Nutritional and endocrine-metabolic aberrations in women with functional hypothalamic amenorrhea

The development of functional hypothalamic amenorrhea (FHA) in weight-stable, nonathletic women has long been thought to be psychogenic in origin. This study was designed to gain insight into the possibility that nutritional deficits and compensatory endocrine-metabolic adaptations contribute to the development and maintenance of FHA of the psychogenic type. Nutritional intake, insulin sensitivity, and 24-h dynamics of insulin/glucose, cortisol, leptin, somatotropic, and LH axes were simultaneously assessed in eight women with FHA not associated with exercise or weight loss and in eight age-and body mass index-matched regular cycling controls (NC). The percent fat body mass was lower and lean body mass was higher in FHA than in NC (P < 0.05). The FHA subjects scored higher (P < 0.05) on two Eating Disorder Inventory subscales and had a higher (P ( 0.05) Beck depression rating than NC, although all were in the subclinical range. Although daily caloric intake did not differ, FHA consumed 50% less (P < 0.001) fat, twice (P < 0.05) as much fiber, and more carbohydrate (P < 0.05) compared to NC. During the feeding phase of the day, FHA exhibited lower glucose (P < 0.05) and insulin (P < 0.01) levels than NC, and the degree of hypoinsulinemia was directly related to relative dietary fat (r = 0.73). Although 24-h mean GH levels did not differ, the pattern of GH release in FHA was distinctly altered from that in NC. GH pulse amplitude was blunted, pulse frequency was accelerated 40% (P < 0.01), and interpulse GH concentrations were elevated 2-fold (P < 0.01) throughout the day for FHA compared to NC. This distorted pattern of GH pulses was associated with a 40% decrease (P < 0.01) in GH-binding protein levels. Levels of the insulin-dependent insulinlike growth factor (IGF)-binding protein-1 (IGFBP-1) were elevated (P < 0.001) during the feeding portion of the day in FHA and were inversely related to insulin (r = -0.50) and directly related to cortisol (r = 0.64) levels for FHA and NC groups together. Although levels of IGF-I and IGFBP-3 did not differ, the elevation of IGFBP-1 levels in FHA resulted in a reduced (P < 0.01) ratio of IGF-I/IGFBP-1, which may decrease the bioactivity and hypoglycemic effect of IGF-I. Twenty-four-hour mean leptin levels and the diurnal excursion of leptin in FHA did not differ from those in NC. LH pulse frequency was slowed 50% (P < 0.001) in FHA, with unaltered pulse amplitude, resulting in 45% lower (P < 0.01) 24-h mean LH levels for FHA compared to NC. LH pulse frequency for the two groups was related positively to insulin (r = 0.80) levels and the ratio of IGF-I/IGFBP-1 (r = 0.70) and negatively with cortisol (r = -0.61) and IGFBP-1 (r = -0.72) concentrations. In summary, we found evidence of subclinical eating disorders in weight-stable, nonathletic women with FHA accompanied by a severe restriction of dietary fat intake. Unbalanced nutrient intake in psychogenic FHA was associated with multiple endocrine-metabolic alterations. Among these, reduced levels of plasma glucose and serum GHBP, a decrease in the ratio of IGF-I/IGFBP-1, accelerated GH pulse frequency, and elevated interpulse GH levels are indicative of a hypometabolic state. In addition, the magnitude of glucoregulatory responses (increased cortisol secretion and decreased insulin/IGF-I action) were directly related to the degree of suppression of GnRH/LH pulse frequency. These results are remarkably similar to those seen in highly trained athletes with FHA(1). Thus, nutritional deficits may represent a common contributing factor to the development and maintenance of multiple neuroendocrine-metabolic aberrations underlying both psychogenic and exercise-related FHA.