Version July 2025
INTRODUCTION —
Type 2 diabetes mellitus is characterized by hyperglycemia, insulin resistance, and relative impairment in insulin secretion. Its pathogenesis is only partially understood, but is heterogeneous, and genetic factors affecting insulin release and responsiveness (as well as sociodemographic, clinical, and lifestyle factors) are important.
The prevalence of and nongenetic risk factors for type 2 diabetes will be reviewed here. The pathogenesis of type 2 diabetes, including genetic susceptibility, and the diagnostic criteria for diabetes are discussed elsewhere. (See "Pathogenesis of type 2 diabetes mellitus" and "Clinical presentation, diagnosis, and initial evaluation of diabetes mellitus in adults".)
PREVALENCE —
Diabetes is estimated to affect 828 million adults worldwide, with a global age-standardized prevalence of approximately 14 percent among adults [1]. Type 2 diabetes represents approximately 98 percent of global diabetes diagnoses, although this proportion varies widely among countries [2]. The prevalence of diabetes among adults in the United States is estimated at 14.7 percent (38.1 million adults) based on National Health and Nutrition Examination Survey (NHANES) data from 2017 to 2020; this includes 11.3 percent of individuals (29.4 million adults) with diagnosed diabetes and 3.4 percent (8.7 million adults) with undiagnosed diabetes [3,4]. Given the marked increase in childhood obesity, there is concern that the prevalence of diabetes will continue to increase substantially. Global data appear to substantiate this concern. Incident cases of diabetes rose by 39 percent from 1990 to 2019 globally [5]. In a separate study, the worldwide incidence rate of type 2 diabetes among adolescents and young adults (aged 15 to 39 years) rose from 117 to 183 per 100,000 population between 1990 and 2019 [6]. The incidence may have stabilized or even decreased since 2010, but whether the decrease is real or artifactual is unclear [7]. (See "Definition, epidemiology, and etiology of obesity in children and adolescents", section on 'Epidemiology'.)
The prevalence of diabetes is higher in certain populations [8,9]. As examples:
●Using data from a national survey for people aged 20 years or older, the prevalence of diagnosed type 2 diabetes in the United States (2018) was 7.5 percent in non-Hispanic White Americans, 9.2 percent in non-Hispanic Asian Americans, 12.5 percent in Hispanic Americans, 11.7 percent in non-Hispanic Black Americans, and 14.7 percent in Native Americans/Alaska Natives [8].
●In an analysis of data from the 2011 to 2014 Behavioral Risk Factor Surveillance System, the prevalence of self-reported diabetes was higher among Asian persons (9.9 percent) and Native Hawaiian or other Pacific Islander individuals (14.3 percent) than in White individuals (8 percent) [10].
●Outside the United States, type 2 diabetes is most prevalent in Polynesia and other Pacific Islands (approximately 25 percent) with similarly high rates in the Middle East and South Asia (Kuwait and Pakistan, in particular) [11,12]. In China, the most populous country in the world, an estimated 13 percent of the adult population has diabetes, with approximately one-half undiagnosed [13,14].
ABNORMAL GLUCOSE METABOLISM —
Abnormal glucose metabolism can be documented years before the onset of overt diabetes. Although the risk of developing type 2 diabetes follows a continuum across all levels of abnormal glycemia, it can be clinically helpful to classify individuals categorically as having impaired fasting glucose (IFG), impaired glucose tolerance (IGT), or a glycated hemoglobin (A1C) level of 5.7 to 6.4 percent (39 to 46 mmol/mol) (table 1) [15,16]. The criteria for defining diabetes and impaired glucose regulation are reviewed in greater detail separately. (See "Clinical presentation, diagnosis, and initial evaluation of diabetes mellitus in adults".)
Although the natural history of IFG and IGT is variable, approximately 25 percent of individuals with either will progress to diabetes over three to five years [15]. Individuals with isolated IFG predominantly have hepatic insulin resistance, whereas those with isolated IGT predominantly have muscle insulin resistance and normal or slightly reduced hepatic insulin sensitivity [15]. Individuals with abnormalities in both IFG and IGT have hepatic and muscle insulin resistance, which confers an even higher risk of progressing to diabetes compared with having only one abnormality. Individuals with additional diabetes clinical risk factors, including obesity and family history, are also more likely to develop diabetes. (See 'Clinical risk factors' below.)
Impaired glucose tolerance — The term IGT describes individuals who, during a 75 g oral glucose tolerance test (OGTT), have blood glucose values between those in people with normal glucose metabolism and those in people with overt diabetes (140 to 199 mg/dL [7.8 to 11 mmol/L]) (table 1). The glycemic levels that define diabetes were chosen based on associated risk of diabetes-specific retinopathy [17]. The rate of progression from IGT to overt diabetes varies among different populations. In six prospective studies, for example, the incidence rates of type 2 diabetes among patients with IGT ranged from 36 to 87 per 1000 person-years [18]. The rates were higher among Hispanic, Pima, and Nauruan people than among White people. Estimates of obesity (including body mass index [BMI], waist-to-hip ratio, and waist circumference) were positively associated with the incidence of type 2 diabetes. In contrast, sex and family history of diabetes were not related to the rate of progression in most studies.
