Examining the role of viral evolution on seasonal influenza incidence

The seasonal occurrence of influenza is well characterized, but lacks a definitive cause. Seasonal variations in P&I mortality in Wisconsin between 1967 and 2004 was examined, with respect to the dominant circulating strains of H3N2, H1N1 and B influenzas.

Using Poisson regression adapted for time series analysis, annual epidemic curve characteristics for outbreak timing, amplitude, duration and average mortality rates were estimated. The characteristics for three population subgroups, 65+ years, 6-64 years and 0-5 years were compared to known circulating strains using ANOVA.

From 1967 to 1969, the A/Hong Kong/1/68 pandemic strain was responsible for an increase in mortality of 1.93 deaths/million in 0 to 5 and 0.58 deaths/millions in those 6 to 64, compared to the average rate throughout the duration of the series (Tukey's HSD, α=0.10). No rise was seen in the 65+ mortality rate; however, it was higher for an 11 year period between 1986 and 1996, when H1N1 viruses of the A/Singapore/6/86-like strain circulated. During this period, the average mortality rate in this subgroup increased by 1.98 deaths/million (Tukey's HSD, α=0.10) compared to the average.

The A/Singapore/6/86 strain, associated with increased mortality for those 65 and older, though showing no increase in other subgroups, was not previously noted for pandemic potential or heightened virulence. These findings suggest that further attention should be given to the role of individual viruses in annual epidemics, as well as the potential impact of a patient's age and immunological history in determining the severity of the disease in a population.