Peak serum luteinising hormone cut‐off during gonadotropin‐releasing hormone analogue test for diagnosing central precocious puberty was lower in girls with obesity as compared with girls with normal weight

Serum luteinising hormone (LH) concentration has been reported to be lower in girls with overweight and obesity (OW/OB) as compared with girls with normal weight (NW). This study aimed to evaluate peak serum LH concentration during gonadotropin‐releasing hormone analogue (GnRHa) test in girls with OW/OB and NW who had central precocious puberty (CPP) and to determine peak serum LH cut‐off for diagnosing CPP in girls with OW/OB.

Central precocious puberty (CPP) in girls is characterised by progressive breast development before 8 years of age, concomitant with increased height velocity and advanced bone age. 1 Both basal and gonadotropin-releasing hormone (GnRH)-or GnRH analogue (GnRHa)-stimulated serum luteinising hormone (LH) concentrations have been used for diagnosing CPP. 2,3Currently, basal serum LH concentrations of 0.2-0.3[4][5][6][7][8] Serum LH concentration has been reported to be lower in both prepubertal and pubertal girls who are overweight and obese as compared with girls with normal weight. 2Several hypotheses have been proposed to explain the finding.In adults, increased body mass index (BMI) was associated with increased clearance of endogenous LH, and thus could result in lower serum LH concentration in individuals with greater BMI. 9 Increased bioavailable oestrogens in obese individuals provide negative feedback on hypothalamus and pituitary gland, and thus, cause reduced serum LH concentrations. 10,11Several mechanisms have been proposed to contribute to increased bioavailable oestrogens in obese children.Decreased hepatic oestrogen metabolism and increased aromatase activity in adipose tissue cause increased serum oestrogen. 10Decreased hepatic sex hormone-binding globulin (SHBG) production, another pathway to increased bioavailable oestrogens, was previously thought to be related to hyperinsulinaemia and insulin resistance generally found in obese children. 12,13However, emerging evidence showed that monosaccharides (glucose and fructose)-induced lipogenesis reduced human SHBG production in transgenic mice and HepG2 hepatoblastoma cells. 14,15Lastly, increased adrenocorticotrophic hormone secretion secondary to increased peripheral metabolism of cortisol in obese individuals causes increased production of adrenal androgens which can be converted to oestrogens. 16,179][20] A few studies which analysed peak serum LH concentrations during GnRH test in CPP girls with OW/OB in comparison with CPP girls with normal weight (NW) found lower peak serum LH concentrations in girls with OW/OB. 21,22A Korean study showed lower peak serum LH concentrations in CPP girls with greater BMI only in those with Tanner stages II and III breasts, but not in girls with Tanner stage IV breasts. 23wever, a Chinese study did not demonstrate the difference of peak serum LH concentrations among girls with different weight statuses. 24erefore, conflicting data regarding the difference in peak serum LH concentrations between CPP girls with OW/OB and NW exist.Health, Ministry of Public Health, Thailand. 25Overweight was defined as BMI SDS of greater than +1 to +2 and obesity was defined as BMI SDS of greater than +2. 25 All enroled girls were classified into NW and OW/OB groups.Breast Tanner stage was determined by two experienced paediatric endocrinologists (P.P. and P. M.) using Marshall and Tanner classification. 26Bone age X-ray results according to Greulich and Pyle standard 27 were collected.
According to our institute's practice, all girls who had neither advanced stages of breast development nor accelerated height velocity with advanced bone age at the time of presentation were observed for 3-6 months following GnRHa testing without depot GnRHa treatment.CPP was diagnosed in girls who had all the findings of progressive breast development, accelerated height velocity and advanced bone age of greater than 1 year when compared to chronological age, independent of GnRHa test results.
Conversely, girls without all aforementioned findings, including girls who had only advanced bone age without progressive breast development and accelerated height velocity, were defined as PT.Subcutaneous GnRHa test using 100 µg of triptorelin acetate (Diphereline ® ; Ipsen Pharma Biotech) was performed in the morning according to the standard protocol of our institute. 7Briefly, following the insertion of intravenous cannula for blood sampling, blood samples for the measurements of basal serum LH, follicle-stimulating hormone (FSH) and oestradiol concentrations were collected.Subcutaneous triptorelin acetate was then administered.Following triptorelin injection, serum LH concentrations were measured at 60, 90 and 120 min, and serum FSH concentration was determined at 120 min, the time point of peak serum FSH concentration achieved. 7Peak serum LH concentration was defined as the maximum LH concentration achieved during the testing.Peak serum LH concentration was usually achieved at 60 min following triptorelin acetate administration in girls with CPP. 7 Chemiluminescent microparticle immunoassay using Architect i2000SR analyzer ® (Abbott) was used for the analysis of serum LH and FSH concentrations.Serum oestradiol concentration was measured by electrochemiluminescence assay using Cobas e602 analyzer ® (Roche).

