Type 2 Diabetes: Why We AreWinning the Battle but Losingthe War? 2015 Kelly West AwardLectureDiabetes Care 2016;39:653–663 | DOI: 10.2337/dc16-0205

Diabetes is among the biggest of the 21st-century global health challenges. In theU.S. and other high-income countries, thanks to investments in science, dedicationto implementing these findings, and measurement of quality of care, there havebeen improvements in diabetes management and declines in rate of diabetescomplications and mortality. This good news, however, is overshadowed by theever-increasing absolute numbers of people with diabetes and its complicationsand the unprecedented growth of diabetes in low- andmiddle-income countries ofthe world. To comprehensively win the war against diabetes requires 1) concertedattention to prevention and 2) expansion of global research to better informpopulation-level policies to curb diabetes but also to better understand individual-and population-level variations in pathophysiology and phenotypes globally so thatprevention and treatment can be tailored. For example, preliminary data show thatthin people in low- and middle-income countries such as India commonly experi-ence type 2 diabetes. Global studies comparing these thin Asian Indians with otherhigh-risk groups such as Pima Indians, a population with a high mean BMI, suggestthat type 2 diabetes may not be a single pathophysiological entity. Pima Indiansmay represent the well-studied phenotype of poor insulin action (type 2A),whereas Asian Indians represent the grossly understudied phenotype of poorinsulin secretion (type 2B). This has major implications for diagnosis, prevention,and treatment and highlights the mismatch between where diabetes burdensoccur (i.e., low- and middle-income countries) and where research happens (i.e.,high-income countries). Correcting this imbalance will advance our knowledgeand arsenal to win the global war against diabetes.

Type 2 diabetes is a prototypical disease at the cross-section of contemporaryglobalization and health. In the U.S. and other high-income countries, some successesare evident in preventing or postponing complications of the disease by better imple-mentation of quality of care. Yet, this “winning of a battle” hides a largerworrying issueof “losing the war,” stemming from the persistently high incidence of diabetes itself athome and from the expanding epidemic worldwide (Fig. 1). The war will not be wonwithout viewing type 2 diabetes in its global context as the world becomes rapidlymore interconnected in the midst of major demographic, economic, and environmen-tal transitions. Although themajority of the disease burden resides in low- andmiddle-income countries, research into diabetes remains concentrated in a few high-incomecountries. This discrepancy between where the preponderance of the disease burdenresides and where the research is conducted hampers our ability to better understand

Emory Global Diabetes Research Center, HubertDepartment of Global Health, Rollins School ofPublic Health, and Department of Medicine,School of Medicine, Emory University, Atlanta, GA

Corresponding author: K.M. Venkat Narayan,knaraya@emory.edu.

© 2016 by the American Diabetes Association.Readers may use this article as long as the workis properly cited, the use is educational and notfor profit, and the work is not altered.

K.M. Venkat Narayan

Diabetes Care Volume 39, May 2016 653

KELLY

WEST

AWARDLEC

TURE

the differences in the pathophysiology, orphenotypes, of the disease. For example,studies in populations, such as Asian In-dians, who have been exposed to gener-ations of undernutrition suggest thattype 2 diabetes may be highly prevalenteven in thin people andmaybedrivennotonly by propensity to fat storage and in-sulin resistancebut also primarily throughinnate and early problems with adequateinsulin secretion. Furthermore, the eti-ology, clinical presentation, diagnosis,treatment, and prevention may differ by

phenotypes. Studies in global settings,allowing for comparisons across popula-tions (e.g., Asian Indians vs. Pima Indians),can better inform differences in pheno-types. It is therefore time to considerthe pathophysiology of type 2 diabetesacross the spectrum, from those domi-nantly affected by insulin resistance (type2A) to those dominantly affectedby insulinsecretion (type 2B). With an increasinglyinterconnectedworld, investment inglobalcollaboration in research and policy will beneeded to win the war against diabetes.

WINNING THE BATTLE

Declines in Rate of ComplicationsAmong People With DiabetesAlthough diabetes (90–95% of which istype 2) remains a daunting public healthproblem, affecting 29.1 million individ-uals and costing $245 billion in the U.S.(1), there have been impressive im-provements in outcomes among peoplewith diabetes in the country over thepast two decades (2) (Fig. 1). Mortalityrates among both men and women withdiabetes in the U.S. have declined sub-stantially between 1997 and 2006 (3).Furthermore, rates of several diabetescomplications have also declined be-tween 1990 and 2010, including the in-cidence of acute myocardial infarction by67.8%, death from hyperglycemic crisisby 64.4%, stroke by 52.7%, amputationby 51.4%, and end-stage renal disease by28.3% (2). Such improvements are notlimited to the U.S., as improvements inoutcomes among people with diabeteshave also been observed in other high-income countries (4,5).

These improvements in diabetes com-plications are likely due to several factors;however, there are three that deservespecial mention: investments in science,institutional orientation toward translationand implementation, and quality-of-carebenchmarking efforts. First, investmentsin science leading to the development ofnew knowledge about the disease, bet-ter diagnostics, and a widening array oftreatment options are all paying off.Namely, large clinical trials such as theDiabetes Control and Complications Trial(DCCT), the UK Prospective DiabetesStudy (UKPDS), the Steno-2 study, andtheir successor mega-trials have helpedto shape our understanding of diabetesmanagement, treatment intensity andtargets, and clinical practice. Second,there has been an increased emphasison the implementation of proven inter-ventions into clinical and public healthpractice and policy. Specifically, attentionhas been given to translational researchto facilitate the implementation ofproven interventions by the Centers forDisease Control and Prevention (CDC),the National Institute of Diabetes and Di-gestive and Kidney Diseases (NIDDK), andthe American Diabetes Association (ADA)(6–8). Large multicenter translationalstudies, such as the Translating ResearchInto Action for Diabetes (TRIAD), funded

Figure 1—Why we are winning the battle but losing the war? NIH, National Institutes of Health.

