Research & Evidence
Can Peer Mentoring Improve Student Mental Health? What the Evidence Supports
Do high-school peer mentors improve mental health? Compare student peer-led studies, adult mentoring research, school belonging and what evidence really shows.
High-school peer mentoring may support connection and everyday school navigation, but research has not established that a typical student-to-student mentoring program consistently improves clinical mental-health outcomes. Direct peer-led school studies remain limited and mixed; many positive mentoring studies instead examine adults mentoring children, a different intervention. 1 3 4
This article reviews what the research actually tested, where it found benefits or limitations, and how schools can communicate evidence accurately without implying any study validated Tagapag. It is general research literacy, not a diagnosis, treatment, safety-response or clinical program guide.
At a glance
School peer mentoring, intergenerational mentoring, peer-led health programs and clinical peer support are not interchangeable.
A 2021 school peer-led review found a small, heterogeneous evidence base and mixed mental-health outcomes.
A 2025 review of peer-led lifestyle programs found no significant pooled benefit in measured mental-health outcomes, but it did not study conventional mentor pairs.
Adult–youth mentoring research may be positive without proving high-school peer-to-peer clinical effectiveness.
CDC connectedness statistics are associations, not causal evidence for a specific school program.
Program quality and student feedback matter, but student safety and clinical care stay with authorized adults.
1. Does high-school peer mentoring improve mental health?
Possibly for some social experiences, but clinical effectiveness is not established. Student mentors can welcome classmates, answer school questions and help them connect with activities. Direct evidence that typical high-school peer-to-peer mentoring reduces diagnosed depression or anxiety remains limited and mixed. 1 6
A positive student experience is important on its own. It should not be turned into a medical claim without a study of that specific intervention, appropriate outcome measures and comparisons. This evidence review is general education, not advice about treating an individual student's symptoms.
2. Four different approaches often get called 'peer support'
Student peer mentoring generally means students helping students navigate school. Intergenerational mentoring uses adult mentors. Peer-led lessons or campaigns train students to teach groups. Clinical peer support can involve trained workers drawing on lived mental-health experience in care settings. 3 7 8
These programs have distinct staffing, purposes, contact patterns and safety requirements. A study of a trained adult mentor or health educator does not establish what a student-to-student matching platform can accomplish. Start by identifying the exact program before comparing outcomes.
3. What did the 2021 direct peer-led evidence review find?
King and Fazel's 2021 systematic review mapped 54 school-based peer-led interventions addressing mental health or well-being. A narrower effectiveness review identified 11 studies with 2,239 participants. Results were mixed; reliable evidence of mental-health improvements was sparse. 1
Programs varied in age, activities, training and outcome measures. The review is directly relevant to claims that any student-led school intervention works, but it does not prove that all peer connections fail. It points to a need for better-defined, well-evaluated programs.
Sources for this section
4. Were benefits the same for mentors and students receiving help?
No. Among seven studies measuring peer leaders, two reported selected improvements such as self-esteem or lower social stress, while one also found increased guilt. Of five assessing peer recipients, two reported some positive outcomes, while another found more learning stress or lower mental-health scores. 1
A mentor's increased confidence is not evidence that the mentee's depression risk fell. Schools should measure the experiences of both sides and consider workload or unwanted responsibility, instead of treating leadership participation as automatically beneficial.
Sources for this section
5. What did a newer 2025 meta-analysis add?
A 2025 review examined seven controlled peer-led health-behavior interventions with 7,060 adolescents at 151 schools. Six involved activity and one diet; the available pooled estimates did not show statistically significant improvement in the measured mental-health and well-being outcomes. 2
These were not ordinary peer mentor pairings. The finding cautions against assuming that putting peers in charge produces a mental-health benefit, but it cannot directly estimate the impact of a school welcome or mentoring program. Null averages also do not exclude benefits in particular settings.
Sources for this section
6. Why do adult mentoring studies show more encouraging results?
A 2019 meta-analysis synthesized 70 studies involving 25,286 young people in intergenerational, one-to-one mentoring programs. It reported a statistically significant overall effect across youth outcomes, alongside expectations of modest effects and limitations in how programs were implemented. The youth averaged about 12 years old. 3
That literature may inform adult supervision and relationship quality, but it is not proof that two high-school students in an app-created match have the same results. Mentor age, preparation, continuity, setting and assessment tools all matter.
Sources for this section
7. What did an earlier school-based mentoring review report?
A 2012 review covered eight studies and 6,072 adolescents, with six contributing to a meta-analysis. Most measured impacts on academics, attendance, behavior, attitudes and self-esteem were very small or not statistically significant; the reported self-esteem estimate was near zero (g = 0.09). 4
This is historical research, not an evaluation of current digital peer mentoring. It does show why even programs situated inside schools should not automatically be described as evidence-based mental-health treatments. A school setting by itself is not a causal mechanism.
