The complex relationship between hypothalamic amenorrhea and polycystic ovary syndrome.

BACKGROUND: Polycystic ovarian morphology (PCOM) is occasionally observed in women with hypothalamic amenorrhea (HA). Although these women with HA/PCOM meet two of the Rotterdam criteria, they are excluded from the diagnosis of polycystic ovary syndrome (PCOS) by having HA. We explored the coexistence of these two disorders in women with HA/PCOM by analyzing their androgen response to gonadotropins and by following their clinical characteristics over time.

METHODS: Baseline and dynamic endocrine profiles during controlled ovarian hyperstimulation for women with HA/PCOM [n = 6, median (interquartile range) age 30 yr (28-31), body mass index (BMI) 19.2 kg/m(2) (18.0-19.2)] were retrospectively compared with those of women with PCOS [n = 10, age 33 (31-34), BMI 24.8 (23.2-27.6)] and normoovulatory controls [n = 20, age 33 (31-35), BMI 21.5(20.3-23.1)]. Long-term outcomes for five women with HA/PCOM were followed during their spontaneous recovery from HA.

RESULTS: With the exception of decreased LH [0.7 (0.3-0.8) vs. 6.0 IU/liter (4.8-7.4); P = 0.003], FSH [3.9 (2.5-5.7) vs. 7.5 IU/liter (5.3-9.5); P < 0.025], and estradiol [20 (14-24) vs. 32 pg/ml (20-39); P < 0.027], baseline endocrine profiles of women with HA/PCOM did not differ significantly from those of normoovulatory controls in terms of 17alpha-hydroxyprogesterone, dehydroepiandrosterone, dehydroepiandrosterone-sulfate, androstenedione, and total testosterone. However, controlled ovarian hyperstimulation with similar doses of gonadotropins resulted in an excess of androgen production compared with the controls [Deltaandrostenedione per dominant follicle 0.30 (0.23-0.37) vs. 0.10 ng/ml (0.05-0.18), P = 0.005; Deltatestosterone per dominant follicle 16 (7-24) vs. 6 ng/dl (2-12), P = 0.04], and these levels were comparable to those of women with PCOS. Recovery from HA/PCOM in some patients was associated with the development of oligomenorrhea and symptoms of androgen excess.

CONCLUSIONS: Women with HA/PCOM may have inherently hyperandrogenic ovaries but are quiescent due to low gonadotropins from the hypothalamic inactivity. The exaggerated ovarian androgen response to low-dose gonadotropin stimulation in these women is consistent with the clinical observation that hyperandrogenism emerges in association with weight gain and the recovery of hypothalamic function. Over time, these patients may fluctuate between symptoms of HA and PCOS, depending on the current status of hypothalamic activity. The fluidity of this transition in HA/PCOM challenges the simple dichotomous definition of PCOS using the Rotterdam criteria, which categorizes the two conditions as being mutually exclusive.