Premium
Age Variation in the Association Between Obesity and Mortality in Adults
Author(s) -
Wang Zhiqiang,
Peng Yang,
Liu Meina
Publication year - 2017
Publication title -
obesity
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.438
H-Index - 199
eISSN - 1930-739X
pISSN - 1930-7381
DOI - 10.1002/oby.21976
Subject(s) - hazard ratio , medicine , obesity , demography , national health and nutrition examination survey , proportional hazards model , body mass index , gerontology , confidence interval , environmental health , population , sociology
Objective The aim of this study was to evaluate the previously reported finding that the association between obesity and mortality strengthens with increasing age. Methods The data were derived from the National Health Interview Survey. Age‐specific hazard ratios of mortality for grade 2/3 obesity (BMI ≥ 35 kg/m 2 ), relative to a BMI of 18.5 kg/m 2 to < 25 kg/m 2 , were calculated by using a flexible parametric survival model (240,184 white men) and Cox proportional hazard models (51,697 matched pairs). Results When the model included interaction terms between obesity and age at the survey, hazard ratios appeared to increase with age if those interaction terms were ignored by fixing age at the survey as a single value. However, when recalculated for adults with various ages at the survey, according to model specifications, hazard ratios were higher for younger adults than for older adults with the same follow‐up duration. Based on matched data, hazard ratios were also higher for younger adults (2.14 [95% CI: 1.90‐2.40] for those 40‐49 years of age) than for older adults (1.22 [95%: 0.91‐1.63] for those 90+ years of age). Conclusions For any given follow‐up duration, the association between obesity and mortality weakens with age. The previously reported strengthening of the obesity‐mortality association with increasing age was caused by the failure to take all the model specifications into consideration when calculating adjusted hazard ratios.