To identify genetic risk factors for incident cardiovascular disease (CVD) among people with type 2 diabetes (T2D).
We conducted a multiancestry time-to-event genome-wide association study for incident CVD among people with T2D. We also tested 204 known coronary artery disease (CAD) variants for association with incident CVD.
Among 49,230 participants with T2D, 8,956 had incident CVD events (event rate 18.2%). We identified three novel genetic loci for incident CVD: rs147138607 (near CACNA1E/ZNF648, hazard ratio [HR] 1.23, P = 3.6 × 10−9), rs77142250 (near HS3ST1, HR 1.89, P = 9.9 × 10−9), and rs335407 (near TFB1M/NOX3, HR 1.25, P = 1.5 × 10−8). Among 204 known CAD loci, 5 were associated with incident CVD in T2D (multiple comparison–adjusted P < 0.00024, 0.05/204). A standardized polygenic score of these 204 variants was associated with incident CVD with HR 1.14 (P = 1.0 × 10−16).
The data point to novel and known genomic regions associated with incident CVD among individuals with T2D.
Introduction
Type 2 diabetes (T2D) is a significant risk factor for cardiovascular disease (CVD), leading to a two- to threefold higher likelihood of developing CVD. CVD is the leading cause of morbidity and mortality in people with diabetes (1,2), and previous studies suggest that life expectancy is reduced by up to 8 years in people with T2D (3). Although CVD mortality rates have declined substantially in the general population in recent decades, this improvement has been less substantial in people with T2D (4).
Beyond conventional risk factors, recent genome-wide association studies (GWAS) have identified at least 204 genetic loci associated with CVD in the general population. These studies have been mostly conducted among people with European ancestry (5), which calls for ancestry-diverse studies. Genetic risk factors of CVD in the general population are mostly thought to be relevant to people with T2D (6). Still, genetic variants of CVD in people with T2D have not been thoroughly investigated. Most studies have been underpowered and were cross-sectional (7,– 9). In this study, we performed a time-to-event GWAS of incident CVD in a large, multiancestry sample of people with T2D, ensuring that the occurrence of T2D preceded any CVD event.
Research Design and Methods
Study Design and Participating Cohorts
This is a meta-analysis of ancestry-specific, cohort-level, time-to-event GWAS for incident CVD in people with T2D, the majority of whom were from the Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) Consortium (10). Details of the methods and study participants can be found in Supplementary Methods. In brief, we studied 49,230 participants with T2D from 16 cohorts and of multiple ancestries. In the case of multiancestry cohorts, participants were grouped into major continental ancestries, resulting in 28 ancestry-specific subgroups (Supplementary Table 1).
Definition of T2D and CVD
T2D was defined for each cohort according to a participant having one or more of the American Diabetes Association criteria (11) (Supplementary Table 2). Participants with known type 1 diabetes or other specific types of diabetes were excluded. To minimize contamination with type 1 diabetes, we excluded people with age at diabetes diagnosis <40 years. CVD was defined as a composite of 1) coronary artery disease (CAD), 2) cerebrovascular disease, and 3) death from a cardiovascular cause (Supplementary Table 3 and Supplementary Methods). An incident CVD event was defined as the first CVD event occurring at least 1 year after T2D diagnosis.
Statistical Analyses
We applied Cox proportional hazards modeling for the time-to-event GWAS (Supplementary Methods). Each single nucleotide variation (SNV) was tested for its association with incident CVD with consideration of observation time and adjustment for covariates. Observation time was defined as years between age at diagnosis of T2D and age at incident CVD for cases or years between age at diagnosis of T2D and age at last follow-up. A time-to-event GWAS was performed for each ancestry-specific cohort subgroup with either the GWASTools R package (12) or the gwasurvivr R package (13) (Supplementary Table 4). The primary analysis included age at diagnosis of T2D and sex as covariates, and significant principal components were used to adjust for population stratification.
Cohort-Level Analysis and Meta-analysis
For each cohort there were specific preimputation genotype quality control criteria (Supplementary Table 5). Genotype imputation was performed with use of a population-specific reference panel. Within each cohort, analysis was performed separately for four major ancestry groups: Admixed African American (AFR), East Asian (EAS), European (EUR), and Admixed Hispanic (HIS). Meta-analysis of cohort-level summary statistics was conducted with an inverse variance–weighted fixed-effects method as implemented in METAL (14). A conventional genome-wide significance threshold was set as P < 5.0 × 10−8 (15,16). The methods for conducting downstream analysis on the significant variants can be found in Supplementary Methods.
