Page 1 of 58 Diabetes

Interaction between type 2 diabetes prevention strategies and genetic determinants of

coronary artery disease on cardiometabolic risk factors

Jordi Merino1,2,3,4,5, Kathleen A Jablonski6, Josep M. Mercader1,2,3, Steven E Kahn7, Ling

Chen1,2,3, Maegan Harden8, Linda M Delahanty2,9, Maria Rosario G. Araneta10, Geoffrey A

Walford1,2, Suzanne B.R. Jacobs1,2,3, Uzoma N. Ibebuogu11, Paul W. Franks12,13,14, William C.

Knowler15, Jose C. Florez1,2,3,4; for the Diabetes Prevention Program Research Group.

1 Center for Genomic Medicine, Massachusetts General Hospital, Boston, MA, USA

2 Diabetes Unit, Massachusetts General Hospital, Boston, MA, USA

3 Programs in Metabolism and Medical & Population Genetics, Broad Institute of MIT and

Harvard, Cambridge, MA, USA

4 Department of Medicine, Harvard Medical School, Boston, MA, USA

5 Research Unit on Lipids and Atherosclerosis, Rovira i Virgili University, IISPV,

CIBERDEM, Reus, Spain

6 Department of Epidemiology and Biostatistics, The Biostatistics Center, Milken Institute

School of Public Health, George Washington University, Rockville, MD, USA

7 Division of Metabolism, Endocrinology & Nutrition, VA Puget Sound Health Care System and

University of Washington, Seattle, WA

8 Genomics Platform, Broad Institute, Cambridge, MA, USA

9 Department of Family Medicine and Public Health, University of California, San Diego, La

Jolla, CA, USA

1

Diabetes Publish Ahead of Print, published online October 21, 2019 Diabetes Page 2 of 58

10 Division of Cardiovascular Diseases, Department of Medicine, University of Tennessee Health

Science Center, Memphis, TN, USA

11 Genetic & Molecular Epidemiology Unit, Lund University Diabetes Centre. Malmo, Sweden

12 Department of Public Health and Clinical Medicine, Umeå University, Umeå, Sweden

13 Department of Nutrition, Harvard TH Chan School of Public Health, Boston, MA, USA

14 Diabetes Epidemiology and Clinical Research Section, National Institute of Diabetes and

Digestive and Kidney Diseases, Phoenix, AZ, USA

Short running title: Genetics and preventive strategies in DPP

Word count: 3,417 words

Number of tables and figures: 5 tables

Corresponding author’s contact information:

Jose C. Florez, MD, PhD

Diabetes Prevention Program Coordinating Center

The Biostatistics Center George Washington University

6110 Executive Blvd, Suite 750

Rockville, MD 20852. USA

Phone: 617 643 3308

Email: [email protected] and [email protected]

2 Page 3 of 58 Diabetes

ABSTRACT

Coronary artery disease (CAD) is more frequent among individuals with dysglycemia.

Preventive interventions for diabetes can improve cardiometabolic risk factors (CRFs), but it is

unclear whether the benefits on CRFs are similar for individuals at different genetic risk for

CAD. We built a 201-variant polygenic risk score (PRS) for CAD and tested for interaction with

diabetes prevention strategies on one-year changes in CRFs in 2,658 Diabetes Prevention

Program participants. We also examined whether separate lifestyle behaviors interact with PRS

on changes in CRFs in each intervention group. Participants in both the lifestyle and metformin

interventions had greater improvement in the majority of recognized CRFs compared to placebo

(P<0.001) irrespective of CAD genetic risk (Pint>0.05). We detected nominal significant

interactions between PRS and dietary quality and physical activity on one-year change in body

mass index, fasting glucose, triglycerides, and HDLc in individuals randomized to metformin or

placebo, but none of them achieved the multiple-testing correction for significance. This study

confirms that diabetes preventive interventions improve CRFs regardless of CAD genetic risk,

and delivers hypothesis-generating data on the varying benefit of increasing physical activity and

improving diet on intermediate cardiovascular risk factors depending on individual CAD genetic

risk profile.

3 Diabetes Page 4 of 58

INTRODUCTION

The risk of coronary artery disease (CAD), the leading cause of disability and mortality worldwide (1), is increased by known cardiometabolic risk factors (CFRs), such as obesity, high blood pressure, impaired lipid and glucose metabolism, and systemic inflammation (2,3). These metabolic features are also present in many individuals with type 2 diabetes, which may contribute to the approximate doubling of CAD risk in persons with diabetes (4,5). A number of studies have demonstrated the effectiveness of control of CRFs in reducing the risk of cardiovascular outcomes among patients with type 2 diabetes (6–10).

Individual risk of CAD and type 2 diabetes reflects the interplay between lifestyle behaviors acting on a backdrop of genetic predisposition (11,12). Previous studies have shown that preventive interventions for type 2 diabetes —including lifestyle intervention programs, increasing physical activity, dietary modifications, and the administration of metformin—can improve CRFs among individuals with dysglycemia (13–15). However, it is unclear whether the benefits of type 2 diabetes preventive interventions on CRFs are similar for individuals at lower or higher genetic risk for CAD.

In the present study, we leveraged data from the Diabetes Prevention Program (DPP) to examine whether type 2 diabetes prevention strategies, either an intensive lifestyle intervention (ILS) or metformin treatment (MET), modify the association between CAD genetic risk and CRFs in participants at high risk of type 2 diabetes. In addition, we investigated the extent to which separate lifestyle behaviors including physical activity, dietary quality, and body weight loss interact with CAD genetic risk on CRFs in each DPP intervention group.

4 Page 5 of 58 Diabetes

METHODS

The Diabetes Prevention Program

The DPP Research Group conducted a multicenter randomized controlled trial in the United

States that tested the effects of ILS and MET interventions on the incidence of diabetes in

glucose-intolerant individuals as described in detail elsewhere (16,17). In brief, a total of 3,234

participants with fasting plasma glucose levels between 5.3 and 6.9 mmol/L, and 2-h plasma

glucose levels between 7.8 and 11.0 mmol/L during a standard 75-g oral glucose tolerance test

were randomized to ILS (n=1,079), MET (850 mg twice daily, n=1,073), or placebo treatment

(PBO, n=1,082). The ILS arm included individual counseling sessions through which

participants were encouraged to achieve and maintain a weight reduction of at least 7% of initial

body weight through a healthy low-calorie, low-fat diet and to engage in physical activity of

moderate intensity, such as brisk walking, for at least 150 minutes per week. A total of 2,658

participants with available DNA, detailed lifestyle information, and CRFs measurements had

paired information at both baseline and one-year follow-up. Each clinical center and the

coordinating center obtained institutional review board approval. The 2,658 included in this

report provided written informed consent for the main study and subsequent genetic

investigations.

Lifestyle behaviors

Specific lifestyle behaviors including changes in physical activity, dietary quality, and weight

loss were assessed at baseline and one year. Self-reported levels of leisure-time physical activity

were assessed at baseline and after one year of follow up with the Modifiable Activity

5 Diabetes Page 6 of 58

Questionnaire (18). The physical activity level was calculated as the product of the duration and frequency of each activity (in hours per week), weighted by an estimate of the metabolic equivalent of that activity (METs) and summed for all activities performed. Usual daily caloric intake during the previous year, including calories from fat, carbohydrate, protein, and other nutrients, was assessed with a modified version of the Block food-frequency questionnaire (19).

We further characterized overall dietary quality at baseline and after one year of follow up using the Alternate Healthy Eating Index 2010 (AHEI-2010) (20). The AHEI-2010 score is based on

11 foods and nutrients emphasizing higher intake of vegetables (excluding potatoes), fruits, whole grains, nuts and legumes, long chain omega-3 fats, and polyunsaturated fatty acids; moderate intake of alcohol; and lower intake of sugar-sweetened drinks and fruit juice, red and processed meats, trans fat, and sodium. Each component is scored from 0 (unhealthiest) to 10

(healthiest) points, with intermediate values scored proportionally. All component scores were summed to obtain a total score ranging from 0 (non-adherence) to 110 (best adherence) points.

Body weight change was defined by the difference between baseline and one-year follow-up.

Baseline and one-year CRF measurements

We considered the following well established risk factors for coronary artery disease at baseline and one-year follow-up: body mass index (BMI), waist circumference (WC), fasting glucose, low density lipoprotein cholesterol (LDLc) high density lipoprotein cholesterol (HDLc), triglycerides (TG), systolic and diastolic blood pressure (SBP, DBP), C-reactive protein (CRP), fibrinogen, and tissue plasminogen activator (tPA). Measurements were performed at baseline and at one-year follow-up (95% of participants completed the one-year follow-up). We also included diabetes incidence as an intermediate CAD risk factor. Diabetes incidence was

6 Page 7 of 58 Diabetes

ascertained at the end of study follow-up. Anthropometric measures included height, weight,

waist circumference and SBP and DBP using standardized methods. Participants fasted for 12

hours the night before blood was drawn from an antecubital vein. Standard blood glucose and

lipid measurements (TG, HDLc, calculated LDLc) were performed at the DPP Central

Biochemistry Laboratory (Northwest Lipid Research Laboratories, University of Washington,

Seattle) using enzymatic methods standardized to the Centers for Disease Control and Prevention

reference methods (21). Measurements of inflammatory markers, including CRP, fibrinogen, and

tPA, were also performed at the DPP Central Biochemistry Laboratory as previously reported

(13,22).

Genotyping and CAD polygenic risk score

We extracted DNA from peripheral blood leukocytes. Genotyping was done with the Human

Core Exome genotyping array from Illumina at the Genomics Platform at the Broad Institute.

Genotypes were called using Birdsuite (https://www.broadinstitute.org/birdsuite/birdsuite-

analysis). A two-stage imputation procedure consisting in pre-phasing the genotypes into whole

chromosome haplotypes followed by imputation itself was conducted. The pre-phasing was

performed using SHAPEIT2 (23). We used 1000 Genomes Phase3 haplotypes as reference panel

(24), and the genotype imputation was done using IMPUTE2 (25). We derived a polygenic risk

score of 204 variants representative of all the 160 CAD loci that had achieved genome-wide

significance for association with CAD in previous association studies published as of December

2017 (26), and used recently to predict the risk of major CAD events among participants with

type 2 diabetes at high cardiovascular risk (27) (Supplementary Table 1). For loci with multiple

independent variants, the variant with the highest significant association with CAD reported in

7 Diabetes Page 8 of 58

literature (lead variant) was selected first, followed by any other variant at that locus

(independent variant) that was not in linkage disequilibrium (r2<0.2) with the lead variant. Three of the 204 CAD risk-increasing variants (rs7797644, rs9365196, and rs9457995) were not available in the DPP GWAS data set (neither were proxies in linkage disequilibrium (r2>0.8)). A total of 138 of the 201 CAD risk-increasing variants considered in the study were genotypes and the remaining 63 imputed at high quality (median info score 0.99, interquartile range 0.97-1).

To calculate the CAD PRS, each variant was recoded as 0, 1, or 2 according to the number of risk alleles (CAD increasing alleles) and weighted by its relative effect size (β coefficient) on

CAD obtained by the literature-based estimates. We calculated the CAD PRS by using the equation: PRS = (β1×SNP1+β2×SNP2+…+β201×SNP201) × (201/sum of the β coefficients), where

SNPi is the risk allele number of each SNP. The CAD PRS ranges from 0 to 402, with each unit corresponding to one average risk allele and higher scores indicating a higher genetic predisposition to CAD. For participants with missing genetic variants, we adjusted the CAD PRS by the number of missing values. Distribution of the weighted CAD PRS fall into the normal distribution curve (Supplementary Fig. 1).

Statistical analysis

Baseline characteristics that are continuous variables are reported as mean ± standard deviation if normally distributed, or as median with 25th and 75th percentiles if not. Categorical variables are presented as frequency. We used generalized linear models to estimate the association of the

CAD polygenic risk score with CRFs at baseline after adjusting for age at randomization, sex, and the top 10 principal components (PCs) for ancestry. Non-normally distributed outcomes were log transformed and presented on the ratio scale (Exp(β)) as the estimated value of CRFs

8 Page 9 of 58 Diabetes

per each 10-units increase in CAD PRS. In this case, the estimated effect size corresponds with a

fractional difference in CRFs. For example, a ratio of 0.9 indicates that the outcome variable

changes by a ratio of 0.9 (i.e., 10% lower) per 10 unit higher PRS. Associations were also tested

for changes from baseline to one-year with interaction terms for intervention arms (ILS or MET

vs. PBO), after adjustment for age at randomization, sex, PCs, and the respective baseline CRFs.

We used Cox proportional hazard models adjusted for the same confounders to investigate the

association between CAD PRS and diabetes incidence and the extent to which type 2 diabetes

preventive interventions modified the association between CAD PRS and diabetes incidence. We

also tested associations of changes in physical activity, dietary quality, and body weight with

one-year change in CRFs in each treatment group separately. We investigated potential

interactions between lifestyle behaviors and PRS on one-year change in CRFs in each treatment

group separately. For statistical significant interactions (P<0.05), we tested the associations

between each 1SD increase in lifestyle variables and one-year change in CRFs among individual

classified as low, intermediate, and high genetic risk on the basis of thirds of the CAD PRS. For

each category, we used general linear models after adjustment for age at randomization, sex, the

top 10 PCs for ancestry, and the respective baseline CRFs. To reject the null hypothesis that type

2 diabetes prevention strategies did not modify the association between CAD genetic risk and

CRFs, a two-sided α-level of 0.05 was used to determine statistical significance. The Statistical

Analysis Software (SAS) version 9.3 was used for all analyses (SAS Institute, Inc., Cary, NC).

9 Diabetes Page 10 of 58

RESULTS

To investigate whether type 2 diabetes prevention strategies modify the association between

CAD genetic risk and CRFs, we used genetic and clinical data collected from 2,658 participants in the DPP. Participants randomly assigned to PBO, ILS, or MET intervention groups displayed no significant differences in baseline characteristics—exception for a lower HDLc and higher

TG in the PBO individuals compared to individuals assigned to MET or ILS (Table 1). There were also no major clinical differences between participants included in this study and all DPP participants (Supplementary Table 2).

