Diabetes in YouthdLookingBackwards to Inform the Future:Kelly West Award Lecture 2017Diabetes Care 2018;41:233–240 | https://doi.org/10.2337/dci17-0031

The Kelly West Award for Outstanding Achievement in Epidemiology is presentedin honor of the memory of Kelly M. West, widely regarded as the “father ofdiabetes epidemiology.” Harry Keen describedWest as characterized by “generosityof spirit, deeply human and humorous, deliberate of address, modest, conciliatoryand untiringly persevering. Few people have done so much to change the landscapeof diabetes” (1). The award and lecture recognize a leading epidemiologist in thefieldof diabetes. Dana Dabelea, MD, PhD, received this award at the American DiabetesAssociation’s 77th Scientific Sessions, 9–13 June 2017, in San Diego, CA. She pre-sented the Kelly West Award Lecture, “Diabetes in YouthdLooking Backwards toInform the Future,” on Sunday, 11 June 2017.

The face of pediatric diabetes has undergone striking changes over several decades.Epidemiology has shown that type1 diabetes incidencehas been increasingworldwide,with wide geographic variation in absolute risk (2). Similarly, more recently, type 2diabetes has been increasing primarily in indigenous youth as never before (2). Inaddition, the obesity epidemic has changed the phenotype of type 1 diabetes, prompt-ing suggestions that some youth have “type 1.5 diabetes,” or even “double diabetes.”Can we identify factors responsible for these changes? We propose that “lookingbackwards” in time and at early stages in the life course of individuals will providenew and useful clues to the etiology and prevention of diabetes.This review focuses on three related themes, and each one has been augmented by

“looking backwards”: 1) surveillance of diabetes in youth, 2) the developmental originsof pediatric obesity and diabetes, and 3) diabetes prevention throughout the lifecourse. Examples of data on diabetes burden in youth, complications, prevalence,and projections are provided. Also discussed is the Developmental Origins of Healthand Disease (DOHaD) paradigm, which posits that various exposures during criticalperiods, such as pregnancy or early life, predispose the fetus and newborn to health ordisease later in life. The integration of basic science in these life-course studiesprovides a uniquewindow into potential biologicmechanisms. The notion of a “viciouscycle of diabetes andobesity” is described, and studies attempting to break the cycle arereviewed. A comprehensive approach to prevention throughout the life course is pro-posed, given data that poor lifestyle habits exist throughout childhood into adulthood.While diabetes can be prevented or delayed in adults, a much earlier primordial pre-vention approach is needed that begins before birth and extends through childhood.

HISTORICAL PERSPECTIVE: “LOOKING BACKWARDS”

Looking backwards in time, at the start of the 20th century, type 1 diabeteswas a rare andrapidly fatal disease. Youthwere thin and usually of white race/ethnicity. The incidence in1900 in Norwaywas 2 per 100,000 per year rising to 7 per 100,000 per year by 1920 (3). In1923, 86% of youth with type 1 diabetes died of diabetic ketoacidosis. However, after

Department of Epidemiology, Colorado School ofPublic Health, and Department of Pediatrics,University of Colorado School of Medicine,Aurora, CO

Corresponding author: Dana Dabelea, dana.dabelea@ucdenver.edu.

© 2018 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. More infor-mation is available at http://www.diabetesjournals.org/content/license.

Dana Dabelea

Diabetes Care Volume 41, February 2018 233

KELLY

WEST

AWARDLEC

TURE

the discovery of insulin, the incidence oftype 1 diabetes began to increase. Overthe past 40 years of the previous century,increases of 2–4% per year, largely acrossEurope, were reported by multiple stud-ies (4,5). In parallel, there was a decreasein the age of onset of type 1 diabetes (3).Type 2 diabetes, not considered a pe-

diatric disease until recently, was oftencalled adult-onset diabetes. Over the pastdecade of the 20th century, there were anumber of reports of obesity-associatedtype 2 diabetes in youth, especially minor-ity youth, in clinic-based studies (6–8)(Fig. 1A). In one of the earliest population-based reports, we showed that prevalencehad increasedover threedecadesbyapprox-imately 80% in Pima Indians (9) (Fig. 1B). Itwas becoming clear that the face of pedi-atric diabetes was changing. However, atthe start of 21st century, there were lim-ited U.S. data on the epidemiology of pe-diatric diabetes, regardless of type.

BURDEN OF TYPE 1 AND TYPE 2DIABETES IN CONTEMPORARY U.S.YOUTH

This motivated the Centers for DiseaseControl and Prevention and the National

Institute of Diabetes and Digestive and Kid-ney Diseases to fund the SEARCH for Di-abetes in Youth (SEARCH) study, coveringall elements of a surveillance system,mon-itoring trends, developing projections ofburden, understanding contemporary clin-icalmanagement, andevaluating the risk ofcomplications. SEARCH has been accom-plishingall thesegoals forover15years (10).

