Geospatial modelling of environmental determinants of malaria in Zimbabwe: A case study from Mudzi district

Understanding the environmental factors influencing malaria transmission is vital to strengthening early warning systems and implementing proactive intervention strategies. While Zimbabwe continues to experience seasonal malaria outbreaks, localized environmental risk assessments remain underexplored, particularly in high-burden districts such as Mudzi. This study employed a geospatial and statistical modelling approach to assess the relationship between environmental variables and malaria incidence in Mudzi District from June 2023 to February 2025. Monthly malaria case data were obtained from all 31 public health facilities across Mudzi, covering the period June 2023 to February 2025. The monthly data was aggregated into custom-defined health access clusters using proximity-based network analysis. Environmental predictors rainfall, land surface temperature, vegetation (NDVI), proximity to water bodies, elevation, and distance to healthcare facilities were derived from MODIS, Sentinel-2, Climate Hazards Group InfraRed Precipitation with Stations (CHIRPS), and ancillary datasets. Negative Binomial Regression (NBR) models were applied to account for overdispersion in count data, incorporating time-lagged variables and interaction effects. Spatial autocorrelation was assessed using Global and Local Moran’s I. Temperature (p < 0.001), distance to water (p < 0.001), distance to health facilities (p < 0.001), and stagnant water body area (p < 0.001) were significant predictors of malaria incidence. Time-lagged temperature effects (Lag 1 and Lag 2) improved model fit and revealed delayed ecological impacts. Vegetation (NDVI) exhibited significance in interaction with elevation (p = 0.003). Local Moran’s I identified spatial outliers, notably low-case clusters adjacent to high-incidence zones, highlighting reporting inconsistencies or intervention effects. Environmental conditions associated with malaria risk in Mudzi District exhibit both spatial and temporal heterogeneity. Predictors such as temperature and water proximity demonstrate lagged and interactive effects on transmission dynamics. These findings support the development of localized, seasonally adaptive malaria surveillance strategies in Zimbabwe.