654 Type 2 Diabetes: Winning the Battle but Losing the War? Diabetes Care Volume 39, May 2016

by the CDC, NIDDK, and the Departmentof Veterans Affairs, have generated valu-able information around key factors toimprove quality of care (9). Factors at thelevel of the health systems (i.e., financing,electronic record systems), disease man-agement strategies (i.e., care coordina-tion, diabetes teams, physician–patientcommunication), physician reimburse-ment (i.e., incentivizing quality as op-posed to volume of services), and thepatient (i.e., reducing out-of-pocket ex-penses, patient education and empower-ment) each positively impact quality ofcare (10–14). Third, the measurementof quality of care, led by the nationalDiabetes Quality Improvement Project(DQIP), working through a coalition of in-fluential private and public national organi-zations, and later to become the NationalDiabetes Quality Improvement Alliance(NDQIA), has helped to focus attention onimplementation. The NDQIA develops,maintains, and promotes the use of an up-dated standardized measurement set (theNDQIA measures) for quality of diabetescare. Monitoring and reporting of qualityof careamongpeoplewithdiabetesnation-ally by the CDC and NIDDK have focusedattention on gaps and documented signifi-cant improvements in diabetes processesand intermediate outcomes (15–17). Forexample, suchdocumentationhas revealedthat control of vascular risk factors, HbA1c,bloodpressure, and LDL cholesterol amongpeoplewith diabetes have all improved inthe period 1999–2000 (17).

LOSING THE WAR

Although the declines in the rates ofdiabetes-related complications during

the past two decades are noteworthy,we are still confronted by dauntingchallenges; namely, 1) the persistentlyhighprevalence and incidence of diabetesin the U.S. (Fig. 2) and 2) the explosion oftype 2 diabetes globally.

Persistently High Prevalence andIncidence of Diabetes in the U.S.The prevalence of diabetes in the U.S.has been rising at least since 1980,while the rising incidence of diabetesseems to have recently plateaued (18).Furthermore, the lifetime risk of diabetesin U.S. has been increasing over time:from 20.7% and 27.5% for males andfemales, respectively, born between1985 and 1989, to 32.8% and 38.5%for those born in 2000 (19), to 40.2%and 39.5% for those born between2000 and 2011 (20). The resulting con-sequences are that overall years spentwith diabetes have increased by 156%for males and 70% for females (20) in thepast 30 years. These increases in lifetimerisk of diabetes are driven by two factors:1) improved survival among those withdiabetes, thanks to better implementa-tion of proven interventions to preventcomplications and delay mortality, and2) continued high diabetes incidence.For a chronic disease such as type 2 di-abetes, even small increases in incidencecan have a dominant impact on lifetimerisk and on population prevalence. As anillustration, the small increase in diabetesincidence in the U.S. between 2000 and2004 resulted in 12millionmoreprojectednumbers of people with diabetes by2050 (21). Furthermore, an estimated86 million people in the U.S. have someform of prediabetes (impaired fasting

glucose [IFG] and/or impaired glucosetolerance [IGT]) (1), and the annualrate of diabetes conversion in thisgroup is severalfold than in peoplewith normoglycemia (22). Thus, theimpending growth of individuals withdiabetes should be of great concern.

As a consequence of the increasingtrend in numbers of people with diabetesin the U.S., even if the rate of complica-tions is decreasing, the absolute numbersof people in the country with complica-tions will continue to rise (Fig. 3). This willhave enormous implications for healthcare systems and payers, as diabetesis already a costly disease (23) and wasthe leading contributor for inflation-adjusted health care spending amongMedicare beneficiaries in 1997–2006in the U.S. (24) and there will be ever-increasing numbers of people with diabe-tes and its complications to be managed.

Explosion of Type 2 Diabetes GloballyA larger worrying story, which the U.S.,with its large immigrant population andglobal interconnectedness, would ig-nore at its own peril, is that diabetes hasemerged as a major public health prob-lem worldwide, with pandemic growthdriven by changing demographic, socio-economic, and lifestyle patterns acrossthe globe. The International DiabetesFederation (IDF) estimates that therewere 415 million people with diabetesin 2015 and projects that the absolutenumber will reach 642 million by 2040,affecting all regions of the world (25)(Fig. 4). Current estimates suggest thatthree-quarters of those affected by dia-betes live in low- andmiddle-incomecoun-tries, where the disease disproportionately

Figure 2—Trends in age-adjusted diagnosed diabetes prevalence and incidence among adults aged 20–79 years, 1980–2012. Data from the NationalHealth Interview Survey. Reproduced with permission from Geiss et al. (18).

care.diabetesjournals.org Narayan 655

affects younger people at the prime oftheir economic productivity. The overallincrease in diabetes prevalence has alsobeen steeper in low- and middle-incomecountries than in affluent high-incomecountries (26). The IDF also estimatesthat globally at least 316 million people(6.9% of the world’s adult population)had IGT in 2013 (27) and that low- andmiddle-income countries will experiencea 50% increase in IGT prevalence by 2035,

compared with 41% increase in high-income countries (27).

HOW TO WIN THE WAR?

Emphasize Prevention and GlobalResearchGiven the ever-growing and worldwideburden of diabetes, driven by high inci-dence and high numbers of peoplewith complications, emphasis on pri-mary prevention is critical (Fig. 1). The

emphasis on the prevention of type 2 di-abetes requires two sets of strategies: 1)effective implementation of proven pre-ventive interventions in people at highrisk (i.e., IGT) in the immediate term and2) aggressive long-term global collabora-tion and research to a) address gaps inknowledge in population-based primaryprevention for improving diet and activityat the societal level and b) advance theunderstanding of differences in patho-physiology, especially in hitherto under-studied but very large populations in therapidly developing economies of theworld.

Effective Implementation of Proven

Preventive Interventions in People

at High Risk

Strong evidence from a number of coun-tries, highlighting the global nature ofscience today, demonstrates that lifestyleintervention and also metformin use inpeople with IGT can prevent or slow pro-gression to diabetes (28–30). Evidencealso exists that interventions applied topeople with IGT are cost-effective andcan reduce diabetes complications, suchas cardiovascular mortality and retinopa-thy, and can improve quality of life (31).

