Impact of Meteorological Factors on Elderly Influenza in Shenzhen and Beijing From 2017 to 2024: A Time Series Study

Abstract

China's heterogeneous climatic conditions and geographic complexity contribute to distinct influenza patterns between northern and southern regions, complicating incidence prediction. Utilizing real-world data, this study aimed to characterize the epidemiology of influenza among elderly populations in Shenzhen and Beijing (representative southern and northern cities) and evaluate the impact of meteorological factors on influenza risk. Laboratory-confirmed influenza cases in individuals aged ≥ 60 years were obtained from China's National Notifiable Infectious Disease Surveillance System (January 2017–April 2024). Spearman correlation analysis identified meteorological factors associated with weekly case counts. Significant factors were incorporated into a distributed lag non-linear model (DLNM) to quantify city-specific, non-linear lagged effects of meteorological variables on influenza risk across pre-pandemic (2017–2019) and post-pandemic (2023–2024) periods. From 2017 to 2024, a total of 31,126 cases of influenza in the elderly were reported in Shenzhen and 22,907 cases were reported in three districts of Beijing. DLNM model fitting showed that each meteorological factor had a nonlinear relationship with influenza incidence. Low temperatures in Shenzhen post-pandemic (lag second weeks: RR = 2.54, 2.20, 2.94), and in Beijing pre-pandemic (lag second weeks: RR = 9.3, 5.23,16.55), post-pandemic (lag third weeks: RR = 9.57, 7.08,12.93) promote the onset of influenza. In Beijing, pre-pandemic high precipitation had an RR of 33.38 (1.48, 7.68) at lag third weeks, which decreased to < 1 post-pandemic. For high precipitation, Shenzhen's pre-pandemic RR at lag third weeks was 9.84 (6.30, 15.37), declining to 5.74 (4.28, 7.70) post-pandemic. Post-pandemic, low atmospheric pressure in Shenzhen (lag third weeks: RR = 1.54, 1.22, 1.94) and high pressure in Beijing (lag third weeks: RR = 5.20, 3.88, 6.96) were associated with elevated influenza risk. The exposure curve for wind speed has a “valley” shape, the RR decreases and then increases as the wind speed changes from low to high, and the risk of influenza is higher for either too high or too low wind speeds. The impact of meteorological factors on influenza among the elderly demonstrates significant spatiotemporal heterogeneity. In both Shenzhen and Beijing, low temperatures and extreme wind speeds (high and low) promote influenza incidence. High precipitation consistently enhances influenza risk in Shenzhen, while in Beijing, its effect shifted from promoting to protective. Post-pandemic, low pressure in Shenzhen and high pressure in Beijing were associated with increased influenza. These findings underscore the importance of integrating regional meteorological and socio-environmental factors into influenza prevention, control, and prediction strategies to more accurately assess epidemic trends and risks.