Letter to the Editor: Influenza Vaccination and Short-Term Risk of Stroke Among Elderly Patients With Chronic Comorbidities in a Population-Based Cohort Study

Abstract

To the Editor,

We read with great interest the article “Influenza vaccination and short-term risk of stroke among elderly patients with chronic comorbidities in a population-based cohort study ”by Wang et al. [1]. While this study shed light on valuable clinical and stroke outcomes of influenza vaccination in a Shanghai Cohort, a few additional elements could have strengthened the article's robustness.

First, the vaccination rate in this cohort was very low (1.25%–1.55%), resulting in only 786 and 589 vaccinated participants in the 2017–2018 and 2018–2019 influenza seasons, respectively. Within these small subgroups, merely 5 and 7 stroke events occurred. Such sparse outcomes render hazard ratio estimates unstable and associated confidence intervals wide, raising concerns about the robustness of the findings [2, 3].

Additionally, despite the application of multivariable adjustment and propensity score matching, residual confounding is highly likely. Vaccinated participants differed significantly from their unvaccinated counterparts—they were younger, more often male, more likely to smoke or drink, more frequently hypertensive, and less often diabetic [4]. Furthermore, important covariates such as medication use (antiplatelets, anticoagulants, antihypertensives, statins), stroke subtype (ischemic vs. hemorrhagic), and prior influenza infection history were not available. These unmeasured factors may strongly bias effect estimates [5].

Thirdly, the exposure definition—vaccination at any time between August and March—does not account for temporal variation in influenza circulation and risk periods [6]. The use of Poisson regression as a sensitivity analysis is also problematic, as extremely low event counts may artificially inflate effect sizes. Additionally, the BMI categories applied do not align with population-specific Chinese cutoffs, which could lead to misclassification [7].

In light of these limitations, the impressive reduction in stroke risk reported should be interpreted with caution. While the hypothesis that influenza vaccination may confer cardiovascular protection is biologically plausible and of great clinical relevance, more definitive evidence will require larger, multicenter cohorts with longer follow-up, precise stroke classification, and comprehensive adjustment for potential confounders.

We appreciate the authors’ contribution and hope that these observations will help guide future investigations in this important area.

In conclusion, I commend the authors for their significant work and recommend that further studies consider these limitations. Doing so would help enhance the strength, consistency, and clinical applicability of the findings.