We first assessed the association between CAD PRS and CRFs before intervention in all three treatment groups combined. At baseline, each 10-risk allele increase in the CAD PRS was associated with higher LDLc (β=0.09 mmol/L [95%CI 0.06; 0.13] P<0.01), and higher DBP

(β=0.52 mmHg [95%CI 0.09; 0.95] P=0.02) after adjustment for age at randomization, sex, and

PCs ancestry markers (Table 2). Adjusted-mean values of baseline lipid levels and DBP across

CAD PRS quartiles are detailed in Supplementary Fig. 2. No additional associations were found between CAD PRS and other baseline CRFs, including glycemic traits, anthropometric measures and inflammation markers (Table 2).

We next investigated the effect of type 2 diabetes preventative strategies, CAD PRS, and the interaction between them on one-year change in CRFs. Participants randomized to ILS had greater improvement in all studied CRFs compared to those in the placebo group (P<0.01 for all)

(Table 3). Individuals randomized to MET displayed a significant improvement in glycemia, anthropometric measures, LDLc, HDLc, CRP, and tPA compared to PBO (Table 3). The PRS did not significantly predict one-year changes in CRFs when we analyzed the entire study population together or in any of the study arms (P>0.05 for all) (Table 4). The interaction

10 Page 11 of 58 Diabetes

between CAD PRS and type 2 diabetes preventive strategies on CRFs outcomes was not

significant (Pint>0.05 for all; Table 5).

We further evaluated associations between changes in physical activity, diet, and body weight

and one-year change in CRFs in each treatment group, and the extent to which lifestyle behaviors

interacted with CAD PRS on one-year change in CRFs. The largest changes in physical activity,

diet, and body weight were observed in ILS compared to PBO (P<0.01 for all) (Supplementary

Table 3). We showed that changes in body weight improved the majority of CRFs parameters

irrespectively of the intervention arm (P<0.01 for all except fibrinogen, supplementary table 4),

and that changes in physical activity and dietary score associated with improved

anthropometrics, blood lipids, and blood pressure measures among participants in the lifestyle

intervention arm (supplementary tables 5 and 6). We detected nominal significant interactions

between PRS and healthy diet and physical activity on one-year change in BMI, fasting glucose,

triglycerides, and HDLc in individuals randomized to metformin or placebo, but none of them

achieved the multiple-testing correction for significance (Supplementary Table 7). Among these

hypothesis-generating interactions, we highlight the association between increasing dietary

quality and one-year changes in BMI among individuals randomized to metformin, which was

more prominent in participants at high genetic risk (Pint=0.01). Changes in BMI per 1 SD

increase in diet quality score were -0.19 (SE 0.10), -0.29 (SE 0.10), and -0.52 (SE 0.11) among

participants at low, intermediate, and high genetic risk, respectively (Supplementary Table 7).

11 Diabetes Page 12 of 58

DISCUSSION

Our findings in the DPP provide evidence on the interplay between genetic factors and type 2 diabetes preventive strategies on intermediate cardiometabolic risk factors. We show that either an intensive lifestyle intervention or metformin has a beneficial effect on one-year change in different CRFs. While a genetic risk score (comprised of 201 variants associated with CAD) does not appear to alter the effectiveness of either intervention, we showed preliminary evidence that increasing physical activity and adhering to a healthy dietary pattern may have a more prominent effect on BMI, fasting glucose and triglycerides in people at high genetic risk that were not assigned to an intensive lifestyle intervention. However, these findings warrant further replication in appropriate randomized clinal studies specially designed to investigate such effects. Taken together, our data suggest that independent of genetic risk, interventions designed to prevent the development of type 2 diabetes in individuals with elevated fasting glucose, impaired glucose tolerance, and overweight/obesity can improve the majority of recognized cardiovascular risk factors, and that among individuals not randomized to an intensive lifestyle intervention, the benefit of increasing physical activity and improving diet may vary depending on individual CAD genetic risk profile.

There are three important findings. First, a polygenic risk score for CAD is associated with baseline lipids levels and diastolic blood pressure, but was not associated with other CAD risk factors such as glycemia or inflammatory markers and did not predict one-year change in these

CRFs during preventive interventions for type 2 diabetes. Our findings that CAD PRS has a strong association with lipid levels is well-aligned with previous studies (28–31), but the positive association with other intermediate risk factors such as diastolic blood pressure or the lack of association with inflammation parameters has not been previously reported as far as we know.

12 Page 13 of 58 Diabetes

While our results need to be interpreted with caution in the context of multiple hypothesis

testing, the most recent data on the genetic overlap between CAD and CRFs suggested that 5-

10% of CAD loci relate to blood pressure (32). The potential mechanism linking overlapping

loci is likely to be via vascular tone regulation and platelet aggregation (28,33), features that

have been predominantly linked with systolic rather than diastolic blood pressure and have

inflammation as a common underlying factor (2). Our findings, in individuals at high risk of

developing type 2 diabetes, where hypertension is a key feature of the metabolic abnormalities

present in individuals with type 2 diabetes, highlight the role of diastolic blood pressure in the

complex overlap between CRFs and the polygenic architecture of CAD.

Second, the type 2 diabetes-preventive intervention strategies we evaluated, metformin treatment

and intensive lifestyle modification, did not interact with CAD genetic risk on one-year change

in CRFs. In other words, participants were likely to benefit similarly from these interventions

despite their genetic susceptibility for CAD. These findings need to be interpreted with caution

since it is possible that highly penetrant single genetic variants of strong effects could interact

with type 2 diabetes prevention strategies on specific CRFs. The rationale to use a combined

PRS in this study relies on the observation that, similar to T2D, for the vast majority of

individuals with CAD, genetic susceptibility results from the cumulative effects of numerous

variants with modest effects disrupting multiple pathways at the same time (34). Our results are

consistent with recent prospective observational studies that have reported that both lifestyle

behaviors and genetic predisposition drive CAD risk without evidence of significant interactions

(35). In addition, the DPP showed that an intensive lifestyle modification is effective for the

prevention of type 2 diabetes regardless of genetic risk based on 34 type-2 diabetes associated

loci (36). Data from Look AHEAD, a long-term randomized clinical trial investigating whether

13 Diabetes Page 14 of 58

an intensive lifestyle intervention for weight loss would decrease cardiovascular morbidity and mortality among individuals with type 2 diabetes, showed that a behavioral weight loss treatment did not alter the association between a CAD genetic risk and CAD (37). Findings from the present study in individuals at high risk of type 2 diabetes support the beneficial effect of early type 2 diabetes prevention strategies regardless of CAD genetic susceptibility.

Third, while our results need to be interpreted in the context of multiple hypothesis testing and lack of replication in independent randomized clinical studies (mainly due to the unavailability of similar resources), we found suggestive evidence that improving dietary quality and increasing physical activity may have a more favorable effect on one-year changes in BMI, fasting glucose, and triglycerides in people at high genetic risk for CAD than in those at low genetic risk. Thus, while an intensive lifestyle intervention aimed to achieve and maintain a weight reduction of at least 7% of initial body weight through diet and physical activity can effectively reduce intermediate cardiometabolic risk factors regardless of genetic risk, the detection of an interaction of CAD genetic risk with dietary quality and physical activity in participants not assigned to the lifestyle intervention suggests a potential additional benefit in those who are at increased genetic risk for CAD. In these individuals, if taking nothing or only metformin for diabetes prevention, the adoption of at least specific lifestyle behaviors such as improving diet or increasing physical activity (even if short of an integral lifestyle intervention) may be particularly beneficial. In other words, a comprehensive lifestyle intervention that achieves 7% weight loss is effective across the entire gradient of CAD genetic risk, whereas other combinations (such as metformin with healthy lifestyles that have a less dramatic effect on weight loss) benefits most those at highest CAD risk. This hypothesis is supported by the observation that we did not see significant interactions between changes in body weight, as a

14 Page 15 of 58 Diabetes

measure of the overall lifestyle modification, and genetic risk for CAD on specific CRFs.

However, further studies are needed to confirm these initial findings.

Several limitations of our study are worth noting. First, the findings are based on a single

randomized clinical trial. We were not able to conduct a replication study due to the lack of

available genetic information in other similar clinical intervention studies that included

individuals at high risk of developing type 2 diabetes (38,39). Second, we did not directly

investigate the association between the CAD PRS and CAD events due to the unavailability of

data in the present study; instead, we used CRFs, which provide early insights on the

atherosclerosis disease process, as a proxy for CAD events. Third, the significant interactions we

observed may be chance observations due to multiple testing or be affected by risk magnification

(i.e., because participants at low risk of a clinical outcome cannot have large absolute risk

reduction, the risk difference is magnified by having a higher baseline risk). Finally, while the

CAD genetic variants included in our PRS were selected on the basis of the most novel

discoveries of genetic variants for CAD risk, we cannot rule out the possibility that a PRS

constructed from a different set of as-yet-undiscovered CAD risk variants will influence response

to type 2 diabetes preventive interventions.

In conclusion, our findings in individuals at high risk of type 2 diabetes provide evidence for the

beneficial effects of type 2 diabetes preventive strategies on cardiometabolic risk factors

regardless of CAD genetic risk profile. Additionally, the effect modification by improving

dietary quality and increasing physical activity and diet on the association of CAD genetic risk

with cardiovascular risk factors among individuals randomized to metformin or placebo

illustrates how early CAD preventive strategies may achieve slightly variable success in

individuals with different genetic susceptibility.

15 Diabetes Page 16 of 58

ACKNOWLEDGMENTS

The Diabetes Prevention Program Research Group acknowledges the commitment and dedication of the participants of the Diabetes Prevention Program and the Diabetes Prevention

Program Outcomes Study

FUNDING During the DPPOS, the National Institute of Diabetes and Digestive and Kidney Diseases

(NIDDK) of the National Institutes of Health provided funding (grant DK-48489) to the clinical centers and the Coordinating Center for the design and conduct of the study, and the collection, management, analysis, and interpretation of the data. The Southwestern American Indian Centers were supported directly by the NIDDK, including its Intramural Research Program, and the

Indian Health Service. The General Clinical Research Center Program, National Center for

Research Resources, and the Department of Veterans Affairs supported data collection at many of the clinical centers. Funding was also provided by the National Institute of Child Health and

Human Development; the National Institute on Aging; the National Eye Institute; the National

Heart, Lung, and Blood Institute; the Office of Research on Women’s Health; the National

Institute on Minority Health and Health Disparities; the Centers for Disease Control and

Prevention; the American Diabetes Association; the Swedish Research Council; and the

European Commission (CoG-2015_681742_NASCENT and H2020-MSCA-IF-2015-703787).

Bristol-Myers Squibb and Parke-Davis provided additional funding and material support during the Diabetes Prevention Program (DPP). Lipha (Merck-Sante) provided medication, and

LifeScan, Inc. donated materials during the DPP and DPPOS.

16 Page 17 of 58 Diabetes

DUALITY OF INTEREST

None.

AUTHOR CONTRIBUTIONS

J.M. conceived the analysis plan and contributed to interpretation, and writing and editing of the

report. K.A.J. conducted statistical analyses, contributed to the interpretation of the results, and

edited and reviewed the report before submission. J.M.M. was responsible of imputation and

contributed to PRS design. J.M.M., S.E.K., L.C., M.H., L.M.D., M.R.G.A., G.A.W., S.B.R.J.,

U.N.I., P.W.F., W.C.K. and J.C.F. contributed to the interpretation of the results and edited and

reviewed the article before submission. S.E.K., L.M.D., E.S., U.N.I., W.C.K., and the DPP

Research Group conducted the clinical trial and obtained the phenotypic data. J.C.F. is the

guarantor of this work and, as such, had full access to all the data in the study and takes

responsibility for the integrity of the data and the accuracy of the data analysis.

PRIOR PRESENTATION

Parts of this study were presented in abstract for at the 77th Scientific Sessions of the American

Diabetes Association, San Diego, CA, June 9 – 13, 2017

DATA AND RESOURCE AVAILABILITY

Data used in this study are available on request at [email protected], or by accessing the

NIDDK Repository.

17 Diabetes Page 18 of 58

REFERENCES

1. Global burden of diasease 2016 study. Global, regional, and national age-sex specific mortality for 264 causes of death, 1980-2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet. 2017;390:1151–210.of Disease Study 2016. Lancet. 2017;390:1151–210.

2. Hansson GK. Inflammation, Atherosclerosis, and Coronary Artery Disease. N Engl J Med. 2005;352:1685–95.

3. Shulman GI. Ectopic Fat in Insulin Resistance, Dyslipidemia, and Cardiometabolic Disease. N Engl J Med. 2014;371:2236–8.

4. Haffner SM, Lehto S, Rönnemaa T, Pyörälä K, Laakso M. Mortality from coronary heart disease in subjects with type 2 diabetes and in nondiabetic subjects with and without prior myocardial infarction. N Engl J Med. 1998;339:229–34.

5. Sarwar N, Gao P, Seshasai SRK, Gobin R, Kaptoge S, Di Angelantonio E, et al. Diabetes mellitus, fasting blood glucose concentration, and risk of vascular disease: a collaborative meta-analysis of 102 prospective studies. Lancet. 2010;375:2215–22.6.

6. UK Prospective Diabetes Study (UKPDS) Group. Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet. 1998;352:837–53.

7. Gaede P, Vedel P, Larsen N, Jensen GVH, Parving H-H, Pedersen O. Multifactorial intervention and cardiovascular disease in patients with type 2 diabetes. N Engl J Med. 2003;348:383–93.

8. Colhoun HM, Betteridge DJ, Durrington PN, Hitman GA, W Neil HA, Livingstone SJ, et al. Primary prevention of cardiovascular disease with atorvastatin in type 2 diabetes in the Collaborative Atorvastatin Diabetes Study (CARDS): multicentre randomised placebo- controlled trial. Lancet. 2004;364:685–96.

