Association of Genetic Variants Related to Gluteofemoral vs Abdominal Fat Distribution With Type 2 Diabetes, Coronary Disease, and Cardiovascular Risk Factors

Abstract

Importance: Body fat distribution, usually measured using waist-to-hip ratio (WHR), is an important contributor to cardiometabolic disease independent of body mass index (BMI). Whether mechanisms that increase WHR via lower gluteofemoral (hip) or via higher abdominal (waist) fat distribution affect cardiometabolic risk is unknown.

Objective: To identify genetic variants associated with higher WHR specifically via lower gluteofemoral or higher abdominal fat distribution and estimate their association with cardiometabolic risk.

Design, setting, and participants: Genome-wide association studies (GWAS) for WHR combined data from the UK Biobank cohort and summary statistics from previous GWAS (data collection: 2006-2018). Specific polygenic scores for higher WHR via lower gluteofemoral or via higher abdominal fat distribution were derived using WHR-associated genetic variants showing specific association with hip or waist circumference. Associations of polygenic scores with outcomes were estimated in 3 population-based cohorts, a case-cohort study, and summary statistics from 6 GWAS (data collection: 1991-2018).

Exposures: More than 2.4 million common genetic variants (GWAS); polygenic scores for higher WHR (follow-up analyses).

Main outcomes and measures: BMI-adjusted WHR and unadjusted WHR (GWAS); compartmental fat mass measured by dual-energy x-ray absorptiometry, systolic and diastolic blood pressure, low-density lipoprotein cholesterol, triglycerides, fasting glucose, fasting insulin, type 2 diabetes, and coronary disease risk (follow-up analyses).

Results: Among 452 302 UK Biobank participants of European ancestry, the mean (SD) age was 57 (8) years and the mean (SD) WHR was 0.87 (0.09). In genome-wide analyses, 202 independent genetic variants were associated with higher BMI-adjusted WHR (n = 660 648) and unadjusted WHR (n = 663 598). In dual-energy x-ray absorptiometry analyses (n = 18 330), the hip- and waist-specific polygenic scores for higher WHR were specifically associated with lower gluteofemoral and higher abdominal fat, respectively. In follow-up analyses (n = 636 607), both polygenic scores were associated with higher blood pressure and triglyceride levels and higher risk of diabetes (waist-specific score: odds ratio [OR], 1.57 [95% CI, 1.34-1.83], absolute risk increase per 1000 participant-years [ARI], 4.4 [95% CI, 2.7-6.5], P < .001; hip-specific score: OR, 2.54 [95% CI, 2.17-2.96], ARI, 12.0 [95% CI, 9.1-15.3], P < .001) and coronary disease (waist-specific score: OR, 1.60 [95% CI, 1.39-1.84], ARI, 2.3 [95% CI, 1.5-3.3], P < .001; hip-specific score: OR, 1.76 [95% CI, 1.53-2.02], ARI, 3.0 [95% CI, 2.1-4.0], P < .001), per 1-SD increase in BMI-adjusted WHR.

Conclusions and relevance: Distinct genetic mechanisms may be linked to gluteofemoral and abdominal fat distribution that are the basis for the calculation of the WHR. These findings may improve risk assessment and treatment of diabetes and coronary disease.

Figure 1.. Associations With Compartmental Fat Mass of Polygenic Scores for Higher Waist-to-Hip Ratio (WHR)

A, Associations with compartmental fat mass for the 202–genetic variants polygenic score for higher WHR are shown. Associations are reported in clinical or standardized units of continuous outcome per 1-SD increase in body mass index (BMI)–adjusted WHR (corresponding to 0.056 ratio units of age-, sex-, and BMI-residualized WHR in the UK Biobank) due to the polygenic score. The statistical significance threshold for analyses reported in this panel was P < .002. B, Associations with compartmental fat mass for the waist- or hip-specific polygenic scores for higher WHR are shown. Associations were estimated in up to 18 330 individuals of European ancestry in the UK Biobank, Fenland, and EPIC-Norfolk studies. Associations are reported in clinical or standardized units of continuous outcome per 1-SD increase in BMI-adjusted WHR (corresponding to 0.056 ratio units of age-, sex-, and BMI-residualized WHR in the UK Biobank) due to the polygenic score used in a given analysis. The statistical significance threshold for analyses reported in this panel was P < .002.

Figure 2.. Associations With Cardiometabolic Risk Factors and Disease Outcomes of Waist- or Hip-Specific Polygenic Scores for Higher Waist-to-Hip Ratio (WHR)

A, Associations with cardiometabolic risk factors for the waist- or hip-specific polygenic scores for higher WHR are shown. Associations are reported in clinical or standardized units of continuous outcome per 1-SD increase in body mass index (BMI)–adjusted WHR (corresponding to 0.056 ratio units of age-, sex-, and BMI-residualized WHR in the UK Biobank) due to the polygenic score used in a given analysis. Data on blood pressure were from the UK Biobank; data on low-density lipoprotein (LDL-C) and triglyceride levels were from the Global Lipids Genetics Consortium; and data on fasting insulin and fasting glucose were from the Meta-analyses of Glucose and Insulin-Related Traits Consortium.^, The statistical significance threshold for analyses reported in this panel was P < .002. B, Associations with type 2 diabetes and coronary artery disease risk for the waist- or hip-specific polygenic scores for higher WHR are shown. Associations are reported in odds ratio or absolute risk increase (ARI) per 1-SD increase in BMI-adjusted WHR (corresponding to 0.056 ratio units of age-, sex-, and BMI-residualized WHR in the UK Biobank) due to the polygenic score used in a given analysis. Associations with type 2 diabetes were estimated in 69 677 cases and 551 081 controls from the DIAGRAM Consortium, EPIC-InterAct, and the UK Biobank. Associations with coronary artery disease were estimated in 85 358 cases and 551 249 controls from the UK Biobank and the CARDIoGRAMplusC4D Consortium. The statistical significance threshold for analyses reported in this panel was P < .006. ^aP value for heterogeneity in association estimates for waist- vs hip-specific polygenic score.