Abstract Urban environmental and climatic conditions are closely linked to pediatric respiratory health, yet their associations with specific infectious diseases such as mycoplasma pneumoniae pneumonia (MPP) remain poorly understood. To address this gap, this study investigated the spatiotemporal dynamics of pediatric MPP across 46 urban subdistricts in Fuzhou, China. Using global and local spatial autocorrelation, space–time scan statistics, and interpretable machine learning models, we assessed both incidence patterns and progression to severe disease, as well as their built-environment correlates. Key findings include: (1) A pronounced seasonality of MPP is observed, with autumn–winter peaks and summer troughs. High-risk clusters are concentrated in central urban areas, whereas peripheral regions exhibit weaker clustering. These patterns are consistent with variations in climate, population density, transport networks, greenspace, and school-term cycles. (2) The autumn–winter period represents the critical risk window, encompassing a large share of cases and repeatedly centering on repeatedly high-incidence subdistricts. Demographically, boys and children aged 5–9 years face the highest risk, whereas the 10–14 age group—particularly girls—shows the lowest incidence. (3) Machine learning identified age, near-road exposure, and selected land-use/amenity intensities as core predictors, especially for severe MPP cases. Age exhibits a non-monotonic incidence profile, while distance from primary roads follows a steep inverse gradient, with elevated risk confined to a 100–150 m roadside buffer zone. School density shows a monotonic positive association, whereas greenspace is a broadly negative association. These findings may help inform seasonally tailored, microscale interventions, strengthened school health monitoring, and targeted urban design in high-prevalence corridors.
Dong et al. (Sat,) studied this question.