Interacting climatic drivers of dengue incidence: a GARMA-based time series analysis in a Brazilian health region.

Dengue is a major public health challenge in Brazil, with significant temporal variability and seasonal outbreaks. In this paper, we investigate the dynamics of the disease in a Brazilian health region, using a generalised autoregressive moving average (GARMA) model, based on the negative binomial distribution, to monthly case counts from 2014 to 2025. Climate covariates such as precipitation, relative humidity, atmospheric pressure, temperature, and wind speed, were considered to determine both individual and joint effects on dengue transmission. The results indicate a nonlinear temporal trend combined with a strong seasonal pattern. A negative and statistically significant interaction between precipitation and humidity was identified, indicating that, while higher values of these variables alone are associated with increased dengue incidence, their combined occurrence at high levels in the preceding month is associated with a relative reduction in case numbers when compared with what would be expected if the effects of both variables acted independently.