Introduction Leptospirosis has emerged as an important health problem in developing countries due to the growth of slum settlements worldwide, where poor sanitation favours rat-borne transmission. Large urban epidemics occur during seasonal periods of heavy rainfall. However, a detailed analysis has not been performed to determine how rainfall, as well as other climatic factors, specifically influences the risk of leptospirosis in these endemic settings.

Methods We analysed data from 2083 leptospirosis cases which were identified during active population-based surveillance performed in the city of Salvador, Brazil between 1996 and 2010. Information on daily rainfall, humidity and temperature were obtained for the same period. A generalised additive model was fitted, using a negative binomial distribution for weekly aggregated data (729 weeks). We incorporated a non-parametric term to estimate the time trend and a sin-cosine term to control for seasonal confounding.

Results Rainfall and humidity were positively associated with the number of cases two weeks later, linearly and without a threshold. Temperature protected, even though the range between maximum and minimum temperatures is small: from 22 to 33°C. A decreasing trend was highly significant, possibly due to intervention of sewerage and garbage collection systems.

Conclusion Leptospirosis is expected to become an increasingly important slum health problem as predicted global climate change and growth of the world's slum population evolves, and models adequate to estimate the impact of both environment and climate variables on incidence of all environmental related diseases should be incorporated in the epidemiologists toolbox.