Conner SC, Lodi S, Lunetta KL, et al. - Since earlier investigations evaluating the link between body mass index (BMI) and atrial fibrillation (AF) did not account for time-varying covariates, which may be influenced by past BMI, researchers showed how the g-formula can account for time-varying confounding in this study including 4,392 participants from the Framingham Heart Study, without AF at 45 to 55 years of age and observed for up to 20 years. Using Cox models, hazard ratios comparing time-varying nonobese vs obese were estimated. They compared nonobese vs obese and 10% annual reduction in BMI (until normal weight is reached) vs natural course, using the g-formula. After accounting for time-varying covariates that relied on past exposure using the g-formula, which Cox models cannot accommodate, a lower rate of AF was observed in correlation with decreased BMI. Context to relative measures of association is provided by absolute measures like the restricted mean survival time difference.