One of the first SEARCH contributionswas providing baseline incidence data oftype 1 and type 2 diabetes in U.S. youthby age and race/ethnicity for year 2002–2003 (Fig. 2) (11). Data showed a strikingrace/ethnicity pattern that goes in oppo-site directions for type 1 versus type 2diabetes, with highest rates of type 1 di-abetes in non-Hispanic whites (NHW), fol-lowed by non-Hispanic blacks (NHB),Hispanics (HISP), and Asian Pacific Is-landers (API), with little type 1 diabetesin American Indians/Alaskan Natives(AIAN). In contrast, the highest rates oftype 2 diabetes are in minority children,AIAN, NHB, HISP, and API, with the lowestrates in NHW youth. Absolute rates oftype 2 diabetes tended to be higher thanrates of type 1 diabetes among most mi-nority groups.

SEARCH recently published the firstcomprehensive assessment of trends inthe incidence rates of type 1 and type 2diabetes in U.S. youth between 2002 and2012 (12) (Fig. 3). Age-, sex-, and race/ethnicity-adjusted type 1 diabetes rates(Fig. 3A) increased on average by 1.8%per year (P 5 0.03). In pairwise compar-isons, the annual rate of increase wasgreater among HISP than among NHW(4.2% vs. 1.2%, P , 0.001). There wasno significant increase among AIAN youth.In contrast, type 2 diabetes rates (Fig. 3B)increased by 4.8% per year on average(P , 0.001), with the largest increasesin AIAN (8.9% per year) and NHB youth(6.3% per year), with no significant in-crease in NHW.

These data allowed us to update theworldwide picture of type 1 diabetes inyouth with contemporary SEARCH infor-mation (Fig. 4). The SEARCH incidencedata are in the middle, following closelypast trends in predominantly white pop-ulations from Allegheny County, PA, andColorado, with northern European andScandinavian countries having higherrates over time (2). Predominantly non-white countries, like China and Japan,although showing increasing rates overtime, have the lowest risk of type 1

diabetes. What does this worldwide pic-ture mean? Genetic differences in popu-lations likely underlie the broad spread inabsolute incidence rates, for example, be-tween Finland and China. However, theincreases have occurred over a shorttime period, making genetic changes inthe populations unlikely. Such changesstrongly suggest the role of environmen-tal factors in the etiology, and many havebeen suggested, including exposures dur-ing pregnancy, type of birth delivery, in-fant diet, gut microbiota, and viruses.

We do know that these increases haveoccurred in parallel with a decrease in thefrequency of high-risk HLA genotypes andan increase in the frequency of moderate/low-risk genotypes in NHW and HISPyouth with type 1 diabetes (13). This sug-gests that, over time, the environmentaltriggers of type 1 diabetes (as yet largely

Figure 1—The changing face of diabetes inyouth. A: Increasing proportion of youth indiabetes clinics with type 2 diabetes between1987 and 1996 (6–8). B: Increasing prevalenceof type 2 diabetes among Pima Indian girlsand boys between two studies conducted be-tween 1967 and 1976 and 1987 and 1996 (9).

Figure 2—A: Incidence rates (per 100,000 peryear) for type 1diabeteswithonset,20 yearsof age in 2002–2003 in SEARCH. Based on1,905 youth with type 1 diabetes in 10millionperson-years at risk. B: Incidence rates (per100,000 per year) for type 2 diabetes withonset ,20 years of age in 2002–2003 inSEARCH (11). Based on 530 youth withtype 2 diabetes in 10 million person-years atrisk. For children aged 0 to 4 years and 5 to9 years, virtually all diabetes was type 1, re-gardless of race/ethnicity, so data are notshown.

234 Kelly West Award Lecture Diabetes Care Volume 41, February 2018

unknown)havebecomemorewidespread,partially alleviating the need for a stronggenetic background as the primary riskfactor.SEARCH data allowed us to explore

trends in type 2 diabetes rates in the con-text of North American youth (Fig. 5).SEARCH rates are shown in solid linesand follow closely prior trends by race/ethnicity reported from Chicago (14)and in First Nation youth from Canada(dotted lines) (15). This indicates thatthe increasing risk of type 2 diabetes inyouth has been happening for severaldecades. Data from the National Healthand Nutrition Examination Survey(NHANES) (16) have shown that the risein type 2 diabetes follows four decades ofgreater overweight and obesity among

U.S. youth, pointing to this environmentalcontribution. Factors such as the increase inmaternal diabetes during pregnancy (17), astrong risk factor for type 2 diabetes in off-spring (9), also have to be considered.

Youth under surveillance by SEARCHclosely approximate the distribution ofthe U.S. Census data for youth aged ,20years in terms of race/ethnicity, age-group, education, and median householdincome (18). This allowed estimation ofthe number of youth with diabetes in2009, using U.S. Census data. Approxi-mately 191,986 youth were identifiedwith physician-diagnoseddiabetes, 166,984with type 1 diabetes, 20,262 with type 2diabetes, and 4,740 with a mixture of“other” types. There were 18,400 youthdiagnosed with type 1 diabetes annually

(incident cases), and 5,100 youth diag-nosed with type 2 diabetes (19).

