To study the feasibility of noninvasive sampling in pediatric patients, we examined the concentrations of LH and FSH in paired serum and urine samples from 65 children (age 0-15 y) with highly sensitive time-resolved immunofluorometric assays. The detection limits of the assays were 0.015 IU/L for LH and 0.018 IU/L for FSH. These sensitivity levels allowed quantification of the low prepubertal LH and FSH concentrations. The correlation between serum and urine gonadotropin values was very good (r = 0.751,~ < 0.001 for FSH; and r = 0.720,~ < 0.001 for LH), and the urine and serum concentrations were very similar. Correction of urinary gonadotropin concentrations Urine sampling has made a comeback in pediatric endocrinologic diagnostics as a result of efforts to avoid repeated, invasive blood sampling. Much interest has been focused in particular on GH in urine (1). RIA methods were not sensitive enough to measure urinary GH levels (2), but measurement of urinary GH levels became possible using more sensitive immunoassays like the immunoradiometric assay (3), sandwich enzyme immunoassay (4), and IFMA (5, 6). The situation has been similar for the gonadotropins. The presence of gonadotropins in prepubertal urine was first demonstrated in the 1960s by Fitschen and Clayton (7) and Kulin et al. (8) using bioassay and by Bagshawe et al. using RIA ( 9 ) . In 1970, FSH and L H were quantitatively measured in urine by Rifkind et al. (10) using RIA. To improve sensitivity, extraction and concentration methods were later included and bet-

Determinations of serum gonadotropin concentrations by ultra-sensitive methods have improved the diagnosis of pubertal disorders. The onset of puberty can be estimated by measuring serum gonadotropin pulsation, but as this requires serial nocturnal blood sampling, it is not a routine investigation. Gonadotropin measurements in first morning voided (FMV) urine samples could reflect the integrated nocturnal gonadotropin secretion and predict pubertal development earlier than daytime serum measurements. We studied the value of urinary LH (U-LH) measurements in FMV urine with reference to serum LH (S-LH) levels using an ultrasensitive time-resolved immunofluorometric assay in samples from 297 children and adolescents (145 boys and 152 girls, aged 5-15 yr) with known pubertal stages (Tanner 1-5). Stage 1 subjects (prepubertal) were divided into 5 age groups to assess whether there is an increase in LH before clinical signs of puberty can be detected. The correlation between FMV urine and S-LH values was good (r = 0.64; P < 0.0001). The 2 oldest groups of prepubertal subjects (11 and 12 yr) had significantly higher (P < 0.001) U-LH concentrations than the 3 younger groups. This difference was less marked for S-LH. A significant increase in FMV U-LH concentration occurs before the first clinical signs of puberty in a sex-independent fashion. Our data indicate that FMV U-LH measurement is a clinically relevant, noninvasive method for the evaluation of pubertal development, and it may be helpful in the investigation of pubertal disorders.

Aims: We studied whether first morning voided (FMV) urinary gonadotropin measurements could be used as a noninvasive alternative to the GnRH test in the assessment of the hypothalamic-pituitary-gonadal function in children. Methods: In a single-center study, we compared FMV urinary gonadotropin concentrations with basal and GnRH-stimulated serum gonadotropin levels in 274 children and adolescents (78 girls, 196 boys) aged 5-17 years referred for growth and pubertal disorders. The concordance between FMV urinary gonadotropin concentrations and GnRH test results was assessed. Results: FMV urinary LH (U-LH), urinary FSH (U-FSH) and their ratios correlated well with the corresponding basal and GnRH-stimulated serum parameters (r ≥ 0.66, p < 0.001). Receiver operating characteristic curve analyses using urinary and serum LH and FSH concentrations showed that FMV U-LH and U-LH/U-FSH performed equally well as the GnRH test in the differentiation of early puberty (Tanner stage 2) from prepuberty (Tanner stage 1) (area under the curve 0.768-0.890 vs. 0.712-0.858). FMV U-LH and U-LH/U-FSH performed equally well as basal serum LH in predicting a pubertal GnRH test result (area under the curve 0.90-0.93). Conclusion: FMV U-LH determination can be used for the evaluation of pubertal development and its disorders, reducing the need for invasive GnRH stimulation tests.

Serum levels of LH and FSH are very low from about 2 yr of age to the onset of puberty, which is heralded by a very sharp increase in LH levels. We studied age-related changes in urinary gonadotropins in a total of 184 boys and girls of various ages. Urinary FSH and LH were measured by ultrasensitive time-resolved immunofluorometric assays. The detection limit was 0.015 IU/L for LH and 0.018 IU/L for FSH. We observed that after an initial drop following the first months of life, urinary LH levels stayed below 0.5 IU/L until age 9 yr in girls and below 1.0 IU/L until age 11 yr in boys, whereas mean urinary FSH levels remained below 3.0 IU/L until age 10 yr in girls and 12 yr in boys. During puberty, mean urinary FSH and LH concentrations increased to about 5 IU/L in boys and 10 IU/L in girls. This corresponds to a 5-fold increase in FSH in both sexes and a 50- to 100-fold increase in LH in boys and girls, respectively. These dynamic changes agree with previous reports regarding serum levels, suggesting that noninvasive urinary gonadotropin measurements can be a viable alternative to serum determinations in the evaluation of gonadotropin secretion during childhood and adolescence.