Global aetiology and epidemiology of type 2 diabetes mellitus and its complications

Abstract

Globally, the number of people with diabetes mellitus has quadrupled in the past three decades, and diabetes mellitus is the ninth major cause of death. About 1 in 11 adults worldwide now have diabetes mellitus, 90% of whom have type 2 diabetes mellitus (T2DM). Asia is a major area of the rapidly emerging T2DM global epidemic, with China and India the top two epicentres. Although genetic predisposition partly determines individual susceptibility to T2DM, an unhealthy diet and a sedentary lifestyle are important drivers of the current global epidemic; early developmental factors (such as intrauterine exposures) also have a role in susceptibility to T2DM later in life. Many cases of T2DM could be prevented with lifestyle changes, including maintaining a healthy body weight, consuming a healthy diet, staying physically active, not smoking and drinking alcohol in moderation. Most patients with T2DM have at least one complication, and cardiovascular complications are the leading cause of morbidity and mortality in these patients. This Review provides an updated view of the global epidemiology of T2DM, as well as dietary, lifestyle and other risk factors for T2DM and its complications.

Full text

The epidemic of diabetes mellitus and its complications
poses a major global health threat. The International
Diabetes Federation (IDF) estimated that 1 in 11 adults
aged 20–79years (415 million adults) had diabetes mel-
litus globally in 2015 (REF.1). This estimate is projected
to rise to 642 million by 2040, and the largest increases
will come from the regions experiencing economic
transitions from low-income to middle-income levels1.
However, these estimates might have under-represented
the true global burden of diabetes mellitus, especially in
regions undergoing rapid epidemiological transitions2.
The reasons for the escalating epidemic of diabetes mel-
litus are multiple, including population ageing, economic
development, urbanization, unhealthy eating habits
and sedentary lifestyles. Over 90% of diabetes mellitus
cases are type2 diabetes mellitus (T2DM)3,4. However,
types of diabetes mellitus are often not distinguished
in population-level estimates; therefore, in this Review,
the term diabetes mellitus refers to all types of diabetes
mellitus unless otherwise specified. Although the genetic
architecture might partially determine an individual’s
response to environmental changes5, the main drivers
of the global epidemic of T2DM are the rise in obesity,
a sedentary lifestyle, energy-dense diets and population
ageing6. Strong evidence indicates that many cases of
T2DM could be prevented by maintaining a healthy body
weight, following a healthy diet, exercising daily for at
least 30 min, avoiding smoking and consuming alcohol
in moderation7,8.
In this Review, we describe the global trends of T2DM
and its complications. We then discuss the roles of major
risk factors, in particular, obesity, lifestyle factors, genetic
predispositions, epigenetics and early developmental fac-
tors in the epidemic of T2DM and its complications. We
highlight evidence from landmark large-scale intervention
trials and longitudinal cohort studies from several coun-
tries and summarize recommendations for preventing
T2DM and its complications.
Global burden of T2DM
T2DM and its complications have contributed tremen-
dously to the burden of mortality and disability world-
wide. For instance, the Global Burden of Disease Study
2013 identified diabetes mellitus (all forms) as the ninth
major cause of reduced life expectancy9. In 2010, it was
estimated that diabetes mellitus caused 3.96 million death s
in adults aged 20–79years during that year (6.8% of
global mortality)10. This estimate was raised to 5.0 mil-
lion deaths due to diabetes mellitus and its complications
during 2015 in an IDF report, which is equivalent to
one death every six seconds1. The incidence of disability
caused by diabetes mellitus has increased substantially
1State Key Laboratory of
Genetic Engineering, School
of Life Sciences, Fudan
University, 2005 Songhu
Road, Shanghai, China.
2Department of Nutrition,
Harvard T.H.Chan School of
Public Health.
3Department of Epidemiology,
Harvard T.H.Chan School of
Public Health, 665
Huntington Avenue, Boston,
Massachusetts 02115, USA.
4Channing Division of
Network Medicine,
Department of Medicine,
Brigham and Women’s
Hospital and Harvard
Medical School, 181
Longwood Avenue, Boston,
Massachusetts 02115, USA.
Correspondence to F.B.H.
nhbfh@channing.harvard.edu
doi:10.1038/nrendo.2017.151
Published online 8 Dec 2017
Global aetiology and epidemiology
of type2 diabetes mellitus and
its complications
Yan Zheng1,2, Sylvia H.Ley2,4 and Frank B.Hu2,3,4
Abstract | Globally, the number of people with diabetes mellitus has quadrupled in the past three
decades, and diabetes mellitus is the ninth major cause of death. About 1 in 11 adults worldwide
now have diabetes mellitus, 90% of whom have type2 diabetes mellitus (T2DM). Asia is a major
area of the rapidly emerging T2DM global epidemic, with China and India the top two epicentres.
Although genetic predisposition partly determines individual susceptibility to T2DM, an
unhealthy diet and a sedentary lifestyle are important drivers of the current global epidemic;
early developmental factors (such as intrauterine exposures) also have a role in susceptibility to
T2DM later in life. Many cases of T2DM could be prevented with lifestyle changes, including
maintaining a healthy body weight, consuming a healthy diet, staying physically active, not
smoking and drinking alcohol in moderation. Most patients with T2DM have at least one
complication, and cardiovascular complications are the leading cause of morbidity and mortality
in these patients. This Review provides an updated view of the global epidemiology of T2DM,
as well as dietary, lifestyle and other risk factors for T2DM and its complications.
REVIEWS
88
|
FEBRUARY 2018
|
VOLUME 14 www.nature.com/nrendo

since 1990, with particularly large increases among
people aged 15–69years11. The Global Burden of
Diseases, Injuries, and Risk Factors Study 2015 estimated
that a high fasting level of glucose was the tenth most
common global risk factor for disability-adjusted life
years (DALYs) in 1990, fourth most common in 2005
and third most common in 2015, accounting for 143
million DALYs in 2015 and a 22% increase in DALYs
from 2005 to 2015 (REF.12).
The onset of diabetes mellitus frequently occurs years
before the actual diagnosis. Globally, 45.8% (or 174.8
million cases) of all diabetes mellitus cases in adults were
estimated to be undiagnosed13; people with undiagnosed
and untreated diabetes mellitus are at a greater risk of
complications than those who are receiving treatment.
Furthermore, medical expenditure for patients with dia-
betes mellitus is up to three times greater than for the
general population without diabetes mellitus14. The IDF
conservatively estimated that in 2015, US$673 billion
(12% of global health expenditure) was spent on treating
diabetes mellitus and its related complications1.
Globally, the number of people living with diabetes
mellitus quadrupled between 1980 and 2014 (REF.15).
Between 2010 and 2030, a 20% increase in the number
of adults with diabetes mellitus in developed countries
and a 69% increase in developing countries has been
predicted16. Asia has emerged as the major area with a
rapidly developing T2DM epidemic. China and India are
the top two epicentres of the global epidemic of T2DM1
(FIG.1). In these countries, the T2DM epidemic is char-
acterized by onset at a lower BMI and younger age than
in Western populations17.
In China, a large-scale population-based survey
was used to estimate that in 2010, >113.9 million
adults (11.6% of the adult population) had diabetes
mellitus and 493.4 million adults (50.1% of the total
population) had prediabetes mellitus (impaired glucose
tolerance, defined as 2-h oral glucose tolerance levels
7.8–11.0 mmol−l, and impaired fasting glucose, defined
as fasting glucose levels 6.1–6.9 mmol−l, according to
the WHO criteria)18. Less than one-third of those with
diabetes mellitus had been previously diagnosed; only
one-quarter of patients with diabetes mellitus had been
treated, and only 39.7% among those treated had blood
levels of HbA1c < 7.0%18. In India, a national study esti-
mated that 62 million individuals had diabetes mel-
litus and 77 million had prediabetes mellitus in 2011
(REF.19). The IDF estimates that India will have 69.2
million patients with diabetes mellitus in 2015, with a
projected rise to 123.5 million by 2040 (REF.1). Based
on data from India, China, Thailand and Malaysia,
the cost of inpatients with diabetes mellitus but no
complications accounted for 11–75% of per-capita
income in 2007, with inpatients who had complications
spending up to three times as much as those without
complications20.
The USA was listed as the country or territory with
the third-highest number of patients with diabetes melli-
tus in 2015 (REF.1) (FIG.1), and half of adults aged 65years
or older had prediabetes mellitus in 2008 (REF.21 ).
Furthermore, the North America and Caribbean region,
where the expenditure for diabetes mellitus per person
is 85-fold that in southeast Asia, spent more on diabe-
tes mellitus treatment than all other regions combined1.
The Pacific nations have a particularly high prevalence
of diabetes mellitus; >30% in American Samoa and 25%
in some other islands in Polynesia and Micronesia15. The
Middle East is another hot spot of the global diabetes
mellitus epidemic, with the prevalence of diabetes melli-
tus among adults ranging from 9.5% in Oman22 to 25.4%
in Saudi Arabia23. Despite a paucity of updated regional
data in Africa, IDF 2015 estimated a regional preva-
lence of 2.1–6.7% in sub-Saharan Africa1. According to
the Global Burden of Disease report, diabetes mellitus
was ranked as one of the leading causes of years of life
lost and has a major impact in Latin American coun-
tries24. On average, 25% of health expenditure in Latin
American countries is spent on treating diabetes melli-
tus and related complications, and the greatest economic
burdens were seen in Mexico and Brazil25. Notably, given
the variations in diagnostic methods and criteria used in
individual reports to identify T2DM, along with the lack
of national data in developing nations, all the current
estimates are likely to be imprecise and are probably an
underestimate of current disease burden26.