PROJECTIONS OF FUTURE BURDEN

SEARCH data have allowed diabetes pro-jections into the future.Markovmodelingwas used to estimate the numbers ofyouthwithdiabetes in theyear2050usingtwo possible scenarios, one of constantfuture incidence and one of annually in-creasing incidence (20). Under the secondscenario (more likely given the evidenceof rising rates reported by SEARCH), theprojected prevalence is likely to increasefrom observed levels in 2010 by almostthreefold for type 1 diabetes and fourfoldfor type 2 diabetes by 2050, with a sub-stantially higher burden among minorityyouth for both types (20).

COMPLICATIONS PATTERNS

The burden of diabetes in youth is impor-tant; however, it is perhaps even moreimportant to determine its consequen-ces. Figure 6 identifies a troublesomeconstellation of complications and co-morbidities among youth with both typesof diabetes. For each complication, withthe exception of cardiac autonomic neu-ropathy, the prevalence is significantlyhigher in type 2 versus type 1 diabetes,especially among minority youth, but theprevalence is high in both types.

At an average age of 21 years and aduration of disease of a little less than8 years, one in three youth with type 1 di-abetes and three in four youth with type 2diabetes had at least one such compli-cation or comorbidity (21).

The data summarized above haveshown 1) increasing incidence of bothtype 1 and type 2 diabetes, especiallyamong minorities; 2) changes in the envi-ronment in which children are born andgrow are likely causes; and 3) there is ahigh burden of early diabetes complica-tions, which are higher in youthwith type 2diabetes and in minority youth. These pat-terns suggest that higher costs and greatersocietal burden are very likely in the next20–30years.Clearly,weneedtounderstandthe causes of these changes to understandhow to develop effective prevention ap-proaches. As these changes are occurringin youth and young adults, taking a devel-opmental origins approach is warranted.

DOHaD

Our approach is based on the DOHaD para-digm, which posits that various exposures

Figure 3—A: Age-, sex-, and race/ethnicity-adjusted type 1diabetes incidence rates (per 100,000 peryear) by race/ethnicity and year of diagnosis (12). Over all race/ethnicity groups, type 1 diabetesrates increased 1.8% per year (dashed line). B: Age-, sex-, and race/ethnicity-adjusted type 2 di-abetes incidence rates (per 100,000 per year) by race/ethnicity and year of diagnosis. Over all race/ethnicity groups, type 2 diabetes rates increased 4.8% per year (dashed line). Adapted with per-mission from Mayer-Davis et al. (12).

care.diabetesjournals.org Dabelea 235

(nutritional, chemical, physical, social, envi-ronmental, etc.) during critical periods,such as during pregnancy or early life,predispose the developing organism tohealth or disease later in life. Such criticalperiods continue to exist during postnatallifedearly life, adolescence, puberty, andagain pregnancydthus potentially creat-ing and perpetuating a transgenerationalcycle of health and disease at the popu-lation level.

DIABETES IN PREGNANCY

There is a substantial body of evidence insupport of this paradigm from studies ofdiabetes in pregnancy. Among the Pima

Indians, offspring of women with diabe-tes during pregnancy had a higher preva-lence of type 2 diabetes and were moreobese than those who were not exposedto diabetes during pregnancy (22). Over70% of people with prenatal exposurehad type 2 diabetes by 25–34 years ofage. Fetal exposure to maternal diabetesin utero was the strongest single risk fac-tor for type 2 diabetes in Pima youth,withan odds ratio (OR) of 10.4. Findings fromthe Pima Indian study provided strong ev-idence that using this approach wouldlead to important findings when appliedto both obesity and diabetes. We found asimilarly strong association in a racial and

ethnically diverse groupof youth and con-trol subjects in the SEARCH Case-Control(SEARCH-CC) study, an OR of 7.3 (23).

Not only were these strong associa-tions but they also reflected specific in-trauterine effects. This was elegantlydemonstrated among the Pima Indiansusing a sib-pair design with siblings bornbefore and after their mothers werediagnosed with diabetes (24). Among28 sib-pairs discordant for exposure to di-abetes in utero, 21 of 28 developed di-abetes after their mother was diagnosedwith diabetes during pregnancy and only7 developed diabetes before the motherwas diagnosed. The OR of 3.7 (P 5 0.02)for the association between type 2 diabe-tes in youth and exposure to maternaldiabetes in utero among sib-pairs virtuallycontrols for the genetic predisposition todiabetes transmitted frommother to off-spring. It also controls for postnatalshared familial risk factors and thus iso-lates the specific intrauterine effects, yetunknown, that are responsible for this as-sociation. The fact that this associationwas not present in sib-pairs born beforeor after the father was diagnosed withdiabetes also reduces the concern for achance finding or birth order effects.

Ultimately, this strong association isimportant because of its public health im-pact. Among the Pima Indians, exposureof the fetus to maternal diabetes duringpregnancy was responsible for 35% oftype 2 diabetes in 5- to 19-year-old childrenbetween 1987 and 1996, approximatelytwice the attributable risk found between1967 and 1976. In a more diverse popu-lation enrolled in SEARCH-CC, we foundthat 47.2% of type 2 diabetes with onsetat #20 years old could be attributed tomaternal diabetes, obesity, and theircombination (23). This also means thatit could be prevented were we successfulin eliminating exposure to diabetes andobesity during pregnancy.

