Causal effect of serum 25 hydroxyvitamin D concentration on cardioembolic stroke: Evidence from two-sample Mendelian randomization

Abstract

Background and aimsThe putative association between serum 25-hydroxyvitamin D concentration [25(OH)D] and the risk of cardioembolic stroke (CES) has been examined in observational studies, which indicate controversial findings. We performed Mendelian randomization (MR) analysis to determine the causal relationship of serum 25(OH)D with the risk of CES. Methods and resultsThe summary statistics dataset on the genetic variants related to 25(OH)D was used from the published GWAS of European descent participants in the UK Biobank, including 417,580 subjects, yielding 143 independent loci in 112 1-Mb regions. GWAS summary data of CES was obtained from GIGASTROKE Consortium, which included European individuals (10,804 cases, 1,234,808 controls). Our results unveiled a causal relationship between 25(OH)D and CES using IVW [OR = 0.82, 95% CI: 0.67–0.98, p = 0.037]. Horizontal pleiotropy was not seen [MR-Egger intercept = 0.001; p = 0.792], suggesting an absence of horizontal pleiotropy. Cochrane's Q [Q = 78.71, p-value = 0.924], Rucker's Q [Q = 78.64, p-value = 0.913], and I2 = 0.0% (95% CI: 0.0%, 24.6%) statistic suggested no heterogeneity. This result remained consistent using different MR methods and sensitivity analyses, including Maximum likelihood [OR = 0.82, 95%CI: 0.67–0.98, p-value = 0.036], Constrained maximum likelihood [OR = 0.76, 95%CI: 0.64–0.90, p-value = 0.002], Debiased inverse-variance weighted [OR = 0.82, 95%CI: 0.68–0.99, p-value = 0.002], MR-PRESSO [OR = 0.82, 95%CI 0.77–0.87, p-value = 0.022], RAPS [OR = 0.82, 95%CI 0.67–0.98, p-value = 0.038], MR-Lasso [OR = 0.82, 95%CI 0.68–0.99, p-value = 0.037]. ConclusionOur MR analysis provides suggestive evidence that increased 25(OH)D levels may play a protective role in the development of cardioembolic stroke. Determining the role of 25(OH)D in stroke subtypes has important clinical and public health implications.