Several barriers need to be overcometo effectively implement these preven-tive interventions for people with predia-betes at scale (22,32). First, despite thehigh prevalence of prediabetes (i.e., oneof every three adults) in the U.S., almost90% of people with prediabetes remainundetected (33). Improved detection ofprediabetes followed by implementationof proven interventions for prevention inhigh-risk groups will slow the expansionof diabetes. Over the years, however,there has been considerable debatearound the evidence for the benefitsand costs of screening for prediabetesand diabetes (34,35). Screening policieshave varied from the liberal position ofthe ADA (36) to a more restrictive policyof the United States Preventive ServicesTask Force (USPSTF) (37). Much newevidence has accumulated over the pastdecade, and current evidence tips thescale toward screening for hyperglycemia(31), which can also be viewed as agateway to diabetes prevention andmanagement (38). TheUSPSTF has recentlyissued a broader set of criteria for screen-ing, which, while moving in the right di-rection, still does not address the issue ofnonobese people with prediabetes (39).

Figure 3—Initiation of treatment for kidney failure related to diabetes (A) and hospital dis-charges for nontraumatic lower-extremity amputation with diabetes as a listed diagnosis (B)in the U.S., 1990–2011. Lines represent age-adjusted rates, and bars represent number of cases.Data from the CDCNational Surveillance System (94) and personal communication (N.R. Burrows).

656 Type 2 Diabetes: Winning the Battle but Losing the War? Diabetes Care Volume 39, May 2016

Screening for prediabetes, however,is only the first step and needs to be cou-pled with strengthening of infrastructureand financing mechanisms to sustainablyand effectively deliver interventions (22).A recent systematic review indicates thattranslation of diabetes prevention in real-life settings is feasible (40), and a varietyof successful initiatives already existto implement diabetes preventionnationally by using existing communityor social resources and networks (41,42).Expanding such initiatives across thecountry is needed, and some interna-tional examplesmay also serve as usefulmodels (43). Although lifestyle inter-ventions should be the primary strategyfor the prevention of diabetes amongpeople with prediabetes, there are dataon effectiveness and cost-effectivenessthat also support the use of metforminin people at high risk (22,44). Yet, indica-tion for metformin use in people withprediabetes to prevent diabetes inci-dence is still not approved, and use ofmetformin among those at high risk ofdiabetes is exceedingly low (45).Although unequivocal evidence exists

for prevention of diabetes in high-riskgroups (i.e., IGT) and should be aggres-sively pursued in the immediate term,data suggest that the impact on diabe-tes prevalence from this approach alonewould be rather modest and the U.S.diabetes prevalence would still be likelyto increase dramatically over the next 20years (46). Therefore, although focusingon implementing currently proven inter-ventions in people with IGT is importantin the short term, a longer-term strategy

for effective prevention will need to gobeyond this narrow high-risk groupand address the major gaps in knowledgeconcerning improving diet andphysical ac-tivity in the population at large. Broaderdiabetes prevention efforts will also needresearch to better understand the patho-physiology of diabetes, so that more tai-lored measures of prevention can beevolved for various subpopulations.

Aggressive Long-term Global Collaboration

and Research

In today’s increasingly globalized world,driven by unprecedented movement oftrade, capital, goods, people, ideas, and

information, every challenge confrontingcomprehensive type 2 diabetes preven-tion is shared across the world in com-mon, and solutions are also likely torequire global collaboration. Clues fromglobal health research can help our sharedunderstanding of pathophysiology andcross-border sharing of best practices inprevention, management, and policy fordiabetes.

Address Gaps in Knowledge in Population-

Based Primary Prevention. Population-based approaches aimed at improvingdiet and physical activity of the entiresociety to primarily prevent diabetesare clearly very appealing. However,few data are available regarding inter-ventions to prevent diabetes in peoplewith normoglycemia (22). Furthermore,multipronged community-based strategiesthat have been tested at the populationlevel for preventing diabetes have thusfar not demonstrated encouraging results(47). There are major impediments tostrategies that can improve healthy life-styles at the societal level. For example,despite knowing that adequate fruit andvegetable intake is recommended, wecannot implement targets because theglobal supply does not meet the demand,and analysis of data indicates that globallythere is, on average, a 22% shortage inthe supply of fruits and vegetables tomeet this demand based on a recom-mended intake of at least five portions

Figure 4—Comparative prevalence of diabetes in people aged 20–79 years by world regions.Data from IDF Diabetes Atlas (27).

Figure 5—Diabetes prevalence, by region of birth and BMI category: National Health InterviewSurvey. Reproduced with permission from Oza-Frank and Narayan (61).

care.diabetesjournals.org Narayan 657

each day. Furthermore, this shortage is ashigh as 58% in low-income countries (48).We therefore need more research in or-der to know what to do to improve thesupply of healthy foods and to make itaffordable.It is nowwidely recognized that type 2

diabetes is occurring with worryingfrequency at younger ages, especiallyaffecting large numbers of people inlow- and middle-income countries. Firstnoted in the late 1990s as an emergingpublic health problem (49,50), data frominvestigations, such as the SEARCH forDiabetes in Youth (SEARCH) study, havenow carefully documented the pre-valence and incidence of youth-onsettype 2 diabetes and highlighted riskfactors for it, notably, ethnicity (e.g.,Native Americans, Asian Pacific Islanders)(51) and childhood obesity (52,53). Astudy of a nationally representativesample of over 7,000 schoolchildren inthe U.S., followed from ages 5 to 14withserial objective measures of anthropom-etry, showed that by age 5, 14.9% of chil-dren were already overweight and 12.5%were obese (54). Importantly, althoughprevalence of obesity increased withage, incidence decreased with age andwas highest at younger ages. Further-more, incidence of obesity from ages 5to 14 was four times higher in childrenwho were overweight at age 5 comparedwith those who were normal weight atthat age. In addition, a large randomizedcontrolled trial of a comprehensiveschool-based program found no differ-ence in risk of obesity between interven-tion and control schools (55). These datataken together suggest that the anteced-ents of childhood obesity and thus ofyoung-onset type 2 diabetes are likely es-tablished well before school age and thatthe window of opportunity for interven-tion is preschool or sooner, perhaps even

in utero (54). However, we need a betterand fuller understanding of the early-lifefactors that predispose individuals andsubpopulations to obesity and type 2diabetes, and global studies are likelyto be of great value, as childhood obesityand young-onset type 2 diabetes arehappening at quickening pace, especiallyin economically transforming low- andmiddle-income countries (56).Advance the Understanding of Differences in

Pathophysiology. Viewing type 2 diabetesin its global context offers exciting andproductive opportunities for a better un-derstanding its complex pathophysiol-ogy. For example, people of South Asiandescent, a population of nearly 1.7 billionpeople worldwide, are at extremely highrisk for type 2 diabetes and have uniquesusceptibilities to the disease (57,58).The prevalence of diabetes in India has

continued to rise dramatically over recentdecades (59), and data now indicate ahigher prevalence in Indians in urbanIndia than in Indian migrants in theU.S. (60). Furthermore, among foreign-born people in the U.S., a fast-growingpopulation of over 36 million, peoplefrom the Indian subcontinent have thehighest diabetes prevalence among allimmigrants to the country (61) (Fig. 5).South Asians have the highest diabetesprevalence among all ethnic groups in theU.S., other than Native Americans (62,63),and develop the disease at younger agesand at lower levels of BMI (61,64–66).

