Accelerometer-Measured Physical Activity Changes After Transitions to Concern About Falling, a Recent Fall, or Both in Older Adults: A Longitudinal NHATS Study

Whether recent falls and concern about falling represent distinct precursors of subsequent free-living activity change remains unclear.

We analyzed National Health and Aging Trends Study Rounds 11-14 using a person-interval design spanning three annual intervals, with each interval restricted to participants who reported neither a recent fall nor concern about falling at the preceding wave. Follow-up states were classified as stable neither, concern only, fall only, or both. Wrist accelerometry provided measures of activity volume and fragmentation, with active minutes per day as a secondary outcome. Survey-weighted models adjusted for the preceding outcome, demographic and health covariates, prior-year falls, and interval; Holm correction was applied across the two primary global tests.

The analytic sample included 873 person-intervals from 437 participants. Transition state was associated with activity volume (Holm-adjusted P = 0.021), whereas evidence for differences in fragmentation was inconclusive (Holm-adjusted P = 0.080). Relative to stable neither, activity volume was 3.18% lower for concern only (95% CI, -7.47% to 1.32%), 6.95% lower for fall only (95% CI, -14.16% to 0.86%), and 12.41% lower for co-occurring fall and concern (95% CI, -20.80% to -3.14%). The co-occurring state was also associated with 41.13 fewer active minutes/day (95% CI, -71.46 to -10.79). Concern-only and fall-only transitions were not clearly different, and concern-related activity restriction was not clearly associated with larger activity changes.

Co-occurring fall and concern showed the clearest association with subsequent loss of device-measured activity volume, while fragmentation changes remained uncertain.