Elevated serum phoenixin levels in boys with central precocious puberty are positively correlated with BMI

Abstract

Objective

This study investigated changes in serum phoenixin (PNX) levels in boys with central precocious puberty (CPP), aiming to evaluate its diagnostic and therapeutic relevance and identify factors influencing its expression.

Method

Serum samples were collected from 43 boys with CPP and 48 age-matched prepubertal controls. Participants were further divided into overweight/obese and normal-weight subgroups based on body mass index (BMI). Levels of phoenixin-14 (PNX-14) and phoenixin-20 (PNX-20) were measured using enzyme-linked immunosorbent assay (ELISA). Correlations with luteinizing hormone (LH), testosterone, and BMI were analyzed. Multiple linear regression identified independent influencing factors, and receiver operating characteristic (ROC) analysis assessed the diagnostic performance of PNX. Twelve CPP boys were followed for 6 months after gonadotropin-releasing hormone analog (GnRHa) treatment.

Results

PNX-20 and PNX-14 levels were significantly higher in the CPP group than in controls (P = 0.001). Among subgroups, the overweight/obese CPP group exhibited the highest levels. Both PNX forms were positively correlated with baseline LH, testosterone, and BMI; PNX-14 was most strongly associated with LH, and PNX-20 with BMI. ROC analysis showed moderate diagnostic value for PNX-14 (AUC: 0.710) and PNX-20 (AUC: 0.696), with a combined AUC of 0.718. Following treatment, levels of PNX-20, LH, follicle-stimulating hormone (FSH), and testosterone significantly declined, while PNX-14 remained unchanged.

Conclusions

PNX is associated with pubertal progression and BMI in boys with CPP. While not suitable as a standalone diagnostic marker, PNX – especially PNX-20 – may serve as a useful adjunct for diagnosis and treatment monitoring.

Introduction

Central precocious puberty (CPP) is a common endocrine disorder, and its incidence has been steadily increasing in recent years, with children showing signs of puberty at progressively younger ages (1, 2). Alongside this trend, the prevalence of childhood obesity has also risen (3). Obesity is known to raise the risk of chronic diseases in adulthood, but it may also have an impact on the timing of puberty and overall reproductive function (4). Emerging research highlights that excessive body weight may accelerate pubertal onset, particularly in girls (5, 6). Obesity is a significant risk factor for early puberty, with metabolic signals such as leptin playing a crucial role. While leptin deficiency may delay puberty, its excess can promote early onset (7, 8). In addition, elevated insulin levels significantly increase LH secretion, accelerating the onset of puberty (9). In boys, however, the association between obesity and pubertal timing is less clear and remains a subject of debate (10). Some studies suggest that obese boys enter puberty earlier (11, 12), while others report delayed puberty in obese boys (13). Furthermore, some studies have not observed any association between obesity and the timing of pubertal onset in boys (14, 15). Thus, the relationship between different metabolic states and the initiation of puberty in males remains an important topic for investigation.

Phoenixin (PNX) is a newly identified neuropeptide encoded by the SMIM20 gene. It exists in two active forms – PNX-14 and PNX-20 – that share biological functions (16). PNX is expressed in several organs involved in the hypothalamic–pituitary–gonadal axis (HPGA), such as the hypothalamus, pituitary gland, and ovaries (16, 17). In vitro studies have demonstrated that PNX, through binding to G protein-coupled receptor 173 (GPR173), activates the cyclic adenosine monophosphate (cAMP)/protein kinase A (PKA) signaling pathway in hypothalamic cell lines (18). This pathway upregulates Kiss1 gene expression, which leads to the synthesis of kisspeptin, a neuropeptide that stimulates GnRH release and subsequently increases GnRH mRNA levels (18, 19, 20). In addition, siRNA-mediated knockdown of PNX in adult female rats has been shown to reduce GnRH receptor expression and delay the estrous cycle (17). This suggests that PNX may participate in reproductive regulation by modulating the expression of GnRH receptors in the pituitary (17). PNX has also been implicated in the regulation of key reproductive hormones, including gonadotropin-releasing hormone (GnRH), LH, FSH, estrogen, and testosterone (17, 21, 22). For instance, intracerebroventricular injection of PNX in female rats increases plasma LH and GnRH levels (21), and PNX-20 administration in male rats elevates serum LH and testosterone levels (22). Furthermore, silencing of GPR173 impairs GnRH-induced LH secretion (21). PNX has been shown to promote follicular growth and maturation, enhancing estradiol production in granulosa cells via activation of the cAMP/PKA signaling pathway and phosphorylation of CREB (23). In women with polycystic ovary syndrome (PCOS), elevated PNX is associated with increased LH and testosterone levels (24), supporting its role in female reproductive regulation. Collectively, these findings suggest that PNX contributes to the initiation of puberty and regulation of reproductive function by acting on kisspeptin neurons and influencing the secretion of GnRH, LH, and FSH. Recent studies have shown that serum PNX levels are significantly higher in girls with CPP compared to their prepubertal peers, indicating that PNX may be involved in triggering pubertal onset (25).

