The final common aetiological pathway to functional hypothalamic anovulation is reduced GnRH drive that manifests in the circulation as a reduction in LH pulse frequency. Several studies have documented that the LH pulse frequency of amenorrhoeic women with FHA is about half that of women in the follicular phase. (Figure 1). Pulsatile administration of GnRH resulted in ovulation and conception in women with FHA. For folliculogenesis to result in ovulation, however, GnRH drive must remain above and below critical thresholds for approximately 12– 14 days. Presumably, intermittent or lesser reductions in GnRH pulse frequency lead to poor follicular development with subsequent luteal insufficiency.
Fig1. Luteinizing hormone pulse patterns in two women with FHA are shown in the top two panels and in two eumenorrheic, ovulatory women who were studied in the early follicular (EF) phase in the bottom two panels. Luteinizing hormone pulse patterns are the best in vivo surrogate for assessing gonadotrophin- releasing hormone activity in humans. To assess luteinizing hormone pulsatility in conditions in which slower frequencies are anticipated, one generally obtains blood samples from a forearm vein at 15- min intervals for 24 h. Computer- driven algorithms are then used to resolve the string of luteinizing hormone determinations into pulses so that frequency and amplitude can be computed. The eumenorrheic women regularly display regular luteinizing hormone pulses of even amplitude whereas the women with FHA display irregular luteinizing hormone pulses with an overall slower frequency and variable amplitude. In general, women with FHA have luteinizing hormone pulse frequency that is roughly half of that observed in eumenorrheic women. The expected nocturnal slowing is seen in the eumenorrheic women whereas some augmentation of luteinizing hormone is seen in women with FHA, a pattern that is somewhat similar to that seen in puberty.
FHA is more than an isolated interruption of the hypothalamic– pituitary– ovarian axis. A constellation of neuroendocrine secretory aberrations has been demonstrated. While cortisol levels are higher in FHA than in eumenorrheic women, the diurnal pattern remains intact. In both eumenorrheic and amenorrhoeic women, cortisol secretion peaks nocturnally. Among women with anovulation, increased cortisol secretion appears to be specific to FHA. Ovarian function is not necessary for GnRH drive to be suppressed by stress. FSH levels in postmenopausal women admitted to the intensive care unit continue to decline after the peak in cortisol levels demonstrating a lag between recovery of the hypothalamic– pituitary– adrenal and resumption of GnRH drive. Elevated cortisol levels have been observed in women with depression and anorexia nervosa but not bulimia. In women athletes, only those with amenorrhoea displayed elevated cortisol levels and reduced luteinizing hormone pulse frequency. Interestingly, women with FHA did not meet criteria for depression. The development of anovulation sufficient to cause amenorrhoea may require greater or more sustained HPA axis activation than generally seen in depression and bulimia. Since energy deficits enhance HPA activation, the extent of energy insufficiency may explain differential rates of FHA in women with depression and bulimia.
Cortisol per se is not likely to be the cause of reduced GnRH pulse frequency in FHA. Acute, exogenous, high- dose administration of cortisol (over 300% of physiological) did not reduce GnRH drive. Hydrocortisone infusions given to mimic moderate (136% increase) and severe (197% increase) stress had no acute impact on gonadotropin secretion. In FHA, the increase in cortisol is in the range of 20– 30%. In underweight anorectics, the increase was in the range of 150%, but after short- term weight recovery, the increase was about 70% of the control group. Typically, weight restoration alone is insufficient to restore menses, despite reduced cortisol levels. Since the increased cortisol secretion observed in women with FHA and anorexia nervosa is lower than the levels needed to independently suppress gonadotropin secretion, other factors must play a causal role. Of note, the decline in LH caused by insulin- induced hypoglycaemia in rhesus monkeys was comparable in monkeys pretreated with metyrapone to block cortisol synthesis and those who were not treated. Similarly, an intravenous infusion of corticotropin- releasing hormone caused a decrease in gonadotropin in monkeys with and without adrenalectomy. While these observations do not obviate a role for cortisol as an indicator of stress, they do suggest that neural concomitants of the stress signal or concomitant metabolic signals, rather than cortisol per se, cause stress- induced suppression of GnRH pulsatility. While an increase in cerebrospinal corticotropin- releasing hormone has been detected in women with anorexia nervosa, women with FHA do not have a detectable increase in cerebrospinal corticotropin- releasing hormone despite an increase in cerebrospinal fluid (CSF) cortisol concentrations. The lack of suppression of CSF CRH in the presence of increased CSF cortisol indicated resistance to feedback inhibition of hypothalamic CRH release by cortisol and suggests that mechanisms other than CRH and cortisol mediate the stress- induced suppression of GnRH.
Cortisol is a hormone with profound metabolic actions and increased secretion produces metabolic mobilization, including inhibition of gluconeogenesis. Cortisol also dampens thyroid- stimulating hormone (TSH) release and causes ‘sick euthyroid syndrome’, presumably as a means of conserving or diverting energy expenditure. As noted earlier, studies have suggested that increased HPA drive predisposes to further endocrine reactivity in response to the sub sequent metabolic challenge of exercise. The converse may also hold. Prior metabolic alterations may predispose to neuroendocrine reactivity in the face of psychogenic challenge. We found that women with FHA showed reduced levels of thyronine and thyroxine despite similar TSH patterns. During recovery from FHA, TSH increased dramatically following a reduction in cortisol levels were reduced. These data indicate that the recovery begins with a decrease in CRH followed by recovery of GnRH and TRH drive.
In a monkey model, metabolic challenge sensitized the hypothalamic– pituitary– ovarian (HPO) axis to the effects of subsequent social stress. Monkeys with the greatest sensitivity to stress had differential serotonergic tone and higher baseline heart rates. Administration of a CRH antagonist reversed the inhibition of LH pulsatility induced by the administration of a potent orexigenic signal, ghrelin. In women, caloric deprivation sufficient to slow LH pulsatility in eumenorrhoeic women dropped circulating glucose and increased circulating cortisol levels, indicating that metabolic stress activates the HPA axis. Women with FHA responded to a graded exercise challenge with a small drop in circulating glucose while eumenorrheic women showed a rise in glucose during exercise. Simultaneously, exercise provoked a larger rise in cortisol in amenorrhoeic as compared to eumenorrhoeic women. These data indicate that the impact of exercise is state dependent. While exercise is often performed to reduce psychological stress, at an endocrine level, it may amplify the stress cascade and heighten stress reactivity. Thus, the management of stress needs to be contextual. Psychogenic solutions need to be found for psychosocial dilemmas and exercise is not a substitute for good mental hygiene.