9. Holman RR, Paul SK, Bethel MA, Neil HAW, Matthews DR. Long-Term Follow-up after Tight Control of Blood Pressure in Type 2 Diabetes. N Engl J Med. 2008;359:1565–76.

10. Cholesterol Treatment Trialists’ (CTT) Collaboration, Fulcher J, O’Connell R, Voysey M,

18 Page 19 of 58 Diabetes

Emberson J, Blackwell L, et al. Efficacy and safety of LDL-lowering therapy among men and women: meta-analysis of individual data from 174,000 participants in 27 randomised trials. Lancet. 2015;385:1397–405.

11. Feero WG, Guttmacher AE, McCarthy MI. Genomics, Type 2 Diabetes, and Obesity. N Engl J Med. 2010;363:2339–50.

12. Franks PW, McCarthy MI. Exposing the exposures responsible for type 2 diabetes and obesity. Science. 2016;354:69–73.

13. Knowler WC, Barrett-Connor E, Fowler SE, Hamman RF, Lachin JM, Walker EA, et al. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med. 2002;346:393–403.

14. Salas-Salvadó J, Bulló M, Estruch R, Ros E, Covas M-I, Ibarrola-Jurado N, et al. Prevention of diabetes with Mediterranean diets: a subgroup analysis of a randomized trial. Ann Intern Med. 2014;160:1–10.

15. Diabetes Prevention Program Research Group. Long-term effects of lifestyle intervention or metformin on diabetes development and microvascular complications over 15-year follow-up: the Diabetes Prevention Program Outcomes Study. Lancet Diabetes Endocrinol. 2015;3:866–75.

16. Diabetes Prevention Program Research Group The Diabetes Prevention Program. Design and methods for a clinical trial in the prevention of type 2 diabetes. Diabetes Care. 1999;22:623–34.

17. The Diabetes Prevention Program: baseline characteristics of the randomized cohort. The Diabetes Prevention Program Research Group. Diabetes Care. 2000;23:1619–29.

18. Kriska, AM.; Caspersen CJ. Introduction to a Collection of Physical Activity Questionnaires. Med Sci Sport Exerc. 1997;29:5–9.

19. Mayer-Davis EJ, Vitolins MZ, Carmichael SL, Hemphill S, Tsaroucha G, Rushing J, et al. Validity and reproducibility of a food frequency interview in a Multi-Cultural Epidemiology Study. Ann Epidemiol. 1999;9:314–24.

20. Chiuve SE, Fung TT, Rimm EB, Hu FB, McCullough ML, Wang M, et al. Alternative

19 Diabetes Page 20 of 58

dietary indices both strongly predict risk of chronic disease. J Nutr. 2012;142:1009–18.

21. Warnick GR. Enzymatic methods for quantification of lipoprotein lipids. Methods Enzymol. 1986;129:101–23.

22. Goldberg RB, Temprosa MG, Mather KJ, Orchard TJ, Kitabchi AE, Watson KE, et al. Lifestyle and metformin interventions have a durable effect to lower CRP and tPA levels in the diabetes prevention program except in those who develop diabetes. Diabetes Care. 2014;37:2253–60.

23. Delaneau O, Marchini J, Zagury J-F. A linear complexity phasing method for thousands of genomes. Nat Methods. 2011;9:179–81.

24. 1000 Genomes Project Consortium A, Auton A, Brooks LD, Durbin RM, Garrison EP, Kang HM, et al. A global reference for human genetic variation. Nature. 2015;526:68–74.

25. Howie BN, Donnelly P, Marchini J. A Flexible and Accurate Genotype Imputation Method for the Next Generation of Genome-Wide Association Studies. PLoS Genet. 2009;5:e1000529.

26. van der Harst P, Verweij N. Identification of 64 Novel Genetic Loci Provides an Expanded View on the Genetic Architecture of Coronary Artery Disease. Circ Res. 2018;122:433–43.

27. Morieri ML, Gao H, Pigeyre M, Shah HS, Sjaarda J, Mendonca C, et al. Genetic Tools for Coronary Risk Assessment in Type 2 Diabetes: A Cohort Study From the ACCORD Clinical Trial. Diabetes Care. 2018;41:2404–13.

28. Nikpay M, Goel A, Won H-H, Hall LM, Willenborg C, Kanoni S, et al. A comprehensive 1000 Genomes–based genome-wide association meta-analysis of coronary artery disease. Nat Genet. 2015;47:1121–30.

29. Webb TR, Erdmann J, Stirrups KE, Stitziel NO, Masca NGD, Jansen H, et al. Systematic Evaluation of Pleiotropy Identifies 6 Further Loci Associated With Coronary Artery Disease. J Am Coll Cardiol. 2017;69:823–36.

30. Willer CJ, Schmidt EM, Sengupta S, Peloso GM, Gustafsson S, Kanoni S, et al. Discovery and refinement of loci associated with lipid levels. Nat Genet. 2013;45:1274–83.

20 Page 21 of 58 Diabetes

31. Deloukas P, Kanoni S, Willenborg C, Farrall M, Assimes TL, Thompson JR, et al. Large- scale association analysis identifies new risk loci for coronary artery disease. Nat Genet. 2013;45:25–33.

32. Khera AV, Kathiresan S. Genetics of coronary artery disease: discovery, biology and clinical translation. Nat Rev Genet. 2017;18:331-44.

33. Ehret GB, Ferreira T, Chasman DI, Jackson AU, Schmidt EM, Johnson T, et al. The genetics of blood pressure regulation and its target organs from association studies in 342,415 individuals. Nat Genet. 2016;48:1171–84.

34. Fuchsberger C, Flannick J, Teslovich TM, Mahajan A, Agarwala V, Gaulton KJ, et al. The genetic architecture of type 2 diabetes. Nature. 2016;536:41–7.

35. Khera AV, Emdin CA, Drake I, Natarajan P, Bick AG, Cook NR, et al. Genetic Risk, Adherence to a Healthy Lifestyle, and Coronary Disease. N Engl J Med. 2016;375:2349– 58.

36. Hivert M-F, Jablonski KA, Perreault L, Saxena R, McAteer JB, Franks PW, 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. 2011;60:1340–8.

37. Look AHEAD Research Group. Prospective association of a genetic risk score and lifestyle intervention with cardiovascular morbidity and mortality among individuals with type 2 diabetes: the Look AHEAD randomised controlled trial. Diabetologia. 2015;58:1803–13.

38. Tuomilehto J, Lindström J, Eriksson JG, Valle TT, Hämäläinen H, Ilanne-Parikka P, et al. Prevention of type 2 diabetes mellitus by changes in lifestyle among subjects with impaired glucose tolerance. N Engl J Med. 2001;344:1343–50.

39. Ramachandran A, Snehalatha C, Mary S, Mukesh B, Bhaskar AD, Vijay V, et al. The Indian Diabetes Prevention Programme shows that lifestyle modification and metformin prevent type 2 diabetes in Asian Indian subjects with impaired glucose tolerance (IDPP- 1). Diabetologia. 2006;49:289–97.

21 Diabetes Page 22 of 58

Table 1. Baseline characteristics of Diabetes Prevention Program participants according to randomization group among those included in the present analysis.

Total PBO MET ILS P (n=2,658) (n=888) (n=880) (n=890) value* Demographics Age, y 50.7 ± 10.7 50.6 ± 10.4 50.9 ± 10.3 50.6 ± 11.4 0.93 Sex, Female n (%) 1,789 (67.3) 607 (68.4) 583 (66.3) 599 (67.3) 0.641 Race/Ethnicity, n (%) White 1476 (55.5) 490 (55.2) 507 (57.6) 479 (53.8) 0.298 African American 537 (20.2) 186 (20.9) 178 (20.2) 173 (19.4) Hispanic 451 (17.0) 147 (16.6) 143 (16.3) 161 (18.1) Other 194 ( 7.3) 65 ( 7.3) 52 ( 5.9) 77 ( 8.7) Current Smoker, n (%) 192 ( 7.2) 69 ( 7.8) 60 ( 6.8) 63 ( 7.1) 0.726 Hyperlipidemia n (%) 136 ( 5.1) 50 ( 5.6) 47 ( 5.3) 39 ( 4.4) 0.486 On lipid lowering med, n (%) 126 (4.7) 45 (5.1) 45 (5.1) 36 (4.0) 0.455 Hypertension, n (%) 427 (16.1) 139 (15.7) 140 (15.9) 148 (16.6) 0.845 Systolic blood pressure, mm Hg 123.7 ± 14.7 123.5 ± 14.4 124.2 ± 15.0 123.5 ± 14.6 0.961 Diastolic blood pressure, mm Hg 78.3 ± 9.3 78.0 ± 9.2 78.3 ± 9.5 78.5 ± 9.1 0.227 Anthropometrics & Lifestyle BMI, kg/m2 34.1 ± 6.6 34.3 ± 6.7 34.0 ± 6.6 34.0 ± 6.6 0.312 Waist circumference, cm 105.3 ± 14.5 105.5 ± 14.3 105.0 ± 14.4 105.4 ± 14.8 0.823 AHEI-2010, units 46.4 ± 10.1 45.9 ± 10.2 44.9 ± 9.9 48.6 ± 9.9 0.001 MET-hours/week# 9.5 (3.8, 20.3) 9.8 (4.0, 21.1) 9.9 (3.8, 21.2) 9.1 (3.8, 18.1) 0.180 Biochemical Fasting glucose, mmol/L 5.93 ± 0.45 5.95 ± 0.47 5.92 ± 0.46 5.92 ± 0.44 0.117 HOMA-IR, units# 6.2 (4.3, 9.0) 6.2 (4.2, 8.8) 6.2 (4.3, 9.1) 6.2 (4.3, 9.1) 0.938 Total cholesterol, mmol/L 5.27 ± 0.94 5.26 ± 0.94 5.26 ± 0.93 5.28 ± 0.95 0.638 LDL-cholesterol, mmol/L 3.23 ± 0.85 3.22 ± 0.86 3.23 ± 0.84 3.24 ± 0.85 0.692 HDL-cholesterol, # 1.14 (0.96, 1.35) 1.11 (0.96, 1.32) 1.14 (0.98,1.32) 1.14 (0.96, 1.35) 0.022 Triglycerides, mmol/L# 1.61 (1.12, 2.31) 1.67 (1.18, 2.37) 1.58 (1.11, 2.25) 1.57 (1.09, 2.27) 0.005 CRP, mg/dl# 0.38 (0.17, 0.77) 0.38 (0.17, 0.80) 0.36 (0.17, 0.74) 0.39 (0.18, 0.76) 0.472 tPA, ng/ml# 11.0 (8.8, 13.4) 10.9 (8.8, 13.5) 10.9 (8.7, 13.2) 11.2 (8.7, 13.5) 0.902 Fibrinogen, umol/L 11.3 ± 2.5 11.4 ± 2.6 11.2 ± 2.5 11.3 ± 2.5 0.840 Genetics Genetic Risk Score, units 172 ± 9 172 ± 9 172 ± 9 173 ± 9 0.222

22 Page 23 of 58 Diabetes

Values are given as the mean ± SD or median (25th, 75th percentile) # except for qualitative variables expressed as n (%). BMI; Body mass index, HOMA-IR; homeostasis model assessment of insulin resistance, CRP; C-reactive protein, tPA, tissue plasminogen activator. P values obtained by Kruskall-Wallis#, ANOVA, and chi-squared tests with 2 d.f. as appropriate.

23 Diabetes Page 24 of 58

Table 2. Baseline association between the genetic risk score and coronary artery disease risk factors.

Cardiometabolic risk factor P  95% CI value BMI, kg/m2 -0.029 -0.319, 0.261 0.845 Waist circumference, cm 0.071 -0.583, 0.724 0.832 Fasting glucose, mmol/L 0.015 -0.014, 0.044 0.306 LDLc, mmol/L 0.094 0.055, 0.133 <0.001 PRS HDLc, # 0.991 0.980, 1.001 0.090

Triglycerides, # 1.017 0.994, 1,041 0.157 Systolic BP, mmHg 0.544 -0.107, 1.197 0.102 Diastolic BP, mmHg 0.524 0.094, 0.953 0.017 CRP, # 1.029 0.982, 1.077 0.230 tPA, # 0.996 0.980, 1.014 0.694 Fibrinogen, umol/L 0.039 -0.077, 0.154 0.511

Table Legend: Each cardiometabolic risk factor (CRFs) represents a separate general linear model on the association between baseline risk factor levels and the increment of 10 risk alleles in the PRS. Regression coefficients for natural log transformed CRFs# are expressed as exp() differences in baseline risk factor levels. BMI; Body mass index, CRP; C-reactive protein, tPA, tissue plasminogen activator. P-value derived by general linear model adjusted for age at randomization, sex, and PCs ancestry markers.

24 Page 25 of 58 Diabetes

Table 3. Effect of type 2 diabetes preventative strategies on one-year change in coronary artery disease risk factors by study intervention.