The available global estimates and predictions of
T2DM highlight the seriousness of the diabetes mellitus
pandemic; however, these estimates have limitations. In
the IDF report, direct nationwide data were lacking in
half of the countries, and their estimates were extrapo-
lated from other similar countries1. The accuracy and
reliability of such extrapolations might be questiona-
ble. In addition, the number of patients with diabetes
mellitus globally by 2015 (415 million) has already far
surpassed what had been predicted in 2000 for 2030 by
both the IDF (324 million) and the WHO (366 million)2.
Therefore, it is important to use these statistics cautiously
and critically, as they are probably underestimates.
With the rising prevalence of childhood obesity in
many countries, the prevalence of T2DM is increasing
in paediatric populations27. Children with T2DM tend
to develop complications in early adulthood27, which
Key points
• Globally, about 1 in 11 adults have diabetes mellitus (90% have type2 diabetes
mellitus (T2DM)), and Asia is the epicentre of this global T2DM epidemic.
• The major driving factors of the global T2DM epidemic include overweight and obesity,
sedentary lifestyle and increased consumption of unhealthy diets containing high
levels of red meat and processed meat, refined grains and sugar-sweetened beverages.
• Given its global influence, it is essential to break the vicious cycle of diabetes mellitus
begetting diabetes mellitus over generations by implementing effective strategies to
prevent gestational diabetes mellitus.
• Among patients with T2DM, cardiovascular complications are the leading cause of
morbidity and mortality, and kidney complications are highly prevalent in patients in
Asia with diabetes mellitus.
• Major clinical trials have demonstrated that diet and lifestyle modifications are
effective in preventing T2DM in high-risk individuals.
• T2DM management strategies including lifestyle modifications, social support and
ensuring medication adherence are key to reducing the incidence of diabetes
mellitus complications.
REVIEWS
NATURE REVIEWS
|
ENDOCRINOLOGY VOLUME 14
|
FEBRUARY 2018
|
89

places a substantial burden on the family and society.
Given its increasing prevalence, T2DM in childhood
has the potential to become a global public health
issue. Globally, the incidence and prevalence of T2DM
in children and adolescents, with data that is pre-
dominantly only available from developed countries,
were found to vary widely depending on ethnicity and
geographical region28. In the USA, the prevalence of
T2DM in children and adolescents has increased by
30.5% between 2001 and 2009 (REF.29), and its inci-
dence has increased 4.8% annually between 2002 and
2012 (REF.30). T2DM disproportionately affects youth
of ethnic minorities in the USA, such as Indigenous
American people, African-American people and
Hispanic people28. T2DM remains fairly uncommon in
children under 10years old, and most youth-onset cases
were found in adolescents (10–19years)31. Even so, data
from China suggest that the prevalence of childhood
T2DM has increased dramatically in the past two
decades32. In India, preliminary data from a national
registry of youth-onset diabetes revealed that 25% of
patients with diabetes mellitus who were <25years old
had T2DM33. In countries such as the USA, Canada
and Australia, the disproportionately higher incidence
of T2DM with a trend of earlier age at onset was evi-
dent among indigenous populations compared with
non-indigenous populations34–36.
Pathophysiology and major risk factors
When the feedback loops between insulin action and insu-
lin secretion do not function properly, the action of insulin
in insulin-sensitive tissues such as liver, muscle and adipose
tissue (insulin resistance in T2DM) and insulin secretion
by pancreatic islet β-cells (β-cell dysfunction in T2DM)
are affected, which results in abnormal blood levels of glu-
cose37 (FIG.2). In T2DM, insulin resistance contributes to
increased glucose production in the liver and decreased
glucose uptake in muscle and adipose tissue at a set insu-
lin level. In addition, β-cell dysfunction results in reduced
insulin release, which is insufficient for maintaining nor-
mal glucose levels38. Both insulin resistance and β-cell dys-
function occur early in the pathogenesis of T2DM, and
their critical importance has been verified longitudinally
in Pima Indian people progressing from normal glucose
tolerance to impaired glucose tolerance toT2DM39.
In the past three decades, advances in epidemiological
research on T2DM have improved our understanding of
a wide range of risk factors for the development of T2DM.
The determinants of T2DM consist of a matrix of genetic,
epigenetic and lifestyle factors (BOX1) that interact with
one another and operate within the larger physical–
sociocultural environment. Although individual predis-
position to T2DM has a strong genetic basis, evidence
from epidemiological studies suggests that many cases
of T2DM can be prevented with lifestyle modifications7,8.
Nature Reviews | Endocrinology
<100 thousand
100 to 500 thousand
0.5 to 1 million
1 to 10 million
10 to 20 million
>20 million
China –
109.6 million
1
India –
69.2 million
2
USA –
29.3 million
3
Figure 1 | Estimated total number of adults (20–79years) living with diabetes mellitus, highlighting the top three
countries or territories for number of adults with diabetes mellitus (20–79years) in 2015. It was estimated that in
2015, 415 million adults aged 20–79years had diabetes mellitus worldwide, and about 46.5% of them lived in three
countries: China, India and the USA. The colour of the country or territory in the map relates to the total number of adults
aged 20–79years living with diabetes mellitus in the area. Figure adapted with permission from REF.1, International
Diabetes Federation Diabetes Atlas. IDF Diabetes Atlas, 7th edn Brussels, Belgium: International Diabetes Federation,
2015 http://www.diabetesatlas.org.
REVIEWS
90
|
FEBRUARY 2018
|
VOLUME 14 www.nature.com/nrendo

Overweight and obesity
The prevalence of T2DM is increasing in parallel with
the escalating incidence of obesity in most developed
countries, such as the USA40, as well as in developing
countries, such as China18,41,42. By contrast, a substan-
tial reduction in the incidence of diabetes mellitus was
observed following a population-wide reduction in body
weight in the early 1990s in Cuba as a result of an eco-
nomic crisis43. Globally, the age-standardized prevalence
of obesity (defined as a BMI ≥30 kg/m2) increased from
3.2% in 1975 to 10.8% in 2014 in men and from 6.4% to
14.9% in women44. If these trends continue, the global
obesity prevalence is estimated to reach 18% in men
and surpass 21% in women by 2025 (REF.44). Excess adi-
posity, assessed by a high BMI, is the single strongest risk
factor for T2DM8,45 and is associated with many meta-
bolic abnormalities that result in insulin resistance46.
In the Nurses’ Health Study, 61% of the T2DM cases
could be attributed to overweight (defined as a BMI
≥25 kg/m2)8. Furthermore, abdominal obesity assessed
by waist circumference or waist-hip ratio predicts T2DM
risk independent of BMI47. Weight gain since young
adulthood, which occurs frequently and gradually dur-
ing the middle life stage, is another independent predic-
tor of T2DM48. In addition, visceral adiposity might be
an independent predictor for T2DM risk49.
In the USA, people of Asian descent are 30–50%
more likely to develop diabetes mellitus at a much
lower BMI than white people50. Such ethnic variations
could be attributed to different fat distributions and
percentages of body fat. For instance, Asian individuals
generally have a higher total body fat percentage at a
given BMI51 and higher visceral adiposity than white
people17. In addition, abdominal (or central) adiposity,
high levels of which increase the risk of T2DM, is highly
prevalent in Asian people52. Such ethnic heterogeneity
in pathophysiology in T2DM could be attributable to
the variations in both genetic background and pheno-
type; for example, Asian people without diabetes mel-
litus generally have poorer β-cell function than white
people without diabetes mellitus53.
At an individual level, treatment of obesity with weight
loss surgery (for example, bariatric surgery) has proven
effective in the prevention and resolution of T2DM54.
However, this approach is expensive and is unlikely to
reverse the current diabetes mellitus epidemic. Thus, pop-
ulation-level strategies for obesity prevention are critical.
To address the dual epidemics of obesity and diabetes mel-
litus, we need to consider the root causes of these diseases,
particularly unhealthy diet and lifestyle choices.
Diet and lifestyle factors
Diet and lifestyle modification is an important aspect
of T2DM prevention. Major clinical trials have demon-
strated that intensive lifestyle interventions can lower
the incidence of diabetes mellitus by 58% compared
with control groups55. Trials have also shown that
these interventions are more effective than pharmaco-
logical interventions55. Landmark clinical trials, such
as the Diabetes Prevention Program in multi-ethnic
Americans55, the Finnish Diabetes Prevention Study56
and the Da Qing IGT and Diabetes Study in China57,
have demonstrated that many cases of T2DM could be
prevented through lifestyle interventions focused on
increasing physical activity and adopting a healthy diet.
Nevertheless, when lifestyle interventions are not feasible,
pharmacological therapy can be considered as a strategy
to prevent the development of T2DM. For example, met-
formin reduced the incidence of T2DM by 31% over an
average follow-up period of 2.8years among high-risk
individuals from the USA who did not have diabetes
mellitus55. Similarly, metformin reduced T2DM risk in
clinical trials in India and China58.
As trial participants are generally high-risk, they do
not represent the general population; therefore, it is dif-
ficult to generalize the results from trials to the general
population. Evaluating the long-term effects of interven-
tions is also difficult owing to high costs of long-term tri-
als and lack of participant adherence to the intervention.
