Mental Health

Mechanisms of impact of mental health peer support in high-, middle- and low-income settings: mediation analysis of the UPSIDES randomised controlled trial.

TL;DR

Peer support affects recovery-related outcomes primarily through intermediate mechanisms of enhanced hope and social inclusion, rather than through direct effects on personal recovery or health and social functioning.

Key Findings

There were no significant direct effects of peer support on personal recovery or health and social functioning.

  • Cross-lagged panel modelling was used to explore longitudinal associations and mediating pathways across four time points: baseline, 4 months, 8 months (end of intervention), and 12-month follow-up.
  • Despite structured peer support delivered over 6–8 months, no significant direct effects emerged on either personal recovery or clinician-rated health and social functioning.
  • Strong autoregressive effects were observed across all variables, indicating high temporal stability of outcomes over time.

Peer support had a significant indirect effect on personal recovery mediated through social inclusion.

  • The indirect effect via social inclusion was β = 0.114 (95% CI [0.049, 0.194], P < 0.05).
  • This was the strongest mediation pathway identified in the study.
  • The sample comprised 565 adults with severe mental health conditions across six sites in Germany, Uganda, Tanzania, India, and Israel.

Peer support had a significant indirect effect on personal recovery mediated through hope.

  • The indirect effect via hope was β = 0.037 (95% CI [0.001, 0.086], P < 0.05).
  • Hope was identified as a secondary but statistically significant mediating pathway between peer support and personal recovery.
  • Standardised, self-report measures of hope were administered at each time point.

Peer support had significant indirect effects on health and social functioning through both social inclusion and hope.

  • The indirect effect via social inclusion on health and social functioning was β = -0.035 (95% CI [-0.064, -0.013], P < 0.05); the negative sign reflects the scoring direction of the clinician-rated functioning measure.
  • The indirect effect via hope on health and social functioning was β = -0.026 (95% CI [-0.052, -0.006], P < 0.05).
  • Health and social functioning was assessed using a clinician-rated instrument, distinguishing it from the self-report measures used for other outcomes.

Empowerment was not identified as a significant mediator between peer support and either personal recovery or health and social functioning.

  • Despite empowerment being theoretically linked to peer support interventions, it did not emerge as a significant mediating pathway in the cross-lagged panel models.
  • Social inclusion and hope were the two variables that showed significant indirect effects, while empowerment did not.
  • Empowerment was measured using standardised self-report instruments at all four time points alongside social inclusion, hope, and personal recovery.

The UPSIDES trial was a multicentre randomised controlled trial conducted across high-, middle-, and low-income country settings.

  • The trial included six sites across five countries: Germany, Uganda, Tanzania, India, and Israel.
  • A total of 565 adults with severe mental health conditions participated.
  • Participants in the intervention group received structured peer support from trained peer workers over a 6- to 8-month period.
  • The study spanned high-, middle-, and low-income settings, making it one of the few multinational peer support trials of its kind.

Individuals with lower baseline levels of hope and social inclusion may particularly benefit from peer support interventions.

  • This finding emerged from interpretation of the mediation results, suggesting that those with the most room for improvement in hope and social inclusion stand to gain the most.
  • The authors highlight this as a practical implication for targeting and designing peer support services.
  • This conclusion supports prioritising social and psychological domains when implementing peer support in mental health services.

What This Means

This research suggests that mental health peer support programs — where people with lived experience of mental health conditions support others with similar experiences — do not directly improve recovery or day-to-day functioning on their own. Instead, the study found that peer support works through two intermediate steps: it first increases a person's sense of hope and their feeling of being socially included in their community, and these improvements in turn lead to better personal recovery and functioning. This finding came from a large international study involving 565 people with severe mental health conditions across Germany, Uganda, Tanzania, India, and Israel, making it one of the most globally diverse studies of peer support to date. The study used a statistical technique called cross-lagged panel modelling to track how these different factors — peer support, hope, social inclusion, empowerment, personal recovery, and functioning — influenced each other over time across four measurement points spanning about a year. Notably, a third potential mechanism, empowerment, did not show the same mediating effect as hope and social inclusion did. The findings were consistent across both self-reported recovery outcomes and clinician-rated functioning measures. This research suggests that peer support programs may be most effective when they are deliberately designed to foster hope and social connection, rather than focusing solely on symptom management or clinical outcomes. It also suggests that people who start with low levels of hope or social inclusion may have the most to gain from peer support. These insights could help mental health services and policymakers design more targeted and effective peer support interventions across diverse global settings.

Have a question about this study?

Citation

Hiltensperger R, Charles A, Goldfarb Y, Haun M, Kalha J, Mahlke C, et al.. (2026). Mechanisms of impact of mental health peer support in high-, middle- and low-income settings: mediation analysis of the UPSIDES randomised controlled trial.. Epidemiology and psychiatric sciences. https://doi.org/10.1017/S2045796026100900