Causal Structural Equation Modeling of Climatic and Environmental Factors Influencing Malaria Incidence in Southeastern Senegal

Abstract

Introduction: Malaria remains a major public health problem in tropical regions due to complex interactions between climatic, environmental, and epidemiological processes. Objective: To improve the quantitative understanding of these mechanisms, this study introduces a causal analysis framework based on structural equation modeling (SEM) to elucidate the direct and indirect pathways through which climatic and environmental variables influence malaria incidence. Methods: We analyzed weekly malaria case data collected in four districts of southeastern Senegal (Kédougou, Salémata, Saraya, and Dianké Makha) over the period 2018–2024 using structural equation modeling and climates factors. Climate predictors included temperature, precipitation, relative humidity, wind speed, atmospheric pressure, cloud cover, and lunar phase, obtained from Visual Crossing, complemented by vegetation indices derived from satellite imagery. Results: The results revealed significant spatial heterogeneity: rainfall and cloud cover had the strongest total effects on malaria incidence in three districts (standardized total effects ranging from 0.51 to 0.66), while high wind speed emerged as the primary determining factor in Kédougou, with a negative association (−0.51). Wind-related factors showed protective effects against malaria transmission, with notable spatial complexity, and atmospheric pressure exhibited consistent indirect effects (0.18 to 0.42), mediated by other climatic predictors, but without a significant direct association. Conclusion: These results underscore the importance of considering indirect causal pathways in environmental epidemiology and highlight district-specific climatic profiles that can inform localized malaria control strategies. The SEM framework presented here offers a rigorous statistical approach to integrating multi-source climatic and environmental data to advance causal inference in vector-borne disease research, ultimately supporting early warning systems and targeted vector control interventions at the district level.