BEYOND DIABETES IN PREGNANCY

However, it is not just diabetes in preg-nancy that needs to be controlled but alsoobesity and other exposures during preg-nancy and postnatal life thatmay contrib-ute to long-term risk. Therefore, we havemoved beyond pregnancy diabetes towardamore comprehensive assessment of early-life exposures.

An example of suchwork is the HealthyStart Study (Exploring the Fuel-MediatedProgramming of Neonatal Growth) study

Figure 4—Temporal trends in type 1 diabetes by geographic location and incidence year (2). Thedashed line is contemporary U.S. data from SEARCH (12).

Figure 5—Temporal trends in incidence rates for type 2 diabetes by geographic area and year ofdiagnosis. SEARCH data are shown in solid lines (2).

236 Kelly West Award Lecture Diabetes Care Volume 41, February 2018

in Colorado, a prebirth cohort of 1,410mother-offspring dyads followed fromearly pregnancy through delivery andinto childhood. We are examining multi-ple nutritional, chemical, physical, andsocial exposures during pregnancy andearly life and exploring their biologic sig-natures (through “omics” research) to de-velop the early-life “exposome” and link itto childhood outcomes, including growth,adiposity, cardiometabolism, neurocogni-tion, and behavior. The study is part ofthe large Environmental Influences onChildhood Health Outcomes (ECHO) con-sortium, a National Institutes of Healtheffort to develop a cohort of over 50,000youth to understand the environmen-tal triggers of many chronic childhooddiseases.The Healthy Start Study has already

produced findings about the role of inutero exposureswithmechanistic insightsand clinical relevance. For example, inde-pendent of prepregnancy BMI, increasedgestationalweight gain in all three trimes-ters was associated with increasedneonatal adiposity, measured by air dis-placement plethysmography (an increasein the percentage of neonatal fat mass of0.55 units for each 0.1 kg of gestationalweight gain per week, P , 0.001) (25).Maternal high- versus low-fat diet wasassociated with 0.8 units higher neona-tal fat mass percent, independent ofprepregnancy BMI, energy intake, andexpenditure (26). Increasing levels oflate-pregnancy physical activity were as-sociatedwith decreased neonatal adipos-ity (41.1 g less neonatal fatmass,P5 0.03)

without significantly reduced lean mass(27). Maternal glucose levels, even withinthe normal range, were found tomediate20% of the association betweenmaternalBMI and neonatal adiposity, whereasother fuels that cross the placenta, suchas triglycerides and free fatty acids, werenot related (28,29).

Healthy Start Study investigators alsocollaborate with basic scientists throughthe BabyBump project (Baby Biology ofIntra-Uterine Metabolic Programming),the mechanistic arm of the Healthy StartStudy. We have isolated, grown, andstored mesenchymal stem cells from um-bilical cord tissue samples collected atbirth. These cells differentiate into nu-merous cell lines, adipocytes, myocytes,astrocytes, etc., and various teams areconducting in vitro and in vivo studiesto elucidate the mechanisms responsi-ble for programming of obesity. Wehave found greater adipogenesis in mes-enchymal stem cells from obese versusnormal-weight mothers mediated by theb-catenin system, providing a potentialpathway through which maternal over-weight programs stem cells toward adi-posity (30).

Similarly, collaboration with genomicexperts is identifying epigenetic signa-tures of in utero exposures, such as ges-tational diabetes mellitus (GDM). Inanother cohort study in Colorado (Ex-ploring Perinatal Outcomes among Chil-dren [EPOCH] Study), we conducted anepigenome-wide association study among10-year-old youth who were exposedand not exposed to GDM and identified

98 differentially methylated regions.Given the strength of association, effectsize, network analysis, replication in cordblood in a subsample, and prior literature,we prioritized nine genes for pyrose-quencing. This validated six out of ninegenes, of which some were also associ-ated with obesity-related outcomes in10-year-old children (31).

ROLE OF EARLY POSTNATALNUTRITION

Although prenatal exposures are clearlyimportant, postnatal ones, especially inthe first 2 years of life (the first 1,000 days,roughly from conception to the secondbirthday), are also important. Data fromthe Pima Indian study have shown a strongprotective effect of breastfeeding (vs.bottle-feeding) in the first 3 months of lifeagainst early-onset type 2 diabeteswith anadjusted OR of 0.41 (32). In the EPOCHcohort, we showed that breastfeedingmodified the effect of exposure to diabetesin utero on obesity outcomes in the off-spring (33). Among those breastfed forless than 6 months, BMI was significantlyhigher in those exposed versus not ex-posed to diabetes in utero; however, inthose breastfed for 6 ormoremonths, theeffect of exposurewas attenuated to non-significance. Similar results were seen forwaist circumference and visceral and subcu-taneous adipose tissue, suggesting a poten-tial postnatal intervention in high-riskinfants.