From a public health perspective, the findings from clin-
ical trials in high-risk populations should be considered
together with evidence from large-scale observational
studies with longer follow-up periods.
Diet. The main evidence from observational and inter-
ventional studies on the associations between the risk
of T2DM and the intake of nutrients and food groups,
as well as dietary patterns, is summarized in TABLE1. A
diet containing high-quality fats and carbohydrates (that
is, low in trans fatty acids, high in polyunsaturated fatty
acids59 and with a low glycaemic index and glycaemic
Nature Reviews | Endocrinology
Factors that affect insulin secretion and action
• Body weight
• Level of physical activity
• Smoking
• Heavy alcohol consumption
• Genetic predisposition
• Gene–environment interaction
Positive risk profile Negative risk profile
Normoglycaemia β-cell dysfunction and
insulin resistance
Adipose tissue Skeletal muscle
Liver
Insulin-mediated
glucose production ↑Insulin-mediated
glucose uptake ↓
Insulin-mediated
glucose uptake ↓
Hyperglycaemia
• Epigenetics
• Gestational diabetes
mellitus
Figure 2 | Pathophysiology of hyperglycaemia in T2DM. Insulin secretion from the
β‑cells in the pancreas normally reduces glucose output by the liver and increases
glucose uptake by skeletal muscle and adipose tissue. Once β‑cell dysfunction in the
pancreas and/or insulin resistance in the liver, skeletal muscle or adipose tissue occur,
hyperglycaemia develops, leading to an excessive amount of glucose circulating in the
blood. The various factors listed at the top affect insulin secretion and insulin action.
T2DM, type 2 diabetes mellitus.
REVIEWS
NATURE REVIEWS
|
ENDOCRINOLOGY VOLUME 14
|
FEBRUARY 2018
|
91

load60) rather than low quality fats and carbohydrates is
more important than the relative quantity of these nutri-
ents for T2DM prevention61. Dietary recommendations
for preventing T2DM typically promote diets rich in
whole grains, fruits, vegetables, nuts and legumes and
low in refined grains, red or processed meat and sugar-
sweetened beverages59. Adherence to a high-quality diet,
such as the Mediterranean diet62, was strongly associated
with a reduced risk of T2DM. In Asian countries, such
as China and South Korea, a rapid nutritional transition
in the past two to three decades that was characterized
by increased energy intake from sugars, animal prod-
ucts and refined grains and reduced consumption of
cereals is a major contributor to the T2DM epidemic59.
Furthermore, undernutrition (for example, exposure
to famine) during early life might increase the risk of
T2DM later in life63, which is discussed in detail in a
subsequent section.
Physical act ivity. Increased physical activity is an essential
component of all effective lifestyle-based trials for the pre-
vention of T2DM. Prospective evidence has shown that
both aerobic exercise and resistance training independently
have beneficial effects on preventing T2DM64. One
study has shown that spending more time on moderate-
intensity and vigorous-intensity physical activity is benefi-
cial for preventing insulin resistance, independent of time
spent sedentary65. By contrast, another study found that
time spent sedentary was associated with an increased risk
of T2DM, regardless of physical activity66.
Smoking. A meta-analysis found a dose–response
relationship between the number of cigarettes smoked
and risk of T2DM, and current smokers had a 45%
higher risk of T2DM than non-smokers67. Mor eov er,
a high level of exposure to second-hand smoke has
been associated with an increased risk of T2DM68.
Smokers are more likely to have central fat accumula-
tion than non-smokers, and smoking is known to induce
insulin resistance and compensatory insulin-secretion
responses69, which could explain the increased risk of
T2DM in people who smoke. Education campaigns to
reduce smoking should be a major public health strategy
to curb the epidemic of T2DM, especially in China and
India, which are epicentres of both T2DM and smoking70.
Alcohol intake. Moderate consumption of alcohol
has been associated with a reduced risk of T2DM71. A
meta-analysis of 20 cohort studies found a U-shaped
relationship between alcohol consumption and T2DM
risk for both sexes, and the lowest risk of diabetes mel-
litus was observed among people who consumed 1-2
drinks per day71. There might be sex differences in the
alcohol–T2DM relationship due to potential sex differ-
ences in alcohol pharmacokinetics (that is, alcohol pro-
cessing and elimination), which depend largely on body
composition72. In a randomized clinical trial involving
postmenopausal women, moderate alcohol consump-
tion (about 25 g per day) for 6weeks improved insulin
sensitivity73. However, the public health messages around
moderate drinking need to be communicated cautiously
within a culturally appropriate context, particularly con-
sidering the steady increase in alcohol consumption in
many Asian countries74 and the health burden of excess
alcohol consumption in eastern Europe75.
Genomics and gene–environment interactions
Even though many cases of T2DM could be prevented
by maintaining a healthy body weight and adhering to a
healthy lifestyle, some individuals with prediabetes mel-
litus are more susceptible to T2DM than others, which
suggests that individual differences in response to life-
style interventions exist76. Substantial evidence from
twin and family studies has suggested a genetic basis
of T2DM77. Over the past decade, successive waves of
T2DM genome-wide association studies have identi-
fied >100 robust association signals, demonstrating the
complex polygenic nature of T2DM5. Most of these loci
affect T2DM risk through primary effects on insulin
secretion, and a minority act through reducing insulin
action78. Individually, the common variants (minor allele
frequency >5%) identified in these studies have only a
modest effect on T2DM risk and collectively explain only
a small portion (~20%) of observed T2DM heritability5.
It has been hypothesized that lower-frequency vari-
ants could explain much of the remaining heritability79.
However, results of a large-scale sequencing study from
the GoT2D and T2D-GENES consortia, published in
2016, do not support such a hypothesis5.
Genetic variants might help reveal possible aetiological
mechanisms underlying T2DM development; however,
the variants identified thus far have not enabled clinical
prediction beyond that achieved with common clinical
measurements, including age, BMI, fasting levels of glu-
cose and dyslipidaemia. A study published in 2014 linked
susceptibility variants to quantitative glycaemic traits and
grouped these variants on the basis of their potential inter-
mediate mechanisms in T2DM pathophysiology: four var-
iants fitted a clear insulin resistance pattern; two reduced
insulin secretion with fasting hyperglycaemia; nine
Box 1 | Major risk factors for T2DM
• Older age
• Non-white ancestry
• Family history of type2 diabetes mellitus (T2DM)
• Low socio-economic status
• Genetic factors (for example, carrying risk alleles in the TCF7L2 gene)
• Components of the metabolic syndrome (increased waist circumference, increased
blood pressure, increased plasma levels of triglycerides, low plasma levels of HDL
cholesterol and small, dense LDL cholesterol particles)
• Overweight or obese (BMI ≥25 kg/m2)
• Abdominal or central obesity (independent of BMI)
• Unhealthy dietary factors (regular consumption of sugary beverages and red meats
and low consumption of whole grains and other fibre-rich foods)
• Cigarette smoking
• Sedentary lifestyle
• History of gestational diabetes mellitus or delivery of neonates >4 kg in weight
• Some medications, such as statins, thiazides and beta-blockers
• Psychosocial stress and depression
REVIEWS
92
|
FEBRUARY 2018
|
VOLUME 14 www.nature.com/nrendo

reduced insulin secretion with normal fasting glycaemia;
and one altered insulin processing80. Considering such evi-
dence, the genetic architecture of T2DM is highly poly-
genic, and thus, substantially larger association studies are
needed to identify most T2DM loci, which typically have
small to modest effect sizes81.
The missing heritability of T2DM could be accounted
for by the interactions between susceptibility loci and
various environmental determinants, whereby the impact
of a given genetic variant is modified by the environ-
mental milieu (and vice versa). Evidence that lifestyle
factors modify the genetic effects on T2DM risk has been
generated from both observational studies and clinical
trials82. However, genetic background might also affect
the individual’s response to lifestyle interventions83. In
addition, replication data are sparse, and comprehensive,
large-scale studies have failed to provide a compelling
basis for the significant interaction effect84,85. This failure
might have occurred because the interaction effects are
of small magnitude or might be due to the limited statis-
tical power and multiple sources of bias and confounding
factors in the current research methods86.
Biomarkers and metabolomics
Over the past two decades, biomarkers from the path-
ways of abnormal adipocyte signalling, subclinical
inflammation, endothelial dysfunction and iron overload
have improved our understanding of the complexity of
T2DM pathophysiology, beyond the classic triumvirate of
β-cell, skeletal muscle and liver87. However, the ability of
these biomarkers to predict future risk of T2DM beyond
anthropometric measures, lifestyle factors and fasting
levels of glucose and lipids is still debatable87.
Within the past 7years, a complementary, novel set of
T2DM biomarkers has largely been generated by metabo-
lomic studies, which systematically analyse metabolites
(low molecular weight biochemicals) in a biological sample.
A meta-analysis of published metabolomics studies that
was published in 2016 revealed that the high circulating
levels of hexoses, branched-chain amino acids, aromatic
amino acids, phospholipids and triglycerides, were asso-
ciated with the incidence of prediabetes mellitus and
T2DM88. As downstream end products, levels of these
metabolites could reflect upstream gene function and
environmental influences, as well as their complex inter-
plays. Of note, the metabolomics-derived indices enable
statistically significant improvement in the prediction of
T2DM risk beyond the use of traditional risk factors89.
