Utility of First-Morning-Voided Urinary Total Luteinizing Hormone in Detecting the Onset of Central Puberty

Previous studies have shown that the first-morning-voided (FMV) urinary luteinizing hormone (U-LH) concentration is a sensitive predictor of pulsatile pituitary gonadotropin secretion [1‒3]. This can be achieved by carefully selecting specific assays that can detect total LH immunoreactivity, that is, all forms of gonadotropin immunoreactivity, including those from subunits and core fragments [4‒10], to avoid falsely low levels of LH immunoreactivity in children and adolescents. Indeed, our recent study revealed that FMV total U-LH measurements can predict the imminent onset of central puberty in girls within 1 year [11]. In another recent study based on FMV total U-LH determinations, we reported that the onset of central (biochemical) puberty occurs at approximately the same age in both sexes [12]. In the same study, we found that some clinically prepubertal boys with a testicular volume of 2 mL, considered prepubertal according to current criteria, exhibited pubertal levels of FMV total U-LH concentration, signaling central pubertal activation [12]. However, the efficacy of FMV U-LH determination in identifying the onset of central puberty has not been validated in large cohorts, which limits its application in clinical settings. Likewise, the sex-independent timing of onset of central puberty remains unconfirmed.

The utility of early morning serum LH (S-LH) as a good predictor of the onset of central puberty was first reported in 2013 by Lee et al. [13] and thereafter in several review papers and guidelines [13‒20]. Indeed, our previous study from the 1990s showed that an increase in FMV total U-LH levels precedes the signs of clinical onset of puberty, but no receiver operating characteristic (ROC) analysis was performed about the ability of FMV total U-LH concentrations to predict the diagnostic early basal morning S-LH level of 0.3 IU/L or above. In this study, we revisited our published data from the 1990s and analyzed them from this perspective.

Our recent study showed not only that the onset of central puberty occurs at around the same age in boys and girls, but also that the time from biochemical to clinical puberty is significantly shorter in girls than in boys [12]. Moreover, some girls present with isolated thelarche, that is, non-progressive breast development prior to an increase in gonadotropin concentrations to pubertal levels. This led us to investigate the serum follicle-stimulating hormone (S-FSH) status in children to determine whether it would be possible to differentiate girls with isolated thelarche from those with “true” progressive central puberty. Differential diagnosis of these two conditions has been a persistent challenge [21, 22].

LH concentrations were determined in FMV urine and early morning serum samples from 297 children (145 boys and 152 girls, aged 5–15 years from Tanner stages 1 through 5) evaluated as part of the COPENHAGEN Puberty Study, with no apparent endocrinologic, nephrologic, oncologic, or neurologic disorders, or any medication before or at the time of sampling. The urine samples were collected during the first Copenhagen Puberty Study, a cross-sectional, single-center study conducted in 1991–1992 involving a total of 1,030 healthy Danish children. The primary findings of the study were published in 1994 [23], and data on urinary gonadotropins in a subset of this population were first reported in 1996 [1]. The subjects were healthy school children who did not receive any medication prior to or at the time of sampling. Samples from menstruating girls were taken outside their periovulatory periods. The subjects emptied their bladders just before bedtime. The entire FMV urine was collected and mixed. Subsequently, an aliquot was frozen and used for later analysis. Serum samples were taken on the same morning. Breast stage 2 or a testicular volume above 3 mL was taken as a definite sign of the onset of puberty. For purposes of ROC analysis regarding comparison of FMV total U-LH and early morning S-LH levels; prepubertal and pubertal categories were formed based on early morning S-LH levels; an S-LH concentration at or above 0.3 IU/L was considered indicative of the onset of central puberty. The research protocol was approved by the Ethical Committees of both Children’s Hospital, Helsinki University Hospital (Helsinki, Finland), and Rigshospitalet, Copenhagen University Hospital (Copenhagen, Denmark). The former analyzed urine samples from the Copenhagen Puberty Study, which were supplied by the latter, using an assay protocol that the former had developed [1].

