We greatly appreciate the interest Dr Kasapoglu has shown in our recent article assessing polycystic ovary syndrome (PCOS) as a risk factor for venous thromboembolism (VTE).

Our report of a possible protective effect of oral contraceptive pill (OCP) use on VTE risk among women with PCOS stems from our observation that OCP use appeared to modify the association between PCOS and VTE, with comparatively larger odds ratio estimates noted among women <35 years of age without report of OCP use than those with report of OCP use. To explain this unexpected finding, we speculated that because OCP suppresses the elevated levels of luteinizing hormone seen in women with PCOS and increases the production of sex hormone–binding globulin (SHBG) by the liver, reduced amounts of plasma testosterone will be available to bind to androgen receptors. We hypothesized that this reduction in androgens along with an increase of SHBG likely improved the impaired fibrinolysis present among women with PCOS since hyperinsulinemia, which can lead to hyperandrogenism, and lower SHBG have been shown to impair fibrinolysis by directly enhancing the secretions of plasminogen activator inhibitor type 1–an inhibitor of fibrinolysis and contributor to thrombosis development.

We agree that our measure of OCP use was limited by lack of information on adherence and timing relative to the VTE event. Comparisons between our findings and those from other studies cited by Dr Kasapoglu is difficult given that we defined PCOS by an algorithm using the presence of hyperandrogenism, ovulatory dysfunction, and/or the presence of polycystic ovaries; however, we believe that our estimates are actually consistent with those of Bird et al, who reported an increased risk of VTE among OCP users who have PCOS but a lower risk of VTE among women with PCOS who use OCPs.

Our intention was not to state that OCP use is definitely protective for VTE risk in women with PCOS but rather to highlight, as Dr Kasapoglu stated, that the association between OCP use and VTE risk in women with PCOS is inconclusive and merits further exploration. Administrative data such as those used in our study and by Bird et al provide limited information on PCOS due to possible misclassification as well as variability in the definition and phenotypic presentation of PCOS ; we concur that large, prospective studies with validated diagnoses are needed to shed more light on this complex syndrome and associated risks.