Developmental origins of T2DM
The thrifty genotype hypothesis postulates that thrifty
genotypes favouring efficient metabolism and storage of
energy were positively selected for as a result of evolution-
ary selection by repeated feast and famine cycles; these
genotypes are maladaptive in many modern environ-
ments90. This hypothesis has been widely used to explain
the disproportionate burden of T2DM among indige-
nous populations (worldwide, >50% of indigenous adults
>35years old are estimated to have T2DM)91. This selec-
tion might have led to increased vulnerability to diabetes
mellitus among indigenous populations at a time of rapid
transition to a high-calorie diet and physical inactivity91.
In contrast to the thrifty genotype hypothesis, the
thrifty phenotype hypothesis (developmental origins) pos-
tulates a mismatch between early developmental environ-
ments (intrauterine) and adulthood environments. This
hypothesis proposes that the adaptations in response to
fetal undernutrition that lead to metabolic and structural
changes (for example, decreased β-cell mass and function
and increased insulin resistance) are beneficial for early
survival but might increase the risk of chronic diseases,
such as T2DM, in adulthood92. Low birthweight, a widely
Table 1 | Associations between nutritional factors the risk of T2DM
Nutritional factor assessed Relative risk (95%CI) Refs
Nutrients
Haeme (iron) 1.31 (1.21–1.43) extreme groups* 143
Glycaemic index 1.19 (1.14–1.24) extreme groups* 60
Glycaemic load 1.13 (1.08–1.17) extreme groups* 60
Docosahexaenoic acid (DHA) or
eicosapentaenoic acid (EPA)
1.04 (0.97–1.10) per 250 mg per day 144
Vegetable fibre 1.04 (0.94–1.15) extreme groups* 145
Fruit fibre 0.96 (0.88–1.04) extreme groups* 145
α‑Linolenic acid 0.93 (0.83–1.04) per 0.5 g per day 144
Magnesium 0.78 (0.73–0.84) extreme groups* 146
Cereal fibre 0.67 (0.62–0.72) extreme groups* 145
VitaminD 0.62 (0.54–0.70) extreme groups* 147
Food groups
Processed red meat 1.51 (1.25–1.83) per 50 g per day 148
Unprocessed red meat 1.19 (1.04–1.37) per 100 g per day 148
Fish or seafood 1.12 (0.94–1.34) per 100 g per day 144
White rice 1.11 (1.08–1.14) per 1 serving per day 149
Green leafy vegetables 0.86 (0.77–0.97) extreme groups* 150
Green leafy vegetables 0.84 (0.74–0.94) extreme groups* 151
Dairy products 0.86 (0.79–0.92) extreme groups* 152
Whole grains 0.68 (0.58–0.81) per 3 servings per day 153
Sugar-sweetened beverages 1.26 (1.12–1.41) extreme groups* 154
Sugar-sweetened beverages 1.18 (1.06–1.32) per 336 g per day 155
Decaffeinated coffee 0.80 (0.70–0.91) extreme groups* 156
Total coffee 0.70 (0.65–0.75) extreme groups* 156
Dietary patterns
Mediterranean diet 0.60 (0.43–0.85) Mediterranean diet
supplemented with extra-virgin olive oil
compared with control group (advice
on a low-fat diet), 0.82 (0.61–1.10)
Mediterranean diet supplemented with
nuts compared with control group
62
Alternate healthy eating index
(AHEI) 2010
0.77 (0.67–0.88) the highest compared with
the lowest quintiles
157
Dietary approaches to stop
hypertension (DASH)
0.75 (0.65–0.85) the highest compared with
the lowest quintiles
157
T2DM, type 2 diabetes mellitus. *For the different categories of nutrients and food groups,
such as the tertiles, quartiles or quintiles, the effect estimates and corresponding 95% CIs of
extreme groups were calculated by comparing the highest and lowest categories in a
meta‑analysis.
REVIEWS
NATURE REVIEWS
|
ENDOCRINOLOGY VOLUME 14
|
FEBRUARY 2018
|
93

used indicator of fetal undernutrition, is associated with
an increased risk of T2DM in adult life93. Epidemiological
evidence from the Dutch Hunger Winter of 1944–1945
(REF.94) and the Chinese famine of 1958–1962 (REF.63)
shows that children born during a famine who are exposed
to intrauterine undernutrition but live in an obeso genic
environment as an adult have an increased risk of chronic
diseases (including T2DM). Epigenetic processes could
be a central underlying mechanism of this thrifty phe-
notype hypothesis, leading to altered feeding behaviour,
insulin secretion and action and even transgenerational
risk transmission95.
Gestational diabetes mellitus, which is a common preg-
nancy complication defined as glucose intolerance with
onset or first recognition during pregnancy96, is another
risk factor that influences T2DM risk in exposed women
and their offspring. The prevalence of gestational diabetes
mellitus varies depending on the diagnostic criteria used
and the study population; for instance, the prevalence is
1.2–3.1% of pregnancies in European countries (except for
Italy) and 1.9–13.7% of pregnancies in the southeast Asia
region97. Women with gestational diabetes mellitus had a
sevenfold increased risk of developing T2DM compared
with those who had a normoglycaemic pregnancy98. In
the offspring of women with gestational diabetes mellitus,
exposure to intrauterine hyperglycaemia is a strong risk
factor for T2DM96. The increasing frequency of expo-
sure to gestational diabetes mellitus inutero, together
with increasing body weight, accounted for most of the
increase in the prevalence of T2DM in Pima Indian chil-
dren99. Given its global influence, it is essential to break
the vicious cycle of diabetes mellitus begetting diabetes
mellitus over generations by implementing effective
strategies to prevent gestational diabetes mellitus.
Other factors
Interest in the role of the gut microbiome in the devel-
opment of T2DM has exploded in the past few years,
and variation in the diversity and composition of the gut
microbiota has been tied to T2DM100. For example, levels
of butyrate-producing bacteria are decreased in the gut
microbiota of patients with T2DM compared with that
of healthy individuals101. In addition, evidence suggests
that ambient air pollution is an emerging risk factor for
T2DM102, especially in developing countries where the
rapid increase in urbanization has introduced high levels
of outdoor and indoor pollution103,104. Furthermore, the
use of some medications, such as statins, thiazides and
beta-blockers, has been associated with an increased risk
of T2DM105.
Epidemiology of complications in T2DM
The complications of diabetes mellitus have traditionally
been divided into macrovascular complications (for exam-
ple, cardiovascular disease (CVD)) and microvascular
complications (for example, complications affecting the
kidney, the retina and the nervous system). Complications
of T2DM are very common, with half of patients with
T2DM presenting with microvascular complications and
27% with macrovascular complications in an observa-
tional study of 28 countries in Asia, Africa, South America
and Europe106. On the basis of cohort studies from
developed countries, the relative risk of microvascular
disorders and macrovascular disorders among patients
with diabetes mellitus was estimated to be at least 10–20
times higher and 2–4 times higher, respectively, than in
people without diabetes mellitus107. In most developing
countries, patients with diabetes mellitus are at a particu-
larly increased risk of developing kidney complications
and stroke (but have a reduced risk of coronary heart dis-
ease) compared with patients in developed countries108.
The large increase in the number of prevalent cases
of diabetes mellitus and undiagnosed diabetes mellitus,
together with advances in the treatment of T2DM mean-
ing that people are living longer with the condition than
they used to, has resulted in a costly increase in the inci-
dence of diabetic complications; for instance, 53% of the
lifetime medical costs of T2DM have been attributed to
treating the major complications of T2DM (nephropathy,
neuropathy, retinopathy, stroke and coronary heart dis-
eases) in the USA109. However, the absence of internation-
ally recognized and standardized classification, definition
or diagnostic criteria for the complications of T2DM
makes it hard to precisely estimate their contributions to
morbidity and mortality26. Patient-centred management
of T2DM involves lifestyle modification and combination
therapy of medication110. In some developed countries,
the management of T2DM, mainly through glycaemic
control and cardiovascular risk management, has resulted
in improved care; however, for the rest of the world, such
data are scarce107.
Cardiovascular disease
CVD, including coronary heart disease, peripheral vas-
cular disease and cerebrovascular disease, is the primary
cause of morbidity and mortality in the USA21. In patients
with T2DM, CVD typically develops 14.6years earlier111,
and with greater severity, than in individuals without dia-
betes mellitus112. Furthermore, individuals with T2DM are
twice as likely to develop CVD as those without T2DM,
independent of age, smoking status, BMI and systolic
blood pressure113, and diabetes mellitus has been associ-
ated with a more than doubled risk of death from vas-
cular causes114. This excess risk disproportionately affects
women115 such that diabetes mellitus eliminates or atten-
uates the reduced risk of CVD that is generally seen in
premenopausal women. Post hoc analysis of data from the
large-scale, randomized clinical trial Action in Diabetes
and Vascular Disease (ADVANCE) has suggested that
patients from Asian countries who have T2DM have a
lower risk of major coronary events than patients from
eastern Europe or Established Market Economies116.
Within Asia, susceptibility to vascular complications
varies across ethnicities and areas. For example, patients
in China with diabetes mellitus had lower rates of cor-
onary artery disease than patients in other countries117,
whereas patients in India who had T2DM had a doubled
risk of coronary artery disease-related deaths compared
with white Europeans who had T2DM, independent of
traditional risk factors118. Coronary artery disease detec-
tion and diabetes mellitus duration in these studies might
partially account for the ethnic differences.
