Main

Leptin is the protein product of the obese (ob) gene. It is secreted as a hormone mainly from adipocytes and interacts with its receptor in the hypothalamus to signal the state of energy stores to the brain. Serum levels of leptin are correlated with the degree of obesity and are regulated by food intake and fasting status(1–4).

Homozygous defects in ob/ob mice or db/db mice, which are unable to produce leptin or to express the leptin receptor, result in adiposity, infertility, short stature, and abnormally high circulating levels of glucose, insulin, and cortisol. Administration of leptin to ob/ob mice corrects these features(5,6). Injection of low doses of leptin leads to an accelerated sexual maturation and earlier onset of reproductive function in female mice(7), although an adequate nutritional status is a prerequisite for pubertal growth and sexual development. Calorie restriction in animals leads to a suppression of reproductive function(7–9). Leptin administration to malnourished prepubertal rats only partially normalizes the delay in pubertal maturation(8,9). Thus, leptin plays an important role in the control of reproductive function in the rodent model and seems to be a signal to allow for sexual maturation to proceed under adequate metabolic conditions(10).

Studies on leptin-deficient humans show that leptin is essential for the initiation of puberty and establishment of secondary sexual characteristics. Patients who are homozygous for a mutation in the leptin receptor gene show no signs of pubertal development and only limited secretion of growth and thyrotropic hormones(11,12). A homozygous missense mutation in the leptin gene results in morbid obesity(13) and impaired sexual function(14).

In humans, longitudinal and cross-sectional studies show increasing leptin levels before the onset of puberty(15–23). In boys, longitudinal and cross-sectional studies show that leptin levels rise by approximately 50% just before the onset of puberty compared with baseline prepubertal levels, and decrease to baseline levels after the initiation of puberty(16,22,23). In girls, leptin levels increase after the onset of puberty, remain constant in midpuberty, and rise to a peak in late puberty(15). In all pubertal stages, leptin concentrations correlate positively with body mass index (BMI) and fat mass(21). It has been demonstrated that leptin levels and body fat mass are inversely related to age at menarche in girls(24). Reference ranges for leptin were constructed in relation to BMI(21). All these investigations emphasize that leptin has a key signaling function in regulating pubertal development, not only in laboratory animals but also in normal children.

The aim of our study was to investigate the relationship between leptin levels and pathologic early onset of pubertal development in girls with central precocious puberty (CPP). In a longitudinal trial, we examined the leptin concentrations in girls with CPP before start of therapy, when the hypothalamic-pituitary-gonadal axis was fully activated, and after 1 and 2 y of treatment with the depot gonadotropin releasing hormone (GnRH) agonist, leuprorelin (leuprolide) acetate, when puberty was completely suppressed.

METHODS

Patients. Fourteen girls (mean age, 6.0 ± 0.6 y; range, 2.1-8.7 y) with progressive idiopathic CPP were studied. This study was approved by the Ethical Committee of the Medical Faculty of the University of Kiel. Main criteria for the diagnosis of CPP were 1) secondary pubertal signs before the age of 8 y, 2) growth rate above the 75th percentile, 3) an acceleration of bone age (BA) more than 1 y above chronological age (CA) and ΔBA/ΔCA > 1.2, and 4) a pubertal response to exogenous GnRH with LH peak > 11 IU/L(25) and a ratio of stimulated LH/stimulated FSH > 1.0(26,27).

GnRH agonist treatment was initiated after an observation period of 6 mo. The absence of an organic course of CPP was demonstrated by magnetic resonance. Patients were treated for progressive CPP within a German multicenter trial with depot leuprorelin acetate (Enantone, Takeda Pharma, Aachen, Germany). For ethical reasons, an untreated patient control group was not available. Injections were given every 30 ± 2 d s.c. in a dose of 0.5 or 1 ampule (1.88 mg for body weight <20 kg and 3.75 mg for body weight >20 kg). Adequate suppression of pituitary-gonadal function was defined as an estradiol level <50 pM and a stimulated plasma LH after GnRH stimulation <5 IU/L.

