[Correlation between low air temperature and influenza incidence in winter in China, 2004-2016]

Objective: To explore the association between low air temperature and influenza incidence in winter in 31 provinces in China. Methods: Influenza incidence and meteorological data were collected from January, 2004 to December, 2016 in 31 provinces in China. Distributed lag nonlinear model was applied to evaluate the association between low air temperature and influenza incidence in early and late winter, after controlling for the long-term trend, as well as the effect of "day of week" . Results: The influenza incidence increased from 4/100 000 in 2004 to 22/100 000 in 2016 in China, showing an overall increasing trend. The incidences of some provinces were lower than 5/100 000, such as Heilongjiang, Liaoning and Jilin, but the incidences of influenza in Gansu, Hebei, Guangdong, Beijing were higher than 30/100 000 in some years. From January to June, the influenza incidence showed a decreasing trend with the increase of air temperature, but the influenza incidence showed a rising trend from July to December with the decrease of air temperature. In early winter, low air temperature had the highest influence on the incidence of influenza in Beijing, the RR was 2.90 (2.57-3.28), but in late winter, low air temperature had the highest influence on the incidence of influenza in Tibet, the RR was 3.37 (2.03-5.58). In early winter, low temperature had the highest influence on the incidence of influenza in northeastern China, the RR was 2.04 (1.48-2.79), but in late winter, low air temperature had the highest influence in southern China, the RR was 1.77 (1.61-1.94). There were area specific difference in the influence of low air temperature on the incidence of influenza between early winter and late winter. Conclusions: The correlation between low air temperature in winter and influenza incidence exists. The influenza of early and late winter on influenza incidence varies with different areas. We should take appropriate protective measures according to local conditions.