Health behaviors, including physical activity (PA), are important across the life course for the prevention of chronic disease, with young adulthood representing a key period for establishing lifelong health behaviors [1]. PA tends to decline during this transitory period [2], yet few studies have reported device-based assessments of PA in young adults over a long-term period.

We conducted a pilot randomized controlled trial, Motivational Interviewing to Promote Healthy Behaviors for Obesity Prevention in Young Adults (MOTIVATE), to determine the feasibility of a 6-month behavioral and educational motivational interviewing (MI)–based intervention. MI is a person-centered counseling method aimed at enhancing intrinsic motivation for behavior change [3]. This paper examines a secondary aim of the trial by evaluating the feasibility of using wearable devices for monitoring PA throughout the 6-month study period. Feasibility outcomes included adherence to wearing the device and descriptive summaries of activity types.

Of 120 individuals assessed for eligibility, 101 participants were randomized, and 5 were excluded due to missing Fitabase data (Supplementary Figure 1 in Multimedia Appendix 1), resulting in a final analytic sample of 96 participants (46 control and 50 intervention). The flow diagram for the parent trial is shown in Supplementary Figure 2 in Multimedia Appendix 1.

Baseline characteristics were similar in control and intervention groups (Table 1). Among 96 participants, the mean age was 22.4 (SD 3.6) years, with 38.5% (n=37) identifying as men, 51.0% (n=49) as women, 2.1% (n=2) as genderqueer/nonbinary, and 8.3% (n=8) preferring not to disclose their gender. At baseline, 60.4% (n=58) of participants reported never having used a fitness tracker or health application. By the end of the study, 57.3% (n=55) indicated that using the Fitbit helped them improve their health behaviors (Supplementary Table 1 in Multimedia Appendix 1).

Most participants (91.7%) had at least one adherent day during the study, including 40 of 46 in the control group and 48 of 50 in the intervention group. The median adherent days were 73.0 (IQR 122.0) for the control group and 80.5 (IQR 121.5) for the intervention group. Adherence declined steadily over the 6-month study period in both groups. The intervention group showed higher adherence than the control group, except during the fifth month, when adherence was similar in both groups (Figure 1A). In terms of PA intensity, there was some variation in light activity, whereas time spent in moderate and vigorous activity remained relatively stable in both groups (Figure 1B). On average, participants spent 2.0 to 2.6 hours per day in light activity, 0.2 to 0.3 hours in moderate activity, and 0.2 to 0.4 hours in vigorous activity.

We observed variability in Fitbit adherence, with the intervention group demonstrating higher adherence than the control group. Adherence declined substantially in both groups over the 6 months, but moderate and vigorous PA levels remained comparable. Additionally, more than half of the participants reported that using the Fitbit helped improve their health behaviors.

These findings are broadly consistent with a previous review among healthy adults that reported that wearable activity trackers were associated with modest increases in PA [9], suggesting that such devices can positively influence behavior, particularly in the short-term. In contrast, a review of children and adolescents found no significant improvements in moderate-to-vigorous activity compared with control groups [10]. This highlights that the effectiveness of wearable devices may vary by population and context. Nonadherence, noncompliance, and reduced engagement with the device over time were common limitations across studies [10], consistent with our observations.

While wearable devices allow continuous monitoring of PA, our results raise questions regarding their long-term feasibility in trial settings. Although moderate and vigorous PA levels appeared stable over time, this may reflect selection bias, as participants with longer follow-up periods were more engaged. Despite gift cards for questionnaire completion (twice per month), adherence declined, suggesting that longer interventions may require additional engagement strategies such as personalized feedback, behavioral prompts, or counseling-based approaches. Given the short duration of previous wearable device studies [10], future trials with longer intervention periods, such as our 6-month design, are needed to assess sustained behavior change. Our findings provide valuable insight into the feasibility of integrating wearables into behavior change interventions. Overall, wearables show clear potential for scalable application in public health, postsecondary education, and primary care settings, where early prevention is crucial.