Diabetes mellitus has reached epidemic proportions, costing an estimated US$1.015 trillion in direct health expenditures worldwide (12% of global health expenditure). Although diabetes mellitus has a variety of forms with differing etiologies (e.g., autoimmune type 1 diabetes mellitus, type 2 diabetes mellitus, gestational diabetes mellitus, monogenic diabetes, drug-induced diabetes, diabetes due to pancreatic diseases or endocrinopathy ), most cases of diabetes mellitus—approximately 90% to 95%—are type 2 diabetes mellitus (hereafter referred to simply as diabetes ). Diabetes affects multiple systems of the body and can result in serious and debilitating complications. In this chapter, we provide an overview of diabetes and its complications, describe the global burden of diabetes, discuss the causal underpinnings of diabetes, and conclude by reviewing strategies to prevent and manage diabetes.
TYPE 2 DIABETES MELLITUS: DEFINITIONS AND OVERVIEW
The maintenance of glucose homeostasis requires appropriate levels of insulin secretion. Diabetes is characterized by a “progressive loss of adequate [pancreatic] β-cell insulin secretion frequently on the background of insulin resistance.” Impaired glucose tolerance and impaired fasting glucose are together termed prediabetes , a precursor to and a risk factor for diabetes. The diagnostic criteria for prediabetes and diabetes are shown in Table 1.1 . The fasting plasma glucose test, 2-h 75-g oral glucose tolerance test, and glycosylated hemoglobin A 1c (HbA 1c ) test are used to screen for diabetes. In the absence of typical symptoms (e.g., polyuria, polydipsia, weight loss), diabetes is diagnosed by two abnormal results from these screening tests.
Table 1.1
American Diabetes Association Diagnostic Criteria for Diabetes and Prediabetes
Data from American Diabetes Association Professional Practice Committee. 2. Diagnosis and Classification of Diabetes: Standards of Care in Diabetes—2024. Diabetes Care . 2023;47(suppl 1):S20–S42.
| Diagnostic Test | ||||
|---|---|---|---|---|
| Fasting Plasma Glucose | OGTT | HbA 1c | ||
| Type 2 diabetes | ≥126 mg/dL (7.0 mmol/L) | ≥200 mg/dL (11.1 mmol/L) | ≥6.5% (48 mmol/mol) | |
| Prediabetes | Impaired fasting glucose |
|
5.7–6.4% (39–47 mmol/mol) | |
| Impaired glucose tolerance |
|
|||
OGTT , Oral glucose tolerance test.
Diabetes is a heterogeneous condition that can manifest as a diversity of phenotypes. Individual differences in factors such as islet biology and autoimmunity, incretin activity, insulin resistance (IR), and adipose tissue distribution may contribute, to varying degrees, to the development of diabetes. There is growing interest in using a “precision medicine” approach to individualize diabetes management according to phenotypic characteristics. Some experts have proposed that diabetes may be subclassified into discrete subgroups (severe insulin-deficient diabetes [SIDD], severe insulin-resistant diabetes [SIRD], mild obesity-related diabetes, and mild age-related diabetes) based on age at diagnosis, body mass index (BMI), glutamic acid decarboxylase antibodies, β-cell function, and IR. The latter two are often expressed as Homeostasis Model of Assessment-β (HOMA-β) and Homeostasis Model of Assessment insulin resistance (HOMA-IR), which are measures estimated from fasting plasma glucose and C-peptide (residual fraction of proinsulin after its cleavage into insulin). These subgroups are associated with different prognoses (e.g., people with SIRD are more likely to develop kidney disease, while people with SIDD often require early insulin treatment and are at the risk of retinopathy) and highlight the heterogeneity of diabetes. Other epidemiological studies have indicated that HOMA-IR and HOMA-β are independently and additively associated with early onset of diabetes and early insulin requirement.
Long-term hyperglycemia and its fluctuations, especially in the presence of other cardiometabolic risk factors, notably obesity, hypertension, and dyslipidemia, can lead to microvascular and macrovascular changes affecting multiple body systems, resulting in fatal and nonfatal complications. The most recognized complications of diabetes include myocardial infarction, stroke, heart failure, peripheral arterial disease, atrial fibrillation, chronic kidney disease, retinopathy including visual loss, and diabetic neuropathy. Diabetes is also associated with other hitherto underappreciated complications, namely, infections (e.g., tuberculosis), liver diseases, mental illness, cancers, and cognitive impairment. For a more thorough discussion of these complications, see Chapter 6 , Chapter 7 , Chapter 8 , Chapter 9 , Chapter 10 , Chapter 15 , Chapter 22 , Chapter 24 , Chapter 25 , Chapter 27 .
