and inflammation by DNA methylation in obesity. JCI Insight. 2016 Nov 17 [cited 2023 Jun 17]; 1(19). https://insight.jci.org/articles/view/87748

Nilsson E, Jansson PA, Perfilyev A, Volkov P, Pedersen M, Svensson MK, et al. Altered DNA methylation and differential expression of genes influencing metabolism and inflammation in adipose tissue from subjects with type 2 diabetes. Diabetes. 2014;63(9):2962–76.

Rodríguez-Rodero S, Menéndez-Torre E, Fernández-Bayón G, Morales-Sánchez P, Sanz L, Turienzo E, et al. Altered intragenic DNA methylation of HOOK2 gene in adipose tissue from individuals with obesity and type 2 diabetes. PLoS ONE. 2017;12(12):e0189153–e0189153.

Wang C, Ha X, Li W, Xu P, Zhang Z, Wang T, et al. Comparative gene expression profile and DNA methylation status in diabetic patients of Kazak and Han people. Medicine. 2018;97(36):e11982–e11982.

Ribel-Madsen R, Fraga MF, Jacobsen S, Bork-Jensen J, Lara E, Calvanese V, et al. Genome-wide analysis of DNA methylation differences in muscle and fat from monozygotic twins discordant for type 2 diabetes. PLoS ONE. 2012;7(12):e51302–e51302.

Andersen E, Ingerslev LR, Fabre O, Donkin I, Altıntaş A, Versteyhe S, et al. Preadipocytes from obese humans with type 2 diabetes are epigenetically reprogrammed at genes controlling adipose tissue function. Int J Obes. 2019;43(2):306–18.

Strausberg RL, Feingold EA, Grouse LH, Derge JG, Klausner RD, Collins FS, et al. Generation and initial analysis of more than 15,000 full-length human and mouse cDNA sequences. Proc Natl Acad Sci U S A. 2002;99(26):16899–903.

Walenta JH, Didier AJ, Liu X, Krämer H. The golgi-associated Hook3 protein is a member of a novel family of microtubule-binding proteins. J Cell Biol. 2001;152(5):923–34.

Massa López D, Thelen M, Stahl F, Thiel C, Linhorst A, Sylvester M, et al. The lysosomal transporter MFSD1 is essential for liver homeostasis and critically depends on its accessory subunit GLMP. Elife. 2019;8:e50025.

Kashoor I, Batlle D. Proximal renal tubular acidosis with and without Fanconi syndrome. Kidney Res Clin Pract. 2019;38(3):267–81.

Yamagata K. Roles of HNF1α and HNF4α in Pancreatic β-Cells. In: Vitamins & Hormones [Internet]. Elsevier; 2014 [cited 2023 Aug 5]. pp. 407–23. https://linkinghub.elsevier.com/retrieve/pii/B9780128001745000168

Hayhurst GP, Lee YH, Lambert G, Ward JM, Gonzalez FJ. Hepatocyte nuclear factor 4α (Nuclear Receptor 2A1) is essential for maintenance of hepatic gene expression and lipid homeostasis. Mol Cell Biol. 2001;21(4):1393–403.

Stoffel M, Duncan SA. The maturity-onset diabetes of the young (MODY1) transcription factor HNF4α regulates expression of genes required for glucose transport and metabolism. Proc Natl Acad Sci USA. 1997;94(24):13209–14.

Vedi M, Smith JR, Thomas Hayman G, Tutaj M, Brodie KC, De Pons JL, et al. 2022 updates to the rat genome database: a findable, accessible, interoperable, and reusable (FAIR) resource. Genetics. 2023;224(1):1042.

Han HS, Kang G, Kim JS, Choi BH, Koo SH. Regulation of glucose metabolism from a liver-centric perspective. Exp Mol Med. 2016;48(3):e218–e218.

Abderrahmani A, Yengo L, Caiazzo R, Canouil M, Cauchi S, Raverdy V, et al. Increased hepatic PDGF-AA signaling mediates liver insulin resistance in obesity-associated type 2 diabetes. Diabetes. 2018;67(7):1310–21.

Nilsson E, Matte A, Perfilyev A, de Mello VD, Käkelä P, Pihlajamäki J, et al. Epigenetic alterations in human liver from subjects with type 2 diabetes in parallel with reduced folate levels. J Clin Endocrinol Metab. 2015;100(11):E1491-1501.

Barres R, Osler ME, Yan J, Rune A, Fritz T, Caidahl K, et al. Non-CpG methylation of the PGC-1alpha promoter through DNMT3B controls mitochondrial density. Cell Metab. 2009;10(3):189–98.

Yuan W, Xia Y, Bell CG, Yet I, Ferreira T, Ward KJ, et al. An integrated epigenomic analysis for type 2 diabetes susceptibility loci in monozygotic twins. Nat Commun. 2014;5:5719–5719.

Kiechl S, Wittmann J, Giaccari A, Knoflach M, Willeit P, Bozec A, et al. Blockade of receptor activator of nuclear factor-κB (RANKL) signaling improves hepatic insulin resistance and prevents development of diabetes mellitus. Nat Med. 2013;19(3):358–63.

Dayeh T, Tuomi T, Almgren P, Perfilyev A, Jansson PA, De Mello VD, et al. DNA methylation of loci within ABCG1 and PHOSPHO1 in blood DNA is associated with future type 2 diabetes risk. Epigenetics. 2016;11(7):482–8.

Chen YT, Lin WD, Liao WL, Tsai YC, Liao JW, Tsai FJ. NT5C2 methylation regulatory interplay between DNMT1 and insulin receptor in type 2 diabetes. Sci Rep. 2020;10(1):16087.

Chen X, Lin Q, Wen J, Lin W, Liang J, Huang H, et al. Whole genome bisulfite sequencing of human spermatozoa reveals differentially methylated patterns from type 2 diabetic patients. Journal of diabetes investigation. 2019

Toperoff G, Aran D, Kark JD, Rosenberg M, Dubnikov T, Nissan B, et al. Genome-wide survey reveals predisposing diabetes type 2-related DNA methylation variations in human peripheral blood. Hum Mol Genet. 2012;21(2):371–83.

Volkov P, Bacos K, Ofori JK, Esguerra JLS, Eliasson L, Rönn T, et al. Whole-genome bisulfite sequencing of human pancreatic islets reveals novel differentially methylated regions in type 2 diabetes pathogenesis. Diabetes. 2017;66(4):1074–85.

García-Calzón S, Perfilyev A, Männistö V, de Mello VD, Nilsson E, Pihlajamäki J, et al. Diabetes medication associates with DNA methylation of metformin transporter genes in the human liver. Clin Epigenetics. 2017;9:102–102.

Zampieri M, Bacalini MG, Barchetta I, Scalea S, Cimini FA, Bertoccini L, et al. Increased PARylation impacts the DNA methylation process in type 2 diabetes mellitus. Clin Epigenet. 2021;13(1):114.

Qie R, Chen Q, Wang T, Chen X, Wang J, Cheng R, et al. Association of ABCG1 gene methylation and its dynamic change status with incident type 2 diabetes mellitus: the Rural Chinese Cohort Study. J Hum Genet. 2021;66(4):347–57.

Krause C, Sievert H, Geißler C, Grohs M, El Gammal AT, Wolter S, et al. Critical evaluation of the DNA-methylation markers ABCG1 and SREBF1 for Type 2 diabetes stratification. Epigenomics. 2019;11(8):885–97.

van Otterdijk SD, Binder AM, Szarc Vel Szic K, Schwald J, Michels KB. DNA methylation of candidate genes in peripheral blood from patients with type 2 diabetes or the metabolic syndrome. PLoS ONE. 2017;12(7):e0180955–e0180955.

Toperoff G, Kark JD, Aran D, Nassar H, Ahmad WA, Sinnreich R, et al. Premature aging of leukocyte DNA methylation is associated with type 2 diabetes prevalence. Clin Epigenetics. 2015;7(1):35–35.

Huang S, Qin P, Chen Q, Zhang D, Cheng C, Guo C, et al. Association of FTO gene methylation with incident type 2 diabetes mellitus: A nested case–control study. Gene. 2021;786: 145585.

Zhang D, Cheng C, Cao M, Wang T, Chen X, Zhao Y, et al. TXNIP hypomethylation and its interaction with obesity and hypertriglyceridemia increase type 2 diabetes mellitus risk: a nested case-control study. J Diabetes. 2020;12(7):512–20.

Hu F, Zhang Y, Qin P, Zhao Y, Liu D, Zhou Q, et al. Integrated analysis of probability of type 2 diabetes mellitus with polymorphisms and methylation of KCNQ1 gene: a nested case-control study. J Diabetes. 2021;13(12):975–86.

Ling C, Del Guerra S, Lupi R, Rönn T, Granhall C, Luthman H, et al. Epigenetic regulation of PPARGC1A in human type 2 diabetic islets and effect on insulin secretion. Diabetologia. 2008;51(4):615–22.

Wittenbecher C, Ouni M, Kuxhaus O, Jähnert M, Gottmann P, Teichmann A, et al. Insulin-like growth factor binding protein 2 (IGFBP-2) and the risk of developing type 2 diabetes. Diabetes. 2019;68(1):188–97.

Chambers JC, Loh M, Lehne B, Drong A, Kriebel J, Motta V, et al. Epigenome-wide association of DNA methylation markers in peripheral blood from Indian Asians and Europeans with incident type 2 diabetes: a nested case-control study. Lancet Diabetes Endocrinol. 2015;3(7):526–34.

Sandovici I, Hammerle CM, Cooper WN, Smith NH, Tarry-Adkins JL, Dunmore BJ, et al. Ageing is associated with molecular signatures of inflammation and type 2 diabetes in rat pancreatic islets. Diabetologia. 2016;59(3):502–11.

Lee YH, Kang ES, Kim SH, Han SJ, Kim CH, Kim HJ, et al. Association between polymorphisms in SLC30A8, HHEX, CDKN2A/B, IGF2BP2, FTO, WFS1, CDKAL1, KCNQ1 and type 2 diabetes in the Korean population. J Hum Genet. 2008;53(11–12):991–8.

Liu Y, Zhou DZ, Zhang D, Chen Z, Zhao T, Zhang Z, et al. Variants in KCNQ1 are associated with susceptibility to type 2 diabetes in the population of mainland China. Diabetologia. 2009;52(7):1315–21.

Hu C, Wang C, Zhang R, Ma X, Wang J, Lu J, et al. Variations in KCNQ1 are associated with type 2 diabetes and beta cell function in a Chinese population. Diabetologia. 2009;52(7):1322–5.

Jonsson A, Isomaa B, Tuomi T, Taneera J, Salehi A, Nilsson P, et al. A variant in the KCNQ1 gene predicts future type 2 diabetes and mediates impaired insulin secretion. Diabetes. 2009;58(10):2409–13.

Yasuda K, Miyake K, Horikawa Y, Hara K, Osawa H, Furuta H, et al. Variants in KCNQ1 are associated with susceptibility to type 2 diabetes mellitus. Nat Genet. 2008;40(9):1092–7.

Unoki H, Takahashi A, Kawaguchi T, Hara K, Horikoshi M, Andersen G, et al. SNPs in KCNQ1 are associated with susceptibility to type 2 diabetes in East Asian and European populations. Nat Genet. 2008;40(9):1098–102.

Müssig K, Staiger H, Machicao F, Kirchhoff K, Guthoff M, Schäfer SA, et al. Association of Type 2 diabetes candidate polymorphisms in KCNQ1 with incretin and insulin secretion. Diabetes. 2009;58(7):1715–20.

Qian Y, Dong M, Lu F, Li H, Jin G, Hu Z, et al. Joint effect of CENTD2 and KCNQ1 polymorphisms on the risk of type 2 diabetes mellitus among Chinese Han population. Mol Cell Endocrinol. 2015;407:46–51.

Been LF, Ralhan S, Wander GS, Mehra NK, Singh J, Mulvihill JJ, et al. Variants in KCNQ1 increase type II diabetes susceptibility in South Asians: a study of 3,310 subjects from India and the US. BMC Med Genet. 2011;12(1):18.

Hu Y, Shi P, He K, Zhu YQ, Yang F, Yang M, et al. Methylation of Tcf712 promoter by high-fat diet impairs β-cell function in mouse pancreatic islets. Diabetes Metab Res Rev. 2018;34(4): e2980.

Flurkey K, Currer JM, Harrison D. Mouse models in aging research. In: The mouse in biomedical research. Elsevier; 2007, pp. 637–72.

Papadopoulou S, Edlund H. Attenuated Wnt signaling perturbs pancreatic growth but not pancreatic function. Diabetes. 2005;54(10):2844–51.

Kumar A, Mittal R. Mapping Txnip: key connexions in progression of diabetic nephropathy. Pharmacol Rep. 2018;70(3):614–22.

Minn AH, Hafele C, Shalev A. Thioredoxin-interacting protein is stimulated by glucose through a carbohydrate response element and induces β-cell apoptosis. Endocrinology. 2005;146(5):2397–405.

Tarling EJ, Edwards PA. ATP binding cassette transporter G1 (ABCG1) is an intracellular sterol transporter. Proc Natl Acad Sci USA. 2011;108(49):19719–24.

Westerterp M, Bochem AE, Yvan-Charvet L, Murphy AJ, Wang N, Tall AR. ATP-binding cassette transporters, atherosclerosis, and inflammation. Circ Res. 2014;114(1):157–70.

Wilson PW, McGEE DL, Kannel WB. Obesity, very low density lipoproteins, and glucose intolerance over fourteen years: The Framingham Study. Am J Epidemiol. 1981;114(5):697–704.

Shah UJ, Xie W, Flyvbjerg A, Nolan JJ, Højlund K, Walker M, et al. Differential methylation of the type 2 diabetes susceptibility locus KCNQ1 is associated with insulin sensitivity and is predicted by CpG site specific genetic variation. Diabetes Res Clin Pract. 2019;148:189–99.

Mattei AL, Bailly N, Meissner A. DNA methylation: a historical perspective. Trends Genet. 2022;38(7):676–707.

Anastasiadi D, Esteve-Codina A, Piferrer F. Consistent inverse correlation between DNA methylation of the first intron and gene expression across tissues and species. Epigenetics Chromatin. 2018;11(1):37.

Ball MP, Li JB, Gao Y, Lee JH, LeProust EM, Park IH, et al. Targeted and genome-scale strategies reveal gene-body methylation signatures in human cells. Nat Biotechnol. 2009;27(4):361–8.

Berumen J, Orozco L, Betancourt-Cravioto M, Gallardo H, Zulueta M, Mendizabal L, et al. Influence of obesity, parental history of diabetes, and genes in type 2 diabetes: a case-control study. Sci Rep. 2019;9(1):2748.

Andrade S, Morais T, Sandovici I, Seabra AL, Constância M, Monteiro MP. Adipose tissue epigenetic profile in obesity-related dysglycemia: a systematic review. Front Endocrinol. 2021;29(12): 681649.

Ryan J, Wrigglesworth J, Loong J, Fransquet PD, Woods RL. A systematic review and meta-analysis of environmental, lifestyle, and health factors associated with DNA methylation age. J Gerontol Ser A. 2020;75(3):481–94.

Oblak L, Van Der Zaag J, Higgins-Chen AT, Levine ME, Boks MP. A systematic review of biological, social and environmental factors associated with epigenetic clock acceleration. Ageing Res Rev. 2021;69: 101348.

Krolevets M, Cate VT, Prochaska JH, Schulz A, Rapp S, Tenzer S, et al. DNA methylation and cardiovascular disease in humans: a systematic review and database of known CpG methylation sites. Clin Epigenet. 2023;15(1):56.

Department of Economic and Social Affairs UN. India to overtake China as world’s most populous country in April 2023, United Nations projects. [cited 2023 Jun 8]. https://www.un.org/en/desa/india-overtake-china-world-most-populous-country-april-2023-united-nations-projects

Anjana RM, Unnikrishnan R, Deepa M, Pradeepa R, Tandon N, Das AK, et al. Metabolic non-communicable disease health report of India: the ICMR-INDIAB national cross-sectional study (ICMR-INDIAB-17). Lancet Diabetes Endocrinol. 2023;11(7):474–89.

Hwang JY, Lee HJ, Go MJ, Jang HB, Choi NH, Bae JB, et al. Genome-wide methylation analysis identifies ELOVL5 as an epigenetic biomarker for the risk of type 2 diabetes mellitus. Sci Rep. 2018;8(1):14862–14862.

Zou L, Yan S, Guan X, Pan Y, Qu X. Hypermethylation of the PRKCZ Gene in Type 2 diabetes mellitus. J Diabetes Res. 2013;2013:721493–721493.

Chen YT, Liao JW, Tsai YC, Tsai FJ. Inhibition of DNA methyltransferase 1 increases nuclear receptor subfamily 4 group A member 1 expression and decreases blood glucose in type 2 diabetes. Oncotarget. 2016;7(26):39162–70.

Davegårdh C, Säll J, Benrick A, Broholm C, Volkov P, Perfilyev A, et al. VPS39-deficiency observed in type 2 diabetes impairs muscle stem cell differentiation via altered autophagy and epigenetics. Nat Commun. 2021;12(1):2431.

Whytock KL, Pino MF, Sun Y, Yu G, De Carvalho FG, Yeo RX, et al. Comprehensive interrogation of human skeletal muscle reveals a dissociation between insulin resistance and mitochondrial capacity. Am J Physiol-Endocrinol Metabol. 2023;325(4):E291-302.

Download references

Acknowledgements

AM is supported by the Wellcome Trust/DBT India Alliance Fellowship [Grant Number IA/CPHI/19/1/504593]. We thank Ms. Ramya for her insightful comments.

This research did not receive any specific grant from any funding agency in the public, commercial or not-for-profit sector.

Author information

Priyanka Chinya Nataraj

Present address: Vedantu, Bangalore, India

Authors and Affiliations

Research Scholar, Manipal Academy of Higher Education, Manipal, India

Nikhil Nadiger

Division of Nutrition, St. John’s Research Institute, St. John’s Medical College, St Johns National Academy of Health Sciences, Sarjapura Road, Koramangala, Bangalore, 560034, India

Nikhil Nadiger, Jyothisha Kana Veed, Priyanka Chinya Nataraj & Arpita Mukhopadhyay

You can also search for this author in PubMed   Google Scholar

Contributions

AM was involved in conceptualization, review and editing of the manuscript. NN, JKV and PCN were involved in data extraction, formal analysis, investigation and writing the manuscript. All authors have read and approved the final manuscript.

Corresponding author

Correspondence to Arpita Mukhopadhyay .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Ethics approval and consent to participate

Not applicable.

Consent for publication

Conflict of interest.

The authors declare no conflict of interest.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1 search strategy for the systematic review of dna methylation association with t2dm, 13148_2024_1670_moesm2_esm.docx.

Additional file 2 Qualitative assessment of research articles included in the review based on the New Castle Ottawa Scale (NOS)

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Nadiger, N., Veed, J.K., Chinya Nataraj, P. et al. DNA methylation and type 2 diabetes: a systematic review. Clin Epigenet 16 , 67 (2024). https://doi.org/10.1186/s13148-024-01670-6

Download citation

Received : 20 January 2024

Accepted : 11 April 2024

Published : 16 May 2024

DOI : https://doi.org/10.1186/s13148-024-01670-6

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Type 2 diabetes
• DNA methylation
• Epigenome-wide association studies
• Epigenetics

Clinical Epigenetics

ISSN: 1868-7083

• Submission enquiries: [email protected]
• General enquiries: [email protected]

• Open access
• Published: 20 May 2024

HbA1c changes in a deprived population who followed or not a diabetes self-management programme, organised in a multi-professional primary care practice: a historical cohort study on 207 patients between 2017 and 2019

• Sarah Ajrouche 1 ,
• Lisa Louis 1 ,
• Maxime Esvan 2 ,
• Anthony Chapron 1 , 2 ,
• Ronan Garlantezec 3 &
• Emmanuel Allory 1 , 2 , 4  

BMC Endocrine Disorders volume  24 , Article number:  72 ( 2024 ) Cite this article

155 Accesses

Metrics details

Diabetes self-management (DSM) helps people with diabetes to become actors in their disease. Deprived populations are particularly affected by diabetes and are less likely to have access to these programmes. DSM implementation in primary care, particularly in a multi-professional primary care practice (MPCP), is a valuable strategy to promote care access for these populations. In Rennes (Western France), a DSM programme was designed by a MPCP in a socio-economically deprived area. The study objective was to compare diabetes control in people who followed or not this DSM programme.

The historical cohort of patients who participated in the DSM programme at the MPCP between 2017 and 2019 ( n  = 69) was compared with patients who did not participate in the programme, matched on sex, age, diabetes type and place of the general practitioner’s practice ( n  = 138). The primary outcome was glycated haemoglobin (HbA1c) change between 12 months before and 12 months after the DSM programme. Secondary outcomes included modifications in diabetes treatment, body mass index, blood pressure, dyslipidaemia, presence of microalbuminuria, and diabetes retinopathy screening participation.