PBO MET ILS Cardiometabolic risk (n=888) (n=880) (n=890) P ILS P MET factor vs PBO vs PBO Mean (95% CI) Mean (95% CI) Mean (95% CI)

BMI, kg/m2 -0.192 (-0.326, -0.057) -0.911 (-1.047, -0.774) -2.465 (-2.601, -2.330) <0.001 <0.001

Waist circumference, cm -0.843 (-1.268, -0.417) -2.153 (-2.582, -1.725) -6.479 (-6.905, -6.053) <0.001 <0.001

Fasting glucose, # 0.001 (-0.006, 0.007) -0.039 (-0.045, -0.032) -0.052 (-0.058, -0.045) <0.001 <0.001

LDLc, mmol/L -0.060 (-0.097, -0.023) -0.116 (-0.154, -0.078) -0.161 (-0.199, -0.124) <0.001 0.038

HDL, mmol/L -0.006 (-0.017, 0.005) 0.021 (0.010, 0.032) 0.038 (0.027, 0.049) <0.001 <0.001

Triglycerides, # -0.060 (-0.084, -0.036) -0.047 (-0.071, -0.023) -0.169 (-0.193, -0.145) <0.001 0.452

Systolic BP, mmHg -0.728 (-1.538, 0.081) -0.856 (-1.672, -0.040) -3.003 (-3.814, -2.192) <0.001 0.828

Diastolic BP, mmHg -0.955 (-1.483, -0.428) -1.259 (-1.791, -0.728) -3.550 (-4.078, -3.022) <0.001 0.427

CRP, # -0.020 (-0.069, 0.029) -0.146 (-0.196, -0.097) -0.396 (-0.445, -0.347) <0.001 <0.001

tPA, ng/mL -0.712 (-0.910, -0.515) -2.049 (-2.249, -1.849) -2.564 (-2.762, -2.367) <0.001 <0.001

Fibrinogen, umol/L 0.084 (-0.039, 0.208) -0.063 (-0.188, 0.061) -0.298 (-0.421, -0.174) <0.001 0.100 Table Legend: Values in this table represents adjusted mean change in cardiometabolic risk factor levels and 95% CIs. For non- normally distributed variables#, we calculated the natural log year 1 value minus natural log baseline value and showed mean change in cardiometabolic risk factor levels and 95% CIs. P values from t-test comparing mean change in ILS or MET to PBO derived by general linear model adjusted for baseline risk factor, age at randomization, sex, and PCs ancestry markers.

25 Diabetes Page 26 of 58

Table 4. Association between coronary artery disease polygenic risk score and one-year change in risk factors by study intervention.

All study participants PBO MET ILS Cardiometabolic risk (n=2,658) (n=888) (n=880) (n=890) factor P P P P  (95% CI)  (95% CI)  (95% CI)  (95% CI) value* value* value* value* BMI, kg/m2 0.013 (-0.084, 0.109) 0.799 0.015 (-0.134, 0.164) 0.841 0.021 (-0.121, 0.163) 0.770 0.001 (-0.204, -0.201) 0.994

Waist circumference, cm -0.027 (-0.330, 0.277) 0.863 -0.008 (-0.474, 0.457) 0.972 -0.212 (-0.724, 0.288) 0.398 0.168 (-0.433, 0.769) 0.584

Fasting glucose, # 0.998 (0.994, 1.004) 0.728 0.999 (0.988, 1.005) 0.414 0.998 (0.991,1.005) 0.526 1.002 (0.995,1.011) 0.502

Diabetes risk 1.021 (0.919, 1.136) 0.696 0.952 (0.809, 1.120) 0.552 1.088 (0.904, 1.309) 0.371 1.068 (0.851, 1.341) 0.571

LDLc, mmol/L 0.010 (-0.017, 0.036) 0.481 0.012 (-0.036, 0.059) 0.632 -0.019 (-0.065, 0.028) 0.424 0.034 (-0.012, 0.080) 0.152

HDLc, mmol/L -0.007 (-0.015, 0.001) 0.075 -0.009 (-0.022, 0.003) 0.147 -0.002 (-0.016, 0.012) 0.822 -0.013 (-0.028, 0.002) 0.094

Triglycerides, # 1.016 (0.999, 1.033) 0.059 1.009 (0.998, 1.038) 0.562 1.011 (0.993, 1.040) 0.427 1.028 (0.999, 1.058) 0.063

Systolic BP, mmHg 0.489 (-0.088, 1.067) 0.097 0.685 (-0.314, 1.683) 0.179 -0.288 (-1.267, 0.692) 0.565 0.951 (-0.073, 1.974) 0.069

Diastolic BP, mmHg 0.085 (-0.291, 0.462) 0.657 0.052 (-0.606, 0.711) 0.876 -0.150 (-0.797, 0.497) 0.649 0.256 (-0.398, 0.911) 0.442

CRP, # 0.981 (0.947, 1.015) 0.272 0.967 (0.913, 1.024) 0.252 1.007 (0.946, 1.067) 0.811 0.971 (0.927, 1.033) 0.354 tPA, ng/mL 0.056 (-0.085, 0.197) 0.440 0.090 (-0.162, 0.024) 0.484 0.081 (-0.155, 0.318) 0.501 0.229 (-0.214, 0.259) 0.850

Fibrinogen, umol/L -0.019 (-0.106, 0.069) 0.679 -0.028 (-0.188, 0.131) 0.728 0.006 (-0.142, 0.154) 0.935 -0.037 (-0.188, 0.115) 0.634

Table Legend: Each cardiometabolic risk factor represents a separate linear regression model on the association between one-year change in cardiometabolic risk factor levels and the increment of 10 risk alleles in the PRS. Regression coefficients for natural log transformed CRFs# are expressed as exp() ratio of one-year change in coronary artery disease risk factors in the entire study

26 Page 27 of 58 Diabetes

population and according to intervention group. Cox models were used to investigate the association between CAD PRS and diabetes incidence (estimates are reported as hazard ratios and 95% CI). BMI; Body mass index, CRP; C-reactive protein, tPA, tissue plasminogen activator. *P-value derived by general linear model adjusted for baseline risk factor, age at randomization, sex, and PCs ancestry markers.

27 Diabetes Page 28 of 58

Table 5. Interaction between coronary artery disease polygenic risk score and intervention group on one-year change in coronary artery disease risk factors.

Cardiometabolic risk ILS vs PBO MET vs PBO

factor Pinteraction Pinteraction BMI, kg/m2 0.617 0.807 Waist circumference, cm 0.670 0.787 Fasting glucose# 0.181 0.394 Diabetes risk 0.166 0.198 LDLc, mmol/L 0.905 0.395 HDLc, mmol/L 0.486 0.586 Triglycerides# 0.649 0.838 Systolic BP, mmHg 0.952 0.157 Diastolic BP, mmHg 0.348 0.818 CRP# 0.745 0.285 tPA, ng/mL 0.470 0.800 Fibrinogen, umol/L 0.687 0.982

Table Legend: Values in this table represents adjusted interaction P values. Pinteraction values were derived by general linear models with interaction terms for intervention arms (ILS or MET vs. PBO), after adjustment for age at randomization, sex, the top 10 PCs for ancestry, and the respective baseline CRFs. For non-normally distributed variables#, we calculated the natural log year 1 value minus natural log baseline value. Cox models were used to investigate interactions between type 2 diabetes prevention strategies and CAD PRS on diabetes incidence. BMI; Body mass index, CRP; C-reactive protein, tPA, tissue plasminogen activator.

28 Page 29 of 58 Diabetes

Supplementary Material

Supplement to: Interaction between type 2 diabetes prevention strategies and genetic

determinants of coronary artery disease on cardiometabolic risk factors.

Jordi Merino, Kathleen A Jablonski, Josep M. Mercader, Steven E Kahn, Ling Chen, Maegan

Harden, Linda M Delahanty, Maria Rosario G. Araneta, Geoffrey A Walford, Suzanne B.R.

Jacobs, Uzoma N. Ibebuogu, Paul W. Franks, William C. Knowler, Jose C. Florez; for the

Diabetes Prevention Program Research Group.

Corresponding author’s contact information:

Jose C. Florez, MD, PhD

Diabetes Prevention Program Coordinating Center

The Biostatistics Center George Washington University

6110 Executive Blvd, Suite 750

Rockville, MD 20852, USA

Phone: 617 643 3308

Email: [email protected] and [email protected] Diabetes Page 30 of 58

Supplementary table 1. Characteristics of genetic variants used to build the coronary artery disease polygenic risk score.

Separate Excel file Page 31 of 58 Diabetes

Supplementary Table 2. Differences in baseline characteristics between the sample of participants included in this study and all Diabetes Prevention Program participants.

DPP Study Sample All DPP P (n=2,658) (n=3,234) value* Demographics Age, y 50.7 ± 10.7 50.6 ± 10.7 0.444 Sex, Female n (%) 1,789 (67.3) 2,191 (67.7) 0.242 Race/Ethnicity, n (%) White 1,476 (55.5) 1,768 (54.7) <0.001 African American 537 (20.2) 645 (19.9) Hispanic 451 (17.0) 508 (15.7) Other 194 ( 7.3) 313 (9.6) Current Smoker, n (%) 192 ( 7.2) 226 (7.0) 0.306 Hyperlipidemia n (%) 136 ( 5.1) 169 (5.2) 0.824 On lipid lowering med, n (%) 126 (4.7) 157 (4.9) 0.567 Hypertension, n (%) 427 (16.1) 520 (16.1) 0.962 Systolic blood pressure, mm Hg 124 ± 15 124 ± 15 0.893 Diastolic blood pressure, mm Hg 78.3 ± 9.3 78 ± 9 0.625 Anthropometrics & Lifestyle BMI, kg/m2 34.1 ± 6.6 34.0 ± 6.7 0.039 Waist circumference, cm 105 ± 15 105 ± 15 0.047 AHEI-2010, units 46.4 ± 10.1 44.2 ± 10.4^ 0.001 MET-hours/week# 9.5 (3.8, 20.2) 9.8 (3.9, 20.6) 0.286 Biochemical Fasting glucose, mmol/L 5.93 ± 0.45 5.91 ± 0.46 <0.001 HOMA-IR, units# 6.18 (4.25, 8.95) 6.16 (4.22, 8.85) 0.252 Total cholesterol, mmol/L 5.27± 0.94 5.26 ± 0.94 0.437 LDL-cholesterol, mmol/L 3.23 ± 0.85 3.23 ± 0.85 0.734 HDL-cholesterol, # 1.18 (0.96, 1.35) 1.10 (0.98, 1.30) 0.380 Triglycerides, mmol/L# 1.61 (1.19, 2.31) 1.59 (1.12, 2.27) 0.236 CRP, mg/dl# 0.38 (0.17, 0.77) 0.37 (0.17, 0.76) 0.010 tPA, ng/ml# 11.0 (8.8, 13.4) 10.9 (8.7, 13.4) 0.078 Fibrinogen, umol/L 11.3 ± 2.5 11.5 ± 2.6 0.305 Values are given as the mean ± SD or median (25th, 75th percentile) # except for qualitative variables expressed as n (%). BMI; Body mass index, HOMA-IR; homeostasis model assessment of insulin resistance, CRP; C-reactive protein, tPA, tissue plasminogen activator. *P values are generated from a general estimating equation models. ^ n=3,175 Diabetes Page 32 of 58

Supplementary Table 3. One-year change in lifestyle components.

PBO MET ILS P Met vs P ILS vs (n=865) (n=852) (n=869) PBO PBO Diet 1-y change AHEI-2010, units 1.49 ± 0.27 0.84 ± 0.27 4.22 ± 0.27 0.661 <0.001 Physical activity 1-y change MET-hours/week 1.28 ± 1.05 1.46 ± 1.06 6.96 ± 1.05 0.588 <0.001 Body weight loss, kg -0.53 ± 0.19 -2.55 ± 0.19 -6.9 ± 0.19 <0.001 <0.001

Table Legend; Values represents adjusted means ± SE. The change in physical activity is not normally distributed and has been ln transformed.

P-value extracted from general linear model adjusted for baseline lifestyle component, age at randomization, sex and ancestry markers.

Eating Index 2010 (AHEI-2010). MET-hours/week (Metabolic equivalents*hours/week). Page 33 of 58 Diabetes

Supplementary Table 4. Association between one-year change in body weight and one-year change in risk factors by study intervention.

PBO MET ILS (n=888) (n=880) (n=890) Cardiometabolic risk factor P P  (95% CI)  (95% CI)  (95% CI) P value* value* value* BMI, kg/m2 -0.11 (-0.012, -0.010) <0.001 -0.011 (-0.012, -0.010) <0.001 -0.011 (-0.012, -0.010) <0.001 Waist circumference, cm -0.681 (-0.743, -0.619) <0.001 -0.626 (-0.703, -0.549) <0.001 -0.756 (-0.806, -0.706) <0.001

Fasting glucose# 0.994 (0.993,0.996) <0.001 0.995 (0.994, 0.996) <0.001 0.995 (0.994, 0.996) <0.001

LDLc, mmol/L -0.006 (-0.014, 0.002) 0.143 -0.011 (-0.019, -0.003) 0.008 -0.016 (-0.021, -0.010) <0.001 HDLc, mmol/L 0.003 (0.005, 0.007) 0.002 0.008 (0.006, 0.010) <0.001 0.004 (0.002, 0.006) <0.001

Triglycerides# 0.981 (0.975, 0.990) <0.001 0.978 (0.974, 0.984) <0.001 0.981 (0.976, 0.989) <0.001

Systolic BP, mmHg -0.456 (-0.612, -0.285) <0.001 -0.368 (-0.536, -0.199) <0.001 -0.374 (-0.494, -0.254) <0.001 Diastolic BP, mmHg -0.220 (-0.329, -0.111) <0.001 -0.197 (-0.309, -0.085) <0.001 -0.245 (-0.321, -0.169) <0.001

CRP# 0.966 (0.957, 0.975) <0.001 0.974 (0.964, 0.983) <0.001 0.964 (0.957, 0.971) <0.001

tPA, ng/mL -0.130 (-0.167, -0.093) <0.001 -0.175 (-0.214, -0.136) <0.001 -0.163 (-0.188, -0.137) <0.001 Fibrinogen, umol/L 0.010 (-0.017, 0.037) 0.412 -0.004 (-0.020, 0.029) 0.784 -0.013 (-0.048, 0.005) 0.151

Table Legend; Each cardiometabolic risk factor represents a separate linear regression model on the association between in cardiometabolic risk factor levels and body weight loss (1kg). Regression coefficients for natural log transformed CRFs# are expressed as exp() ratio of one-year change in coronary artery disease risk factors according to intervention group. BMI; Body mass index, CRP; C-reactive protein, tPA, tissue plasminogen activator. *P-value derived by general linear model adjusted for baseline risk factor, age at randomization, sex, and PCs ancestry markers. Diabetes Page 34 of 58

Supplementary Table 5. Association between one-year change in physical activity and one-year change in risk factors by study intervention.