REVIEWS
94
|
FEBRUARY 2018
|
VOLUME 14 www.nature.com/nrendo

Renal disease
Approximately 10% of deaths in people with T2DM are
attributable to renal failure119. Diabetes mellitus causes
44% of the incident cases of end-stage renal disease
(ESRD) in the USA21. Furthermore, in the USA, about
25% of patients with T2DM have diabetic kidney disease,
which is defined as persistent albuminuria, persistent
reduced estimated glomerular filtration rate or both120.
In China, glomerulonephritis was historically the lead-
ing cause of ESRD; however, diabetes-related chronic
kidney disease was the leading cause of ESRD in the gen-
eral population in 2010 and has been the leading cause
in hospitalized patients since 2011 (REF.1 21). In North
America, diabetes-related ESRD is 80% more prevalent
in patients with T2DM of Asian descent than in patients
who are white122. Furthermore, the risk of diabetic kid-
ney disease is much higher in Asian countries than in
Western countries17. The ADVANCE trial has confirmed
the increased frequency of renal disease in patients with
diabetes mellitus in Asia compared with white patients
in eastern Europe and Established Market Economies116.
Genetic background, lifestyle and patient awareness of
complications might account for these ethnic differences
in renal disease among patients with diabetes mellitus123.
Other complications
The prevalence of diabetic retinopathy is approximately
28.5% in the USA124 and ranges from 16% to 35% in Asian
countries125,126. T2DM is the leading cause of non-traumatic
lower-limb amputations in the USA21. In the UK, about
one in three amputees has diabetes mellitus127, and in
Australia, about half of amputees have diabetes mellitus128.
Directly or indirectly, T2DM might also increase the risks
of disorders in the musculoskeletal, hepatic and digestive
systems, as well as cognitive function and mental health
disorders, and could increase the incidence of some can-
cers, for instance, those of the liver, pancreas and endome-
trium129. Several key comorbidities, such as non-alcoholic
fatty liver disease130, obstructive sleep apnoea131 and
depression132, are associated with T2DM bi-directionally,
and such interrelationships are at least partially caused by
obesity. However, the paucity of population-level data on
the associations of T2DM with these diseases and other
complications, such as infections and neuropathy, is a
major gap in population-level monitoring.
T2DM management
Modification of lifestyle, including weight loss, increasing
physical activity and adopting a healthy diet, remains one
of the first-line strategies for the management of T2DM.
In the Look AHEAD (Action for Health in Diabetes) trial
in the USA133, a 4-year intensive lifestyle intervention
through caloric restriction and increased physical activity
achieved increased weight loss, improved cardiometabolic
risk profiles and a reduced requirement for medication
to control CVD risk factors compared with the control
group (who had diabetes mellitus and received support
and education about lifestyle modifications)134. However,
after a median follow-up period of 9.6years, the trial was
terminated because the intervention did not reduce the
rate of CVD events134.
Another randomized, controlled trial in the USA, the
Health Benefits of Aerobic and Resistance Training in
Individuals with Diabetes (HART-D) trial, found that
HbA1c levels were reduced in the group that under-
took combined resistance and aerobic training after the
9-month exercise program, but not in the group that
undertook either resistance training or aerobic training
alone135. These findings suggest that it is more beneficial
to combine both aerobic and resistance exercises than to
perform only one type of exercise when time available to
exercise is limited136. The post hoc subgroup analysis of
data from the PREDIMED trial, which was conducted
in Spain, revealed that a Mediterranean diet signifi-
cantly reduced CVD risk (by ~30%) in participants
with diabetes mellitus137. Furthermore, a Mediterranean
diet enriched with extra-virgin olive oil might protect
against diabetic retinopathy, but not against diabetic
nephropathy 138.
In addition to lifestyle modification, social support
has an important role in T2DM management as it directly
affects the performance of diabetes mellitus self-care
behaviours and indirectly affects glycaemic control139.
For example, patients whose family members exhibit
non-supportive behaviours have reduced adherence to
diabetes mellitus medication regimens140. Public health
and social interventions through a multifaceted systems
approach, involving structural changes in schools,
workplaces, communities, media and food and beverage
systems, have been proposed to address the pandemic
of obesity 141, and these are also applicable to T2DM
prevention and management.
Conclusions
In the past three decades, T2DM and its complications
have reached epidemic levels, particularly in develop-
ing countries. T2DM is a global crisis that threatens the
health and economy of the world. Approximately 1 in
every 11 adults has T2DM globally, and about 75% of
patients with diabetes mellitus live in developing coun-
tries1. T2DM is associated with an 8-year reduction in
lifespan in the USA142, and also has a negative effect on
quality of life as most patients also have complications.
About 12% of the global health expenditure was spent
on treatment of T2DM and its related complications in
2015 (REF.1).
An accumulating body of evidence from large pro-
spective observational studies and randomized clini-
cal trials indicates that many cases of T2DM could be
prevented by maintaining a healthy body weight with
a focus on maintaining energy balance by engaging
in regular physical activity and consuming a healthy
diet7,8. Preventing and managing gestational diabetes
mellitus to stop the vicious cycle in which diabetes
mellitus begets diabetes mellitus is also key. The mis-
match between early developmental environment
(for example, fetal undernutrition) and obesogenic
adulthood environment is an important risk factor for
T2DM. For T2DM management, lifestyle modification,
social support and medication adherence are impor-
tant for reducing the risk of cardiovascular and other
complications.
REVIEWS
NATURE REVIEWS
|
ENDOCRINOLOGY VOLUME 14
|
FEBRUARY 2018
|
95

1. International Diabetes Federation. IDF Diabetes Atlas
— 7th Edition. DiabetesAtlas http://www.
diabetesatlas.org/ (2015).
2. Zimmet,P.Z. Diabetes and its drivers: the largest
epidemic in human history? Clin. Diabetes Endocrinol.
3, 1 (2017).
3. Holman,N., Young,B. & Gadsby,R. Current
prevalence of type1 and type2 diabetes in adults and
children in the UK. Diabet Med. 32, 1119–1120
(2015).
4. Bruno,G. etal. Incidence of type1 and type2
diabetes in adults aged 30–49years: the population-
based registry in the province of Turin, Italy. Diabetes
Care 28, 2613–2619 (2005).
5. Fuchsberger,C. etal. The genetic architecture of
type2 diabetes. Nature 536, 41–47 (2016).
6. Chatterjee,S., Khunti,K. & Davies,M.J. Type2
diabetes. Lancet 389, 2239–2251 (2017).
7. Schellenberg,E.S., Dryden,D.M., Vandermeer,B.,
Ha,C. & Korownyk,C. Lifestyle interventions for
patients with and at risk for type2 diabetes: a
systematic review and meta-analysis. Ann. Intern.
Med. 159, 543–551 (2013).
8. Hu,F.B. etal. Diet, lifestyle, and the risk of type2
diabetes mellitus in women. N.Engl. J.Med. 345,
790–797 (2001).
9. GBD 2013 Mortality and Causes of Death
Collaborators. Global, regional, and national age-sex
specific all-cause and cause-specific mortality for 240
causes of death, 1990–2013: a systematic analysis
for the Global Burden of Disease Study 2013. Lancet
385, 117–171 (2015).
10. Roglic,G. & Unwin,N. Mortality attributable to
diabetes: estimates for the year 2010. Diabetes Res.
Clin. Pract. 87, 15–19 (2010).
11. Global Burden of Disease Study 2013 Collaborators.
Global, regional, and national incidence, prevalence,
and years lived with disability for 301 acute and
chronic diseases and injuries in 188 countries,
1990–2013: a systematic analysis for the Global
Burden of Disease Study 2013. Lancet 386,
743–800 (2015).
12. GBD 2015 Risk Factors Collaborators. Global,
regional, and national comparative risk assessment of
79 behavioural, environmental and occupational, and
metabolic risks or clusters of risks, 1990–2015: a
systematic analysis for the Global Burden of Disease
Study 2015. Lancet 388, 1659–1724 (2016).
13. Beagley,J., Guariguata,L., Weil,C. & Motala,A.A.
Global estimates of undiagnosed diabetes in adults.
Diabetes Res. Clin. Pract. 103, 150–160 (2014).
14. Rubin,R.J., Altman,W.M. & Mendelson,D.N.
Health care expenditures for people with diabetes
mellitus, 1992. J.Clin. Endocrinol. Metab. 78
,
809A
–
809F (1994).
15. NCD Risk Factor Collaboration. Worldwide trends in
diabetes since 1980: a pooled analysis of 751
population-based studies with 4.4 million participants.
Lancet 387, 1513–1530 (2016).
16. Shaw,J.E., Sicree,R.A. & Zimmet,P.Z. Global
estimates of the prevalence of diabetes for 2010 and
2030. Diabetes Res. Clin. Pract. 87, 4–14 (2010).
17. Kong,A.P. etal. Diabetes and its comorbidities —
where east meets west. Nat. Rev. Endocrinol. 9,
537–547 (2013).
18. Xu,Y. etal. Prevalence and control of diabetes in
Chinese adults. JAMA 310, 948–959 (2013).
19. Anjana,R.M. etal. Prevalence of diabetes and
prediabetes (impaired fasting glucose and/or impaired
glucose tolerance) in urban and rural India: phaseI
results of the Indian Council of Medical Research-INdia
DIABetes (ICMR-INDIAB) study. Diabetologia 54,
3022–3027 (2011).
