Objectively Assessed Weight Change and All-Cause Mortality among Community-Dwelling Older People

Abstract

Later life changes in body weight may be associated with an increased risk of mortality in older adults. The objective of this study was to examine whether weight change over four years was associated with a 17-year mortality risk in older adults. Participants were 1664 community-dwelling adults aged ≥65 years in the longitudinal Enquete de Sante’ Psychologique-Risques, Incidence et Traitement (ESPRIT) study. Outcomes were all-cause mortality, cardiovascular disease (CVD) and cancer mortality. Weight change was defined as difference between weight at baseline and 4 years, categorised into: weight stable (±<5% weight change), weight loss (≥5%) and weight gain (≥5%). Association between weight change and mortality risk was evaluated using Cox proportional hazards models. Over 17 years of follow-up (median 15 years), 565 participants died. Compared to stable weight participants, those with ≥ 5% weight loss had an increased risk of all-cause mortality (HR: 1.24, 95% CI: 1.00−1.56, p = 0.05) and CVD mortality (HR: 1.53, 95% CI: 1.10−2.14, p = 0.01), but not cancer mortality (HR: 0.83, 95% CI: 0.50−1.39, p = 0.49). Weight gain of ≥5% was not associated with increased mortality (HR: 1.05, 95% CI: 0.76−1.45, p = 0.74). Weight monitoring in older adults could help identify weight loss at its early stages to better target interventions to maintain nutritional reserve and prevent premature mortality.

Keywords: aged; ageing; body weight; body weight maintenance; healthy aging; mortality; older adults; weight change; weight gain; weight loss.

Figure 1

The risk of all-cause mortality according to percentage weight change, estimated by multivariate Cox regression based on restricted cubic splines. Log hazard ratios (log-HRs) with 95% confidence intervals for all-cause mortality and weight change in percentage. The model includes age as time scale, sex, marital status, education level, smoking status, alcohol consumption, depression, diabetes and baseline BMI. Subjects with over than 30 % weight change were not included in plot (n = 2).