Relative importance of four functional measures as predictors of 15-year mortality in the older Dutch population

Decreased physical function is known to raise mortality risk. Little is known about how different physical function measures compare in predicting mortality risk in older men and women. The objective of this study was to compare four, objective and self-reported, physical function measures in predicting 15-year mortality risk in older men and women. Data were used from the Longitudinal Aging Study Amsterdam (LASA), an ongoing cohort study in a population-based sample of the older Dutch population, sampled from municipal records. The 1995-96 cycle, including 727 men and 778 women aged 65-88 years, was considered as the baseline. Mortality was followed up through September 1, 2011. Physical function measures were: lower-body performance (chair stands test, walk test and tandem stand); handgrip strength (grip strength dynamometer); lung function (peak expiratory flow rate); functional limitations (self-report of difficulties in performing six activities of daily living). Cox proportional hazard models were used to determine the predictive value of each physical function measure for 15-year mortality risk, adjusted for demographic, lifestyle and health variables as potential confounders. 1031 participants (68.5%) had died. After adjustments for confounders, in models assessing single functional measures, peak flow was the strongest predictor of all-cause mortality in men (HR 1.76, CI 1.38-2.26, CI) and lower-body performance in women (HR 1.97,CI 1.40-2.76, CI). In a model including all four functional measures only peak flow was statistically significant in predicting mortality in both genders (men HR 1.54,CI 1.18-2.01 and women HR 1.45,CI 1.08-1.94). In women, lower-body performance (HR 1.66, CI 1.15-2.41) followed by grip strength (HR 1.38, CI 1.02-1.89), and in men, functional limitations (HR 1.43, CI 1.14-1.8) were the other significant predictors of all-cause mortality. Both objective and self-reported measures of physical functioning predicted all-cause mortality in a representative sample of the older Dutch population to different extents in men and women. Peak flow contributed important unique predictive value for mortality in both men and women. In women, however, lower-body performance tests had better predictive ability. A second-best predictor in men was self-reported functional limitations. Peak flow, and possibly one of the other measures, may be used in clinical practice for assessment in the context of time constraints.