In particular, diabetes is a risk factor for severe COVID-19 and associated mortality. The relationship between diabetes and COVID-19 may be mediated by hyperglycemia, inflammation, altered immune status, and renin-angiotensin-aldosterone system activation. In the Emerging Risk Factor Collaboration study of 820,900 people in 97 prospective cohorts, individuals with diabetes had a 1.3- to 3-fold increased risk of death from conditions ranging from cardiovascular diseases and kidney failure to infections, mental disorders, and liver disease. Due to their propensity for developing these life-threatening conditions, people with diabetes die earlier than those without. For example, a person diagnosed with diabetes at the age of 50 years might lose around 6 years of life, although the prognosis can be improved by managing multiple risk factors.
Although a loss of β-cell secretion to maintain normoglycemia is inherent to diabetes in general, people diagnosed with diabetes at a young age have an especially accelerated rate of β-cell function loss. Young-onset diabetes is often defined as diabetes diagnosed before the age of 40 years. People with young-onset diabetes have a more aggressive course and longer duration of disease, thus resulting in a greater risk of cardiovascular-kidney and other complications of diabetes than people diagnosed with diabetes later in life. The risk of complications is also greater than similarly aged people with type 1 diabetes, often due to concomitant cardiometabolic risk factors with delayed or suboptimal intervention. The incidence and prevalence rates of young-onset diabetes are particularly high among Asian, Black, and Indigenous populations in many countries, including the United States and Canada.
There is strong evidence that treatment of multiple risk factors simultaneously in patients with diabetes improves outcomes. Meta-analyses suggest that reducing HbA 1c by 0.9%, systolic blood pressure by 10 mm Hg, and low-density lipoprotein (LDL) cholesterol by 1 mmol/L might reduce the risk of cardiovascular disease and all-cause mortality by 10% to 20%. In the Swedish National Diabetes Registry, having a higher number of risk factors within target ranges (HbA 1c < 7%, systolic BP < 140 mm Hg, LDL < 2.5 mmol/L, high physical activity, no smoking) was associated with marked reduction in risk of mortality, myocardial infarction, and stroke in a graded manner. Clinical practice guidelines recommend that patients with diabetes undergo regular examination to detect silent complications (e.g., foot and eye examinations) and assess management of modifiable risk factors including lifestyle (e.g., physical activity, diet, smoking, sleep, emotional wellness) and cardiometabolic risk factors (e.g., blood glucose, HbA 1c , blood pressure, LDL cholesterol, triglycerides, body weight, kidney function including urinary albumin) and early use of organ-protective drugs (e.g., angiotensin-converting enzyme inhibitor or angiotensin receptor blocker, statins). The achievement of these goals may be optimized by integrating assessment, education, and care services in a structured, team-based setting. Furthermore, people at high risk of cardiovascular disease and those with established atherosclerotic cardiovascular disease (ASCVD), chronic kidney disease, or heart failure should be treated with a sodium-glucose cotransporter-2 inhibitor (if heart failure or chronic kidney disease predominates) or glucagon-like peptide-1 receptor agonist (if ASCVD predominates and especially in obese patients). Both of these drugs have been proven to prevent the onset or reduce the progression of these adverse outcomes.
GLOBAL BURDEN OF DIABETES
Prevalence and Incidence
The number of people with diabetes has sharply increased from 151 million (2000) to 58,937 million (2024) worldwide (age 20–79 years; Table 1.2 ). On current trends, the number of people with diabetes will rise to around 853 million by 2050. This increase is attributable to population growth and aging, as well as a rising burden of obesity. The prevalence rate of diabetes is defined as the proportion of a population with diabetes at a point in time. Globally, the prevalence rate of diabetes is 11.1% (age 20–79 years), and this figure is projected to rise to 13% by 2050. In the United States, there are around 38.5 million people with diabetes, and the prevalence rate is 15.7% (age 20–79 years). Over 80% of people with diabetes live in low- and middle-income countries (LMICs), such as China, India, and Pakistan ( Table 1.3 ). The standardized prevalence of diabetes (which uses statistical adjustment to account for differences in age distribution across populations) is highest in Pakistan (31.4%), Marshall Islands (25.7%), Kuwait (25.6%), and Samoa (25.4%). Globally, around two-thirds of people with diabetes reside in urban areas, which have a higher prevalence (12.7%) of diabetes than rural areas (8.8%). This urban-rural disparity is in part due to the rapid changes in environment and behaviors associated with urbanization and economic development.
