Oral estrogen therapy has been associated with increased risk of venous thromboembolism in postmenopausal women in observational studies and randomized controlled trials. The relationship between transdermal estrogen and venous thromboembolism is less clear. Scarabin and colleagues conducted a case-control study to investigate whether transdermal estrogen therapy increases the risk of venous thromboembolism.

They identified 155 consecutive cases of unexplained venous thromboembolism in post-menopausal women 45 to 70 years of age who were treated at seven French teaching hospitals. Exclusions from the study included a history of venous thromboembolism, heart disease, or stroke; cancer; systemic inflammatory illness; or recent surgery or prolonged bed rest. Women with any contraindication to estrogen therapy also were excluded. Venous thromboembolism was confirmed by ultrasonography, and for each patient with venous thromboembolism, up to three control subjects were identified from women of the same age and area of residence who were admitted to the same hospital for conditions not thought to be associated with estrogen therapy. Patients with elevated blood pressure, obesity, or diabetes were matched to similar control subjects. Patients and control subjects were interviewed to gather data about demographic and health issues, especially gynecologic history and hormone use.

Although the overall proportion of women at high risk for venous thromboembolism was similar in both groups (28 percent of women with venous thromboembolism compared with 26 percent of control subjects), patients with venous thromboembolism were significantly more likely to have varicose veins, obesity, or a family history of venous thromboembolism. Oral estrogen was used by 21 percent of women with venous thromboembolism and 7 percent of control subjects, and transdermal estrogen was used by 19 percent of patients with venous thromboembolism and 24 percent of control subjects. After statistical adjustment for confounding factors, the odds ratio for venous thromboembolism in women using oral estrogen was 3.5 compared with 0.9 in women using transdermal preparations (see the accompanying table). The risk did not appear to change with the duration of transdermal estrogen use.

The authors conclude that no association was found between transdermal estrogen use and venous thromboembolism, but that the increased risk in patients taking oral estrogen was confirmed. They note that the difference could be attributed to the reduced plasma estrone-to-estradiol ratios achieved by transdermal preparations. This ratio is approximately one with transdermal estrogen but closer to five with oral preparations. Oral preparations also increase plasma concentrations of clotting factors and have other effects on coagulability. The authors call for further study of the role of transdermal estrogen therapy in postmenopausal women.