Teachers as first-line mental health and psychosocial support providers in war-affected Ukrainian communities: changes after an MHPSS training programme across communities with different levels of security threat

Abstract Background. Teachers in war-affected Ukraine carry a dual burden of personal exposure and first-line psychosocial support for pupils. Evidence on what community-based training changes in them, and under which security conditions, is scarce. Methods. Single-group evaluation (first two of three waves) of a 30-hour in-service MHPSS training programme in six communities of three oblasts spanning a security threat index from 2.00 to 4.59 on a 0–5 scale. Three hundred teachers (50 per community; 86.7% women) were assessed before training (June 2026) and after it (November 2026) on subjective well-being, positive mental health, psychological safety of the school collective, stress and recovery (with a cumulative exposure checklist and a social desirability scale) and mental health literacy. Change was estimated at the individual and the community level with new exposure as a covariate; subgroup contrasts were tested against regression to the mean, and sensitivity analyses and reliable change indices were reported. Results. Retention was complete. Ten of 21 outcomes changed robustly; two further changes did not survive community-level and sensitivity checks. Mental health literacy rose most (Δ = 4.67, d_z = 0.73, d_av = 0.56; high level 58% → 79%), followed by physical well-being (d_z = 0.54), stress reactions (−0.46) and recovery actions (0.46); positive mental health and total well-being changed little, and psychological safety, a pre-specified negative control, not at all. Between waves 23% of teachers (46% in the high-threat cities) reported new stressful events; each was associated with 0.44 additional points of stress reactions and statistically accounted for the smaller change in those cities. Every anticipated subgroup advantage matched regression to the mean and was not significant in the hypothesised direction after baseline adjustment. Social desirability rose from 15% to 21%; principal effects were robust to its control. Reliable individual change was shown by 2–10% of teachers. Internal consistency fell at the second wave, to zero or below for three short subscales, without evidence of careless responding. Conclusions. Change after the programme followed a content-specific profile consistent with the controlled literature, and the smaller change in stress in front-line communities was associated with continuing exposure. Without a comparison group these changes cannot be attributed to the programme, and none of the anticipated subgroup advantages was distinguishable from regression to the mean. Evaluations under ongoing war should measure exposure at every wave, test differentiated claims against regression to the mean and report reliable change. Keywords. mental health and psychosocial support; teachers; armed conflict; Ukraine; mental health literacy; resilience; subjective well-being; psychological safety.