PLoS ONE (Jan 2020)

Association of acute myocardial infarction with influenza: A nationwide observational study.

  • Moman A Mohammad,
  • Johan Tham,
  • Sasha Koul,
  • Rebecca Rylance,
  • Cecilia Bergh,
  • David Erlinge,
  • Ole Fröbert

DOI
https://doi.org/10.1371/journal.pone.0236866
Journal volume & issue
Vol. 15, no. 8
p. e0236866

Abstract

Read online

IntroductionInfluenza may precipitate cardiovascular disease, but influenza typically peaks in winter, coinciding with other triggers of myocardial infarction (MI) such as low air temperature, high wind velocity, low atmospheric pressure, and short sunshine duration.ObjectiveWe aimed to determine the relationship of week-to-week variation in influenza cases and acute MI, controlling for meteorological factors in a nationwide population.MethodsWeekly laboratory-confirmed influenza case reports were obtained from the Public Health Agency of Sweden from 2009 to 2016 and merged with the nationwide SWEDEHEART MI registry. Weekly incidence of MI was studied with regard to number of influenza cases stratified into tertiles of 0-16, 17-164, and >164 cases/week. Incidence rate ratios (IRR) were calculated using a count regression model for each category and compared to a non-influenza period as reference, controlling for air temperature, atmospheric pressure, wind velocity, and sunshine duration.ResultsA total of 133562 MI events was reported to the registry during the study period. Weeks with influenza cases were associated with higher incidence of MI than those without in unadjusted analysis for overall MI, ST-elevation MI and non ST-elevation MI independently. During the influenza season, weeks with 0-16 reported cases/week were not associated with MI incidence after adjusting for weather parameters, adjusted IRR for MI was 1.03 (95% CI 1.00-1.06, P = 0.09). However, weeks with more cases reported were associated with MI incidence: 17-163 reported cases/week, adjusted IRR = 1.05 (95% CI 1.02-1.08, P = 0.003); and for ≥164 cases/week, the IRR = 1.06 (95% CI 1.02-1.09, P = 0.002). Results were consistent across a large range of subgroups.ConclusionsIn this nationwide observational study, we found an association of incidence of MI with incidence of influenza cases beyond what could be explained by meteorological factors.