South Asians also exhibit other uniquefeatures in terms of diabetes develop-ment. For example, South Asians werefive to nine times more likely than otherracial/ethnic groups to exhibit dysglycemiaand dyslipidemia (characterized by low

Table 1—Comparison of diabetes in Pima Indians and Asian Indians

Pima Indians Asian Indians (Chennai, India)

Prevalence of diabetes* Very high (50% by age 55 years) Very high (50% by age 55 years)

Obesity profile Very obese (mean BMI 33.7 kg/m2, mean waistcircumference 108.6 cm)

Relatively thinner (mean BMI 25.7 kg/m2, mean waistcircumference 83.1 cm)

Glucose profile Higher mean 2-h plasma glucose Higher mean fasting plasma glucose

Prediabetes distribution Nearly two-thirds of prediabetes is isolated IGT Over two-thirds of prediabetes is isolated IFG

Insulin resistance vs.secretion

2–4.5 times more insulin resistant than Asian Indiansacross BMI and glucose strata

One-half to one-third of insulin secretion in Pima Indiansacross BMI and glucose strata

*Data from Gila River Indian community in Arizona cohort (96), data from the National Health and Nutrition Examination Survey (NHANES) andCentre for Cardiometabolic Risk Reduction in South Asia (CARRS) surveillance study (U. Gujral, personal communication), and comparisons betweenthe Gila River Indian community in Arizona and the CARRS surveillance study (97).

Figure 6—Heterogeneity of type 2 diabetes: a hypothesis.

658 Type 2 Diabetes: Winning the Battle but Losing the War? Diabetes Care Volume 39, May 2016

HDL cholesterol and high triglycerides),even at a healthy BMI (67). The SouthAsian populationmay offer unique oppor-tunities to investigate the causes oftype 2 diabetes and cardiometabolicmetabolic diseases in people who arerelatively thin, a phenomenon being re-ported in other populations either livingin or whose ancestry is from parts ofthe world undergoing recent economicdevelopment (68,69). This concept of aunique South Asian phenotype has beendescribed by others (70,71) and goesback several centuries. Diabetes wasdescribed in ancient times in the prein-sulin era by the Egyptians, Arabs, Greeks,and Indians. A sixth-century IndianAyurvedic textbook, Charaka Samhita,even identified two forms of diabetes:“There are two forms of diabetes(Madhu Meha or honey urine): one as-sociated with emaciation, dehydration,polyuria, and lassitude, and the otherwith stout build, gluttony, obesity, andsleepiness” (72).An important area for future investi-

gation of phenotypes relevant to type 2diabetes prevention and treatment isthe role of innate or early problemswith insulin secretion. In fact, someauthors have recently called for a reclas-sification of diabetes with a greaterb-cell–centric approach (73). Geneticstudies add to this discourse and revealtwo insights: 1) that there are only a fewoverlaps between genes identified fortype 2 diabetes and those for obesity,thereby indicating that pathophysiologicalprocesses other than obesity and insulinresistancemay be important in the devel-opment of diabetes (74), and 2) that themajority of type 2 diabetes genes seem tobe related to the b-cell rather than to in-sulin resistance. South Asians may be agood population group to investigatethe role of b-cell dysfunction in patho-genic progression of type 2 diabetes.Population-based data comparing

U.S. Pima Indians and Asian Indiansfrom Chennai, India, have revealed severalintriguing findings, notably, the relativelypoorer insulin secretion at baseline and30 and 120 min after an oral glucosetolerance test and the relatively higherproportion of isolated IFG (Table 1). Asisolated IFG is primarily the result of he-patic insulin resistance, with early-phase,stationary impairment in b-cell function(75), and is diagnosed as a fasting plasmaglucose $5.6 mmol/L and ,7.0 mmol/L

(76), these comparisons indicate thatthese two high prevalence groups mayhave very different pathways to devel-oping diabetes. The Asian Indian type 2diabetes phenotype may involve greaterinnate insulin secretion problems, both inthe fasting state and in response to glu-cose challenges. This has implications forprevention and treatment. In a trial of astepwise strategy of lifestyle interventionand metformin, when required, amongSouth Asians with prediabetes, there wasan overall relative risk reduction of 32% of

progression to diabetes. Importantly, inthe group with isolated IFG, the relativerisk reduction was a meager 12% (77).Similarly, a study from Japan has alsoshown a null effect of lifestyle interventionamong Japanese people with isolatedIFG (78). The results of these studies indi-cate that measures to improve insulin ac-tion may have limited effect in peoplewhose primary problemmay be with in-sulin secretion and that other types ofinterventions may be needed in thesepopulations.

Figure 7—World map sized to relative proportion of country population (A), prevalence ofunderweight in children (B), diabetes prevalence (C ), and birth cohorts (D) available forresearch (95).

care.diabetesjournals.org Narayan 659

Other investigations also lend supportto the possibility that b-cell function maybe declining very early in the natural his-tory among South Asians. For example,in a cross-sectional study of South AsianIndians with different glycemic status,b-cell function was reduced in thosewith even mild dysglycemia (e.g., fastingglucose 100–126 mg/dL and/or 2-hglucose 140–199 mg/dL) regardless ofage, adiposity, insulin sensitivity, or familyhistory (79). b-Cell function was alsomore strongly associated with prediabe-tes and type 2 diabetes than with insulin

resistance in young adults in Chennai,India, and in a cohort of South AsianIndian migrants in the U.S. (80,81).Comparisons between South Asiansand other racial/ethnic groups havealso been informative. For example, ananalysis from the Whitehall II cohortstudy in the U.K. suggested that SouthAsians may have a poorer b-cell reserverelative to Europeans (82), and a study intheU.S. showed South Asians had poorerb-cell function compared with whites,blacks, Hispanics, and Chinese Ameri-cans (83). A study in the Netherlands

suggested that family members of SouthAsians with type 2 diabetes may havepoorer b-cell adaptation than similarDutch individuals (84). Although informa-tive, most of the published data suggest-ing that South Asian populations mayhave susceptibility for poor b-cell func-tion are from cross-sectional studies,and investments in longitudinal multi-ethnic studies are needed to establishcausality and mechanism.

