Testing the Shamiri Intervention and Its Components With Kenyan Adolescents During the COVID-19 Pandemic: Outcomes of a Universal, 5-Arm Randomized Controlled Trial.

Abstract

Objective: Mental health problems are prevalent among African adolescents, but professional treatment capacity is limited. Shamiri, an efficient lay provider-delivered intervention, has significantly reduced depression and anxiety symptoms in previous randomized controlled trials (RCTs). This trial investigated effects of the full Shamiri intervention and its components (growth-only, gratitude-only, and values-only) against a study skills control.

Method: In a 5-group RCT with adolescents from Kenyan high schools, anxiety, depression, and well-being were self-reported through 8-month follow-up. The RCT occurred immediately after an unanticipated government-mandated COVID-19 shutdown forced 3 years of schoolwork into 2 years, escalating academic pressures.

Results: Participants (N = 1,252; 48.72% female) were allocated to: growth (n = 249), gratitude (n = 237), values (n = 265), Shamiri (n = 250), and study skills (n = 251) conditions. Longitudinal multilevel models showed that, across all conditions, anxiety scores significantly improved at midpoint (B = -0.847), end point (B = -2.948), 1-month (B = -1.587), 3-month (B = -2.374), and 8-month (B = -1.917) follow-ups. Depression scores also improved significantly at midpoint (B = -0.796), end point (B = -3.126), 1-month (B = -2.382), 3-month (B = -2.521), and 8-month (B = -2.237) follow-ups. Well-being scores improved significantly at midpoint (B = 1.73), end point (B = 3.44), 1-month (B = 2.21), 3-month (B = 1.78), and 8-month (B = 1.59) follow-ups. Symptom reduction with Shamiri matched that of pre-COVID-19 trials, but symptom reduction with study skills far outpaced that of trials before the COVID-19-related school shutdown (31% greater anxiety reduction and 60% greater depression reduction). Thus, in contrast to previous RCTs, this COVID-19-era trial showed no significant differences between outcomes in any intervention and active control groups.

Conclusion: Our RCT conducted during a post-COVID-19 period of heightened academic pressure produced unexpected results. Improvements in youth-reported anxiety and depression were consistent with previous trials for Shamiri, but markedly larger than in previous trials for study skills. Control interventions teaching life skills may produce mental health benefits when they convey skills of particular contextual relevance.

Plain language summary: A large five-group randomized controlled trial involved comparing the Shamiri Intervention to its component interventions (growth mindset, gratitude, and values affirmation) and a study-skills control. In contrast to previous studies of Shamiri, similar effects were observed across all groups (p<.05). This trial was conducted right after an unanticipated government-mandated COVID shutdown forced three years of schoolwork into two. Benchmarking analyses against previous trials showed approximately equal effects of Shamiri over time, but a 31% greater anxiety reduction and 60% greater depression reduction for the study-skills condition; this highlights the potential of interventions teaching highly relevant life-skills for improving mental health.

Clinical trial registration: Five-Arm Shamiri Trial; https://pactr.samrc.ac.za/; PACTR202104716135752.