Abstract Background Sequential physical activity interventions hold promise in developing effective interventions when adapted to participants’ physical activity. Adaptive interventions can be further strengthened by understanding the moderators of their effects on physical activity outcomes. Purpose To determine whether baseline personal characteristics moderated the impact of adaptive interventions on physical activity in the Working Women Walking trial. Methods This study analyzed data from a sequential multiple assignment randomized trial (SMART) targeting 18- to 70-year-old women employed at an urban medical center who self-reported low physical activity (N = 301). Four adaptive interventions included combinations of Initial Treatments (Fitbit or Fitbit + Text Messages) and Augmented Treatments for nonresponders (Motivational Calls or Group Meetings). Personal characteristics, including demographics, health status, self-reported physical activity, and neighborhood walkability, were measured at baseline. Moderation of intervention effects was tested across ActiGraph-assessed physical activity at baseline, 9-10 weeks (end of Initial Phase), 35-36 weeks (end of Augmented Phase), and 49-50 weeks (end of Maintenance). Results There was no evidence that personal characteristics moderated the differential impact of the Initial or Augmented Treatments on device-assessed steps or moderate-to-vigorous physical activity. Independent of the Initial Treatment, Black race, higher baseline body mass index, and higher baseline self-reported moderate-to-vigorous physical activity were associated with lower increases in physical activity (P .05) during the Initial Phase. Lower neighborhood walkability was associated with relative decreases in physical activity during the Maintenance Phase. Conclusions Although personal characteristics may have altered the degree to which participants increased their physical activity, they did not alter the differential impact of the adaptive interventions. Clinical Trial Registration Number ClinicalTrials.gov identifier: NCT03558828.
Daniel et al. (Thu,) studied this question.