Abstract
BACKGROUND: Digital health interventions and gamification hold promise for managing chronic diseases, but evidence comparing their efficacy, long-term effectiveness, and cost-efficiency with those of usual care is limited. Moreover, there is a growing need for randomized controlled trials (RCTs) evaluating digital physical activity interventions to incorporate idiographic approaches and intensive longitudinal assessments that capture individual variability and the dynamic nature of behavior change. OBJECTIVE: This 2-arm parallel RCT with embedded N-of-1 analyses examined whether a digital intervention integrating gamification and telecoaching (Kiplin program) outperformed a supervised, face-to-face, adapted physical activity program (usual care) in improving physical activity, clinical outcomes, and cost-effectiveness among adults with obesity and type 2 diabetes (T2D). METHODS: We randomized (1:1) 50 patients with obesity or T2D (mean age 47.90, SD 12.49 years; 37/50, 74% female) referred to the University Hospital of Clermont-Ferrand, France, to either the Kiplin digital program or the usual care group. Both programs lasted 3 months, with a 6-month follow-up. The Kiplin intervention included 2 face-to-face and 20 online supervised sessions and 3 mobile app games, whereas the control group completed a standard hospital-based adapted physical activity (APA) program with 3 individual face-to-face, supervised sessions per week (36 sessions). The primary outcome was the change in daily step count, measured objectively and continuously via wearable devices from baseline to the end of the intervention. Secondary outcomes included changes in accelerometer-assessed physical activity, quality of life, body composition, physical capacities, and daily steps over 9 months. Program adherence was also evaluated. Mixed-effects models and generative additive models were conducted to analyze both between- and within-person evolutions in physical activity. A cost-utility analysis was computed to compare the cost-effectiveness of the programs. RESULTS: Compared with usual care, Kiplin participants achieved greater increases in daily steps during both the 3-month intervention (+1085 steps/day) and follow-up (+1775 steps/day), with sustained effects over time. Idiographic analyses revealed marked heterogeneity, showing substantial between- and within-person variability, with 9 participants exhibiting nonlinear patterns and divergent individual trajectories, with some participants showing no improvement. No significant group differences were observed in secondary clinical outcomes, except for change in moderate-to-vigorous physical activity, in favor of the Kiplin group. Cost-utility analyses showed no significant difference between programs. Kiplin participants attended an average of 14.68 of 22 possible APA sessions and engaged in an average of 2.6 games. In contrast, usual care patients attended an average of 30.27 of 36 APA sessions. CONCLUSIONS: This study demonstrates the potential of digital gamified interventions to promote and sustain physical activity, offering an alternative to face-to-face programs. However, individual heterogeneity in the response to the intervention highlights the need for screening tools and tailored approaches. Further large-scale studies are warranted to evaluate the long-term clinical and economic impact of such interventions.