Individuals who have only IGT generally do not develop clinically significant microvascular complications of diabetes such as retinopathy and nephropathy [19]. They are, however, at substantially increased risk (when compared with matched individuals with normal glucose tolerance) for developing macrovascular disease (eg, coronary artery disease). (See "Clinical presentation, diagnosis, and initial evaluation of diabetes mellitus in adults", section on 'A1C, FPG, and OGTT as predictors of cardiovascular risk'.)
Impaired fasting glucose — IFG is defined as a fasting blood sugar of 100 to 125 mg/dL (5.6 to 7 mmol/L) (table 1). IFG increases the risk of developing type 2 diabetes [20].
Although fasting glucose levels <100 mg/mL (5.55 mmol/L) are considered normal by the 2003 criteria of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus, individuals with fasting glucose values in the higher quintiles of the normal range are at increased risk for developing type 2 diabetes. In a prospective cohort study (over 46,500 participants followed for a mean of 81 months), the overall incidence of diabetes in those with normal fasting glucose was low (4 percent) [21]. However, those with fasting plasma glucose of 95 to 99 mg/dL (5.3 to 5.5 mmol/L) had increased risk of incident diabetes compared with those with fasting glucose <85 mg/dL (4.7 mmol/L; hazard ratio [HR] 2.33, 95% CI 1.95-2.79) [21].
Similar results were reported in a study of 13,163 healthy male Israeli army recruits [22]. Those with fasting plasma glucose levels >87 mg/dL (4.83 mmol/L) had a greater risk of diabetes compared with those in the lowest quintile with fasting glucose levels <81 mg/dL (4.5 mmol/L). The risk of diabetes was even greater (HR 8.23, 95% CI 3.6-19.0) in those with high-normal fasting glucose levels (91 to 99 mg/dL [5.1 to 5.5 mmol/L]) in combination with elevated serum triglycerides (>150 mg/dL) compared with participants with fasting glucose <86 mg/dL (4.8 mmol/L) and serum triglycerides <150 mg/dL; thus, a combination of parameters may identify individuals for whom preventive measures would be most effective.
Glycated hemoglobin — A1C measurements are also helpful in predicting diabetes (table 1) [16,23-25]. In a systematic review of 16 prospective studies examining the relationship between A1C and future incidence of diabetes mellitus, risk of diabetes increased sharply with A1C across the range of 5 to 6.5 percent (31 to 48 mmol/mol) [26]. For persons with A1C between 5.5 to 6.0 percent and 6.0 to 6.5 percent, the projected five-year risk of diabetes ranged from 9 to 25 and 25 to 50 percent, respectively. In the largest prospective cohort study of 26,563 women without diagnosed diabetes followed for 10 years, baseline A1C, at levels considered to be within the normal range, was an independent predictor of future type 2 diabetes [24]. In those individuals with baseline A1C in the highest quintile (A1C >5.22 percent [34 mmol/mol]), the adjusted relative risk (RR) of diabetes was 8.2, 95% CI 6.0-11.1, compared with participants in the lowest A1C quintile.
The international standardization of the A1C assay and biological and patient-specific factors (eg, low red cell turnover in iron deficiency anemia, rapid red cell turnover in patients treated with erythropoietin, hemoglobinopathies) that may cause misleading results are reviewed in detail elsewhere. (See "Measurements of chronic glycemia in diabetes mellitus", section on 'Glycated hemoglobin (A1C)'.)
SOCIODEMOGRAPHIC RISK FACTORS —
Extensive research has demonstrated that multiple sociodemographic factors are associated with the development of type 2 diabetes, including age, race and ethnicity, socioeconomic status, and other social determinants of health [27].
Age — Compared with individuals in all other age groups, adults aged ≥65 years have the highest prevalence of diabetes (29.2 versus 4.8 percent among those ages 18 to 44 and 18.9 percent among those 45 to 64 years) [3]. Aging increases the risk of developing type 2 diabetes due to both progressively impaired insulin secretion and increased insulin resistance, changes often mediated through obesity and sarcopenia [28]. In older adults, changes in diet and lower physical activity can lead to decreased lean body mass and increased adiposity, particularly visceral adiposity [28], contributing to insulin resistance.
Sex — Based on analysis of 2017 to 2020 data from the National Health and Nutrition Examination Survey (NHANES), men have a slightly higher age-adjusted prevalence of total diabetes (diagnosed and undiagnosed, 14.2 percent) compared with women (12.4 percent) [29]. However, women tend to have a higher prevalence of undiagnosed diabetes.
Race and ethnicity — Certain racial and ethnic populations in the United States are more likely to have diabetes compared with non-Hispanic White individuals. In an analysis of 2017 to 2020 data from the NHANES, the age-adjusted prevalence of total diabetes (using A1C or fasting plasma glucose) was 11.2 percent among non-Hispanic White, 16.8 percent among non-Hispanic Black, 16.4 percent among Asian, and 17.6 percent among Hispanic participants [29]. Similarly, data from the prospective Nurses' Health Study (NHS) collected over 20 years found that the risk for developing diabetes in women, corrected for body mass index (BMI), was higher for Asian, Hispanic, and Black Americans (relative risk [RR] 2.26, 1.86, and 1.34, respectively) compared with White Americans [30].
