A large genome-wide association study identified genetic variants at the ABO blood group locus that seem to be significantly associated with early-onset ischemic stroke.

The study, published in Neurology, found that these variants had stronger effects in early-onset stroke compared with late-onset stroke and may indicate an increased role of prothrombotic mechanisms in younger stroke patients.

Researchers analyzed genetic data from 16,730 cases of early-onset stroke (defined as occurring from age 18-59 years) and 599,237 controls across 48 studies. The study population included 990 individuals of African ancestry; 11,114 of European ancestry; 189 of Hispanic ancestry; 230 of Afro-Caribbean ancestry; and 4,207 of pan-Asian ancestry.

Methods

The study combined data from the Early Onset Stroke Consortium and additional sites. Brain imaging confirmation of ischemic stroke was utilized. Genotypes were imputed using the TOPMed reference panel.

The primary analysis was a transethnic meta-analysis, with a parallel European-only analysis. Logistic regression tested associations between variants and stroke, adjusting for sex and ancestry. Effect sizes were compared between early-onset and late-onset stroke using a Wald test.

ABO blood groups were assigned using two single-nucleotide polymorphisms (SNPs). Genetic correlations were estimated using LD Score Regression. A polygenic risk score for venous thromboembolism (VTE) was created using PRSice software. Colocalization analyses used the coloc package.

Power calculations indicated 80% power to detect odds ratios (ORs) ranging from 1.09 to 1.20 for common variants (minor allele frequencies [MAF] >5%) at genome-wide significance threshold.

Findings

Two genome-wide significant loci were identified, both mapping to the ABO gene. The lead variants tagged blood subgroups O1 and A1:

• rs975381715 (O1-tagging): OR = 0.88 (95% confidence interval [CI] = 0.85-0.92; P = 3.72 × 10−10)
• rs8176685 (A1-tagging): OR = 1.16 (95% CI = 1.10-1.21; P = 2.06 × 10−9)

These associations were significantly stronger for early-onset stroke compared with late-onset stroke:

• O1 variant: OR = 0.88 vs 0.96 (P for interaction = .0004)
• A1 variant: OR = 1.16 vs 1.05 (P for interaction = .001)

The researchers estimated that 6% of early-onset stroke cases in Europeans could be attributed to the A1-tagging variant, compared with only 2% of late-onset cases.

Analysis of ABO blood group distribution revealed:

• Early-onset stroke cases: 48.4% in group A, 4.8% in group AB, 11.3% in group B, 35.5% in group O
• Late-onset stroke cases: 45.2% in group A, 4.7% in group AB, 11.0% in group B, 39.1% in group O
• Controls: 44.4% in group A, 4.3% in group AB, 10.1% in group B, 41.1% in group O

In Europeans, associations of ABO subgroups with early-onset stroke and late-onset stroke were:

• O1 (rs529565): early-onset OR = 0.88 (P = 4.31 × 10−14); late-onset OR = 0.96 (P = .036)
• A1 (rs635634): early-onset OR = 1.16 (P = 6.54 × 10−13); late-onset OR, 1.05 (P = .044)
• B (rs1137827): early-onset OR = 1.05 (P = .324); late-onset OR = 1.02 (P = .591)

To explore potential mechanisms, the team examined associations with VTE. They found stronger associations of ABO variants with early-onset VTE vs late-onset VTE, as well as a higher genetic correlation between early-onset stroke and VTE (0.376 ± 0.153, P = .014) compared with late-onset stroke and VTE (0.032 ± 0.198, P = .87). A polygenic risk score for VTE was more strongly associated with early-onset stroke (OR = 1.13 per standard deviation increase; 95% CI = 1.10-1.16; P < 5.21 × 10−16) than late-onset stroke (OR = 1.04; 95% CI = 1.01-1.08; P = .010).

The study also revealed evidence of colocalization between early-onset stroke and VTE, as well as with plasma levels of von Willebrand factor and factor VIII, at the ABO locus. Strong evidence for shared causal SNPs was found between early-onset stroke and VTE, von Willebrand factor, and factor VIII (posterior probability >99% for all pairs).

Stroke subtype analysis, available for 69.5% of all stroke cases, identified 11 SNPs associated with various subtypes at genome-wide significance. Sample sizes ranged from 886 to 5,149 for transethnic analysis and 376 to 1,502 for European-only analysis.

Additional findings included:

• SHKBP1 rs118091666: genome-wide significant but rare (MAF = 0.018 in Europeans)
• 19 "suggestive" loci with subthreshold significance (P < 1 × 10−6) in transethnic analysis
• 14 loci with subthreshold significance in European-only analysis

The UK Biobank provided replication:

• ABO rs529565 (O1): early-onset stroke OR = 0.93 (P = .02); late-onset stroke OR = 0.95 (P = .02)
• ABO rs2519093 (A1): early-onset stroke OR = 1.10 (P = .03), late-onset stroke OR = 1.05 (P = .07)

These findings suggest prothrombotic mechanisms may play a larger role in early-onset stroke compared with late-onset stroke. The authors noted that while the effect sizes are too small for immediate clinical use, the results could inform future research into gene-environment interactions and rare variant studies targeting prothrombotic pathways.

Conflict of interest declarations were not made available at time of publishing.