Association of Known 204 CAD Variants
We tested 204 previously reported CAD variants identified in the general population for association with incident CVD in people with T2D (5,6). With adjustment for multiple comparisons, the association’s significance threshold was set as P < 0.00024 (0.05/204). In addition, a weighted polygenic score based on these 204 known CAD variants was constructed as previously described (6) and tested for association with incident CVD in people with T2D.
Results
Study Overview
A total of 49,230 people with T2D, who had not developed CVD either at the time of T2D diagnosis or within 1 year thereafter, were included in the analysis (Supplementary Table 6). Individuals with European ancestry comprised ∼63.2% (N = 31,118) of the participants, while the remaining 36.8% (N = 18,112) were of non-European ancestry (AFR 22.6%, n = 11,124; HIS 8.8%, 4,325; EAS 5.4%, 2,663). Among 49,230 participants with T2D, 8,956 developed incident CVD (event rate 18.2%) over a mean follow-up duration ranging from 3.2 to 33.7 years. Detailed clinical characteristics of the participants can be found in Supplementary Table 7.
Loci for Incident CVD in People With T2D
We tested 15,471,776 SNVs with overall minor allele frequency (MAF) ≥1% for association with incident CVD. A plot of expected-by-observed association statistics showed minimal inflation (λGC = 1.09 for variants with MAF ≥1%) (Supplementary Fig. 1A). We identified three SNVs associated with incident CVD in people with T2D at genome-wide significance (Supplementary Fig. 1B and Table 1). The variant rs147138607 (chromosome 1 [chr1]:181855562:G>C, MAF 10.7%) had a hazard ratio (HR) for incident CVD in T2D of 1.23 (95% CI 1.15–1.32, P = 3.6 × 10−9) and resides in an intergenic region between the genes CACNA1E and ZNF648 (Fig. 1A). The second most significant variant, rs77142250 (chr4:11444867:T>C, MAF 1.3%), was present at low frequency (1.3%) only in those of African ancestry, had an HR 1.89 (95% CI 1.52–2.35, P = 9.9 × 10−9), and resides near the gene HS3ST1 (Fig. 1B). The third variant, rs335407 (chr6:155665441:C>T, MAF 5.5%), had an HR of 1.25 (95% CI 1.16–1.35, P = 1.58 × 10−8) and resides in an intergenic region between the genes TFB1M and NOX3 (Fig. 1C). Results from the downstream analysis of these three variants can be found in Supplementary Results.
Genetic variants significantly associated with incident CVD in people with T2D in basic model
| chr | POS (rsID) | Effect allele | Ancestry | Frequency | HR (95% CI) | P | Het P | Sample size |
|---|---|---|---|---|---|---|---|---|
| 1 | 181855562 (rs147138607) | G>C | European/European American | 0.018 | 1.20 (1.00–1.44) | 0.047 | 0.756 | 24,457 |
| African American | 0.127 | 1.22 (1.12–1.33) | 2.3 × 10−6 | 0.381 | 8,929 | |||
| Hispanic/Latinx | 0.065 | 1.26 (1.03–1.55) | 0.027 | 0.198 | 3,163 | |||
| East Asian | 0.050 | 1.55 (1.07–2.25) | 0.021 | 0.469 | 2,511 | |||
| Combined | 0.107 | 1.23 (1.15–1.32) | 3.6 × 10−9 | 0.713 | 39,060 | |||
| 4 | 11444867 (rs77142250) | T>C | African American | 0.013 | 1.89 (1.52–2.35) | 9.9 × 10−9 | 0.363 | 9,748 |
| 6 | 155665441 (rs335407) | C>T | European/European American | 0.027 | 1.33 (1.19–1.50) | 1.4 × 10−3 | 0.664 | 29,910 |
| African American | 0.084 | 1.18 (1.05–1.31) | 3.8 × 10−3 | 0.891 | 7,765 | |||
| Hispanic/Latinx | 0.033 | 1.34 (0.99–1.81) | 0.055 | 0.596 | 3,163 | |||
| East Asian | 0.026 | 0.92 (0.45–1.88) | 0.810 | 0.541 | 2,511 | |||
| Combined | 0.055 | 1.25 (1.16–1.35) | 1.5 × 10−8 | 0.859 | 43,349 |
| chr | POS (rsID) | Effect allele | Ancestry | Frequency | HR (95% CI) | P | Het P | Sample size |
|---|---|---|---|---|---|---|---|---|
| 1 | 181855562 (rs147138607) | G>C | European/European American | 0.018 | 1.20 (1.00–1.44) | 0.047 | 0.756 | 24,457 |
| African American | 0.127 | 1.22 (1.12–1.33) | 2.3 × 10−6 | 0.381 | 8,929 | |||
| Hispanic/Latinx | 0.065 | 1.26 (1.03–1.55) | 0.027 | 0.198 | 3,163 | |||
| East Asian | 0.050 | 1.55 (1.07–2.25) | 0.021 | 0.469 | 2,511 | |||
| Combined | 0.107 | 1.23 (1.15–1.32) | 3.6 × 10−9 | 0.713 | 39,060 | |||
