This is the first report, to the best of our knowledge, on the performance of urinary gonadotropins in a cohort of children that have been on follow-up for signs of puberty progression. Longitudinal cohort studies in peripubertal children are scarce, and there is a significant knowledge gap in this area. The results of this study confirm that FMV U-LH determinations, which reflect the integrated LH pulse secretion into the bloodstream during sleep, are a sensitive method to detect HPG axis activation already before the manifestation of clinical signs of puberty. This has been demonstrated in previous cross-sectional and longitudinal studies on the clinical aspects of urinary gonadotropin measurements [3, 4, 12,13,14]. The widespread use of non-invasive and sensitive methods for the assessment of gonadotropin secretion is becoming increasingly important because of the continuing trend toward an earlier onset of puberty in both sexes, particularly in girls, mainly due to environmental factors; indeed, the proportion of girls with non-progressive isolated thelarche has increased during recent decades among early maturing girls [15, 16]. Evaluation of gonadotropin secretion activity by quantification of FMV total U-LH concentrations by sensitive assays can be beneficial for differentiation between isolated thelarche and progressive central precocious puberty.

In addition, abnormal timing of pubertal development has been reported to be associated with adverse health and psychosocial outcomes. For example, early age at menarche is associated with an increased risk of obesity, type 2 diabetes [17], and cardiovascular disease [18] later in life. Other reported associations with early menarche include increased risk of breast cancer [19] and all-cause mortality [20]. These are not only important for the individual, but also have potentially large public health implications, especially given the secular trend toward earlier onset of puberty [21, 22].

In this study, it was possible to differentiate between the EPP and LPP girls by both FMV U-LH and U-FSH determinations before the LPP girls showed physical signs of puberty. FMV total U-LH concentrations at or above 0.8 IU/L were shown to be a good predictor of clinical signs of puberty across prepubertal and pubertal girls according to the results of this study. Our study results also showed that FMV total U-LH concentrations above 0.3 IU/L were an excellent predictor of imminent HPG axis activation because there is naturally a time lag between HPG axis activation and the manifestation of the physical signs of puberty. Another interesting finding was that the FMV total U-LH concentrations were not only above the aforementioned cut-off value of 0.3 IU/L in all the five LPP girls differentiating them from all the five EPP girls, but also at or above 0.6 IU/L in four of the five LPP girls, two of whom were even younger than some of the EPP girls. These findings imply that the cut-off values for FMV total U-LH concentrations signifying forthcoming HPG axis activation and the manifestation of clinical signs of puberty are approximately 0.3 and 0.8 IU/L; accordingly, the FMV total U-LH concentration to indicate HPG axis activation, i.e., the onset of central puberty, should be in the 0.3–0.8 IU/L range, presumably around 0.6 IU/L according to the results of this limited semi-longitudinal study. Thus, ROC curve analyses from this exploratory study need to be replicated in larger cohorts.

A novel aspect of this study design is also the use of the overnight fold change (FMV/LNV ratio) in U-LH-ir and U-FSH-ir levels as a conceptual analogy to a GnRH stimulation test setting. LNV and FMV urine gonadotropin concentrations represented the basal level and the AUC of the GnRH test, respectively. The overnight fold change in total urinary LH or FSH concentrations, which reflects the change in the respective gonadotropin pulse activity during the whole course of overnight sleep, appears to be one of the parameters that add value to the assessment of pubertal development. In this particular study with a rather limited number of EPP and LPP subjects, detection of a transient increase in the overnight fold change in total U-LH-ir levels did not appear to be superior to FMV U-LH determinations alone in predicting imminent puberty. However, based on our results, since elevated overnight change in U-LH appears to be a parameter that is uniquely associated with LPP, but not with EPP, EP, or LP girls, it can be used as an adjunct parameter to evaluate the central activation of puberty in cases that remain borderline according to FMV total U-LH results. While our results suggest that the value of overnight change in U-LH is of limited value in the clinical setting, future studies should clarify whether the change can add sensitivity and specificity to FMV U-LH in different clinical scenarios. This phenomenon reflects the well-known physiological pattern of increased gonadotropin secretion exclusively during the night in early puberty, followed by loss of diurnal variation in the later stages of puberty. In addition, the overnight fold change in total U-LH-ir levels or the total U-LH/U-FSH ratio appears to be predictive of imminent menarche, as the increase in total U-LH concentration during the course of overnight sleep until morning is significantly higher in EP subjects when compared to that in LP subjects (Fig. 2C). Future studies may determine whether the transition from the premenarcheal stage to menarche is associated with a specific pattern of urinary gonadotropin secretion.

Furthermore, little is known about the critical ratio between LH and FSH during the pubertal transition in both sexes despite a large number of previous studies since the 1970s on the relevance of the LH:FSH ratio in the physiology as well as in the diagnostics of pubertal disorders [23,24,25]. An earlier study of ours revealed the sex difference in U-LH/U-FSH ratio in peripubertal boys and girls as well as at different pubertal stages, confirming the relative FSH dominance in girls compared to boys [4]. The results of the current study suggest that imminent menarche may be related to an overnight increase rate in total U-LH/U-FSH ratio or in FMV total U-LH-ir levels remaining at or below 21 and 24%, respectively, which should be confirmed by further longitudinal research to reveal the estimated time to menarche at the same time.

The main strength of the current study is its semi-longitudinal design which enabled clinical signs of puberty to be detected in some initially prepubertal children over the course of a 1-year follow-up. The high sensitivity and specificity of the test method were another important strength of the current study. On the other hand, the small sample size and the lack of serum gonadotropin measurements were major limitations of this study. In addition, mass spectrometric determination of estradiol and testosterone would have been a better alternative to RIA, the sensitivity of which may still be too low for the determination of these steroid hormones in the pediatric age group.

It can be concluded that FMV total U-LH and U-FSH above 0.3 IU/L and 2.5 IU/L, respectively, may serve as cut-off values to predict the manifestation of the clinical signs of puberty within 1 year. Moreover, FMV total U-LH above 0.8 IU/L in our cohort was exclusively found in those exhibiting clinical signs of puberty and may serve as a threshold that differentiates girls with central precocious puberty from those still prepubertal. The suggested range and cut-off for FMV total U-LH concentrations (0.3–0.8 IU/L and 0.6 IU/L, respectively) signifying the critical FMV total U-LH-ir levels associated with the release of the central brake on the GnRH pulse generator should be verified by further larger studies involving early morning serum LH and FSH determinations and comparisons against the gold standard (such as early morning serum LH concentration cut-off at 0.3 IU/L) in children and adolescents of both sexes. Thereafter, FMV total U-LH and U-FSH determinations can be used as the first-line tool in the evaluation of pubertal development as well as for screening and diagnostics of pubertal disorders; hence, the need for invasive GnRH stimulation tests or early morning serum LH determinations may be decreased substantially. In addition, an overnight increase in U-LH-ir levels or in the U-LH/U-FSH ratio remaining at or below 1.2-fold in an early pubertal girl may be a sign of imminent menarche, the presumed timing of which can be unraveled by longitudinal studies. Our hypothesis that integrated gonadotropin activity during overnight sleep is reflected as an increase in integrated total U-LH or U-FSH immunoreactivity during overnight sleep also requires further investigation in a larger cohort with power analysis for conclusive ROC analyses.