Dual-task Gait Variability At Baseline Predicts Impact Of A Multi-modal Intervention In Sedentary Older Adults

Many older adults know - but do not follow - recommended physical activity guidelines. Moreover, the effectiveness of physical activity intervention is highly variable across individuals due to differences in genetics, lifestyles, and environmental exposures. Elucidation of factors associated with this heterogeneous treatment effect are thus needed to inform personalized physical activity interventions within this population. PURPOSE: To determine if and what baseline factors predict the effect of physical activity intervention in older adults. METHODS: We conducted a secondary analysis of a pilot randomized controlled clinical trial designed to enhance physical activity in inactive older adults. Participants wore a Fitbit throughout the entire study period. Baseline physical activity was measured as daily steps over 2 weeks. All participants then received 8 weeks of intervention comprising 4, bi-weekly ‘coaching’ sessions. Participants also received 10, once-daily sessions of transcranial direct current stimulation (tDCS) or sham stimulation during the initial 2 weeks of intervention. Physical, cognitive, and patient-reported outcomes were assessed at baseline and after intervention. Linear regressions tested whether baseline variables predicted treatment effect (i.e. change in average daily steps during baseline to during the 8-week intervention). RESULTS: 28 participants completed the study (tDCS: N = 15, 78 ± 8y/o, 14F; sham: N = 13, 83 ± 6y/o, 11F). Participants with better(lower) baseline stride time variability while walking and engaging in a serial subtraction task exhibited greater increase in daily step counts during intervention, after controlling for age, sex, tDCS group, and study site (β = -36915, p = 0.01). No other baseline variable predicted treatment effect. CONCLUSIONS: Dual-task gait variability was the sole predictor of intervention effectiveness, suggesting that an older adults’ capacity to perform a complex cognitive-motor task may reflect their capacity to modulate behavior and increase physical activity over time. NIA K01AG075252, Boston Roybal Pilot Grant, OAIC Pepper Center Pilot Grant