Time-varying confounders in association between general and central obesity and coronary heart disease: Longitudinal targeted maximum likelihood estimation on atherosclerosis risk in communities study

Abstract

Aim

This study examines the association between general and central obesity and the risk of cardiovascular diseases, utilizing the Targeted Maximum Likelihood Estimator (TMLE) method to account for time-varying covariates and also we compares the findings with those derived from conventional regression methods in the Atherosclerosis Risk in Communities (ARIC) cohort study.

Methods

We considered 15,792 participants 45–75 years of age registered in the Atherosclerosis Risk in Communities study, visit 1 and followed to visit 4. General obesity defined as body mass index and central obesity defined as Waist Circumference (WC), Waist-Hip-Ratio (WHR), Waist-Height-Ratio (WHtR), Body Shape Index (BSI) and Body Roundness Index (BRI). The effect of obesity on Coronary Heart Disease (CHD) was estimated and compared by Longitudinal Targeted Maximum Likelihood Estimation (LTMLE) and generalized linear model.

Results

The effects of BMI, adjusted for baseline and time-varying confounders, was 1.15 (95 %CI =1.00, 1.34). About the gender groups, the effect of BMI for males and females was 1.17 (95 %CI =0.97,1.40) and 1.19 (95 %CI =0.94,1.52), respectively. Considering age groups, the effect of BMI was 1.21 (95 %CI =0.95, 1.53) and 1.13 (95 %CI = 0.93, 1.36) for age ≤ 54 years and age > 54 years, respectively. With regards to central obesity, the BSI and WC were shown the strongest effects, respectively. Among females and age group≤54 years, WHtR was associated with a higher incidence of CHD.

Conclusions

According to the results, the appropriate index for obesity varies based on gender and age. Knowledge about this difference will help to experts to implement appropriate interventions.