Modeling Nonlinear and Delayed Climate Effects on Malaria Incidence Using Distributed Lag Non-Linear Models: Evidence from Kano State, Nigeria

Malaria remains a major public health challenge in northern Nigeria, with transmission dynamics strongly influenced by climatic variability. This study investigated the nonlinear and lagged effects of key meteorological variables on malaria incidence in Kano State using a Distributed Lag Non-Linear Model (DLNM). Monthly malaria incidence data obtained from the District Health Information System and corresponding meteorological variables (rainfall, relative humidity, and temperature) from the Nigerian Meteorological Agency were analyzed over the period 2015–2025. Descriptive and time-series analyses revealed pronounced seasonal patterns, with malaria incidence peaking during and shortly after the rainy season. Spearman correlation and cross-correlation analyses indicated that rainfall and relative humidity were positively associated with malaria incidence, exhibiting significant lag effects of approximately 1–3 months, while temperature showed weaker and more complex relationships. Model comparison demonstrated that the DLNM outperformed both Poisson and Negative Binomial models, achieving the lowest AIC (2,589.80) and BIC (2,699.95), indicating superior model fit. Exposure–lag–response analysis revealed that moderate rainfall and specific humidity ranges significantly increased malaria risk after 2–4 months, whereas extreme conditions reduced transmission. Temperature effects were nonlinear, with lower temperatures associated with higher malaria risk at longer lag periods. The model further demonstrated strong predictive performance, accurately capturing seasonal transmission patterns with a Mean Absolute Percentage Error (MAPE) of 5.85%. These findings highlight the importance of incorporating nonlinear and delayed climatic effects in malaria modelling and underscore the potential of DLNM-based approaches for climate-informed early warning systems and targeted public health interventions in malaria-endemic regions.