Table 1.2
Numbers of People With Diabetes (in Millions), 2024 and 2050 (20–79 Years), Globally
Data from IDF Diabetes Atlas, 10th Edition Committee. IDF Diabetes Atlas . 10th ed. International Diabetes Federation; 2021.
| Region | 2024 | 2050 |
|---|---|---|
| Middle East and North Africa | 84.7 | 162.6 |
| North America and Caribbean | 56.2 | 68.1 |
| Western Pacific | 253.8 | |
| Southeast Asia | 106.9 | 184.5 |
| South and Central America | 35.4 | 51.5 |
| 65.6 | 72.4 | |
| Sub-Saharan Africa | 24.6 | 59.5 |
| World | 588.7 | 852.5 |
Table 1.3
Countries With the Highest Number of People With Diabetes (20–79 Years), 2024 and 2050
Data from IDF Diabetes Atlas, 11th Edition Committee. IDF Diabetes Atlas . 11th ed. International Diabetes Federation; 2025.
| Country | 2021 (Millions) | Country | 2045 (Millions) |
|---|---|---|---|
| China | 148.0 | China | 168.3 |
| India | 89.8 | India | 156.7 |
| United States | 38.5 | Pakistan | 70.2 |
| Pakistan | 34.5 | United States | 43.0 |
| Indonesia | 20.4 | Indonesia | 28.6 |
| Brazil | 16.6 | Egypt | 24.7 |
| Bangladesh | 13.9 | Brazil | 24.0 |
| Mexico | 13.6 | Bangladesh | 23.1 |
| Egypt | 13.2 | Mexico | 19.9 |
| Japan | 10.8 | Turkey | 14.1 |
A substantial proportion of people with diabetes are undiagnosed. This proportion ranges from around 28.9% in high-income countries to 45.5% to 58.7% in LMICs. Barriers to diagnosing diabetes include poor awareness among the public and healthcare providers, insufficient health system infrastructure, and lack of capacity or processes to detect diabetes. Delayed diagnosis increases the risk of presenting with advanced diabetes and severe complications.
The increasing prevalence of diabetes is an urgent public health issue with both societal and personal implications, calling for a multipronged strategy to reduce its burden. The prevalence rate is the proportion of people with diabetes among survivors at any one time. Thus in an aging society with a declining death rate, prevalence may continue to rise despite a falling number of new (incident) cases. Thus incidence rates are a preferred indicator to evaluate the impact of prevention strategies for controlling the epidemic of diabetes. The incidence rate of diabetes is defined as the number of new cases of diabetes divided by the person-time at risk over a particular period. An analysis of 23 high- and middle-income countries suggested that from 2010 to 2018, the incidence of diabetes had a declining trend in 19 of these countries (−1.1% to −10.8% per year), with an increasing trend in four of these countries (0.9%–5.6% per year). Incidence trends in LMICs have not been similarly studied due to insufficient data. Together with the declining trends of death rates among people with diabetes in high-income countries, these figures strongly demonstrate the preventable nature of diabetes and its complications.
