OBJECTIVE: High baseline YKL-40 is associated with later development of ischemic stroke, but not with myocardial infarction. Whether high YKL-40 levels are associated with increased risk of venous thromboembolism is presently unknown. We tested the hypothesis that observationally and genetically high YKL-40 is associated with increased risk of venous thromboembolism in the general population.

APPROACH AND RESULTS: Cohort and Mendelian randomization studies in 96 110 individuals from the Danish general population, with measured plasma levels of YKL-40 (N=21 647) and CHI3L1 rs4950928 genotype (N=94 579). From 1977 to 2013, 1489 individuals developed pulmonary embolism, 2647 developed deep vein thrombosis, and 3750 developed venous thromboembolism (pulmonary embolism and deep vein thrombosis). For the 91% to 100% versus 0% to 33% YKL-40 percentile category, the multifactorially adjusted hazard ratio was 2.38 (95% confidence interval, 1.25-4.55) for pulmonary embolism, 1.98 (1.09-3.59) for deep vein thrombosis, and 2.13 (1.35-3.35) for venous thromboembolism. Compared with rs4950928 GG homozygosity, presence of C-allele was associated with a doubling (CG) or tripling (CC) in YKL-40 levels, but not with risk of venous thromboembolism. A doubling in YKL-40 was associated with a multifactorially adjusted observational hazard ratio for pulmonary embolism of 1.17 (1.00-1.38) and a genetic odds ratio of 0.97 (0.76-1.23). Corresponding risk estimates were 1.28 (1.12-1.47) observationally and 1.11 (0.91-1.35) genetically for deep vein thrombosis and 1.23 (1.10-1.38) observationally and 1.08 (0.92-1.27) genetically for venous thromboembolism.

CONCLUSIONS: High YKL-40 levels were associated with a 2-fold increased risk of venous thromboembolism, but the association was not causal.