20. Goldhaber-Fiebert,J.D. etal. Inpatient treatment of
diabetic patients in Asia: evidence from India, China,
Thailand and Malaysia. Diabet Med. 27, 101–108
(2010).
21. National Center for Chronic Disease Prevention and
Health Promotion, Division of Diabetes Translation.
National diabetes fact sheet: national estimates and
general information on diabetes and prediabetes in
the United States, 2011. Centers for Disease Control
and Prevention https://www.cdc.gov/diabetes/pubs/
pdf/methods11.pdf (2011).
22. Al-Siyabi,H., Al-Anquodi,Z., Al-Hinai,H. & Al-Hinai,S.
Nizwa Healthy Lifestyle Project Evaluation Report
2010 (Ad Dakhiliyah, Oman: Ministry of Health,
2010).
23. Al-Rubeaan,K. etal. Epidemiology of abnormal
glucose metabolism in a country facing its epidemic:
SAUDI-DM study. J.Diabetes 7, 622–632 (2015).
24. Lozano,R. etal. Global and regional mortality from
235 causes of death for 20 age groups in 1990 and
2010: a systematic analysis for the Global Burden of
Disease Study 2010. Lancet 380, 2095–2128 (2012).
25. Arredondo,A. Type2 diabetes and health care costs
in Latin America: exploring the need for greater
preventive medicine. BMC Med. 12, 136 (2014).
26. Zimmet,P., Alberti,K.G., Magliano,D.J. &
Bennett,P.H. Diabetes mellitus statistics on
prevalence and mortality: facts and fallacies. Nat. Rev.
Endocrinol. 12, 616–622 (2016).
27. Chen,L., Magliano,D.J. & Zimmet,P.Z. The
worldwide epidemiology of type2 diabetes mellitus —
present and future perspectives. Nat. Rev. Endocrinol.
8, 228–236 (2011).
28. Nadeau,K.J. etal. Youth-onset type2 diabetes
consensus report: current status, challenges, and
priorities. Diabetes Care 39, 1635–1642 (2016).
29. Dabelea,D. etal. Prevalence of type1 and type2
diabetes among children and adolescents from 2001
to 2009. JAMA 311, 1778–1786 (2014).
30. Mayer-Davis,E.J. etal. Incidence trends of type1 and
type2 diabetes among youths, 2002–2012. N.Engl.
J.Med. 376, 1419–1429 (2017).
31. Fazeli Farsani,S., van der Aa,M.P., van der
Vorst,M.M., Knibbe,C.A. & de Boer,A. Global
trends in the incidence and prevalence of type2
diabetes in children and adolescents: a systematic
review and evaluation of methodological approaches.
Diabetologia 56, 1471–1488 (2013).
32. Fu,J. & Prasad,H.C. Changing epidemiology of
metabolic syndrome and type2 diabetes in Chinese
youth. Curr. Diab Rep. 14, 447 (2014).
33. Praveen,P.A. etal. Registry of Youth Onset Diabetes in
India (YDR): rationale, recruitment, and current status.
J.Diabetes Sci. Technol. 10 , 1034–1041 (2016).
34. Ley,S.H. etal. Metabolic syndrome and its
components as predictors of incident type2 diabetes
mellitus in an Aboriginal community. CMAJ 180,
617–624 (2009).
35. Reinehr,T. Type2 diabetes mellitus in children and
adolescents. World J.Diabetes 4, 270–281 (2013).
36. Australian Institute of Health and Welfare. Type2
diabetes in Australia’s children and young people: a
working paper (Australian Institute of Health and
Welfare, 2014).
37. Stumvoll,M., Goldstein,B.J. & van Haeften,T.W.
Type2 diabetes: principles of pathogenesis and
therapy. Lancet 365, 1333–1346 (2005).
38. Reaven,G.M. Banting lecture 1988. Role of insulin
resistance in human disease. Diabetes 37, 1595–1607
(1988).
39. Weyer,C., Bogardus,C., Mott,D.M. & Pratley,R.E.
The natural history of insulin secretory dysfunction
and insulin resistance in the pathogenesis of type2
diabetes mellitus. J.Clin. Invest. 104, 787–794
(1999).
40. Centers for Disease Control and Prevention, Division
of Diabetes Translation. Maps of diabetes and obesity
in 1994, 2000, and 2014 (Centers for Disease
Control and Prevention, 2016).
41. Tian,Y. etal. BMI, leisure-time physical activity, and
physical fitness in adults in China: results from a series
of national surveys, 2000–2014. Lancet Diabetes
Endocrinol. 4, 487–497 (2016).
42. Yang,W. etal. Prevalence of diabetes among men and
women in China. N.Engl. J.Med. 362, 1090–1101
(2010).
43. Franco,M. etal. Population-wide weight loss and
regain in relation to diabetes burden and
cardiovascular mortality in Cuba 1980-2010:
repeated cross sectional surveys and ecological
comparison of secular trends. BMJ 346, f1515
(2013).
44. NCD Risk Factor Collaboration. Trends in adult body-
mass index in 200 countries from 1975 to 2014: a
pooled analysis of 1698 population-based
measurement studies with 19.2 million participants.
Lancet 387, 1377–1396 (2016).
45. Hu,F.B. in Obesity Epidemiology (ed. Hu,F.B.)
149–173 (Oxford Univ. Press, 2008).
46. Sinha,R. etal. Assessment of skeletal muscle
triglyceride content by 1H nuclear magnetic resonance
spectroscopy in lean and obese adolescents:
relationships to insulin sensitivity, total body fat,
and central adiposity. Diabetes 51, 1022–1027
(2002).
47. Carey,V.J. etal. Body fat distribution and risk of
non-insulin-dependent diabetes mellitus in women.
The Nurses’ Health Study. Am. J.Epidemiol. 145,
614–619 (1997).
48. Zheng,Y. etal. Associations of weight gain from early
to middle adulthood with major health outcomes later
in life. JAMA 318, 255–269 (2017).
49. Lebovitz,H.E. & Banerji,M.A. Point: visceral
adiposity is causally related to insulin resistance.
Diabetes Care 28, 2322–2325 (2005).
50. Lee,J.W., Brancati,F.L. & Yeh,H.C. Trends in the
prevalence of type2 diabetes in Asians versus whites:
results from the United States National Health
Interview Survey, 1997–2008. Diabetes Care 34,
353–357 (2011).
51. Deurenberg,P., Deurenberg-Yap,M. & Guricci,S.
Asians are different from Caucasians and from each
other in their body mass index/body fat percent
relationship. Obes. Rev. 3, 141–146 (2002).
52. Chan,J.C. etal. Diabetes in Asia: epidemiology, risk
factors, and pathophysiology. JAMA 301, 2129–2140
(2009).
53. Narayan,K.M. Type2 diabetes: why we are winning
the battle but losing the war? 2015 Kelly West award
lecture. Diabetes Care 39, 653–663 (2016).
54. Carlsson,L.M. etal. Bariatric surgery and prevention
of type2 diabetes in Swedish obese subjects. N.Engl.
J.Med. 367, 695–704 (2012).
55. Knowler,W.C. etal. Reduction in the incidence of
type2 diabetes with lifestyle intervention or
metformin. N.Engl. J.Med. 346, 393–403 (2002).
56. Tuomilehto,J. etal. Prevention of type2 diabetes
mellitus by changes in lifestyle among subjects with
impaired glucose tolerance. N.Engl. J.Med. 344,
1343–1350 (2001).
57. Pan,X.R. etal. Effects of diet and exercise in
preventing NIDDM in people with impaired glucose
tolerance. The Da Qing IGT and Diabetes Study.
Diabetes Care 20, 537–544 (1997).
58. Tong,Y.Z. etal. Consensus on the prevention of
type2 diabetes in Chinese adults. Chin. Med. J. 130,
600–606 (2017).
59. Ley,S.H., Hamdy,O., Mohan,V. & Hu,F.B.
Prevention and management of type2 diabetes:
dietary components and nutritional strategies. Lancet
383, 1999–2007 (2014).
60. Bhupathiraju,S.N. etal. Glycemic index, glycemic
load, and risk of type2 diabetes: results from 3 large
US cohorts and an updated meta-analysis. Am. J.Clin.
Nutr. 100, 218–232 (2014).
61. Schulze,M.B. & Hu,F.B. Primary prevention of
diabetes: what can be done and how much can be
prevented? Annu. Rev. Public Health 26, 445–467
(2005).
62. Salas-Salvado,J. etal. Prevention of diabetes with
Mediterranean diets: a subgroup analysis of a
randomized trial. Ann. Intern. Med. 160, 1–10 (2014).
63. Li,Y. etal. Exposure to the Chinese famine in early life
and the risk of hyperglycemia and type2 diabetes in
adulthood. Diabetes 59, 2400–2406 (2010).
64. Grontved,A., Rimm,E.B., Willett,W.C.,
Andersen,L.B. & Hu,F.B. A prospective study of
weight training and risk of type2 diabetes mellitus in
men. Arch. Intern. Med. 172, 1306–1312 (2012).
65. Ekelund,U., Brage,S., Griffin,S.J. & Wareham,N.J.
Objectively measured moderate- and vigorous-
intensity physical activity but not sedentary time
predicts insulin resistance in high-risk individuals.
Diabetes Care 32, 1081–1086 (2009).