Sources for this section
8. Where does school connectedness fit?
CDC defines school connectedness broadly: students feel that adults and peers care about them and their learning. In the 2021 national high-school YRBS, 61.5% reported feeling connected to others at school; connected students had lower prevalence of several adverse experiences, including poor mental health. 5 9
This was a cross-sectional observational association, not a controlled test of a peer mentoring program. The number is 2021, published 2023, not a new 2025 estimate. It shows a compelling school-climate question without determining the direction of cause.
9. Can a successful match be equated with school belonging?
No. School connection depends on peers, adults, teaching, student activities, classroom norms and many other factors. A student might find a helpful mentor but still feel excluded from class; another might feel connected through friends and teams without formal mentoring. 5 6 9
CDC describes opportunities for group projects, respectful listening, peer discussion and welcoming activities. These are useful possible practices, not experimental evidence that one particular student-matching formula improves clinical outcomes.
10. Does clinical peer support evidence transfer to school volunteers?
A 2024 scoping review examined 15 studies of youth peer support in integrated mental-health services and related settings. These interventions often involved trained peer workers and lived experience of mental-health challenges, with varying outcomes and implementation conditions. 8
An ordinary high-school volunteer helping a new classmate locate clubs is not a clinical peer worker. A school should not present student mentors as therapists or import claims from specialized treatment services without evidence that the context and intervention are comparable.
Sources for this section
11. What study design can support causal claims?
A carefully designed comparison—sometimes a randomized assignment of schools or eligible participants—can test whether a specific program causes a measured difference. Baseline measures, follow-up, attrition, participant selection and prespecified outcomes matter. Studies with no comparison group offer a much weaker basis for causal statements. 1 2 3 4
Ask five questions: Who participated? What actually happened? What did the comparison group receive? What outcome was measured? How much uncertainty remained? An impressive sample size is no substitute for asking whether a study directly answers the claim.
Sources for this section
12. What can a school measure without pretending to deliver therapy?
A supervised mentoring program can measure appropriate process outcomes: voluntary enrollment, introductions, meaningful follow-up, requests for a new match, student satisfaction and whether participants can identify a trusted staff contact. Belonging or engagement questions may also be informative when planned and governed responsibly. 6 7
A dashboard count of messages is not a clinical improvement score. Even a student saying 'I feel more welcome' is descriptive feedback unless a valid design rules out alternative explanations and defines a precise outcome. Schools need approved privacy and evaluation practices.
Sources for this section
13. Why are before-and-after testimonials insufficient?
Imagine new ninth graders say they feel more comfortable in October than in August. This may reflect adjusting naturally, meeting classmates, joining activities, support from teachers or the mentoring program—or a combination. Without an appropriate comparison, a school cannot isolate how much change mentoring caused. 1 4 5
A transparent program report can say that participants reported greater comfort in a particular voluntary survey. It should not convert that into 'the program prevents depression' or imply all students experienced the same change.
14. What quality practices still make sense?
MENTOR's Fifth Edition emphasizes purpose, participant preparation, mentor training, inclusive recruitment, support for relationships, program leadership and evaluation. These research-informed standards are helpful for organizing a program even though they are not a clinical effectiveness trial of high-school peers. 7
Students should have choices, accessible ways to participate, a clear school adult contact and a respectful way to end or change a mentoring relationship. Program staff—not other students or an algorithm—remain accountable for student welfare and any safeguarding policy.
Sources for this section
15. Can peer-led programs have unwanted effects?
Yes, at least some studied groups reported unfavorable findings. The 2021 school peer-led review notes effects such as greater guilt or learning stress in certain studies. This is not a claim that most peer mentoring is harmful; it is a reason to track unintended burdens as well as hoped-for benefits. 1
Avoid assigning students constant emotional availability, forcing personal disclosure or making mentors feel responsible for resolving distress. Adult supervision and an appropriate handoff route help keep mentoring centered on ordinary peer connection.
16. What can Tagapag legitimately say about this research?
Tagapag is intended to help schools organize peer mentoring, student connections, goals and check-ins within a school-run program. The cited systematic reviews and CDC studies did not test Tagapag. Therefore they cannot establish Tagapag-specific effects on depression, anxiety or other clinical mental-health outcomes. 1 3 5
The defensible message is that schools value peer connection and can evaluate a supervised program's real experience. Research on supportive relationships provides context; clinical treatment and school safety decisions remain with appropriately qualified adults.
17. How should schools explain the evidence to families?
Say what the program actually offers: student introductions, approachable peer guidance, a staff coordinator and ways to report ordinary feedback. Distinguish intended benefits, student feedback and outcomes supported by controlled research. All three are useful but are not synonyms. 3 6 7
For personal mental-health concerns, schools should direct students and families to authorized professionals and established support paths. This research article is not medical guidance, a confidentiality policy or a replacement for school-specific safeguarding review.
