Factor V Leiden and third-generation oral contraceptives

The authors respond to Carolyn Westhoffs critique of their study on the association between exposure to oral contraceptives and deep vein thrombosis (DVT). They found an apparent multiplication of risks which was more clearly present for oral contraceptives containing a third-generation progestogen than for those containing a second-generation progestogen. The risks seem to multiply but there is no reason why they should. The authors explain that the appearance of independence of such risks is merely the result of the use of multiplicative models that is relative risks or logistic models. These models are a simple result of a mathematical convenience and have no intrinsic biological meaning which becomes clearer when looking at the same data on an additive scale. The authors agree with Westhoff that the biological interpretation of interaction is considerably more difficult. They do not however believe that they overstated misinterpreted or misled by suggesting that the observed difference in risk among factor V Leiden positive individuals may offer a clue to the biological mechanism underlying the increased risk of venous thrombosis of third-generation oral contraceptives.