Obesity and the Emergence of Sleep-Wake Gonadotropin Secretion in Girls during Early Pubertal Development

Abstract

It has been clearly established that obesity is associated with reduced gonadotropin secretion in men and women. In normal women, early studies involving relatively few subjects had shown that serum LH levels were significantly reduced in obese individuals compared with levels of normal-weight controls (1, 2). Subsequent larger studies that examined the relationship of LH secretory dynamics to body weight included women with polycystic ovary syndrome (PCOS) as well as normal women (3, 4). In particular, LH secretion in both groups of obese women was characterized by reduced basal levels and decreased pulse amplitude, whereas LH frequency appeared unaffected. Recently, it was demonstrated that morbidly obese women with normal menstrual function exhibited a similar pattern of LH suppression compared with that of normal women (5). In adolescent girls, there are essentially no data concerning the influence of obesity on the rate and amplitude of LH secretion. However, during puberty the progressive maturation of gonadotropin secretion in normal girls through various stages of development has been well documented (6). Given the observed impact of obesity on LH secretion in adults, it seems logical to investigate this relationship at a time when activation of hypothalamic-pituitary interaction is emerging. This notion is underscored by the association of obesity with an earlier onset of puberty and precocity, increased risk of chronic anovulation with menstrual irregularities, and greater likelihood of PCOS. The last of these conditions commonly manifests clinical symptomatology during or soon after puberty. In this issue of JCEM, McCartney et al. (7) have extended their efforts to define the role of obesity on gonadotropin secretion during pubertal development. In previous studies using a cross-sectional design, they showed that during early puberty (Tanner 1–3), obese girls had significantly increased serum total and free testosterone levels compared with nonobese controls accompanied by corresponding decreases of circulating LH concentrations (8, 9). In addition, obese girls exhibited hyperinsulinemia and increased homeostasis model of assessment values compared with that of normal-weight controls, which may have accounted, in part, for hyperandrogenemia in the presence of lowered serum LH. In early pubertal girls, diurnal variation was observed for both testosterone and progesterone with levels lower at night and rising in the early morning hours. The incremental change in both hormones was significant in nonobese girls but failed to achieve statistical significance in obese girls likely because of too few numbers. Nevertheless, the heightened level of progesterone in obese girls may have contributed to reduced LH levels in early puberty. This notion relates to the insensitivity of the GnRH pulse generator to progesterone negative feedback in women with PCOS. In normal women, progesterone administration inhibits LH secretion, e.g. pulse frequency. Prior studies by this group have shown that androgen may interfere with this feedback mechanism as suggested by reduction of rapid LH release in response to progesterone administration in women with PCOS treated with an antiandrogen, which implies that hyperandrogenemia may contribute to increased LH secretion in this disorder. This proposed mechanism is suggested in young adolescents because some girls with PCOS treated with progesterone failed to exhibit a reduced rate of LH release (10). Because their earlier studies had been restricted to basal serum LH measurements, McCartney et al. (7) proceeded to conduct an extensive frequent sampling study of obese [body mass index (BMI) 95th percentile] and nonobese (BMI 95th percentile) girls at Tanner stages 1–5 and found differences in LH secretion within and between groups during early puberty (11). The experimental design employed a sampling interval of 1900–2300 h to denote daytime secretion and 2300–0300 h and 0300– 0700 h to reflect nighttime release. As expected, LH pulse frequency, amplitude, and mean concentrations in nonobese Tanner 1–2 girls were, in general, greater with sleep than values observed in awake subjects. In Tanner 1–2 obese girls, there was little change in LH secretion over the same intervals of study.