The classic interventional studies of Bullen et al demonstrated that in normal women with no history of cycle irregularity a significant increase in exercise could induce menstrual cycle dysfunction ranging from short or inadequate luteal phases to anovulation and amenorrhea, which is termed hypothalamic amenorrhea (HA). This study also showed that the combination of weight loss and exercise was more deleterious than exercise alone. While the conclusions of this study focused on the role of exercise and nutrition, it is likely that some degree of stress also contributed to the findings as the study was conducted at a camp and changes in living situations per se are associated with the onset cycle disturbances, with the classic example being young women in their freshman year in college. The association of increased exercise and decreased nutritional intake was confirmed in multiple other cross-sectional studies. Loucks and colleagues quantified the interaction between energy intake and energy expenditure that resulted in disrupted GnRH secretion. The authors developed the concept of energy availability which they defined as dietary energy intake minus exercise energy expenditure. In well-controlled studies in a laboratory setting they demonstrated a threshold of energy availability at which pulsatile LH secretion, used as a marker of GnRH pulse frequency, was altered in healthy women who had undergone 5 days of decreased energy availability during their early follicular phase (EFP). In these studies, LH pulse decreased in conjunction with an energy availability of 20 and 10 kcal/kg lean body mass per day (LBM*day) compared with neutral energy availability of 45 kcal/kg LBM*day or 30 kcal/kg LBM*day 67. Further studies showed exposure to energy intakes of 0%, -8%, -22% and -42% over 3 months was associated with menstrual cycle disruption in the form of inadequate luteal phases, intermittent ovulation and amenorrhea. Taken together, these studies suggest that the change in LH pulse frequency with short-term energy deprivation predicts menstrual cycle disturbance when energy deprivation is more chronic although the minimum degree of chronic energy deficit required for cycle disturbance (-22%) may be less than would have been predicted by the short-term energy deficit studies and thus. the results of short-term studies would provide a conservative marker of potential risk. Mechanisms Linking Decreased Energy Availability and Stress to Inhibition of GnRH Decreased Energy Availability - The primary function of the hypothalamus is the regulation of homeostasis. In the setting of decreased energy availability, due to a nutritional deficit or an increase in energy expenditure, conservation of energy is achieved by hypothalamic coordination of various neuroendocrine axes to redirect the available energy towards the crucial systems for survival. Moreover, the magnitude of energy deficit is more predictive of suppression of reproductive function than weight loss per se. It is well documented in women that reproductive function is one of the first systems to be suppressed in association with energy deprivation via disruption of GnRH secretion. It is hypothesized that short-term survival is prioritized over reproduction due to the high energy demands of pregnancy and breastfeeding. With energy deprivation, the drive for energy intake is increased while energy expenditure is decreased in an attempt to restore balance. Leptin, ghrelin and adiponectin are highly specific signals that not only signal energy availability at the level of the hypothalamus as part of an adaptive response to restore energy balance, but are also linked to control of GnRH. Leptin provides a crucial link between energy balance and the hypothalamic control of reproduction; however, individual susceptibility of the reproductive axis to absolute leptin concentrations or changes in leptin induced by changes in energy availability has not been determined. Ghrelin, a peptide hormone, is primarily secreted by the stomach, is inversely related to BMI, acts at the hypothalamus to stimulate both appetite and growth hormone secretion, and appears to play a significant role in both short- and long-term regulation of energy homeostasis. Most importantly for this study, ghrelin administration to healthy women results a significant decrease in mean LH and LH pulse frequency, but as for leptin, it is unknown whether absolute ghrelin concentrations or changes in ghrelin induced by changes in energy availability can predict individual susceptibility of the reproductive axis decreased energy availability. Like leptin, adiponectin is adipocytokine secreted by adipose tissue. Adiponectin increases satiety and reduces energy expenditure. Adiponectin receptors are present on GnRH neurons and adiponectin inhibits GnRH secretion and thereby inhibits GnRH-stimulated LH secretion. Adiponectin is not acutely affected by meals and may thus be a more stable biomarker to associate with susceptibility to reduced energy intake. There is now considerable evidence linking activation of neuroendocrine mechanisms by stress to inhibition of reproductive signaling, only some of which is mediated by the inhibitory effects of cortisol on gonadotropin secretion at the pituitary level. Activation of the stress axis as evidenced by increased cortisol is associated with acute energy deprivation in a dose responsive fashion; however this is not the case for more chronic energy deficits. Cortisol levels in serum, urine and cerebral spinal fluid are greater in women with HA compared to normally cycling controls. In a long term follow-up study of women with HA, there was an inverse relationship between circulating cortisol and LH levels and those women who had not recovered reproductive function had significantly higher fasting plasma cortisol levels than those women who had recovered. Cortisol is also increased with exercise and is highest in exercising amenorrheic women. Cortisol and leptin are inversely related in HA, however leptin administration caused no change in cortisol levels in healthy men and significantly improves reproductive function in women, with no change in cortisol levels. There is, thus, considerable information implicating stress pathways in the etiology of HA. In addition, stress pathways may also be involved in inhibitory effects of nutritional deficits on central reproductive function. Other hormonal signals that appear to be involved in linking metabolism to reproduction include the thyroid hormone axis and insulin.