Colleagues at the Barbara Davis Centerfor Childhood Diabetes in Colorado arealso studying infant and early-life diet inrelation to type 1 diabetes risk. Data fromthe Diabetes Autoimmunity Study in theYoung (DAISY) suggest that timing of nu-tritional exposures is relevant for type 1diabetes risk (34). Early (before 4months)and late (after 6 months) introduction ofany solid foods and early and late intro-duction to any cereals all increase the riskof type 1 diabetes from 1.7-fold to over3-fold, while breastfeeding at introduc-tion of cereals reduces risk by about50%. Thus, postnatal nutrition is impor-tant for type 1 diabetes risk as well.

Given the evidence of the importanceof good nutrition and active lifestyles, it isunfortunate that most U.S. youth do notmeet the recommendations for eating2.5–6.5 cups of fruits and vegetableseach day, do not eat theminimumrecom-mended amounts of whole grains (2–3ounces each day), eat more than the

Figure 6—Prevalence of diabetes complications and comorbidities by type of outcome and race/ethnicity, adjusted to 21 years of age (21). Other, NHB, HISP, API, and AIAN combined.

care.diabetesjournals.org Dabelea 237

recommended maximum daily intake ofsodium (1,500–2,300 mg each day), andhave 40% of daily calories coming from“empty calories” (sugars and solid fats).In addition, only 29% percent of highschool students had participated in atleast 60 min per day of physical activityon each of the 7 days before the survey(35). This suggests that a successful ap-proach to prevention of obesity and di-abetes, as well as asthma, heart disease,and selected cancers, requires a compre-hensive life-course approach.

A COMPREHENSIVE APPROACH TOPREVENTION THROUGHOUT THELIFE COURSE

There is now definitive evidence from theDiabetes Prevention Program (DPP) study,which we are proudly part of, that metfor-min and especially lifestyle interventionsreduced diabetes risk in adults by 31–58% (36). These effects were sustainedover 15 years (37), with heart diseaseand stroke risk factors also reduced withfewer medications. In addition, we havenow evidence, informed by surveillancesystems, that the annual rate of diabetesdevelopment in U.S. adults, after a steadyincrease, may be decreasing for the firsttime in 20 years, by about 5.4% per year(38). There are now 1.4million fewer newreported cases in the U.S. each year. Wemay have turned the corner, perhaps as aconsequence of DPP and DPP-like (39)translational programs, though contraryopinions exist about these trends (40).Unfortunately, we are seeing an oppositetrend of similar magnitude in pediatric di-abetes. To us, this means that we need toincrease or shift the focus on preventionearlier in life, working toward “primor-dial” prevention.The notion of a “vicious cycle of diabe-

tes and obesity” was first described byPettitt and Knowler (41) (Fig. 7). Youngwomen who are obese or have diabetesduring pregnancy have a higher risk thattheir offspring will also be obese or de-velop diabetes later in life. As these youthenter young adulthood, obese women orwomen with diabetes transmit this in-creased risk to the next generation. It islikely that this vicious cycle must be bro-ken earlier and at different levels duringthe life course if we are to make greaterstrides toward reduction of diabetes andobesity. There are multiple points alongthe circle where attempts have beenmade, as shown in Fig. 7. Some of these

include 1) controlling diabetes or hyper-glycemia in pregnancy, 2) preventing obe-sity and diabetes in pregnancy, and 3)preventing obesity and diabetes in youth.

One strategy has been to control dia-betes or hyperglycemia in pregnancy.A randomized trial of glucose controlamong women with mild GDM was con-ducted, the Australian Carbohydrate In-tolerance Study in Pregnant Women(ACHOIS) (42). The intervention includeddietary and lifestyle advice, and a usual-care groupwas blinded to the “diagnosis”of GDM, which was not treated unlesssymptomatic. ACHOIS confirmed thatGDM has relatively rare but serious ad-verse perinatal effects and that treatmentis beneficial in avoiding them. However,in an observational follow-up of offspringat ages 4 to 5 years born to mothers inACHOIS, no effect was seen on offspringBMI z-scores between groups (43). Whythese results were negative is uncertain.It may be because GDM was mild in allmothers in both arms of the trial or thatexcess obesity among offspring may onlyoccur at older ages. A trial from the U.S.Maternal-Fetal Medicine Units Networkfound that treatment of mild GDM didnot significantly reduce the frequencyof a composite outcome that includedstillbirth or perinatal death and severalneonatal complications; it did reduce therisksof fetal overgrowth, shoulderdystocia,cesarean delivery, and hypertensive disor-ders (44). Further research examining the

later consequences of GDM is clearly re-quired, given the strength of observa-tional evidence discussed above.

Another related strategy is to preventobesity and diabetes in pregnancy. Anumber of older studies have tried exer-cise (45), diet (46), or a combination (47).There is evidence of a trend to lower ges-tational weight gain in the combined in-terventions, but there was limited effecton GDM and no effects on offspring birthweight orobesitymarkers in the next gen-eration. The Finnish Gestational DiabetesPrevention Study (RADIEL) trial random-ized high-risk pregnant women (priorGDMor obese) using individualized coun-seling on diet, physical activity, andweight control. Control subjects receivedstandard antenatal care. There was a 39%reduction in GDM incidence among inter-vention women compared with usual-carewomen (48). This is the first study to seeGDM effects; however, no differenceswere seen in infant size between groups.There is more to be done here to includeinterventions that are more effective, ascompliance has not always been satisfac-tory, and with longer offspring follow-up.

