Assessing the Association Between Relative BMI and Mortality. (7th June 2021)

• Record Type: Journal Article
• Authors: Mestre, Luis; Pavela, Gregory; Yi, Nengjun; Xun, Pengcheng; Allison, David
• Is Part Of: Current developments in nutrition. Volume 5(2021)Supplement 2
• Page Start: 36
• Page End: 36
• Publication Date: 2021-06-07
• Subjects: Nutrition -- Periodicals; Nutritional Physiological Phenomena; Nutrition; Periodicals; Periodicals; Fulltext; Internet Resources; Periodicals; 612.3
• DOI: 10.1093/cdn/nzab033_036
• Languages: English
• ISSNs: 2475-2991

Abstract

Abstract: Objectives: The objective is to determine the association of relative BMI and mortality rate. Methods: Varying-coefficient generalized additive proportional hazards model was used to analyze the association between BMI and mortality. This model allows the coefficient to vary as smooth functions of other variables. We show that the varying-coefficient model provides an efficient way for jointly assessing the effects of BMI and the cohort characteristics on the mortality rate to capture the idea of relative BMI. Results: All considered predictors are significantly associated with the mortality rate. The smooth terms and the interactions of BMI with mean BMI and standard deviation BMI were non-linear. Absolute BMI (individual BMI) has significant interactions with the mean and standard deviation BMI. Conclusions: There is an association of a person with BMI in the normal range with lower mortality while their BMI in the low or high extreme had higher mortality. A person's BMI is associated with the interactions that have with the mean BMI and standard deviation of its cohort, which can explain, to some extent, the non-linear relationship between BMI and mortality rate. Funding Sources: NIH-NIA