Effects of Using a Smartphone App Combined With Behavior Change Techniques on the Level of Physical Activity Among Adults and Older Adults: Sequential Multiple Assignment Randomized Trial.

Background: The use of tools, such as smartphone apps, to increase the level of physical activity (PA) decreases over time. Adaptive intervention trials have been recommended to test technology-based interventions owing to the possibility of adapting interventions based on individual responses. Objective: This study aimed to investigate the effects of using a smartphone app combined with behavior change techniques on the PA level in adults and older adults (assessed using the step count). Moreover, the study investigated the time spent in sedentary behavior and time spent in moderate-to-vigorous PA (MVPA). Methods: In this single-blinded, sequential multiple assignment randomized trial, participants were randomized into 3 groups during a 24-week intervention (group 1: app with tailored messages; group 2: app with tailored messages plus gamification I; and control group: educational information). In the sixth week, participants from groups 1 and 2 were classified as responders and nonresponders according to their average daily step count. Nonresponders were rerandomized among the other groups, adding a second type of gamification (group 3: app with tailored messages plus gamification II). After another 6 weeks, participants were reassessed and advised to keep monitoring their step count with the app, but without interference from the researchers. Face-to-face assessments were conducted. The behavior change techniques included app features (goal setting, auto-monitoring, ranking, and virtual badges) and researcher-provided resources (tailored messages and in-person sessions of PA). The intervention effects were analyzed using linear mixed models. Results: The study included 53 participants (control group: n=17, group 1: n=17, group 2: n=19; mean age 44.0, SD 12.7 years). Groups 1 and 2 had 63% (10/16) and 47% (7/15) responders, respectively (P=.38). Regarding the PA level, participants from group 1 showed increases in the average daily step count at all assessments (final vs initial: B=797.2 steps/day, 95% CI 475.3-1119.1; P<.001; follow-up vs initial: B=2097.6 steps/day, 95% CI 1577.2-2618.1; P<.001). All participants showed a reduction in the time spent in sedentary behavior at the final assessment compared with the initial assessment (B=-70.8 min/week, 95% CI -88.8 to -52.9; P<.001), without differences among groups. The time spent in MVPA varied across time among all participants. Regardless of the initial group and allocation in the second randomization, responders from groups 1 and 2 showed a constant increase in the average daily step count (week 6 vs week 1: B=1548.0 steps/day, 95% CI 1407.4-1688.6; P<.001; week 12 vs week 1: B=1720.3 steps/day, 95% CI 1568.8-1871.7; P<.001; week 12 vs week 6: B=172.3, 95% CI 20.8-323.8; P=.03). Conclusions: The adaptive intervention protocol using a smartphone app with behavior change techniques increased participants' PA levels. Stepping up behavior change techniques and progressively offering new stimuli may contribute to a change in behavior regarding PA.