66. Rockette-Wagner,B. etal. The impact of lifestyle
intervention on sedentary time in individuals at high
risk of diabetes. Diabetologia 58, 1198–1202
(2015).
67. Willi,C., Bodenmann,P., Ghali,W.A., Faris,P.D. &
Cornuz,J. Active smoking and the risk of type2
diabetes: a systematic review and meta-analysis.
JAMA 298, 2654–2664 (2007).
68. Hayashino,Y. etal. A prospective study of passive
smoking and risk of diabetes in a cohort of workers:
the High-Risk and Population Strategy for
Occupational Health Promotion (HIPOP-OHP) study.
Diabetes Care 31, 732–734 (2008).
69. Reaven,G. & Tsao,P.S. Insulin resistance and
compensatory hyperinsulinemia: the key player
between cigarette smoking and cardiovascular
disease? J.Am. Coll. Cardiol. 41, 1044–1047
(2003).
70. GBD 2015 Tobacco Collaborators. Smoking
prevalence and attributable disease burden in 195
countries and territories, 1990–2015: a systematic
analysis from the Global Burden of Disease Study
2015. Lancet 389, 1885–1906 (2017).
71. Baliunas,D.O. etal. Alcohol as a risk factor for type2
diabetes: a systematic review and meta-analysis.
Diabetes Care 32, 2123–2132 (2009).
REVIEWS
96
|
FEBRUARY 2018
|
VOLUME 14 www.nature.com/nrendo

72. Mumenthaler,M.S., Taylor,J.L., O’Hara,R. &
Yesavage,J.A. Gender differences in moderate drinking
effects. Alcohol Res. Health 23, 55–64 (1999).
73. Joosten,M.M., Beulens,J.W., Kersten,S. &
Hendriks,H.F. Moderate alcohol consumption
increases insulin sensitivity and ADIPOQ
expression in postmenopausal women: a
randomised, crossover trial. Diabetologia 51,
1375–1381 (2008).
74. Ezzati,M. & Riboli,E. Behavioral and dietary risk
factors for noncommunicable diseases. N.Engl.
J.Med. 369, 954–964 (2013).
75. Powles,J.W., Zatonski,W., Vander Hoorn,S. &
Ezzati,M. The contribution of leading diseases and
risk factors to excess losses of healthy life in Eastern
Europe: burden of disease study. BMC Public Health
5, 116 (2005).
76. Böhm,A., Weigert,C., Staiger,H. & Haring,H.U.
Exercise and diabetes: relevance and causes for
response variability. Endocrine 51, 390–401 (2016).
77. Almgren,P. etal. Heritability and familiality of type2
diabetes and related quantitative traits in the Botnia
Study. Diabetologia 54, 2811–2819 (2011).
78. McCarthy,M.I. Genomics, type2 diabetes, and
obesity. N.Engl. J.Med. 363, 2339–2350 (2010).
79. Manolio,T.A. etal. Finding the missing heritability of
complex diseases. Nature 461, 747–753 (2009).
80. Dimas,A.S. etal. Impact of type2 diabetes
susceptibility variants on quantitative glycemic traits
reveals mechanistic heterogeneity. Diabetes 63,
2158–2171 (2014).
81. Flannick,J. & Florez,J.C. Type2 diabetes: genetic
data sharing to advance complex disease research.
Nat. Rev. Genet. 17, 535–549 (2016).
82. Franks,P.W., Pearson,E. & Florez,J.C. Gene-
environment and gene-treatment interactions in
type2 diabetes: progress, pitfalls, and prospects.
Diabetes Care 36, 1413–1421 (2013).
83. Hagberg,J.M., Jenkins,N.T. & Spangenburg,E.
Exercise training, genetics and type2 diabetes-
related phenotypes. Acta Physiol. 205, 456–471
(2012).
84. Langenberg,C. etal. Gene-lifestyle interaction and
type2 diabetes: the EPIC interact case-cohort study.
PLoS Med. 11, e1001647 (2014).
85. Hivert,M.F. etal. Updated genetic score based on 34
confirmed type2 diabetes loci is associated with
diabetes incidence and regression to normoglycemia
in the diabetes prevention program. Diabetes 60,
1340–1348 (2011).
86. Franks,P.W. & McCarthy,M.I. Exposing the
exposures responsible for type2 diabetes and obesity.
Science 354, 69–73 (2016).
87. Meigs,J.B. Multiple biomarker prediction of type2
diabetes. Diabetes Care 32, 1346–1348 (2009).
88. Guasch-Ferre,M. etal. Metabolomics in prediabetes
and diabetes: a systematic review and meta-analysis.
Diabetes Care 39, 833–846 (2016).
89. Tripathy,D. etal. A novel insulin resistance index to
monitor changes in insulin sensitivity and glucose
tolerance: the ACT NOW study. J.Clin. Endocrinol.
Metab. 100, 1855–1862 (2015).
90. Neel,J.V. Diabetes mellitus: a “thrifty” genotype
rendered detrimental by “progress”? Am. J.Hum.
Genet. 14, 353–362 (1962).
91. Cunningham,M. The state of the World’s indigenous
peoples. Chapter V, health. United Nations http://www.
un.org/esa/socdev/unpfii/documents/SOWIP/en/
SOWIP_chapter5.pdf (2009).
92. Hales,C.N. & Barker,D.J. The thrifty phenotype
hypothesis. Br. Med. Bull. 60, 5–20 (2001).
93. Whincup,P.H. etal. Birth weight and risk of type2
diabetes: a systematic review. JAMA 300,
2886–2897 (2008).
94. Ravelli,A.C. etal. Glucose tolerance in adults after
prenatal exposure to famine. Lancet 351, 173–177
(1998).
95. Ong,T.P. & Ozanne,S.E. Developmental
programming of type2 diabetes: early nutrition and
epigenetic mechanisms. Curr. Opin. Clin. Nutr. Metab.
Care 18, 354–360 (2015).
96. Coustan,D.R. Gestational diabetes mellitus. Clin.
Chem. 59, 1310–1321 (2013).
97. Chen,L., Mayo,R., Chatry,A. & Hu,G. Gestational
diabetes mellitus: its epidemiology and implication
beyond pregnancy. Curr. Epidemiol. Rep. 3, 1–11
(2016).
98. Bellamy,L., Casas,J.P., Hingorani,A.D. &
Williams,D. Type2 diabetes mellitus after gestational
diabetes: a systematic review and meta-analysis.
Lancet 373, 1773–1779 (2009).
99. Dabelea,D. etal. Increasing prevalence of type II
diabetes in American Indian children. Diabetologia
41, 904–910 (1998).
100. Ussar,S. etal. Interactions between gut microbiota,
host genetics and diet modulate the predisposition to
obesity and metabolic syndrome. Cell Metab. 22,
516–530 (2015).
101. Qin,J. etal. A metagenome-wide association study of
gut microbiota in type2 diabetes. Nature 490,
55–60 (2012).
102. Rao,X., Montresor-Lopez,J., Puett,R.,
Rajagopalan,S. & Brook,R.D. Ambient air pollution:
an emerging risk factor for diabetes mellitus. Curr.
Diab Rep. 15, 603 (2015).
103. Gall,E.T., Carter,E.M., Earnest,C.M. & Stephens,B.
Indoor air pollution in developing countries: research
and implementation needs for improvements in global
public health. Am. J.Public Health 103, e67–72
(2013).
104. World Health Organization. Ambient air pollution: a
global assessment of exposure and burden of disease.
WHO http://apps.who.int/iris/bitstream/10665/
250141/1/9789241511353-eng.pdf (2016).
105. Anyanwagu,U., Idris,I. & Donnelly,R. Drug-induced
diabetes mellitus: evidence for statins and other drugs
affecting glucose metabolism. Clin. Pharmacol. Ther.
99, 390–400 (2016).
106. Litwak,L. etal. Prevalence of diabetes complications
in people with type2 diabetes mellitus and its
association with baseline characteristics in the
multinational A1chieve study. Diabetol Metab. Syndr.
5, 57 (2013).
107. Gregg,E.W., Sattar,N. & Ali,M.K. The changing face
of diabetes complications. Lancet Diabetes Endocrinol.
4, 537–547 (2016).
108. Zimmet,P.Z., Magliano,D.J., Herman,W.H. &
Shaw,J.E. Diabetes: a 21st century challenge. Lancet
Diabetes Endocrinol. 2, 56–64 (2014).
109. Zhuo,X., Zhang,P. & Hoerger,T.J. Lifetime direct
medical costs of treating type2 diabetes and diabetic
complications. Am. J.Prev. Med. 45, 253–261
(2013).
110 . Reusch,J.E. & Manson,J.E. Management of type2
diabetes in 2017: getting to goal. JAMA 317,
1015–1016 (2017).
111. Booth,G.L., Kapral,M.K., Fung,K. & Tu,J.V.
Relation between age and cardiovascular disease in
men and women with diabetes compared with non-
diabetic people: a population-based retrospective
cohort study. Lancet 368, 29–36 (2006).
112 . Beckman,J.A., Paneni,F., Cosentino,F. &
Creager,M.A. Diabetes and vascular disease:
pathophysiology, clinical consequences, and medical
therapy: part II. Eur. Heart J. 34, 2444–2452
(2013).
113 . Sarwar,N. etal. Diabetes mellitus, fasting blood
glucose concentration, and risk of vascular disease: a
collaborative meta-analysis of 102 prospective
studies. Lancet 375, 2215–2222 (2010).