PBO MET ILS (n=888) (n=880) (n=890) Cardiometabolic risk factor P P  (95% CI)  (95% CI)  (95% CI) P value* value* value* BMI, kg/m2 -0.024 (-0.051, 0.016) 0.108 -0.015 (-0.022, 0.015) 0.458 -0.030 (-0.053, -0.011) 0.008 Waist circumference, cm -0.084 (-0.195, 0.027) 0.136 -0.032 (-0.160, 0.097) 0.629 -0.227 (-0.439, -0.016) 0.035

Fasting glucose# 0.097 (0.991,1.003) 0.416 0.999 (0.997, 1.001) 0.837 -0.998 (0.995, 1.001) 0.087

LDLc, mmol/L 0.003 (-0.008, 0.015) 0.554 -0.011 (-0.023, 0.001) 0.081 -0.005 (-0.021, 0.011) 0.563 HDLc, mmol/L 0.001 (-0.003, 0.003) 0.907 -0.002 (-0.006, 0.001) 0.249 0.001 (-0.004, 0.006) 0.693

Triglycerides# 0.996 (-0.980, 1.003) 0.235 0.997 (0.990, 1.005) 0.439 1.002 (0.991, 1.012) 0.759

Systolic BP, mmHg -0.158 (-0.398, 0.081) 0.196 -0.150 (-0.399, 0.099) 0.238 -0.448 (-0.810, -0.087) 0.015 Diastolic BP, mmHg -0.021 (-0.178, 0.137) 0.798 -0.101 (-0.265, 0.063) 0.226 -0.314 (-0.545, -0.084) 0.008

CRP# 1.009 (0.995, 1.023) 0.187 1.004 (0.989, 1.013) 0.613 0.990 (0.967, 1.012) 0.368

tPA, ng/mL 0.006 (-0.054, 0.067) 0.837 -0.027 (-0.091, 0.038) 0.420 -0.037 (-0.119, 0.046) 0.388 Fibrinogen, umol/L 0.035 (-0.004, 0.073) 0.076 0.001 (-0.024, 0.051) 0.482 0.001 (-0.005, 0.005) 0.959

Table Legend; Each cardiometabolic risk factor represents a separate linear regression model on the association between in cardiometabolic risk factor levels and increment of 10 MET-hours/week. Regression coefficients for natural log transformed CRFs# are expressed as exp() ratio of one-year change in coronary artery disease risk factors according to intervention group. BMI; Body mass index, CRP; C-reactive protein, tPA, tissue plasminogen activator. *P-value derived by general linear model adjusted for baseline risk factor, age at randomization, sex, and PCs ancestry markers. Page 35 of 58 Diabetes

Supplementary Table 6. Association between one-year change in dietary score and one-year change in risk factors by study intervention.

PBO MET ILS (n=888) (n=880) (n=890) Cardiometabolic risk factor P P  (95% CI)  (95% CI)  (95% CI) P value* value* value* BMI, kg/m2 -0.060 (-0.108, -0.013) 0.011 -0.071 (-0.150, -0.006) 0.005 -0.094 (-0.158, -0.026) <0.001 Waist circumference, cm -0.314 (-0.774, 0.146) 0.181 -0.703 (-1.241, -0.192) 0.007 -1.294 (-1.852, -0.192) <0.001

Fasting glucose# 0.994 (0.983,1.005) 0.231 0.997 (0.982, 1.010) 0.712 0.995 (0.988, 1.004) 0.253

LDLc, mmol/L -0.011 (-0.058, -0.035) 0.648 0.030 (-0.016, 0.076) 0.206 -0.041 (-0.084, 0.001) 0.056 HDLc, mmol/L 0.003 (-0.001, 0.015) 0.662 0.013 (-0.001, 0.027) 0.056 0.022 (0.008, 0.036) 0.002

Triglycerides# 0.990 (0.961, 1.002) 0.539 0.966 (0.938, 0.996) 0.018 0.951 (0.928, 0.989) 0.008

Systolic BP, mmHg -0.222 (-1.127, 0.762) 0.658 -0.811 (-1.794, -0.173) 0.106 -0.798 (-1.765, -0.170) 0.106 Diastolic BP, mmHg -0.018 (-0.665, 0.629) 0.957 -0.783 (-1.429, -0.136) 0.018 -0.256 (-0.877, -0.365) 0.419

CRP# 0.933 (0.886, 0.990) 0.019 0.996 (0.943, 1.064) 0.914 0.938 (0.863, 0.980) 0.011

tPA, ng/mL -0.126 (-0.375, 0.123) 0.320 -0.103 (-0.338, 0.132) 0.388 -0.165 (-0.388, 0.057) 0.145 Fibrinogen, umol/L -0.017 (-0.174, 0.139) 0.829 -0.172 (-0.320, -0.024) 0.023 -0.133 (-0.276, 0.010) 0.067

Table Legend; Each cardiometabolic risk factor represents a separate linear regression model on the association between in cardiometabolic risk factor levels and increment of 10 units in the dietary score. Regression coefficients for natural log transformed CRFs# are expressed as exp() ratio of one-year change in coronary artery disease risk factors according to intervention group. BMI; Body mass index, CRP; C-reactive protein, tPA, tissue plasminogen activator. *P-value derived by general linear model adjusted for baseline risk factor, age at randomization, sex, and PCs ancestry markers. Diabetes Page 36 of 58

Supplementary Table 7. Modification of coronary artery disease genetic risk effect on one-year change in intermediate risk factors by lifestyle components.

Low Intermediate High One-year genetic risk genetic risk genetic risk change Pinteraction One-year change Treatment in risk factor  (SE)  (SE)  (SE) in LS component arm BMI, kg/m2 Diet MET -0.191 (0.095) -0.288 (0.103) -0.515 (0.114) 0.012 BMI, kg/m2 PA MET -0.003 (0.018) -0.261 (0.098) -0.323 (0.123) 0.014

Triglycerides# Diet MET 0.996 (0.960, 1.031) 0.969 (0.925, 1.016) 0.934 (0.892, 0.977) 0.024

Fasting glucose# PA PBO 1.005 (0.991, 1.021) 0.993 (0.983, 1.002) 0.989 (0.977, 1.000) 0.012

HDLc, mmol/L PA PBO 0.022 (0.009) -0.011 (0.009) -0.012 (0.008) 0.033

Table Legend: Predicted one-year changes on cardiometabolic risk factors by 1SD increase in lifestyle behaviors among individuals at low, intermediate, and high genetic risk. Pinteraction values were derived by general linear models with interaction terms for CAD PRS and lifestyle behaviors (dietary quality score or physical activity), after adjustment for age at randomization, sex, the top 10 principal components for ancestry, and the respective baseline cardiac risk factors. For non-normally distributed variables#, we calculated the natural log year 1 value minus natural log baseline value. incidence. Displayed only nominal significant interactions. BMI, body mass index; MET, metformin; PBO, placebo. Page 37 of 58 Diabetes

Supplementary Fig. 1. Coronary artery disease polygenic risk score distribution.

Figure legend; Q-Q plot for the distribution of the CAD polygenic risk score. The y axis displays CAD polygenic risks core values. The x axes represent quartiles of the distribution in the CAD polygenic risk score. Diabetes Page 38 of 58

Supplementary Fig. 2. Adjusted-mean values of baseline lipid and diastolic blood pressure levels according quartiles of the genetic risk score.

A B 3.5 81

3.4 80

3.3 79

3.2 78 LDLc, mmol/L 3.1 mmHgBP, Diastolic 77

3.0 76 GRS Q1 GRS Q2 GRS Q3 GRS Q4 GRS Q1 GRS Q2 GRS Q3 GRS Q4

Figure legend; Baseline mean concentrations of intermediate risk factors according to the number of risk alleles at identified CAD loci, weighted by effect size in an aggregate genotype score. The right y axis displays adjusted mean and 95% confidence interval values of LDLc (A) and DBP (B) according to PRS quartiles. P-value was obtained from general linear models after adjustment for age at randomization, sex, and ancestry markers. P < 0.001 for LDLc and 0.027 for DBP. Page 39 of 58 Diabetes

DPP Research Group members

Pennington Biomedical Research Center (Baton Hermes J. Florez, MD Rouge, LA) Anna Giannella, RD, MS George A. Bray, MD* Lascelles Kirby, MS Iris W. Culbert, BSN, RN, CCRC** Carmen Larreal Catherine M. Champagne, PhD, RD Valerie McLymont, RN Jadell Mendez Barbara Eberhardt, RD, LDN Juliet Ojito, RN Frank Greenway, MD Arlette Perry, PhD Patrice Saab, PhD Fonda G. Guillory, LPN April A. Herbert, RD The University of Texas Health Science Center (San Michael L. Jeffirs, LPN Antonio, TX) Betty M. Kennedy, MPA Steven M. Haffner, MD, MPH* Jennifer C. Lovejoy, PhD Maria G. Montez, RN, MSHP, CDE** Laura H. Morris, BS Carlos Lorenzo, MD, PhD Lee E. Melancon, BA, BS Arlene Martinez, RN, BSN, CDE Donna Ryan, MD Deborah A. Sanford, LPN University of Colorado (Denver, CO) Kenneth G. Smith, BS, MT Richard F. Hamman, MD, DrPH* Lisa L. Smith, BS Patricia V. Nash, MS** Julia A. St. Amant, RTR Lisa Testaverde, MS** Richard T. Tulley, PhD Denise R. Anderson, RN, BSN Paula C. Vicknair, MS, RD Larry B. Ballonoff, MD Donald Williamson, PhD Alexis Bouffard, MA, Jeffery J. Zachwieja, PhD B. Ned Calonge, MD, MPH Lynne Delve University of Chicago (Chicago, IL) Martha Farago, RN Kenneth S. Polonsky, MD* James O. Hill, PhD Janet Tobian, MD, PhD* Shelley R. Hoyer, BS David Ehrmann, MD* Bonnie T. Jortberg, MS, RD, CDE Margaret J. Matulik, RN, BSN** Dione Lenz, RN, BSN Bart Clark, MD Marsha Miller, MS, RD Kirsten Czech, MS David W. Price, MD Catherine DeSandre, BA Judith G. Regensteiner, PhD Ruthanne Hilbrich, RD Helen Seagle, MS, RD Wylie McNabb, EdD Carissa M. Smith, BS Ann R. Semenske, MS, RD Sheila C. Steinke, MS Brent VanDorsten, PhD Jefferson Medical College (Philadelphia, PA) Jose F. Caro, MD* Joslin Diabetes Center (Boston, MA) Pamela G. Watson, RN, ScD* Edward S. Horton, MD* Barry J. Goldstein, MD, PhD* Kathleen E. Lawton, RN** Kellie A. Smith, RN, MSN** Ronald A. Arky, MD Jewel Mendoza, RN, BSN** Marybeth Bryant Renee Liberoni, MPH Jacqueline P. Burke, BSN Constance Pepe, MS, RD Enrique Caballero, MD John Spandorfer, MD Karen M. Callaphan, BA Om P. Ganda, MD University of Miami (Miami, FL) Therese Franklin Richard P. Donahue, PhD* Sharon D. Jackson, MS, RD, CDE Ronald B. Goldberg, MD* Alan M. Jacobsen, MD Ronald Prineas, MD, PhD* Lyn M. Kula, RD Patricia Rowe, MPA** Margaret Kocal, RN, CDE Jeanette Calles, MSEd Maureen A. Malloy, BS Paul Cassanova-Romero, MD Maryanne Nicosia, MS, RD Diabetes Page 40 of 58

Cathryn F. Oldmixon, RN Pamela A. Schinleber, RN, MS Jocelyn Pan, BS, MPH Marizel Quitingon Massachusetts General Hospital (Boston, MA) Stacy Rubtchinsky, BS David M. Nathan, MD* Ellen W. Seely, MD Charles McKitrick, BSN** Dana Schweizer, BSN Heather Turgeon, BSN** Donald Simonson, MD Kathy Abbott Fannie Smith, MD Ellen Anderson, MS, RD Caren G. Solomon, MD, MPH Laurie Bissett, MS, RD James Warram, MD Enrico Cagliero, MD Jose C. Florez, MD, PhD+ VA Puget Sound Health Care System and Linda Delahanty, MS, RD University of Washington (Seattle, WA) Valerie Goldman, MS, RD Steven E. Kahn, MB, ChB* Alexandra Poulos Brenda K. Montgomery, RN, BSN, CDE** Wilfred Fujimoto, MD University of California-San Diego (La Jolla, CA) Robert H. Knopp, MD Jerrold M. Olefsky, MD* Edward W. Lipkin, MD Elizabeth Barrett-Connor, MD* Michelle Marr, BA Mary Lou Carrion-Petersen, RN, BSN** Dace Trence, MD Steven V. Edelman, MD Robert R. Henry, MD University of Tennessee (Memphis, TN) Javiva Horne, RD Abbas E. Kitabchi, PhD, MD, FACP* Simona Szerdi Janesch, BA Mary E. Murphy, RN, MS, CDE, MBA** Diana Leos, RN, BSN William B. Applegate, MD, MPH Sundar Mudaliar, MD Michael Bryer-Ash, MD William Polonsky, PhD Sandra L. Frieson, RN Jean Smith, RN Raed Imseis, MD Karen Vejvoda, RN, BSN, CDE, CCRC Helen Lambeth, RN, BSN Lynne C. Lichtermann, RN, BSN St. Luke’s-Roosevelt Hospital (New York, NY) Hooman Oktaei, MD F. Xavier Pi-Sunyer, MD* Lily M.K. Rutledge, RN, BSN Jane E. Lee, MS** Amy R. Sherman, RD, LD David B. Allison, PhD Clara M. Smith, RD, MHP, LDN Nancy J. Aronoff, MS, RD Judith E. Soberman, MD Jill P. Crandall, MD Beverly Williams-Cleaves, MD Sandra T. Foo, MD Carmen Pal, MD Northwestern University’s Feinberg School of Kathy Parkes, RN Medicine (Chicago, IL) Mary Beth Pena, RN Boyd E. Metzger, MD* Ellen S. Rooney, BA Mariana K. Johnson, MS, RN** Gretchen E.H. Van Wye, MA Catherine Behrends Kristine A. Viscovich, ANP Michelle Cook, MS Marian Fitzgibbon, PhD Indiana University (Indianapolis, IN) Mimi M. Giles, MS, RD David G. Marrero, PhD* Deloris Heard, MA Melvin J. Prince, MD* Cheryl K.H. Johnson, MS, RN Susie M. Kelly, RN, CDE** Diane Larsen, BS Yolanda F. Dotson, BS Anne Lowe, BS Edwin S. Fineberg, MD Megan Lyman, BS John C. Guare, PhD David McPherson, MD Angela M. Hadden Mark E. Molitch, MD James M. Ignaut, MA Thomas Pitts, MD Marcia L. Jackson Renee Reinhart, RN, MS Marian S. Kirkman, MD Susan Roston, RN, RD Kieren J. Mather, MD Beverly D. Porter, MSN Page 41 of 58 Diabetes

