0162 Sleep Moderated Intervention Efficacy in Two Randomized Controlled Trials Aimed at Increasing Physical Activity Among Inactive Latinas

Abstract Introduction Sleep and physical activity (PA) have a bidirectional relationship, with regular PA improving sleep quality and adequate sleep enhancing energy, motivation, and performance for PA. Sufficient sleep and PA improve health across various domains but many women, especially Latinas, do not meet sleep and/or PA guidelines. While targeted interventions have been designed and tested, research is needed to elucidate determinants of intervention success. This study examined sleep as a possible moderator of the intervention effect on PA outcomes in two randomized controlled trials (RCTs) designed to increase moderate-to-vigorous PA (MVPA) among inactive Latinas. Methods We conducted a secondary analysis of data from two RCTs (n=205 and n=195). Using a series of generalized mixed effects models with subject-specific intercepts, we tested the hypothesis that sleep moderated the intervention effect on PA outcomes. Models simultaneously regressed minutes/week of MVPA at each follow-up on time, group, time*group, sleep, time*sleep, group*sleep, and time*group*sleep. Interest was in understanding if effects of the intervention differed by sleep and if differential effects based on sleep also differed over time. Sleep was quantified both continuously and dichotomously (whether or not participants slept at least 7 hours). Analyses were run separately for each RCT to help establish consistency of findings. Results Sleep (whether operationalized as continuous or dichotomous) was a significant moderator of intervention effect on PA outcomes over time in both RCTs, p’s<.05. The intervention effect was stronger for participants who slept at least 7 hours per day; longer sleep led to increased experienced benefits from the intervention whereas participants with insufficient sleep performed similarly across time regardless of group assignment. Conclusion To our knowledge, this is the first study to examine sleep as an effect modifier of PA intervention success. Given our findings, that sleep indeed moderated the intervention effect in two different RCTs with different samples, future PA interventions should consider tailoring intervention content or materials based on characteristics of participant sleep. The implications of our findings are particularly relevant for PA interventions targeting Latinas since they are at high risk for both insufficient sleep and PA. Support (if any) National Institutes of Health/National Cancer Institute (Grant Numbers: 5R01CA159954 and 2R01CA159954)