Peer-led vs teacher-led: what the research actually says about school wellbeing programs
Most schools trying to stand up a wellbeing program have to make an early design decision: who delivers it. Teachers, school counsellors, a clinical partner, or trained peers. The conversation gets pitched as a question of resourcing (peer-led is cheaper) or of safeguarding (adult-led is safer), but the real question is which structure actually helps students.
The short answer from the evidence: peer-led programs show strong engagement signals in older adolescents, particularly for help-seeking and reducing stigma. Clinically meaningful change in depression or anxiety symptoms? Weaker, and inconsistent across the studies. If you want symptom-level change, the evidence behind teacher-led and clinician-led programs is stronger but still not uniform. And if you're measuring anything, the choice of screener matters more than most program coordinators realize.
Here's the longer version.
What the PLOS One systematic review found
The single best starting point on this question is King and Fazel's 2021 systematic review in PLOS One, Examining the mental health outcomes of school-based peer-led interventions on young people: A scoping review of range and a systematic review of effectiveness. The review pulled together 11 studies published between 1990 and 2020 that evaluated peer-led mental health interventions in secondary schools across multiple countries. Fully cited, open access, worth reading in full.
Headline findings from the review:
- Peer-led interventions consistently improved help-seeking attitudes and knowledge about mental health. Students who participated were more likely to know when and where to seek help and reported lower stigma around talking about mental health. These effects were measured across most of the included studies.
- The evidence for change in depression and anxiety symptom scores was weaker. Some studies found small improvements, some found null effects, and a minority found that intervention groups and control groups moved in parallel. The review authors were explicit that symptom-level change shouldn't be inferred from the broader help-seeking signal.
- Methodological quality varied widely. Several studies lacked active control groups, used self-designed outcome measures, or ran for too short a duration to detect longer-term effects. The review flagged this as the main reason confidence in effect sizes should stay modest.
- Program structure mattered. Interventions that trained peer leaders for longer (20+ hours) and provided ongoing supervision outperformed one-off workshops. This is consistent with the broader implementation-science literature: dose and fidelity are the biggest levers.
The review's recommendation, worded directly: peer-led interventions have a place in school wellbeing programs, particularly as a way to reduce stigma and improve help-seeking, but should not be relied upon as the primary vehicle for symptom reduction without further evidence.
How this compares to teacher-led and clinician-led programs
Teacher-led cognitive-behavioral interventions have a longer evidence base. The Penn Resilience Program, for example, has been evaluated in multiple randomized trials and shows small-to-moderate effects on depressive symptoms in adolescents (APA summary of the evidence). Similar trials of Coping Cat for childhood anxiety, delivered by trained school staff, show stronger symptom-level effects than the peer-led comparators.
Clinician-led programs, typically delivered by a school-based psychologist or a visiting therapist, consistently outperform both peer-led and teacher-led models on clinical outcomes. This isn't surprising: a trained clinician running a manualized CBT protocol is the closest analogue to what gets studied in clinical trials. The catch is that school districts rarely have the clinician headcount to deliver a full program to a student population, which is why this isn't the default.
A reasonable synthesis: peer-led is good for reach and stigma, teacher-led is good for cost-effective symptom change when fidelity is high, clinician-led is the best individual option but constrained by availability. Real-world programs increasingly combine the three, with peers handling awareness and help-seeking, teachers running structured curricula, and clinicians taking the small number of students who screen positive for targeted intervention.
The PHQ-9 vs PHQ-A question (and why it matters here)
If you're going to measure the effect of a wellbeing program on depression symptoms, you have to pick a screener. The most common one in adult clinical settings is the PHQ-9, a nine-item self-report instrument validated in primary care populations. It's fast (under two minutes), free, translated into dozens of languages, and clinicians are used to interpreting its scores.
There's a problem, though. The PHQ-9 was validated on adults. The specific wording of several items, the reference period of two weeks, and the suicide-risk item (item 9) were developed for an adult population. When schools use the PHQ-9 with students, particularly students under 13, they're using an instrument outside its validation population.
The PHQ-A (sometimes called the PHQ-9 Modified for Adolescents, or PHQ-9A) is a teen-adapted version. Johnson, Harris, Spitzer, and Williams introduced it in 2002 in the Journal of Adolescent Health (DOI link). The adaptation involves rewording several items for age-appropriate language, adding items related to school functioning, and clarifying the suicide-risk item. Validation studies in adolescent populations show the PHQ-A performs adequately, though the evidence base is thinner than the adult PHQ-9.
Why does this matter for a peer-led vs teacher-led decision? Because the way a program's effect size is measured depends on the instrument. A peer-led intervention that produces a real effect on adolescent depression may register as null on an adult-validated PHQ-9 and as a small-but-meaningful effect on a PHQ-A. Several of the studies in the 2021 systematic review used adult instruments with adolescents, which is one of the methodological flags the review authors raised.