Paris J. Roach, MD Frederick L. Brancati, MD, MHS Nancy D. Rowland, BS, MS Jeanne M. Clark, MD Madelyn L. Wheeler, RD Jeanne B. Charleston, RN, MSN Janice Freel Medstar Research Institute (Washington, DC) Katherine Horak, RD Robert E. Ratner, MD* Dawn Jiggetts Gretchen Youssef, RD, CDE** Deloris Johnson Sue Shapiro, RN, BSN, CCRC** Hope Joseph Catherine Bavido-Arrage, MS, RD, LD Kimberly Loman Geraldine Boggs, MSN, RN Henry Mosley Marjorie Bronsord, MS, RD, CDE Richard R. Rubin, PhD Ernestine Brown Alafia Samuels, MD Wayman W. Cheatham, MD Kerry J. Stewart, EdD Susan Cola Paula Williamson Cindy Evans Peggy Gibbs University of New Mexico (Albuquerque, NM) Tracy Kellum, MS, RD, CDE David S. Schade, MD* Claresa Levatan, MD Karwyn S. Adams, RN, MSN** Asha K. Nair, BS Carolyn Johannes, RN, CDE** Maureen Passaro, MD Leslie F. Atler, PhD Gabriel Uwaifo, MD Patrick J. Boyle, MD Mark R. Burge, MD University of Southern California/UCLA Research Janene L. Canady, RN, CDE Center (Alhambra, CA) Lisa Chai, RN Mohammed F. Saad, MD* Ysela Gonzales, RN, MSN Maria Budget** Doris A. Hernandez-McGinnis Sujata Jinagouda, MD** Patricia Katz, LPN Khan Akbar, MD Carolyn King Claudia Conzues Amer Rassam, MD Perpetua Magpuri Sofya Rubinchik, MD Kathy Ngo Willette Senter, RD Amer Rassam, MD Debra Waters, PhD Debra Waters Kathy Xapthalamous Albert Einstein College of Medicine (Bronx, NY) Harry Shamoon, MD* Washington University (St. Louis, MO) Janet O. Brown, RN, MPH, MSN** Julio V. Santiago, MD* (deceased) Elsie Adorno, BS Samuel Dagogo-Jack, MD, MSc, FRCP, FACP* Liane Cox, MS, RD Neil H. White, MD, CDE* Jill Crandall, MD Samia Das, MS, MBA, RD, LD** Helena Duffy, MS, C-ANP Ana Santiago, RD** Samuel Engel, MD Angela Brown, MD Allison Friedler, BS Edwin Fisher, PhD Crystal J. Howard-Century, MA Emma Hurt, RN Stacey Kloiber, RN Tracy Jones, RN Nadege Longchamp, LPN Michelle Kerr, RD Helen Martinez, RN, MSN, FNP-C Lucy Ryder, RN Dorothy Pompi, BA Cormarie Wernimont, MS, RD Jonathan Scheindlin, MD Elissa Violino, RD, MS Johns Hopkins School of Medicine (Baltimore, Elizabeth Walker, RN, DNSc, CDE MD) Judith Wylie-Rosett, EdD, RD Christopher D. Saudek, MD* Elise Zimmerman, RD, MS Vanessa Bradley, BA** Joel Zonszein, MD Emily Sullivan, MEd, RN** Tracy Whittington, BS** University of Pittsburgh (Pittsburgh, PA) Caroline Abbas Trevor Orchard, MD* Diabetes Page 42 of 58

Rena R. Wing, PhD* Cyndy Edgerton, RD Gaye Koenning, MS, RD** Jacqueline M. Ghahate M. Kaye Kramer, BSN, MPH** Justin Glass, MD Susan Barr, BS Martia Glass, MD Miriam Boraz Dorothy Gohdes, MD Lisa Clifford, BS Wendy Grant, MD Rebecca Culyba, BS Robert L. Hanson, MD, MPH Marlene Frazier Ellie Horse Ryan Gilligan, BS Louise E. Ingraham, MS, RD, LN Susan Harrier, MLT Merry Jackson Louann Harris, RN Priscilla Jay Susan Jeffries, RN, MSN Roylen S. Kaskalla Andrea Kriska, PhD David Kessler, MD Qurashia Manjoo, MD Kathleen M. Kobus, RNC-ANP Monica Mullen, MHP, RD Jonathan Krakoff, MD Alicia Noel, BS Catherine Manus, LPN Amy Otto, PhD Sara Michaels, MD Linda Semler, MS, RD Tina Morgan Cheryl F. Smith, PhD Yolanda Nashboo (deceased) Marie Smith, RN, BSN Julie A. Nelson, RD Elizabeth Venditti, PhD Steven Poirier, MD Valarie Weinzierl, BS Evette Polczynski, MD Katherine V. Williams, MD, MPH Mike Reidy, MD Tara Wilson, BA Jeanine Roumain, MD, MPH Debra Rowse, MD University of Hawaii (Honolulu, HI) Sandra Sangster Richard F. Arakaki, MD* Janet Sewenemewa Renee W. Latimer, BSN, MPH** Darryl Tonemah, PhD Narleen K. Baker-Ladao, BS Charlton Wilson, MD Ralph Beddow, MD Michelle Yazzie Lorna Dias, AA Jillian Inouye, RN, PhD George Washington University Biostatistics Center Marjorie K. Mau, MD (DPP Coordinating Center Rockville, MD) Kathy Mikami, BS, RD Raymond Bain, PhD* Pharis Mohideen, MD Sarah Fowler, PhD* Sharon K. Odom, RD, MPH Tina Brenneman** Raynette U. Perry, AA Solome Abebe Julie Bamdad, MS Southwest American Indian Centers (Phoenix, AZ; Jackie Callaghan Shiprock, NM; Zuni, NM) Sharon L. Edelstein, ScM William C. Knowler, MD, DRPH*+ Mary Foulkes, PhD Norman Cooeyate** Yuping Gao Mary A. Hoskin, RD, MS** Kristina L. Grimes Carol A. Percy, RN, MS** Nisha Grover Kelly J. Acton, MD, MPH Lori Haffner, MS Vickie L. Andre, RN, FNP Steve Jones Rosalyn Barber Tara L. Jones Shandiin Begay, MPH Richard Katz, MD Peter H. Bennett, MB, FRCP John M. Lachin, ScD Mary Beth Benson, RN, BSN Pamela Mucik Evelyn C. Bird, RD, MPH Robert Orlosky Brenda A. Broussard, RD, MPH, MBA, CDE James Rochon, PhD Marcella Chavez, RN, AS Alla Sapozhnikova Tara Dacawyma Hanna Sherif, MS Matthew S. Doughty, MD Charlotte Stimpson Roberta Duncan, RD Page 43 of 58 Diabetes

Marinella Temprosa, PhD Robert L Hanson, MD, MPH3 Fredricka Walker-Murray Willian C. Knowler, MD, DrPH3 Toni I. Pollin, PhD4 Central Biochemistry Laboratory (Seattle, WA) Alan R. Shuldiner, MD4 Santica Marcovina, PhD, ScD* Kathleen Jablonski, PhD5 Greg Strylewicz, PhD** Paul W. Franks, PhD, MPhil, MS6,7,8 F. Alan Aldrich Marie-France Hivert, MD9

Carotid Ultrasound 1=Massachusetts General Hospital Dan O’Leary, MD* 2=Broad Institute 3=NIDDK CT Scan Reading Center 4=University of Maryland Elizabeth Stamm, MD* 5=Coordinating Center 6=Lund University, Sweden Epidemiological Cardiology Research Center- 7=Umeå University, Sweden Epicare (Winston-Salem, NC) 8=Harvard School of Public Health Pentti Rautaharju, MD, PhD* 9=Université de Sherbrooke Ronald J. Prineas, MD, PhD*/* Teresa Alexander Charles Campbell, MS Sharon Hall Yabing Li, MD Margaret Mills Nancy Pemberton, MS Farida Rautaharju, PhD Zhuming Zhang, MD

Nutrition Coding Center (Columbia, SC) Elizabeth Mayer-Davis, PhD* Robert R. Moran, PhD**

Quality of Well-Being Center (La Jolla, CA) Ted Ganiats, MD* Kristin David, MHP* Andrew J. Sarkin, PhD* Erik Groessl, PhD

NIH/NIDDK (Bethesda, MD) R. Eastman, MD Judith Fradkin, MD Sanford Garfield, PhD

Centers for Disease Control & Prevention (Atlanta, GA) Edward Gregg, PhD Ping Zhang, PhD

University of Michigan (Ann Arbor, MI) William H. Herman, MD, MPH

+Genetics Working Group Jose C. Florez, MD, PhD1,2 David Altshuler, MD, PhD1,2 Liana K. Billings, MD1 Ling Chen, MS1 Maegan Harden, BS2 Diabetes Page 44 of 58

Supplementary table 1. Characteristics of genetic variants used to build the coronary artery disease genetic risk score.

Note: List of 202 independent CAD SNPs used to build the coronary artery disease genetic risk score. Relative effect size on CAD obtained by the literature-based estimates (Van der Harst P. et al., Circ Research 2018). From the original list, a close proxy (LD=1), rs149076167 (22193768) was substituted for rs6984210 (8:22033615_C_G). # SNPs and proxies were not available in the DPP GWAS dataset (rs7797644, rs9365196, rs9457995). Chr, chromosome; SNP, single nucleotide polymorphism; EA, effect allele associated with coronary artery disease; EAF, effect allele frequency; OR, odds ratio; CAD, coronary artery disease; T2D, Position based on hg19. r2 denotes the quality of imputation for variants non-directly genotyped.

CHR BP (Hg19) MarkerName rsID 1 2252205 1:2252205_C_T rs36096196 1 3325912 1:3325912_A_C rs2493298 1 38461319 1:38461319_A_C rs61776719 1 55496039 1:55496039_C_T rs11206510 1 55505647 1:55505647_G_T rs11591147 1 56966350 1:56966350_A_G rs17114046 1 56986303 1:56986303_A_G rs147055617 1 57020650 1:57020650_C_G rs17416285 1 109821511 1:109821511_G_T rs602633 1 115753482 1:115753482_A_G rs11806316 1 151762308 1:151762308_C_G rs11810571 1 154422067 1:154422067_C_T rs4845625 1 169094459 1:169094459_C_T rs1892094 1 200646073 1:200646073_C_T rs6700559 1 201872264 1:201872264_C_T rs2820315 1 210468999 1:210468999_C_T rs60154123 1 222823529 1:222823529_A_C rs17465637 1 230845794 1:230845794_A_G rs699 2 19942473 2:19942473_A_G rs16986953 2 21291529 2:21291529_A_G rs668948 2 44073881 2:44073881_C_T rs6544713 2 44081627 2:44081627_G_T rs4076834 2 45896437 2:45896437_A_C rs582384 2 85809989 2:85809989_C_T rs1561198 2 145270592 2:145270592_A_G rs6740731 2 145286559 2:145286559_G_T rs17678683 2 145801461 2:145801461_C_T rs2252641 2 164957251 2:164957251_A_G rs12999907 2 188196469 2:188196469_C_T rs840616 2 203893999 2:203893999_A_C rs115654617 2 216291359 2:216291359_C_T rs17517928 2 216304384 2:216304384_C_T rs1250229 2 218669225 2:218669225_C_T rs61741262 2 218683154 2:218683154_A_G rs2571445 2 227100698 2:227100698_G_T rs2972146 2 233633460 2:233633460_A_G rs1801251 2 238223955 2:238223955_A_G rs11677932 Page 45 of 58 Diabetes

3 14901525 3:14901525_C_G rs13079221 3 46688562 3:46688562_A_G rs7633770 3 48193515 3:48193515_C_T rs7617773 3 49448566 3:49448566_A_C rs7623687 3 124450081 3:124450081_C_G rs4678145 3 132257961 3:132257961_G_T rs10512861 3 136069472 3:136069472_G_T rs667920 3 138092889 3:138092889_C_T rs185244 3 153839866 3:153839866_C_G rs12493885 3 156852592 3:156852592_C_G rs4266144 3 172115902 3:172115902_A_G rs12897 4 3449652 4:3449652_A_G rs16844401 4 57838583 4:57838583_G_T rs17087335 4 77416627 4:77416627_A_G rs12500824 4 81181072 4:81181072_A_T rs10857147 4 82587050 4:82587050_A_G rs11099493 4 96117371 4:96117371_A_T rs3775058 4 120909501 4:120909501_A_C rs7678555 4 146782837 4:146782837_A_C rs35879803 4 148281001 4:148281001_C_G rs4593108 4 148400819 4:148400819_A_C rs6842241 4 156436517 4:156436517_A_C rs13118820 4 156635309 4:156635309_A_G rs7692387 4 169687725 4:169687725_G_T rs7696431 5 9556694 5:9556694_C_T rs1508798 5 55860781 5:55860781_A_G rs3936511 5 121413208 5:121413208_C_T rs1800449 5 131667353 5:131667353_A_G rs273909 5 142516897 5:142516897_C_T rs246600 6 1617143 6:1617143_C_T rs9501744 6 12756658 6:12756658_A_T rs1412748 6 12903957 6:12903957_A_G rs9349379 6 22598259 6:22598259_C_T rs7766436 6 31919578 6:31919578_A_G rs2072633 6 34618893 6:34618893_A_G rs2814993 6 35034800 6:35034800_C_G rs17609940 6 36638636 6:36638636_A_G rs1321309 6 39174922 6:39174922_C_T rs10947789 6 43758873 6:43758873_A_G rs6905288 6 57160572 6:57160572_G_T rs9367716 6 82612271 6:82612271_A_C rs4613862 6 126717064 6:126717064_A_G rs1591805 6 134209837 6:134209837_C_T rs2327429 6 134214227 6:134214227_A_G rs2327433 6 150997401 6:150997401_C_T rs17080091 6 160465291 6:160465291_A_G rs688359 Diabetes Page 46 of 58