Measurements. Standing height was determined by stadiometer. BA was estimated centrally by one experienced investigator by the method of Greulich and Pyle(28). Prospective adult height was calculated according to Bayley and Pinneau(29). German longitudinal normative data were used as standards for height and growth velocity(30,31). The normative data of Rolland-Cachera et al.(32) were used for BMI.

Expected leptin levels and leptin SD score, both adjusted for BMI, were calculated according to the following equations: (Equation 1) as reported by Blum et al.(21).

Plasma gonadotropins and estradiol and serum IGF-1 were determined by RIA or enzyme immunoassay at the Endocrine Laboratory of the University Department of Pediatrics in Kiel. Sensitivity of LH and FSH assay was 0.3 IU/L. A value of 0.2 IU/L was assigned to LH and FSH levels below the detection limit. The detection limit of estradiol was 5 pg/mL. A value of 4 pg/mL was assigned to samples below the detection limit.

Serum leptin was measured by RIA. Assay components and assay procedure were described previously(21). Assay sensitivity was 0.03 ng/mL. Intraassay and interassay variability was 0.8 and 8.5%, respectively (n = 10)(21).

Statistics. Statistical comparison of data at start of treatment and 1 and 2 y after start of therapy was performed by ANOVA. In those cases in which the data showed a significant deviation from a normal distribution, Friedman's ranked analysis of variance was applied. For evaluation of statistical differences between two different times, the Student-Newman-Keuls test was used as a post hoc test (SigmaStat 1.0, Jandel Scientific, Erkrath, Germany). Multiple stepwise regression analysis was performed to test which parameters contributed to the variation of leptin levels. Data are presented as mean and SEM, and a p < 0.05 was considered significant.

RESULTS

The clinical characteristics of the 14 girls with CPP are summarized in Table 1. Mean CA and BA at start of treatment were 6.0 ± 0.6 y and 9.5 ± 0.7 y, respectively. During the pretreatment observation period, ΔBA/ΔCA was 2.52 ± 0.4. Patients showed a pubertal growth rate of 13.0 ± 1.5 cm/y, equivalent to a velocity of 7.8 ± 1.7 height SD score. The stimulated LH/FSH ratio at the time of diagnosis was 1.8 ± 0.2, indicating central activation of the hypothalamic-pituitary-gonadal axis.

Table 1 Clinical, auxological, and hormonal characteristics of girls with CPP before treatment and during treatment with GnRH agonist
Full size table

During GnRH agonist therapy, the BMI SD score for CA increased significantly from 1.9 ± 0.5 to 2.7 ± 0.5 after 1 y and to 2.8 ± 0.4 after 2 y of treatment (p < 0.05 versus pretreatment). Sufficient suppression of the hypothalamic-pituitary-gonadal axis was achieved in all patients. GnRH-stimulated LH levels >5 IU/L (n = 83 GnRH tests with n = 15 not completely suppressed) and estradiol levels >50 pM (n = 109 estradiol levels with n = 7 >50 pM) were only seen during the first few months of treatment.

Serum leptin levels and expected leptin values of healthy girls of the same BMI referenced to our patients are demonstrated in Table 2. There was a significant increase in serum leptin levels after 1 y (8.1 ± 1.7 ng/mL; p < 0.05) and 2 y of treatment (11.2 ± 3.0 ng/mL; p < 0.05) compared with pretreatment levels (7.5 ± 2.3 ng/mL).

Table 2 Leptin values of girls with CPP before treatment and during treatment with GnRH agonist
Full size table

Serum leptin levels were in the normal range for BMI and pubertal stage in the majority of the patients on the three occasions they were measured. A leptin level more than +2 SD was found in two of 14 patients at pretreatment, in no patients after 1 y of treatment, and in one of 15 patients after 2 y of treatment (Fig. 1). No patient showed a leptin levels less than -2 SD. Mean leptin SD score adjusted for BMI showed no significant changes during the period of observation.

