ASSOCIATIONS OF SEASONAL INFLUENZA ACTIVITY WITH METEOROLOGICAL PARAMETERS IN TEMPERATE AND SUBTROPICAL CLIMATES: GERMANY, ISRAEL, SLOVENIA AND SPAIN

Previous studies and observations have shown that the timing of seasonal influenza epidemics vary across latitude, suggesting the role of climatic and environmental conditions in influenza circulation. This study assessed the association of meteorological parameters with influenza activity in 3 study sites with temperate climate (Ljubljana in Slovenia, Castilla y Leon in Spain, and Berlin in Germany) and 6 district-level sites with subtropical climate (Israel). We used weekly influenza surveillance laboratory results from year 2000 in Spain and from year 2007 in other sites, up to year 2011. We obtained daily data for air temperature from ground station, precipitation from NASA Tropical Rainfall Measuring Missions, specific humidity (SH) and solar radiation from NASA Global Land Data Assimilation System. Both the satellite and climate model data had 0.25° spatial resolutions. We used binomial regression to model the weekly proportion of cases testing positive for influenza, with the meteorological parameters as the covariates. Using significance level of 0.05, we found the following significant associations. An increase in influenza proportion in Ljubljana was significantly associated with a decrease in temperature. In Castilla y Leon, increased influenza positivity was significantly associated with decreased SH and solar radiation. Influenza positivity in Berlin was significantly associated with decreased SH. In Israel, influenza positivity in 5 districts in the northern part was significantly associated with decreased SH while one of these districts was also significantly associated with increased rainfall. Influenza in the South district was significantly associated with decreased temperature. Our findings suggest that the relationship between influenza activity and meteorological parameters is complex and may be location dependent. Although our study only considered outdoor conditions, the significant associations revealed here generally support the theory that low absolute humidity and temperature increase influenza activity. Furthermore, these significant meteorological parameters could be predictive of influenza activity and should be incorporated into country-specific influenza transmission models.

_{*M. Brombarg (Israel CDC) contributed to this work.}