HbA1c was significantly improved in the exposed group after the programme ( p  < 0.01). The analysis did not find any significant between-group difference in socio-demographic data, medical history, comorbidities, and treatment adaptation.

Conclusions

These results, consistent with the international literature, promote the development of DSM programmes in primary care settings in deprived areas. The results of this real-life study need to be confirmed on the long-term and in different contexts (rural area, healthcare organisation).

Peer Review reports

Introduction

Diabetes is a chronic disease that has doubled in prevalence in the last three decades [ 1 ] and is now one of the ten first causes of death worldwide [ 2 ]. Currently, 463 million people have diabetes worldwide (4.5 million in France) and this number could rise to 700 million by 2045 [ 3 ]. Diabetes incidence has increased dramatically, particularly that of type 2 diabetes mellitus that accounts for 90% of all cases [ 4 ]. Diabetes is associated with high morbidity index and altered quality of life [ 5 , 6 , 7 ]. Its prevalence has particularly increased in low-income and disadvantaged socio-economic groups [ 8 , 9 , 10 ], and even more in developed countries [ 11 ]. Its prevalence was twice as high in people receiving universal health coverage (UHC) [ 8 ] in whom it was also associated with worse glycaemic control [ 12 ] and more complications [ 13 , 14 , 15 ]. Higher diabetes prevalence was also found in some immigrant populations. For example, in metropolitan France, the risk of diabetes is 2.5 times higher in women who came from a North African country than in non-immigrant women [ 9 ]. Therefore, the population’s contextual and cultural characteristics need to be considered when developing preventive actions, such as Diabetes Self-Management (DSM) programmes [ 16 , 17 ].

DSM education brings together the knowledge and skills that make people more aware about their health and their health choices by offering specific training, support and coaching [ 18 ]. DSM education enables people with diabetes to acquire and maintain skills to manage diabetes, resulting in quality of life improvement, increasing active role with the healthcare providers (HCP), and better adherence to treatment/follow-up and prevention of complications [ 19 , 20 ]. The objective of DSM education is to make patients more autonomous and to produce a complementary effect to the usual pharmacological interventions [ 19 ]. It is an ongoing process, adapted to the disease course and the patient's lifestyle [ 21 ]. Although their effectiveness is acknowledged, particularly for type 2 diabetes mellitus [ 22 , 23 , 24 ], participation in DSM programmes in group settings is still limited among people with diabetes [ 25 ], especially in deprived populations. This difficult access is partly explained by their living conditions and socio-cultural background that complicate access to programmes and the will to change lifestyle habits [ 26 ]. Another explanation is that the current DSM programmes were not developed by taking into account the social and cultural background of the targeted populations [ 27 ].

The accessibility issues to DSM programmes and the obstacles to DSM practice are a major research topic [ 28 , 29 ]. Furthermore, the fact that DSM education is mostly organised in hospitals [ 30 , 31 ] may constitute an additional obstacle [ 32 ]. In 2014, in France, only 3.9% of self-management programmes were run in primary care settings, compared with 82% in a hospital structure [ 18 ]. Primary care now appears to be the preferred place for promoting access to care and reducing social inequalities in health [ 27 ]. Multi-professional Primary Care Practices (MPCP) bring together medical/paramedical professionals and social services around a common health project to improve inter-professional collaboration and access to care for the population [ 33 ]. Therefore, they seem suitable places for developing prevention programmes due to their accessibility based on their geographical position, relational proximity with the habitants, better cultural knowledge by the HCP and capacity to break down social isolation [ 34 ]. MPCPs are an opportunity to integrate DSM education in primary care and they could become reference structures in this field [ 35 , 36 ].

In Rennes, the Villejean district is one of the five socio-economically deprived areas of the city. The median income is estimated at 670 euros (vs 1628 euros in the whole city), 38.3% of the population is unemployed, and 51% of < 20-year-old people receive UHC [ 37 ]. In 2015, 71 HCPs of this district decided to create the "Rennes Nord-Ouest" MPCP and developed a collective DSM programme for their patients with diabetes (supplementary files 1 and 2). In accordance with the recommendations, DSM programmes must be evaluated [ 18 ]. The value of this programme was initially demonstrated from the users’ point of view [ 34 ]. This qualitative study in 2020 also showed that in the first year of the DSM programme, participants were from nine different countries and 80% were considered as socio-economically deprived. This assessment must be continued by including quantitative biomedical parameters, as described in the international literature [ 38 ]. In Europe, several randomised controlled trials have demonstrated the benefit of group DSM for improving glycaemic control in non-deprived populations, such as the X-PERT study [ 39 ] and the DESMOND study [ 40 ]. In the United States, two randomised control trials carried out by community health workers in clinics found a significative effect of DSM programmes among socially deprived immigrant people with diabetes [ 41 , 42 ]. However, we did not find any study on similar interventions for deprived people carried out in MPCPs.

The main objective of this study in a socio-economically deprived area was to compare diabetes control in a group that participated in a DSM programme run by an MPCP and in a group that did not receive this intervention.

Study design

This was an historical exposed/non-exposed cohort study to assess the effect of a DSM intervention in primary care, carried out by a MPCP located in a socio-economically deprived area of Rennes, France.

Description of the intervention

The programme targeted ≥ 18-year-old people with diabetes to improve or develop self-care skills and change their eating habits. The DSM programme was designed and implemented by the "Rennes Nord-Ouest" MPCP, in the Villejean district, Rennes, France, in 2017. Patients were included in the programme upon suggestion by one of the MPCP HCPs involved in their care (e.g. general practitioner (GP), nurse, pharmacist, chiropodist), even if their own GP was not working at the MPCP. HCP of the MPCP recruited participants during their usual consultations. Refusal to participate was not recorded. Only interested patients had a BEPI (Bilan educatif partagé initial, patient-centred educational assessment) (supplementary file 3) with a HCP of the team before the DSM programme start to fix personal objectives that were used to prepare a personalized attendance programme to the different workshops.The programme consisted of seven to nine workshops that lasted 1–2 h and were held on weekdays between 9am and 5pm over a period of 1–2 months. The MPCP received annual funding from the local health authority (Agence régionale de santé) to cover the intervention running costs, and the training and remuneration of the involved HCPs.

Exposed and non-exposed groups

The exposed group (receiving the intervention) included ≥ 18-year-old patients with type 1 or type 2 diabetes who were followed by at least one HCP in the MPCP and who participated in the DSM programme between 2017 and 2020. All the 75 patients who participated in the programme (at least BEPI completion) were eligible. If some had participated in more than one annual session, only their first participation was considered.

The non-exposed group included all the patients selected from the SOPHIA database of the GPs whose patients were in the exposed group. SOPHIA is a free diabetes support service set up by the French public health insurance in 2008 to offer remote coaching (emails, personal online space, and telephone follow-up with a nurse) adapted to the needs of people with diabetes in order to help them live better with their disease. This service was offered to all patients at the MPCP (i.e. people in the exposed and non-exposed groups). The SOPHIA database includes ≥ 18-year-old patients with type 1 and 2 diabetes who are registered with a GP, have long duration disease (LDD) status for diabetes, are affiliated to the public health insurance, and had at least three prescriptions for anti-diabetic drugs in the year of the intervention.

Each patient in the exposed group was randomly matched to two control patients based on sex (male or female), diabetes type (type 1 or type 2), year of birth (before 1960 or after; median calculated in the exposed group) and whether their GP was a MPCP member. The intervention date was the BEPI date.

The exclusion criteria for the exposed and non-exposed groups were: GP’s or patient’s refusal to participate in the study, patients unable to read and write in French, lack of follow-up during the study period (patient arrived at the practice after the intervention date, or left before), haemoglobinopathy that does not allow HbA1c monitoring, gestational diabetes, and drug-induced diabetes.

Study endpoints

The primary outcome was glycated haemoglobin change (HbA1c in %) between 12 months before and 12 months after the intervention start date (i.e. the BEPI date).

Secondary outcomes were modifications in diabetes treatment, body mass index (BMI; in kg/m2), systolic and diastolic blood pressure (in mmHg), lipid profile (low density lipoprotein C, LDLc, in mmol/L), microalbuminuria, and screening for diabetic retinopathy between before and after the intervention.

Data collection

Data were collected by two residents in general practice in 11 practices (21 GPs who followed the participants) after the intervention, between March and December 2021. Data were extracted from computerised medical records (consultations with clinical examination, laboratory work-up results, and specialist letters) from the practice professional software. Data were collected for the years 2017 to 2020, and as close as possible to the target dates (12 months before and 12 months after the intervention) to obtain at least two distinct values, particularly in terms of kidney function, lipid levels and microalbuminuria.

To characterise the two groups, each patient’s socio-demographic data (year of birth, sex, profession, education level, and socio-professional categories) and medical history (diabetes type and duration, other associated LDD) were collected. Concerning chronic treatment, prescriptions close to the target dates were identified to determine the diabetes treatments (metformin, other oral drugs, GLP-1 analogues, or insulin). Prescriptions for statins, angiotensin converting enzyme inhibitors, or related drugs were also retained.

Lastly, mentions of ophthalmological consultations (specialist’s letters or key words) were searched in the different consultations within the study interval.

Statistical analysis

Patient characteristics were expressed as n (%) for categorical variables and mean ± standard deviation (SD) for continuous variables. For univariate comparison between (exposed and non-exposed) groups, the Student’s t or Mann–Whitney-Wilcoxon’s test was used for continuous variables and the χ2 or Fisher’s exact test for categorical variables.

Outcome changes over time were analysed using generalised linear mixed models. A sensitivity analysis was performed for the primary outcome using a model adjusted for sex, age, BMI, and education level. Multiple imputation was used to account for missing values. Fifty imputed datasets were created and combined using standard between/within-variance techniques. Statistical analyses were computed at the two-sided α level of 5% with SAS version 9.4 (SAS Institute, Cary, North Carolina, USA).

Ethical aspects and legislation

This study was approved by the Rennes University Hospital ethics committee on 14 June 2021 (Number 21.77–2, supplementary file 5). It complied with the reference methodology MR-004 defined by the French committee on personal data protection (Commission Nationale Informatique et Libertés; CNIL) and with the European General Data Protection Regulation (GDPR).

Among the 75 patients who completed a BEPI between 2017 and 2019, 24 GP’s were identified. Three GP’s refused to participate; each of them had one patient who had the BEPI. As three other patients with a BEPI refused to participate to the study, the exposed group was composed of 69 patients (Fig.  1 ). In the SOPHIA database, 488/560 patients followed by the GPs of the patients in the exposed group did not participate in the intervention. Therefore, a participation rate of 13% to the DSM programme could be estimated. Among them, 149 were selected by random 2:1 matching. After excluding 11 patients, 138 patients were included in the non-exposed group. With the 69 patients of the exposed group, 207 patients were included in the study.

Description of the study population (Table  1 )

The analysis did not find any significant difference between groups concerning socio-demographic characteristics, age at diabetes diagnosis [49 (± 12) years for the exposed group and 49 (± 13) years for the non-exposed group], and percentage of patients with diabetes discovered < 1 year before the intervention date [ n  = 13 (19.1%) for the exposed group and n  = 22 (17.3%) for the non-exposed group]. Education level and percentage of retired patients [ n  = 29 (42%) for the exposed group and n  = 43 (37.7%) for the non-exposed group] were comparable between groups. Presence of another known LDD [ n  = 29 (42%) in the exposed group and n  = 57 (41.3%) in the non-exposed group], mean number of LDDs per patient and their nature, and comorbidities (hypertension, dyslipidaemia, known diabetic nephropathy, known diabetic retinopathy or obesity) were not significantly different between groups.

Pre-intervention data (Table  2 )

Pre-intervention weight, BMI and blood pressure were not significantly different between groups. Among treatments, only prescription of GLP-1 analogues was higher in the exposed group than non-exposed group [ n  = 12 (17.6%) vs n  = 6 (4.3%); p  = 0.01]. Among laboratory data, the mean HbA1c level was significantly higher in the exposed than non-exposed group [8.3% ± 2.2 vs 7.1% ± 1.2; p  < 0.01], and more patients had nephropathy with microalbuminuria in the exposed than non-exposed group [ n  = 19 (33.9%) vs n  = 17 (17.9%); p  = 0.02]. Adherence to the annual ophthalmological follow-up was higher in the exposed than non-exposed group [ n  = 39 (72.2%) vs n  = 48 (44.4%); p  < 0.01].

Post-intervention changes (Table  3 , Fig.  2 )

HbA1c (%) change over time (24 months) in the exposed and non-exposed groups

After the intervention, the mean HbA1c decreased by 0.73% [-1.13; -0.33] in the exposed group and increased by 0.35% [0.07; 0.63] in the non-exposed group ( p  < 0.01) (primary endpoint). All the secondary endpoints were similar between groups (supplementary file 6). In the secondary analyses, HbA1c change difference in the two groups after exclusion of patients with type 1 diabetes was still significant ( p  < 0.01) and remained also after the sensitivity analysis adjusted for sex, age, BMI and education level ( p  < 0.01).

The main result of our study is the significant difference in HbA1c change ( p  < 0.01) between the exposed group and the non-exposed group at 12 months post-intervention (i.e. DSM programme). This result is consistent with the literature. The systematic review by Odgers-Jewell et al. found that DSM education in groups efficiently reduced HbA1c by 0.3% at 12 months and up to 36 months [ 38 ]. Like in our study, there was no significant difference in BMI, blood pressure and LDLc change between exposed and non-exposed groups during the same period. The TIME randomised controlled trial on the long-term effectiveness of a programme for low-income populations in Houston community clinics found improvements in HbA1c at 12, 18 and even 24 months post-intervention [ 43 ]. Compared with the exposed group, HbA1c level in the non-exposed group (conventional medical follow-up) worsened. Similarly, the randomised controlled trial by Trento et al. [ 44 ] showed a progressive increase over 5 years in the HbA1c of controls compared with individuals receiving group DSM education in a hospital. In our study, the pre-intervention HbA1c and microalbuminuria were significantly higher in the intervention group, suggesting that patients who participated in the programme had more unbalanced and complicated diabetes. Hadjiconstantinou et al. found that patients with higher HbA1c (> 7%) benefit more from DSM programmes, as observed for our participants [ 29 ]. In this perspective article, the authors stressed that better outcomes were observed in groups that included participants with higher baseline HbA1c, younger age (< 65 years), and a higher proportion of ethnic minorities, like in our population. The lack of significant between-group difference in HbA1c and microalbuminuria after the intervention (supplementary file 6), combined with the analysis of variance for HbA1c, may indicate that the DSM intervention has a catch-up effect between groups, bringing both populations to same level. Indeed, while HbA1c decreased by 0.73% [-1.13; -0.33] in the exposed group, it increased by 0.35% [0.07; 0.63] in the non-exposed group ( p  < 0.01). Insulin prescription alone cannot explain this result because changes in insulin prescription were similar between groups ( p  = 0.54) and the HbA1c change difference remained also after the subgroup analysis adjusted for insulin prescription ( p  < 0.01). One hypothesis to be considered is that HCPs might have preferentially proposed the DSM programme to patients with badly controlled diabetes, although this was not an objective of the programme. In an interdisciplinary literature review, Carey et al. suggested the concept of " proportionate universalism " according to which health actions should be universal, but with a scale and intensity proportionate to the patients’ disadvantage level [ 45 ]. " Proportionate universalism " would be a way to move towards more equity in health by rebalancing situations without stigmatising population groups. Continuity of care in general practice allows practitioners to reduce social inequalities in health. Gray et al., in a systematic review of observational studies between 1996 and 2017, highlighted that increased continuity of care by doctors is associated with lower mortality rate in their patients [ 46 ]. Similarly, Sandvik et al. described the GP’s contribution to the life expectancy of their patients through the implementation of informal (access to all the patient's information), longitudinal (transcending the various disease episodes), and interpersonal (the relationship of trust established between patient and GP) continuity [ 47 ].

Another important finding in our study was the significant higher adherence to the ophthalmological follow-up in the exposed group than in the non-exposed group (72.2% versus 44% before the intervention and 72% versus 38.1% after the intervention). This may be explained by a closer follow-up of patients in the exposed group by their GP/other HCPs. However, this does not seem to have had an effect on baseline HbA1c that was higher in the exposed group. Additionally, our exposed group may have had a lower level of health literacy (i.e. the set of individual and environmental conditions for a patient to understand and process health information) [ 48 ]. This could explain why the GP better followed these patients and, for instance, might have been more likely to ask the secretary of the practice to organise an appointment with the specialist rather than delegating this task directly to the patient. According to the French national health council (Haut Conseil de la Santé Publique; HCSP), " people with low literacy level are 1.5 to 3 times more likely to be in unfavourable health conditions than people with higher literacy level " [ 27 ]. This could explain the initial difference in HbA1c level between groups. A qualitative study on the health literacy level of participants in a DSM programme in a socio-economically deprived area of Montpellier (south of France) highlighted the diversity of health literacy profiles that coexisted in that area [ 49 ]. Moreover, low health literacy is more likely to be observed among people with low income, belonging to ethnic minorities, or migrant populations [ 27 ]. Our exposed group included mainly patients from a practice in an area with elevated socio-economic difficulties and consequently people with more precarious profiles.

Strengths and limitations

To our knowledge, this is the first French study that evaluated the effect on HbA1c of a DSM intervention carried out by an MPCP in a socio-economically deprived area. Another of its strengths is that patients were from different general practices in this deprived area and their medical records were fully accessible. Moreover, our exclusion criteria included absence of follow-up during the study period or the presence of a pathology that did not allow HbA1c monitoring. The aim was to optimise data collection, especially for the primary outcome (HbA1c changes). Our study also has several limitations including missing data, potential residual cofounding, and potential selection bias. First, data were missing for some variables, especially education level and participation rate. Education level is not routinely collected in medical records. We assumed that this variable was missing at random and consequently we used the multiple imputation method to deal with this issue. The obtained results were in accordance with the main analysis. Second, other information (e.g. private health insurance status, marital and family situation, country of birth, understanding of written French, financial situation) was not present in the medical records. These missing data would have allowed matching the two groups also for these socio-economic variables. In our opinion, to develop research in primary care in France, the healthcare organisation needs to think how the patients’ socio-economic data could be collected using the GP’s professional software tools. Moreover, the study retrospective nature did not allow collecting other potential cofounding variables, for instance participation in other DSM programmes or individual data about deprivation for both groups. In addition, we used a logistic regression to take into account potential confounding factors collected in our study. Alternatively, we could have used a propensity score to take into account the non-random allocation of the intervention in our study. However, the performance of these two methods is similar in observational studies [ 50 , 51 , 52 ]. Lastly, we did not know why some patients with diabetes followed at this MPCP did not participate in the DSM programme (refusal rate and reasons for this choice). Therefore, we could not exclude, in addition to a possible reversion to the mean, a selection bias because our exposed group may constitute a subgroup of the population with diabetes more committed to better control their HbA1c.

Our findings suggest that HbA1c improved after participation in a DSM programme led by an MPCP in a socio-economically deprived area. This needs to be confirmed by a prospective study, but it should already encourage the development of DSM targeted to deprived populations in primary care.

Availability of data and materials

The datasets used and analysed in the current study are available from the corresponding author on reasonable request.

Abbreviations

Angiotensin Converting Enzyme Inhibitor

Angiotensin II Receptor Antagonist

Agence Régionale de Santé (Local Health Authority)

Bilan Educatif Partagé Initial (Initial patient-centred educational assessment)

Body Mass Index

Chronic Kidney Disease Epidemiology

Commission Nationale de l’Informatique et des Libertés (French Committee on Data Protection)

Complémentaire Santé Solidaire

Diastolic Blood Pressure

Diabetes Self-Management

General Data Protection Regulation

Glomerular Filtration Rate

Glucagon-Like Peptide-1

General Practitioner

Haute Autorité de Santé (French Health Authority)

Glycated Haemoglobin fraction A1c

HealthCare Provider

Haut Conseil de la Santé Publique (National Health Council)

Institut National de la Statistique et des Etudes Economiques (National institute of statistic and economic studies)

Long Duration Disease

Low Density Lipoprotein C

Multi-professional Primary Care Practice

Type two Diabetes Mellitus

Randomised Controlled Trial

Règlement Général sur la Protection des Données (General data protection framework)

Systolic Blood Pressure

Universal Health Coverage

Organisation mondiale de la Santé. Rapport mondial sur le diabète. Genève: Organisation mondiale de la Santé; 2016 [cited 2021 Nov 12]. 86 p. Available from: https://apps.who.int/iris/handle/10665/254648

L’OMS lève le voile sur les principales causes de mortalité et d’incapacité dans le monde : 2000–2019. [cited 2024 Mar 14]. Available from: https://www.who.int/fr/news/item/09-12-2020-who-reveals-leading-causes-of-death-and-disability-worldwide-2000-2019

Saeedi P, Petersohn I, Salpea P, Malanda B, Karuranga S, Unwin N, et al. Global and regional diabetes prevalence estimates for 2019 and projections for 2030 and 2045: Results from the International Diabetes Federation Diabetes Atlas, 9th edition. Diabetes Res Clin Pract. 2019;157:107843.

