Dominant role of abdominal adiposity in circulating lipoprotein, lipid, and metabolite levels in UK Biobank: Mendelian randomization study

Abstract

Background: The causal impact of excess adiposity on systemic metabolism is unclear. We used multivariable Mendelian randomization to compare the direct effects of total adiposity (using body mass index (BMI)) and abdominal adiposity (using waist-to-hip-ratio (WHR)) on circulating lipoproteins, lipids, and metabolites with a five-fold increase in sample size over previous studies. Methods: We used new metabolic data on 109,532 UK Biobank participants. BMI and WHR were measured in 2006-2010, during which EDTA plasma was collected. Plasma samples were used in 2019-2020 to quantify 249 metabolic traits with high-throughput nuclear magnetic resonance spectroscopy including subclass-specific lipoprotein concentrations, apolipoprotein B, cholesterol and triglycerides, plus pre-glycemic and inflammatory metabolites. We used two-stage least squares regression models with genetic risk scores for BMI and WHR as instruments to estimate the total (unadjusted) and direct (mutually adjusted) effects of BMI and WHR on metabolic traits. We also estimated the effects of BMI and WHR on statin use, and examined interaction of main effects by sex, statin use, and age as a proxy for medication use. Results: Higher BMI (per standard deviation (SD) or 4.8 kg/m2) was estimated to moderately decrease apolipoprotein B and low-density lipoprotein (LDL) cholesterol before and after adjustment for WHR, whilst higher BMI increased triglycerides before but not after WHR adjustment. Estimated effects of higher WHR (per SD, or 0.090 ratio-unit) on lipoproteins, lipids, and metabolites were often larger than those of BMI, but null for LDL cholesterol, and attenuations were minimal upon adjustment for BMI. Patterns of effect estimates differed by sex, e.g., only BMI independently increased triglycerides among men, whereas only WHR independently increased triglycerides among women. Higher BMI and WHR (per SD) were each estimated to directly increase the relative odds of using statins (by 3.49 (95% CI = 3.42, 3.57) times higher for WHR). These patterns were most pronounced among women, and there was strong evidence that the effects of BMI and WHR on metabolic traits differed by statin use and age. Among the youngest adults (38-53 years, statin use 5%), higher BMI and WHR (per SD) each modestly increased LDL cholesterol (0.04 SD, 95% CI = -0.01, 0.08 for total effect of BMI and 0.10 SD, 95% CI = 0.02, 0.17 for total effect of WHR). This estimate for BMI fully attenuated, and the estimate for WHR remained unchanged, upon mutual adjustment. These direct effects on LDL cholesterol were more inverse for BMI and less positive for WHR at intermediate ages (54-62 years, statins 17%) and older ages (63-73 years, statins 29%) where the mutually adjusted effects of BMI and WHR on LDL cholesterol had reversed to -0.19 SD (95% CI = -0.27, -0.11) and -0.05 SD (95% CI = -0.16, 0.06), respectively. Conclusions: Our results suggest that abdominal adiposity has a dominant role in driving the metabolic harms of excess adiposity, particularly among women. Our findings also suggest that apparent effects of adiposity on lowering LDL cholesterol are explained by an effect of adiposity on statin use.