Temperature-Based Predictions for West Nile Virus Outbreaks in Endemic Regions of Continental Croatia

Weather conditions, especially temperature, rainfall, and humidity, affect the transmission and spread of West Nile virus (WNV). This study investigated the effects of weather patterns on WNV activity in Croatia using climatological, equine seroprevalence, and human case data collected in two endemic continental Croatian regions with high WNV activity from 2015 to 2024. Overall equine WNV IgG prevalence was significantly higher in East Croatia (30.34%) than in Central Croatia (10.90%) and increased over time in both regions with a similar temporal pattern, indicating a shared upward trend in viral circulation. Higher equine seroprevalence was observed in areas with confirmed recent equine infections (IgM positive) within the same transmission season. In addition, human cases and recent equine infections were significantly associated with higher equine seroprevalence in the following season, with increases of 4% and 7%, respectively. In contrast to precipitation and humidity, temperature was significantly associated with human WNV cases, whereas no comparable effect was found in horses. Temperatures in February, April, May, and October emerged as key predictors, and the model including mean April and May temperatures showed the best predictive performance for human WNV cases, further supported by analysis of the epidemic year 2018. However, when 2018 was excluded, the effects of temperature remained significant only for May and July, with increased May temperatures emerging as the most important predictor of WNV activity in the upcoming transmission season.