Article   PubMed   Google Scholar  

Emerging Risk Factors Collaboration, Sarwar N, Gao P, Seshasai SRK, Gobin R, Kaptoge S, et al. Diabetes mellitus, fasting blood glucose concentration, and risk of vascular disease: a collaborative meta-analysis of 102 prospective studies. Lancet Lond Engl. 2010;375(9733):2215–22.

Article   Google Scholar  

Druet C, Roudier C, Romon I, Assogba F, Bourdel-Marchasson I, Eschwege E, et al. Échantillon national témoin représentatif des personnes diabétiques, Entred 2007–2010. In Saint-Maurice: Institut de veille sanitaire; 2012 [cited 2021 Oct 18]. p. 8. Available from: https://www.santepubliquefrance.fr/maladies-et-traumatismes/diabete/documents/rapport-synthese/echantillon-national-temoin-representatif-des-personnes-diabetiques-entred-2007-2010.-caracteristiques-etat-de-sante-prise-en-charge-et-poids-ec

Fagot-Campagna A, Romon I, Fosse S, Roudier C. Prévalence et incidence du diabète, et mortalité liée au diabète en France – Synthèse épidémiologique. Saint-Maurice (Fra) : Institut de veille sanitaire, novembre 2010, 12 p. Disponible sur : www.invs.sante.fr .

Rao Kondapally Seshasai S, Kaptoge S, Thompson A, Di Angelantonio E, Gao P, Sarwar N, et al. Diabetes mellitus, fasting glucose, and risk of cause-specific death. N Engl J Med. 2011;364(9):829–41.

Article   PubMed   PubMed Central   Google Scholar  

Mandereau-Bruno L, Fosse-Edorh S. Prévalence du diabète traité pharmacologiquement (tous types) en France en 2015. Disparités territoriales et socio-économiques. Bull Epidémiol Hebd. 2017;(27-28):586–91. http://invs.santepubliquefrance.fr/beh/2017/27-28/2017_27-28_3.html .

Fosse S, Fagot-Campagna A. Prévalence du diabète et recours aux soins en fonction du niveau socio-économique et du pays d’origine en France métropolitaine. Enquête décennale santé 2002-2003 et enquêtes santé et protection sociale 2002 et 2004. Saint-Maurice: Institut de veille sanitaire; 2011. 78 p. Disponible à partir de l’URL : http://www.invs.sante.fr .

Larrañaga I, Arteagoitia JM, Rodriguez JL, Gonzalez F, Esnaola S, Piniés JA, et al. Socio-economic inequalities in the prevalence of Type 2 diabetes, cardiovascular risk factors and chronic diabetic complications in the Basque Country. Spain Diabet Med. 2005;22(8):1047–53.

Agardh EE, Ahlbom A, Andersson T, Efendic S, Grill V, Hallqvist J, et al. Explanations of socioeconomic differences in excess risk of type 2 diabetes in Swedish men and women. Diabetes Care. 2004;27(3):716–21.

Bowker SL, Mitchell CG, Majumdar SR, Toth EL, Johnson JA. Lack of insurance coverage for testing supplies is associated with poorer glycemic control in patients with type 2 diabetes. CMAJ Can Med Assoc J J Assoc Medicale Can. 2004;171(1):39–43.

Bihan H, Laurent S, Sass C, Nguyen G, Huot C, Moulin JJ, et al. Association Among Individual Deprivation, Glycemic Control, and Diabetes Complications: The EPICES score. Diabetes Care. 2005;28(11):2680–5.

Barnichon C, Ruivard M, Philippe P, Vidal P, Teissonière M. Diabète de type 2 et précarité : une étude cas-témoins. Rev Médecine Interne. 2011;32(8):467–71.

Article   CAS   Google Scholar  

Li X, Sundquist J, Forsberg PO, Sundquist K. Association between neighbourhood deprivation and heart failure among patients with diabetes mellitus: A 10-year follow-up study in Sweden. J Card Fail. 2020;26(3):193–9.

Attridge M, Creamer J, Ramsden M, Cannings-John R, Hawthorne K. Culturally appropriate health education for people in ethnic minority groups with type 2 diabetes mellitus. Cochrane Database Syst Rev. 2014;2014(9):CD006424.

PubMed   PubMed Central   Google Scholar  

Alzubaidi H, Mc Namara K, Browning C. Time to question diabetes self-management support for Arabic-speaking migrants: exploring a new model of care. Diabet Med J Br Diabet Assoc. 2017;34(3):348–55.

Haute Autorité de santé. Éducation thérapeutique du patient (ETP) : évaluation de l’efficacité et de l’efficience dans les maladies chroniques [on line]. 2018 [cited 10 may 2024]. Avaible at: https://www.has-sante.fr/upload/docs/application/pdf/2018-11/mc_238_synthese_litterature_etp_vf.pdf.

Organisation mondiale de la Santé. Bureau régional de l’Europe. Education thérapeutique du patient : programmes de formation continue pour professionnels de soins dans le domaine de la prévention des maladies chroniques [on line]. 1998 [consulted 10th may 2024]. Available at : https://iris.who.int/bitstream/handle/10665/345371/9789289055987-fre.pdf?sequence=1&isAllowed=y .

Tourette-Turgis C. L’éducation thérapeutique du patient: La maladie comme occasion d’apprentissage. Paris: De Boeck; 2017.

Google Scholar  

Haute Autorité de Santé. [cited 2024 Mar 14]. Stratégie médicamenteuse du contrôle glycémique du diabète de type 2. Available from: https://www.has-sante.fr/jcms/c_1022476/fr/strategie-medicamenteuse-du-controle-glycemique-du-diabete-de-type-2

Trento M, Gamba S, Gentile L, Grassi G, Miselli V, Morone G, et al. Rethink Organization to iMprove Education and Outcomes (ROMEO): a multicenter randomized trial of lifestyle intervention by group care to manage type 2 diabetes. Diabetes Care. 2010;33(4):745–7.

Yamaoka K, Tango T. Efficacy of Lifestyle Education to Prevent Type 2 Diabetes: A meta-analysis of randomized controlled trials. Diabetes Care. 2005;28(11):2780–6.

He X, Li J, Wang B, Yao Q, Li L, Song R, et al. Diabetes self-management education reduces risk of all-cause mortality in type 2 diabetes patients: a systematic review and meta-analysis. Endocrine. 2017;55(3):712–31.

Article   CAS   PubMed   Google Scholar  

Horigan G, Davies M, Findlay-White F, Chaney D, Coates V. Reasons why patients referred to diabetes education programmes choose not to attend: a systematic review. Diabet Med J Br Diabet Assoc. 2017;34(1):14–26.

Manuello P. Inégalités sociales, maladies chroniques et éducation thérapeutique du patient. Soins. 2017;815:14–8.

Haut Conseil de la Santé Publique. Évaluation des programmes d’éducation thérapeutique des patients 2010–2014. Paris: Haut Conseil de la Santé Publique; 2015 Oct [cited 2021 Oct 23]. Available from: https://www.hcsp.fr/explore.cgi/avisrapportsdomaine?clefr=528

Albano MG, Crozet C, d’Ivernois JF. Analysis of the 2004–2007 literature on therapeutic patient education in diabetes: results and trends. Acta Diabetol. 2008;45(4):211–9.

Hadjiconstantinou M, Quinn LM, Tippins F, Schreder S, Khunti K, Davies MJ. A perspective piece on Diabetes Self-Management Education and Support (DSMES) programmes for under-represented groups with T2DM in the UK. Br J Diabetes. 2021;21(1):3–10.

Jacquat D, Morin A. Éducation thérapeutique du patient. Propositions pour une mise en œuvre rapide et pérenne. Juin 2010. Hegel. 2011;3(3):52–7.

Morel A, Lecoq G, Jourdain-Menninger D. Evaluation de la prise en charge du diabète [on line]. 2012 [consulted 10th of may 2024]. Available at: https://www.vie-publique.fr/files/rapport/pdf/124000256.pdf .

Betancourt JR, Green AR, Carrillo JE, Ananeh-Firempong O. Defining cultural competence: a practical framework for addressing racial/ethnic disparities in health and health care. Public Health Rep Wash DC 1974. 2003;118(4):293–302.

Article L1411–11 - Code de la santé publique - Légifrance. [cited 2021 Jun 20]. Available from: https://www.legifrance.gouv.fr/codes/article_lc/LEGIARTI000031930722/

Allory E, Lucas H, Maury A, Garlantezec R, Kendir C, Chapron A, et al. Perspectives of deprived patients on diabetes self-management programmes delivered by the local primary care team: a qualitative study on facilitators and barriers for participation, in France. BMC Health Serv Res. 2020;20(1):855.

Fournier C. Les maisons de santé pluriprofessionnelles, une opportunité pour transformer les pratiques de soins de premier recours: place et rôle des pratiques préventives et éducatives dans des organisations innovantes [thèse de doctorat]. Paris, France: Université Paris Sud; 2015.

Blanchard A, Fiquet L, Le Gall V, Maury A, Allory E. Collaboration interprofessionnelle et maison de santé pluriprofessionnelle. Représentations de participants à un programme d’éducation thérapeutique du patient. Exercer. 2020;165:292–8.

Rennes Métropole. Contrat de ville de la métropole rennaise, 2015 > 2020. Plan d’actions territorial de Villejean. Direction Associations Jeunesse Egalité / Mission Egalité; [cited 2021 Sep 12]. Available from: https://metropole.rennes.fr/sites/default/files/inline-files/Q10_-_contrat_de_Ville_-_plan_d_actions_Villejean_0.pdf

Odgers-Jewell K, Ball LE, Kelly JT, Isenring EA, Reidlinger DP, Thomas R. Effectiveness of group-based self-management education for individuals with Type 2 diabetes: a systematic review with meta-analyses and meta-regression. Diabet Med J Br Diabet Assoc. 2017;34(8):1027–39.

Deakin TA, Cade JE, Williams R, Greenwood DC. Structured patient education: the Diabetes X-PERT Programme makes a difference. Diabet Med. 2006;23(9):944–54.

Davies MJ, Heller S, Skinner TC, Campbell MJ, Carey ME, Cradock S, et al. Effectiveness of the diabetes education and self management for ongoing and newly diagnosed (DESMOND) programme for people with newly diagnosed type 2 diabetes: cluster randomised controlled trial. BMJ. 2008;336(7642):491–5.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Prezio EA, Cheng D, Balasubramanian BA, Shuval K, Kendzor DE, Culica D. Community Diabetes Education (CoDE) for uninsured Mexican Americans: a randomized controlled trial of a culturally tailored diabetes education and management program led by a community health worker. Diabetes Res Clin Pract. 2013;100(1):19–28.

Vaughan EM, Hyman DJ, Naik AD, Samson SL, Razjouyan J, Foreyt JP. A Telehealth-supported, Integrated care with CHWs, and MEdication-access (TIME) Program for Diabetes Improves HbA1c: a Randomized Clinical Trial. J Gen Intern Med. 2021;36(2):455–63.

Vaughan EM, Johnson E, Naik AD, Amspoker AB, Balasubramanyam A, Virani SS, et al. Long-Term Effectiveness of the TIME Intervention to Improve Diabetes Outcomes in Low-Income Settings: a 2-Year Follow-Up. J Gen Intern Med. 2022;37(12):3062–9.

Trento M, Passera P, Borgo E, Tomalino M, Bajardi M, Cavallo F, et al. A 5-Year Randomized Controlled Study of Learning, Problem Solving Ability, and Quality of Life Modifications in People With Type 2 Diabetes Managed by Group Care. Diabetes Care. 2004;27(3):670–5.

Carey G, Crammond B, De Leeuw E. Towards health equity: a framework for the application of proportionate universalism. Int J Equity Health. 2015;15(14):81.

Pereira Gray DJ, Sidaway-Lee K, White E, Thorne A, Evans PH. Continuity of care with doctors-a matter of life and death? A systematic review of continuity of care and mortality. BMJ Open. 2018;8(6):e021161.

Sandvik H, Hetlevik Ø, Blinkenberg J, Hunskaar S. Continuity in general practice as predictor of mortality, acute hospitalisation, and use of out-of-hours care: a registry-based observational study in Norway. Br J Gen Pract. 2022;72(715):e84–90.

Margat A, De Andrade V, Gagnayre R. « Health Literacy » et éducation thérapeutique du patient : Quels rapports conceptuel et méthodologique? Educ Thérapeutique Patient - Ther Patient Educ. 2014;6(1):10105.

Masson E. EM-Consulte. [cited 2022 Feb 28]. Littératie en santé et précarité : optimiser l’accès à l’information et aux services en santé. L’expérience de Solidarité Diabète. Available from: https://www.em-consulte.com/article/1193261/litteratie-en-sante-et-precarite-optimiser-l-acces

Arbogast PG, Ray WA. Performance of disease risk scores, propensity scores, and traditional multivariable outcome regression in the presence of multiple confounders. Am J Epidemiol. 2011;174(5):613–20.

Biondi-Zoccai G, Romagnoli E, Agostoni P, Capodanno D, Castagno D, D’Ascenzo F, et al. Are propensity scores really superior to standard multivariable analysis? Contemp Clin Trials. 2011;32(5):731–40.

Shah BR, Laupacis A, Hux JE, Austin PC. Propensity score methods gave similar results to traditional regression modeling in observational studies: a systematic review. J Clin Epidemiol. 2005;58(6):550–9.

Download references

Acknowledgements

We thank the Rennes Nord-Ouest primary care practice (managed by the association “Avenir Santé Villejean Beauregard”). We thank all the study participants and their GPs who gave their consent to the use of their health data. We thank the French network of University Hospitals HUGO (‘Hôpitaux Universitaires du Grand Ouest’) that supported this article.

No specific funding for this research.

Author information

Authors and affiliations.

Department of General Practice, Univ Rennes, 2, Avenue du Pr Léon Bernard, RENNES Cedex, 35043, France

Sarah Ajrouche, Lisa Louis, Anthony Chapron & Emmanuel Allory

CHU Rennes, Inserm CIC 1414 (Centre d’Investigation Clinique), Rennes, 35000, France

Maxime Esvan, Anthony Chapron & Emmanuel Allory

CHU de Rennes, Univ Rennes, Inserm, EHESP (Ecole des Hautes Etudes en Santé Publique), Irset - UMR_S 1085, Rennes, 35000, France

Ronan Garlantezec

LEPS (Laboratoire Educations et Promotion de la Santé), University of Sorbonne Paris Nord, UR 3412, Villetaneuse, F-93430, France

Emmanuel Allory

You can also search for this author in PubMed   Google Scholar

Contributions

SA, LL, ME, EA: Substantial contributions to the conception and design of the work SA, LL: data acquisition SA, LL, ME, EA: data analysis SA, LL, ME, RG, EA: data interpretation SA, LL, ME, RG, EA: drafted the manuscript SA, LL, ME, AC, RG, EA: substantively revised it.

Corresponding author

Correspondence to Emmanuel Allory .

Ethics declarations

Ethics approval and consent to participate.

Each participant was given an information letter and informed consent was obtained (supplementary file 4). This study was approved by the Rennes University Hospital ethics committee on June 14, 2021 (Number 21.77–2) (supplementary file 5). All procedures performed in the study involving human participants were in accordance with the national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Supplementary material 1., rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Ajrouche, S., Louis, L., Esvan, M. et al. HbA1c changes in a deprived population who followed or not a diabetes self-management programme, organised in a multi-professional primary care practice: a historical cohort study on 207 patients between 2017 and 2019. BMC Endocr Disord 24 , 72 (2024). https://doi.org/10.1186/s12902-024-01601-9

Download citation

Received : 27 March 2024

Accepted : 07 May 2024

Published : 20 May 2024

DOI : https://doi.org/10.1186/s12902-024-01601-9

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Diabetes self-management education
• Deprived population
• Diabetes mellitus
• Primary healthcare

BMC Endocrine Disorders

ISSN: 1472-6823

• General enquiries: [email protected]

• Open access
• Published: 14 May 2024

Estimating effects of whole grain consumption on type 2 diabetes, colorectal cancer and cardiovascular disease: a burden of proof study

• Houpu Liu 1 ,
• Jiahao Zhu 1 ,
• Rui Gao 1 ,
• Lilu Ding 1 ,
• Ye Yang 1 ,
• Wenxia Zhao 1 ,
• Xiaonan Cui 2 ,
• Wenli Lu 3 ,
• Jing Wang 1   na1 &
• Yingjun Li 1   na1  

Nutrition Journal volume  23 , Article number:  49 ( 2024 ) Cite this article

562 Accesses

1 Altmetric

Metrics details

Previous studies on whole grain consumption had inconsistent findings and lacked quantitative assessments of evidence quality. Therefore, we aimed to summarize updated findings using the Burden of Proof analysis (BPRF) to investigate the relationship of whole grain consumption on type 2 diabetes (T2D), colorectal cancer (CRC), stroke, and ischemic heart disease (IHD).

We conducted a literature search in the Medline and Web of Science up to June 12, 2023, to identify related cohort studies and systematic reviews. The mean RR (relative risk) curve and uncertainty intervals (UIs), BPRF function, risk-outcome score (ROS), and the theoretical minimum risk exposure level (TMREL) were estimated to evaluate the level of four risk-outcome pairs.

In total, 27 prospective cohorts were included in our analysis. Consuming whole grain at the range of TMREL (118.5–148.1 g per day) was associated with lower risks: T2D (declined by 37.3%, 95% UI: 5.8 to 59.5), CRC (declined by 17.3%, 6.5 to 27.7), stroke (declined by 21.8%, 7.3 to 35.1), and IHD (declined by 36.9%, 7.1 to 58.0). For all outcomes except stroke, we observed a non-linear, monotonic decrease as whole grain consumption increased; For stroke, it followed a J -shaped curve (the greatest decline in the risk of stroke at consuming 100 g whole grain for a day). The relationships between whole grain consumption and four diseases are all two-star pairs (ROS: 0.087, 0.068, 0.062, 0.095 for T2D, CRC, stroke, and IHD, respectively).

Consuming 100 g of whole grains per day offers broad protective benefits. However, exceeding this threshold may diminish the protective effects against stroke. Our findings endorse replacing refined grains with whole grains as the main source of daily carbohydrates.

Registry and registry number for systematic reviews or meta-analyses

We have registered our research in PROSPERO, and the identifier of our meta-analyses is CRD42023447345 .

Peer Review reports

Introduction

Whole grains have been widely endorsed as a superior substitute for primary energy and carbohydrate sources in daily dietary guidelines because of their high dietary fiber content and numerous bioactive compounds [ 1 ]. The Global Burden of Disease Study 2019 (GBD 2019) has reported that lower intake of whole grain accounted for 1,844,836 (95% uncertainty interval [UI]: 2,338,609–921,291) deaths and 42.5 million (53.2–17.5) disability-adjusted life years (DALYs) [ 2 ]. The large estimated burden demonstrated the importance of fully appreciating the relationship between whole grain consumption and potentially related health outcomes and of further improving the strength of evidence supporting the understanding of those relationships.

Increasing evidence has found that a high intake of whole grains is related to a reduction in the risk of type 2 diabetes (T2D), colorectal cancer (CRC), ischemic heart disease (IHD), and stroke [ 2 , 3 ]. However, regarding CRC, T2D and IHD, previous studies, including dose-response meta-analysis or cohort studies, exhibit variations in their consumption ranges. This complicates the comparability and consolidation of evidence [ 3 , 4 , 5 , 6 ]. Besides, in relation to stroke, recent meta-analyses have presented inconsistent findings [ 7 , 8 ]. Although there is an increasing body of evidence supporting the positive impact of consuming whole grains on health, the challenge lies in accurately estimating RR associated with varying levels of consumption. This limitation hinders the ability of decision-makers to fully comprehend the strength of the connection between consuming whole grains and various health outcomes.

Burden of proof risk function (BPRF) is a new meta-analysis method that can quantitatively estimate the level of risk closest to the null hypothesis [ 9 ]. Hitherto, most of meta-regression studies applied given fixed knots to fit the spline models or forced a log-linear assumption to simplify statistical analysis. However, such a method may limit their ability to capture the effects of whole grain consumption on health outcomes, as the relationship between increasing whole grain intake and its impact on health might not be straightforward: it could lead to slight decreases in positive effects, or it could even become harmful if the consumption of whole grains goes beyond a certain point [ 3 , 10 ]. Unlike existing methods, BPRF relaxed the conventional assumption of a log-linear shape in risk functions, and instead applied a data-driven approach to determine the relationship of risk-outcome pairs using a quadratic spline. Thus, BPRF can help to identify the ‘true’ shape of the risk function [ 11 ]. In addition, existing methods, such as Grading of Recommendations, Assessment, Development and Evaluations (GRADE) or NutriGRADE, are commonly applied to assess the quality of the underlying evidence [ 12 ]. However, such methods are unable to extend to quantify variation in true effect size caused by bias from covariates or other limitations of the evidence [ 11 ]. Nevertheless, BPRF can synthesize available evidence in algorithm to calculate uncertainty inclusive of between-study heterogeneity.