The racial and ethnic disparities in diabetes prevalence and incidence may be related in part to modifiable risk factors. As an example, in a retrospective analysis of a cohort study in 4251 Black and White young adults without diabetes at baseline (median follow-up 30 years), the racial disparity in diabetes risk was largely attenuated after adjustment for clinical risk factors (eg, BMI, waist circumference, blood pressure) [31]. These clinical factors are influenced by psychosocial and environmental determinants of behaviors and health care access, reviewed immediately below [32].
Socioeconomic status — Socioeconomic status is a multidimensional construct that includes educational attainment, economic status (eg, income), and occupation. Socioeconomic status influences access to material and social resources, including health care, housing, transportation, food, and political power and control [32]. Lower socioeconomic status has been consistently associated with a higher risk of developing diabetes [32,33]. For example, in an analysis using NHANES data from 2017 to 2018, the prevalence of total diabetes (diagnosed and undiagnosed) was 11.6 percent among individuals who attained a college or more advanced degree versus 19.6 percent among those attaining less than high school education [34]. In a separate analysis using National Health Interview Survey (NHIS) data from 2011 to 2014, age-standardized prevalence of physician-diagnosed diabetes was 5.0 percent among those attaining college education or higher versus 11.0 percent among those with less than high school education [35]. Further, diabetes prevalence was 5.5 percent among those with income >400 percent of federal poverty level versus 10.9 percent among those with income <100 percent of federal poverty level. These differences in diabetes prevalence have widened from 1999 to 2002 to 2011 to 2014.
In the United States, racial and ethnic disparities in socioeconomic status contribute to the racial and ethnic disparities in diabetes prevalence [32]. For example, Black Americans have lower educational attainment and double the unemployment rate compared with White Americans [32]. The 2019 median and mean family wealth levels of Black American families were reported to be less than 15 percent those of White families.
Neighborhood factors — Neighborhoods and their physical and social environments impact where people live, work, play, learn, worship, and age. Multiple cross-sectional studies have demonstrated connections between various neighborhood characteristics, such as green spaces, walkability, and food availability, and the prevalence of diabetes [27]. However, the cross-sectional nature of many of these studies precludes firm conclusions about causality.
Structural racism — Structural racism refers to "the totality of ways in which societies foster racial discrimination through mutually reinforcing systems of housing, education, employment, earnings, benefits, credit, media, health care and criminal justice" [36]. These systems, in turn, reinforce discriminatory values and behaviors and inequitable distribution of resources. Although examining the impact of structural racism on the development of diabetes is challenging, a study that used historic redlining as a proxy measure of structural racism found that redlining was associated with higher prevalence of diabetes at the census tract level [37].
CLINICAL RISK FACTORS
Family history — Compared with individuals without a family history of type 2 diabetes, those with a family history in any first-degree relative have a two- to three-fold increased risk of developing diabetes [38,39]. The risk of type 2 diabetes is higher (five- to sixfold) in those with both a maternal and paternal history of type 2 diabetes [38,39]. The risk is likely mediated through genetic, anthropometric (body mass index [BMI], waist circumference), and lifestyle exposures (diet, physical activity, smoking) factors. The genetics of type 2 diabetes is reviewed separately. (See "Pathogenesis of type 2 diabetes mellitus", section on 'Genetic susceptibility'.)
Obesity — The risk of impaired glucose tolerance (IGT) or type 2 diabetes rises with increasing body weight (figure 1) [40-44]. In an analysis of five National Health and Nutrition Examination Survey (NHANES) data sets spanning over thirty years, increase in BMI over time was the most important of the three covariates studied (age, race/ethnicity, BMI) for the increase in diabetes prevalence, accounting for approximately 50 percent of the increase in diabetes prevalence in males and 100 percent in females [45]. In addition, the NHS demonstrated an approximately 100-fold increased risk of incident diabetes over 14 years in nurses whose baseline BMI was >35 kg/m2 compared with those with BMI <22 kg/m2 [46].
The risk of diabetes associated with body weight appears to be modified by age. In a prospective cohort study of over 4000 adults >65 years of age, the risk of diabetes associated with the highest tertile in BMI was greater in individuals <75 years of age compared with those ≥75 years (hazard ratio [HR] 4.0 versus 1.9) [47].
Obesity acts at least in part by inducing resistance to insulin-mediated peripheral glucose uptake, which is an important component of type 2 diabetes, likely unmasking the part of the population with limited insulin secretion [48-50]. Reversal of obesity decreases the risk of developing type 2 diabetes and, in patients with established disease, improves glycemic management and can lead to remission. (See "Prevention of type 2 diabetes mellitus", section on 'Lifestyle intervention' and "Medical nutrition therapy for type 2 diabetes mellitus" and "Initial management of hyperglycemia in adults with type 2 diabetes mellitus", section on 'Weight management'.)
