Long-term Follow-up of Functional Hypothalamic Amenorrhea and Prognostic Factors

Abstract

Functional hypothalamic amenorrhea (FHA), caused by deficient output of gonadotropin-releasing hormone, may present with an inadequate luteal phase, anovulation with menstrual irregularity, or actual amenorrhea. The underlying mechanisms remain uncertain but there appear to be numerous neuroendocrine abnormalities in FHA, including hyperfunction of the hypothalamic-pituitary-adrenal axis, leading to increased secretion of corticoids and opioids as well as heightened dopaminergic tone and increased nocturnal melatonin secretion. The investigators followed 93 women with FHA over 7 to 9 years (average, 8.1 years), at which time 65 patients (70.7%) had recovered. The FHA diagnosis was based on amenorrhea for 6 months or longer, the hormonal findings, and a radiologically normal sella turcica. In no case did 5 days of oral medroxyprogesterone acetate induce menstrual bleeding. Patients underwent transabdominal or transvaginal pelvic ultrasonography, which excluded polycystic ovary syndrome. Depending on their wish to conceive, patients received either estrogen replacement therapy or oral contraception. Compared with control women, those with FHA had significantly lower levels of gonadotropins, prolactin, estradiol, and thyrotropin, and higher levels of cortisol. The major historical factors associated with FHA were psychological stress and competitive athletic activity, but these factors did not correlate with the likelihood of recovery. Recovered women had higher baseline body mass indices, but age and the duration of amenorrhea were not factors. The only significant hormonal differences were higher androstenedione and lower cortisol levels in recovered women. Recovery did correlate with an increased body mass index at follow-up, but not with the ultrasonographic appearance of the ovaries (whether multifollicular or not). Recovery rates were 74% in women given hormone replacement therapy, 42% in those receiving oral contraception, and 80% in untreated patients. The respective mean recovery times were 25, 34, and 16.5 months. Twelve recovering women had 14 spontaneous pregnancies during follow-up, all but one of which occurred in the setting of hormone replacement therapy. On multivariate logistic regression analysis, baseline body mass index as well as baseline plasma cortisol and A levels predicted recovery from FHA (Fig. 1). FHA is frequently a result of psychological or physical stress, which, along with an inappropriate diet, may alter energy balance and reduce the body mass index. The disorder is reversible, as was the case in more than two thirds of the present patients. The role of various treatments remains unclear, but a stable or increasing body mass index does seem to be a factor in recovery.