To precisely quantify the health effects of whole grain consumption, a meta-regression analysis was conducted on the evidence from prospective cohort studies. This study focused specifically on four health outcomes (T2D, CRC, IHD, and stroke) linked to whole grain consumption, as reported by the GBD study [ 13 ].

Our protocol has been registered in International Prospective Register of Systematic Reviews (PROSPERO, identifier: CRD42023447345 ). We followed a standard framework of Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline to report our results [ 14 ].

Search strategy and selection criteria

We searched data published in English in the MEDLINE and Web of Science for systematic review, and cohort studies from 1 January, 2000 to 12 June, 2023, using standard search strings (Supplementary Table 1 ). A reference list of included publications was also manually screened to identify additional cohort studies. Titles and abstracts were screened by two reviewers (H Liu and J Wang), with discrepancies being reconciled through consulting a third author (Y Li).

Only prospective observational studies (for both incidence and mortality) published in English were included. Studies should report a relative risk ratio (RR), odds ratio (OR) or hazard ratio (HR) of the associations between whole grain consumption and at least one of the four outcomes. Additionally, they should specify the amount of whole grain consumption in both the reference group and the alternate group for comparison.

Retrospective studies, conference abstracts, ecological studies, case reports, case-series, letters to the editor, conference proceedings, umbrella reviews, systematic reviews or meta-analyses as well as studies conducted in animals, children, or adolescents were excluded. Besides, we excluded studies that failed to report whole grain consumption without grams or servings equivalent, such as studies that used aggregated “diet scores” as a measure of consumption, and those that only reported specific subtypes of grains were also excluded. And studies reporting outcomes outside the scope of interest, such as all-cause mortality, or lacking specificity such as cardiovascular disease or diabetes mellitus, have been excluded.

Data extraction

For each study, we collected the information of the eligible studies including the first author’s name, location, population characteristics (age, sex, race, and sample size), follow-up period, exposure definition, exposure assessment method, outcome definition, outcome ascertainment method, and covariates used in the study. Data were extracted by one author (H Liu) and checked by another author (J Wang) for accuracy. Besides, we also collected data on the range of exposure, sample size, person-years, number of events and risk estimate (RRs, HRs or ORs) and its corresponding uncertainty to conduct BPRF analysis. The uniform extraction procedures are shown in Supplementary Table 2 .

We used a framework of BPRF methodology developed by Zheng et al. to assess the risk of bias in included studies [ 11 , 15 , 16 ]. For each included study, we extracted information concerning aspects of study design that could potentially bias the reported effect size and coded this information into study-level covariates [ 11 ]. These study-level covariates are followed as: follow-up time (≤ 10 months and > 10 months), exposure definitions, outcome definitions, effect size measures (HRs, RRs or ORs), the endpoint of outcome events (incidence or mortality), frequency of exposure measurements (single or repeat), outcome ascertainment methods (administrative records or self-reports), and the level of adjustment for relevant confounders (creating cascading dummy variables standing for the number of confounders adjusted in risk regression model from selected studies, and the minimum threshold for confounder adjustment for age and sex) [ 11 ]. These covariates would be further adjusted in our BPRF analysis if they significantly biased our estimated risk functions.

In addition to these covariates, we selected four common study characteristics that are highly relevant and likely to introduce bias, in order to evaluate the study quality [ 9 , 17 , 18 ]. These characteristics include the representativeness of the study population (whether it represents the general population or specific sub-groups such as high-risk populations), outcome confirmation, exposure mesurement and assessment, and control for confounding factors [ 11 ]. The quality score for each selected study was calculated by summing the scores across these four domains.

Statistical methods

The estimates for our primary indicators of this work are mean RRs across a range of exposures, BRPFs, ROSs and star ratings for each risk-outcome pair. And the exposure unit was standardized to grams of consumption per day before synthesis. For each study that reported means or quantiles consumption rather than ranges of whole grain consumption, midpoint of defined quantile as the cutoff for intake intervals was used [ 10 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 ]. When the quantile dose range didn’t have a specific endpoint, and mean and standard deviation weren’t available, we assumed a consumption level of 0 g per day as the lowest amount [ 22 , 24 , 25 , 28 , 30 , 31 , 32 , 33 ]. For the upper limit of consumption, we used the range from the closest quartile or tertile within the cohort. In addition, we used 30 g to evaluate one serving of whole grain consumption if the value of a serving was not stated [ 34 ].

Estimating the shape of relationships between whole grain consumption and four health outcomes

We firstly modeled the mean log-RR (a measure of effect size) curve with MR-BRT (a Bayesian meta-regression tool) developed at the Institute for Health Metrics and Evaluation (IHME) [ 13 , 35 ], and followed a uniform analysis procedure to select model specifications for all dietary risks, which is described by Zheng et al. [ 15 ]. For protective risk factors with hypothesis of monotonic deceresing, the final models were run applying quadratic splines with two internal knots and a linearity prior on the right tail [ 11 , 15 ]. However, for the J-shaped risk curve, we employed quadratic splines with three internal knots and a linearity prior on both the right and left tails, without a monotonic prior [ 11 ]. Besides, to avoid the influence of extreme data and reduce publication bias, we trimmed 10% of data for each outcome as outliers [ 11 ].

Following the GRADE approach, we created binary covariates based on the the extracted information about specific study characteristics to identify potential sources of systematic bias within our included datasets. A step-wise Lasso approach were applied to assess the significance of these bias covariates at a threshold of 0.05. If the bias covariates were found to be significant, they were selected for adjustment in the final log-RR model.

To evaluate and adjust between-study heterogeneity, we quantified common sources of bias across the selected covariates that were likely to cause bias. And we calculated 95% UIs for each mean risk curve both with between-study heterogeneity incorporated (a ‘conservative’ UI) and without between-study heterogeneity incorporated (a ‘conventional’ UI) based on the selected biased covariates. Only the UIs that include between-study heterogeneity are presented in our main results unless specified.

Based on the aforementioned models, we then adjusted the selected bias covariates to decrease the variation in model residuals arising from differences in study quality and analysis. However, we primarily applied empirical evidence to choose bias covariate related to potential publication or reporting biases, which may ignore some confounders. Thus, Egger’s regression was conducted to detect publication bias, which estimated correlation between the study residuals and standard deviation of the corresponding data points. Funnel plots of the residuals of the risk function and standard deviations were generated to inspect reporting bias visually. And P value was used to assess the statistical significance of a risk for publication and/or reporting bias.

Estimating the TMREL/minimum risk exposure level

To draw robust conclusions about health benefits of whole grain consumption, we calculated the theoretical minimum risk exposure level (TMREL) of all potential outcomes linked to consuming whole grains. TMREL aligns with real-world consumption patterns supported by the included data, enabling an estimation of the average risk associated with whole grain intake. For protective risk factors, the lower bound of TMREL is defined as the 85th percentile of the lower limit within the highest consumption range across all studies. whereas the upper bound of TMREL is determined as the 85th percentile of the midpoint within the highest consumption range across all studies [ 15 ].

Estimating BPRF value, risk-outcome score (ROS) and star rating

Using the mean RR curves that incorporated between-study heterogeneity into uncertainty estimate, we estimated the BPRF from a conservative risk function. The BPRF was defined as the 5th (for harmful) or 95th percentile risk curve that is closest to the null. Afterwards, we calculated the ROS, which was equivalent to the mean log-BPRF averaged value over the 15th and 85th percentiles of the distribution of whole grain consumption. This value can give conservative interpretations regarding the association between whole grain consumption and four health outcomes [ 15 , 36 ]. Then, the ROSs of risk-outcome pairs were converted into a comparison across risk-outcome pairs and a star rating (from one to five) was assigned based on the quantitative assessment of the association, where a one-star rating indicating a non-significant relationship based on the conservative interpretation, two-star through five-star ratings implying a decrease in risk with average exposure (compared to no exposure). And the ranges of ROS in 0–0.1398 stands for two-star pairs, > 0.1398–0.4055 for three-star pairs, > 0.4055–0.6152 for four-star pairs and greater than 0.6152 for five-star pairs for protective risks [ 11 ].

Sensitivity analyses

To strengthen our estimates on the association between whole grain intake and four health outcomes and reduce the impact of outliers, we used trimming analysis with the Least Trimmed Squares (LTS) method. This method automatically identifies and removes outliers within the model’s likelihood. In our study, we trimmed the top 10% of data points that deviated the most from the expected dose-response curve as part of sensitivity analysis [ 11 , 35 ].

Dose-response analysis on whole grain consumption and four health outcomes was conducted by applying MR-BRT tool which included several Python packages (limetr 0.0.5, mrtool 0.0.1, IPOPT 1.2.0). And we executed BPRF analysis in Visual Studio Code with extensions of R version 4.2.1 and Python 3.9.0.

Study identification

A total of 3118 articles were found using search strings, and of those we identified 28 population-based prospective cohort studies, presenting a total of 184 estimates of effect sizes for associations between whole grain consumption and the four included health outcomes [ 10 , 25 , 26 , 28 , 29 , 32 , 33 ]. Details of the literature search are shown in Supplementary Fig.  1 . Eleven studies were from the United States, including data mainly from the HPFS, NHS, NHSII, the ATBC cohort, and the Cancer Prevention Study‑II Nutrition Cohort [ 10 , 20 , 21 , 27 , 28 , 29 , 32 , 38 , 39 , 41 , 42 ]. Fifteen of 28 studies were from the European population [ 19 , 22 , 23 , 24 , 25 , 26 , 30 , 33 , 25 , 43 , 44 , 45 , 46 , 47 ]. Besides, one study (the PURE cohort) collected information covering 21 countries (including the regions of North America and Europe, South America, Africa, the Middle East, South Asia, South East Asia, and China) [ 48 ] and one study reported the role of whole grain consumption on IHD in Chinese population [ 40 ]. Detailed information about the included cohorts is displayed in Supplementary Table 3 .

Characteristics of included studies

Of 28 included publications in the BPRF analysis, a total of eight studies investigated the association between whole grains and T2D [ 10 , 28 , 29 , 32 , 33 , 37 , 46 , 47 ], seven for CRC [ 19 , 20 , 21 , 30 , 42 , 43 , 44 ], six for both IHD and stroke [ 23 , 26 , 39 , 40 , 41 , 48 ], three for IHD [ 25 , 38 , 45 ], three for stroke [ 24 , 27 , 49 ], and one for IHD, stroke and CRC [ 22 ]. The median follow-up time of all included studies was 13.5 years (range: 6–25.8 years).

All included publications used dietary records or recalls, or food frequency questionnaires to collect data regarding whole grain intake. In total, seventeen publications used baseline data of whole grain intake in their analysis (single measurement) [ 19 , 23 , 24 , 25 , 26 , 30 , 33 , 39 , 25 , 43 , 44 , 45 , 48 ], whereas ten considered the average whole grain intake throughout the follow-up (i.e., based on multiple measurements) as the main exposure [ 10 , 20 , 21 , 22 , 27 , 37 , 38 , 40 , 41 , 42 ]. Three studies took self-report records to assess outcomes [ 28 , 29 , 32 ], and 25 studies used administrative medical records [ 10 , 19 , 21 , 22 , 25 , 26 , 27 , 30 , 33 , 37 , 41 , 43 , 44 , 49 ]. Three studies used mortality as the endpoint [ 26 , 39 , 43 ], and the rest studies considered incidence as the endpoint. 8 studies reported effect sizes with RRs [ 19 , 20 , 24 , 26 , 30 , 32 , 33 , 49 ], seventeen studies reported HRs [ 22 , 23 , 25 , 27 , 37 , 38 , 39 , 40 , 41 , 48 ], one study reported ORs [ 46 ], and one study reported incidence rate ratios (IRRs) [ 44 ]. The detailed information is presented in Supplementary Tables 4 – 7 .

Estimation of the shape of whole grains with T2D, CRC, IHD and stroke

Using BPRF methodology, our analyses revealed a correlation between higher whole grain intake and a reduced risk across all the outcomes considered. Figures  1 , 2 , 3 and 4 depict the BPRF curves for each risk-outcome pair, while Table  1 presents the results of the dose-response analysis.

BPRF analysis on the association between whole grain consumption and T2D. a, log RR function. b, RR function. c, modified funnel plot showing the residuals (relative to zero) on the x-axis and the estimated s.d. that includes reported s.d. and between-study heterogeneity on the y-axis

BPRF analysis on the association between whole grain consumption and CRC. a, log RR function. b, RR function. c, modified funnel plot showing the residuals (relative to zero) on the x-axis and the estimated s.d. that includes reported s.d. and between-study heterogeneity on the y-axis

BPRF analysis on the association between whole grain consumption and IHD. a, log RR function. b, RR function. c, modified funnel plot showing the residuals (relative to zero) on the x-axis and the estimated s.d. that includes reported s.d. and between-study heterogeneity on the y-axis

BPRF analysis on the association between whole grain consumption and stroke. a, log RR function. b, RR function. c, modified funnel plot showing the residuals (relative to zero) on the x-axis and the estimated s.d. that includes reported s.d. and between-study heterogeneity on the y-axis

Specifically, our analysis revealed that the associations between whole grain consumption and the risk of T2D, CRC and IHD all exhibited non-linear, monotonically decreasing trends (Figs.  1 , 2 and 3 ). In regard to T2D (Fig.  1 a and b), the sharpest decline in risk was noted at daily consumption of 50 g, with a reduction of 34.3% (95% UI including between-study heterogeneity: 5.3 to 55.7), compared to no whole grain consumption (at 0 g per day). Nonetheless, the reduction in risk tapered off to a mere 1.2% (0.2–1.6) when comparing a consumption level of 90 g per day to 50 g per day. With respect to CRC (Fig.  2 a and b), the largest reduction in CRC risk was identified when comparing the risk between an intake of 0 g per day and of 80 g per day, showcasing a noteworthy decline of 17.3%. we observed only marginal additional reductions in risk when consumption is beyond 80 g per day. As for IHD, the steepest decline of 32.1% (95% UI inclusive of between-study heterogeneity of 6.0 to 51.8) in IHD risk was observed when comparing risk between an intake of 0 g per day and of 30 g per day, with more modest marginal declines in IHD risk when consumption levels greater than 30 g per day (Fig.  3 a and b).

Different from the aforementioned results, a J -shape association was found between whole grain consumption and the risk of stroke (Fig.  4 a and b). The greatest reduction in stroke risk, observed at an intake of 100 g per day, was 24.6% (95% UI including between heterogeneity: 8.8 to 38.8). The mean risk of stroke at 60 g per day was 14.1% (7.0 to 21.3 including between-study heterogeneity) higher than at 100 g per day. And it was 1.5% (0.3 to 2.0) higher at 120 g per day compared to 100 g per day.

Additionally, the BPRF estimated ROSs for IHD, T2D, CRC and stroke of 0.095, 0.087, 0.068 and 0.062, respectively, which were applied to explore the average health benefits across the universe of whole grain consumption. Such estimates indicated that the consuming whole grains, on average, was related to a 9.9% decreased risk of IHD, a 9.1.% lower risk of T2D, a 7.0% lower risk of CRC and a 6.4% lower risk of stroke compared to a 0 g of whole grain intake. The star ratings of the four risk-outcome pairs all correspond to a two-star rating. After adjusting for between-study heterogeneity, the relationships still achieved statistical significance.

TMREL level of whole grain consumption

Based on observed exposure levels reported in the included studies, a TMREL of 118.5 g to 148.1 g per day, corresponding to approximately 4–5 servings per day, was estimated (detailed input information presented in Supplementary Table 8 ). Compared to TMREL (118.5–148.1 g), consuming no whole grains was associated with a 37.3% (5.8 to 59.5, inclusive of between-study heterogeneity) greater mean risk of T2D, a 17.3% (95% UI inclusive of between-study heterogeneity of 6.5 to 27.7) greater risk of CRC, a 36.9% (95% UI inclusive of between-study heterogeneity of 7.1 to 58.0) greater mean risk of IHD, and a 21.8% greater mean risk of stroke (95% UI inclusive of between-study heterogeneity of 7.3 to 35.1).

Sensitivity analysis and publication bias

The sensitivity analyses showed that trimming had significant effects on the ROS and reporting bias of the association between whole grain consumption and T2D and IHD. Without trimming, the ROS of T2D is -0.136, and a significant publication bias was detected using Egger’s regression ( P  = 0.023, as shown in Supplementary Fig.  2 c). With respect to IHD, the results without trimming have reported a ROS of -0.273 and statistically significant evidence of small-study bias ( P for Egger’s regression = 0.028, Supplementary Fig.  3 c). However, both of the two health outcomes were found two significant study-level bias covariates (T2D: age of the population and outcome ascertainment methods; IHD: exposure measurement and outcome ascertainment methods). after removing outliers (T2D: 4 [ 31 , 50 ], IHD: 5 [ 24 , 50 ]) and adjusted for the selected bias covariates, no evidence of publication bias was observed (Figs.  1 c and 3 c). On the other hand, for CRC and stroke, trimming had a minor impact on the results. In both our analyses, with and without trimming, no significant evidence of publication bias was detected and no bias covariates were identified (Figs.  2 c and 4 c; and Supplementary Fig.  4 c, 5 c).

In this analysis, we applied a BPRF framework, which takes into account between-study heterogeneity, to quantify the association between whole grain consumption and four health outcomes. Our results suggested that increasing the intake of whole grains was significantly related to a reduction in the risk of CRC, T2D, IHD and stroke. When comparing TMREL (118.5–148.1 g per day) with a daily intake of 0 g of whole grains, the risk reductions for four diseases (T2D, CRC, IHD, and stroke) were 37.2%, 27.3%, 26.9%, and 21.8%, respectively. For all outcomes except stroke, we observed that mean risk exhibited a non-linear, monotonic decrease as whole grain consumption increased. However, the relationship between whole grains and stroke is like a J -shaped as the risk increased with exposure levels above or below a global minimum. Based on a conservative interpretation of available data (the averaged BPRF value), we found a slight decline in the risk of stroke, CRC, T2D, and IHD compared to no whole grain intake (by at least 6.4%, 7.0%, 9.1%, and 9.9%, respectively). The converted grade ratings of our evidence were all two-star ratings.

The protective role of whole grain consumption on CRC risk is well-documented. The World Cancer Research Fund/American Institute for Cancer Research (WCRF/AICR) has provided strong evidence on the protective role of whole grain consumption at 90 g/day on the risk of CRC (RR: 0.83; 95% CI: 0.79 to 0.89) [ 51 ]. Similarly, Schwingshackl et al. also reported an RR of 0.84 (95% CI: 0.78 to 0.90) of whole grain consumption at 90 g per day, and such evidence was rated as moderate according to the NutriGrade recommendation [ 5 ]. In consistent with previous studies, we found that consuming 90 g whole grains per day was also related with a similar reduction in the risk of CRC (RR: 0.83; 95% UI: 0.73 to 0.95), although the range of confidence interval was relatively wide (due to the consideration of between-study heterogeneity). Furthermore, as computing the mean across the universe of studies is appropriate to estimate the relationship between risk and outcome [ 52 ], we calculated the averaged BPRF value of 0.932, which is corresponding to a decline in the mean risk of CRC by 7.0%. Additionally, the star rating considering between-study heterogeneity of the whole grains-CRC pair is two-star. This estimate implied that the strength of the association evidence was relatively weak, in contrast to the WCRF/ARIC assessment, which categorized the evidence as “convincing” [ 11 ]. In fact, the extent of biases in nutritional epidemiological studies, including substantial residual confounding and selective reporting, can significantly impact the accuracy of health risks estimates related to the studied nutrients. Furthermore, the observational findings from prospective cohort studies exhibit considerable variation across different research endeavors [ 53 ]. Therefore, it is important to consider the strength of the association and employ a quantitative approach to assess consistency (that is, between-study heterogeneity) when evaluating evidence. Additionally, it is advisable to adopt a more conservative interpretation [ 54 ]. Our risk assessment indicated that increasing whole grain consumption can slightly reduce the risk of CRC, after correcting for biases due to factors such as study design, the representativeness of the study population, control for confounding, and so on.