Fresh serum samples were analyzed in duplicate for LH using DELFIA immunofluorometric assays (IFMA) [24, 25], utilizing reagents provided by Wallac (Turku, Finland). Estradiol and testosterone levels were determined using radioimmunoassays from Diagnostic Products Corp., Los Angeles, CA, USA, and Immunodiagnostic System Ltd., Boldon, UK, respectively. The estradiol assay had a detection threshold of 18 pmol/L, while for the testosterone assay, it was 0.23 nmol/L. For the estradiol and testosterone assays, intra-assay and inter-assay variabilities were reported as 7.5% and 8.4%, and 3.8% and 8.6%, respectively. Urine specimens were preserved at −20°C without any preservatives for around 2 years before being analyzed. Urinary LH levels were analyzed in duplicate using DELFIA^® IFMA by Wallac [8]. LH measurements were conducted using two approaches; the LH Delfia, which identifies only complete LH molecules, and the LHspec Delfia, which can detect intact LH molecules, β-subunits, and certain fragments. The detection threshold for the LHspec assay was noted as 0.012 IU/L, with intra-assay and inter-assay variabilities of 5.7% and 6.4%, respectively. The standards used in the LH and LHspec assays were aligned with the WHO 2nd IS 80/552 standard. From previous research findings, adjusting hormone concentrations for urinary density or creatinine to offset urine excretion rate variability did not enhance the correlation with serum levels [26]. Notably, in very diluted urine samples, this correlation was notably disrupted, likely due to excessive adjustment for urinary density or creatinine. Consequently, hormone concentration measurements were not adjusted for urine excretion rate variability.

The detection limits were set at 0.015 IU/L for the LH assay and 0.012 IU/L for the LHspec assay, determined by the concentration that equals the average plus two standard deviations from twelve duplicate measurements of the baseline standard. In the statistical analysis, any concentrations falling below these detection limits were assigned a value of 0.01 IU/L. ROC analysis was used as the statistical tool to calculate different cutoff points for total U-LH levels, predicting the cutoff value determined for early morning S-LH concentrations (≥0.3 IU/L). The area under the ROC curve was used as a summary measure to average the diagnostic accuracy across the range of test values. The detection limit (0.012 IU/L for the total LH assay) was defined as the concentration corresponding to the value of the mean plus 2 SD of 12 duplicates of the zero standard. For statistical evaluation, concentrations below the detection limit were assigned a value of 0.01 IU/L. The association between sex and timing of the onset of central puberty was investigated by comparing the number of children or adolescents classified as prepubertal and pubertal at each year of age (year 00–year 99) according to the criteria based on FMV total U-LH ≥0.6 IU/L and S-LH ≥0.3 IU/L separately using χ² and Fisher’s Exact tests. Differences were considered statistically significant at p < 0.05.

ROC analysis revealed that total U-LH cutoff levels at 0.60 and 0.63 IU/L in girls and boys, respectively, predicted the presence of a S-LH concentration of 0.3 IU/L or more with a sensitivity of 97.4% and a specificity of 90.6%. The same ROC analysis provided higher cutoff levels for U-LH corresponding to 0.78 IU/L in boys and 0.79 IU/L in girls, respectively, for a higher level of specificity (94.7%) at the expense of a slightly lower level of sensitivity (94.1%). The area under the curve was 0.98 in boys and 0.99 in girls (shown in Fig. 1).