| RESULTS
There were 971 girls enroled.Their median (IQR) age at breast onset was 7.4 (7.0, 7.8) years.Of 971 girls, 497 and 474 patients had Tanner stages II and III breasts, respectively.There were 337 (34.7%) girls with overweight (N = 216) and obesity (N = 121).The remaining 634 (65.3%) girls had NW.In OW/OB group, 205 of them were classified as CPP and 132 patients had PT.In the NW group, 395 patients were diagnosed as having CPP and the remaining 239 patients had PT (Figure 1).Comparing between NW and OW/OB groups, there was no significant difference in age at breast onset (7.4 (6.9, 7.8) vs. 7.5 (7.0, 7.9) years, p = .083).Girls with OW/OB had more advanced bone age at the time of GnRHa testing than girls with NW (9.4 (8.8, 10.5) vs. 8.8 (7.8, 9.4) years, p < .001).Clinical characteristics of all patients in each subgroup are presented in Table 1.
The profiles of serum LH and FSH changes during GnRHa testing in girls with CPP and PT are presented in Figure 2. Peak serum LH and FSH concentrations were achieved at 60 and 120 min, respectively, following subcutaneous triptorelin acetate administration in girls with CPP.The distribution of the peak serum LH concentrations in girls with CPP and PT is presented in Figure 3.
In girls with CPP, their basal and peak serum LH and FSH concentrations, and LH:FSH ratio during GnRHa testing were not different between NW and OW/OB groups for either Tanner stage II or Tanner stage III breasts (Table 1).
Additionally, subgroup analyses according to breast Tanner stages did not show correlations between BMI SDS and hormonal profiles in both Tanner stages II and III breasts.A significant correlation between basal and peak serum LH concentrations (r = .68,p < .01)was found.Comparing between NW and OW/OB groups, OW/OB group had a slightly greater correlation between basal and peak serum LH concentrations than NW group did (r = .71,p < .01 vs. r = .65,p < .01).
The diagnostic accuracy of peak serum LH cut-offs for diagnosing CPP was determined in both girls with NW and OW/OB.The AUCs (95% CI) of peak serum LH in all girls, girls with NW, and girls with OW/OB were 0.920 (0.902-0.937), 0.899 (0.874-0.923) and 0.955 (0.935-0.975), p < .001,respectively.Peak serum LH concentration of 5 IU/L, which is the current widely used cut-off for diagnosing girls with CPP, 2 provided the sensitivity of 76% and specificity of 92% for diagnosing CPP when analysed all 971 patients.
In the NW group, peak serum LH concentration of 5 IU/L had the sensitivity of 75% and specificity of 90%.However, in the OW/OB group, peak serum LH cut-off of 5 IU/L had greater sensitivity and specificity of 79% and 95%, respectively.Lower peak serum LH cutoffs to 4 and 4.5 IU/L in the OW/OB group provided greater sensitivity at 86% and 82%, respectively, with acceptable and comparable specificity at 93%-94%.In contrast, peak serum LH cut-offs of 4 and 4.5 IU/L in the NW group had comparable sensitivity of 85% and 81%, respectively, but the specificities of 7.5 (7.0, 7.9) 7.5 (6.9, 7.9) .719 Age at GnRHa test (years) both cut-offs were lower at 76% and 85%, respectively.Subgroup analyses according to Tanner stages of breasts provided comparable results (Table 2 and Figure 4).
Lowering peak serum LH cut-off from 5 to 4 IU/L, 15 more girls in the OW/OB group who were classified as CPP (7.3%, 162 to 177 out of 205) would have been defined as CPP by GnRHa test results.
In the meantime, 3 girls with OW/OB who were classified as PT