Fat distribution — The distribution of excess adipose tissue is another important determinant of the risk of insulin resistance and type 2 diabetes. The degree of insulin resistance and the incidence of type 2 diabetes are highest in those individuals with central or abdominal obesity, as measured by waist circumference or waist-to-hip circumference ratio (figure 2) [44,47,51,52]. Intra-abdominal (visceral) fat rather than subcutaneous or retroperitoneal fat appears to be of primary importance in this regard. Why the pattern of fat distribution is important and the relative roles of genetic and environmental factors in its development are not known [51,52]. (See "Obesity in adults: Prevalence, screening, and evaluation", section on 'Waist circumference' and "Obesity: Genetic contribution and pathophysiology", section on 'Body fat distribution'.)
Birth and childhood weight — An apparent U-shaped relationship exists between birth weight and risk of type 2 diabetes. This issue is discussed in detail elsewhere. (See "Pathogenesis of type 2 diabetes mellitus", section on 'Role of intrauterine development'.)
Above-average childhood BMI also is a risk factor for diabetes, independent of birth weight [53,54]. Remission of overweight or obesity before puberty appears to negate the risk. In a population-based study from Denmark, men who had overweight at seven years of age, but who had normal weight by 13 years of age (and remained with normal weight), had a similar risk of developing type 2 diabetes in adulthood as men who never had overweight as children or in early adulthood [54]. Remission of overweight after age 13 years but before early adulthood (17 to 26 years) was associated with increased risk, but risk was lower than that among men who had overweight at every age. (See "Epidemiology, presentation, and diagnosis of type 2 diabetes mellitus in children and adolescents", section on 'Risk factors'.)
Gestational diabetes — The risk for type 2 diabetes is higher in women who have had gestational diabetes [55-58]. These women have defects in both insulin secretion and insulin action, the severity of which correlates with the future risk of diabetes [55,56]. In a meta-analysis of observational studies, the cumulative incidence of type 2 diabetes in women with and without gestational diabetes was 16 and 2 percent, respectively, by 10 years (relative risk [RR] 8.09, 95% CI 4.34-15.08) [58]. (See "Gestational diabetes mellitus: Glucose management, maternal prognosis, and follow-up", section on 'Maternal prognosis'.)
Hypertension — High blood pressure is associated with risk of diabetes [59,60]. For example, in one large prospective cohort study, women with self-reported high-normal (130 to 139/85 to 89 mmHg) or elevated (≥140/90 mmHg or on antihypertensive therapy) blood pressure were at increased risk of developing diabetes compared with women with normal blood pressure (multivariate adjusted hazard ratios [HRs] 1.4, 95% CI 1.2-1.7 and 2.0, 95% CI 1.8-2.3 for high-normal and elevated blood pressure, respectively) [60]. The association persisted after adjustment for several metabolic dysfunction variables, such as BMI, hypercholesterolemia, age, exercise, smoking, and family history of diabetes. However, these results do not prove causality, and other potential mediators of the association (eg, insulin resistance, antihypertensive medications, or genetic polymorphisms linking endothelial dysfunction, immune function, and type 2 diabetes) were not included in the statistical analysis.
Cardiovascular disease — Heart failure and myocardial infarction appear to be associated with an increased risk of type 2 diabetes. In one study of 2616 patients with coronary artery disease, those with advanced heart failure (New York Heart Association [NYHA] class III) had nearly twice the risk of developing diabetes during 6 to 12 years of follow-up (17 versus 8 percent in NYHA class I patients; RR 1.7, 95% CI 1.1-2.6) [61]. Progressive body weight gain is an unlikely explanation, as weight loss is common in severe heart failure. (See "Heart failure: Clinical manifestations and diagnosis in adults".)
Similar findings were noted in a retrospective analysis of 8291 patients with myocardial infarction and without diabetes [62]. During a mean observation period of three years, 12 percent developed diabetes, representing an annual incidence rate of 3.7 percent compared with 0.8 to 1.6 percent in population-based cohorts. Independent predictors of diabetes included markers of metabolic dysfunction (BMI, hypertension, high triglycerides, low high-density lipoprotein [HDL], smoking) and medications (diuretics, beta-blockers, lipid-lowering drugs).
Obstructive sleep apnea — A bidirectional relationship is evident between obstructive sleep apnea (OSA) and type 2 diabetes. Thus, OSA increases risk of incident type 2 diabetes and is highly prevalent among individuals with existing diabetes. Risk of diabetes with OSA may increase with OSA severity and appears independent of BMI [63]. Nonetheless, studies evaluating the impact of OSA treatment on glucose metabolism in people with either prediabetes or type 2 diabetes have yielded inconsistent findings [63,64].
Hyperuricemia — Several prospective studies have found an association between higher levels of serum uric acid and an increased risk of developing type 2 diabetes [65-69]. After controlling for other diabetes risk factors (eg, BMI, alcohol consumption, smoking, physical activity) the elevated risk was attenuated but remained significant. Proposed mechanisms underlying this increased risk include endothelial dysfunction, oxidative stress, and insulin resistance [70]. Although the association is plausible, these observational studies do not prove causality.