Management
There are substantial gaps in the management of cardiometabolic risk factors in people with diabetes. In the United States, initial improvements in diabetes control observed from 1999 to 2010 have now stalled or partially reversed. The rate of achieving an HbA 1c <7% declined from 57.4% (2007–10) to 50.5% (2015–18), while the rate of attaining all glycemic, blood pressure, and LDL cholesterol targets has plateaued at 22.2% (2015–18). Young adults and people without health insurance were less likely to meet treatment targets. A study of 28 LMICs showed that out of all people with diabetes, only 44% were aware of their diagnosis, 28% were treated, and 23% had an HbA 1c <8%—a pattern known as the care cascade. In a study of 49 countries across Africa, the Middle East, South Asia, South and Central America, Asia, and Eastern Europe, the rate of achieving an HbA 1c <7% declined from 36.0% (2005) to 30.1% (2017). Barriers to attaining treatment targets include insufficient investment and capacity in health systems and suboptimal organization of diabetes care for implementing risk assessment and management programs to allow for early intervention. Other reasons include socioeconomic factors, high medication costs, complex treatment regimens, and inadequate patient empowerment with suboptimal self-management and poor medication adherence. In Hong Kong, a cosmopolitan city in South China with 7 million people, the introduction of a territory-wide diabetes risk assessment and management program in 2000 focusing on early intervention and patient empowerment was accompanied by progressive improvement in glycemic control against a backdrop of 50% to 70% decline in death and major cardiovascular event rates in 0.8 million people with diabetes in 2000 to 2016.
Outcomes
Data from high-income countries suggest that the incidence of cardiovascular-kidney complications and mortality in people with diabetes has fallen since 1990. In the United States, the incidence of myocardial infarction and stroke have each dropped by around two-thirds and one-half, respectively, from 1990 to 2010. However, the greatest declines were limited to people aged 75 years or older. Similar overall declines have been reported in other high-income countries, such as Australia, Canada, and Sweden. The incidence of end-stage kidney disease among people with diabetes has declined by around 28% in the United States (1990–2010) and by 6% per year in China (2000–12). Nevertheless, in recent years, the number of people with end-stage kidney disease has grown by 11% in the United States and by 40% to 700% in other regions, including Russia, the Philippines, Malaysia, Korea, Australia, Taiwan, and Scotland. This increase may be in part due to increasing survival among people with diabetes. In a study of 16 high-income countries from 1995 to 2016, all-cause mortality dropped in most countries by 0.5% to 4.2% per year, with the largest decreases in East Asian regions such as Hong Kong, South Korea, Taiwan, and Singapore.
In LMICs, there are limited data on the incidence of diabetes outcomes. It is estimated that the incidence of cardiovascular complications and premature mortality among people with diabetes is around two to three times higher than in high-income countries. In the 10 LMICs with the highest number of people with diabetes (China, India, Brazil, Mexico, Indonesia, Egypt, Pakistan, Bangladesh, Turkey, Thailand), diabetes may cause over 3 million premature deaths every year between the ages of 30 and 69 years. Nearly half of these deaths are attributable to cardiovascular causes.
CAUSAL UNDERPINNINGS OF TYPE 2 DIABETES
IR, often due to obesity, and impaired insulin secretion due to pancreatic β-cell dysfunction are crucial factors and occur early in the pathogenesis of diabetes. Under physiological conditions, IR leads to compensatory hyperplasia of pancreatic β-cells to promote insulin secretion for maintaining glucose homeostasis and preventing the progression from normal glucose tolerance to prediabetes and overt diabetes. Obesity and associated inflammation can induce IR, increase hepatic glucose production, and decrease peripheral glucose uptake in skeletal muscle and adipose tissue, leading to hyperglycemia.
In people with poor β-cell reserves due to factors such as genetic predisposition, this abnormal milieu of glucolipotoxicity and low-grade inflammation can impair pancreatic β-cell function and decrease insulin secretion, leading to hyperglycemia and eventually overt diabetes. Of note, β-cell mass in people with prediabetes and diabetes is approximately 60% and 40%, respectively, compared to people without prediabetes or diabetes. Persistent hyperglycemia in at-risk populations with prediabetes can further decrease β-cell mass and its functional capacity, thus making it harder for the body to maintain glucose homeostasis. In the past two decades, advances in diabetes research have confirmed that the pathogenesis of diabetes involves a complex interaction of genetic, epigenetic, and environmental factors.
