Association of influenza seasonality with temperature related indicators

E.Lofgren, N. Fefferman, G.Sedmak, J.Gorski, E.N.Naumova

Background

• Influenza seasonality is one of the best characterized, and least understood, aspects of the dynamics of non-pandemic influenza infection.

Objective

• To examine the association of seasonal infection rates with ambient temperature and temperature-related factors.

Method

• A data set of 2619 laboratory confirmed cases of influenza in Milwaukee, WI, adjusted for population, was analyzed against an array of climatological variables obtained from NOAA for the period between Jan. 1967 & June 2005.
  
• Using regression modeling adapted for time series studies and contour mapping techniques, the relationship between influenza incidence and temperature related indicators was elucidated.

Results

• Influenza A infection was found to be strongly associated with a decrease in monthly temperature (r=0.418, p <.0001).
  
• Incidence was greatest in the month after the coldest month of the year (18.1±3.7 cases/mill. persons/mnth), and during relatively warm winters.
  
• The typical two-year cycle of seasonality was not observed in these data, and Influenza B had a two month delay in occurrence compared to Influenza A.

Discussion

• These preliminary results suggest a complex relationship between virus, host and environment than had been previously been considered in the context of seasonality.
  
• We are supplementing the analysis with influenza mortality and hospitalization data to further explore the full nature of Influenza seasonality, on an epidemiological, mathematical, and virological level.