Central adiposity as a predictor of mortality in older adults: Identification of cutoffs using generalized additive models

Abstract

Background: Obesity is associated with premature mortality in adults; however, this association has been inconsistent in the older adult population. In addition, there is a lack of specific cutoff points for indicators of negative health outcomes in older adults. Methods: This is a prospective study with 796 non-institutionalized older adults. Data on sociodemographic characteristics, lifestyle, food consumption, and nutritional status were obtained at baseline. Generalized additive models were used to identify cutoff points for the waist circumference (WC) and waist-to-height ratio (WHtR) and Cox proportional hazards models to assess the independent association between adiposity and mortality. Results: Over the 9 years of follow-up, 197 deaths (24.7%) occurred, of which 51.8% were men, with a mean age of 76.1 ± 9.0 years. Older adults at higher risk of death had WHtR of <0.52 or ≥0.63 and WC of <83 cm or ≥101 cm. An increased risk of death was observed in older adults with high WC (HR: 2.03 95% CI: 1.20–3.41) and high WHtR (HR: 1.51 95% CI: 1.01–2.26) in the adjusted models, and an increase in WC was a risk factor for higher CVD mortality (HR: 2.09, 95% CI: 1.12–3.88) in the adjusted models. Conclusion: Adiposity was associated with an increased risk of death in older adults. In view of these results and considering the lack of cutoff points for anthropometric indices in Brazilian older adults, further studies are needed to confirm the WC and WHtR cutoff values found in this study.