The association of smoking with venous thrombosis in women and its modification by exogenous hormones

Abstract

Background: The evidence for an association between smoking and venous thrombosis (VT) remains inconsistent. In particular, the nature of this association has not been reported among users of hormone therapy (HT). Methods: We studied 2125 women with validated VT and 5749 controls from a large healthcare system in Washington State. Smoking status (current, former, never) was assessed from medical records review and, for a subset, also by telephone interview. The agreement between the 2 measures was excellent (Kappa 0.84). Current hormone use was calculated from pharmacy databases. We used multivariate logistic regression to evaluate the association between smoking and VT, and its modification by the use of HT. Adjustment variables included BMI, race, diabetes and pregnancy. Results: Our sample comprised mostly postmenopausal white women, with a mean age of 65 years and a smoking prevalence of 10%. Current smokers were at higher risk of VT than non-smokers (OR 1.21, 95%CI 1.02-1.46). This risk did not differ between women not using and using HT: OR 1.2 (95%CI 1.0-1.5) and 1.2 (95%CI 0.8-1.9), respectively (p for interaction = 0.73). A subgroup analysis among women in their first year of use of HT suggested a higher risk for current smokers, compared to non-smokers, however without evidence of an interaction (OR 3.3, 95%CI 0.9-12.6, p for interaction = 0.13). In former smokers, an association between smoking and VT was only present during the first year after smoking cessation (OR 2.7, 95%CI 2.0-3.8), without suggestion of an increased risk thereafter. Conclusion: Among women in the US, smoking was a weak risk factor for VT and was not modified in women who had used HT for >1 year. Further studies are needed to elucidate the association of smoking with VT during the first year of treatment with HT.