With respect to IHD, the evidence stemming from previous meta-analyses has displayed a lack of consistency. For instance, Hu H et al. only found a linear association with 3 knots percentiles (25th, 50th, and 75th) selected [ 8 ]. However, Bechthold A, et al. provided evidence of a non-linear dose-response association ( P non-linearity <0.001) for IHD using three fixed knots at 10%, 50%, and 90% through the total distribution of the reported intake [ 7 ]. The disparities in their findings could be potentially due to variations in the selection of different knot placements along the estimated risk function curve, which might influence on the resulting accuracy of a spline approximation of a curve [ 55 ]. On the other hand, BPRF analysis, according to the given degree and number of knots, automatically sampled a set of knot placements for a feasible knot distribution, evaluated each resulting model by computing its fit and curvature, and then aggregated the final model as a weighted combination of the ensemble to mitigate the effect of spline parameter selection results and draw a robust conclusion. With this methodology, we found a non-linear, monotonic decline association between whole grain consumption and IHD.

In the case of stroke, previous meta-analyses generated mixed results. Bechthold et al. observed no association between whole grain intake and the risk of stroke in the non-linear dose-response analysis [ 7 ]. Conversely, Aune et al. observed a protective role of whole grain consumption on stroke risk, but this role was only significant in their non-linear dose-response analysis, and the risk curve exhibited a J -shaped pattern [ 6 ]. The difference between these studies might be partially attributable to different included studies [ 6 , 7 ]. Our analysis, including the results of newly published studies (the PURE study, China Kadoorie Biobank study and UK Biobank study) and applying BPRF methodology (free of log-linear hypothesis), found a J -shaped relationship between whole grain consumption and stroke, and we observed the greatest reduction in stroke risk observed at an intake of 100 g per day (RR: 0.75, 95%UI: 0.62 to 0.92). Unlike our analysis, both of the aforementioned studies assumed the association between whole grains and stroke to be log-linear [ 6 , 7 ], which might be inappropriate. A log-linear association implies that a fixed increment of health roles of whole grain consumption (for example, 30 g/day) remains constant across all levels of intake; however, an increase in consumption from 0 to 120 g/day would not have the same impact as an increase from 240 to 360 g/day, especially considering that excessive consumption may cause health issues such as overweight [ 56 ].

Our analyses support the need for stronger efforts and policies to encourage increased whole grain consumption as a means to reduce the risk of chronic diseases. Whole grains are well-known for their abundance of dietary fiber and nutrients. However, they can also be a notable source of food-borne contaminants. Nonetheless, current evidence suggests that increasing whole grain consumption could improve public health [ 57 ]. We estimated a TMREL of 118.1–148.5 g per day as the high consumption levels of whole grain intake in the real world, and such estimates are in line with the recommended intake of whole grains promoted by the GBD and the World Health Organization (WHO) [ 58 ], which is at least 125 g per day [ 59 ]. To address both individual and environmental health, the Lancet EAT Commission recommends a primarily plant-based diet, including 232 g of whole grains per day to reduce the carbon footprint of animal-based foods [ 60 ]. Nevertheless, our analysis solely took the individual-level health benefits into consideration, and the potential environmental benefits of increased whole grain consumption were not evaluated. Based on our analyses, particularly the notable protective roles observed with daily consumption of 100 g of whole grains against the risk of stroke, it seems that incorporating a minimum of three servings of whole grains per day has the potential to lower the risk of chronic diseases.

Our study employed BPRF methodology to estimate the association between whole grain consumption and four health outcomes. Compared to traditional meta-analysis methods, this method could quantify between-study heterogeneity, and infer flexible risk functions. It does so without imposing a log-linear hypothesis, which may exaggerate risks at higher exposure levels and overlook crucial details at lower exposure levels. With this methodology, we have found that the risk curves for whole grain consumption and IHD, CRC and T2D displayed decreasing marginal returns, indicating that as whole grain intake increases, the incremental health benefits of whole grains decrease. In addition, quantifications of between-study heterogeneity and corrections for biases due to study design in the methods can contribute to a conservative interpretation and a better understanding of the protective role of whole grain consumption in real-world settings. Thirdly, by estimating RRs associated with consuming whole grains at the TMREL (in correspondence to high real-world consumption levels), we were able to provide sufficient evidence to justify more robust efforts and policies promoting increased whole grain consumption to reduce chronic disease risk, especially with regard to CRC, T2D, IHD and stroke. In general, our analysis results indicate that improving whole grain consumption is beneficial toward enhancing public health.

Although the methodological framework addressed by Zheng et al. overcame many of the limitations in existing meta-analysis approaches, this study still has several limitations. Firstly, all studies included in our analysis were observational, and we were unable to definitively assess causality. Besides, this study mainly focused on total or whole grain consumption, and the impacts of different specific subtypes of whole grains on health outcomes may vary. For example, previous reviews have indicated that increasing whole-grain breakfast cereals, other than whole-grain bread, may decrease the risk of stroke [ 6 ]; furthermore, oats or oatmeal are linked to lower all-cause mortality but show no impact on T2D and CVD incidence [ 6 , 61 ]. Thus, further prospective cohort studies and randomized clinical trials focusing on different subtypes of whole grains and their associations with specific chronic diseases are required. Besides, most of the studies included were from the US and Europe, which limited the ability to make evidence-based recommendations, as dietary patterns can vary significantly between Asian and Western populations [ 62 ]. With respect to the Asian population, rather than whole grain consumption, most of studies investigated the role of refined grain consumption in the form of white rice and noodles [ 63 ], and further studies are needed to explore the association between whole grains and health outcomes on populations in Asia. Thirdly, the associations between whole grains and risks of different stroke types may be heterogeneous [ 24 ]. Unfortunately, we couldn’t investigate these associations separately due to a lack of reported data on stroke types in available studies.

In conclusion, the present study demonstrates that the consumption of whole grains plays a protective role in the risks of CRC, T2D, IHD and stroke, and the BPRF analysis, which did not rely on log-linear assumptions, revealed non-linear associations between whole grain intake and the four diseases of interests. The star ratings converted by ROSs for all four outcomes are all two stars, indicating that the associations between whole grain intake and CRC, T2D, IHD and stroke remain significant. The current body of evidence justifies the need for increased efforts and policies to promote higher whole grain consumption for the betterment of public health.

Data availability

No datasets were generated or analysed during the current study.

Cheng Z, Qiao D, Zhao S, Zhang B, Lin Q, Xie F. Whole grain rice: updated understanding of starch digestibility and the regulation of glucose and lipid metabolism. Compr Rev Food Sci Food Saf. 2022;21(4):3244–73.

Article   CAS   PubMed   Google Scholar  

Collaborators GBDRF. Global burden of 87 risk factors in 204 countries and territories, 1990–2019: a systematic analysis for the global burden of Disease Study 2019. Lancet. 2020;396(10258):1223–49.

Article   Google Scholar  

Reynolds A, Mann J, Cummings J, Winter N, Mete E, Te Morenga L. Carbohydrate quality and human health: a series of systematic reviews and meta-analyses. Lancet. 2019;393(10170):434–45.

Aune D, Norat T, Romundstad P, Vatten LJ. Whole grain and refined grain consumption and the risk of type 2 diabetes: a systematic review and dose-response meta-analysis of cohort studies. Eur J Epidemiol. 2013;28(11):845–58.

Schwingshackl L, Schwedhelm C, Hoffmann G, Knuppel S, Laure Preterre A, Iqbal K, et al. Food groups and risk of colorectal cancer. Int J Cancer. 2018;142(9):1748–58.

Aune D, Keum N, Giovannucci E, Fadnes LT, Boffetta P, Greenwood DC, et al. Whole grain consumption and risk of cardiovascular disease, cancer, and all cause and cause specific mortality: systematic review and dose-response meta-analysis of prospective studies. BMJ. 2016;353:i2716.

Article   PubMed   PubMed Central   Google Scholar  

Bechthold A, Boeing H, Schwedhelm C, Hoffmann G, Knuppel S, Iqbal K, et al. Food groups and risk of coronary heart disease, stroke and heart failure: a systematic review and dose-response meta-analysis of prospective studies. Crit Rev Food Sci Nutr. 2019;59(7):1071–90.

Hu H, Zhao Y, Feng Y, Yang X, Li Y, Wu Y, et al. Consumption of whole grains and refined grains and associated risk of cardiovascular disease events and all-cause mortality: a systematic review and dose-response meta-analysis of prospective cohort studies. Am J Clin Nutr. 2023;117(1):149–59.

Article   PubMed   Google Scholar  

Dai X, Gil GF, Reitsma MB, Ahmad NS, Anderson JA, Bisignano C, et al. Health effects associated with smoking: a Burden of Proof study. Nat Med. 2022;28(10):2045–55.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Hu Y, Ding M, Sampson L, Willett WC, Manson JE, Wang M, et al. Intake of whole grain foods and risk of type 2 diabetes: results from three prospective cohort studies. BMJ. 2020;370:m2206.

Zheng P, Afshin A, Biryukov S, Bisignano C, Brauer M, Bryazka D, et al. The Burden of Proof studies: assessing the evidence of risk. Nat Med. 2022;28(10):2038–44.

Schwingshackl L, Knuppel S, Schwedhelm C, Hoffmann G, Missbach B, Stelmach-Mardas M, et al. Perspective: NutriGrade: a Scoring System to assess and judge the Meta-evidence of Randomized controlled trials and Cohort studies in Nutrition Research. Adv Nutr. 2016;7(6):994–1004.

Collaborators GBDCRF. The global burden of cancer attributable to risk factors, 2010-19: a systematic analysis for the global burden of Disease Study 2019. Lancet. 2022;400(10352):563–91.

Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71.

Stanaway JD, Afshin A, Ashbaugh C, Bisignano C, Brauer M, Ferrara G, et al. Health effects associated with vegetable consumption: a Burden of Proof study. Nat Med. 2022;28(10):2066–74.

Lescinsky H, Afshin A, Ashbaugh C, Bisignano C, Brauer M, Ferrara G, et al. Health effects associated with consumption of unprocessed red meat: a Burden of Proof study. Nat Med. 2022;28(10):2075–82.

Siemieniuk R, Guyatt G. What is GRADE? BMJ Best Practice. 2021. https://bestpractice.bmj.com/info/us/toolkit/learn-ebm/what-is-grade/ .

The GRADE Working Group. GRADE handbook. 2013. https://gdt.gradepro.org/app/handbook/handbook.html .

Pietinen P, Malila N, Virtanen M, Hartman TJ, Tangrea JA, Albanes D, et al. Diet and risk of colorectal cancer in a cohort of Finnish men. Cancer Causes Control. 1999;10(5):387–96.

He X, Wu K, Zhang X, Nishihara R, Cao Y, Fuchs CS, et al. Dietary intake of fiber, whole grains and risk of colorectal cancer: an updated analysis according to food sources, tumor location and molecular subtypes in two large US cohorts. Int J Cancer. 2019;145(11):3040–51.

Hullings AG, Sinha R, Liao LM, Freedman ND, Graubard BI, Loftfield E. Whole grain and dietary fiber intake and risk of colorectal cancer in the NIH-AARP Diet and Health Study cohort. Am J Clin Nutr. 2020;112(3):603–12.

Thompson AS, Tresserra-Rimbau A, Karavasiloglou N, Jennings A, Cantwell M, Hill C, et al. Association of Healthful Plant-based Diet Adherence with Risk of Mortality and Major Chronic diseases among adults in the UK. JAMA Netw Open. 2023;6(3):e234714.

Sonestedt E, Hellstrand S, Schulz CA, Wallstrom P, Drake I, Ericson U, et al. The association between carbohydrate-rich foods and risk of cardiovascular disease is not modified by genetic susceptibility to dyslipidemia as determined by 80 validated variants. PLoS ONE. 2015;10(4):e0126104.

Tektonidis TG, Akesson A, Gigante B, Wolk A, Larsson SC. A Mediterranean diet and risk of myocardial infarction, heart failure and stroke: a population-based cohort study. Atherosclerosis. 2015;243(1):93–8.

Helnaes A, Kyro C, Andersen I, Lacoppidan S, Overvad K, Christensen J, et al. Intake of whole grains is associated with lower risk of myocardial infarction: the Danish Diet, Cancer and Health Cohort. Am J Clin Nutr. 2016;103(4):999–1007.

Johnsen NF, Frederiksen K, Christensen J, Skeie G, Lund E, Landberg R, et al. Whole-grain products and whole-grain types are associated with lower all-cause and cause-specific mortality in the scandinavian HELGA cohort. Br J Nutr. 2015;114(4):608–23.

Juan J, Liu G, Willett WC, Hu FB, Rexrode KM, Sun Q. Whole grain consumption and risk of ischemic stroke: results from 2 prospective cohort studies. Stroke. 2017;48(12):3203–9.

van Dam RM, Hu FB, Rosenberg L, Krishnan S, Palmer JR. Dietary calcium and magnesium, major food sources, and risk of type 2 diabetes in U.S. black women. Diabetes Care. 2006;29(10):2238–43.

Li J, Glenn AJ, Yang Q, Ding D, Zheng L, Bao W, et al. Dietary protein sources, mediating biomarkers, and incidence of type 2 diabetes: findings from the women’s Health Initiative and the UK Biobank. Diabetes Care. 2022;45(8):1742–53.

Larsson SC, Giovannucci E, Bergkvist L, Wolk A. Whole grain consumption and risk of colorectal cancer: a population-based cohort of 60,000 women. Br J Cancer. 2005;92(9):1803–7.

Wirstrom T, Hilding A, Gu HF, Ostenson CG, Bjorklund A. Consumption of whole grain reduces risk of deteriorating glucose tolerance, including progression to prediabetes. Am J Clin Nutr. 2013;97(1):179–87.

Meyer KA, Kushi LH, Jacobs DR Jr., Slavin J, Sellers TA, Folsom AR. Carbohydrates, dietary fiber, and incident type 2 diabetes in older women. Am J Clin Nutr. 2000;71(4):921–30.

Montonen J, Knekt P, Jarvinen R, Aromaa A, Reunanen A. Whole-grain and fiber intake and the incidence of type 2 diabetes. Am J Clin Nutr. 2003;77(3):622–9.

Aune D, Chan DS, Lau R, Vieira R, Greenwood DC, Kampman E, et al. Dietary fibre, whole grains, and risk of colorectal cancer: systematic review and dose-response meta-analysis of prospective studies. BMJ. 2011;343:d6617.

Zheng P, Barber R, Sorensen RJD, Murray CJL, Aravkin AY. Trimmed constrained mixed effects models: formulations and algorithms. J Comput Graphical Stat. 2021;30(3):544–56.

Akbaraly TN, Sabia S, Shipley MJ, Batty GD, Kivimaki M. Adherence to healthy dietary guidelines and future depressive symptoms: evidence for sex differentials in the Whitehall II study. Am J Clin Nutr. 2013;97(2):419–27.

Maukonen M, Harald K, Kaartinen NE, Tapanainen H, Albanes D, Eriksson J, et al. Partial substitution of red or processed meat with plant-based foods and the risk of type 2 diabetes. Sci Rep. 2023;13(1):5874.

Hu Y, Willett WC, Manson JAE, Rosner B, Hu FB, Sun Q. Intake of whole grain foods and risk of coronary heart disease in US men and women. BMC Med. 2022;20(1):192.

Jacobs DR Jr., Andersen LF, Blomhoff R. Whole-grain consumption is associated with a reduced risk of noncardiovascular, noncancer death attributed to inflammatory diseases in the Iowa women’s Health Study. Am J Clin Nutr. 2007;85(6):1606–14.

Yang J, Du H, Guo Y, Bian Z, Yu C, Chen Y, et al. Coarse Grain Consumption and Risk of Cardiometabolic diseases: a prospective cohort study of Chinese adults. J Nutr. 2022;152(6):1476–86.

Steffen LM, Jacobs DR Jr., Stevens J, Shahar E, Carithers T, Folsom AR. Associations of whole-grain, refined-grain, and fruit and vegetable consumption with risks of all-cause mortality and incident coronary artery disease and ischemic stroke: the atherosclerosis risk in communities (ARIC) Study. Am J Clin Nutr. 2003;78(3):383–90.

Um CY, Campbell PT, Carter B, Wang Y, Gapstur SM, McCullough ML. Association between grains, gluten and the risk of colorectal cancer in the Cancer Prevention Study-II Nutrition Cohort. Eur J Nutr. 2020;59(4):1739–49.

Skeie G, Braaten T, Olsen A, Kyro C, Tjonneland A, Nilsson LM, et al. Whole grain intake and survival among scandinavian colorectal cancer patients. Nutr Cancer. 2014;66(1):6–13.

Egeberg R, Olsen A, Loft S, Christensen J, Johnsen NF, Overvad K, et al. Intake of wholegrain products and risk of colorectal cancers in the Diet, Cancer and Health cohort study. Br J Cancer. 2010;103(5):730–4.

Rautiainen S, Levitan EB, Orsini N, Åkesson A, Morgenstern R, Mittleman MA, et al. Total antioxidant capacity from diet and risk of myocardial infarction: a prospective cohort of women. Am J Med. 2012;125(10):974–80.

Wirström T, Hilding A, Gu HF, Östenson CG, Björklund A. Consumption of whole grain reduces risk of deteriorating glucose tolerance, including progression to prediabetes. Am J Clin Nutr. 2013;97(1):179–87.

Kyrø C, Tjønneland A, Overvad K, Olsen A, Landberg R. Higher whole-grain intake is Associated with Lower Risk of type 2 diabetes among Middle-aged men and women: the Danish Diet, Cancer, and Health Cohort. J Nutr. 2018;148(9):1434–44.

Swaminathan S, Dehghan M, Raj JM, Thomas T, Rangarajan S, Jenkins D, et al. Associations of cereal grains intake with cardiovascular disease and mortality across 21 countries in Prospective Urban and rural epidemiology study: prospective cohort study. BMJ. 2021;372:m4948.

Mizrahi A, Knekt P, Montonen J, Laaksonen MA, Heliovaara M, Jarvinen R. Plant foods and the risk of cerebrovascular diseases: a potential protection of fruit consumption. Br J Nutr. 2009;102(7):1075–83.

Kyro C, Tjonneland A, Overvad K, Olsen A, Landberg R. Higher whole-grain intake is Associated with Lower Risk of type 2 diabetes among Middle-aged men and women: the Danish Diet, Cancer, and Health Cohort. J Nutr. 2018;148(9):1434–44.

Vieira AR, Abar L, Chan DSM, Vingeliene S, Polemiti E, Stevens C, et al. Foods and beverages and colorectal cancer risk: a systematic review and meta-analysis of cohort studies, an update of the evidence of the WCRF-AICR continuous Update Project. Ann Oncol. 2017;28(8):1788–802.

Murray CJ, Ezzati M, Lopez AD, Rodgers A, Vander Hoorn S. Comparative quantification of health risks conceptual framework and methodological issues. Popul Health Metr. 2003;1(1):1.

Trepanowski JF, Ioannidis JPA, Perspective. Limiting dependence on Nonrandomized studies and improving randomized trials in Human Nutrition Research: why and how. Adv Nutr. 2018;9(4):367–77.

Ioannidis JPA. The challenge of reforming nutritional epidemiologic research. JAMA. 2018;320(10):969–70.

Yeh R, Nashed YSG, Peterka T, Tricoche X. Fast automatic knot Placement Method for Accurate B-spline curve fitting. Comput Aided Des. 2020;128.

San-Cristobal R, Navas-Carretero S, Martinez-Gonzalez MA, Ordovas JM, Martinez JA. Contribution of macronutrients to obesity: implications for precision nutrition. Nat Rev Endocrinol. 2020;16(6):305–20.

Thielecke F, Nugent AP. Contaminants in Grain-A Major Risk for Whole Grain Safety? Nutrients. 2018;10(9).

World Health Organization. Diet, nutrition and the prevention of chronic diseases: Report of a joint WHO/FAO expert consultation. 2018.

Collaborators GBD. Health effects of dietary risks in 195 countries, 1990–2017: a systematic analysis for the global burden of Disease Study 2017. Lancet. 2019;393(10184):1958–72.

Willett W, Rockstrom J, Loken B, Springmann M, Lang T, Vermeulen S, et al. Food in the Anthropocene: the EAT-Lancet Commission on healthy diets from sustainable food systems. Lancet. 2019;393(10170):447–92.

Wehrli F, Taneri PE, Bano A, Bally L, Blekkenhorst LC, Bussler W et al. Oat intake and risk of type 2 diabetes, Cardiovascular Disease and all-cause mortality: a systematic review and Meta-analysis. Nutrients. 2021;13(8).

Zhou BF, Stamler J, Dennis B, Moag-Stahlberg A, Okuda N, Robertson C, et al. Nutrient intakes of middle-aged men and women in China, Japan, United Kingdom, and United States in the late 1990s: the INTERMAP study. J Hum Hypertens. 2003;17(9):623–30.

Muthayya S, Sugimoto JD, Montgomery S, Maberly GF. An overview of global rice production, supply, trade, and consumption. Ann N Y Acad Sci. 2014;1324:7–14.

Download references

This study was supported by the Key Discipline of Zhejiang Province in Public Health and Preventative Medicine (First Class, Category A), Hangzhou Medical College.