Polycystic ovary syndrome — Polycystic ovary syndrome is associated with an increased risk for type 2 diabetes, independent of BMI, particularly in women with a first-degree relative with type 2 diabetes. This topic is reviewed separately. (See "Clinical manifestations of polycystic ovary syndrome in adults", section on 'IGT/type 2 diabetes'.)
Metabolic syndrome — Patients with metabolic syndrome, including those without hyperglycemia as a diagnostic element, are at particularly high risk for type 2 diabetes. (See "Metabolic syndrome (insulin resistance syndrome or syndrome X)", section on 'Risk of type 2 diabetes'.)
LIFESTYLE FACTORS
Diet
Dietary patterns — Dietary patterns affect the risk of type 2 diabetes mellitus. Consumption of red meat, processed and ultra-processed foods, and sugar-sweetened beverages is associated with an increased risk of type 2 diabetes in multiple studies [71-83]. In contrast, consumption of a diet high in fruits, vegetables, legumes, nuts, whole grains, and olive oil (such as the Mediterranean diet) is associated with a reduced risk [71-74,84-98]. Reduced risk of type 2 diabetes may also be associated with adherence to a healthy cardiac diet (high in cereal fiber and polyunsaturated fat, and low in trans fat and glycemic load) consisting of whole grain, fruits, vegetables, and low-fat dairy products combined with sodium restriction (such as the Dietary Approaches to Stop Hypertension [DASH] diet) [99]. A recent meta-analysis of large cohorts with over 100,000 participants found that consumption of foods with a high glycemic index was associated with an increased incidence of type 2 diabetes (relative risk [RR] between the lowest and highest quintiles: 1.27 [95% CI 1.21-1.34]) [100]. A similar association was seen between consumption of foods with a high glycemic load and type 2 diabetes (RR 1.15 [95% CI 1.09-1.21]). However, many of the studies linking diet to diabetes risk are observational and thus subject to confounding. In addition, most studies have used food frequency questionnaires to capture dietary patterns, but none of the specific foods examined can be considered in isolation. For example, higher meat intake is associated with more saturated fat intake, relatively lower fruit and vegetable intake, and frequently, higher body mass index (BMI). Although some lifestyle and dietary factors are considered in multivariable analysis, other unmeasured lifestyle or dietary factors may account for the findings in these observational studies [72,73,101-103]. (See "Medical nutrition therapy for type 2 diabetes mellitus" and "Healthy diet in adults" and "Prevention of type 2 diabetes mellitus", section on 'Lifestyle intervention'.)
Coffee consumption — Long-term coffee consumption may be associated with a decreased risk of type 2 diabetes [104-110]. In a systematic review of nine cohort studies (combined n = 193,473), compared with those with minimal coffee consumption (<2 cups per day), diabetes risk was lower in participants who drank >6 cups daily (RR 0.65, 95% CI 0.54-0.78) and those who consumed four to six cups daily (RR 0.72, 95% CI 0.62-0.83) [106]. Similar findings were demonstrated in a prospective study of over 88,000 young women in the Nurses' Health Study (NHS) that showed a dose-response relationship between increased coffee consumption and lower diabetes risk [111]. Associations were similar for noncaffeinated and caffeinated coffee. However, these observational data do not establish a cause-and-effect relationship, and UpToDate authors do not recommend increasing coffee intake as a diabetes prevention strategy.
Vitamins and micronutrients
Vitamin D — Several prospective observational studies have shown an inverse relationship between circulating 25-hydroxyvitamin D levels and risk of type 2 diabetes. Obesity is also associated with low 25-hydroxyvitamin D concentrations, and a relationship between vitamin D deficiency and type 2 diabetes may be confounded by obesity. This topic is reviewed in detail elsewhere. (See "Vitamin D and extraskeletal health", section on 'Diabetes' and "Prevention of type 2 diabetes mellitus", section on 'Vitamin D'.)
Selenium — Although animal models suggest that low doses of the antioxidant selenium may improve glucose metabolism, these findings have not been demonstrated in humans [112]. In an exploratory analysis of the Nutritional Prevention of Cancer trial, 1202 individuals who did not have diabetes at baseline were randomly assigned to selenium (200 mcg daily) or placebo and evaluated for incident type 2 diabetes [113]. After 7.7 years of follow-up, the cumulative incidence of diabetes was higher in those taking selenium than placebo (incidence 12.6 versus 8.4 cases per 1000 person-years, respectively, hazard ratio [HR] 1.55, 95% CI 1.03-2.33). Thus, selenium supplementation does not confer benefit and may increase the risk of type 2 diabetes. Potential mechanisms for this association are unknown, but may be related to the effects of selenium on glucagon (stimulatory) and insulin-like growth factor 1 (inhibitory) [114].
Iron — An association between serum ferritin levels [115,116], high iron intake [117], and type 2 diabetes has been reported, but the association is not well understood. Low iron diets are not recommended.
