Assessing temporal causal effects of climate on dengue incidence in Peninsular Malaysia

Climatic factors such as temperature, humidity, and rainfall significantly influence dengue transmission. However, their lagged and immediate effects on dengue incidence in Peninsular Malaysia remain poorly understood. This study aimed to analyze these relationships using advanced time-series methods.

Monthly dengue cases (2013-2020) were obtained from the Ministry of Health Malaysia, while climate data were sourced from the Malaysia Meteorological Department. A Vector Autoregressive (VAR) model was used to assess lagged associations, while Granger causality tests explored temporal relationships. Instantaneous causality tests evaluated concurrent interactions.

The VAR model revealed that lagged dengue cases significantly influenced current cases ( ${\beta }_{t-2}=0.3853,P=0.0133:{\beta }_{t-5}=-0.3538,P=0.0120$ ) while temperature at lag 4 ( ${\beta }_{t-4}=-1643,P=0.0119$ ) had a delayed negative impact. Granger causality analysis indicated that temperature significantly predicted dengue incidence ( $F=1.7581,P=0.01342$ ), while humidity and rainfall showed no significant temporal effects. Instantaneous tests revealed strong immediate associations between temperature ( $X^2=26.828,P=<0.001$ ), humidity ( $X^2=30.099,P=<0.001$ ), and rainfall ( $X^2=24.77,P=<0.001$ ) with dengue incidence.

Temperature plays a critical role in both lagged and immediate dengue dynamics, while humidity and rainfall exhibit stronger immediate effects. These findings underscore the importance of integrating climate data, particularly, temperature into dengue early warning systems to enhance outbreak prediction accuracy.