Mental health symptoms, perceived support-network effectiveness, and unmet need for care following an extreme climate event: A serial cross-sectional study from the CONECT-RS project

Epidemiological data on mental health collected while a climate disaster unfolds remain rare. Using a serial cross-sectional design with continuous online collection, we examined population-averaged differences in psychiatric symptoms, perceived support-network effectiveness, and psychological-care status across three periods after the 2024 floods in Southern Brazil: T1 (<1 month), T2 (1-3 months), and T3 (3-5 months). Generalized estimating equations estimated adjusted between-period differences in symptom scores and period-specific differences in their concurrent association with perceived support. Across 3838 response records from non-probability survey respondents, adjusted PHQ-9 and GAD-7 scores did not differ between T2 and T1 but were lower at T3 (B = -2.09, 95% CI -2.67 to -1.52; B = -2.37, 95% CI -2.88 to -1.86). PCL-5, assessed only at T2 and T3, was lower at T3 (B = -6.61, 95% CI -8.07 to -5.16). Reporting an effective support network was associated with lower concurrent symptom scores. This association differed across periods for depression (T2 × support B = -1.30, 95% CI -2.26 to -0.34) and anxiety (B = -1.24, 95% CI -2.10 to -0.37) but not for PTSD symptoms. Odds of reporting effective support were lower after T1. Among respondents who perceived a need for care, lower income, older age, and not reporting an effective support network were associated with higher odds of unmet need. These findings highlight socioeconomic inequalities in symptom burden and unmet need after a climate disaster and support evaluating community-based support and access strategies.