Author information

Jing Wang and Yingjun Li contributed equally to this work.

Authors and Affiliations

Department of Epidemiology and Health Statistics, School of Public Health, Hangzhou Medical College, 481 Binwen Road, Hangzhou, 310053, China

Houpu Liu, Jiahao Zhu, Rui Gao, Lilu Ding, Ye Yang, Wenxia Zhao, Jing Wang & Yingjun Li

Department of Radiology, Key Laboratory of Cancer Prevention and Therapy, Tianjin Medical University Cancer Institute & Hospital, National Clinical Research Center for Cancer, Tianjin, China

Xiaonan Cui

Department of Epidemiology and Health Statistics, School of Public health, Tianjin Medical University, Tianjin, China

You can also search for this author in PubMed   Google Scholar

Contributions

All authors contributed to the study’s conception and design. Liu Houpu, Wang, Jing, and Li, Yingjun conceived the study, searched the literature and performed data extraction. Liu, Houpu and Zhu, Jiahao had the idea for the article, performed the main analysis and written the first manuscript. Gao Rui, Yang Ye and Zhao Wenxia conducted the sensitivity analysis and inspected the analysis results. Ding Lilu, Cui, Xiaonan, and Lu, Wenli provided statistical expertise. Wang, Jing and Li, Yingjun reviewed the original manuscript and critically revised the work. All authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Corresponding authors

Correspondence to Jing Wang or Yingjun Li .

Ethics declarations

Ethics approval and consent to participate.

Not applicable.

Consent for publication

Patient and public involvement, competing interests.

The authors declare no competing interests.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Liu, H., Zhu, J., Gao, R. et al. Estimating effects of whole grain consumption on type 2 diabetes, colorectal cancer and cardiovascular disease: a burden of proof study. Nutr J 23 , 49 (2024). https://doi.org/10.1186/s12937-024-00957-x

Download citation

Received : 13 December 2023

Accepted : 07 May 2024

Published : 14 May 2024

DOI : https://doi.org/10.1186/s12937-024-00957-x

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Whole grains
• Bayes Theorem
• Cardiovascular diseases
• Diabetes Mellitus, type 2
• Colorectal neoplasms

Nutrition Journal

ISSN: 1475-2891

• General enquiries: [email protected]

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 14 May 2024

Association of type 2 diabetes with family history of diabetes, diabetes biomarkers, mental and physical disorders in a Kenyan setting

• David M. Ndetei 1 , 2 , 3 ,
• Victoria Mutiso 2 , 3 ,
• Christine Musyimi 2 , 3 ,
• Pascalyne Nyamai 2 , 3 ,
• Cathy Lloyd 4 &
• Norman Sartorius 5  

Scientific Reports volume  14 , Article number:  11037 ( 2024 ) Cite this article

208 Accesses

Metrics details

• Endocrinology
• Risk factors

This study aimed to determine the degree of family relations and associated socio-demographics characteristics, clinical/physical and mental disorders in type 2 diabetes mellitus in a Kenyan diabetes clinic. This study was part of a large multicentre study whose protocol and results had been published. It took place at the outpatient diabetes clinic at a County Teaching and Referral Hospital in South East Kenya involving 182 participants. We used a socio-demographic questionnaire, the Hamilton Depression (HAM-D) and PHQ-9 rating scales for depression, the MINI International Neuropsychiatric Interview (MINI; V5 or V6) for DSM-5 diagnoses, the WHO-5 Well-being scale and Problem Areas in Diabetes Scale (PAID). We extracted from the notes all physical conditions. We enquired about similar conditions in 1st and 2nd degree relatives. Descriptive, Chi-square test, Fisher’s exact test, one way ANOVA, and Multinomial logistic regression analysis were conducted to test achievements of our specific aims. Of the 182 patients who participated in the study, 45.1% (82/182) reported a family history of diabetes. Conditions significantly ( p  < 0.05) associated with a degree of family history of diabetes were retinopathy, duration of diabetes (years), hypertension, and depressive disorder. On average 11.5% (21/182) scored severe depression (≥ 10) on PHQ-9 and 85.2% (115/182) scored good well-being (≥ 13 points). All DSM-5 psychiatric conditions were found in the 182 patients in varying prevalence regardless of relations. In addition, amongst the 182 patients, the highest prevalence was poor well-being on the WHO quality of life tool. This was followed by post-traumatic disorders (current), suicidality, and psychotic lifetime on DSM-5. The least prevalent on DSM-5 was eating disorders. Some type 2 diabetes mellitus physical disorders and depression have increased incidence in closely related patients. Overall, for all the patients, the prevalence of all DSM-5 diagnoses varied from 0.5 to 9.9%.

Similar content being viewed by others

Magnitude of depression and its associated factors among patients with diabetes mellitus at public hospitals in Southwest Ethiopia, 2021

Prevalence and factors associated with diabetes-related distress in type 2 diabetes patients: a study in Hong Kong primary care setting

Association between mental health comorbidity and health outcomes in type 2 diabetes mellitus patients

Introduction.

Family history is a non-modifiable risk factor for diabetes 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 . The risk of developing type 2 diabetes mellitus (T2DM) increases approximately two to four times when either or both parents have T2DM 5 . Between 60 and 68.8% of diabetes patients have at least one family member with a history of diabetes 2 , 6 . Paternal history is significantly associated with higher chances of having T2DM 1 . An early age onset of T2DM is more likely if a family member had also an early onset of diabetes 2 , 8 , 9 , 10 .

A positive family history of diabetes is associated with increased levels of obesity, impaired glucose tolerance, fasting triglycerides, hemoglobin AIc (HbA1c), insulin dose per kilogram, lower levels of high-density lipoprotein cholesterol 3 , 8 , 11 , 12 , a greater waist to hip ratio as well as greater body mass index (BMI) 13 and a high prevalence of diabetes complications, particularly retinopathy and dyslipidemia compared to those without a relative with diabetes 9 . More specifically and in addition, there is an impact on leptin, (a hormone that regulates fat storage in the body) 14 , a high prevalence of hypertension, and lower low-density lipoprotein (LDL) cholesterol levels in those with fathers with T2DM as opposed to those with mothers with diabetes 15 . Various physical conditions are associated with diabetes. These include cardiovascular diseases 16 , 17 , hypertension 18 , thyroid abnormality, and diabetes complications such as retinopathy, neuropathy, and stroke 17 , 19 , 20 as well as high levels of biomarkers such as hemoglobin AIc (HbAIc) and cholesterol 21 .

Mental disorders such as schizophrenia, major depressive disorder, and bipolar disorder are associated with a family history of diabetes 22 , 23 , 24 . The risk of developing diabetes is three times higher in individuals with schizophrenia than in the normal population 23 . Siblings of schizophrenic parents are more likely to develop T2DM than those whose parents do not have schizophrenia 25 .

T2DM is also associated with anxiety, Post Traumatic Stress Disorder (PTSD), depression 26 , 27 , 28 , and eating disorders 29 , 30 , 31 . Research in this area is lacking in a Kenyan setting and is urgently required in order to inform clinical practice and potential community-based interventions.

Studies in African countries on the association between diabetes and family history have largely confirmed the global trends, showing an increased frequency of T2DM in persons with a family history of diabetes and an early onset of diabetes between 18 and 30 years 32 , 33 , 34 , 35 . Significantly higher blood glucose levels have been reported in those with a maternal family history of diabetes than in those without such a history 36 . Kenyan studies have found that people with T2DM are likely to have a positive family history specifically a first-degree relative and are also likely to develop diabetes early in life 37 , 38 . First-degree relations include an individual's biological parents, siblings, and children. Second-degree relatives include an individual's grandparents, grandchildren, uncles, aunts, nephews, nieces, and half-siblings. No study in Kenya has examined how physical conditions and mental disorders are comorbid in patients with T2DM or has examined the degree of family relations and how these vary with socio-demographics, measures of well-being, stress levels related to diabetes, and the prevalence of DSM-5 diagnoses in the Type 2 Diabetes. This information would inform an integrated approach to management. This study sought to fill these gaps.

The primary objective of this study was to determine the degree of family relations and associated socio-demographic characteristics, physical and mental disorders in people with T2DM. The secondary objective was to determine the overall prevalences of physical disorders and mental disorders in T2DM regardless of family relations.

The primary specific aims were:

To determine the relationships between social demographics in T2DM in different degrees of family relations

To determine the patterns of physical disorders and physical characteristics of T2DM in different degrees of family relations

To determine the mental health and disorders associated with T2DM in different degrees of family relations

To determine the independent predictors of T2DM in different degrees of family relations

The secondary specific aims were:

To determine the overall prevalence of physical disorders in T2DM

To determine the overall prevalence of mental disorders (stress, wellbeing, and psychiatric disorders) in T2DM

To determine the independent predictors of depression in T2DM

Study design and setting

This study was part of a larger multicentre study whose protocol has been published previously 39 . It took place between September 2013 and May 2015 at an outpatient diabetes clinic in one of the County Teaching and Referral Hospital in South East Kenya approximately 60 Kilometres from Nairobi. The clinic is run by a diabetologist and a team trained in diabetes management, offering psychoeducation, and counselling.

Study participants

Between September 2013 and May 2015, a sample of consecutive outpatient clinic attendees with T2DM were invited to participate in the study. Inclusion criteria were adults aged 18–65 with T2DM diagnosed at least 12 months earlier and able to give informed consent. Exclusion criteria included: communication and cognitive difficulties; life threatening or serious conditions in the previous 6 months and being an inpatient (as this may have indicated a serious condition); pregnant women or in the first 6 months post-partum clinic; substance use dependency or a current schizophrenic illness. All patients who met the inclusion criteria and did not have any exclusion criteria consented to the study and were included.

The trained research assistant completed a form that contained information from the medical records such as age, duration of diabetes, and presence/history of diabetes complications i.e. cardiovascular disease, retinopathy, peripheral neuropathy, peripheral vascular disease, and renal disease and associated disorders as well as the most recent measurement of blood pressure, HbA1c, height and weight.

For this study, we recorded the family history of T2DM in the following:

History of diabetes in 1st degree relatives (parent or sibling)

History of diabetes in 2nd degree relatives (grandparents, aunt, uncle, and cousin)

History of diabetes in both 1st and 2nd degree relatives

Study instruments

A standardised template for extracting information from the medical records on socio-demographic data and various medical complications known to be associated with T2DM, and laboratory indicators of T2DM was utilised. We also enquired about the history of smoking.

The following psychometric instruments were administered by a trained research assistant: (i) the Patient Health Questionnaire (PHQ-9), (ii) the Hamilton Depression (HAM-D) rating scale, (iii) the WHO-5 wellbeing scale, (iv) the Problem Areas in Diabetes Scale (PAID) and (v) the MINI International Neuropsychiatric Interview (MINI; V5 or V6). The psychometric properties of these instruments have been described in the protocol for this study 39 but also summarized here for quick reference. The PHQ-9 consists of 9 items on a 4-point Likert-type scale (0 = not at all; 1 = several days; 2 = more than half the days; 3 = nearly every day) with a total score ranging from 0–27. It has good psychometric properties and has been used extensively in many culturally diverse countries 40 . PHQ-9 scores with cut-off points of 1, 5, 10, 15, and 20 represent minimal, mild, moderate, moderately severe, and severe depression, respectively. Moderate to severe depressive symptomatology was defined as PHQ-9 scores >  10 , as this was a research study rather than clinical practice where a significant level of symptoms would usually be considered as PHQ-9 scores above 15 41 . The Hamilton Depression (HAM-D) Rating Scale has been considered a gold standard in depression studies and a preferred scale in the evaluation of depression treatment 42 .

It is the most widely employed depression scale on a global scale 43 and has been administered to various patient populations ranging from psychiatric, medical, and other research settings 44 . The HAM-D Rating Scale is a 17-item tool that takes 20–30 min to administer and scored between 0 and 4 points. Scores of 0–7 indicate normal, 8–16 indicate mild depression, 17–23 moderate depression, and counts over 24 are indicative of severe depression 42 . It has good psychometric properties with sufficient reliability (internal, inter-rater, and retest safety) and efficacy (convergent, discriminant, and predictive validity) 44 . The WHO-5 wellbeing scale is a 5-item questionnaire that measures a person’s overall psychological wellbeing 45 . The items are ‘I have felt cheerful and in good spirits’, ‘I have felt calm and relaxed’, ‘I have felt active and vigorous’, ‘I woke up feeling fresh and rested’, and ‘My daily life has been filled with things that interest me’. Poor wellbeing was defined as WHO-5 scores <  13 . The PAID is a 20-item questionnaire which measures the extent of diabetes-related emotional distress 46 . Items include ‘feeling overwhelmed with your diabetes’ and ‘feelings of guilt or anxiety when you get off track with your diabetes management’. Moderate-severe levels of diabetes-related distress are defined as scores (standardized to 100) >  40 46 . The MINI has been widely used in a range of different populations—including those with serious illnesses and in community surveys and is a reliable diagnostic tool according to DSM-V criteria 47 . It can be administrated by trained non-mental health specialists. Individuals diagnosed with depression (or other psychiatric disorders such as anxiety disorders) were advised to consult their physician for further assessment and treatment with a view to referral to the hospital psychiatric services. If any individual indicated suicidality (question 9 on the PHQ-9) immediate referral was made to the psychiatric service at the hospital.

Ethical consideration

Ethical approval was granted by the Kenyatta National Hospital—University of Nairobi (KNH-UoN) Ethics and Research Committee (ERC) (#P470/09/2013). All methods were performed in accordance with relevant guidelines and regulations as per the World Medical Association Declaration of Helsinki—ethical principles for medical research involving human subjects. Informed written consent was obtained from participants. For illiterate participants, informed written consent was obtained from their guardian/legally authorised representative.

Data analysis

This was performed using SPSS version 21 (IBM, Chicago, IL). All continuous variables were tested for normality using the Shapiro–Wilk test. Basic descriptive statistics in the form of frequency, percentage, mean, and standard deviation were carried out. The chi-square test or Fisher's exact test were used where appropriate to analyze the difference in the prevalence between family history of diabetes across different categories of socio-demographics, physical and mental disorder variables. Differences in levels of continuous variables were examined using the one way ANOVA for parametric data. Multinomial logistic regression was employed to identify the impact of a family history of diabetes on the risk factors of diabetes in the participants. Statistical significance was considered at p value < 0.05.

Socio-demographic characteristics

Table 1 summarizes the socio-demographic characteristics (frequencies and percentages) of the participants and the association between the degree of family history of diabetes and socio-demographic characteristics.

The mean age was 50.1 (± 11.1) years. The majority of respondents were female (74.2%), married/co-habiting (78.6%), had a regular income household (66.3%), were daily/weekly exercisers (74.6%) and non-smokers (90.7%), with the smallest proportion living in an urban area (18.1%) and the biggest proportion having access to health services (90.1%).

Of the 182 study participants, 45.1% (82/182) reported a family history of diabetes. The prevalence of diabetes in 1st degree relatives (parent, sibling) and 2nd degree relatives (grandparent, aunt, uncle, cousin) was 24.2% (44/182) and 12.1% (22/182) respectively; 8.8% (16/182) reported a family history of diabetes in both 1st degree and 2nd degree relatives.

The degree of family history of diabetes was not significantly ( p  > 0.05) associated with any socio-demographic variable.

Physical conditions and clinical characteristics in family relations

Table 2 summarizes the associations between the degree of family history of diabetes and physical conditions/clinical characteristics while Fig.  1 summarizes various physical conditions in descending prevalence.

Prevalence of the various physical conditions in descending order (N = 182).

The physical conditions significantly ( p  < 0.05) associated with the degree of family history of diabetes were retinopathy, duration of diabetes (years), and history of hypertension. The clinical characteristics significantly ( p  < 0.05) associated with the degree of family history of diabetes were HbA1C (%) and hypertension.

Mental disorders

Table 3 summarizes the association between the degree of family history of diabetes and mental disorders, mean scores of HAM-D, WHO-5 Well-being, PAID, and PHQ-9. It also summarizes the various DSM-5 diagnoses.

Only depressive symptoms (as measured by the HAM-D) were significantly ( p  = 0.030) associated with the degree of family history of diabetes. PHQ-9 unlike HAM-D did not reveal any significant trends ( p  > 0.05). All other measures were not significantly associated with a family history of diabetes ( p  > 0.05).

Independent predictors of T2DM in family relations

Table 4 summarizes the predictors of T2DM in different degrees of family relations.

Participants who had diabetes in both 1st and 2nd degree relatives had 6.28 increased odds of having retinopathy compared with participants who did not have a family history of diabetes. Diabetes in both 1st and 2nd degree relatives was associated with a higher duration of diabetes (years) and higher HbA1C (%).

Diabetes in 1st degree relatives was associated with higher HAM-D total scores.

PHQ-9 depression symptoms prevalence.

Figure  2 depicts the prevalence of various depression symptoms measured by PHQ-9.

Prevalence of PHQ-9 aspects in descending order (N = 182).

Most respondents had experienced profound fatigue or low energy levels, with over half indicating trouble with sleep patterns. Notably, a significant portion, comprising 15.40% of respondents, reported thoughts of being better off dead or of hurting themselves in some way.

Diabetes type 2 regardless of family relation

Table 5 summarizes the prevalence of the various aspects of mental health disorders as measured by the various instruments used in all the 182 patients attending the clinic, regardless of family relations. The prevalence of these various conditions is summarized in Fig.  3 in descending order. HAM-D was by far the most common mental health disorder while eating disorders (bulimia and anorexia) were the least with suicidality occupying the third position in the descending order, while elevated PAID was among the least.

Prevalence of HAM-D, poor WHO wellbeing, PHQ-9, PAID and DSM-5 mental disorders in descending order (N = 182).

Table 6 summarizes the independent predictors of depression in Diabetes. These predictors are diabetic foot problems, poor WHO-5 Wellbeing, and suicidality.

This report serves two main purposes: to provide context-appropriate evidence for Kenya to support the holistic and liaison approach to the management of T2DM and secondly to contribute to the global data pool by offering recommendations that can be replicated in similar contexts.

To our knowledge, this is the first Kenyan cohort study that reports different genetic loading (family history in different degrees of relations) and the significant independent predictors of T2DM and the associations between T2DM and socio-demographic characteristics, physical conditions, and mental disorders. As far as we were able to establish, this is not just a first for Kenya but also in Africa.

Family history

The finding of 45.1% of family history is lower than the reported 60–68.9% in the literature. This discrepancy could be attributed to the selection of the research participants in various studies. Ours was an outpatient clinic that excluded those admitted and presumably with severe forms of T2DM and possibly higher genetic loading. However, the finding of 45% is still significant for the Kenyan context, given that it is a non-modifiable contributor, hence the need for concerted efforts to focus on modifiable factors that are feasible in the Kenyan situation with limited resources, besides genetic counseling.

Social-demographics

There were no significant differences between a family history of diabetes and all the socio-demographic variables studied, nor was any socio-demographic variable a predictor of T2DM. It is noteworthy that smoking status was not associated with any type of T2DM family history. This could be a reflection of no history of smoking in the cohort studied, a practice that should be encouraged and no doubt the policy in Kenya to put social pressure against smoking and also counseling at the clinic. Another unexpected finding though not reaching a significant level was that of only 25% males of the total clinic patients. This could be explained as a gender preference to attend this public facility or a reflection of the differential gender prevalence of diabetes in the communities served by this public facility. A further possible explanation is a trend though not significant that the overwhelming majority (84–87%) of females, as opposed to 12.5% -22.7% of males, had a family history of T2DM. Mixed methods studies are required to explain these findings.

Physical conditions and biomarkers

Our study has shown that the higher the genetic loading the higher the association of retinopathy with T2DM in 1st and 2nd degree relatives compared with other levels of family history. Additionally, the highest association with diabetes in both 1st and 2nd degree relatives was found for the duration of diabetes in years, hypertension, and two specific biomarkers—HbAIC (%) and blood pressure (BP). BP and by extension hypertension can be easily monitored in the community, with the support of a relative, using easily available and affordable but reliable and valid BP monitors at home or the nearest health facility. This is an efficient way of monitoring and preventing T2DM, especially in those with a high genetic loading of diabetes. There is a new policy for every Kenyan family and all the individuals in that family to be reached at their homes on a regular basis by the newly created cadre of Primary Health promoters. They will not only attend to health promotion through awareness and attend to minor ailments but also take blood pressure. This community approach to monitoring blood pressure if successful is likely to have a critical impact on diabetes. Routine screening for blood pressure achieves extra significance given that 16.5% of our study patients were aged 60 + on age group. It is at this age group where various dementing conditions increase and hypertension is a risk factor for dementia 48 , 49 . The same principle applies to a routine determination of HbAIC (%) in those with the highest family loading of genetic risk for diabetes. In the Kenyan situation, blood samples for these can be taken at the nearest facility, and analysis carried out in that or the nearest available facility. Routine liaison consultation with the easily available ophthalmic clinical officers, (with the option to refer) for ophthalmoscopy is required for all patients with T2DM and more mandatory for patients with the highest genetic family history of T2DM in all diabetes clinics everywhere. Good history taking on the duration of diabetes is a routine practice that is reemphasized.

