Combined Self-Reported and Device-Measured Physical Activity Assessment and Disability Incidence in Older Adults

Abstract

Objectives: To clarify the relationship between self-reported and device-measured physical activity (PA) and disability incidences in community-dwelling older adults.

Design: Prospective cohort study.

Setting and participants: In total, 4504 (median age 75 [range, 72-79] years; women, 52.1%) community-dwelling older adults were included in this study.

Methods: Device-measured moderate-to-vigorous PA (MVPA) was measured using a triaxial accelerometer. Self-reported MVPA was assessed using the Japanese version of the International Physical Activity Questionnaire, Short Form. The cutoff MVPA was 150 min/wk. PA was divided into 4 groups based on self-reported and device-measured assessments: (1) self-reported/device-measured inactive, (2) self-reported active/device-measured inactive, (3) self-reported inactive/device-measured active, and (4) self-reported/device-measured active. The disability incidence, defined as a long-term care insurance certification of any level indicating a requirement for support or continuous care, was prospectively determined using data extracted from the Japanese long-term care insurance system.

Results: Of the participants, 892 (19.8%) developed disabilities within 62 (range, 60-64) months. The prevalence of self-reported inactive/device-measured inactive, self-reported active/device-measured inactive, self-reported inactive/device-measured active, and self-reported active/device-measured active were 884 (34.9%), 441 (17.4%), 693 (27.3%), and 515 (20.3%) participants, respectively. Multivariate Cox regression analysis showed that self-reported/device-measured active, self-reported inactive/device-measured active, and self-reported active/device-measured inactive independently decreased the disability incidence risk (self-reported/device-measured active: hazard ratio [HR], 0.555, 95% confidence interval [CI], 0.412-0.749; self-report inactive/device-measured active: HR, 0.656; 95% CI, 0.518-0.831; self-reported active/device-measured inactive: HR, 0.718; 95% CI, 0.556-0.928). In the post hoc analysis, those who were self-reported active/device-measured inactive were divided into 2 groups according to involvement in sports. Those who played sports had a lower disability risk; however, the difference was not significant.

Conclusions and implications: Combined self-reported and device-measured PA assessments may allow for disability risk stratification in older adults. The findings will also help in creating methods for evaluating PA levels to prevent disability incidences among older adults.

Keywords: Physical activity; device-measured assessment; disability incidence; older adults; self-reported assessment.