Influenza–pneumonia epidemiologic puzzle solved
By Kirsty Oswald, medwireNews Reporter
01 July 2013
Sci Transl Med 2013; 5: 191ra84

medwireNews: Influenza has a very strong but transient effect on the risk for Streptococcus pneumoniae infection, increasing susceptibility by a remarkable 100-fold - but only in the week following infection, research shows.

The researchers, reporting in Science Translational Medicine, say the findings help to explain why epidemiologic studies have failed to find a consistent pattern between the two pathogens at the population level, despite it being long established that influenza exacerbates pneumococcal infection risk in individuals.

The team, led by Sourya Shrestha (University of Michigan, Ann Arbor, USA), used an adapted SIRS (susceptible, infected, recently recovered) model of pneumococcal transmission with influenza incidence as a covariate and applied it to weekly records of influenza and pneumococcal pneumonia hospitalizations in Illinois, USA before and after the introduction of the pneumococcal vaccine.

They then tested three mutually exclusive hypotheses: transmission impact, in which individuals infected with pneumococcal pneumonia contribute more to transmission if they have recently been infected with influenza; susceptibility impact, in which individuals infected with influenza are more susceptible to pneumococcal pneumonia; and pathogenesis impact, in which influenza infection leads to increased severity of subsequent pneumococcal pneumonia, increasing the odds of detection.

The authors found no evidence to support the transmission or severity impact models. By contrast, susceptibility impact had a significant effect, such that influenza infection increased the risk for pneumococcal pneumonia 85 to 115-fold. However, there was no evidence of an interaction beyond 1 week of influenza infection.

During peak flu season this translated to 40% of pneumococcal pneumonia being attributable to influenza, whereas outside of peak season, it accounted for less than 1% of cases. This resulted in an annual attributable fraction of 2% to 10%, consistent with prior epidemiologic accounts.

Moreover, using simulated datasets, the authors show that interannual variations in influenza peaks have only a modest effect on pneumococcal pneumonia peaks, providing a further cause for the lack of association in previous epidemiologic studies.

The association between influenza outbreaks and pneumonia incidence has been reported for over 2 centuries, explain the authors, and more recently, pneumonia cases spiked during the A/H1N1 influenza epidemic in 2009.

But while studies in animals and at the individual level have confirmed this, epidemiologic studies have found only moderate effects or none at all.

The authors say that traditional statistical techniques may mask the patterns they identified using their novel approach, through which they have "unpacked the population-level implications of immune-mediated processes shown to occur at the level of the individual."

medwireNews ( is an independent clinical news service provided by Springer Healthcare Limited. © Springer Healthcare Ltd; 2013

Free abstract

Friendly links

Follow me on Twitter