Figure 1
figure 1

Individual serum leptin levels in girls with CPP at pretreatment and after 1 and 2 y of treatment with depot leuprorelin acetate. Measured leptin concentration (upper panel); leptin SD score (SDS) adjusted for BMI (lower panel).

Full size image

In a stepwise regression analysis model with BMI, BA, CA, basal and stimulated LH, estradiol, dehydroepiandrosterone, androstenedione, and clinical pubertal signs, BMI was the only parameter that did show a significant correlation with leptin (p = 0.006, Table 3, Fig. 2).

Table 3 Stepwise regression analysis of plasma leptin and clinical and hormonal parameters in girls with CPP
Full size table
Figure 2
figure 2

Serum leptin levels in girls with CPP according to BMI before initiation of therapy (A) and after 1 y of treatment with depot leuprorelin acetate (B). The closed line represents the linear regression between serum leptin levels and BMI. The 95% confidence limits are shown by the dotted lines.

Full size image

DISCUSSION

We investigated serum leptin concentrations in girls with CPP before the start of treatment and after 1 and 2 y of therapy with depot leuprorelin acetate. This longitudinal study enabled us to elucidate the role of leptin in pathologically early pubertal development and during its suppression.

Several investigations in animals confirm that leptin signals adequate metabolic conditions for sexual maturation to proceed(7–10). Furthermore, investigations in normal children showed that serum leptin levels increase just before the obvious onset of pubertal signs(15,16,21,22).

We found that leptin concentrations had a tendency to be slightly but not significantly higher both at pretreatment and during treatment than the expected leptin concentrations of healthy girls (Table 2; Fig. 1). On the average, leptin serum concentrations were 1.4-fold higher in girls with CPP than in healthy girls of the same BMI(21). Similarly, the mean leptin SD score adjusted for BMI was invariably greater than zero both at diagnosis and during treatment, although the majority of CPP patients had leptin levels in the upper normal range (Fig. 1). To our knowledge, other investigators have not calculated expected leptin levels adjusted for BMI(33).

We found a significant increase in measured and expected leptin concentrations throughout the period of treatment with the depot GnRH agonist while the hypothalamic-pituitary-gonadal axis was permanently suppressed (Table 2). Therefore, suppressive treatment did not affect the tendency for higher than the expected leptin levels. The mean leptin SD score adjusted for BMI remained unchanged during treatment; in particular, no decrease was seen. This absence of decrease in leptin concentrations is in contrast to findings of decreased leptin levels in ovariectomized rats(34) and postmenopausal women(35,36). The assumption that lack of estrogens is responsible for the decrease in leptin is discussed controversially. In fact, other investigators did not find a relationship between estrogen and decreased leptin(37–39). Our findings of a slight increase of leptin levels after initiation of puberty are similar to the observation in healthy girls, who show a continuous rise of leptin concentration after pubertal onset(15,20). In contrast, in healthy boys, leptin concentrations decrease to prepubertal levels after initiation of puberty(16,22,23). Particularly in boys with CPP, a suppression of precocious testosterone production leads to an increase in leptin concentration(33).

In girls with CPP, leptin concentration appeared to be unaffected by the functional status of the hypothalamic-pituitary-gonadal axis and vice versa. Serum leptin concentrations were related to BMI only. Thus, BMI is the major predictive factor for leptin both in CPP patients and in healthy girls(21). In principal, our results confirm those of Palmert et al.(33), who found that estrogen did not play a major role in the regulation of leptin levels. Our study emphasizes that leptin concentrations should be referenced to the major influencing variables BMI and sex in patients with disorders of pubertal development and in cases of disease-specific changes in BMI.

It is well-known that ballet dancers, marathon runners, and patients with chronic diseases suffer from impaired reproductive function(40), which may be related to low leptin levels(10). Thus, leptin is signaling to the brain as to whether or not the conditions for sexual development are adequate. However, in the case of CPP, in which patients may be either lean or obese, it seems unlikely that the onset of CPP is directly induced by leptin. Furthermore, treatment with GnRH agonists seems to have no influence on leptin concentrations in girls with CPP. Thus, we have no evidence that alterations of leptin are related to premature onset of puberty in children.