Peer-led family sessions cut stress in refugee adults

Refugees carry among the highest burdens of mental health problems yet frequently underuse the services available to them, and interventions designed for people already diagnosed do little to stop symptoms developing in the first place. That gap has driven interest in prevention programmes that can be delivered by trained members of the community rather than clinicians, in people's own homes and languages. Problem Management Plus for Immigrants is an adaptation of the World Health Organization's Problem Management Plus, bundling psychoeducation, problem-solving, behavioural activation, social support and networking, and mind-body exercises. The question here was whether such a programme improves mental, social and emotional well-being in a resettled community.

This pilot randomised controlled trial enrolled Bhutanese adults aged 18 or older resettled in Massachusetts who scored 14 or below on the Patient Health Questionnaire-9, deliberately targeting people below the threshold for more severe depression. Families were randomised to the intervention (58 families) or to a talk programme with a community support services pamphlet (58 families), giving 232 participants from 116 families in total. Trained community interventionists delivered five sessions in family settings. Primary outcomes were perceived stress on the Cohen Perceived Stress Scale-10 and anxiety and depression on the Hopkins Symptoms Checklist-25, measured at baseline, six weeks and three months. Secondary outcomes included hair cortisol, coping, coping self-efficacy, social support and network, family conflict resolution and family satisfaction, analysed with linear mixed-effects models adjusted for baseline scores, age, duration of residence, marital status and chronic disease history.

All 232 recruited participants were retained for the whole project, an unusually complete follow-up. The intervention group showed significantly greater decreases than controls in stress, anxiety and depressive symptom scores at both the six-week and three-month assessments. Secondary outcomes moved in the same direction with large effect sizes: coping, family conflict resolution, self-efficacy, family satisfaction and social networking were all significantly higher in the intervention group at both time points (Cohen's d greater than 0.8, p less than .01). Notably, hair cortisol concentrations did not differ significantly between groups at baseline or three months, so the biological stress marker did not track the self-reported improvements.

The trial supports the wider case that task-shifted, peer-delivered psychological interventions can work outside specialist clinics, and extends it to a prevention framing in people who are not severely symptomatic. The divergence between large self-reported gains and an unchanged hair cortisol signal is the interesting tension: it may reflect the limits of cortisol as an outcome measure, the short follow-up, or expectancy effects in an unblinded psychosocial trial. The authors are clear this is a pilot and that a large-scale randomised trial across diverse refugee groups is needed before the model is scaled nationally. Durability beyond three months is untested.

This is a reason to think that structured, community-delivered support built around problem-solving, activation and social connection can shift how stressed, anxious and low people feel, even without professional therapists. It also illustrates a recurring pattern in stress research: questionnaires and biology do not always agree, and an unchanged cortisol measure alongside improved mood is a genuine finding rather than a failure to report. For anyone reading it as a general lesson, the ingredients — practical problem-solving, doing more of what matters, and strengthening social networks — are the same components that appear across mainstream psychological approaches. The caveats about pilot size, one community and short follow-up all apply.