COVID-19 and stroke MR study: data errors, timing gaps, statistical flaws.

Publication date: Jun 30, 2025

This correspondence critiques the Mendelian randomization study by Zhang et al. on the causal link between COVID-19 and ischemic stroke, highlighting three methodological concerns. First, the study incorrectly attributes the eQTL data to the GTEx database, whereas the actual source is the eQTLGen Consortium, necessitating correction for accurate data provenance. Second, the outcome data (collected prior to 2018) predate the COVID-19 pandemic (post-2019), violating the relevance assumption in instrumental variable analysis. This temporal mismatch may render observed associations biologically implausible, reflecting genetic pleiotropy or residual confounding rather than causal effects of SARS-CoV-2 infection. Third, the absence of multiple testing correction elevates the risk of false positives, undermining the validity of subsequent gene pathway analyses. Addressing these issues-revising data attribution, clarifying temporal limitations, and enhancing statistical rigor-would strengthen the study’s reliability and translational implications.

Open Access PDF

Semantics

Original Article