Genetic, Epigenetic, and Gene-Environmental Interactions
Although the natural history of diabetes has been traditionally described as a process of increasing age-related IR followed by declining β-cell function and impaired insulin secretion, it has been postulated that people at risk for diabetes may have a primary genetic defect in β-cell function. In accordance with this hypothesis, family history is a strong and independent risk factor for diabetes that can interact with other lifestyle factors to bring about earlier development of diabetes at ages <40 years. Family history as a risk factor not only reflects genetic inheritance but also shares environment and health-related behaviors within families and their effects on previous generations ( Fig. 1.1 ). Nutritional or metabolic factors can lead to temporal changes in the expression of genetic factors (so-called metabolic imprinting), which can persist throughout life. The “thrifty phenotype hypothesis” describes how maternal undernutrition (as occurs during famine or in less severe food shortages) changes the intrauterine environment, leading to fetal adaptations that can affect the child throughout life. For the same mother, siblings born before and after the mother developed gestational diabetes had very different risks for glucose intolerance, despite sharing similar genetic profiles. In a similar vein, self-reported ethnicity or race is sometimes interpreted as a proxy for genetic ancestry. However, interethnic differences in culture and socioeconomic determinants may interact with other environmental, psychological, and behavioral factors to influence the risk of diabetes. This complexity is supported by the poor transferability of polygenic risk scores developed in European populations to other ethnic groups living in countries at different stages of economic development.
Complex gene-environment interactions in the pathogenesis of type 2 diabetes mellitus (T2DM).
In the presence of genetic predisposition, environmental factors such as nutrition, physical activity, chemicals, viruses, and sleep dysregulation can affect the transcriptional and regulatory processes (via methylation, chromatin modeling, or histone modifications), leading to the development and progression of T2DM. Several key environmental and health-related behaviors factors, candidate loci with evidence of gene-lifestyle interactions (such as MTIF3 , FADS , FTO , TBC1D4 , and MTNR1B ), and target organs that can affect glycemia and adipose tissue regulation are shown here.
From Franks, P.W., Poveda, A. Lifestyle and precision diabetes medicine: will genomics help optimise the prediction, prevention and treatment of type 2 diabetes through lifestyle therapy?. Diabetologia 60, 784–792 (2017). https://doi.org/10.1007/s00125-017-4207-5 .
With the advent of single- and multiancestry genome-wide association studies (GWAS), exome sequencing, and targeted array genotyping, a total of 237 genetic loci with 338 distinct association signals known to influence quantitative glycemic traits (including glucose, insulin, HbA 1c levels) and β-cell mass or function have been identified since 2005. These results indicate that diabetes is a polygenic disease and that the patterns of genetic variation may inform risk stratification and prognostication for implementing personalized medicine in the prevention and treatment for diabetes. Of note, known risk variants are generally common variants with a minor allele frequency of >0.05 with modest effect sizes. By combining these risk variants into a polygenic risk score, statistical power can be maximized to capture information on novel mechanistic insights into disease pathophysiology, clinical heterogeneity of disease predisposition and progression, and responses to targeted interventions. A polygenic risk score based on the findings of a large-scale GWAS estimated a 20% heritability for the risk of diabetes among Europeans, albeit with poor transferability to other ethnic groups with considerably lower explained variance. Thus there remain unresolved questions over the generalizability of such data to non-European populations and the extent to which genetic data adds to disease prediction apart from family history, clinical data, and sociodemographic information.
Even within the same family or same ethnic group, individuals exhibit variable susceptibility to diabetogenic environmental factors with varying outcomes depending on the individual genetic predisposition. This relationship is referred to as the “gene × environment (G×E) interaction.” The G×E interaction occurs throughout life and is well illustrated by considering the Pima Indians, a Native American population living in the Southwestern United States and Mexico. Compared with Pima Indians living in the United States, Pima Indians living in Mexico are more physically active (mean physical activity of 27.4 vs. 7.6 hours/week) and less obese (prevalence of obesity, 13.2% vs. 69.3%). Although both populations share a similar genetic predisposition for diabetes from an ethnic perspective, compared with Pima Indians living in the United States, Pima Indians living in Mexico have a lower prevalence of diabetes (8.0% vs. 37.5%). In the United Kingdom (UK) Biobank study, compared with those with low genetic risk and healthy behaviors, people with high genetic risk and healthy behaviors had a hazard ratio of 1.94 (95% confidence interval [CI], 1.30–2.90) for the risk of diabetes. The independent and additive effects of health-related behaviors and genetic factors are further evidenced by the progressive increase in hazard ratio to 6.27 (95% CI, 4.53–8.68) in people with high genetic risk and intermediate lifestyle and 15.46 (95% CI, 10.82–22.08) among those with high genetic risk and poor lifestyle.
