4 - (SYM 105) Testing the Shamiri Intervention and Its Components for Kenyan Adolescents: A Five-arm Randomized Controlled Trial

Background: Brief, school-based interventions that teach CBT principles through positively focused intervention components have shown promise in settings with elevated stigma and few mental health professionals. One such intervention, Shamiri, has significantly reduced depression and anxiety symptoms in several randomized controlled trials (RCTs) with Kenyan high school students (Osborn et al., 2020; Osborn & Venturo-Conerly et al., 2022). Shamiri includes three intervention components, growth mindset, gratitude, and values affirmation, which may have differential effects on mental health. In a previous four-arm RCT, single session versions of these three interventions were each tested, and values and growth outperformed an active control at reducing anxiety (Venturo-Conerly et al., 2022).

Methods: This well powered five-arm RCT was conducted to compare the effectiveness of each component (growth, gratitude, and values) to the full Shamiri intervention and to an active control. Anxiety (GAD-7), depression (PHQ-8), and wellbeing (S-WEMWBS) scores were reported by all N=1,252 participants from four Kenyan high schools at baseline, midpoint, and 1-, 3-, and 8-month follow-up. Multiple imputation was used to address missing data, and mixed-effects models were run in R to analyze change in each symptom outcome over time, controlling for age, gender, and baseline score.

Results: Over time all groups showed statistically significant (p< .05) improvements on measures of anxiety, depression, and wellbeing, yet there were no significant differences among the different conditions in symptom improvement over time. Additionally, outcomes were not significantly related to intervention fidelity (p’s >0.10), nor did findings differ when analyzing boys’ and girls’ schools separately.

Discussion: This fully-powered study showed no differences between any of the five intervention groups tested. This stands in contrast to past research, which has shown both the Shamiri intervention and the values and growth conditions outperforming the active control (Osborn et al., 2020; Osborn & Venturo-Conerly, 2022; Venturo-Conerly et al., 2022). The intervention was delivered early in the COVID-19 pandemic when intervention training and delivery were particularly challenging, and two schools experienced fires during the study period, yet intervention fidelity and school appear not to have influenced effects. Avenues for continuing to investigate under what conditions interventions show effectiveness will be discussed.