Climatic heterogeneity in meteorological associations with childhood acute respiratory symptoms across four South Asian countries.

Background: Childhood acute respiratory symptoms (ARSs) remain an important public health concern in South Asia, but meteorological associations may not be transferable across climatically diverse settings. Studies that rely on pooled regional estimates or single exposure metrics may therefore obscure meaningful heterogeneity. We examined whether associations of childhood ARSs with temperature and vapour pressure deficit (VPD) differed across macroclimatic regimes in four South Asian countries. Methods: We analysed Demographic and Health Survey data from India, Bangladesh, Nepal, and Timor-Leste, initially comprising 244,437 children aged <5 years aggregated into 33,407 survey cluster-month units. Childhood ARS was defined using a harmonised symptom-based measure based on caregiver-reported fast breathing and chest-related breathing difficulty during the preceding two weeks. Monthly temperature and VPD were assigned from ERA5 reanalysis data. We estimated the associations within country-macroclimate strata using mixed-effects binomial logistic regression models adjusted for demographic, socioeconomic, and household environmental covariates. Results: Meteorological associations with childhood ARS varied substantially across climatic regimes. Temperature-related associations were heterogeneous in both direction and magnitude across strata. By contrast, the clearest and most robust inverse association was observed for VPD in India's arid macroclimate (odds ratio (OR) = 0.56; 95% confidence interval (CI) = 0.44, 0.71). In humid tropical settings, associations for both temperature and VPD were largely null. In India's arid macroclimate, comparing the highest vs lowest VPD tertile corresponded to a risk difference of -1.28 percentage points (95% CI = -2.14, -0.71), equivalent to approximately 13 fewer ARS cases per 1000 children surveyed. Conclusions: Meteorological associations with childhood ARS were strongly context-dependent and should not be assumed to be uniform across climatic regimes. In this multicountry analysis, VPD showed the clearest association in India's arid macroclimate, whereas temperature-related associations were more heterogeneous across settings. These findings highlight limitations of pooled regional interpretations and support climate-stratified approaches to meteorological risk assessment in child health surveillance. Keywords: childhood acute respiratory symptoms; South Asia; climatic heterogeneity; vapour pressure deficit; temperature; child health surveillance.