Walking and talking as predictors of falls in the general population: the Leiden 85-Plus Study.

Abstract

To compare the value of dual tasking in predicting falling in the general population of oldest old with that of easy-to-administer single tasks. Prospective population-based follow-up study. Municipality of Leiden, the Netherlands. Representative cohort of 380 individuals, all aged 85 at baseline. During enrollment, walking time over a 12-meter distance was measured, as well as the verbal fluency to recite names of animals or professions during a 30-second period. In the dual task, performance was assessed when participants combined walking with reciting names. Incidence of falls and fractures was assessed by interviewing participants and checking their medical histories. After 1 year of follow-up, 42% of the participants reported one or more falls, and 4% suffered a fracture. Total walking time, number of steps, and verbal fluency were all strongly related to incident falls (P for trend for all <.01), but dual-task performance was not a better predictor for incident falls than single-task performance. The dual-task test in this study had no predictive value above that of a single-task test to predict falling. Dual tasks with more-sensitive measures of impaired dual-task execution might have better test characteristics. In this study, history of falls and performance on an easy-to-administer single walking task identified old persons at higher risk for falling who could benefit from fall preventive strategies.