Review: Meta-analysis on mindfulness-based interventions for adolescents' stress, depression, and anxiety in school settings: a cautionary tale.

BACKGROUND: Mindfulness-based interventions (MBIs) have been applied in school settings for adolescents with symptoms of stress, depression, and anxiety; however, general conclusions of the efficacy of such interventions remain unclear. This meta-analysis reviewed randomized-controlled MBI trials for stress, depression, and anxiety in school settings. METHODS: One hundred one records were included after removing duplicates. Nine studies met inclusion criteria, totalling 5046 adolescents aged 12-18. Eighteen comparisons between an MBI and a control group were analyzed. RESULTS: The overall effect for symptoms of the 17 observations including stress, depression, and anxiety resulted in a significant improvement with a small effect size (k = 17, n = 3721, Hedge's g = .33, CI 95% .17-.49 p < .01). Subgroup analysis revealed that when MBIs were compared to an active control group effects were not significant (k = 5, n = 2753, Hedge's g = .27, CI 95% -.03-.57 p = .08), and when compared to an inactive control group the effect was significant with a small effect size (k = 5, n = 1065, Hedge's g = .38, CI 95% .02-.75 p < .05). Analysis of the interventions on a per symptom basis yielded a significant and moderate effect size for perceived stress (k = 7, n = 1116, Hedge's g = .55, CI 95% .31-.79 p < .01); however, there were no significant effects for depression (k = 6, n = 3172, Hedge's g = .20, CI 95% -.05-.44 p < .01) and anxiety (k = 4, n = 837, Hedge's g = .19, CI 95% -.14-.53 p = .25). CONCLUSIONS: The impact of MBIs in school settings for adolescents yielded a significant improvement for stress, but did not for depression and anxiety. The effects were significant when compared to inactive controls, but not when compared to active controls. Implications of these findings are discussed.