Changes in Depression Symptoms Following Conflict Truce: A Mixed-Effects Analysis of Displacement and Time

Background: Armed conflict and displacement have profound effects on children's mental health. While empirical evidence highlights the adverse impacts of war, fewer studies explore potential improvements in well-being during post-conflict periods such as truces or ceasefires. This study examines whether children’s depressive symptoms declined following a truce and how these changes varied based on displacement history.

Methods: We used data from a longitudinal evaluation of Semillas de Apego, a psychosocial program serving families affected by armed conflict in Colombia. This study employed a quasi-experimental design, using the 2019 national truce as a natural intervention. The analytic sample included 1,376 children aged one and five assessed at three timepoints between June 2018 and March 2019: Baseline (BL, June 2018), Follow-up 1 (F1, ~3 months later), and Follow-up 2 (F2, ~9 months later, post-truce). Thus, only F2 captured post-truce depression score outcome. Depression scores (standardized z-scores) were measured and transformed using a caregiver-reported adapted from the Child Trauma Symptom Checklist for Young Children. To assess within-subject changes over time, three paired t-tests were conducted: BL vs. F1, F1 vs. F2, and BL vs. F2. Next, we estimated a linear mixed-effects regression model for repeated measures and between-individual variance. A random intercept by participant ID was included to account for individual variability.

Results: The Mean of depression scores across waves were, BL (M = 0.037), F1 (M = -0.001), and F2 (M = -0.065). The comparison of depression scores across time points showed the following results: BL vs. F1 (t = 1.35, p = 0.18), F1 vs. F2 (t = 1.44, p = 0.15), and BL vs. F2 (t = 3.42, p < 0.001). The linear mixed-effects model revealed no significant difference in depression at F1 compared to baseline (β = -0.037, p = 0.184). However, a significant reduction in depression was observed at F2 compared to baseline (β = -0.122, p = 0.002). The frequency of forced displacement was positively associated with depression scores (β = 0.116, p < 0.001). The significant interaction between time (at post-truce period) and displacement was observed, showing a significant post-truce reduction in depression (β = -0.139, p = 0.010). Model comparison results justify that the use of mixed effect model given its overall model fit of mixed effect model is significantly better than a standard linear regression (χ² (1) = 498.77, p < 0.001).

Conclusions and Implications: Findings suggest that ceasefire periods may offer critical opportunities for psychological recovery, particularly among moderately displaced children. However, children with more severe displacement histories may require ongoing mental health support. These findings reveal the importance of disaggregating conflict-affected populations when evaluating post-conflict recovery interventions.

Keywords: Child mental health, Post-conflict recovery, Displacement, Depression symptoms, Ceasefire impact.