Chromium — Chromium deficiency is generally limited to hospitalized patients with increased catabolism and metabolic demands in the setting of malnutrition. Other patients at risk for chromium deficiency include patients with short bowel syndrome, burns, traumatic injuries, or those on parenteral nutrition without appropriate trace mineral supplementation. Associations have been suggested between low chromium levels and impaired glucose tolerance (IGT) and unfavorable lipid profiles. Clinical trials have shown conflicting results, but generally suggest that chromium supplementation may improve glycemia in individuals with diabetes but not those with normal glucose tolerance [118]. (See "Overview of dietary trace elements", section on 'Chromium'.)
Magnesium — In meta-analyses of prospective cohort studies, risk of type 2 diabetes was lower in adults with higher magnesium intake [90,119]. Sources of dietary magnesium include nuts, whole grains, and green leafy vegetables.
Physical activity — A sedentary lifestyle lowers energy expenditure, promotes weight gain, and increases the risk of type 2 diabetes [120]. Among sedentary behaviors, prolonged television watching is consistently associated with the development of obesity and diabetes [121].
Physical inactivity, even without weight gain, appears to increase the risk of type 2 diabetes. In a cohort study of Swedish men, low aerobic capacity and muscle strength at 18 years of age was associated with an increased risk of type 2 diabetes 25 years later, even among men with normal BMI [122]. In contrast, being physically active (including frequent walking) appears to be associated with a reduced risk of type 2 diabetes [123]. In a meta-analysis of 10 cohort studies, walking at faster speeds was associated with a graded decrease in the risk of type 2 diabetes [124]; compared with easy/casual walking, the RR of type 2 diabetes was 0.85 (95% CI 0.70-1.00) for average/normal walking, 0.76 (95% CI 0.65-0.87) for fairly brisk walking, and 0.61 (95% CI 0.49-0.73) for brisk/striding walking (>6.4 km/hour). Importantly, unmeasured or residual confounding may affect these findings; for example, physical activity levels may be associated with dietary patterns, smoking, or other lifestyle factors.
Physical activity of moderate intensity reduces the incidence of new cases of type 2 diabetes, regardless of the presence or absence of IGT. (See "Prevention of type 2 diabetes mellitus", section on 'Exercise'.)
Smoking — Several large prospective studies have raised the possibility that cigarette smoking increases the risk of type 2 diabetes [125-133]. In a meta-analysis of 25 prospective cohort studies, current smokers had an increased risk of developing type 2 diabetes compared with nonsmokers (pooled adjusted RR 1.4, 95% CI 1.3-1.6) [134]. The risk appears to be graded, with increasing risk as the number of cigarettes smoked per day and pack-year history rises. In one study, the risk was also increased for nonsmokers who had been exposed to secondhand smoke, compared with those who had not been exposed [131].
While a definitive causal association has not been established, a relationship between cigarette smoking and diabetes mellitus is biologically possible based on several observations:
●Smoking increases blood glucose concentration after an oral glucose challenge [135].
●Smoking may impair insulin sensitivity [136].
●Cigarette smoking has been linked to increased abdominal fat distribution and greater waist-to-hip ratio that may adversely affect glucose tolerance [137,138]. (See 'Fat distribution' above.)
The effect of smoking cessation on diabetes risk is variable and may depend on individual patient factors. Smoking cessation may reduce diabetes risk by reducing systemic inflammation. On the other hand, smoking cessation is often associated with weight gain, which will increase the risk of diabetes. (See "Pathogenesis of type 2 diabetes mellitus", section on 'Role of diet, obesity, and inflammation'.)
In an analysis of three cohort studies in the United States (mean follow-up 19.6 years), smoking cessation was associated with an increased risk of type 2 diabetes compared with continuing to smoke (HR 1.22, 95% CI 1.12-1.32) [139]. The risk peaked five to seven years after cessation and did not drop to that among individuals who had never smoked until 30 years after quitting. The increased risk of diabetes was directly proportional to weight gain. Nevertheless, those who stopped smoking had significantly lower rates of overall and cardiovascular mortality compared with current smokers, irrespective of weight gain.
Similar findings were noted in other cohort studies [140-142]. The increased risk of type 2 diabetes after smoking cessation does not outweigh the overall benefits. Smoking cessation efforts should be accompanied by additional lifestyle interventions, such as increasing physical activity and or making dietary changes to maintain or reduce body weight.
Alcohol – Excess alcohol use has been associated with increased risk of type 2 diabetes. In contrast, moderate alcohol intake (defined for females and males as <2 and <3 drinks per day, respectively) has been associated with a lower risk of type 2 diabetes. However, some of these studies may be limited by comparisons to current rather than lifelong nondrinkers who may include persons with former heavy alcohol use [143]. (See "Overview of the risks and benefits of alcohol consumption", section on 'Diabetes mellitus'.)
Sleep duration — Sleep quantity, quality, and chronotype may be associated with development of type 2 diabetes, but causality is uncertain [144,145]. In a meta-analysis of 10 prospective observational studies, compared with approximately eight hours of nightly sleep, short (≤5 to 6 hours nightly) and long (>8 to 9 hours nightly) duration of sleep were significantly associated with an increased risk of type 2 diabetes (RR 1.28 and 1.48, respectively) [146]. Similarly, in a large cohort of United Kingdom residents, moderately short (<5 hours) and extremely short (three to four hours) sleep duration compared with normal (seven to eight hours) sleep duration were each associated with type 2 diabetes (adjusted HR 1.16 [95% CI 1.05-1.28] and 1.41 [95% CI 1.19-1.68], respectively) [145]. Importantly, analyses were adjusted for BMI, physical activity levels, and general dietary habits.
