Abstract P200: Effects Of A 2-year Randomized Controlled Trial Physical Activity Intervention On Accelerometer-measured Sleep Characteristics Among Older Adults In Senior Center Communities

Abstract

Background: Targeting daily increases in physical activity (PA) is a common lifestyle strategy for cardiovascular disease (CVD) risk reduction. While less targeted in lifestyle interventions, poor sleep is also consistently linked to CVD. Emerging evidence connects changes in daily PA to beneficial changes in sleep; however, little is known about the impact of community-based PA interventions on habitual sleep among older adults. We leveraged accelerometer data from the Peer Empowerment Program for Physical Activity in Low Income & Minority Seniors (PEP4PA) randomized controlled trial to examine the effect of a PA intervention on accelerometer-measured sleep characteristics among older adults in low income and diverse senior center communities. Methods: PEP4PA participants (N=476; mean age=71.0 ± 9 years, 75% female, 60% low income, 36% non-white) were recruited from senior centers and randomized to a healthy control condition(n=209) or a peer-lead walking intervention (N=267). The peer-led PA intervention included self-monitoring techniques, health coach counseling, group walks, and community advocacy to improve walking conditions. Participants wore Actigraph GT3X+ accelerometers on the wrist at baseline, 6, 12, and 24 months. Data were processed for sleep characteristics including weekly average sleep duration and sleep efficiency. Mixed effects regression models provided estimates of the intervention effect on sleep characteristics at each time point. Models were adjusted for age, sex, and race. Three way interaction terms (time*treatment*sleep duration status) were used to assess differential intervention effects by short (<7) and long (>8 hours) sleep duration status at baseline. Results: At baseline, the average sleep duration of participants was 7 hours (SD: 1.07 hours) and the average sleep efficiency was 87.0% (SD: 6.2%). Compared to participants in the control group, participants in the intervention group had shorter sleep durations (7.1 hours vs 7.0 hours; p value: 0.08) and poorer sleep efficiency 86.0% vs 88.2%; p<0.001) at baseline. At 12 or 24-months, there was no intervention effect on participants’ average sleep duration or efficiency. Although not significant, at 6 months, long sleepers in the intervention group experienced an average 6-minute reduction in sleep duration (-B: 6.25, p=0.45) compared to those in the control group. Conclusions: Over the 2-year period, the PA intervention did not appear to have an impact on participants’ average sleep duration or quality. More investigation is needed into the temporal and daily relations of PA and sleep among older adults in interventions to understand if increases in daily PA are associated with better sleep characteristics at night.