| 4 | 11444867 (rs77142250) | T>C | African American | 0.013 | 1.89 (1.52–2.35) | 9.9 × 10−9 | 0.363 | 9,748 |
| 6 | 155665441 (rs335407) | C>T | European/European American | 0.027 | 1.33 (1.19–1.50) | 1.4 × 10−3 | 0.664 | 29,910 |
| African American | 0.084 | 1.18 (1.05–1.31) | 3.8 × 10−3 | 0.891 | 7,765 | |||
| Hispanic/Latinx | 0.033 | 1.34 (0.99–1.81) | 0.055 | 0.596 | 3,163 | |||
| East Asian | 0.026 | 0.92 (0.45–1.88) | 0.810 | 0.541 | 2,511 | |||
| Combined | 0.055 | 1.25 (1.16–1.35) | 1.5 × 10−8 | 0.859 | 43,349 |
Three distinct genetic loci increased risk of incident CVD among individuals with T2D with genome-wide significance in time-to-event analysis (P < 5.0 × 10−8). Het P, significance of heterogeneity; POS, position in GRCh37/hg19; rsID, reference single nucleotide polymorphism identifier.
Regional association plots for the three genome-wide significant variants. A: rs147138607, near CACNA1E and ZNF648. B: rs77142250, near HS3ST1. C: rs335407, near TFB1M and NOX3. The hash marks above the panels represent the position of each SNP that was genotyped or imputed. The negative log10 of P values from the Cox regression is shown on the y-axis. Estimated recombination rates are plotted to reflect recombination hot spots. The single nucleotide polymorphisms (SNPs) in linkage disequilibrium with the most significant single nucleotide polymorphism are color coded to represent their strength of linkage disequilibrium based on European ancestry for A and C and African ancestry for B.
Regional association plots for the three genome-wide significant variants. A: rs147138607, near CACNA1E and ZNF648. B: rs77142250, near HS3ST1. C: rs335407, near TFB1M and NOX3. The hash marks above the panels represent the position of each SNP that was genotyped or imputed. The negative log10 of P values from the Cox regression is shown on the y-axis. Estimated recombination rates are plotted to reflect recombination hot spots. The single nucleotide polymorphisms (SNPs) in linkage disequilibrium with the most significant single nucleotide polymorphism are color coded to represent their strength of linkage disequilibrium based on European ancestry for A and C and African ancestry for B.
Role of Known CAD Variants in People With T2D
Among the 204 CAD variants identified in the general population (Supplementary Table 8), we observed nominally significant associations with consistent direction of effect for CAD in people with T2D for 35 SNVs, which included 5 that were significant after Bonferroni correction (P < 0.00024, 0.05/204) (Fig. 2A). For the 204 variants, we further observed consistency in the direction of association for risk of CVD between the general population and people with T2D (Fig. 2B). The CAD polygenic score consisting of these 204 variants was associated with increased CVD in people with T2D, with an estimated HR of 1.14 (95% CI 1.12–1.16) per 1-SD increase (Table 2). We showed that the association between the CAD polygenic score and CVD differed by ancestry groups (nonsignificant in East Asians with use of European-derived summary statistics).
Association of 204 previously identified CAD variants with incident CVD in people with T2D. A: Quantile-quantile plot showing the distribution of the observed P values for the 204 CAD variants with risk of incident CVD in people with T2D against the expected distribution under the null hypothesis. Red dots highlight five variants that were significantly associated with incident CVD after Bonferroni correction. B: Comparison of the effect size of 204 known CAD variants in the general population and incident CVD in people with T2D. Effect size of the known 204 CAD variants for prevalent CAD in the general population (x-axis, β-coefficient from logistic regression analysis) and incident CVD in people with T2D (y-axis, β-coefficient from Cox regression analysis) is plotted. Red dots highlight 35 variants that were nominally (P < 0.05) associated with incident CVD and had same direction of association in the general population.