The biochemical markers (FMV total U-LH ≥0.6 IU/L and early morning S-LH ≥0.3 IU/L) identified the onset of puberty earlier than the clinical Tanner criteria, except S-LH in girls (Table 1). The cutoff level of FMV total U-LH (0.6 IU/L) identified the onset of central puberty occurring at approximately the same ages (9.00–9.99 years) in both sexes (Table 1). Based on the increase in FMV total U-LH (or S-LH) levels, central pubertal development was activated at the mean age of 10.3 (10.3) in boys and age 10.5 (10.6) in girls. The youngest age at which a rise in both FMV total U-LH levels above 0.6 IU/L and in early morning S-LH levels above 0.3 IU/L was observed was 10.2 years in both sexes, whereas it was not possible to detect the onset of central puberty by morning S-LH in girls younger than those in the 10.00–10.99 age group. The lag time from biochemical to clinical onset of puberty was markedly shorter in girls than in boys (Table 1). At the age of 13 years, but not at earlier ages, 100% of the boys and girls were classified as pubertal by both biochemical markers and the clinical Tanner criteria (Table 1). Overall, higher percentages of clinically prepubertal boys and girls were found to have achieved central onset of puberty according to the biochemically pubertal cutoff level of FMV total U-LH (0.6 IU/L) than by the corresponding cutoff level of early morning S-LH (0.3 IU/L) across each yearly age group (shown in Table 1; Fig. 2c, d).

The increase in FMV total U-LH levels to or above the level of 0.6 IU/L preceded the onset of clinical puberty in all adolescents (shown in Fig. 2), except in six clinically pubertal girls aged 10–12 years of age. Early morning S-LH levels in five of these girls were at prepubertal levels (below 0.3 IU/L) and in the sixth one at a level of 0.32 IU/L (shown in Fig. 2b). The S-FSH concentrations of these six clinically pubertal but biochemically prepubertal girls, ranging between 2.3 and 2.7 IU/L, were significantly higher than those measured in both biochemically and clinically prepubertal girls of the same 10–12-year-old age group and significantly lower than those measured in both biochemically and clinically pubertal girls (p = 0.039 and p = 0.018, respectively). The 95% confidence intervals of S-FSH concentrations in pubertal but biochemically prepubertal girls did not overlap with those in biochemically and clinically prepubertal or biochemically and clinically pubertal girls (Table 2).

With the advent of ultrasensitive assays, which can detect LH concentrations as low as 0.01 IU/L, a basal morning serum LH concentration at 0.3 IU/L or above is now considered to be diagnostic for central precocious puberty, reflecting a GnRH-stimulated LH concentration greater than 5 IU/L [27, 28]. ROC analysis of our data revealed that a cutoff value of 0.6 IU/L for FMV total U-LH determinations predicts central pubertal activation with exceptionally high sensitivity and high specificity. Meanwhile, a higher cutoff value of 0.8 IU/L enhances the specificity to a markedly high level, while maintaining a substantially high sensitivity. Therefore, FMV total U-LH determination could serve as a clinically useful, noninvasive tool for screening and managing pubertal disorders [27, 28]. On the other hand, early morning S-LH level determinations are known to miss approximately 10% of cases with the onset of puberty, mainly in girls, depending on the assay method [14, 27]. As early morning S-LH with levels of 0.3 IU/L or higher do not always accurately predict progressive puberty, FMV total U-LH level at 0.6 IU/L may offer a more sensitive threshold for evaluating the onset of central puberty. Total FMV U-LH determinations may potentially replace early morning S-LH determinations or other invasive investigations such as the GnRH-stimulation test. Further comparative studies are essential to establish this potential.

As a secondary finding of this retrospective analysis of our 1996 dataset, we confirmed our recent finding that central pubertal development, using total U-LH ≥0.6 IU/L and S-LH ≥0.3 IU/L as cutoffs, is activated at around the same age in boys and girls [12]. Indeed, in this re-analysis, the first rise in FMV total U-LH over the cutoff level of 0.6 IU/L occurred at around 10 years of age in both sexes, which is in line with our recent report on the sex-independent timing of the onset of central puberty revealed by nocturnal LH concentrations [12]. Conversely, a notably higher percentage of girls than boys progressed into clinical puberty at the age of 10–11 years, establishing a marked sex difference in the lag time, exceeding 1 year, between biochemical activation of gonadotropin secretion and the emergence of clinical signs of puberty, a set of findings that reaffirms the results presented in our previous study [12]. Indeed, in clinical practice, managing prepubertal girls with pubertal FMV U-LH levels may pose some challenges, yet the phenomenon of increased U-LH preceding physical puberty signs has been well established and supported by numerous studies since its initial report in 1996 [1, 2, 11, 12].