| DISCUSSION
To the best of our knowledge, this is the first study which demonstrated that peak serum LH cut-off for diagnosing CPP in girls with OW/OB was lower than that in girls with NW with comparable diagnostic accuracy.Previous studies of serum LH cutoffs for diagnosing CPP in girls did not consider the effect of obesity on serum LH cut-offs. 2This study showed that peak serum LH cut-off of 5 IU/L had sensitivity of 76% and specificity of 92% in diagnosing girls with CPP, which is in agreement with the previous studies. 18,28Interestingly, this study demonstrated that lower   9 These could partly support the finding of our study with regard to the interaction between BMI and serum LH cut-off.
There has been no pharmacokinetic data of triptorelin in children.In healthy adult volunteers, following subcutaneous injection of triptorelin at a dose of 0.1 mg, time to maximum triptorelin concentration was 0.63 ± 0.26 h with a peak plasma triptorelin concentration of 1.85 ± 0.23 ng/mL. 29However, no pharmacokinetic data of individuals with OW/OB are available.
[23] However, no significant differences in peak serum LH concentrations between CPP girls with OW/OB and NW were demonstrated in this study, which is in agreement with a Chinese study which also showed no differences in peak serum LH concentrations among CPP girls with normal weight, overweight and obesity. 24They hypothesised that the correlation between BMI and peak serum LH concentration T A B L E 2 Sensitivity and specificity of different peak serum luteinising hormone (LH) cut-offs for diagnosing central precocious puberty in normal-weight and overweight/obese girls with Tanner stages II and III breasts.Bone age of CPP girls with OW/OB was significantly greater than that of CPP girls with NW despite having comparable chronological age.The mechanism underlying the advancement of bone age in girls with OW/OB is unclear.Alterations in androgens, oestrogens and SHBG as well as increased insulin resistance and secretion may contribute to the finding. 32Additionally, increased growth hormone and insulin-like growth factor 1 in obese children might also play a role. 33e strength of this study was that a large number of girls with premature breast development were included, so subgroup analyses according to each Tanner stage were practicable.However, we acknowledge the limitation of the study that relatively small proportions of girls with OW/OB were included.This limited the analysis of the degree of OW/OB in this study.
In conclusion, lower peak serum LH cut-off to 4 IU/L for diagnosing CPP in girls with OW/OB should be considered to avoid underdiagnosis of the condition.
Additionally, there have been no studies concerning the peak serum LH cut-off for diagnosing CPP girls who are overweight and obese.This study aimed to evaluate peak serum LH concentrations following GnRHa test of CPP girls with OW/OB in comparison with those of CPP girls with NW, and to define the peak serum LH cut-off for diagnosing CPP in girls who are overweight and obese.

