Delayed and nonlinear associations and predictive causal signals between climate factors, air pollutants, and influenza epidemics in 48 countries

Influenza remains a recurring threat to global health with pandemic potential, yet the predictive causal associations of climate factors and air pollutants on its transmission remains not fully quantified. Thus, this study aimed to quantify the nonlinear and delayed effects of these factors, as well as their predictive causal signals, on global influenza epidemics. This study is a longitudinal time-series analysis conducted across 48 countries from 2022 to 2024. Weekly influenza positivity rates were derived from the FluNet database and integrated with climate variables and air pollutants. Spatiotemporal Bayesian models were constructed separately for the Northern and Southern Hemispheres to quantify the nonlinear and lagged effects of climate variables and air pollutants while accounting for socioeconomic conditions. Predictive causal relationships were further assessed using convergent cross mapping to identify potential environmental drivers of influenza transmission. In Northern Hemisphere countries, the highest odds ratio (OR) for influenza was 1.74 (95% CI: 1.34–2.26) at an SO 2 concentration of 3100 μg/m 3 with a 1.6-week lag, and 2.36 (95% CI: 1.53–3.64) at a temperature of –27 °C with a 2.2-week lag. In Southern Hemisphere countries, the highest OR was 1.72 (95% CI: 0.65–4.58) at an OC concentration of 5800 μg/m 3 with a 3.2-week lag, and 1.31 (95% CI: 1.00–1.71), corresponding to a temperature of 14 °C with a lag of 2 weeks. CCM analysis identified PM 2.5 in dust and wind speed as the dominant drivers among air pollutants and climate factors, respectively. PM 2.5 in dust influenced influenza in 19 countries, while wind speed did so in 15 countries across the study area. Climate variables and air pollutants exhibited nonlinear and lagged effects on influenza transmission, with key air pollutants varying across hemispheres. Predictive causal signals of selected climate variables were further modulated by specific socioeconomic conditions. These findings highlight the importance of integrating environmental and socioeconomic conditions, providing practical guidance for implementing targeted influenza control strategies.