Isolated thelarche presents a challenge in the differential diagnosis of true central puberty in girls [29‒32]. In this cohort, the S-FSH levels in the clinically pubertal but biochemically prepubertal girls were higher than those in the biochemically and clinically prepubertal girls, and lower than those in the biochemically and clinically pubertal girls. FSH may be involved in the onset of breast development in girls, potentially by inducing sufficient estrogen levels to stimulate glandular breast tissue growth [29, 32]. Verification of the potential usefulness of FMV urinary FSH (U-FSH) concentrations in a sufficiently large population of clinically pubertal but biochemically prepubertal girls may provide insights into this matter; however, it might be premature to consider it a noninvasive diagnostic approach for differential diagnosis at this stage. The subgroup of girls and boys at Tanner stage 2 in our study was significantly limited in size, highlighting the clear necessity for prospective studies on the onset of puberty in peripubertal healthy subjects, as well as in those with central precocious or delayed puberty.

In our recently published study, we revealed that a commercially available immunochemiluminometric assay offers a viable alternative to the recently discontinued DELFIA IFMA for detecting all types of U-LH immunoreactivity [33], promising significant advancements in the development of noninvasive, effective, and scalable methods for monitoring total U-LH variations across different physiological and clinical contexts documented in our earlier studies: minipuberty in newborns, the onset of central puberty in peripubertal children, puberty-related disorders in adolescents, and the fertility cycle in women, with a particular focus on post-peak changes [1‒3, 8‒12, 34, 35]. Furthermore, findings from this recently published study demonstrate that another commercially available chemiluminescent microparticle immunoassay serves as a valuable tool for assessing non-intact LH immunoreactivity levels by calculating the arithmetic difference between total and intact LH concentrations [33].

We conclude that an FMV total U-LH concentration of 0.6 IU/L or above can be considered a sign of central activation of pubertal onset in both sexes and may replace early morning serum LH determinations in this regard. Our findings suggest two different FMV total U-LH cutoff levels, as specified by ROC analysis. The former, set at 0.6 IU/L, can be proposed for screening purposes, providing a potential tool for early identification in clinical settings. The latter, the 0.8 IU/L level, which can be considered for management purposes, requires additional research to establish its effectiveness and applicability in clinical practice. In fact, before relying exclusively on FMV total U-LH measurements in the management of children with an abnormal timing of puberty, these findings should be confirmed in patients showing signs of precocious or delayed puberty. The utility of concurrent S-FSH or noninvasive FMV U-FSH determination should also be studied more extensively before it is recommended for use in the differential diagnosis of girls with isolated thelarche. Thus, FMV urinary gonadotropin determinations can be proposed as a clinically useful, noninvasive, and practical, thus a highly scalable method for the evaluation of activation of the HPG-axis.

The authors express their gratitude to the patients and their parents or custodians for their understanding and cooperation in this study.

The study was conducted in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of the University of Helsinki and Helsinki University Hospital (Ethics Committee Approval HYKS/52-M/1991, HUS/2426/2018; Research Permit HUS/54/2019). Written informed consent was obtained from all parents or guardians and directly from the children and adolescents as they are all over 6 years of age.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as potential conflicts of interest.

A part of the study was funded by grants from the Suomen Lääketieteen Säätiö (Finnish Medical Foundation, Grant Nos. 3583, 5393) and Lastentautien Tutkimussäätiö (Foundation for Pediatric Research in Finland). The funder had no role in the design, data collection, data analysis, or reporting of the study.

Conceptualization, methodology, validation, investigation, and resources: A.D., A.J., and K.M.M.; funding acquisition, formal analysis, data curation, and writing – original draft preparation: A.D. and M.H.; writing – review and editing, visualization, and project administration: A.D., M.H., A.J., and K.M.M.; and supervision: A.J., M.H., and K.M.M. All authors have read and agreed to the published version of the manuscript.

The data that support the findings of this study are not publicly available due to some indirect information that could compromise the privacy of research participants but are available from the corresponding author (A.D.) upon reasonable request.