Spatially varying correlation between environmental conditions and human leptospirosis in Sarawak, Malaysia

The spatial distribution of environmental conditions may influence the dynamics of vector-borne diseases like leptospirosis. This study aims to investigate the global and localised relationships between leptospirosis with selected environmental variables. The association between environmental variables and the spatial density of geocoded leptospirosis cases was determined using global Poisson regression (GPR) and geographically weighted Poisson regression (GWPR). A higher prevalence of leptospirosis was detected in areas with higher water vapour pressure (exp((a) over cap): 1.12; 95% CI: 1.02 - 1.25) and annual precipitation (exp((a) over cap): 1.15; 95% CI: 1.02 - 1.31), with lower precipitation in the driest month (exp((a) over cap): 0.85; 95% CI: 0.75 - 0.96) and the wettest quarter (exp((a) over cap): 0.88; 95% CI: 0.77 - 1.00). Water vapor pressure (WVP) varied the most in the hotspot regions with a standard deviation of 0.62 (LQ: 0.15; UQ; 0.99) while the least variation was observed in annual precipitation (ANNP) with a standard deviation of 0.14 (LQ: 0.11; UQ; 0.30). The reduction in AICc value from 519.73 to 443.49 indicates that the GWPR model is able to identify the spatially varying correlation between leptospirosis and selected environmental variables. The results of the localised relationships in this study could be used to formulate spatially targeted interventions. This would be particularly useful in localities with a strong environmental or socio-demographical determinants for the transmission of leptospirosis.