The evaluation of quantitative risk factors of venous thromboembolism is characterized by several unsolved problems. The majority of quantitative components of hemostasis are dependent on age, sex, and hormone intake. The cut-off values determined in case-control studies depend on the specific patient/control group and can not be generalized on individuals with other characteristics. Furthermore, the approach does not give reference values dependent on age, sex, and hormone intake which are necessary for risk estimations in clinical practice.
To overcome these disadvantages, we used a multiple regression analysis to create a system of reference values which change continuously depending on age, sex, and hormone intake according to the parameter distribution in healthy controls and calculated the relative risk of hemostatic components (729 patients with first VTE and 675 healthy controls). A significantly increased risk for venous thrombosis was associated with deficiency of protein S activity (odds ratio (OR) 2.8 to 6.1, p=0.0007), free protein S concentration (OR 2.7 to 20.4, p=0.0001), protein C activity (OR 3.2 to 9.4, p=0.0001), antithrombin activity (OR up to 75, p=0.0001), and increased levels of fibrinogen (OR 3.8, p=0.0001), factor VIII:C (OR 3.3, p=0.0001), factor IX (OR 2.3, p=0.0001), factor XI (OR 2.9, p=0.0001), vWF activity (OR 2.4, p=0.0001), and vWF antigen (OR 3.5, p=0.0001).
In contrast to previously published studies, this new approach gives clinically important cut-off values dependent on age, sex, and hormone intake and allows to identify patients at increased risk for venous thrombosis on an individualized basis. Particularly parameters which are highly dependent on age and sex, like protein S activity, can be characterized more precisely using the described procedure. This comprehensive analysis demonstrates that, apart from well-known risk determinants, deficiency of protein S and increased values of FI, FVIII:C, FIX, FXI, vWF, and vWF-Ag are risk determinants predicting venous thrombosis.