Working Women Walking Program: A Sequential Multiple Assignment Randomized Trial

Background : Different interventions have shown efficacy in improving physical activity in women. This study aimed to determine the most effective adaptive intervention combining 4 efficacious treatments (Fitbit, text messages, personal calls, and group meetings) for improving physical activity. Methods : The Working Women Walking program used a Sequential Multiple Assignment Randomized Trial (SMART) design with 3 phases: initial (weeks 1–8), augmented (weeks 9–34), and maintenance (weeks 35–50). Low-active women aged 18–70 working at an academic medical center were recruited. In the initial phase, participants were randomized to Fitbit or Fitbit + text messages. After 8 weeks, nonresponders were rerandomized to an augmented treatment (personal calls or group meetings). The primary outcomes (steps/day and minutes of moderate/vigorous physical activity [MVPA]/day via ActiGraph) were assessed at baseline and 8, 34, and 50 weeks. Results : The study had 301 women (age: 45.1 [11.6] y). During the initial phase, there were no differences by initial treatment for change in steps ( P = .78) or MVPA ( P = .60). During the augmented phase, there were no differences by augmented treatment among nonresponders for change in steps ( P = .95) or MVPA ( P = .78). A significant overall increase was seen in steps (752/d; P < .001, d = 0.56) and MVPA at 8 weeks (4.1 min/d; P < .001, d = 0.48), and this was sustained at 34 and 50 weeks. Conclusions : Text messages, personal calls, and group meetings did not lead to increased physical activity. Using Fitbit and goal setting (a constant) appeared to have some benefit for many of the women in improving physical activity.