Did COVID-19 change the trajectory of children's internalising symptoms? A meta-analysis of longitudinal studies and moderating factors

The COVID-19 pandemic disrupted children's and adolescents' daily lives worldwide, raising concerns about changes in internalising symptoms. Prior reviews have documented elevated depression and anxiety symptoms during the pandemic, but cross-sectional designs have limited conclusions about within-person change and its moderators. To address this gap, we conducted a meta-analysis of longitudinal studies assessing internalising symptoms before and during the COVID-19 pandemic. Eighty peer-reviewed studies (187 effect sizes) contributed data from youth aged 5–17 years across 20 countries. Standardized mean change (SMC) effect sizes were synthesised using random-effects models with robust variance estimation. Guided by Bronfenbrenner's ecological systems theory, we examined moderators spanning individual, family/school, contextual, and temporal domains. Overall, internalising symptoms increased modestly from pre- to peri-pandemic assessments (SMC = 0.32, 95% CI [0.11, 0.53]; multilevel SMC = 0.34, 95% CI [0.15, 0.53]). Effects varied widely across studies (I 2 = 97.9%), indicating highly heterogeneous symptom trajectories, and suggesting that pooled estimates should be interpreted cautiously. Larger increases were observed in studies including parent-report measures and in contexts of full school closures and highly stringent government responses. Age, gender/sex, region, COVID prevalence, and assessment timing were not significant moderators. Although there were larger increases for composite scores across broad internalising and depressive symptoms than for anxiety or stress, there were no statistically robust between-domain differences. Together, these results indicate that internalising symptoms increased modestly during the COVID-19 pandemic, but that youth experiences were highly context-dependent, underscoring the importance of considering ecological conditions—such as school closures and policy responses—when evaluating pandemic-related mental health change.