2 |
MATERIALS AND METHODS Medical records of girls who had breast development before 8 years of age and underwent subcutaneous GnRHa test between October 2007 and April 2022 at the Pediatric Endocrine Unit, Faculty of Medicine Ramathibodi Hospital (a tertiary teaching hospital), Mahidol University, Bangkok, Thailand, were retrospectively reviewed.All girls had either Tanner stage II or III breasts at the time of GnRHa test performance.Girls with benign premature thelarche (PT) developed before 3 years of age, gonadotropin-independent precocious puberty, no clinical course data of progression of breast development and chronic illnesses were excluded.Clinical data including ages at breast onset and at the time of GnRHa test performance, weight, height, BMI, Tanner stages of breast development at the time of GnRHa test performance and GnRHa test results were collected.Standard deviation scores (SDSs) of weight, height and BMI were calculated using the 'National growth references for Thai children aged 5-19 years, the year 2020', Bureau of Nutrition, Department of The study protocol was approved by the Ethics Committee on Human Research of the Faculty of Medicine Ramathibodi Hospital, Mahidol University, Bangkok, Thailand (MURA2021/886) and conformed to the Declaration of Helsinki.SPSS for Windows software program, version 24, was used for the statistical analysis.Categorical data were presented in number and percentage.All continuous data were non-normally distributed, so they were expressed as median (interquartile range [IQR]).Mann-Whitney U test was used for the comparisons of variables between the NW and OW/OB groups.Spearman's correlation was used to describe the correlations between BMI SDS and other variables.The receiver operating characteristic curves of GnRHa test results were constructed to assess the areas under the curves (AUCs) with 95% confidence interval (CI).The sensitivity and specificity of different serum LH cut-offs for CPP diagnosis were determined.A p <.05 was considered statistically significant.

1
Classification of all enroled patients.CPP, central precocious puberty; PT, premature thelarche.T A B L E 1 Clinical characteristics and gonadotropin-releasing hormone analogue (GnRHa) test results of all enroled patients with central precocious puberty (CPP) and premature thelarche (PT) according to their Tanner stages of breasts and body mass index (BMI) categories.

F I G U R E 2
Serum luteinising hormone (LH) and follicle-stimulating hormone (FSH) concentrations versus time (min) after 0.1 mg subcutaneous triptorelin acetate injection.Solid lines represent girls with central precocious puberty and dashed lines represent girls with premature thelarche.Data are presented as median and interquartile range.F I G U R E 3 Distribution of peak serum luteinising hormone (LH) concentrations among the enroled girls.Black bars represent girls with central precocious puberty and striped bars represent girls with premature thelarche.Each bar demonstrates the number of girls who had peak serum LH concentrations between the consecutive peak serum LH concentrations, for example, the bars between 1 and 2 IU/L represent the number of girls who had peak serum LH concentrations between 1 and 2 IU/L.*There were three girls in the premature thelarche group who had high peak serum LH concentrations of 21, 28, and 33 IU/L.Those girls had regression of breast development within 6-month of follow-up without depot gonadotropin-releasing hormone analogue treatment.
linear.Several factors, including serum oestradiol, adrenal androgens, fasting insulin and glucose concentrations, might serve as factors that influenced peak serum LH concentrations.This study included only girls with relatively early stages of puberty (Tanner stages II and III breasts) whose peak serum LH concentrations are not as high as those of girls with later stages of puberty. 31Therefore, differences in peak serum LH concentrations between girls with OW/OB and NW, if any, might not be demonstrated.F I G U R E 4 Diagnostic accuracies of peak serum luteinising hormone (LH) and follicle-stimulating hormone (FSH) concentrations, and peak serum LH:FSH ratio for diagnosing central precocious puberty in girls with normal weight, and overweight and obesity.(A) peak serum LH in all enroled girls; (B) peak serum LH in girls with Tanner stage II breasts; (C) peak serum LH in girls with Tanner stage III breasts; (D) peak serum FSH in all enroled girls; (E) peak serum FSH in girls with Tanner stage II breasts; (F) peak serum FSH in girls with Tanner stage III breasts; (G) peak serum LH:FSH ratio in all enroled girls; (H) peak serum LH:FSH ratio in girls with Tanner stage II breasts; (I) peak serum LH:FSH ratio in girls with Tanner stage III breasts.Sensitivity (sens.) is shown in solid lines and specificity (spec.) is shown in dashed lines.F I G U R E 4 (Continued) Data are expressed as median (interquartile range).Auxological data were obtained on the day of GnRHa testing.