Very limited data support a causal relationship between short sleep duration and development of diabetes. In a crossover study in 38 women aged 20 to 75 years with baseline sleep duration of seven to nine hours nightly, the effect of reduced sleep duration on insulin sensitivity was examined [147]. Participants underwent sequential, six-week phases of sleep maintenance (usual sleep time maintained) and sleep restriction (sleep time reduced by 1.5 hours nightly). Sleep restriction led to increases in fasting insulin concentration and homeostasis model assessment of insulin resistance, indicating diminished insulin sensitivity. These changes were independent of changes in adiposity and more pronounced in postmenopausal compared with premenopausal participants. Potential mechanisms linking short sleep duration and diabetes risk may include changes in appetite-regulating hormones, melatonin secretion, or behavioral changes associated with shorter sleep duration (eg, an increased time window for eating). In addition, obstructive sleep apnea (OSA) may shorten sleep duration and is known to increase the risk of type 2 diabetes. (See 'Obstructive sleep apnea' above and "Clinical presentation and diagnosis of obstructive sleep apnea in adults", section on 'Complications'.)
ENVIRONMENTAL EXPOSURES —
Epidemiologic studies have reported an increased risk of type 2 diabetes after exposure to some environmental toxins and contaminants [148-151]. As examples:
●Chronic exposure to inorganic arsenic in drinking water (adjusted odds ratio [OR] 3.58, 95% CI 1.18-10.83, for type 2 diabetes in individuals at the 80th versus the 20th percentiles for the level of total urinary arsenic) [152]. (See "Arsenic exposure and chronic poisoning".)
●Exposure to bisphenol A, a monomer used to make hard, polycarbonate plastics, and some epoxy resins (adjusted OR 1.39, 95% CI 1.21-1.60, per one standard deviation increase in urinary bisphenol A concentration) [153]. (See "Occupational and environmental risks to reproduction in females: Specific exposures and impact", section on 'Bisphenol A and other phenols'.)
●Chronic exposure to organophosphate and chlorinated pesticides (OR 1.17, 95% CI 0.99-1.38, for type 2 diabetes in those with the highest quartile of cumulative days of use compared with lowest quartile) [154].
MEDICATIONS —
A large number of drugs can impair glucose tolerance or cause overt diabetes mellitus; they act by decreasing insulin secretion, increasing hepatic glucose production, or causing resistance to the action of insulin (table 2). This topic is reviewed elsewhere. (See "Pathogenesis of type 2 diabetes mellitus", section on 'Drug-induced hyperglycemia'.)
OTHER
Breastfeeding — Breastfeeding has been associated with a decreased risk of maternal type 2 diabetes [155,156]. As an example, in two large cohorts from the Nurses' Health Study (NHS), with data collected prospectively in 83,585 parous women and retrospectively in 73,418, each additional year of lactation reduced the risk of diabetes in women who had been pregnant within the prior 15 years by 14 to 15 percent [155]. Risk reduction began to accrue with a minimum of six months of lactation, and longer durations of breastfeeding per pregnancy were associated with a greater benefit. In this study, the incidence of diabetes in women with a history of gestational diabetes was not affected by lactation. However, in a subsequent prospective study of women with recent gestational diabetes, breastfeeding reduced the two-year incidence of type 2 diabetes mellitus [157]. (See "Gestational diabetes mellitus: Obstetric issues and management", section on 'Breastfeeding'.)
Endogenous sex hormones — Levels of endogenous sex hormones may influence the risk of type 2 diabetes differently in males and females. A systematic review found that, after adjusting for body mass index (BMI), high testosterone levels were associated with an increased risk for type 2 diabetes in women but a decreased risk in men [158]. Decreased levels of sex hormone-binding globulin (SHBG) were associated with an increased risk for type 2 diabetes; this association was stronger in women than in men. In a subsequent study that included a genotype analysis, SHBG polymorphisms were associated with plasma levels of SHBG and were predictive of risk of type 2 diabetes in males and females [159]. Carriers of an rs6257 allele had lower plasma levels of SHBG and increased risk of type 2 diabetes, whereas carriers of an rs6259 variant allele had higher plasma levels and lower risk. Sex differences in the action of testosterone on lipolysis and insulin secretion have been demonstrated in preclinical studies.
PREDICTION MODELS —
Several diabetes-prediction models incorporate clinical risk factors and/or metabolic factors to generate a prediction score [160]. These models vary in complexity and most have not been validated in varied populations.
