Improving health behaviors and symptoms in youth with mental health disorders: The GamerFit RCT

We tested GamerFit, a theory-based health coaching and exergaming intervention delivered via mHealth app for feasibility and preliminary efficacy to improve physical activity (PA), sleep, perceptions of barriers and support, and mental health outcomes in youth with heterogeneous/comorbid mental health disorders (HCMHD).

A convenience sample of youth ages 13-17 with HCMHD were recruited via clinical referral, listservs, and media platforms. The primary outcome of feasibility was assessed relative to the original intervention, which was not delivered via mHealth app. Groupings were randomized 1:1 to the 12-wk GamerFit intervention arm (GamerFit) or active comparator arm (AC) using single-blind design to assess preliminary efficacy. GamerFit participants used the mHealth app and a Fitbit^™ to follow a progressive exergaming and gamified step program. AC participants were given PA/sleep tips and Fitbit^™ to track PA/sleep. Intervention feasibility was assessed for the GamerFit group using process data and parental/participant report; PA duration/intensity (actigraphy, self-report), sleep duration/quality (actigraphy, self-report), sleep hygiene, perceptions of social support, self-regulation, positive/negative affect, and global quality of life were measured at wks 0, 12 and 16. Mixed effects linear models were used to account for the repeated measures correlation over time with an unstructured covariance matrix. The covariates in the model included the main effect for time and treatment as well as the interaction of these effects.

62 participants were randomized (15.0 ± 1.5 avg age, 24 % female-identifying, 21 % non-white, 74 % on medication). GamerFit averaged 88 % coaching attendance, 34.8 min/wk exergaming, and 8033 steps/day, exceeding the original intervention's feasibility benchmarks for coaching attendance and steps per day, but not exergaming minutes per week. Parental/participant acceptability/accessibility was high. Days/wk of PA increased significantly in GamerFit vs. AC (diff 2.0 ± 0.9, 95 % CI, p = 0.04). GamerFit significantly improved sleep hygiene, sleep quality, barriers to exercise, emotional and informational support, affect, and reduced unhealthy days; AC did not. GamerFit showed clinically meaningful improvements on all self-regulation sub-scales; AC did not.

GamerFit shows promise to improve health behaviors and outcomes among youth with HCMHD. An effectiveness RCT should take place in a more racially/ethnically diverse population.