Even where there is no significant association with a family history of T2DM, our findings suggest there is a need for liaison practice, especially with renal and cardiology expertise. This expertise is usually but not always, available at all the 47 County Referral and Teaching hospitals in Kenya including the hospital where this study took place. While all physical conditions associated with T2DM were found in this cohort, only diabetic foot problems predicted depression. The holistic approach in that clinic could have mitigated other physical conditions as predictors of depression.

Although there was co-morbidity of diabetes with various mental disorders including alcohol abuse and dependence, WHO-5 wellbeing and diabetic stress, only depression, as determined by HAM-D was significant but less common in those with the highest level of genetic loading i.e., in both 1st and 2nd degree relatives.

Unlike HAM-D, PHQ-9 did not show any significant trends, suggesting the HAM_D scale is probably more sensitive and also the possibility that it is more valid than PHQ-9 in the type of patients we studied. While we do not have a conclusive explanation for this finding, we note that our sample size was small so no strong inferences could be made. Nevertheless, we venture a plausible explanation.

Firstly, if there are other family members with diabetes you are less likely to be depressed or anxious because there is support around you to help with your diabetes, therefore, less diabetes distress and more knowledge and understanding of diabetes.

However, we do not know whether individuals were living alone, an unlikely possibility in the Kenyan social-cultural context, if not they could still have family contacts through the still operational extended family and family social support systems in Kenya, though, this is diminishing towards nuclear centered families. It is also possible—that if there was a more laissez-faire attitude towards diabetes in relatives, then that might also lead to lower levels of anxiety and stress. On the other side, this attitude could at the same time lead to poorer glycemic control and so increased risk for microvascular disease. Either way, there are important implications for practice—screening for diabetes as well as depression, and improved knowledge of the risks of diabetes. The depression could be secondary to the onset of T2DM and most likely related to the burden of care in patients with T2DM.

The prevalence of various mental disorders found in this study was less than has been reported previously in the wider non-diabetic general clinical population in Kenya during a past study 50 .

Although there were no significant associations of all other types of mental disorders with a family history of T2DM, the high co-morbidity, ranging up to 13.6% and with a particular note of suicidality, calls for liaison with mental health experts in the management of T2DM. Apart from the findings on family relations, there are other incidental but clinically important findings. Of note is that although the association with psychotic conditions did not achieve significance, these psychotic conditions could negatively affect the overall management of T2DM. It is likely that the patients with these symptoms were treatment naive or not yet diagnosed and had therefore not received appropriate treatment for their psychosis. We therefore recommend routine screening for mental disorders using easily self-administered tests for all patients attending diabetes clinics. This self-screen is recommended because diabetologists are not necessarily experts in mental health and may not have the time to take a full history or make a diagnosis using a clinician-administered tool. Secondly, more importantly, the patients themselves may not be aware of, or may not feel able to report their mental health problems. Thirdly, joint management of diabetes and any mental disorder may have a better outcome for both conditions. This is feasible at local health center facilities, which are widely accessible at the community level, using stepwise upward referrals to the higher levels where there is the necessary expertise. Recommendations for treatment can then be provided using a stepwise downward referral process so that the patients can be managed in their communities. This will enhance the availability and accessibility of services and benefit capacity building in skills at the grassroots level.

The low-level prevalence of emotional stress (2.7%) does not allow us to test significant associations. While being diagnosed with diabetes can cause anxiety and depression and lead to emotional distress, the cause-effect could also be bi-directional—i.e. diagnosis leading to emotional distress or conversely emotional distress from other unidentified factors such as physical conditions leading to anxiety and distress. This calls for a qualitative approach that explores at a clinical level any directional relationship in a particular patient.

This finding of 2.7% prevalence of diabetes-related emotional distress is one of the lowest as compared to 12.8–46% reported in the literature 51 , 52 , 53 . We speculate that this is a reflection of the type of engagement of the patients that goes beyond the prescription of drugs in that particular clinic. It is the integrated management of diabetes that we speculate reduces emotional distress within a setting where the patients are fully educated on their conditions and management. It is likely that the levels of emotional distress would be similar to those reported in the literature for other situations and clinics that do not incorporate such holistic practices. If indeed that is the case, then it is a reflection of good practice in that specific clinic which could be replicated elsewhere.

Only combined methods—quantitative and qualitative have the potential to delineate these associations. Overall, our findings suggest the need for screening for depression, WHO-5 wellbeing, and suicidality in routine clinical management of T2DM at least in all patients with T2DM. Any positive screening findings should be integrated into the management of the patient.

Conclusions

Family relationships are important in both physical disorders and depression, suggesting shared genetic predisposition, and/or modulation by shared environmental factors. Depression emerges as the common mental disorder in individuals with Type 2 Diabetes, irrespective of relational factors. Additionally, all examined patients exhibited various mental health concerns and DSM-5 disorders. This Kenyan study contributes to the global database on the topic of Types of diabetes and family relations and associated mental and physical conditions. We have achieved all our aims for this study.

Based on all the achieved general and specific aims, we have suggested some clinical and community health practices and policies.

Limitations and recommendations to overcome the limitation

This study was carried out in a cohort of patients attending a diabetes clinic and therefore does not reflect the wider population of people with T2DM. This study excluded those untreated patients in the community or where clinics do not provide psychoeducation as in this clinic. Conversely, this holistic approach could be replicated in other clinics and contexts.

Secondly, we could not establish any directional relationships using the quantitative methods, given that our data is cross-sectional. Only mixed qualitative and quantitative methods could address this.

Although we achieved our aims, we recommend more studies at the community level to include those who may have T2DM and go for other services or are not treated by a specialist. Such a study though necessary for better understanding would be expensive and would require more complicated logistics.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Wang, C. et al. Association between parental history of diabetes and the incidence of type 2 diabetes mellitus differs according to the sex of the parent and offspring’s body weight: A finding from a Japanese worksite-based cohort study. Prev. Med. (Baltim.) 81 , 49–53 (2015).

Article   Google Scholar  

Gopalakrishnan, S. & Geetha, A. R. U. Study on the impact of family history of diabetes among type 2 diabetes mellitus patients in an urban area of Kancheepuram district, Tamil Nadu. Int. J. Community Med. Public Health 4 , 4151–4156 (2017).

Liu, Q., Yuan, J., Bakeyi, M., Li, J., Zhang, Z., Yang, X., et al. Development and validation of a nomogram to predict type 2 diabetes mellitus in overweight and obese adults: A prospective cohort study from 82938 adults in China. Int. J. Endocrinol. 2020 (2020).

Nagarathna, R., Bali, P., Anand, A., Srivastava, V., Patil, S., Sharma, G., et al. Prevalence of diabetes and its determinants in the young adults Indian population-call for yoga intervention. Front. Endocrinol. (Lausanne) . 846 (2020).

Papazafiropoulou, K. A., Papanas, N., Melidonis, A. & Maltezos, E. Family history of type 2 diabetes: Does having a diabetic parent increase the risk?. Curr. Diabet. Rev. 13 (1), 19–25 (2017).

Article   CAS   Google Scholar  

Sheu, W. H. et al. Family aggregation and maternal inheritance of Chinese type 2 diabetes mellitus in Taiwan. Zhonghua yi xue za zh= Chin. Med. J. Free China ed. 62 (3), 146–151 (1999).

CAS   Google Scholar  

Hemminki, K., Li, X., Sundquist, K. & Sundquist, J. Familial risks for type 2 diabetes in Sweden. Diabet. Care 33 (2), 293–297. https://doi.org/10.2337/dc09-0947 (2009).

Alharithy, M. K., Alobaylan, M. M., Alsugair, Z. O. & Alswat, K. A. Impact of family history of diabetes on diabetes control and complications. Endocr. Pract. 24 (9), 773–779 (2018).

Article   PubMed   Google Scholar  

Maghbooli, Z., Pasalar, P., Keshtkar, A., Farzadfar, F. & Larijani, B. Predictive factors of diabetic complications: A possible link between family history of diabetes and diabetic retinopathy. J. Diabet. Metab. Disord. 13 (1), 1–5 (2014).

Silverman-Retana, O. et al. Effect of familial diabetes status and age at diagnosis on type 2 diabetes risk: A nation-wide register-based study from Denmark. Diabetologia. 63 (5), 934–943 (2020).

Thorn, L. M. et al. Effect of parental type 2 diabetes on offspring with type 1 diabetes. Diabetes Care. 32 (1), 63–68 (2009).

Article   PubMed   PubMed Central   Google Scholar  

Tan, J. T., Tan, L. S. M., Chia, K. S., Chew, S. K. & Tai, E. S. A family history of type 2 diabetes is associated with glucose intolerance and obesity-related traits with evidence of excess maternal transmission for obesity-related traits in a South East Asian population. Diabet. Res. Clin. Pract. 82 (2), 268–275 (2008).

Parkkola, A., Turtinen, M., Härkönen, T., Ilonen, J. & Knip, M. Family history of type 2 diabetes and characteristics of children with newly diagnosed type 1 diabetes. Diabetologia. 64 (3), 581–590 (2021).

Article   CAS   PubMed   Google Scholar  

Koebnick, C. et al. Combined association of maternal and paternal family history of diabetes with plasma leptin and adiponectin in overweight Hispanic children. Diabet. Med. 25 (9), 1043–1048 (2008).

Article   CAS   PubMed   PubMed Central   Google Scholar  

Papazafiropoulou, A. et al. Familial history of diabetes and clinical characteristics in Greek subjects with type 2 diabetes. BMC Endocr. Disord. 9 (1), 1–7 (2009).

Shirey, K. et al. Symptoms of depression among patients attending a diabetes care clinic in rural western Kenya. J. Clin. Transl. Endocrinol. 2 (2), 51–54 (2015).

PubMed   PubMed Central   Google Scholar  

Atlas, D. International Diabetes Federation. IDF Diabetes Atlas 7th edn. (Belgium International Diabetes Federation, 2015).

Google Scholar  

Amankwah-Poku, M., Amoah, A. G. B., Sefa-Dedeh, A. & Akpalu, J. Psychosocial distress, clinical variables and self-management activities associated with type 2 diabetes: A study in Ghana. Clin. Diabet. Endocrinol. 6 (1), 1–10 (2020).

Kêkê, L. M. et al. Body mass index and childhood obesity classification systems: A comparison of the French, International Obesity Task Force (IOTF) and World Health Organization (WHO) references. Rev. Epidemiol. Sante Publique. 63 (3), 173–182 (2015).

Gahlan, D., Rajput, R., Gehlawat, P. & Gupta, R. Prevalence and determinants of diabetes distress in patients of diabetes mellitus in a tertiary care centre. Diabet. Metab. Syndr. Clin. Res. Rev. 12 (3), 333–336 (2018).

Rehman, S. U., Shakaib, A., Rashid, S., Gary, T. L. & Brancati, F. L. Regarding depressive symptoms and metabolic control in African–Americans with type 2 diabetes/Response to Rehman et al.. Diabet. Care 23 (10), 1596 (2000).

Van Welie, H. et al. The prevalence of diabetes mellitus is increased in relatives of patients with a non-affective psychotic disorder. Schizophr. Res. 143 (2–3), 354–357 (2013).

Lamberti, J., Crilly, J., Maharaj, K., Olson, D. & Costea, O. Prevalence of adult-onset diabetes among outpatients receiving antipsychotic drugs. Schizophr. Res. 1 (60), 360 (2003).

Su, M.-H. et al. Familial aggregation and shared genetic loading for major psychiatric disorders and type 2 diabetes. Diabetologia. 65 , 800–810 (2022).

Huang, M.-H. et al. Increased risk of type 2 diabetes among the siblings of patients with schizophrenia. CNS Spectr. 24 (4), 453–459 (2019).

Wong, H., Singh, J., Go, R. M., Ahluwalia, N. & Guerrero-Go, M. A. The effects of mental stress on non-insulin-dependent diabetes: determining the relationship between catecholamine and adrenergic signals from stress, anxiety, and depression on the physiological changes in the pancreatic hormone secretion. Cureus. 11 (8), e5474 (2019).

Van der Feltz-Cornelis, C. M. et al. Effect of interventions for major depressive disorder and significant depressive symptoms in patients with diabetes mellitus: A systematic review and meta-analysis. Gen. Hosp. Psychiatry 32 (4), 380–395 (2010).

Chaturvedi, S. K. et al. More anxious than depressed: prevalence and correlates in a 15-nation study of anxiety disorders in people with type 2 diabetes mellitus. Gen. Psychiatry. 32 (4), e100076 (2019).

Winston, A. P. Eating disorders and diabetes. Curr. Diabet. Rep. 20 (8), 1–6 (2020).

Harris, S. R., Carrillo, M. & Fujioka, K. Binge-eating disorder and type 2 diabetes: A review. Endocr. Pract. 27 (2), 158–164 (2021).

Alessi, J. et al. Mental health in the era of COVID-19: Prevalence of psychiatric disorders in a cohort of patients with type 1 and type 2 diabetes during the social distancing. Diabetol. Metab. Syndr. 12 (1), 1–10 (2020).

Asiimwe, D., Mauti, G. O. & Kiconco, R. Prevalence and risk factors associated with type 2 diabetes in elderly patients aged 45–80 years at Kanungu District. J. Diabet. Res. 6 , 1–9 (2020).

Endris, T., Worede, A. & Asmelash, D. Prevalence of diabetes mellitus, prediabetes and its associated factors in Dessie Town, Northeast Ethiopia: A community-based study. Diabet. Metab. Syndr. Obes. Targets Ther. 12 , 2799 (2019).

Omar, S. M., Musa, I. R., ElSouli, A. & Adam, I. Prevalence, risk factors, and glycaemic control of type 2 diabetes mellitus in eastern Sudan: A community-based study. Ther. Adv. Endocrinol. Metab. 10 , 2042018819860071 (2019).

Erasmus, R. T. et al. Importance of family history in type 2 black South African diabetic patients. Postgrad. Med. J. 77 (907), 323–325 (2001).

Chetty, R. R. & Pillay, S. Glycaemic control and family history of diabetes mellitus: Is it all in the genes?. J. Endocrinol. Metab. Diabet. S. Afr. 26 (2), 66–71 (2021).

El-busaidy, H., Dawood, M., Kasay, A., Mwamlole, C., Koraya, N. & Parpia, H. How serious is the impact of type II diabetes in rural Kenya? (2014).

Kiraka, G. N., Kunyiha, N., Erasmus, R. & Ojwang, P. J. Family history as a risk for early-onset type 2 diabetes in Kenyan patients. Afr. J. Diabet. Med. 22 (2), 15–17 (2014).

Lloyd, C. E. et al. The INTERPRET–DD study of diabetes and depression: A protocol. Diabet. Med. 32 (7), 925–934 (2015).

Kroenke, K., Spitzer, R. L. & Williams, J. B. W. The PHQ-9: Validity of a brief depression severity measure. J. Gen. Intern. Med. 16 (9), 606–613 (2001).

Petrak, F., Baumeister, H., Skinner, T. C., Brown, A. & Holt, R. I. G. Depression and diabetes: Treatment and health-care delivery. Lancet Diabet. Endocrinol. 3 (6), 472–485 (2015).

Hamilton, M. A rating scale for depression. J. Neurol. Neurosurg. Psychiatry 23 (1), 56 (1960).

Vindbjerg, E., Makransky, G., Mortensen, E. L. & Carlsson, J. Cross-cultural psychometric properties of the Hamilton Depression Rating Scale. Can. J. Psychiatry 64 (1), 39–46 (2019).

Rohan, K. J. et al. A protocol for the Hamilton Rating Scale for Depression: Item scoring rules, Rater training, and outcome accuracy with data on its application in a clinical trial. J. Affect. Disord. 200 , 111–118 (2016).

Newnham, E. A., Hooke, G. R. & Page, A. C. Monitoring treatment response and outcomes using the World Health Organization’s Wellbeing Index in psychiatric care. J. Affect. Disord. 122 (1–2), 133–138 (2010).

Polonsky, W. H. et al. Assessment of diabetes-related distress. Diabet. Care. 18 (6), 754–760 (1995).

Sheehan, D. V. et al. The Mini-International Neuropsychiatric Interview (MINI): The development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. J. Clin. Psychiatry 59 (20), 22–33 (1998).

PubMed   Google Scholar  

Musyimi, C. W., Muyela, L. A., Mutiso, V. N., Mutunga, E. & Ndetei, D. M. Understanding dementia care pathways for policy development and service planning in Kenya. Dementia. 22 , 1027–1037 (2023).

Comas-Herrera, A., Lorenz-Dant, K., Ferri, C., Govia, I., Sani, T., Jacobs, R., et al. Supporting people living with dementia and their carers in low-and middle-income countries during COVID-19. LTCcovid Org. 4 (2020).

Ndetei, D. M. et al. The prevalence of mental disorders in adults in different level general medical facilities in Kenya: A cross-sectional study. Ann. Gen. Psychiatry 8 (1), 1–8 (2009).

Lloyd, C. E. et al. Prevalence and correlates of depressive disorders in people with Type 2 diabetes: Results from the International Prevalence and Treatment of Diabetes and Depression (INTERPRET-DD) study, a collaborative study carried out in 14 countries. Diabet. Med. 35 (6), 760–769 (2018).

Perrin, N. E., Davies, M. J., Robertson, N., Snoek, F. J. & Khunti, K. The prevalence of diabetes-specific emotional distress in people with Type 2 diabetes: A systematic review and meta-analysis. Diabet. Med. 34 (11), 1508–1520 (2017).

Ramkisson, S., Pillay, B. J. & Sartorius, B. Diabetes distress and related factors in South African adults with type 2 diabetes. J. Endocrinol. Metab. Diabet. S. Afr. 21 (2), 35–39 (2016).

Download references

Acknowledgements

All the study participants and the staff of Machakos County Referral and Teaching Hospital Diabetes Clinic, the Association for the Improvement of Mental Health Programs, Switzerland, and the Africa Mental Health Research and Training Foundation.

Author information

Authors and affiliations.

Department of Psychiatry, University of Nairobi, Nairobi, Kenya

David M. Ndetei

Africa Mental Health Research and Training Foundation, Mawensi Road, Off Elgon Road, Mawensi Garden, P.O. Box 48423-00100, Nairobi, Kenya

David M. Ndetei, Victoria Mutiso, Christine Musyimi & Pascalyne Nyamai

World Psychiatric Association Collaborating Centre for Research and Training, Nairobi, Kenya

The Open University, Milton Keynes, UK

Cathy Lloyd

Association for the Improvement of Mental Health Programmes (AMH), Geneva, Switzerland

Norman Sartorius

You can also search for this author in PubMed   Google Scholar

Contributions

D.M.N.—conceptualization, literature review, drafting; V.M.—study implementation; C.M.—ethical issues, data quality control; P.N.—literature review and drafting; C.L.—conceptualization and assisted with data analysis; N.S.—conceptualization. All authors read and approved the final manuscript.

Corresponding author

Correspondence to David M. Ndetei .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Ndetei, D.M., Mutiso, V., Musyimi, C. et al. Association of type 2 diabetes with family history of diabetes, diabetes biomarkers, mental and physical disorders in a Kenyan setting. Sci Rep 14 , 11037 (2024). https://doi.org/10.1038/s41598-024-61984-6

Download citation

Received : 12 January 2024

Accepted : 13 May 2024

Published : 14 May 2024

DOI : https://doi.org/10.1038/s41598-024-61984-6

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Family relations
• Type 2 diabetes mellitus
• Physical conditions
• Kenyan setting

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Ozempic's Semaglutide Shows Promise in Reducing Kidney Disease Risk for Type 2 Diabetes Patients

Patients with type 2 diabetes who take semaglutide, the active ingredient in the popular drug Ozempic, may be less likely to develop complications from chronic kidney disease, according to a new study.

The Effects of Semaglutide on Kidney Diseases

Novo Nordisk published its research, which involved 3,533 patients from around the world, in the New England Journal of Medicine  and presented it at the European Renal Association meeting in Stockholm. It showed that patients who received semaglutide had a 24% lower risk of death from cardiovascular and kidney disease compared to those who received a placebo.

Vlado Perkovic, a kidney researcher at the University of New South Wales Sydney, said, "These results show a lot of promise for changing how we treat people who are at high risk of complications related to diabetes ." Novo Nordisk wants permission from the Food and Drug Administration (FDA) to add information about Ozempic's use in people with chronic kidney disease to the drug's label.

About 850 million people worldwide have chronic kidney disease. This change could help millions of patients, especially in the US, where more than 1 in 7 adults have the condition.

Novo Nordisk's executive medical director, Michael Radin, said the trial demonstrates their commitment to bettering the lives of those with renal disease and type 2 diabetes.

