Effect modification of general and central obesity by sex and age on cardiovascular outcomes: Targeted maximum likelihood estimation in the atherosclerosis risk in communities study

To elucidate the effect modification of general and central obesity by sex and age on the risk of cardiovascular events. The analysis included 14,983 males and females aged 45–75 years from the Atherosclerosis Risk in Communities study. Obesity was defined with body mass index (BMI), waist-to-hip ratio (WHR) and body shape index (BSI) which categorized the participants as obese and non-obese. Targeted maximum likelihood estimation was used to estimate the risk ratio (RR) with the tmle package in R software. After adjustment, the strongest effect of BMI on CHD was in females (RR (95%CI): 1.26 (1.11, 1.42)) and in age>54 (RR (95%CI): 1.16 (1.06, 1.27)) and for HF it was in age>54 (RR (95%CI): 1.18 (1.10, 1.26)) and in females (RR (95%CI): 1.17 (1.08, 1.28)). Regarding central obesity, WHR (RR (95%CI): 1.19 (1.05, 1.34)) had the strongest effects on CHD for males and BSI (RR (95%CI): 1.140 (1.02, 1.26)) for age ≤ 54, and for HF the WHR (RR (95%CI): 1.22 (1.10, 1.36)) and BSI (RR (95%CI): 1.18 (1.07, 1.30)) had the strongest effects for age≤54, respectively. Among males and age≤54, WHR index was associated with a higher risk of CHD and HF while BMI was so for females and age>54. •Body mass index is frequently disapproved for its inadequate capacity to differentiate between fat mass and other body mass.•Different fat and body mass definition result in different risk of CHD and HF for males and females.•The central obesity provides better prediction than global obesity for both outcomes.•Targeted maximum likelihood estimation provide better estimation of true values (causal effect) than conventional methods.