Association of 204 previously identified CAD variants with incident CVD in people with T2D. A: Quantile-quantile plot showing the distribution of the observed P values for the 204 CAD variants with risk of incident CVD in people with T2D against the expected distribution under the null hypothesis. Red dots highlight five variants that were significantly associated with incident CVD after Bonferroni correction. B: Comparison of the effect size of 204 known CAD variants in the general population and incident CVD in people with T2D. Effect size of the known 204 CAD variants for prevalent CAD in the general population (x-axis, β-coefficient from logistic regression analysis) and incident CVD in people with T2D (y-axis, β-coefficient from Cox regression analysis) is plotted. Red dots highlight 35 variants that were nominally (P < 0.05) associated with incident CVD and had same direction of association in the general population.
Association of polygenic score of 204 known CAD variants and incident CVD in people with T2D
| Ancestry | HR | 95% CI | Het P | P |
|---|---|---|---|---|
| European/European American | 1.18 | 1.14–1.21 | 0.010 | <1.0 × 10−16 |
| African American | 1.10 | 1.05–1.15 | 0.255 | 8.3 × 10−5 |
| Hispanic/Latinx | 1.10 | 1.03–1.18 | 0.474 | 0.0031 |
| East Asian | 0.99 | 0.88–1.13 | 0.586 | 0.982 |
| Overall | 1.14 | 1.12–1.16 | 0.002 | <1.0 × 10−16 |
| Ancestry | HR | 95% CI | Het P | P |
|---|---|---|---|---|
| European/European American | 1.18 | 1.14–1.21 | 0.010 | <1.0 × 10−16 |
| African American | 1.10 | 1.05–1.15 | 0.255 | 8.3 × 10−5 |
| Hispanic/Latinx | 1.10 | 1.03–1.18 | 0.474 | 0.0031 |
| East Asian | 0.99 | 0.88–1.13 | 0.586 | 0.982 |
| Overall | 1.14 | 1.12–1.16 | 0.002 | <1.0 × 10−16 |
Polygenic score of 204 CAD variants discovered from the general population was associated with increased risk of incident CVD in people with T2D. HR for 1-SD increase in PRS. Het P, significance of heterogeneity.
Conclusions
In this study, we sought to identify novel genetic loci associated with incident CVD in people with T2D by performing a time-to-event GWAS. We discovered three distinct SNVs that reached genome-wide significance: rs147138607, between CACNA1E and ZNF648; rs77142250, near HS3ST1; and rs335407, between TFB1M and NOX3. We found that most CAD variants already known from cross-sectional GWAS in the general population were also associated with incident CVD events in people with T2D. Furthermore, a polygenic score composed of 204 CAD variants was associated with incident CVD. To the best of our knowledge, this is the first large-scale genetic association study with investigation of genetic risk factors of incident CVD specifically in people with T2D.
The main objective of this study was to identify genetic variants that could explain the excess risk of CVD in people with T2D. We show that for people with T2D there is enrichment of genetic risk factors of CAD observed in the general population: 1) there was an excess number of common single variants known to be associated with CAD in people with T2D and 2) polygenic score composed of these variants was significantly associated with incident CVD in people with T2D. Furthermore, we identified genetic loci associated with incident CVD, specifically in people with T2D. These variants were not identified as genetic risk factors of CVD in the general population. Taken together, we show that the excess CVD risk for people with T2D is conferred at least in part by the excess of known CAD variants and variants with effects specifically in the context of T2D. However, further research is required to quantify the excess risk conferred by these genetic risk factors.
The strengths of this study include the use of time-to-event GWAS for incident CVD rather than performance of a conventional case-control analysis. This study also benefits from the fact that we included samples from different ancestries and performed a multiancestry meta-analysis (36.8% non-European). Multiancestry meta-analysis is known to increase power where association signal is shared across ancestry groups and improves fine-mapping resolution. Interestingly, all three variants showed significance for African ancestry, while rs147138607 and rs335407 were also nominally significant for European ancestry. It should be noted that the rs77142250 variant was exclusively present in individuals with African ancestry, with a low MAF of 1.3%. Moreover, investigators of recent multiancestry GWAS advocate a more stringent level of statistical significance, set at P < 5.0 × 10−9 (17). Therefore, our findings require further replication.
In conclusion, we discovered three loci that are associated with incident CVD and show that known CAD variants identified in the general population are also enriched in people with T2D. While these findings necessitate validation, they offer insight into the increased risk of CVD observed in individuals with T2D.
