Association between different adiposity measures and all-cause mortality risk among centenarians: A prospective cohort study

Abstract

Little evidence on the association between adiposity measures and all-cause mortality was observed among centenarians, and no targeted development of optimal weight recommendations for them. To comprehensively assess the association between adiposity indices and all-cause mortality among centenarians. This prospective population-based cohort study included 1002 centenarians registered in 18 counties and cities in Hainan Province from June 2014 to May 2021. The age of participants at baseline was provided by the civil affairs bureau and verified before enrollment. All-cause mortality was rigorously confirmed as the primary outcome. BMI was calculated by height and weight. BRI was calculated by height and waist circumference. At baseline, the mean (SD) age was 102.8±2.7 years, and 180 participants (18.0%) were men. The median follow-up time was 5.0 (4.8-5.5) years, with 522 deaths. In BMI categories, compared with the lowest group (mean BMI=14.2kg/m2), the highest group (mean BMI=22.2kg/m2) had lower mortality (hazard ratio [HR], 0.61; 95%CI, 0.47-0.79) (P for trend=0.001). In BRI categories, compared with the lowest group (mean BRI=2.3), the highest group (mean BRI=5.7) had lower mortality (hazard ratio [HR], 0.66; 95%CI, 0.51-0.85) (P for trend=0.002), and the risk did not decrease after BRI exceeded 3.9 in women. Higher BRI was associated with lower HRs after adjusting for interaction with comorbidities status. E-values analysis suggested robustness to unmeasured confounding. BMI and BRI were inverse linear associated with mortality risk in the whole population, while BRI was observed to be J-shaped in women. The interaction of lower multiple complication incidence and BRI had a significant effect on the reduced risk of all-cause mortality.