Simple clinical models may be more effective in predicting diabetes than complex models [161,162]. As an example, in the Framingham Offspring Study, several models to predict incident diabetes were compared [162]. The simple clinical model included information typically available at clinic evaluations, such as age, parental history of diabetes, body mass index (BMI), blood pressure, high-density lipoprotein (HDL), triglycerides, and impaired fasting glucose (IFG). Each of the metabolic syndrome traits (elevated blood pressure and triglyceride concentrations, low HDL levels, and IFG), obesity, and parental history were highly associated with developing diabetes. Adding more complex measurements (oral glucose tolerance, insulin sensitivity, insulin resistance) did not improve the model, nor did adding a genotype score based on the presence of a number of risk alleles confirmed to be associated with type 2 diabetes [163].
Machine learning analysis of UK Biobank data was used to identify the most significant factors that predict subsequent type 2 diabetes. The most important predictors included A1C, followed by BMI, waist circumference, blood glucose, family history of diabetes, gamma-glutamyl transferase, waist-hip ratio, HDL cholesterol, age, and urate [164]. These clinical and biological factors better predicted type 2 diabetes than diet, physical activity, or socioeconomic status.
In other models, the addition of genetic data to the simple clinical model (and other clinical models) had a minimal effect on prediction of type 2 diabetes [165,166]. In one such model, genetic data were incorporated based on low and high genetic risk groups (quintiles with the lowest and highest number of risk alleles, respectively) [165]. The improvement in prediction was too small to allow for individual risk prediction. Thus, at the current time, there is insufficient evidence to support genotyping for risk assessment in clinical practice. The genetics of type 2 diabetes, including a discussion of the associated risk alleles, are reviewed elsewhere. (See "Pathogenesis of type 2 diabetes mellitus", section on 'Genetic susceptibility'.)
SOCIETY GUIDELINE LINKS —
Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Diabetes mellitus in adults" and "Society guideline links: Diabetes mellitus in children".)
INFORMATION FOR PATIENTS —
UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5th to 6th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10th to 12th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.
Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)
●Basics topics (see "Patient education: Type 2 diabetes (The Basics)" and "Patient education: Treatment for type 2 diabetes (The Basics)" and "Patient education: Lowering your risk of prediabetes and type 2 diabetes (The Basics)")
●Beyond the Basics topics (see "Patient education: Type 2 diabetes: Overview (Beyond the Basics)" and "Patient education: Type 2 diabetes: Treatment (Beyond the Basics)" and "Patient education: Exercise and medical care for people with type 2 diabetes (Beyond the Basics)")
SUMMARY
●Abnormal glucose metabolism – Patients with impaired fasting glucose (IFG), impaired glucose tolerance (IGT), or a glycated hemoglobin (A1C) level of 5.7 to 6.4 percent (39 to 46 mmol/mol) are at increased risk of developing type 2 diabetes (table 1). Patients with both IFG and IGT have hepatic and muscle insulin resistance, which confers an increased risk of progressing to diabetes compared with having only one abnormality. Although most of the high-risk groups have been defined categorically (eg, IFG or IGT), the risk for developing diabetes follows a continuum across the entire spectrum of subdiabetic glycemic values. Higher fasting or two-hour oral glucose tolerance test (OGTT) values or higher A1C values convey higher risk than lower values. (See 'Abnormal glucose metabolism' above.)
●Sociodemographic risk factors – Extensive research has demonstrated that multiple sociodemographic factors are associated with the development of type 2 diabetes, including age, race and ethnicity, socioeconomic status, and other social determinants of health. (See 'Sociodemographic risk factors' above.)
●Clinical risk factors – Clinical conditions including cardiovascular disease, obstructive sleep apnea (OSA), and hypertension are associated with increased risk of incident type 2 diabetes. In females, a history of polycystic ovary syndrome or gestational diabetes also increases diabetes risk. (See 'Clinical risk factors' above.)
•Obesity – Obesity is the most important modifiable risk factor for type 2 diabetes. (See 'Obesity' above.)
•Family history – Compared with individuals without a family history of type 2 diabetes, those with a family history in any first-degree relative have a two- to three-fold increased risk of developing diabetes. Genetic susceptibility is an important contributor to the risk of developing diabetes. Insulin resistance and impaired insulin secretion in type 2 diabetes have a substantial genetic component. (See 'Family history' above and "Pathogenesis of type 2 diabetes mellitus", section on 'Genetic susceptibility'.)
●Lifestyle factors – Body weight, insulin resistance, and impaired insulin secretion can also be influenced by lifestyle factors, such as diet, physical activity, smoking, alcohol consumption, and sleep duration. (See 'Lifestyle factors' above.)
•Dietary patterns – Adherence to a diet high in fruits, vegetables, nuts, whole grains, and olive oil is associated with a lower risk of type 2 diabetes. (See 'Diet' above.)
•Physical activity – A sedentary lifestyle lowers energy expenditure, promotes weight gain, and increases the risk of type 2 diabetes. (See 'Physical activity' above.)
●Prevention – Identification of individuals at risk for diabetes is important as lifestyle modification (predominantly exercise and weight loss) successfully decreases the development of diabetes. Medical therapy for diabetes prevention also may be reasonable in selected patients. (See "Prevention of type 2 diabetes mellitus".)
ACKNOWLEDGMENT —
The UpToDate editorial staff acknowledges David McCulloch, MD, who contributed to earlier versions of this topic review.