Semaglutide, a GLP-1 agonist, is a gastrointestinal hormone that regulates blood sugar and hunger. Ozempic was initially approved as a type 2 diabetic therapy in 2017.

In 2021, they repackaged it as Wegovy to help overweight or obese individuals who also suffer from another chronic medical condition lose weight. New research suggests that GLP-1 agonists also benefit the heart.

This has sparked more research on their possible therapeutic uses for conditions including addiction, sleep apnea, and Parkinson's disease. In March, the FDA gave Wegovy its all-clear to lower overweight individuals' heart disease risk.

READ ALSO: Ozempic Reduces Alcohol and Nicotine Cravings, Novo Nordisk to Investigate Further

Ozempic's Potential to Lower the Risk of Diseases

The study's findings showed that Ozempic and Wegovy could slow the development of chronic kidney disease  and lower the risk of major cardiovascular events, kidney failure, stroke, and death in the participants. Compared to the placebo group, semaglutide had a 24% lower risk of events related to kidney disease, an 18% lower risk of significant heart problems, and a 20% lower risk of death.

Based on the good results of the kidney disease study, Novo Nordisk plans to ask the FDA to expand the label later this year. If passed, these changes could make it much easier to care for people with diabetes and kidney disease.

Researchers are still looking into how semaglutide might help treat a lot of different illnesses, such as gout, liver disease, cancer, dementia, and Parkinson's disease. For people with type 2 diabetes and chronic kidney disease, the future looks bright.

The study is a big step forward in treating diabetes. It gives millions worldwide hope for better results and quality of life. As Novo Nordisk continues to try to get governmental approvals for semaglutide and find new uses for it, diabetes management is about to take a giant leap forward.

RELATED ARTICLE: Ozempic Babies: Do Semaglutide-Based Drugs Reduce Contraceptives' Effectiveness?

Check out more news and information on Medicine & Health  in Science Times.

Most Popular

What Dissolves Ear Wax Fast?

Neuralink's First User Admits He Wasn’t Aware Implant Thread Could Pull Out From Brain, Was Initially Worried But Not Anymore

World's First Head Transplant Machine Unveiled: BrainBridge Start-Up Reveals Ambitious Robotic System in New Spooky Video

NASA Turns On Psyche’s Electric Hall Thrusters

Nightmares Could Signal Early Onset of Brain Autoimmune Disorders, Study Finds

Latest stories.

Howler Monkeys in Mexico Are Dying, Dropping From Trees Due to Intense Heat Wave

Largest Planet-Forming Disk Resembles Giant Cosmic Butterfly and Hamburger-Like Sandwich

Singapore Airlines Incident: Could Climate Change Be Worsening Air Turbulence?

Search for Aliens: 5 Reasons We Still Don't Encounter Extraterrestrials

Car Accident Risk? Electric Vehicles Have Higher Likelihood of Hitting Pedestrians Compared to Petrol Cars, Study Reveals

Subscribe to the science times.

Sign up for our free newsletter for the Latest coverage!

Recommended Stories

China Announces Its Radar Technology for Space Weather Forecast After Releasing First Batch of Results

Pluto's Icy Exterior May Hide a Deep, Dense Underground Ocean, Study Suggests

Potentially Habitable Exoplanet Discovered Just 40 Light Years Away; Rare Exo-Venus World Possesses Earth-Like Temperature

SpaceX Launches New Batch of 23 Starlink Satellites, Expanding Global Internet Coverage

Diet Review: Ketogenic Diet for Weight Loss

Finding yourself confused by the seemingly endless promotion of weight-loss strategies and diet plans? In this series , we take a look at some popular diets—and review the research behind them .

What is it?

The ketogenic or “keto” diet is a low-carbohydrate, fat-rich eating plan that has been used for centuries to treat specific medical conditions. In the 19 th century, the ketogenic diet was commonly used to help control diabetes. In 1920 it was introduced as an effective treatment for epilepsy in children in whom medication was ineffective. The ketogenic diet has also been tested and used in closely monitored settings for cancer, diabetes, polycystic ovary syndrome, and Alzheimer’s disease.

However, this diet is gaining considerable attention as a potential weight-loss strategy due to the low-carb diet craze, which started in the 1970s with the Atkins diet (a very low-carbohydrate, high-protein diet, which was a commercial success and popularized low-carb diets to a new level). Today, other low-carb diets including the Paleo, South Beach, and Dukan diets are all high in protein but moderate in fat. In contrast, the ketogenic diet is distinctive for its exceptionally high-fat content, typically 70% to 80%, though with only a moderate intake of protein.

How It Works

The premise of the ketogenic diet for weight loss is that if you deprive the body of glucose—the main source of energy for all cells in the body, which is obtained by eating carbohydrate foods—an alternative fuel called ketones is produced from stored fat (thus, the term “keto”-genic). The brain demands the most glucose in a steady supply, about 120 grams daily, because it cannot store glucose. During fasting, or when very little carbohydrate is eaten, the body first pulls stored glucose from the liver and temporarily breaks down muscle to release glucose. If this continues for 3-4 days and stored glucose is fully depleted, blood levels of a hormone called insulin decrease, and the body begins to use fat as its primary fuel. The liver produces ketone bodies from fat, which can be used in the absence of glucose. [1]

When ketone bodies accumulate in the blood, this is called ketosis. Healthy individuals naturally experience mild ketosis during periods of fasting (e.g., sleeping overnight) and very strenuous exercise. Proponents of the ketogenic diet state that if the diet is carefully followed, blood levels of ketones should not reach a harmful level (known as “ketoacidosis”) as the brain will use ketones for fuel, and healthy individuals will typically produce enough insulin to prevent excessive ketones from forming. [2] How soon ketosis happens and the number of ketone bodies that accumulate in the blood is variable from person to person and depends on factors such as body fat percentage and resting metabolic rate. [3]

What is ketoacidosis?

There is not one “standard” ketogenic diet with a specific ratio of macronutrients ( carbohydrates , protein , fat ). The ketogenic diet typically reduces total carbohydrate intake to less than 50 grams a day—less than the amount found in a medium plain bagel—and can be as low as 20 grams a day. Generally, popular ketogenic resources suggest an average of 70-80% fat from total daily calories, 5-10% carbohydrate, and 10-20% protein. For a 2000-calorie diet, this translates to about 165 grams fat, 40 grams carbohydrate, and 75 grams protein. The protein amount on the ketogenic diet is kept moderate in comparison with other low-carb high-protein diets, because eating too much protein can prevent ketosis. The amino acids in protein can be converted to glucose, so a ketogenic diet specifies enough protein to preserve lean body mass including muscle, but that will still cause ketosis.

Many versions of ketogenic diets exist, but all ban carb-rich foods. Some of these foods may be obvious: starches from both refined and whole grains like breads, cereals, pasta, rice, and cookies; potatoes, corn, and other starchy vegetables; and fruit juices. Some that may not be so obvious are beans , legumes, and most fruits. Most ketogenic plans allow foods high in saturated fat, such as  fatty cuts of meat , processed meats, lard, and butter, as well as sources of unsaturated fats , such as nuts, seeds, avocados, plant oils, and oily fish. Depending on your source of information, ketogenic food lists may vary and even conflict.

• Strong emphasis on fats at each meal and snack to meet the high-fat requirement. Cocoa butter, lard, poultry fat, and most plant fats (olive, palm, coconut oil) are allowed, as well as foods high in fat, such as avocado, coconut meat, certain nuts (macadamia, walnuts, almonds, pecans), and seeds (sunflower, pumpkin, sesame, hemp, flax).
• Some dairy foods may be allowed. Although dairy can be a significant source of fat, some are high in natural lactose sugar such as cream, ice cream, and full-fat milk so they are restricted. However, butter and hard cheeses may be allowed because of the lower lactose content.
• Protein stays moderate. Programs often suggest grass-fed beef (not grain-fed) and free-range poultry that offer slightly higher amounts of omega-3 fats, pork, bacon, wild-caught fish, organ meats, eggs, tofu, certain nuts and seeds.
• Most non-starchy vegetables are included: Leafy greens (kale, Swiss chard, collards, spinach, bok choy, lettuces), cauliflower, broccoli, Brussels sprouts, asparagus, bell peppers, onions, garlic, mushrooms, cucumber, celery, summer squashes.
• Certain fruits in small portions like berries. Despite containing carbohydrate, they are lower in “net carbs”* than other fruits.
• Other: Dark chocolate (90% or higher cocoa solids), cocoa powder, unsweetened coffee and tea, unsweetened vinegars and mustards, herbs, and spices.

Not Allowed

• All whole and refined grains and flour products, added and natural sugars in food and beverages, starchy vegetables like potatoes, corn, and winter squash.
• Fruits other than from the allowed list, unless factored into designated carbohydrate restriction. All fruit juices.
• Legumes including beans, lentils, and peanuts.
• Although some programs allow small amounts of hard liquor or low carbohydrate wines and beers, most restrict full carbohydrate wines and beer, and drinks with added sweeteners (cocktails, mixers with syrups and juice, flavored alcohols).

*What Are Net Carbs? “Net carbs” and “impact carbs” are familiar phrases in ketogenic diets as well as diabetic diets. They are unregulated interchangeable terms invented by food manufacturers as a marketing strategy, appearing on some food labels to claim that the product contains less “usable” carbohydrate than is listed. [6] Net carbs or impact carbs are the amount of carbohydrate that are directly absorbed by the body and contribute calories. They are calculated by subtracting the amount of indigestible carbohydrates from the total carbohydrate amount. Indigestible (unabsorbed) carbohydrates include insoluble fibers from whole grains, fruits, and vegetables; and sugar alcohols, such as mannitol, sorbitol, and xylitol commonly used in sugar-free diabetic food products. However, these calculations are not an exact or reliable science because the effect of sugar alcohols on absorption and blood sugar can vary. Some sugar alcohols may still contribute calories and raise blood sugar. The total calorie level also does not change despite the amount of net carbs, which is an important factor with weight loss. There is debate even within the ketogenic diet community about the value of using net carbs.

Programs suggest following a ketogenic diet until the desired amount of weight is lost. When this is achieved, to prevent weight regain one may follow the diet for a few days a week or a few weeks each month, interchanged with other days allowing a higher carbohydrate intake.

The Research So Far

The ketogenic diet has been shown to produce beneficial metabolic changes in the short-term. Along with weight loss, health parameters associated with carrying excess weight have improved, such as insulin resistance, high blood pressure, and elevated cholesterol and triglycerides. [2,7] There is also growing interest in the use of low-carbohydrate diets, including the ketogenic diet, for type 2 diabetes. Several theories exist as to why the ketogenic diet promotes weight loss, though they have not been consistently shown in research: [2,8,9]

• A satiating effect with decreased food cravings due to the high-fat content of the diet.
• A decrease in appetite-stimulating hormones, such as insulin and ghrelin, when eating restricted amounts of carbohydrate.
• A direct hunger-reducing role of ketone bodies—the body’s main fuel source on the diet.
• Increased calorie expenditure due to the metabolic effects of converting fat and protein to glucose.
• Promotion of fat loss versus lean body mass, partly due to decreased insulin levels.

The findings below have been limited to research specific to the ketogenic diet: the studies listed contain about 70-80% fat, 10-20% protein, and 5-10% carbohydrate. Diets otherwise termed “low carbohydrate” may not include these specific ratios, allowing higher amounts of protein or carbohydrate. Therefore only diets that specified the terms “ketogenic” or “keto,” or followed the macronutrient ratios listed above were included in this list below. In addition, though extensive research exists on the use of the ketogenic diet for other medical conditions, only studies that examined ketogenic diets specific to obesity or overweight were included in this list. ( This paragraph was added to provide additional clarity on 5.7.18. )

• A meta-analysis of 13 randomized controlled trials following overweight and obese participants for 1-2 years on either low-fat diets or very-low-carbohydrate ketogenic diets found that the ketogenic diet produced a small but significantly greater reduction in weight, triglycerides, and blood pressure, and a greater increase in HDL and LDL cholesterol compared with the low-fat diet at one year. [10] The authors acknowledged the small weight loss difference between the two diets of about 2 pounds, and that compliance to the ketogenic diet declined over time, which may have explained the more significant difference at one year but not at two years (the authors did not provide additional data on this).
• A systematic review of 26 short-term intervention trials (varying from 4-12 weeks) evaluated the appetites of overweight and obese individuals on either a very low calorie (~800 calories daily) or ketogenic diet (no calorie restriction but ≤50 gm carbohydrate daily) using a standardized and validated appetite scale. None of the studies compared the two diets with each other; rather, the participants’ appetites were compared at baseline before starting the diet and at the end. Despite losing a significant amount of weight on both diets, participants reported less hunger and a reduced desire to eat compared with baseline measures. The authors noted the lack of increased hunger despite extreme restrictions of both diets, which they theorized were due to changes in appetite hormones such as ghrelin and leptin, ketone bodies, and increased fat and protein intakes. The authors suggested further studies exploring a threshold of ketone levels needed to suppress appetite; in other words, can a higher amount of carbohydrate be eaten with a milder level of ketosis that might still produce a satiating effect? This could allow inclusion of healthful higher carbohydrate foods like whole grains, legumes, and fruit. [9]
• A study of 39 obese adults placed on a ketogenic very low-calorie diet for 8 weeks found a mean loss of 13% of their starting weight and significant reductions in fat mass, insulin levels, blood pressure, and waist and hip circumferences. Their levels of ghrelin did not increase while they were in ketosis, which contributed to a decreased appetite. However during the 2-week period when they came off the diet, ghrelin levels and urges to eat significantly increased. [11]
• A study of 89 obese adults who were placed on a two-phase diet regimen (6 months of a very-low-carbohydrate ketogenic diet and 6 months of a reintroduction phase on a normal calorie Mediterranean diet) showed a significant mean 10% weight loss with no weight regain at one year. The ketogenic diet provided about 980 calories with 12% carbohydrate, 36% protein, and 52% fat, while the Mediterranean diet provided about 1800 calories with 58% carbohydrate, 15% protein, and 27% fat. Eighty-eight percent of the participants were compliant with the entire regimen. [12] It is noted that the ketogenic diet used in this study was lower in fat and slightly higher in carbohydrate and protein than the average ketogenic diet that provides 70% or greater calories from fat and less than 20% protein.

Potential Pitfalls

Following a very high-fat diet may be challenging to maintain. Possible symptoms of extreme carbohydrate restriction that may last days to weeks include hunger, fatigue, low mood, irritability, constipation, headaches, and brain “fog.” Though these uncomfortable feelings may subside, staying satisfied with the limited variety of foods available and being restricted from otherwise enjoyable foods like a crunchy apple or creamy sweet potato may present new challenges.

Some negative side effects of a long-term ketogenic diet have been suggested, including increased risk of kidney stones and osteoporosis, and increased blood levels of uric acid (a risk factor for gout). Possible nutrient deficiencies may arise if a variety of recommended foods on the ketogenic diet are not included. It is important to not solely focus on eating high-fat foods, but to include a daily variety of the allowed meats, fish, vegetables, fruits, nuts, and seeds to ensure adequate intakes of fiber, B vitamins, and minerals (iron, magnesium, zinc)—nutrients typically found in foods like whole grains that are restricted from the diet. Because whole food groups are excluded, assistance from a registered dietitian may be beneficial in creating a ketogenic diet that minimizes nutrient deficiencies.

Unanswered Questions

• What are the long-term (one year or longer) effects of, and are there any safety issues related to, the ketogenic diet?
• Do the diet’s health benefits extend to higher risk individuals with multiple health conditions and the elderly? For which disease conditions do the benefits of the diet outweigh the risks?
• As fat is the primary energy source, is there a long-term impact on health from consuming different types of fats (saturated vs. unsaturated) included in a ketogenic diet?
• Is the high fat, moderate protein intake on a ketogenic diet safe for disease conditions that interfere with normal protein and fat metabolism, such as kidney and liver diseases?
• Is a ketogenic diet too restrictive for periods of rapid growth or requiring increased nutrients, such as during pregnancy, while breastfeeding, or during childhood/adolescent years?

Bottom Line

Available research on the ketogenic diet for weight loss is still limited. Most of the studies so far have had a small number of participants, were short-term (12 weeks or less), and did not include control groups. A ketogenic diet has been shown to provide short-term benefits in some people including weight loss and improvements in total cholesterol, blood sugar, and blood pressure. However, these effects after one year when compared with the effects of conventional weight loss diets are not significantly different. [10]

Eliminating several food groups and the potential for unpleasant symptoms may make compliance difficult. An emphasis on foods high in  saturated fat  also counters recommendations from the Dietary Guidelines for Americans and the American Heart Association and may have adverse effects on blood LDL cholesterol. However, it is possible to modify the diet to emphasize foods low in saturated fat such as olive oil, avocado, nuts, seeds, and fatty fish.

A ketogenic diet may be an option for some people who have had difficulty losing weight with other methods.  The exact ratio of fat, carbohydrate, and protein that is needed to achieve health benefits will vary among individuals due to their genetic makeup and body composition. Therefore, if one chooses to start a ketogenic diet, it is recommended to consult with one’s physician and a dietitian to closely monitor any biochemical changes after starting the regimen, and to create a meal plan that is tailored to one’s existing health conditions and to prevent nutritional deficiencies or other health complications. A dietitian may also provide guidance on reintroducing carbohydrates once weight loss is achieved.

A modified carbohydrate diet following the Healthy Eating Plate model may produce adequate health benefits and weight reduction in the general population. [13]

• Low-Carbohydrate Diets
• David Ludwig clears up carbohydrate confusion
• The Best Diet: Quality Counts
• Other Diet Reviews
• Paoli A, Rubini A, Volek JS, Grimaldi KA. Beyond weight loss: a review of the therapeutic uses of very-low-carbohydrate (ketogenic) diets. Eur J Clin Nutr . 2013 Aug;67(8):789.
• Paoli A. Ketogenic diet for obesity: friend or foe?. Int J Environ Res Public Health . 2014 Feb 19;11(2):2092-107.
• Gupta L, Khandelwal D, Kalra S, Gupta P, Dutta D, Aggarwal S. Ketogenic diet in endocrine disorders: Current perspectives. J Postgrad Med . 2017 Oct;63(4):242.
• von Geijer L, Ekelund M. Ketoacidosis associated with low-carbohydrate diet in a non-diabetic lactating woman: a case report. J Med Case Rep . 2015 Dec;9(1):224.
• Shah P, Isley WL. Correspondance: Ketoacidosis during a low-carbohydrate diet. N Engl J Med . 2006 Jan 5;354(1):97-8.
• Marcason W. Question of the month: What do “net carb”, “low carb”, and “impact carb” really mean on food labels?. J Am Diet Assoc . 2004 Jan 1;104(1):135.
• Schwingshackl L, Hoffmann G. Comparison of effects of long-term low-fat vs high-fat diets on blood lipid levels in overweight or obese patients: a systematic review and meta-analysis. J Acad Nutr Diet . 2013 Dec 1;113(12):1640-61.
• Abbasi J. Interest in the Ketogenic Diet Grows for Weight Loss and Type 2 Diabetes. JAMA . 2018 Jan 16;319(3):215-7.
• Gibson AA, Seimon RV, Lee CM, Ayre J, Franklin J, Markovic TP, Caterson ID, Sainsbury A. Do ketogenic diets really suppress appetite? A systematic review and meta‐analysis. Obes Rev . 2015 Jan 1;16(1):64-76.
• Bueno NB, de Melo IS, de Oliveira SL, da Rocha Ataide T. Very-low-carbohydrate ketogenic diet v. low-fat diet for long-term weight loss: a meta-analysis of randomised controlled trials. Br J Nutr . 2013 Oct;110(7):1178-87.
• Sumithran P, Prendergast LA, Delbridge E, Purcell K, Shulkes A, Kriketos A, Proietto J. Ketosis and appetite-mediating nutrients and hormones after weight loss. Eur J Clin Nutr . 2013 Jul;67(7):759.
• Paoli A, Bianco A, Grimaldi KA, Lodi A, Bosco G. Long term successful weight loss with a combination biphasic ketogenic mediterranean diet and mediterranean diet maintenance protocol. Nutrients . 2013 Dec 18;5(12):5205-17.
• Hu T, Mills KT, Yao L, Demanelis K, Eloustaz M, Yancy Jr WS, Kelly TN, He J, Bazzano LA. Effects of low-carbohydrate diets versus low-fat diets on metabolic risk factors: a meta-analysis of randomized controlled clinical trials. Am J Epidemiol . 2012 Oct 1;176(suppl_7):S44-54.

Terms of Use

The contents of this website are for educational purposes and are not intended to offer personal medical advice. You should seek the advice of your physician or other qualified health provider with any questions you may have regarding a medical condition. Never disregard professional medical advice or delay in seeking it because of something you have read on this website. The Nutrition Source does not recommend or endorse any products.