Do all adolescents benefit equally from school-based mental health interventions? A subgroup analysis by depression severity at 12 and 18 months follow-up.

Background: School-based mental health interventions show modest effects in reducing depression symptoms. However, most trials report only average effects that may mask important differences across symptom severity levels, which are important in assessing long term risk, particularly as sub-threshold depression is a risk factor for clinical depression. Aims: This study examined whether the effects of DISCOVER, a brief CBT-based self-referral workshop for adolescents, at 12- and 18-month follow-ups differed according to participants' baseline depression severity (clinical, subthreshold, or normal range). Methods: This is an exploratory secondary analysis of the BESST cluster randomised controlled trial involving 900 adolescents from 57 UK schools randomised to DISCOVER or treatment-as-usual (TAU). Participants were categorised by baseline depression severity using established Mood and Feelings Questionnaire (MFQ) cut-offs: clinical (>27), subthreshold (17-27), or normal (<17). We conducted multilevel mixed-effects multinomial logistic regression to examine transitions between severity categories and mixed-effects linear models for continuous outcomes (depression, anxiety, wellbeing, sleep quality, resilience) for each severity group. Results: Among adolescents with clinical-level depression at baseline (DISCOVER n=142, TAU n=172), DISCOVER significantly increased the odds of achieving normal-range symptoms at 12 months (RRR = 3.12, 95% CI: 1.48-15.97, p=0.006) and 18 months (RRR = 2.63, 95% CI: 1.11-6.24, p=0.029). Small-to-moderate effects were maintained for depression (d=0.44 at 12 months, d=0.30 at 18 months) and anxiety (d=0.38 at 12 months, d=0.30 at 18 months). For adolescents with subthreshold depression symptoms at baseline (DISCOVER n=170, TAU n=157), DISCOVER did not show benefits on depression but demonstrated specific anxiety benefits at 12 months (d=0.34, p=0.004). Conclusions: DISCOVER produces small to moderate sustained benefits for adolescents with clinical-level depression, with effects maintained through critical post-school transitions. The intervention shows more limited benefits for those with subthreshold symptoms, achieving specific anxiety reduction but not broader improvement. These findings suggest that self-referral school-based interventions are effective over the longer term for adolescents with clinical-level symptoms, while highlighting the need for tailored approaches for subthreshold presentations.