The AACAP screening tools resource has a longer list of adolescent-validated alternatives, including the RCADS and the Strengths and Difficulties Questionnaire (SDQ). The choice isn't a purely academic one. It determines whether the program's effect shows up at all.
Practical implications for a counsellor designing a program
What does all of this add up to if you're a school counsellor or district lead trying to make real decisions this term? A few things.
1. Don't pick peer-led purely because it's cheaper
Peer-led programs have real strengths and real limits. The strength is reach: you can train 20 peer leaders and touch 400 students. The limit is symptom-level change, which the evidence doesn't support as a primary mechanism. If your goal is to reduce stigma and increase help-seeking (worthy goals on their own), peer-led is a defensible choice with evidence behind it. If your goal is to reduce student depression scores, peer-led alone probably won't do it, and you should plan a multi-tier approach.
2. Train your peer leaders more than you think you need to
The programs that showed the strongest effects in the 2021 review trained peer leaders for 20+ hours and provided ongoing supervision. The ones that showed null effects frequently trained peer leaders for under 10 hours with no follow-up. If your budget constrains training dose, shrink the program's reach rather than the training depth. Fewer well-trained peer leaders outperform more thinly-trained ones.
3. Match your screener to your population
If your students are under 15, use an adolescent-validated screener. The PHQ-A and RCADS are both reasonable starting points. If your students are 15-18, the adult PHQ-9 is defensible but you should flag the age-appropriateness question in any evaluation write-up. Using an adult instrument on 12-year-olds and then reporting the result as if it were comparable to an adult population's scores is a common methodological trap. Vendors who sell wellbeing platforms frequently make this mistake. Ask for the validation citation before deploying any screener at scale.
4. Plan for a stepped-care model
The combined evidence from peer-led, teacher-led, and clinician-led programs points to a stepped-care approach. Universal peer-led education for all students (awareness, stigma reduction, help-seeking). Targeted teacher-led curricula for students showing early signs on a screener (a validated one, see above). Clinician-led intervention for the smaller group with elevated symptom scores. This isn't a new idea, it's roughly the model described in the SAMHSA comprehensive approach to school mental health, but the evidence keeps reinforcing that no single tier works well on its own.
5. Commit to measuring more than symptom change
If you're running a peer-led component and you measure only depression scores, the evidence says you'll likely under-count the program's real effect. Help-seeking behavior, knowledge of warning signs, stigma attitudes, and willingness to refer a friend are all mediators that peer-led programs move reliably. Include them in your evaluation. Instruments like the General Help-Seeking Questionnaire are short, free, and validated for adolescent use.
The gaps in the evidence
Worth being honest about what we still don't know:
- Cultural and contextual generalizability. Most of the studies in the 2021 review were conducted in North America, the UK, and Australia. Whether peer-led models transfer to different cultural contexts is less clear.
- Long-term outcomes. Most studies measured at 3 or 6 months post-intervention. Whether the help-seeking effects persist at 12 or 24 months is largely unknown.
- Dose-response relationships. We know 20+ hours of peer training outperforms 10 hours, but the optimal dose and the diminishing-returns point haven't been mapped.
- Digital adaptations. Whether a peer-led model delivered via a digital platform produces the same effects as in-person peer-led is an open question, and a commercially important one given the direction of the sector.
Those gaps matter because they bound how confident any practitioner should be in a program design choice. The evidence is strong enough to rule some choices out (peer-led with 4-hour training, or an adult PHQ-9 on 12-year-olds), and strong enough to support a stepped-care combination model. It isn't strong enough to prescribe one specific curriculum or screener over another with certainty.
One more thing: vendor claims
If you're evaluating a digital wellbeing platform for your school and the vendor claims "peer-led, evidence-based," ask three follow-ups. First: which studies specifically, cited to a journal. Second: what was the validation population for the outcome measure used, and does it match your students' age. Third: what's the training dose for the peer leaders in their model, and is it closer to the 20-hour benchmark or the 4-hour one.
A vendor with real evidence will answer those three cleanly. A vendor running on marketing language will pivot, reframe, or offer testimonials. The pivot is the answer.
Further reading
- King, T., & Fazel, M. (2021). Examining the mental health outcomes of school-based peer-led interventions on young people. PLOS One.
- Johnson, Harris, Spitzer, & Williams (2002). The patient health questionnaire for adolescents. Journal of Adolescent Health.
- AACAP Mental Health Screening Tools, with separate sections for adult and adolescent validation.
- SAMHSA: Advancing Mental Health Care in Schools, a practitioner-facing overview of the stepped-care model.
- RCADS (Revised Children's Anxiety and Depression Scale) via the UK's CORC, free to use.
