Influenza Vaccination and Risk of Ischemic Stroke: A Population-Based Case-Control Study

Abstract

Background and objectives:To assess the relationship between influenza vaccination in the general population and risk of a first ischemic stroke (IS) during pre-epidemic, epidemic and post-epidemic periods.Methods:A nested case-control study was carried out in a Spanish primary care database over 2001-2015. Subjects aged 40-99 years with at-least 1-year registry and no history of stroke or cancer were selected to conform the source cohort, from which incident IS cases were identified and classified as cardioembolic or non-cardioembolic. Five controls per case were randomly selected, individually matched with cases for exact age, sex and date of stroke diagnosis (index date). A patient was considered vaccinated when he/she had a recorded influenza vaccination at least 14 days before the index date within the same season. Adjusted odds ratios (AOR) and their respective 95% confidence intervals (CI) were computed through a conditional logistic regression. Pneumococcal vaccination was used as a negative control.Results:From a cohort of 3,757,621 patients, we selected 14,322 incident IS cases (9,542 non-cardioembolic and 4,780 cardioembolic) and 71,610 matched controls. Of them, 41.4% and 40.5%, respectively, were vaccinated yielding a crude OR of 1.05(95%CI:1.01-1.10). Vaccinated subjects presented a higher prevalence of vascular risk factors, diseases and comedication than non-vaccinated and, after full adjustment, the association of influenza vaccination with IS yielded an AOR of 0.88(95%CI:0.84-0.92) was found, appearing early (AOR15-30 days=0.79;95%CI:0.69-0.92) and slightly declining over time (AOR>150 days=0.92;95%CI:0.87-0.98). A reduced risk of similar magnitude was observed with both types of IS, in the three epidemic periods and in all subgroups analyzed (men, women, subjects below and over 65 years of age, and subjects with intermediate and high vascular risk). By contrast, pneumococcal vaccination was not associated with a reduced risk of IS (AOR=1.08;95%CI:1.04-1.13).Discussion:Results are compatible with a moderate protective effect of influenza vaccine on IS appearing early after vaccination. The finding that a reduced risk was also observed in pre-epidemic periods suggests that either the “protection” is not totally linked to prevention of influenza infection, or it may be partly explained by unmeasured confounding factors.