Modeling the effect of climatic factors on malaria incidences using an auto-regressive generalized linear model with lagged covariates

Abstract

Malaria remains a significant public health challenge globally, with its transmission intricately associated with climatic factors. Understanding the time lag effect of these variables on malaria incidence is crucial for developing effective control strategies. The research used monthly malaria incidence data obtained from the Uganda Ministry of Health, along with climate data from the Uganda National Meteorological Authority to investigate the impact of climatic factors, specifically rainfall, temperature, humidity and wind, on malaria incidence in Arua District, using monthly data from 2020 to 2024. To capture both immediate and delayed effects, the study employed an autoregressive generalized linear model (AR-GLM) with lagged covariates. The results indicate that increased rainfall and humidity with specific delays are positively associated with the incidence of malaria, while changes in temperature show a complex and unstable relationship. In addition, the AR-GLM model, which incorporates these time-dependent effects, outperforms a standard GLM, as evidenced by lower Akaike Information Criterion (AIC) and Mean Absolute Percentage Error (MAPE) values. Residual analysis confirms that AR-GLM adequately captures the temporal dependencies in the malaria data, showing no significant autocorrelation in residuals. The findings contribute to the development of effective malaria control and prevention strategies tailored to the specific temporal dynamics of the climate-malaria interactions in Arua district, ultimately aiming to improve public health outcomes.