The frequent clustering of multiple risk factors in people with or at risk of developing diabetes can have additive effects on one another on impairing multiple organ function. Since genetic factors cannot be modified, there is a need to monitor modifiable clinical markers (e.g., blood glucose, HbA 1c , blood pressure, lipid profile, BMI, waist circumference, kidney function including albuminuria, and estimated glomerular filtration rate) and health-related behaviors (e.g., diet, physical activity, sleep, mood, alcohol use, tobacco use) to delay the onset of diabetes and prevent complications. Together with other emerging risk factors such as pathway-specific biomarkers, the microbiome, and psychosocial factors, it may be possible to build a unique profile of each individual, describing the potential causes, trajectory, and consequences of diabetes to increase the precision of prediction, prevention, prognostication, and treatment of this condition.
Overweight, Obesity, and Associated Health-Related Behaviors
Overweight and obesity during childhood and adolescence can track into adult life, leading to early onset and worsening of associated cardiometabolic risk factors. Exposure to an abnormal metabolic milieu from childhood and adolescence can lead to hyperinsulinemia with progressive β-cell dysfunction and the development of young-onset diabetes and premature cardiovascular disease. Of note, there are interethnic differences in body fat distribution, with East and South Asians having increased visceral adiposity, which is closely associated with inflammation. Together with their propensity to have lower pancreatic β-cell function, East and South Asians have been shown to have lower glucose disposal rates and a higher prevalence of diabetes than non-Asians, given the same BMI.
Obesity and hyperglycemia interact to increase the risk of premature mortality. A meta-analysis of 230 prospective cohort studies reported a J-shaped relationship between BMI and all-cause mortality with the lowest risk observed among people with a BMI of 20 to 22 kg/m 2 . Among healthy never-smokers, every 5 kg/m 2 increment in BMI was associated with a 21% increased risk for all-cause mortality. In the Korean National Health Insurance Service cohort involving 15,149,275 adults, after a mean follow-up of 13.7 years, a similar J-shaped relationship among BMI, fasting plasma glucose, and all-cause mortality was also observed. Among people aged 18 to 44 years and using the overweight (BMI 25–27.4 kg/m 2 ) group with normoglycemia as the reference group, those with the lowest BMI <20 kg/m 2 and normoglycemia had a hazard ratio of 1.29 (95% CI, 1.25–1.33) for all-cause mortality. Using the same reference group, people with diabetes and overweight (BMI 25–27.4 kg/m 2 ) had a hazard ratio of 2.59 (95% CI, 2.39–2.82), and the leanest group (BMI < 20 kg/m 2 ) had a hazard ratio of 11.18 (95% CI, 10.20–12.25). The increased hazard ratio associated with low BMI might be confounded by other unmeasured factors, such as malnutrition or poor health, although BMI has been shown to be closely correlated with β-cell function, while central obesity is often considered a surrogate of visceral obesity with IR.
Excessive caloric intake and poor diet quality, such as low intake of dietary fiber and high intake of foods with high glycemic index, notably sugar-sweetened beverages and foods with high content of saturated and trans fats, increase the risk of obesity, diabetes, and associated cardiometabolic diseases. Consumption of a healthy diet with balanced nutrition has been shown to improve insulin action, reduce hepatic glucose production, and counteract the effect of obesity on the metabolic milieu. A high-quality diet, especially maintained over the long term, is associated with a 26% reduced risk for diabetes. This efficacy is augmented when combined with high levels of leisure-time exercise and is associated with a 45% risk reduction for diabetes. Both the DASH (Dietary Approaches to Stop Hypertension, a low-fat, high-fiber diet rich in vegetables, fruit, and low-fat dairy products) and Mediterranean-style diet (high intake of vegetables, beans, fruits, nuts, fish, and olive oils with a low consumption of meat, high-fat dairy products, and processed foods), supplemented with extra virgin olive oil or mixed nuts, can improve multiple cardiometabolic risk factors, as well as reduce the risk for diabetes and cardiovascular diseases. Increased physical activity is also associated with improved insulin sensitivity and cardiometabolic risk factors with a reduced risk of diabetes. Evidence on the efficacy of lifestyle interventions for the prevention of diabetes and its complications is further discussed in Chapter 3 .
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
