Background: Obesity is a major modifiable risk factor for heart failure (HF), but body mass index (BMI) does not distinguish biologically distinct fat depots. Whether imaging-derived adipose tissue depots capture specific cardiometabolic pathways underlying HF risk beyond conventional anthropometric measures remains uncertain. Methods: We performed two-sample Mendelian randomization (MR) and multivariable MR mediation analyses using published genome-wide association study summary statistics. General adiposity traits included BMI, waist-to-hip ratio (WHR), and WHR adjusted for BMI from GIANT and UK Biobank meta-analyses of up to 694,649 individuals. MRI-derived visceral adipose tissue (VAT), abdominal subcutaneous adipose tissue (ASAT), and gluteofemoral adipose tissue (GFAT) were derived from 38,965 UK Biobank participants. HF outcome data were obtained from the HERMES consortium, including 1,946,349 individuals and 153,174 HF cases. Cardiometabolic mediators included type 2 diabetes (T2D), systolic blood pressure (SBP), LDL cholesterol, HDL cholesterol, and triglycerides. Primary analyses used inverse-variance weighted MR, with sensitivity analyses and directionality testing. Mediation was estimated using joint multivariable MR conditioning on significant cardiometabolic mediators.
Background: Obesity is a major modifiable risk factor for heart failure (HF), but body mass index (BMI) does not distinguish biologically distinct fat depots. Whether imaging-derived adipose tissue depots capture specific cardiometabolic pathways underlying HF risk beyond conventional anthropometric measures remains uncertain. Methods: We performed two-sample Mendelian randomization (MR) and multivariable MR mediation analyses using published genome-wide association study summary statistics. General adiposity traits included BMI, waist-to-hip ratio (WHR), and WHR adjusted for BMI from GIANT and UK Biobank meta-analyses of up to 694,649 individuals. MRI-derived visceral adipose tissue (VAT), abdominal subcutaneous adipose tissue (ASAT), and gluteofemoral adipose tissue (GFAT) were derived from 38,965 UK Biobank participants. HF outcome data were obtained from the HERMES consortium, including 1,946,349 individuals and 153,174 HF cases. Cardiometabolic mediators included type 2 diabetes (T2D), systolic blood pressure (SBP), LDL cholesterol, HDL cholesterol, and triglycerides. Primary analyses used inverse-variance weighted MR, with sensitivity analyses and directionality testing. Mediation was estimated using joint multivariable MR conditioning on significant cardiometabolic mediators. Results: Among MRI-derived adipose depots, ASAT was the only trait significantly associated with HF risk (odds ratio [OR] 1.64 per 1-SD increase; 95% CI 1.40-1.93; FDR q<0.001). VAT showed a positive but imprecise association (OR 1.38; 95% CI 0.87-2.20), and GFAT was not associated with HF. Among general adiposity measures, BMI (OR 1.65; 95% CI 1.58-1.71) and WHR (OR 1.30; 95% CI 1.23-1.38) were robustly associated with HF, whereas WHR adjusted for BMI was not. ASAT was significantly associated with T2D, SBP, HDL cholesterol, and triglycerides, but not LDL cholesterol. In joint multivariable MR, 67.9% of ASAT's HF effect was mediated through T2D, SBP, HDL cholesterol, and triglycerides (95% CI 49.2-86.8%). In contrast, BMI demonstrated only 8.5% mediation (95% CI -7.4 to 24.4%), and WHR showed non-significant mediation of 36.7% (95% CI -8.3 to 81.7%). Conclusions: MRI-derived abdominal subcutaneous adipose tissue captures a biologically coherent cardiometabolic signal underlying HF risk that is diluted by conventional anthropometric measures. ASAT may represent an imaging biomarker of metabolic syndrome-mediated HF risk and could support more precise risk stratification and mechanistically targeted prevention in HF.