Changes in haemostatic variables induced by oral contraceptives containing 50 micrograms or 30 micrograms oestrogen: absence of dose-dependent effect on PAI-1 activity

Abstract

Several studies have suggested a dose-response relation between the oestrogen content of oral contraceptive (OC) and the risk of both venous thrombosis and arterial disease, when oestrogen doses were higher than 50 micrograms. However, there is no clear epidemiological evidence for a decrease in thrombotic risk with formulations containing less than 50 micrograms oestrogen. Therefore, we investigated haemostatic variables in users of OC containing either 30 micrograms (35 women) or 50 micrograms (29 women) ethinyl estradiol as compared with non users (64 women) matched for age and smoking status. Mean values of antithrombin activity were significantly lower in 30 micrograms or 50 micrograms oestrogen users than in non users (96% and 98% vs 105%, respectively, p < 0.001), but they were not significantly different between the two groups of OC users. There was a significant increase in mean values of factor VII antigen in women taking either 30 micrograms or 50 micrograms oestrogen as compared with non users (96% and 101% vs 85%, respectively, p < 0.005). Although the difference between both groups of OC users was not significant, a positive linear trend in factor VII levels was observed within the 0-50 micrograms oestrogen range (p < 0.001). Mean levels of fibrinogen were slightly higher in 30 micrograms or 50 micrograms oestrogen users than in non users (2.71, 2.66 g/l vs 2.55 g/l, respectively), but there was no significant difference between the three groups.(ABSTRACT TRUNCATED AT 250 WORDS)

PIP: In Paris, France, clinicians compared data on 64 women aged 19-40 who used combined oral contraceptives (OCs) for 6-200 months with data on 64 healthy women who did not use OCs for the last two months and who were matched for age and smoking status to investigate activity of plasminogen activator inhibitor 1 (PAI-1), factor VII antigen, fibrinogen concentration, and antithrombin activity in users of OCs containing either 30 mcg or 50 mcg estrogen and in nonusers. OC users exhibited lower mean values of PAI-1 activity than nonusers (4.63-4.89 vs. 6.47 AU/ml; p 0.02). There was no dose-dependent effect of estrogen on PAI-1 activity, however. Antithrombin activity values were much lower in OC users than nonusers (96-98% vs. 105%; p 0.001). The difference between the two groups of OC users was not significant, however. The mean values of factor VII antigen in women using either 30 mcg or 50 mcg estrogen were higher than those for nonusers (96% and 101% vs. 85%, respectively; p 0.005). The difference in factor VII antigen values between the two OC groups was not significant, yet there was a positive linear trend in factor VII levels within the 0-50 mcg estrogen range (p 0.001). No significant difference in the mean fibrinogen levels between the three groups (30 mcg estrogen OC group, 50 mcg estrogen OC group, and nonusers) was observed. Hemostatic variables were not significantly different between 30 mcg estrogen OCs containing 100 mcg, 150 mcg, or 200 mcg levonorgestrel. The researchers could not conduct a valid assessment of the progestogen effect in 50 mcg estrogen OCs due to the wide range of different types of progestogens. These findings suggest an alteration of blood coagulation and fibrinolysis in OC users within the 30-50 mcg estrogen range. Estrogen appears to have a dose-dependent effect on factor VII but no significant effect on PAI-1 activity and other markers of thrombogenic risk and arterial disease risk.