OBJECTIVE: Cerebral venous thrombosis (CVT) is an uncommon form of stroke affecting mostly young individuals. Although genetic factors are thought to play a role in this cerebrovascular condition, its genetic etiology is not well understood. METHODS: Genome-wide association study performed to identify genetic variants influencing susceptibility to CVT. A two-stage genome-wide study was undertaken in 882 Europeans diagnosed with CVT and 1205 ethnicity-matched control subjects divided into discovery and independent replication datasets. RESULTS: In the overall case-control cohort, we identified highly significant associations with 37 SNPs within 9q34.2 region. The strongest association was with rs8176645 (combined P = 9.15 × 10(-24) ; OR = 2.01, 95%CI: 1.76-2.31). The discovery set findings were validated across an independent European cohort. Genetic risk score for this 9q34.2 region increases CVT risk by a pooled estimate OR = 2.65 (95%CI: 2.21-3.20, P = 2.00 × 10(-16) ). SNPs within this region were in strong linkage disequilibrium (LD) with coding regions of the ABO gene. ABO blood group was determined using allele combination of SNPs rs8176746 and rs8176645. Blood groups A, B or AB, were at 2.85 times (95%CI: 2.32-3.52, P = 2.00 × 10(-16) ) increased risk of CVT compared with individuals with blood group-O. INTERPRETATION: We present the first chromosomal region to robustly associate with a genetic susceptibility to CVT. This region more than doubles the likelihood of CVT, a risk greater than any previously identified thrombophilia genetic risk marker. That the identified variant is in strong LD with the coding region of the ABO gene with differences in blood group prevalence provides important new insights into the pathophysiology of CVT. This article is protected by copyright. All rights reserved.Read less <