Meteorological predictors of dengue hemorrhagic fever incidence in Cilacap Regency, Indonesia: A time series analysis
Khaidar Ali, Serius Miliyani Dwi Putri, Damairia Hayu Parmasari, Yemima Shamura Vividia, I Made Dwi Mertha Adnyana, Dwi Sarwani Sri RezekiObjective:
To analyze the nonlinear relationship between meteorological predictors and dengue hemorrhagic fever, and to quantify the meteorologically attributable fraction of dengue hemorrhagic fever (DHF) during 2021-2025 in Cilacap Regency.
Methods:
This ecological time series design used sequential three-stage analytical tools, including a cross-correlation function (CCF) analysis applied to identify lag structures, Granger causality within a vector autoregression (VAR) framework used to confirm directional temporal precedence, and a quasi-Poisson Distributed Lag Nonlinear Model (DLNM) with natural cubic spline crossbases over a 28-day lag window used to estimate the exposure-response relationship. Autoregressive terms alongside long-term trends and seasonal splines were incorporated to account for autocorrelation and potential confounding.
Results:
Granger causality testing confirmed unidirectional of DHF relationships across all four predictors, including mean temperature, rainfall, relative humidity, and wind speed (all
Conclusions:
Rainfall and temperature were the meteorological predictors most strongly associated with DHF incidence in Cilacap Regency. This finding provides evidence for actionable vector control program and early warning system of DHF to consider the utilization of meteorological predictor to prevent DHF outbreak in Cilacap.