A Dyadic Approach for a Remote Physical Activity Intervention in Adults With AD and Their Caregivers

This study will compare 2 strategies for the delivery of an intervention to increase moderate physical activity (MPA) in community dwelling adults with AD and their caregiver (dyads); real-time group video conferencing (RGV) vs. enchanced usual care (EUC). Adults (age ≥ 55 yrs.) with mild to moderate AD (n=100) and their caregiver (n=100) will be randomized (1:1) to one of the 2 intervention arms for an 18-mo. trial (6 mos. active intervention, 6 mos. maintenance, 6 mos no contact). Cohorts of ~20 dyads will be recruited and computer randomized. Dyads will be stratified by the sex of the person with AD, and sequentially randomized by the study statistician with equal allocation to the RGV or EUC arms. Participants in both arms will be provided with an iPad® for intervention delivery, Fitbit (Fitbit Inc., San Francisco, CA) for self-monitoring MPA, and will be asked to complete 150 min of MPA/wk. Dyads in the RGV arm will be asked to complete three 45 min sessions that include aerobic, resistance, and balance/coordination exercises delivered by a trained health coach via Zoom® software on an iPad during mos. 0-6, and 1 session/wk. during mos. 7-12 to groups of 5-8 dyads in their home. Dyads in the EUC arm, will be given a recommended exercise plan to follow own their own. Dyads in both arms will be provided with written materials regarding exercise and physical activity from the National Institute on Aging and will be asked to complete brief (15-20 min) FaceTime meetings (0-6 mos.= 2/mo.; 7-12 mos.= 1/mo.) with the heath coach to discuss progress, provide support and receive additional guidance on how to increase MPA. All outcomes will be collected by trained research assistants who are blinded to the study condition. The primary outcome, total MPA, will be assessed, in both individuals with AD and caregivers, by Actigraph at baseline, 3, 6, 12 & 18 mos. All secondary/exploratory outcomes will be assessed at the individual with AD or caregivers' home at baselines, 6, 12, and 18 months. Secondary outcomes for the individual with AD are sedentary time, functional fitness, activities of daily living, quality of life, residential transition, and cognitive function. Secondary outcomes for the caregiver are sedentary time, functional fitness, quality of life, and caregiver burden. The exploratory outcomes are age, sex, BMI, attendance (exercise/support sessions), use of pre-recorded videos, PA self-monitoring, peer interactions during group sessions, caregiver support, type and quality of dyadic relationship, and number of caregivers. This trial is powered to detect a between-arm difference (RGV vs. EUC) of 10 min./d. in MPA. This difference represents additional 70 mins. of MPA/wk. in the RGV arm. Power analysis shows that 84 participants (42/arm) would provide 81% power to test an overall between-arm difference across time, i.e., group effect. This sample size would also provide ≥ 80% power to detect a between-arm difference in change, i.e., group-by-time interaction, as small as f = 0.10. Thus, conservatively assuming a high attrition rate of 20%, the study team will recruit 100 dyads at baseline to assure the final sample size requirements are achieved, i.e., final N ≥ 84 (thus, power ≥ 80%) with attrition up to 20%. General mixed modeling for repeated measures will be utilized to evaluate the primary aim to compare total MPA (min/wk.) across the 6 mo. active intervention in adults with AD and their caregiver randomized to the RGV and EUC interventions. A similar mixed modeling analysis will be conducted to examine the secondary aim. General mixed models will be fitted for the two arms combined to examine the association for the process variables/participant characteristics with MPA. However, if there is a significant between-arm difference in MPA, the investigators will determine whether the process variables/participant characteristics amplify or attenuate the RGV effect, i.e., moderation, by testing a 2-way interaction with the group effect and/or a 3-way interaction with the group-by-time interaction term.