The role of early b-cell dysfunction inthe pathophysiology of diabetes may beespecially important in populations orsubgroups where prevalence of type 2diabetes ishighat lowBMIs, aphenomenonthat is quite common in populations ofemerging economies, where nutritiontransitions are overlapping with centu-ries of undernutrition (68,69). For exam-ple, in a population-based study inChennai, India, approximately 20% ofpeople with BMI ,18.5 kg/m2 hadtype 2 diabetes (U. Gujral, personal com-munication) and overall diabetes preva-lence was high, even though more thantwo-thirds of men and half of womenhad BMI,25 kg/m2. Intriguingly, histor-ical data suggest that the South Asianpopulation may have had good nutritionstatus in the Mesolithic period, as indi-cated by their tallness (85), and that thepopulation may have been growing un-dernourished for generations. This maypoint to the role of maternal and/orearly childhood nutrition, as the windowof maximal opportunity for height, amarker of maternal and early nutritionand possibly of metabolic capacity, maybe in the first 3 years of life (86,87). Anumber of studies also support therole of transgenerational metabolic path-ways linking early malnourishment withdiabetes risk (88,89).

Time for New Classification of Type 2Diabetes?Although both insulin resistance andimpaired insulin secretion are importantin the pathophysiology of type 2 diabetes,studies in populations, such as SouthAsians, especially in comparison withgroups such as the Pima Indians, raisethe question of whether it is time toseriously consider heterogeneity withintype2diabetesdevelopmentanddiagnosis.In particular, could there be different phe-notypic presentations of type 2 diabetesdescribed by the relative roles of insulin re-sistance and b-cell function? Is it time to

Figure 7—Continued.

660 Type 2 Diabetes: Winning the Battle but Losing the War? Diabetes Care Volume 39, May 2016

consider the pathophysiology of type 2diabetes across the spectrum, from thosedominantly affected by insulin resistance(type 2A) to those dominantly affected byinsulin secretion (type 2B)? For example,the traditional and well-studied pheno-type of type 2 diabetes, mainly amanifes-tation of increased metabolic load,defined largely by impaired insulin ac-tion, and driven by adiposity, physicalinactivity, and adult dietary factors, maybe a distinct type (Fig. 6), whereas the farlesser studied phenotype, characterizedby reduced metabolic capacity and im-paired b-cell function and driven byhitherto incompletely understood fac-tors, such as maternal and child nutri-tion, reduced b-cell mass, microbiomes,and endocrine disruptors, may be an-other. Studies across global populationsmay shed further insights into type 2 di-abetes phenotypes and may point to dif-ferences in etiology, pathophysiology,diagnosis, prevention, and treatment bydiabetes subtypes.

CONCLUSIONS

The reductions in complication ratesamong people with type 2 diabetes inthe U.S. and in high-income countriesare encouraging, but the war againstdiabetes needs a new era of informedprevention through expanded globalcollaboration in research and policy.Although the global burden of type 2diabetes points to the need for aggres-sive global collaboration in science toadvance prevention and treatment,there is currently a huge mismatch be-tweenwhere the diabetes burdens resideglobally and where research is concen-trated (Fig. 7). Correcting this imbalancewill be important, and research intotype 2 diabetes and related noncom-municable diseases in low- and middle-income countries offers huge scientificopportunities (90). Perhaps there arelessons to be learned for type 2 diabetesand other noncommunicable diseasesfrom how global collaborations in scienceand policy have helped with the fightagainst HIV and AIDS (91,92). We arenow in a highly interconnected globalworld and all problemsdwhether it isclimate change, water security, HIV, orEboladcut across national borders inways unimaginable, and efficient and ef-fective solutions need expanded globalengagement in science (93). The waragainst type 2 diabetes is no exception

and cannot be comprehensively wonwithout the investment in collaborativeglobal diabetes research.

Acknowledgments.K.M.V.N. especially thanksDr. Peter Bennett for inspiring him into diabetesepidemiology andDr.WilliamKnowler for teach-ing him the rigors of science and is grateful to theinnumerable colleagues at the Phoenix Epide-miology and Clinical Research Branch at theNIDDK, at the Division of Diabetes Translationat the CDC, in the several national and interna-tional multicenter studies he has had the privi-lege to be involved with, and at the Emory GlobalDiabetes Research Center. K.M.V.N. also thanksMohammed Ali, Lisa Staimez, Shivani Patel, MaryBeth Weber, Unjali Gujral, Samrawit Yisahak, andMark Hutcheson for their valuable comments onthe manuscript.Duality of Interest. No potential conflicts ofinterest relevant to this article were reported.

References1. Division of Diabetes Translation, NationalCenter for Chronic Disease Prevention andHealth Promotion, Centers for Disease Controland Prevention. National Diabetes StatisticsReport, 2014 [Internet]. Available from http://www.cdc.gov/diabetes/pubs/statsreport14/national-diabetes-report-web.pdf. Accessed 23January 20162. Gregg EW, Li Y, Wang J, et al. Changes indiabetes-related complications in the UnitedStates, 1990-2010. N Engl J Med 2014;370:1514–15233. Gregg EW, Cheng YJ, Saydah S, et al. Trendsin death rates among U.S. adults with andwithout diabetes between 1997 and 2006: find-ings from the National Health Interview Survey.Diabetes Care 2012;35:1252–12574. Kennon B, Leese GP, Cochrane L, et al.Reduced incidence of lower-extremity ampu-tations in people with diabetes in Scotland: anationwide study. Diabetes Care 2012;35:2588–25905. Ikonen TS, Sund R, Venermo M, Winell K.Fewer major amputations among individualswith diabetes in Finland in 1997-2007: a pop-ulation-based study. Diabetes Care 2010;33:2598–26036. Narayan KM, Benjamin E, Gregg EW, Norris SL,Engelgau MM. Diabetes translation research:where are we and where do we want to be? AnnIntern Med 2004;140:958–9637. Narayan KM, Gregg EW, Engelgau MM,et al. Translation research for chronic disease:the case of diabetes. Diabetes Care 2000;23:1794–17988. Garfield SA, Malozowski S, Chin MH, et al.;Diabetes Mellitus Interagency CoordinatingCommittee (DIMCC) Translation ConferenceWorking Group. Considerations for diabetestranslational research in real-world settings.Diabetes Care 2003;26:2670–26749. TRIAD Study Group. Health systems, patientsfactors, andquality of care for diabetes: a synthesisof findings from the TRIAD study. Diabetes Care2010;33:940–94710. Kerr EA, Gerzoff RB, Krein SL, et al. Diabetescare quality in the Veterans Affairs Health Care

