Purpose: Estrogen has been found to increase the risk of deep vein thrombosis (DVT), which is associated with venous ulcers. A matched case-control study was therefore performed to quantify the effect of hormone replacement therapy (HRT) on the occurrence of venous ulcers of the lower limbs.
Methods: Women presenting to a participating vascular surgery department between January and December 1997 with a first open venous ulcer served as cases. Controls were sampled among women with sub-acute conditions such as skin problems, cold, headache/migraine, sore throat and mild ear infections, and were matched on referral site and age (+/- 5 years). Subjects were eligible for this study if they were postmenopausal, and excluded if they had non-palpable pedal pulse or any chronic active diseases. The study consisted of an interviewer-administered risk factor questionnaire. Conditional multivariate logistic regression analyses were performed adjusting for history of DVT and unbalanced distributions in important covariates.
Results: Forty-four cases and 80 matched controls were recruited. The mean age of participants was 64 years. HRT users were significantly thinner (27% with body mass index (BMI = (kg/height (m2)) > or = 27 vs. 41%, p = 0.04) and less likely to have a history of DVT (12 vs. 33%, p = 0.02) and varicose veins (VV) (52 vs. 69%, p = 0.04) than non-users. A matched univariate logistic regression analysis was performed to quantify the effect of HRT on venous ulcers and generated an odds ratio (OR) of 0.14 with 95% confidence interval of (0.04, 0.49)95%: adjusting for prior DVT resulted in an OR of 0.29 (0.09, 0.93)95%. When adjusting for BMI, history of VV, exercise, education, and smoking, the OR observed was 0.03 (0.01, 0.13)95%: further adjustment for prior DVT gave an OR estimate of 0.06 (0.01, 0.50)95%CI.
Conclusions: Our study demonstrates that HRT has a protective effect on the occurrence of venous ulcers of the lower limbs. Adjusting for prior DVT moves the estimate closer to the null. DVT is a strong contraindication for HRT and therefore must be considered in order to eliminate confounding by contraindication.