Article Information
Acknowledgments. A full list of principal Cardiovascular Health Study (CHS) investigators and institutions can be found at CHS-NHLBI.org. The authors thank the other investigators, the staff, and the participants of the Multi-Ethnic Study of Atherosclerosis (MESA) for valuable contributions. A full list of participating MESA investigators and institutes can be found at https://www.mesa-nhlbi.org. The authors acknowledge the Penn Medicine BioBank (PMBB) for providing data and thank the patient-participants of Penn Medicine who consented to participate in this research program. The authors also thank the PMBB team and Regeneron Genetics Center for providing genetic variant data for analysis. The authors thank the other investigators, the staff, and the participants of the Reasons for Geographic and Racial Differences in Stroke (REGARDS) study for valuable contributions. A full list of participating REGARDS investigators and institutions can be found at http://www.regardsstudy.org.
R.D.J. is deceased.
S.S.R. is an editor of Diabetes Care but was not involved in any of the decisions regarding review of the manuscript or its acceptance.
Funding. S.H.K. was supported by a National Research Foundation of Korea grant funded by the Korean Ministry of Science and ICT (RS-2023-00262002) and by a Ministry of Food and Drug Safety grant (23212MFDS202) in 2023. J.M. was supported by American Diabetes Association grant No. 7-21-JDFM-005. J.M.M. is supported by American Diabetes Association Innovative or Clinical Translational Award 1-19-ICTS-068, American Diabetes Association grant 11-22-ICTSPM-16, and National Human Genome Research Institute grant U01HG011723. M.O.G. was supported by the National Institute of Diabetes and Digestive and Kidney Disease (R01-DK109588 and P30-DK063491) and National Center for Advancing Translational Sciences (UL1TR001420, UL1TR001881), National Institutes of Health, U.S. Department of Health and Human Services, and by the Eris M. Field Chair in Diabetes Research. R.B.H.-C. was supported by the National Institute of General Medical Sciences of the National Institutes of Health under award no. T32GM100842. S.M.D. is supported by U.S. Department of Veterans Affairs Clinical Research and Development award IK2-CX001780. R.M. was supported by the National Heart, Lung, and Blood Institute (NHLBI), National Institutes of Health (R01HL142809 and R01HL159514); the American Heart Association (22TPA969625); and the Wild Family Foundation. This CHS research was supported by NHLBI contracts HHSN268201200036C, HHSN268200800007C, HHSN268201800001C, N01HC55222, N01HC85079, N01HC85080, N01HC85081, N01HC85082, N01HC85083, N01HC85086, and 75N92021D00006 and NHLBI grants U01HL080295, R01HL085251, R01HL087652, R01HL105756, R01HL103612, R01HL120393, and U01HL130114, with additional contribution from the National Institute of Neurological Disorders and Stroke, National Institutes of Health. Additional support was provided through R01AG023629 from the National Institute on Aging. The provision of genotyping data was supported in part by the National Center for Advancing Translational Sciences, Clinical and Translational Science Institute (CTSI) grant UL1TR001881, and National Institute of Diabetes and Digestive and Kidney Diseases Diabetes Research Centers grant DK063491 to the Southern California Diabetes Endocrinology Research Center. The Framingham Heart Study (FHS) is conducted and supported by the National Heart, Lung, and Blood Institute (NHLBI) in collaboration with Boston University (contract nos. N01-HC-25195, HHSN268201500001I, and 75N92019D00031). The Jackson Heart Study (JHS) is supported by and conducted in collaboration with Jackson State University (HHSN268201800013I), Tougaloo College (HHSN268201800014I), the Mississippi State Department of Health (HHSN268201800015I), and the University of Mississippi Medical Center (HHSN268201800010I, HHSN268201800011I, and HHSN268201800012I), contracts from the NHLBI and the National Institute on Minority Health and Health Disparities. The Korean Genome and Epidemiology Study (KoGES) analysis was supported by a grant from the Korea Health Technology R&D Project through the Korea Health Industry Development Institute, funded by the Ministry of Health and Welfare (grant HI15C3131). MESA and the MESA SNP Health Association Resource (SHARe) projects are conducted and supported by the NHLBI in collaboration with MESA investigators. Support for MESA is provided by contracts 75N92020D00001, HHSN268201500003I, N01-HC-95159, 