Finally, another strategy is to preventobesity and diabetes in youth. A Cochranreview included55 studies targeting children6–12 years of age (49) and found thatprograms were effective at reducingadiposity, though by only 20.15 kg/m2

(95%CI20.21 to20.09), with substantialheterogeneity between programs in all

Figure 7—The vicious cycle of transgenerational obesity and diabetes, including potential timeswhen the cycle might be broken through effective interventions.

238 Kelly West Award Lecture Diabetes Care Volume 41, February 2018

age-groups. Several promising policiesand strategies included changes in schoolcurriculum that include healthy eating,physical activity and body image, in-creased sessions for physical activity, im-provements in nutritional quality of theschool food supply, and others showingthat effective programs must be deeplyembedded in school culture.One of our recent contributions in this

area aims to reduce risk factors for type 2diabetes in American Indian youthdtheTribal Turning Point Program (TTPP), a pri-mordial prevention feasibility trial. In thispilot, 60 prepubertal 8- to 10-year-oldyouth and their caregivers were random-ized to a lifestyle intervention or a controlcondition. The lifestyle intervention hadthree components, a group-based DPP-inspired active learning curriculum focusedon behavior change, toolboxmaterials, andindividual-level motivational interviewing.After 8 months of follow-up with excellentretention, all obesity outcomes substan-tially and significantly decreased in the in-tervention compared with control group(50). Given the initial success of the pilot,we are now actively expanding the studyand will conduct a fully powered trialwith implementation and disseminationcomponents.

IMPLICATIONS FOR FUTURERESEARCH

To understand better the burden and riskof youth-onset diabetes, the effects ofprevention programs, the developmentand burden of complications, and latermortality, sustainable surveillance will berequired, such as conducted by SEARCH,perhapswith expansion formore completecoverage of high-risk minorities.To explore comprehensively the role

of the environmental exposome and itsbiologic pathways and to identify poten-tially causal associations, longitudinal(pre)birth cohort studies are crucial. Ex-pansions to consortia, such as The Envi-ronmental Determinants of Diabetes inYoung (TEDDY) and ECHO studies, and ad-dition of ones specifically targeted at obe-sity and diabetes prevention are needed.In order to improve our understanding

of mechanisms, animal studies with rapidtranslation to humans and mechanisticstudies nested in population cohorts arerobust ways to move our understandingforward more quickly.Finally, we must focus on breaking the

vicious cycle of diabetes and obesity at

multiple levels, with an increased focuson primordial prevention, which will re-quire effective interventions, includingrandomized clinical trials and transla-tional studies.

Acknowledgments. The author especiallythanks the mentors and colleagues who havehelped develop critical thinking and encouragedher along the way: Peter H. Bennett, NationalInstitutes of Health, National Institute of Di-abetes and Digestive and Kidney Diseases(NIDDK), Phoenix, AZ; Richard F. Hamman, Col-oradoSchoolofPublicHealth,Aurora,CO;TrevorJ. Orchard, University of Pittsburgh School ofPublic Health, Pittsburgh, PA; and her closecolleague, Elizabeth J. Mayer-Davis, GillingsSchool of Global Public Health, Chapel Hill, NC.Shealso thanks theSEARCHforDiabetes inYouthfaculty and staff; the many students, fellows,colleagueswhohaveworked to developmuchofthe work in the article; and her loving family.Funding. Funding for this work came frommultiple National Institutes of Health and Cen-ters for Disease Control and Prevention (CDC)grants, including CDC U18 DP006139, NIDDKUC4 DK108173, R01DK076648, R01DK068001,R01DK048375, R01DK100340, R34DK096403,and R01DK59184.Duality of Interest. No potential conflicts of in-terest relevant to this article were reported.

References1. Gale EAM. Kelly West (1925-1980) [Internet],13 August 2014. Available from https://doi.org/1014496/dia1104696428.8. Accessed 4 Septem-ber 20172. DabeleaD, HammanRF, KnowlerWC.Diabetesin youth. In Diabetes in America. 3rd. ed. CowieCC, Casagrande SS,Menke A, et al., Eds. Bethesda,MD, National Institutes of Health, 2017, p. 15.1–15.543. Gale EA. The rise of childhood type 1 diabetesin the 20th century. Diabetes 2002;51:3353–33614. Patterson C, Guariguata L, Dahlquist G, SolteszG, Ogle G, Silink M. Diabetes in the young - aglobal view and worldwide estimates of numbersof childrenwith type 1 diabetes. Diabetes Res ClinPract 2014;103:161–1755. Kostraba JN, Gay EC, Cai Y, et al. Incidence ofinsulin-dependent diabetes mellitus in Colorado.Epidemiology 1992;3:232–2386. Pinhas-Hamiel O, Dolan LM, Daniels SR,Standiford D, Khoury PR, Zeitler P. Increased in-cidenceof non-insulin-dependentdiabetesmellitusamong adolescents. J Pediatr 1996;128:608–6157. Scott CR, Smith JM, Cradock MM, Pihoker C.Characteristics of youth-onset noninsulin-dependentdiabetes mellitus and insulin-dependent diabetesmellitus at diagnosis. Pediatrics 1997;100:84–918. Hale DE, Danney KM. Non-insulin dependentdiabetes in Hispanic youth (type 2Y) [Abstract].Diabetes 1998;47(Suppl. 1):A829. Dabelea D, Hanson RL, Bennett PH, Roumain J,Knowler WC, Pettitt DJ. Increasing prevalence oftype II diabetes in American Indian children. Dia-betologia 1998;41:904–91010. Hamman RF, Bell RA, Dabelea D, et al.;SEARCH for Diabetes in Youth Study Group. The

