As for men, anticonvulsant medications also affect endocrine reproductive function in women in clinically important ways, including increasing levels of sex hormone-binding globulin (SHGB). There are additional considerations that differ from the effects on men, however.

Polycystic ovary syndrome (PCOS)

Much attention has been focused on the association between valproate and PCOS in women.122

The NIH diagnostic criteria for PCOS include:

  • ovulatory dysfunction, polymenorrhea, oligomenorrhea, or amenorrhea
  • clinical evidence of hyperandrogenism and/or hyperandrogenemia
  • exclusion of other endocrinopathies

Cystic ovaries are often present but are not necessary to meet the diagnostic criteria. PCOS is also associated with obesity.

In a small group of women with epilepsy, Isojarvi and colleagues reported that PCOS was more common in those treated with sodium valproate (approximately 45%) than in those treated with carbamazepine (20%).122 In a second report, they noted that some features of PCOS improved when valproate was replaced with lamotrigine.123 These features were:

  • body-mass index
  • fasting serum insulin
  • serum testosterone
  • number of ovarian cysts

These findings suggest that valproate causes a cascade of symptoms that starts with weight gain, insulin resistance, and hyperinsulinemia, which then causes hyperandrogenism,124 leading to PCOS.

The association between PCOS and valproate use is controversial, however, since these findings have not been borne out by other authors, who found no association between PCOS and valproate use.125 Valproate therapy is associated with elevations in testosterone and triglyceride levels in women with epilepsy, although PCOS in association with valproate was not found in this study population.126

Hyperinsulinemia may indeed cause hyperandrogenism by directly stimulating ovarian steroidogenesis127 and by inhibiting the synthesis of SHBG, with consequent increase in the availability of bioactive androgens.

Others argue, however, that PCOS is related to the underlying epilepsy and that the difference in the incidence of PCOS between women treated with valproate and those treated with other AEDs may be due to a beneficial effect upon PCOS by hepatic enzyme-inducing AEDs such as barbiturates, phenytoin, and carbamazepine.128 Among women with temporal lobe epilepsy (TLE), PCOS is less common in treated women (13%) than in untreated women (30%) when treatment does not include sodium valproate.129 Hepatic enzyme-inducing AEDs lower biologically active androgen levels, whereas valproate does not. AEDs other than valproate, therefore, may treat epilepsy-related hyperandrogenism and thus PCOS. This mechanism could contribute to the higher occurrence of PCOS in women with epilepsy who are treated with valproate, which has no such beneficial effect.128

It is also possible that both mechanisms pertain:

  1. Valproate may, by causing hyperinsulinemia, cause or exacerbate PCOS in some women with epilepsy.
  2. AEDs such as barbiturates, phenytoin, and carbamazepine may treat PCOS by treating hyperandrogenism.

Other effects

Other effects of valproate on reproduction may include amenorrhea and pubertal arrest seen transiently after initiation of valproate therapy.130,131 Valproate also has been associated with anovulatory cycles in women with primary generalized epilepsy.132 In this study, however, most women with this epilepsy type were taking valproate, so separating the relationship between and ovulation and epilepsy type and valproate use is difficult.132

Adapted from: Klein P and Herzog AG. Endocrine aspects of partial seizures. In: Schachter SC, Schomer DL, eds. The comprehensive evaluation and treatment of epilepsy. San Diego, CA: Academic Press; 1997. p. 207-232.
With permission from Elsevier (

Authored By: 
Pavel Klein MD
Andrew G. Herzog MD
Reviewed By: 
Cynthia L. Harden MD
Sunday, February 1, 2004