75N92020D00005, N01-HC-95160, 75N92020D00002, N01-HC-95161, 75N92020D00003, N01-HC-95162, 75N92020D00006, N01-HC-95163, 75N92020D00004, N01-HC-95164, 75N92020D00007, N01-HC-95165, N01-HC-95166, N01-HC-95167, N01-HC-95168, N01-HC-95169, UL1-TR-000040, UL1-TR-001079, UL1-TR-001420, UL1TR001881, DK063491, and R01HL105756. Funding for SHARe genotyping was provided by NHLBI contract N02-HL-64278. This study was also supported in part by NHLBI contracts R01HL151855 and R01HL146860 and National Institute of Diabetes and Digestive and Kidney Diseases contract UM1DK078616. The PMBB is approved under institutional review board protocol no. 813913 and supported by Perelman School of Medicine at University of Pennsylvania, a gift from the Smilow family, and the National Center for Advancing Translational Sciences under Clinical and Translational Science Award no. UL1TR001878. J.W.J. is an Established Clinical Investigator of the Netherlands Heart Foundation (grant 2001 D 032). Support for genotyping was provided by the 7th Framework Programme of the European Commission (grant 223004) and by the Netherlands Genomics Initiative (Netherlands Consortium for Healthy Aging grant 050-060-810). This REGARDS research was supported by NHLBI grants R01HL136666 and T32HL007457. The REGARDS study is supported by a cooperative agreement, U01 NS041588, from the National Institute of Neurological Disorders and Stroke. Representatives of the funding agency were involved in the review of the manuscript but not directly involved in the collection, management, analysis, or interpretation of data. The Women’s Health Initiative (WHI) program is funded by the NHLBI through contracts 75N92021D00001, 75N92021D00002, 75N92021D00003, 75N92021D00004, and 75N92021D00005. Scientific Computing Infrastructure at Fred Hutch is funded by Office of Research Infrastructure Programs grant S10OD028685. The WHI Diabetes Ancillary Study was funded by National Institute of Diabetes and Digestive and Kidney Diseases grant R01DK125403 (principal investigator S.L.). The Women’s Genome Health Study (WGHS) is partly funded by the NHLBI (HL043851 and HL080467) and the National Cancer Institute (CA047988 and UM1CA182913).
This publication does not represent the views of the Department of Veterans Affairs or the U.S. government. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of Neurological Disorders and Stroke or the National Institutes of Health.
Duality of Interest. S.M.D. receives research support from Renalytix, outside the current work. A Prospective Study of Pravastatin in the Elderly at Risk (PROSPER) was supported by an investigator-initiated grant obtained from Bristol-Myers Squibb. The WGHS receives funding for genotyping from Amgen. No other potential conflicts of interest relevant to this article were reported.
Author Contributions. S.H.K. conceived and designed the study, conducted data analysis, participated in results discussion, and wrote the manuscript. R.B.H.-C., D.A.D., D.E.C., J.M., P.W., J.A.B., J.Y., X.G., F.A., and M.M. conducted data analysis, participated in results discussion, and wrote the manuscript. M.S., J.M.M., J.H., H.M.T.V., Z.L., N.D.A., S.G., N.L.T., L.A.L., N.W., K.L.W., S.T., S.L., R.J.F.L., R.J., P.H.S., N.R.H., M.M.B., A.C.M., A.P.R., J.E.M., N.S.C., L.K.C., Y.-D.I.C., K.D.T., M.G., J.v.M., and A.N.P. conducted data analysis, participated in results discussion, and edited the manuscript. R.D.J. provided data and participated in results discussion. B.M.P., R.N., R.D., K.S.P., J.W.J., M.K., A.C., S.S.R., S.M.D., C.H., N.H.C., M.R.I., J.S.P., and G.N.N. provided data, participated in results discussion, and edited the manuscript. R.S., M.O.G., J.C.F., D.I.C., S.R.H., C.K., and P.S.d.V. participated in study design, participated in results discussion, and edited the manuscript. S.L. and R.M. performed data analysis and wrote the manuscript. J.D., C.-T.L., J.I.R., and J.B.M. conceived and designed the study, participated in results discussion, and edited the manuscript. S.H.K. and J.B.M. are the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Prior Presentation. Parts of this study were presented in abstract form at the 81st Scientific Sessions of the American Diabetes Association, New Orleans, LA, 25–29 June 2021.
Handling Editors. The journal editors responsible for overseeing the review of the manuscript were Cheryl A.M. Anderson and Adrian Vella.
This article contains supplementary material online at https://doi.org/10.2337/figshare.25444327.
