Attribution of Causes of Weight Loss and Weight Gain to 3-Year Mortality in Older Adults: Results From the Longitudinal Aging Study Amsterdam

Abstract

Weight loss is associated with a higher mortality risk in old age, but the underlying cause may impact this association. We examined associations between causes of intentional and unintentional weight loss and weight gain and mortality. We used data of five triannual examination rounds of the Longitudinal Aging Study Amsterdam (age ≥ 55 years, n = 2,645) and two examination rounds of a new cohort (n = 909). Self-reported weight loss or gain and causes were measured during a personal interview. Time-dependent Cox regression was used to model the association between weight loss and gain causes and subsequent 3-year mortality. At baseline, 16% reported weight loss (mean = 4.7 kg, SD = 3.7) in 6 months. After adjustment for potential confounders, an increased mortality risk was observed for unintentional weight loss due to medical reasons (<72 years: hazard ratio = 2.43 [95% confidence interval: 1.52-3.88]; ≥72 years: 1.62 [1.23-2.14]), unknown reasons (1.98 [1.49-2.62]), and change in eating pattern (1.89 [1.12-3.18]). No association was found for unintentional weight loss due to social reasons, intentional weight loss (dieting or physical activity), or weight gain. Weight loss due to medical or social reasons was often regained in subsequent 3 years while weight loss due to other causes was not. Weight loss due to social reasons was not associated with mortality suggesting that not all unintentional weight loss is harmful. The increased mortality risk of other causes of unintentional weight loss may be related to underlying disease. Intentional weight loss was not associated with mortality.