6 160679400 6:160679400_A_C rs624249 6 160863532 6:160863532_C_T rs2048327 6 160911596 6:160911596_A_G rs147555597 6 161005610 6:161005610_C_T rs55730499 6 161111700 6:161111700_C_T rs186696265 6 161013013 6:161013013_C_T rs1405708866 6 161143608 6:161143608_C_T rs4252120 7 1937261 7:1937261_A_G rs10267593 7 12261911 7:12261911_A_G rs11509880 7 19049388 7:19049388_A_G rs2107595 7 45077978 7:45077978_A_G rs2107732 7 107244545 7:107244545_C_T rs10953541 7 117332914 7:117332914_A_G rs975722 7 129663496 7:129663496_C_T rs11556924 7 139757136 7:139757136_G_T rs10237377 7 150690176 7:150690176_C_T rs3918226 8 18286997 8:18286997_C_T rs6997340 8 19800529 8:19800529_C_G rs6997330 8 19824667 8:19824667_C_T rs15285 8 22033615 8:22033615_C_G rs6984210 8 106565414 8:106565414_A_G rs10093110 8 126490972 8:126490972_A_T rs2954029 9 21706571 9:21706571_A_G rs896655 9 21970916 9:21970916_C_T rs3731249 9 22062012 9:22062012_A_C rs4977754 9 22073996 9:22073996_G_T rs1855185 9 22098619 9:22098619_A_G rs2891168 9 22113324 9:22113324_C_G rs13301964 9 110517794 9:110517794_C_T rs944172 9 113169775 9:113169775_C_T rs111245230 9 124420173 9:124420173_C_T rs885150 9 136149399 9:136149399_A_G rs507666 10 12303813 10:12303813_C_T rs61848342 10 30317073 10:30317073_A_G rs9337951 10 44480811 10:44480811_G_T rs1870634 10 44777560 10:44777560_C_G rs1657346 10 82251514 10:82251514_C_T rs17680741 10 91004886 10:91004886_A_G rs2246942 10 104638480 10:104638480_C_T rs3740390 10 105693644 10:105693644_A_G rs4918072 10 124237612 10:124237612_A_G rs4752700 11 5701074 11:5701074_A_C rs11601507 11 9751196 11:9751196_A_G rs10840293 11 10745394 11:10745394_G_T rs11042937 11 13301548 11:13301548_A_T rs1351525 11 43696917 11:43696917_G_T rs7116641 Page 47 of 58 Diabetes

11 65391317 11:65391317_A_G rs12801636 11 75274150 11:75274150_G_T rs590121 11 75284334 11:75284334_G_T rs659418 11 100624599 11:100624599_A_G rs7947761 11 103660567 11:103660567_C_T rs974819 11 116648917 11:116648917_C_G rs964184 12 7175872 12:7175872_C_T rs11838267 12 20220033 12:20220033_C_G rs10841443 12 54513915 12:54513915_C_G rs11170820 12 57527283 12:57527283_C_T rs11172113 12 90013089 12:90013089_C_T rs2681492 12 95355541 12:95355541_A_G rs7306455 12 111884608 12:111884608_C_T rs3184504 12 118265441 12:118265441_G_T rs11830157 12 121416988 12:121416988_A_G rs2244608 12 124427306 12:124427306_A_T rs11057401 12 125307053 12:125307053_A_G rs11057830 13 28973621 13:28973621_A_G rs9319428 13 33058333 13:33058333_A_G rs9591012 13 110818102 13:110818102_A_T rs11617955 13 110960943 13:110960943_A_G rs3809346 13 111040681 13:111040681_A_G rs11838776 13 111049623 13:111049623_C_T rs9515203 13 113631780 13:113631780_A_C rs1317507 14 58794001 14:58794001_A_G rs2145598 14 75446879 14:75446879_C_G rs10131894 14 94838142 14:94838142_G_T rs112635299 14 100133942 14:100133942_C_T rs2895811 14 100148961 14:100148961_C_T rs8003602 15 65024204 15:65024204_A_G rs6494488 15 67450305 15:67450305_A_G rs17228058 15 79139000 15:79139000_C_T rs11637783 15 79117133 15:79117133_G_T rs6495335 15 89574218 15:89574218_A_G rs8042271 15 91416550 15:91416550_A_C rs17514846 15 96146414 15:96146414_A_C rs17581137 16 56995236 16:56995236_A_C rs1800775 16 72130815 16:72130815_A_C rs1050362 16 75462055 16:75462055_A_G rs12930452 16 81906423 16:81906423_A_G rs7199941 16 83045790 16:83045790_A_G rs7500448 17 2126504 17:2126504_C_G rs216172 17 2170216 17:2170216_C_T rs170041 17 17543722 17:17543722_A_G rs12936587 17 27941886 17:27941886_A_G rs13723 17 30033514 17:30033514_C_T rs76954792 Diabetes Page 48 of 58

17 40257163 17:40257163_C_T rs2074158 17 45013271 17:45013271_C_T rs17608766 17 47047868 17:47047868_A_G rs3895874 17 47440466 17:47440466_A_G rs16948048 17 62387091 17:62387091_C_T rs1867624 18 47229717 18:47229717_A_C rs9964304 18 57838401 18:57838401_A_G rs663129 19 8429323 19:8429323_A_G rs116843064 19 11202306 19:11202306_G_T rs6511720 19 11277232 19:11277232_A_G rs4804573 19 17855763 19:17855763_C_G rs73015714 19 32882020 19:32882020_A_T rs12976411 19 41832231 19:41832231_A_G rs12980942 19 41825191 19:41825191_C_T rs73045269 19 41851509 19:41851509_A_C rs4803455 19 45412079 19:45412079_C_T rs7412 19 45422946 19:45422946_A_G rs4420638 19 46190268 19:46190268_A_G rs1964272 20 33313566 20:33313566_C_T rs6088590 20 33764554 20:33764554_A_G rs867186 20 39924279 20:39924279_A_G rs6102343 20 44586023 20:44586023_C_T rs3827066 20 57714025 20:57714025_C_T rs260020 21 30533076 21:30533076_A_G rs2832227 21 35593827 21:35593827_A_G rs28451064 22 24763476 22:24763476_C_T rs5760368 Page 49 of 58 Diabetes

Supplementary table 1. Characteristics of genetic variants used to build the coronary artery disease genetic risk score.

Note: List of 202 independent CAD SNPs used to build the coronary artery disease genetic risk score. Relative effect size on CAD obtained by the literature-based estimates (Van der Harst P. et al., Circ Research 2018). From the original list, a close proxy (LD=1), rs149076167 (22193768) was substituted for rs6984210 (8:22033615_C_G). # SNPs and proxies were not available in the DPP GWAS dataset (rs7797644, rs9365196, rs9457995). Chr, chromosome; SNP, single nucleotide polymorphism; EA, effect allele associated with coronary artery disease; EAF, effect allele frequency; OR, odds ratio; CAD, coronary artery disease; T2D, Position based on hg19. r2 denotes the quality of imputation for variants non-directly genotyped.

locus Effect Allele Non-Effect Allele Freq EA Imputed (r2) MORN1 T C 0.1483 N PRDM16 A C 0.1355 N FHL3 A C 0.5325 N PCSK9 T C 0.8209 N PCSK9 G T 0.0164 Y (1.00) PLPP3(PPAP2B) A G 0.9062 Y (1.00) PLPP3(PPAP2B) G A 0.9173 Y (1.00) PLPP3(PPAP2B) G C 0.8906 Y (1.00) PSRC1(SORT1) G T 0.2343 N NGF G A 0.3668 N TDRKH G C 0.1898 N IL6R T C 0.4338 N ATP1B1 C T 0.4956 N DDX59,CAMSAP2 C T 0.4715 N LMOD1 T C 0.3012 N HHAT T C 0.1501 N MIA3 C A 0.3135 N AGT G A 0.561 N OSR1(AK097927) A G 0.1095 N APOB A G 0.8142 N ABCG8,ABCG5 T C 0.3038 Y (1.00) ABCG8,ABCG5 T G 0.9337 Y (1.00) PRKCE A C 0.5251 N VAMP8,VAMP5 T C 0.4583 N ZEB2,TEX41 A G 0.174 N ZEB2,TEX41 G T 0.9105 N ZEB2,TEX41 C T 0.5299 N FIGN A G 0.8189 N CALCRL C T 0.3609 N WDR12,NBEAL1 A C 0.1168 N FN1 C T 0.2515 N FN1 T C 0.2722 N TNS1 T C 0.8917 N TNS1 A G 0.3895 N LOC646736 T G 0.6526 N KCNJ13,GIGYF2 A G 0.3645 Y (1.00) COL6A3 G A 0.3157 Y (1.00) Diabetes Page 50 of 58

FGD5 C G 0.4306 Y (1.00) SNORD77,ALS2CL A G 0.4061 Y (1.00) CDC25A T C 0.6693 Y (1.00) RHOA A C 0.858 Y (1.00) UMPS,ITGB5 C G 0.1284 Y (1.00) DNAJC13 G T 0.1433 Y (1.00) STAG1 T G 0.7751 Y (1.00) MRAS T C 0.1625 Y (1.00) ARHGEF26 C G 0.864 Y (1.00) CCNL1 G C 0.6774 Y (1.00) FNDC3B G A 0.5914 Y (0.99) HGFAC,RGS12 A G 0.0695 Y (0.99) REST,NOA1 T G 0.2043 Y (0.99) SHROOM3 A G 0.3565 Y (0.99) FGF5 T A 0.7208 Y (0.99) HNRNPD A G 0.6924 Y (0.99) UNC5C A T 0.2341 Y (0.97) MAD2L1 C A 0.7166 Y (0.97) ZNF827 C A 0.329 Y (0.94) EDNRA C G 0.8035 Y (0.94) EDNRA A C 0.1641 Y (0.99) GUCY1A3,MAP9 A C 0.3606 Y (0.99) GUCY1A3,MAP9 G A 0.1892 Y (0.99) PALLD T G 0.5098 Y (0.99) SEMA5A T C 0.8134 Y (0.99) LOC101928448 G A 0.823 Y (0.99) LOX T C 0.1728 Y (0.99) SLC22A4-SLC22A5 G A 0.8754 Y (0.99) ARHGAP26 T C 0.4633 Y (0.99) FOXC1 C T 0.1309 Y (0.99) PHACTR1 T A 0.4619 Y (1.00) PHACTR1 G A 0.5907 Y (1.00) HDGFL1 T C 0.2903 Y (0.99) C2 G A 0.4468 Y (0.99) ANKS1A,C6orf16 A G 0.1445 Y (0.86) ANKS1A,C6orf16 G C 0.1972 Y (0.86) CDKN1A,PANDAR A G 0.4907 Y (1.00) KCNK5 T C 0.7671 Y (1.00) VEGFA A G 0.5807 N PRIM2 G T 0.3155 N RP11-379B8.1 A C 0.5314 N CENPW A G 0.4927 N TCF21 T C 0.695 N TCF21 G A 0.9033 N PLEKHG1 C T 0.0839 N IGF2R G A 0.6127 N Page 51 of 58 Diabetes

LPA,PLG,LPAL2,SLC22A3 C A 0.3855 N LPA,PLG,LPAL2,SLC22A3 C T 0.6295 N LPA,PLG,LPAL2,SLC22A3 A G 0.0088 N LPA,PLG,LPAL2,SLC22A3 T C 0.0693 N LPA,PLG,LPAL2,SLC22A3 T C 0.0131 N LPA,PLG,LPAL2,SLC22A3 C T 0.9833 N LPA,PLG,LPAL2,SLC22A3 T C 0.7203 Y (1.00) MAD1L1 G A 0.2012 Y (1.00) TMEM106B A G 0.3612 N HDAC9 A G 0.1767 N CCM2 G A 0.0874 N 7q22(BCAP29) C T 0.233 N CFTR,CCTNBP2 G A 0.595 N ZC3HC1 C T 0.3624 N PARP12 G T 0.3608 N NOS3 T C 0.0732 N NAT2 T C 0.3025 N LPL C G 0.0403 N LPL C T 0.2651 N BMP1 G C 0.939 N ZFPM2 G A 0.4167 N TRIB1 A T 0.5402 N CDKN2B,CDKN2A A G 0.5975 N CDKN2B,CDKN2A T C 0.027 N CDKN2B,CDKN2A A C 0.9502 N CDKN2B,CDKN2A G T 0.0797 N CDKN2B,CDKN2A G A 0.5243 N CDKN2B,CDKN2A G C 0.9798 N KLF4 C T 0.7197 N SVEP1 C T 0.9666 N DAB2IP C T 0.7304 Y (0.99) ABO A G 0.1847 Y (0.94) CDC123 C T 0.6355 Y (0.93) KIAA1462 A G 0.3178 Y (0.93) CXCL12 G T 0.3511 Y (1.00) CXCL12 G C 0.1176 Y (0.99) TSPAN14 T C 0.7152 Y (0.99) LIPA G A 0.64 Y (0.96) AS3MT,CYP17A1,CNNM2 C T 0.1066 Y (1.00) STN1 A G 0.2713 Y (1.00) HTRA1 G A 0.5508 Y (0.99) TRIM5,TRIM22 A C 0.0739 Y (1.00) SWAP70 A G 0.5527 Y (1.00) MRVI1,CTR9 T G 0.5121 Y (0.92) ARNTL T A 0.3101 Y (0.88) HSD17B12 G T 0.6877 N Diabetes Page 52 of 58