114 . Seshasai,S.R. etal. Diabetes mellitus, fasting
glucose, and risk of cause-specific death. N.Engl.
J.Med. 364, 829–841 (2011).
115 . Regensteiner,J.G. etal. Sex differences in the
cardiovascular consequences of diabetes mellitus: a
scientific statement from the american heart
association. Circulation 132, 2424–2447 (2015).
116 . Clarke,P.M. etal. Event rates, hospital utilization, and
costs associated with major complications of diabetes:
a multicountry comparative analysis. PLoS Med. 7,
e1000236 (2010).
117 . Chi,Z.S., Lee,E.T., Lu,M., Keen,H. & Bennett,P.H.
Vascular disease prevalence in diabetic patients in
China: standardised comparison with the 14 centres
in the WHO Multinational Study of Vascular Disease
in Diabetes. Diabetologia 44, S82–S86 (2001).
118 . Forouhi,N.G., Sattar,N., Tillin,T., McKeigue,P.M. &
Chaturvedi,N. Do known risk factors explain the
higher coronary heart disease mortality in South Asian
compared with European men? Prospective follow-up
of the Southall and Brent studies, UK. Diabetologia
49, 2580–2588 (2006).
119 . van Dieren,S., Beulens,J.W., van der Schouw,Y.T.,
Grobbee,D.E. & Neal,B. The global burden of
diabetes and its complications: an emerging
pandemic. Eur. J.Cardiovasc. Prev. Rehabil. 17,
S3–S8 (2010).
120. Afkarian,M. etal. Clinical manifestations of kidney
disease among US adults with diabetes, 1988–2014.
JAMA 316, 602–610 (2016).
121. Zhang,L. etal. Trends in chronic kidney disease in
China. N.Engl. J.Med. 375, 905–906 (2016).
122. McNeely,M.J. & Fujimoto,W.Y. in The Epidemiology
of Diabetes Mellitus (eds Ekoé,J.-M., Rewers,M.,
Williams,R. & Zimmet,P.) 323–337 (John Wiley &
Sons, 2008).
123. Wilkinson,E. etal. Lack of awareness of kidney
complications despite familiarity with diabetes: a
multi-ethnic qualitative study. J.Ren Care 37, 2–11
(2011).
124. Zhang,X. etal. Prevalence of diabetic retinopathy in
the United States, 2005–2008. JAMA 304,
649–656 (2010).
125. Wong,T.Y. etal. Prevalence and risk factors for
diabetic retinopathy: the Singapore Malay Eye Study.
Ophthalmology 115 , 1869–1875 (2008).
126. Jee,D., Lee,W.K. & Kang,S. Prevalence and risk
factors for diabetic retinopathy: the Korea National
Health and Nutrition Examination Survey
2008–2011. Invest. Ophthalmol. Vis. Sci. 54,
6827–6833 (2013).
127. Schofield,C.J. etal. Mortality and hospitalization in
patients after amputation: a comparison between
patients with and without diabetes. Diabetes Care 29,
2252–2256 (2006).
128. Lim,T.S. etal. Outcomes of a contemporary
amputation series. ANZ J.Surg. 76, 300–305 (2006).
129. Giovannucci,E. etal. Diabetes and cancer: a
consensus report. Diabetes Care 33, 1674–1685
(2010).
130. Cali,A.M. etal. Glucose dysregulation and hepatic
steatosis in obese adolescents: is there a link?
Hepatology 49, 1896–1903 (2009).
131. Jordan,A.S., McSharry,D.G. & Malhotra,A. Adult
obstructive sleep apnoea. Lancet 383, 736–747
(2014).
132. Tabak,A.G., Akbaraly,T.N., Batty,G.D. &
Kivimaki,M. Depression and type2 diabetes: a causal
association? Lancet Diabetes Endocrinol. 2, 236–245
(2014).
133. Wadden,T.A. etal. The Look AHEAD study: a
description of the lifestyle intervention and the
evidence supporting it. Obesity 14, 737–752 (2006).
134. Wing,R.R. etal. Cardiovascular effects of intensive
lifestyle intervention in type2 diabetes. N.Engl.
J.Med. 369, 145–154 (2013).
135. Church,T.S. etal. Effects of aerobic and resistance
training on hemoglobin A1c levels in patients with
type2 diabetes: a randomized controlled trial. JAMA
304, 2253–2262 (2010).
136. Sigal,R.J. & Kenny,G.P. Combined aerobic and
resistance exercise for patients with type2 diabetes.
JAMA 304, 2298–2299 (2010).
137. Estruch,R. etal. Primary prevention of cardiovascular
disease with a Mediterranean diet. N.Engl. J.Med.
368, 1279–1290 (2013).
138. Diaz-Lopez,A. etal. Mediterranean diet, retinopathy,
nephropathy, and microvascular diabetes
complications: a post hoc analysis of a randomized
trial. Diabetes Care 38, 2134–2141 (2015).
139. Osborn,C.Y., Rivet Amico,K., Fisher,W.A.,
Egede,L.E. & Fisher,J.D. An information-motivation-
behavioral skills analysis of diet and exercise behavior
in Puerto Ricans with diabetes. J.Health Psychol. 15,
1201–1213 (2010).
140. Mayberry,L.S. & Osborn,C.Y. Family support,
medication adherence, and glycemic control among
adults with type2 diabetes. Diabetes Care 35,
1239–1245 (2012).
141. Barry,C.L., Gollust,S.E. & Niederdeppe,J. Are
Americans ready to solve the weight of the nation?
N.Engl. J.Med. 367, 389–391 (2012).
142. Mozaffarian,D. etal. Executive summary: heart
disease and stroke statistics — 2016 update: a report
from the American Heart Association. Circulation
133, 447–454 (2016).
143. Zhao,Z. etal. Body iron stores and heme-iron intake
in relation to risk of type2 diabetes: a systematic
review and meta-analysis. PLoS ONE 7, e41641
(2012).
144. Wu,J.H. etal. Omega-3 fatty acids and incident
type2 diabetes: a systematic review and meta-
analysis. Br. J.Nutr. 107, S214–S227 (2012).
145. Schulze,M.B. etal. Fiber and magnesium intake and
incidence of type2 diabetes: a prospective study and
meta-analysis. Arch. Intern. Med. 167, 956–965
(2007).
146. Dong, J. Y., Xun, P. & Qin, L. Q. Magnesium intake and
risk of type 2 diabetes: meta-analysis of prospective
cohort studies. Diabetes Care 34, 2116–2122 (2011).
147. Song,Y. etal. Blood 25-hydroxy vitaminD levels and
incident type2 diabetes: a meta-analysis of prospective
studies. Diabetes Care 36, 1422–1428 (2013).
REVIEWS
NATURE REVIEWS
|
ENDOCRINOLOGY VOLUME 14
|
FEBRUARY 2018
|
97

148. Pan,A. etal. Red meat consumption and risk of type2
diabetes: 3 cohorts of US adults and an updated meta-
analysis. Am. J.Clin. Nutr. 94, 1088–1096 (2011).
149. Hu,E.A., Pan,A., Malik,V. & Sun,Q. White rice
consumption and risk of type2 diabetes: meta-
analysis and systematic review. BMJ 344, e1454
(2012).
150. Carter,P., Gray,L.J., Troughton,J., Khunti,K. &
Davies,M.J. Fruit and vegetable intake and incidence
of type2 diabetes mellitus: systematic review and
meta-analysis. BMJ 341, c4229 (2010).
151. Cooper,A.J. etal. Fruit and vegetable intake and
type2 diabetes: EPIC-InterAct prospective study and
meta-analysis. Eur. J.Clin. Nutr. 66, 1082–1092
(2012).
152. Tong,X., Dong,J.Y., Wu,Z.W., Li,W. & Qin,L.Q.
Dairy consumption and risk of type2 diabetes
mellitus: a meta-analysis of cohort studies. Eur. J.Clin.
Nutr. 65, 1027–1031 (2011).
153. Aune,D., Norat,T., Romundstad,P. & Vatten,L.J.
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. 28, 845–858 (2013).
154. Malik,V.S. etal. Sugar-sweetened beverages and
risk of metabolic syndrome and type2 diabetes: a
meta-analysis. Diabetes Care 33, 2477–2483
(2010).
155. Romaguera,D. etal. Consumption of sweet beverages
and type2 diabetes incidence in European adults:
results from EPIC-InterAct. Diabetologia 56,
1520–1530 (2013).
156. Ding,M., Bhupathiraju,S.N., Chen,M., van
Dam,R.M. & Hu,F.B. Caffeinated and
decaffeinated coffee consumption and risk of type2
diabetes: a systematic review and a dose-response
meta-analysis. Diabetes Care 37, 569–586
(2014).
157. de Koning,L. etal. Diet-quality scores and the risk of
type2 diabetes in men. Diabetes Care 34, 1150–1156
(2011).
Acknowledgements
Y.Z. was supported by fellowship 7-12-MN-34 from the
American Diabetes Association.
Author contributions
Y.Z. and F.B.H. researched data for the article, contributed to
discussion of the content, wrote the article and reviewed and/
or edited the manuscript before submission. S.H.L. contrib-
uted to discussion of the content and reviewed and/or edited
the manuscript before submission.
Competing interests statement
The authors declare no competing interests.
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.
REVIEWS
98
|
FEBRUARY 2018
|
VOLUME 14 www.nature.com/nrendo