Systemand commercialmanaged care: the TRIADstudy. Ann Intern Med 2004;141:272–28111. Mangione CM, Gerzoff RB, Williamson DF,et al.; TRIAD Study Group. The associationbetween quality of care and the intensity ofdiabetes disease management programs. AnnIntern Med 2006;145:107–11612. Schmittdiel JA, Uratsu CS, Karter AJ, et al.Why don’t diabetes patients achieve recom-mended risk factor targets? Poor adherenceversus lack of treatment intensification. J GenIntern Med 2008;23:588–59413. Ettner SL, Thompson TJ, Stevens MR, et al.;TRIAD Study Group. Are physician reimburse-ment strategies associated with processes ofcare and patient satisfaction for patients withdiabetes in managed care? Health Serv Res2006;41:1221–124114. Karter AJ, Stevens MR, Herman WH, et al.;Translating Research Into Action for DiabetesStudy Group. Out-of-pocket costs and diabetespreventive services: the Translating ResearchInto Action for Diabetes (TRIAD) study. DiabetesCare 2003;26:2294–229915. Saaddine JB, Engelgau MM, Beckles GL,Gregg EW, Thompson TJ, Narayan KM. A diabetesreport card for the United States: quality of carein the 1990s. Ann Intern Med 2002;136:565–57416. Saaddine JB, Cadwell B, Gregg EW, et al.Improvements in diabetes processes of careand intermediate outcomes: United States,1988-2002. Ann Intern Med 2006;144:465–47417. Ali MK, Bullard KM, Saaddine JB, Cowie CC,Imperatore G, Gregg EW. Achievement of goalsin U.S. diabetes care, 1999-2010. N Engl J Med2013;368:1613–162418. Geiss LS, Wang J, Cheng YJ, et al. Prevalenceand incidence trends for diagnosed diabetesamong adults aged 20 to 79 years, UnitedStates, 1980-2012. JAMA 2014;312:1218–122619. Narayan KM, Boyle JP, Thompson TJ,Sorensen SW, Williamson DF. Lifetime risk fordiabetes mellitus in the United States. JAMA2003;290:1884–189020. Gregg EW, Zhuo X, Cheng YJ, Albright AL,Narayan KM, Thompson TJ. Trends in lifetimerisk and years of life lost due to diabetes inthe USA, 1985-2011: a modelling study. LancetDiabetes Endocrinol 2014;2:867–87421. Narayan KM, Boyle JP, Geiss LS, SaaddineJB, Thompson TJ. Impact of recent increase inincidence on future diabetes burden: U.S.,2005-2050. Diabetes Care 2006;29:2114–211622. Narayan KM, Williamson DF. Prevention oftype 2 diabetes: risk status, clinic, and community.J Gen Intern Med 2010;25:154–15723. American Diabetes Association. Economiccosts of diabetes in the U.S. in 2007. DiabetesCare 2008;31:596–61524. Thorpe KE, Florence CS, Howard DH, Joski P.The impact of obesity on rising medical spending.Health Aff (Millwood) 2004;23(Suppl.):W4-480–4-48625. InternationalDiabetesFederation. IDFDiabetesAtlas. 7th ed. Brussels, Belgium: InternationalDiabetes Federation, 201526. Hwang CK, Han PV, Zabetian A, Ali MK,Narayan KM. Rural diabetes prevalence quintu-ples over twenty-five years in low- and middle-income countries: a systematic review and

care.diabetesjournals.org Narayan 661

meta-analysis. Diabetes Res Clin Pract 2012;96:271–28527. International Diabetes Federation. IDF Dia-betes Atlas. 6th ed. Brussels, Belgium, Interna-tional Diabetes Federation, 201328. Knowler WC, Barrett-Connor E, Fowler SE,et al.; Diabetes Prevention Program ResearchGroup. Reduction in the incidence of type 2 di-abetes with lifestyle intervention or metformin.N Engl J Med 2002;346:393–40329. Tuomilehto J, Lindstrom J, Eriksson JG,et al.; Finnish Diabetes Prevention Study Group.Prevention of type 2 diabetes mellitus bychanges in lifestyle among subjects with im-paired glucose tolerance. N Engl J Med 2001;344:1343–135030. Ramachandran A, Snehalatha C, Mary S,Mukesh B, Bhaskar AD, Vijay V; Indian DiabetesPrevention Programme (IDPP). The Indian Dia-betes Prevention Programme shows that life-style modification and metformin prevent type 2diabetes in Asian Indian subjects with impairedglucose tolerance (IDPP-1). Diabetologia 2006;49:289–29731. Narayan KM, Gujral UP. Evidence tips thescale toward screening for hyperglycemia. Dia-betes Care 2015;38:1399–140132. Fradkin JE, Roberts BT, Rodgers GP. What’spreventing us from preventing type 2 diabetes?N Engl J Med 2012;367:1177–117933. Centers for Disease Control and Prevention(CDC). Awareness of prediabetes–United States,2005-2010.MMWRMorbMortal Wkly Rep 2013;62:209–21234. Engelgau MM, Narayan KM, Herman WH.Screening for type 2 diabetes. Diabetes Care2000;23:1563–158035. Echouffo-Tcheugui JB, Ali MK, Griffin SJ,Narayan KM. Screening for type 2 diabetesand dysglycemia. Epidemiol Rev 2011;33:63–8736. American Diabetes Association. Standardsof medical care in diabetesd2010. DiabetesCare 2010;33(Suppl. 1):S11–S6137. U.S. Preventive Services Task Force. Screen-ing for type 2 diabetes mellitus in adults: U.S.Preventive Services Task Force recommenda-tion statement. Ann Intern Med 2008;148:846–85438. Narayan KM, Weber MB. Screening for hy-perglycemia: the gateway to diabetes preven-tion and management for all Americans. AnnIntern Med 2015;162:795–79639. Siu AL; U.S. Preventive Services Task Force.Screening for abnormal blood glucose andtype 2 diabetes mellitus: U.S. Preventive Ser-vices Task Force recommendation statement.Ann Intern Med 2015;163:861–86840. Ali MK, Echouffo-Tcheugui J, WilliamsonDF. How effective were lifestyle interventionsin real-world settings that were modeled onthe Diabetes Prevention Program? Health Aff(Millwood) 2012;31:67–7541. Ackermann RT, Finch EA, Brizendine E, ZhouH, Marrero DG. Translating the Diabetes Pre-vention Program into the community. TheDEPLOY Pilot Study. Am J Prev Med 2008;35:357–36342. Sattin RW, Williams LB, Dias J, et al. Commu-nity trial of a faith-based lifestyle interventionto prevent diabetes among African-Americans.J Community Health 2016;41:87–96