18. Five checks for any headline saying 'mentoring works'
First identify who the mentors were (same-age students, older adults or trained health workers). Second identify the program design. Third identify what improved: friendship, connectedness, confidence, mental-health symptoms or something else. Fourth examine the comparison and follow-up. Fifth ask whether the claim names a specific product not studied. 1 2 3 8
For practical school implementation, read Peer Mentor Matching, School Connectedness Research, and Starting a School Peer Mentoring Program.
Sources for this section
- King & Fazel (2021), school-based peer-led mental-health systematic review
- Brinsley et al. (2025), peer-led adolescent health-behavior trial meta-analysis
- Raposa et al. (2019), meta-analysis of intergenerational youth mentoring
- Murphy et al. (2024), youth peer support within integrated mental-health care scoping review
Common questions, answered
- Can peer mentoring prevent teen depression?
There is not consistent direct evidence that typical high-school peer mentoring prevents diagnosed depression. Treat it as school connection support, not clinical care.
- What did the 2021 school peer-led review find?
It mapped 54 studies and synthesized 11 eligible effectiveness studies. Mental-health results for leaders and recipients were mixed and limited.
- What did the 2025 review find?
Seven controlled peer-led lifestyle studies showed no significant pooled effect for several mental-health/well-being outcomes; they were not ordinary mentor-matching trials.
- What is the difference between adult mentoring and peer mentoring?
Intergenerational mentoring has an older adult mentor. High-school peer mentoring pairs students, usually around school navigation, encouragement or belonging.
- Is feeling connected the same as improving mental health?
No. School connectedness is an experience associated with several outcomes, while clinical mental health is measured differently.
- What year was the CDC school connectedness statistic?
The 61.5% national connectedness figure comes from a 2021 survey published in 2023.
- Can message counts show a program works?
They show participation or activity, not clinical effectiveness or causal improvements in mental health.
- How can a school learn whether mentoring helps?
Plan a responsible evaluation with appropriate comparison, consent, valid outcomes, follow-up, privacy controls and professional input when needed.
- Do peer mentors replace counselors?
No. Student mentors provide ordinary connection and guidance; professional assessment, care and safety remain with qualified adults.
- Could mentoring cause additional stress?
Some peer-led studies reported unfavorable outcomes. Programs should check for burden, provide adult support and allow students to change or leave a match.
- Has Tagapag been studied in these trials?
No. The referenced peer-led and mentoring studies did not evaluate Tagapag or prove Tagapag-specific mental-health outcomes.
- What should schools promise to families?
Describe the actual school-led mentoring activities, safeguards and plans for evaluating participation—not unsupported treatment or clinical outcomes.
Sources & further reading
These references inform the discussion. Research on school connectedness or other mentoring programs is not evidence that Tagapag has produced the same outcomes.
- King & Fazel (2021), school-based peer-led mental-health systematic review
54 mapped studies; 11 effectiveness studies (2,239 participants), mixed results including some unfavorable outcomes
- Brinsley et al. (2025), peer-led adolescent health-behavior trial meta-analysis
Seven school trials (7,060 adolescents), mainly activity rather than mentoring; no significant pooled mental-health benefit
- Raposa et al. (2019), meta-analysis of intergenerational youth mentoring
70 adult-to-youth mentoring studies (25,286 participants), modest overall youth outcomes, NOT high-school peer-to-peer mentoring
- Wood & Mayo-Wilson (2012), school-based mentoring systematic review
Eight school mentoring studies (6,072 adolescents), very small and largely nonsignificant effects; not exclusively peer student programs
- CDC (2023), 2021 national YRBS school-connectedness analysis
2021 cross-sectional survey: 61.5% school connectedness; associations do not establish causation
- CDC (2024), Peer Connection and Support
General respectful peer interaction and group activities, not proof of clinical or platform outcomes
- MENTOR (2025), Elements of Effective Practice, Fifth Edition
Research-informed mentor-program standards, relationship monitoring and evaluation; not peer-program clinical efficacy trial
- Murphy et al. (2024), youth peer support within integrated mental-health care scoping review
15 studies in clinical youth services and some education; different peer-worker role and health-care setting
- CDC (2024), School Connectedness and Student Well-Being
Schoolwide belonging and relationship guidance; not causal proof that Tagapag works
Related reading
View all resourcesFor Schools
Peer Mentor Matching in High School: Preferences, Safety, and Fair Access
A school-led framework for respectful peer mentor matching that prioritizes student choice, accessible introductions and accountable adult support.
Research & Evidence
What Research Says About School Connectedness and Belonging
Belonging matters, but research findings need context. Here is what the CDC says—and what that does not prove about any one program.
For Schools
How to Start a Peer Mentoring Program at Your High School
From defining a purpose and recruiting mentors to preparing staff and keeping relationships active, here is a practical starting point.