SEARCH for Diabetes in Youth study: rationale,findings, and future directions. Diabetes Care2014;37:3336–334411. Dabelea D, Bell RA, D’Agostino RB Jr, et al.;Writing Group for the SEARCH for Diabetes inYouth StudyGroup. Incidence of diabetes in youthin the United States. JAMA 2007;297:2716–272412. Mayer-Davis EJ, Lawrence JM, Dabelea D,et al.; SEARCH for Diabetes in Youth Study. Inci-dence trends of type 1 and type 2 diabetes amongyouths, 2002-2012. N Engl J Med 2017;376:1419–142913. Vehik K, Hamman RF, Lezotte D, et al. Trendsin high-risk HLA susceptibility genes among Colo-rado youth with type 1 diabetes. Diabetes Care2008;31:1392–139614. Smith TL, Drum ML, Lipton RB. Incidence ofchildhood type I and non-type 1 diabetes mellitusin a diverse population: the Chicago ChildhoodDiabetes Registry, 1994 to 2003. J Pediatr Endo-crinol Metab 2007;20:1093–110715. Dean HJ, Sellers EA, Young K. Type 2 diabetesin youth inManitoba, Canada, 1986 to 2002. Can JDiabetes 2003;27:449–45416. Ogden CL, Carroll MD, Lawman HG, et al.Trends in obesity prevalence among childrenand adolescents in the United States, 1988-1994through 2013-2014. JAMA 2016;315:2292–229917. Dabelea D, Snell-Bergeon JK, Hartsfield CL,Bischoff KJ, Hamman RF, McDuffie RS; KaiserPermanente of Colorado GDM Screening Pro-gram. Increasing prevalence of gestational diabe-tes mellitus (GDM) over time and by birth cohort:Kaiser Permanente of Colorado GDM ScreeningProgram. Diabetes Care 2005;28:579–58418. Dabelea D, Mayer-Davis EJ, Saydah S, et al.;SEARCH for Diabetes in Youth Study. Prevalence oftype 1 and type 2 diabetes among children andadolescents from 2001 to 2009. JAMA 2014;311:1778–178619. Pettitt DJ, Talton J, Dabelea D, et al.; SEARCHfor Diabetes in Youth Study Group. Prevalence ofdiabetes mellitus in U.S. youth in 2009: TheSEARCH for Diabetes in Youth Study. DiabetesCare 2014;37:402–40820. Imperatore G, Boyle JP, Thompson TJ, et al.;SEARCH for Diabetes in Youth Study Group. Projec-tionsof type1and type2diabetesburden in theU.S.populationaged,20 years through2050: dynamicmodeling of incidence, mortality, and populationgrowth. Diabetes Care 2012;35:2515–252021. DabeleaD, Stafford JM,Mayer-Davis EJ, et al.;SEARCH for Diabetes in Youth Research Group.Association of type 1 diabetes vs type 2 diabetesdiagnosedduring childhood and adolescencewithcomplications during teenage years and youngadulthood. JAMA 2017;317:825–83522. Dabelea D, Knowler WC, Pettitt DJ. Effect ofdiabetes in pregnancy on offspring: follow-up re-search in the Pima Indians. J Matern Fetal Med2000;9:83–8823. Dabelea D, Mayer-Davis EJ, Lamichhane AP,et al. Association of intrauterine exposure to ma-ternal diabetes andobesitywith type 2diabetes inyouth: the SEARCH Case-Control Study. DiabetesCare 2008;31:1422–142624. Dabelea D, Hanson RL, Lindsay RS, et al. In-trauterine exposure to diabetes conveys risks fortype 2 diabetes and obesity: a study of discordantsibships. Diabetes 2000;49:2208–221125. Starling AP, Brinton JT, Glueck DH, et al. Asso-ciations of maternal BMI and gestational weight

