Assessing Temporal Trends and Ecological Associations Between PM2.5 Exposure, Urbanization, and Under-Five Mortality in Somalia, 1990–2023: An Ecological Time-Series Analysis

Abstract Background Ambient fine particulate matter (PM2.5) is a significant environmental health concern, and rapid urbanization may have an impact on child survival by affecting health-care access, sanitation, housing, poverty, inequality, and environmental exposure. Somalia has a unique situation in which these processes coexist with recurring drought, displacement, violence, and humanitarian disasters. Objective To study temporal trends in PM2.5 exposure, urbanization, and under-five mortality in Somalia from 1990 to 2023, as well as to investigate ecological relationships between yearly variations in PM2.5 exposure, urbanization, and under-five mortality. Methods This ecological national time-series analysis included 34 yearly observations. Kendall rank tests, Theil-Sen slopes, and linear trend models with heteroskedasticity and autocorrelation (HAC; Newey-West) were used to examine temporal trends. Raw Pearson and Spearman correlations were seen as exploratory. HAC standard errors were used in the main association model to regress first differences in under-five mortality on first differences in PM2.5 and urbanization. Sensitivity and supporting analyzes included robust Huber regression and influence, as well as residual, autocorrelation, and stationarity tests. Results PM2.5 exposure increased over time (Kendall τ = 0.387, p = 0.001; Theil-Sen slope = 0.090 µg/m³/year, 95% CI 0.042–0.143), as did urbanization (τ = 1.000, p < 0.001; slope = 0.661 percentage points/year, 95% CI 0.555-0.800), while under-five mortality declined (τ=−0.674, p < 0.001; slope = − 2.30 deaths per 1,000 live births/year, 95% CI − 3.33 to − 1.30). Annual variations in PM2.5 (β = 1.28, 95% CI − 8.81 to 11.36, p = 0.804) and urbanization (β=−2.85, 95% CI − 25.29 to 19.58, p = 0.803) did not substantially impact under-five mortality. Robust sensitivity analysis yielded the same general conclusion. Conclusion Long-term increases in PM2.5 exposure and urbanization were associated with lower under-five mortality in Somalia, but year-to-year first-difference analyzes did not replicate the basic ecological trends. These data suggest a careful interpretation of temporal ecological connections but do not show causation.