PCNX3 G A 0.2364 N SERPINH1 T G 0.3006 N SERPINH1 T G 0.8477 N ARHGAP42 G A 0.7214 N PDGFD,DYNC2H1 T C 0.3082 N APOA1-A5-A4-C3,ZNF259 G C 0.8434 N C1S T C 0.8679 N LOC156393 G C 0.3409 N HOXC4 G C 0.9292 N LRP1 C T 0.5903 N ATP2B1 C T 0.8141 N NDUFA12 G A 0.1018 N SH2B3,ATXN2,HNF1A T C 0.4445 N KSR2 G T 0.6156 N HNF1A G A 0.6703 N SCARB1,CCDC92 T A 0.3021 N SCARB1,CCDC92 A G 0.1465 N FLT1 A G 0.3121 N N4BP2L2 G A 0.3413 N COL4A1/A2 T A 0.1164 N COL4A1/A2 A G 0.4275 N COL4A1/A2 A G 0.2705 N COL4A1/A2 T C 0.7392 N MCF2L A C 0.2608 N ARID4A G A 0.5758 N TMED10 G C 0.4528 N SERPINA1,SERPINA2 G T 0.0193 N HHIPL1,CYP46A1 C T 0.5834 N HHIPL1,CYP46A1 C T 0.2681 N OAZ2,RBPMS2 A G 0.8117 N SMAD3 A G 0.779 N ADAMTS7 T C 0.5698 N ADAMTS7 T G 0.6026 N MFGE8-ABHD2 G A 0.0745 N FURIN A C 0.4605 N LINC00924 (15q26.2) A C 0.754 N CETP C A 0.4989 N DHX38,TXNL4B A C 0.3755 N CFDP1 G A 0.4202 N PLCG2 A G 0.3982 N CDH13 A G 0.7594 N SMG6,SRR C G 0.3507 N SMG6,SRR C T 0.2943 N RASD1, SMCR3, PEMT G A 0.4379 N CORO6,ANKRD13B G A 0.5066 N (17q11.2) T C 0.2197 N Page 53 of 58 Diabetes

DHX58,KAT2A C T 0.8152 N GOSR2 C T 0.8632 N UBE2Z,GIP A G 0.5553 N UBE2Z,ZNF652 G A 0.645 N PECAM1 T C 0.6209 N ACAA2 C A 0.716 N PMAIP1,MC4R A G 0.2471 N ANGPTL4 G A 0.0202 N LDLR G T 0.1136 N LDLR G A 0.4676 N MAP1S,FCHO1 G C 0.794 N ZNF507,LOC400684 A T 0.9408 N TGFB1,CCDC97 G A 0.1583 N TGFB1,CCDC97 T C 0.1545 N TGFB1,CCDC97 C A 0.4954 N APOE,APOC1,TOMM4 C T 0.079 N APOE,APOC1,TOMM4 G A 0.8191 N SNRPD2 G A 0.4807 N NCOA6 C T 0.6143 N PROCR A G 0.8971 N ZHX3 A G 0.2467 N PCIF1,ZNF335 T C 0.1437 N ZNF831 T C 0.1309 N MAP3K7CL G A 0.8258 N MRPS6 A G 0.1239 N POM121L9P,ADORA2A T C 0.9738 N Diabetes Page 54 of 58

Note: List of 202 independent CAD SNPs used to build the coronary artery disease genetic risk score. Relative effect size on CAD obtained by the literature-based estimates (Van der Harst P. et al., Circ Research 2018). From the original list, a close proxy (LD=1), rs149076167 (22193768) was substituted for rs6984210 (8:22033615_C_G). # SNPs and proxies were not available in the DPP GWAS dataset (rs7797644, rs9365196, rs9457995). Chr, chromosome; SNP, single nucleotide polymorphism; EA, effect allele associated with coronary artery disease; EAF, effect allele frequency; OR, odds ratio; CAD, coronary artery disease; T2D, Position based on hg19. r2 denotes the quality of imputation for variants non-directly genotyped.

OR CAD (95%CI) P-value 1.05 1.03 1.06 1.33E-08 1.05 1.04 1.07 9.97E-10 1.04 1.03 1.05 7.67E-10 1.04 1.03 1.06 7.31E-10 1.27 1.21 1.34 1.86E-22 1.10 1.08 1.12 8.04E-28 1.08 1.06 1.11 5.73E-14 1.06 1.05 1.08 5.26E-13 1.11 1.09 1.12 3.63E-58 1.03 1.02 1.05 4.16E-10 1.05 1.03 1.06 3.24E-11 1.04 1.03 1.05 6.69E-15 1.04 1.03 1.05 3.14E-12 1.03 1.02 1.04 1.77E-08 1.04 1.03 1.05 1.10E-10 1.05 1.03 1.06 2.58E-08 1.07 1.06 1.09 8.21E-28 1.02 1.01 1.04 1.85E-05 1.08 1.06 1.11 1.06E-16 1.06 1.04 1.07 7.22E-17 1.05 1.04 1.06 1.84E-18 1.08 1.06 1.10 2.78E-11 1.03 1.02 1.05 7.64E-09 1.05 1.04 1.06 1.90E-23 1.05 1.03 1.06 2.77E-09 1.05 1.03 1.07 4.04E-07 1.04 1.03 1.05 5.41E-13 1.05 1.03 1.06 6.31E-11 1.03 1.02 1.04 7.78E-07 1.11 1.09 1.13 2.20E-32 1.04 1.03 1.06 1.46E-10 1.04 1.03 1.06 1.58E-13 1.06 1.04 1.08 2.03E-08 1.04 1.03 1.05 1.56E-12 1.03 1.02 1.04 1.94E-07 1.03 1.02 1.04 2.13E-06 1.03 1.02 1.05 2.64E-08 Page 55 of 58 Diabetes

1.04 1.03 1.05 1.02E-10 1.03 1.02 1.04 1.10E-08 1.04 1.03 1.05 3.48E-11 1.05 1.03 1.06 5.46E-08 1.07 1.05 1.08 1.45E-14 1.04 1.03 1.06 3.00E-09 1.05 1.04 1.06 6.08E-15 1.07 1.05 1.08 2.40E-17 1.05 1.04 1.07 3.95E-11 1.04 1.02 1.05 1.40E-08 1.04 1.02 1.05 1.21E-09 1.04 1.02 1.06 0.0006748 1.03 1.02 1.05 3.75E-07 1.03 1.02 1.04 1.64E-10 1.05 1.04 1.06 1.37E-14 1.04 1.03 1.05 2.49E-10 1.04 1.03 1.05 7.63E-09 1.05 1.04 1.06 1.23E-14 1.03 1.01 1.04 1.09E-05 1.06 1.04 1.07 8.93E-15 1.08 1.06 1.09 6.95E-24 1.04 1.03 1.05 3.29E-10 1.07 1.05 1.08 3.34E-23 1.03 1.02 1.04 3.09E-08 1.05 1.03 1.06 7.64E-13 1.04 1.02 1.05 3.74E-08 1.03 1.02 1.05 5.41E-05 1.05 1.03 1.07 4.74E-10 1.04 1.03 1.05 6.53E-17 1.05 1.03 1.06 8.42E-08 1.04 1.02 1.05 6.74E-09 1.11 1.10 1.13 2.72E-76 1.05 1.03 1.06 7.93E-13 1.04 1.03 1.05 3.88E-12 1.05 1.03 1.06 3.74E-10 1.02 1.01 1.04 0.002218 1.03 1.02 1.04 3.48E-08 1.04 1.03 1.05 7.63E-12 1.04 1.03 1.05 4.00E-12 1.03 1.02 1.04 1.20E-08 1.03 1.02 1.04 6.58E-10 1.04 1.03 1.05 1.17E-10 1.07 1.06 1.08 3.56E-31 1.05 1.03 1.07 8.84E-08 1.06 1.04 1.07 6.09E-09 1.05 1.04 1.06 7.07E-15 Diabetes Page 56 of 58

1.05 1.04 1.07 3.32E-18 1.07 1.06 1.08 1.58E-34 1.47 1.38 1.57 1.74E-32 1.37 1.34 1.40 9.78E-154 1.73 1.64 1.82 1.99E-97 1.61 1.54 1.69 2.06E-91 1.05 1.04 1.06 1.71E-18 1.04 1.02 1.05 1.88E-08 1.03 1.02 1.05 3.27E-08 1.08 1.06 1.09 1.25E-24 1.06 1.04 1.08 3.64E-08 1.03 1.02 1.04 4.47E-06 1.03 1.02 1.04 4.17E-08 1.06 1.05 1.07 1.37E-23 1.03 1.02 1.05 6.53E-09 1.11 1.09 1.14 1.35E-20 1.03 1.02 1.05 4.60E-09 1.13 1.09 1.16 3.87E-11 1.04 1.03 1.05 1.63E-12 1.08 1.06 1.11 1.04E-11 1.03 1.02 1.04 1.88E-08 1.05 1.04 1.06 1.68E-22 1.04 1.03 1.05 7.27E-12 1.15 1.11 1.19 2.37E-15 1.10 1.07 1.13 3.26E-12 1.11 1.09 1.14 1.35E-19 1.19 1.17 1.20 2.27E-207 1.19 1.14 1.25 2.28E-14 1.04 1.03 1.05 3.59E-12 1.06 1.03 1.09 0.0002208 1.04 1.02 1.05 7.86E-10 1.06 1.04 1.07 3.04E-14 1.04 1.03 1.05 6.38E-10 1.06 1.04 1.07 1.73E-17 1.06 1.05 1.07 3.50E-24 1.08 1.07 1.10 1.66E-23 1.04 1.03 1.06 1.72E-11 1.06 1.04 1.07 5.15E-20 1.07 1.05 1.09 4.67E-15 1.04 1.03 1.05 9.63E-10 1.03 1.02 1.04 8.02E-11 1.08 1.06 1.10 5.61E-13 1.04 1.03 1.05 6.90E-12 1.02 1.01 1.03 0.000397 1.04 1.02 1.05 1.28E-08 1.03 1.02 1.04 1.03E-08 Page 57 of 58 Diabetes

1.04 1.03 1.05 2.29E-11 1.04 1.02 1.05 1.49E-10 1.05 1.03 1.06 2.17E-08 1.04 1.03 1.05 3.57E-09 1.06 1.05 1.07 1.12E-28 1.06 1.04 1.07 1.15E-12 1.05 1.04 1.07 6.15E-10 1.05 1.03 1.06 2.86E-14 1.09 1.06 1.11 9.28E-13 1.02 1.01 1.03 1.04E-05 1.06 1.04 1.07 8.64E-15 1.05 1.03 1.07 5.89E-09 1.06 1.05 1.07 1.03E-25 1.01 1.00 1.02 0.03443 1.05 1.04 1.06 1.46E-18 1.03 1.01 1.04 4.03E-05 1.07 1.05 1.08 1.91E-18 1.04 1.03 1.05 5.36E-11 1.04 1.03 1.05 2.74E-10 1.08 1.06 1.10 6.94E-18 1.04 1.03 1.06 8.51E-17 1.04 1.03 1.06 1.47E-11 1.06 1.05 1.07 3.89E-23 1.04 1.03 1.05 8.21E-12 1.03 1.02 1.04 4.26E-08 1.03 1.02 1.05 4.81E-09 1.15 1.10 1.20 8.44E-10 1.04 1.03 1.05 5.44E-15 1.06 1.04 1.07 2.89E-16 1.04 1.02 1.05 3.90E-08 1.06 1.04 1.07 6.38E-17 1.07 1.06 1.08 1.81E-31 1.06 1.05 1.08 4.12E-30 1.06 1.04 1.09 3.58E-07 1.06 1.05 1.07 9.86E-27 1.04 1.02 1.05 1.21E-08 1.03 1.02 1.04 1.66E-06 1.04 1.03 1.05 2.92E-11 1.05 1.04 1.06 8.95E-16 1.04 1.03 1.05 9.36E-13 1.06 1.04 1.07 1.61E-16 1.04 1.03 1.05 2.24E-11 1.05 1.04 1.06 4.11E-17 1.03 1.02 1.04 9.51E-10 1.04 1.02 1.05 5.35E-10 1.04 1.03 1.05 1.19E-08 Diabetes Page 58 of 58

1.04 1.03 1.06 9.62E-08 1.05 1.03 1.06 8.20E-10 1.03 1.02 1.04 3.23E-08 1.04 1.03 1.05 5.23E-13 1.03 1.02 1.04 1.88E-10 1.04 1.03 1.05 1.14E-09 1.02 1.01 1.04 0.0001357 1.15 1.10 1.20 3.57E-10 1.12 1.10 1.14 2.52E-35 1.04 1.03 1.05 3.09E-10 1.05 1.04 1.07 2.24E-13 1.01 0.98 1.03 0.4549 1.06 1.04 1.07 2.51E-14 1.06 1.05 1.08 2.11E-14 1.05 1.04 1.06 2.36E-17 1.15 1.12 1.17 2.14E-35 1.08 1.07 1.10 1.98E-26 1.03 1.02 1.04 1.31E-08 1.03 1.02 1.04 2.18E-09 1.06 1.04 1.08 6.84E-12 1.04 1.02 1.05 1.12E-08 1.04 1.03 1.06 4.40E-09 1.05 1.04 1.07 7.96E-10 1.04 1.03 1.05 4.16E-09 1.11 1.09 1.13 2.58E-33 1.11 1.07 1.15 2.39E-08