care.diabetesjournals.org Dabelea 239

gain with neonatal adiposity in the Healthy StartStudy. Am J Clin Nutr 2015;101:302–30926. Shapiro AL, Kaar JL, Crume TL, et al. Mater-nal diet quality in pregnancy and neonatal adipos-ity: the Healthy Start Study. Int J Obes 2016;40:1056–106227. Harrod CS, Chasan-Taber L, Reynolds RM,et al. Physical activity in pregnancy and neonatalbody composition: the Healthy Start Study. Ob-stet Gynecol 2014;124:257–26428. Crume TL, Shapiro AL, Brinton JT, et al. Ma-ternal fuels and metabolic measures during preg-nancy and neonatal body composition: the HealthyStart Study. J Clin Endocrinol Metab 2015;100:1672–168029. Shapiro ALB, Schmiege SJ, Brinton JT, et al.Testing the fuel-mediated hypothesis: maternalinsulin resistance and glucose mediate the asso-ciation betweenmaternal and neonatal adiposity,the Healthy Start Study. Diabetologia 2015;58:937–94130. Boyle KE, Patinkin ZW, Shapiro AL, Baker PR2nd, Dabelea D, Friedman JE. Mesenchymal stemcells from infants born to obese mothers exhibitgreater potential for adipogenesis: the HealthyStart BabyBUMP Project. Diabetes 2016;65:647–65931. Zhang W, Yang IV, Kechris KJ, Fingerlin TE,Dabelea D. Epigenetic marks of in utero exposureto gestational diabetes (GDM) and childhood ad-iposity outcomes: the EPOCH Study [Abstract]. Di-abetes 2016;65(Suppl. 1):A42432. Pettitt DJ, Forman MR, Hanson RL, KnowlerWC, Bennett PH. Breastfeeding and incidence ofnon-insulin-dependent diabetes mellitus in PimaIndians. Lancet 1997;350:166–16833. Crume TL, Ogden L, Maligie M, et al. Long-term impact of neonatal breastfeeding on child-hood adiposity and fat distribution among

children exposed to diabetes in utero. DiabetesCare 2011;34:641–64534. Frederiksen B, Kroehl M, LambMM, et al. In-fant exposures and development of type 1 diabe-tes mellitus: the Diabetes Autoimmunity Study inthe Young (DAISY). JAMA Pediatr 2013;167:808–81535. Centers for Disease Control and Prevention.Childhood nutrition facts [article online], 2017. Avail-able from https://www.cdc.gov/healthyschools/nutrition/facts.htm. Accessed 9 August 201736. 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;346:393–40337. Diabetes Prevention Program ResearchGroup. Long-term effects of lifestyle interventionor metformin on diabetes development and mi-crovascular complications over 15-year follow-up:the Diabetes Prevention Program OutcomesStudy. Lancet Diabetes Endocrinol 2015;3:866–87538. Geiss LS, Wang J, Cheng YJ, et al. Prevalenceand incidence trends for diagnosed diabetesamong adults aged 20 to 79 years, United States,1980-2012. JAMA 2014;312:1218–122639. Ely EK, Gruss SM, Luman ET, et al. A nationaleffort toprevent type2 diabetes: participant-levelevaluation of CDC’s National Diabetes PreventionProgram. Diabetes Care 2017;40:1331–134140. Selvin E, Ali MK. Declines in the incidence ofdiabetes in the U.S.-real progress or artifact? Di-abetes Care 2017;40:1139–114341. Pettitt DJ, Knowler WC. Diabetes and obesityin the Pima Indians: a cross-generational viciouscycle. J Obesity Weight Regul 1988;7:6542. Crowther CA, Hiller JE, Moss JR, McPhee AJ,Jeffries WS, Robinson JS; Australian Carbohydrate

Intolerance Study in Pregnant Women (ACHOIS)Trial Group. Effect of treatment of gestational di-abetes mellitus on pregnancy outcomes. N Engl JMed 2005;352:2477–248643. GillmanMW, Oakey H, Baghurst PA, VolkmerRE, Robinson JS, Crowther CA. Effect of treatmentof gestational diabetes mellitus on obesity in thenext generation. Diabetes Care 2010;33:964–96844. Landon MB, Spong CY, Thom E, et al.; EuniceKennedy Shriver National Institute of Child Healthand Human Development Maternal-Fetal Medi-cine Units Network. A multicenter, randomizedtrial of treatment for mild gestational diabetes.N Engl J Med 2009;361:1339–134845. Han S,Middleton P, Crowther CA. Exercise forpregnant women for preventing gestational dia-betesmellitus. Cochrane Database Syst Rev 2012;7:CD00902146. Tieu J, Shepherd E, Middleton P, CrowtherCA. Dietary advice interventions in pregnancyfor preventing gestational diabetes mellitus. Co-chrane Database Syst Rev 2017;1:CD00667447. O’Brien CM, Grivell RM, Dodd JM. Systematicreview of antenatal dietary and lifestyle interven-tions in women with a normal body mass index.Acta Obstet Gynecol Scand 2016;95:259–26948. Koivusalo SB, Rono K, Klemetti MM, et al.Gestational diabetes mellitus can be preventedby lifestyle intervention: the Finnish GestationalDiabetes Prevention Study (RADIEL): a random-ized controlled trial. Diabetes Care 2016;39:24–3049. Waters E, de Silva-Sanigorski A, Hall BJ, et al.Interventions for preventing obesity in children.Cochrane Database Syst Rev 2011;12:CD00187150. Sauder KA, DabeleaD, Bailey-CallahanR, et al.Targeting risk factors for type 2 diabetes in Amer-ican Indian youth: the Tribal Turning Point pilotstudy. Pediatr Obes. 21 June 2017 [Epub ahead ofprint]. https://doi.org/10.1111/ijpo.12223

240 Kelly West Award Lecture Diabetes Care Volume 41, February 2018