Body-composition predictors of mortality in women aged ≥75 y: data from a large population-based cohort study with a 17-y follow-up

Abstract

The role of body composition as a risk factor for death remains controversial in older persons. We determined the role of body-composition variables in mortality in older women. Longitudinal analyses were performed in a prospective cohort study of older women. Participants were 4574 community-dwelling women aged ≥ 75 y at the baseline visit (between January 1992 and April 1994). Several body-composition variables were assessed by using anthropometric measures and dual-energy X-ray absorptiometry at the baseline visit. The main outcome was overall mortality. Body-composition variables were body mass index (BMI; in kg/m(2)), hip circumference, waist circumference, waist-to-hip ratio, fat mass/height(2), lean mass/height(2), percentage of fat mass, percentage of lean mass, and the lean mass:fat mass ratio. The mean (± SD) age at baseline was 80.2 ± 3.8 y. During the 17.7 y (IQR: 17.2-18.1 y) of follow-up, 2876 women died. U-shaped in crude analyses and reversed J-shaped relations in adjusted analyses between BMI, hip and waist circumferences, fat mass/height(2), and risk of death were shown. Adjusted risk of death was significantly higher in participants with BMI ≤ 24.6 and fat mass/height(2) ≤ 8.2 kg/m(2). There was a negative linear association between fat mass (%) and risk of death: a 10% increase in fat mass was associated with a 12% reduction of mortality risk (adjusted HR: 0.88; 95% CI: 0.84, 0.92; P < 0.001). Linear and statistically significant relations were shown between lean mass/height(2) and risk of death in crude but not adjusted analyses. Risk of mortality was consistently higher in older women with low adiposity. No lean mass indicator was associated with risk of death. Clinicians should be alerted by low adiposity in older women.