43. Saaristo T, Moilanen L, Korpi-Hyovalti E,et al. Lifestyle intervention for prevention oftype 2 diabetes in primary health care: one-year follow-up of the Finnish National DiabetesPrevention Program (FIN-D2D). Diabetes Care2010;33:2146–215144. Chatterjee R, Narayan KM, Lipscomb J,Phillips LS. Screening adults for pre-diabetesand diabetes may be cost-saving. DiabetesCare 2010;33:1484–149045. Moin T, Li J, Duru OK, et al. Metformin pre-scription for insured adults with prediabetesfrom 2010 to 2012: a retrospective cohortstudy. Ann Intern Med 2015;162:542–54846. Gregg EW, Boyle JP, Thompson TJ, BarkerLE, Albright AL, Williamson DF. Modeling theimpact of prevention policies on future diabetesprevalence in the United States: 2010-2030.Popul Health Metr 2013;11:1847. Satterfield DW, Volansky M, Caspersen CJ,et al. Community-based lifestyle interventionsto prevent type 2 diabetes. Diabetes Care 2003;26:2643–265248. Siegel KR, Ali MK, Srinivasiah A, Nugent RA,Narayan KM. Do we produce enough fruits andvegetables to meet global health need? PLoSONE 2014;9:e10405949. Fagot-Campagna A, Pettitt DJ, EngelgauMM, et al. Type 2 diabetes among North ado-lescents: an epidemiologic review and a publichealth perspective. J Pediatr 2000;136:664–67250. Narayan KM, Fagot-Campagna A, ImperatoreG. Type 2 diabetes in children: a problem lurkingfor India? Indian Pediatr 2001;38:701–70451. Dabelea D, Bell RA, D’Agostino RB Jr, et al.;Writing Group for the SEARCH for Diabetes inYouth Study Group. Incidence of diabetes inyouth in the United States. JAMA 2007;297:2716–272452. Hannon TS, Rao G, Arslanian SA. Childhoodobesity and type 2 diabetes mellitus. Pediatrics2005;116:473–48053. Kaufman FR. Type 2 diabetes mellitus inchildren and youth: a new epidemic. J PediatrEndocrinol Metab 2002;15(Suppl. 2):737–74454. Cunningham SA, Kramer MR, Narayan KM.Incidence of childhood obesity in the UnitedStates. N Engl J Med 2014;370:403–41155. Foster GD, Linder B, Baranowski T, et al.;HEALTHY Study Group. A school-based interven-tion for diabetes risk reduction. N Engl J Med2010;363:443–45356. Lobstein T, Jackson-Leach R, Moodie ML,et al. Child and adolescent obesity: part of abigger picture. Lancet 2015;385:2510–252057. Gujral UP, Pradeepa R, Weber MB, NarayanKM, Mohan V. Type 2 diabetes in South Asians:similarities and differences with white Cauca-sian and other populations. Ann N Y Acad Sci2013;1281:51–6358. Sattar N, Gill JM. Type 2 diabetes in migrantsouth Asians: mechanisms, mitigation, andmanagement. Lancet Diabetes Endocrinol2015;3:1004–101659. Mohan V, Deepa M, Deepa R, et al. Seculartrends in the prevalence of diabetes and im-paired glucose tolerance in urban South India–the Chennai Urban Rural Epidemiology Study(CURES-17). Diabetologia 2006;49:1175–117860. Gujral UP, Narayan KM, Pradeepa RG, et al.Comparing type 2 diabetes, prediabetes, and

