Risk of influenza after Previous Acute Respiratory Illness in a Cohort of Households with Children


Background. Understanding interactions between the agents that cause respiratory infections can provide valuable insights into pathogenesis of acute respiratory illness (ARI). Viral-bacterial interactions are well studied, particularly influenza and S. pneumoniae. Recent studies have suggested viral competition and innate immunity as mechanisms of viral-viral interactions, though these studies are less common.

Methods. We used prior ARI as a proxy for specific respiratory pathogens to examine potential interaction between previous infection and the risk of influenza over 3 years of observation in a prospective cohort of households with children. Regression models used robust standard errors to account for clustering within individuals.

Results. For those with a prior ARI, risk of influenza tended to increase as the time between illnesses increased, though some variability was observed across study years (figure). Previous ARI at any point (RR: 1.29 95% CI 1.03-1.62) and onset of ARI greater than 8 weeks prior to current illness (RR: 1.55 95% CI 1.17-2.01) were significantly associated with an increased risk of influenza compared to those with no previous ARI. However, after adjusting for calendar time between illness onset and peak influenza transmission the association was attenuated for all outcomes and was reversed for those with prior ARI at any point (RR: 0.79 95% CI.0.63-0.98). In fully adjusted models, the risk of influenza was significantly lower for illnesses with an ARI that began 4-8 weeks prior to the current illness (RR: 0.63 95% CI 0.43-0.93) compared to those without a previous ARI.

Conclusion. The observed increase in risk of influenza with any prior ARI appears to be driven by illnesses that occurred more than 8 weeks before the onset of the influenza illness. However, this finding could also be due to underlying susceptibly that is not captured by high-risk health conditions. Significantly lower risk for those with an ARI that began between 4 and 8 weeks prior to the current illness may imply competition among pathogens that cause ARI and warrants further study. Additional analyses will examine interactions among specific viruses specific. These results suggest that calendar time is an important confounder when examining the effects of respiratory pathogen interaction on virus specific outcomes.

Open Forum Infectious Diseases