Background and objectives: The number of pediatric patients diagnosed with influenza types A and B is increasing annually, especially in temperate regions such as Shanghai (China). The onset of pandemic influenza viruses might be attributed to various ambient meteorological factors including temperature, relative humidity (Rh), and PM 1 concentrations, etc. The study aims to explore the correlation between the seasonality of pandemic influenza and these factors. Methods: We recruited pediatric patients aged from 0 to 18 years who were diagnosed with influenza A or B from July 1st, 2017 to June 30th, 2019 in Shanghai Children's Medical Centre (SCMC). Ambient meteorological data were collected from the Shanghai Meteorological Service (SMS) over the same period. The correlation of influenza outbreak and meteorological factors were analyzed through preliminary Pearson's r correlation test and subsequent time-series Poisson regression analysis using the distributed lag non-linear model (DLNM). Results: Pearson's r test showed a statistically significant correlation between the weekly number of influenza A outpatients and ambient meteorological factors including weekly mean, maximum, minimum temperature and barometric pressure (P < 0.001), and PM 1 (P < 0.01). While the weekly number of influenza B outpatients was statistically significantly correlated with weekly mean, maximum and minimum temperature (P < 0.001), barometric pressure and PM 1 (P < 0.01), and minimum Rh (P < 0.05). Mean temperature and PM 1 were demonstrated to be the statistically significant variables in the DLNM with influenza A and B outpatients through time-series Poisson regression analysis. A U-shaped curve relationship was noted between the mean temperature and influenza A cases (below 15°C and above 20°C), and the risks increased for influenza B with mean temperature below 10°C. PM 1 posed a risk after a concentration of 23 ppm for both influenza A and B. High PM 1 , low and the high temperature had significant effects upon the number of influenza A cases, whereas low temperature and high PM 1 had significant effects upon the number of influenza B cases.

No abstract