their associated risk factors in Asian Indians inIndia and in the U.S.: the CARRS and MASALAStudies. Diabetes Care 2015;38:1312–131861. Oza-Frank R, Narayan KM. Overweight anddiabetes prevalence among US immigrants. AmJ Public Health 2010;100:661–66862. Venkataraman R, Nanda NC, Baweja G,Parikh N, Bhatia V. Prevalence of diabetes mel-litus and related conditions in Asian Indians liv-ing in the United States. Am J Cardiol 2004;94:977–98063. Lee JWR, Brancati FL, Yeh H-C. Trends in theprevalence of type 2 diabetes in Asians versuswhites: results from the United States NationalHealth Interview Survey, 1997-2008. DiabetesCare 2011;34:353–35764. Oza-Frank R, Ali MK, Vaccarino V, NarayanKM. Asian Americans: diabetes prevalenceacross U.S. and World Health Organizationweight classifications. Diabetes Care 2009;32:1644–164665. Mather HM, Keen H. The Southall DiabetesSurvey: prevalence of known diabetes in Asiansand Europeans. Br Med J (Clin Res Ed) 1985;291:1081–108466. Creatore MI, Moineddin R, Booth G, et al.Age- and sex-related prevalence of diabetesmellitus among immigrants to Ontario, Canada.CMAJ 2010;182:781–78967. Patel S, Shivashankar R, Ali MK, et al. Is the“South Asian phenotype” unique to SouthAsians? Comparing cardiometabolic risk factorsin the CARRS and NHANES studies. Glob Heart.In press68. Maskarinec G, Grandinetti A, Matsuura G,et al. Diabetes prevalence and body mass indexdiffer by ethnicity: the Multiethnic Cohort. EthnDis 2009;19:49–5569. Chan WB, Tong PCY, Chow CC, et al. Theassociations of body mass index, C-peptideand metabolic status in Chinese type 2 diabeticpatients. Diabet Med 2004;21:349–35370. Unnikrishnan R, Anjana RM, Mohan V. Di-abetes in South Asians: is the phenotype differ-ent? Diabetes 2014;63:53–5571. Enas EA, Mohan V, Deepa M, Farooq S,Pazhoor S, Chennikkara H. The metabolic syn-drome and dyslipidemia among Asian Indians: apopulation with high rates of diabetes and pre-mature coronary artery disease. J CardiometabSyndr 2007;2:267–27572. MacFarlane I, Bliss M, Jackson J, Williams G.Diabetes in its historical and social context. InTextbook of Diabetes. 2nd ed. Hoboken, NJ,Blackwell Scientific Publications, 199773. Schwartz SS, Epstein S, Corkey BE, GrantSFA, Gavin JR 3rd, Aguilar RB. The time is rightfor a new classification system for diabetes: ra-tionale and implications of the b-cell-centricclassification schema. Diabetes Care 2016;39:179–18674. McCarthy MI. Genomics, type 2 diabetes,and obesity. N Engl J Med 2010;363:2339–235075. Abdul-Ghani MA, Tripathy D, DeFronzo RA.Contributions of b-cell dysfunction and insulinresistance to the pathogenesis of impaired glu-cose tolerance and impaired fasting glucose. Di-abetes Care 2006;29:1130–113976. Nathan DM, Davidson MB, DeFronzo RA,et al.; American Diabetes Association. Im-paired fasting glucose and impaired glucose

662 Type 2 Diabetes: Winning the Battle but Losing the War? Diabetes Care Volume 39, May 2016

tolerance: implications for care. Diabetes Care2007;30:753–75977. Weber MB, Harish R, Staimez LR, et al. Re-duction in diabetes incidence differs by predia-betes type in a randomized translational trial ofprevention [Abstract]. Diabetes 2015;64(Suppl.1A):LB4678. Saito T, Watanabe M, Nishida J, et al.;Zensharen Study for Prevention of Lifestyle DiseasesGroup. Lifestyle modification and prevention oftype 2 diabetes in overweight Japanese with im-paired fasting glucose levels: a randomizedcontrolled trial. Arch Intern Med 2011;171:1352–136079. Staimez LR, Weber MB, Ranjani H, et al. Evi-dence of reduced b-cell function in Asian In-dians with mild dysglycemia. Diabetes Care2013;36:2772–277880. Mohan V, Amutha A, Ranjani H, et al. Asso-ciations of b-cell function and insulin resistancewith youth-onset type 2 diabetes and prediabe-tes among Asian Indians. Diabetes Technol Ther2013;15:315–32281. Gujral UP, Narayan KM, Kahn SE, KanayaAM. The relative associations of b-cell functionand insulin sensitivity with glycemic status andincident glycemic progression in migrant AsianIndians in the United States: the MASALA study.J Diabetes Complications 2014;28:45–5082. Ikehara S, Tabak AG, Akbaraly TN, et al. Agetrajectories of glycaemic traits in non-diabeticSouth Asian and white individuals: the

Whitehall II cohort study. Diabetologia 2015;58:534–54283. Kanaya AM, Herrington D, Vittinghoff E,et al. Understanding the high prevalence ofdiabetes in U.S. south Asians compared withfour racial/ethnic groups: the MASALA andMESA studies. Diabetes Care 2014;37:1621–162884. Jainandunsing S, Ozcan B, Rietveld T, et al.Failing beta-cell adaptation in South Asian fam-ilies with a high risk of type 2 diabetes. ActaDiabetol 2015;52:11–1985. Lukacs JR, Pal JN. Skeletal variation amongMesolithic people of the Ganga Plains: new ev-idence of habitual activity and adaptation toclimate. Asian Perspect 2003;42:329–35186. Victora CG, de Onis M, Hallal PC, BlossnerM, Shrimpton R. Worldwide timing of growthfaltering: revisiting implications for interven-tions. Pediatrics. 2010;125:e473–48087. Stein AD, Wang M, Martorell R, et al.; Co-horts Group. Growth patterns in early childhoodand final attained stature: data from five birthcohorts from low-andmiddle-income countries.Am J Hum Biol 2010;22:353–35988. HanafiMY, Saleh MM, Saad MI, AbdelkhalekTM, Kamel MA. Transgenerational effects of obe-sity and malnourishment on diabetes risk in F2generation. Mol Cell Biochem 2016;412:269–28089. Hardikar AA, Satoor SN, Karandikar MS,et al. Multigenerational undernutrition in-creases susceptibility to obesity and diabetes

that is not reversed after dietary recuperation.Cell Metab 2015;22:312–31990. Jaacks LM, Ali MK, Bartlett J, et al. Global non-communicable disease research: opportunities andchallenges. Ann Intern Med 2015;163:712–71491. Narayan KM, Ali MK, Koplan JP. Global non-communicable diseases–where worlds meet. NEngl J Med 2010;363:1196–119892. Narayan KM, Ali MK, del Rio C, Koplan JP,Curran J. Global noncommunicable diseases–lessons from the HIV-AIDS experience. N Engl JMed 2011;365:876–87893. Richmond G. Global science engagement.Science 2016;351:42794. Centers for Disease Control and Prevention.Data and statistics, diabetes [Internet]. Avail-able from http://www.cdc.gov/diabetes/data/.Accessed 16 February 201695. Worldmapper. The world as you’ve neverseen it before [Internet]. Available from http://www.worldmapper.org/. Accessed 25 January201696. Pavkov ME, Hanson RL, Knowler WC,Bennett PH, Krakoff J, Nelson RG. Changing pat-terns of type 2 diabetes incidence among PimaIndians. Diabetes Care 2007;30:1758–176397. Staimez LR, Deepa M, Ali MK, Mohan V,Narayan KM, Hanson RL. The tale of two Indians:a comparison of beta-cell function and insulinresistance between Pima Indians and Asian In-dians [Abstract]. Diabetes 2014;63(Suppl. 1):A400

care.diabetesjournals.org Narayan 663