However, to date, there has been no research exploring this relationship in boys. Therefore, this study aims to compare serum levels of PNX-14 and PNX-20 in normal-weight and overweight/obese boys with CPP, as well as in age-matched prepubertal controls. We seek to determine whether PNX levels are influenced by BMI or pubertal development and to explore the potential of PNX as a diagnostic marker for CPP and an indicator of treatment efficacy in boys.

Materials and methods

Sample collection and storage

Fasting venous blood samples (2–3 mL) were collected from all participants. The samples were centrifuged at 2,100 g for 15 min, and the serum was separated and stored at −20°C until analysis.

Study participants

Serum samples and clinical records were collected from 43 boys diagnosed with idiopathic central precocious puberty (ICPP) at the Pediatric Outpatient Department of the First Affiliated Hospital of Guangxi Medical University between September 2020 and September 2024 (Group A). Forty-eight age-matched prepubertal boys undergoing routine health evaluations were recruited as the control group (Group B). Of the CPP group, 12 boys received GnRHa therapy and were monitored over 6 months: four were classified as overweight/obese and eight as having normal weight. The GnRHa administered was triptorelin acetate (Ferring Pharmaceuticals, Hoechst, Germany) via subcutaneous injection. The standard initial dose was 3.75 mg, followed by subsequent doses of 80–100 μg/kg (4 weeks) (maximum dose: 3.75 mg) (26). To examine the influence of BMI, participants in both groups were subdivided into overweight/obese and normal-weight categories: overweight/obese CPP (Group A1), normal-weight CPP (Group A2), overweight/obese control (Group B1), and normal-weight control (Group B2).

Inclusion criteria

Children in the ICPP group met the diagnostic criteria for CPP (26): i) early appearance of secondary sexual characteristics; ii) enlarged gonads, with testicular volume ≥4 mL; iii) peak LH level ≥5.0 IU/L and LH peak/FSH peak ratio ≥0.6 during the GnRHa stimulation test (measured by immunoluminescence assay, ICMA); iv) bone age (BA) advanced by ≥ 1 year compared to chronological age; v) accelerated linear growth exceeding the age-related norm. All CPP cases were first-time diagnoses, with no prior treatment, and cranial MRI scans showed no abnormalities.

NC group: age-matched prepubertal boys undergoing routine health examinations.

Exclusion criteria

i) Incomplete clinical or laboratory data; ii) precocious puberty secondary to organic conditions, including central nervous system disorders, congenital adrenal hyperplasia, McCune-Albright syndrome, or primary hypothyroidism (2); iii) co-existing endocrine or systemic diseases.

Definition of weight categories

Overweight and obese boys were defined as those with a BMI above the 85th percentile, while normal-weight boys had a BMI between the 3rd and 85th percentiles (27).

Clinical and laboratory assessments

Basic clinical data, including age, height, weight, BMI, and Tanner stage, were collected for all participants. Testicular volume was measured using a Prader orchidometer. BMI was calculated as weight (kg) divided by height squared (m²). Bone age radiographs were interpreted by two pediatric endocrinologists with at least 5 years of experience, using the Greulich-Pyle (G P) atlas method. The GnRH stimulation test was performed in the early morning. Suspected precocious puberty boys were administered a GnRHa after an 8 h fasting period at a dose of 2.5 μg/kg (maximum 100 μg). Blood samples were collected at 0, 30, 60, 90, and 120 min post-injection to measure serum FSH and LH levels. Serum LH, FSH, and testosterone levels were quantitatively analyzed using the Mindray CL-2000i chemiluminescence immunoassay system (China). Serum PNX-14 and PNX-20 levels were measured using ELISA kits (Fine Biotech Co, Ltd, China), with sensitivities of 0.938 pg/mL and 9.375 pg/mL, respectively. The intra-assay and inter-assay coefficients of variation were below 5.72 and 4.69% for PNX-14, and 5.07 and 5.5% for PNX-20, respectively.

Statistical analysis

Continuous variables with a normal distribution were expressed as mean ± standard deviation (SD), and comparisons were made using independent t-tests or ANOVA. Non-normally distributed data were expressed as median with interquartile range (median (P25–P75)), and comparisons were conducted using the Mann–Whitney U test or Kruskal–Wallis test. Paired t-tests were used to compare hormone levels before and after GnRHa treatment. Correlations between PNX and other clinical parameters were assessed using Spearman correlation. Multivariate linear regression analysis was conducted to identify the main factors influencing PNX levels. In addition, ROC curves were constructed to assess the diagnostic performance of PNX in distinguishing CPP. The cut-off value was determined based on the maximum Youden index, and sensitivity and specificity were calculated. All statistical analyses were performed using IBM SPSS Statistics (version 25.0, IBM, USA). A P-value of less than 0.05 was considered statistically significant.

Results

Clinical and laboratory characteristics

Among the 43 CPP patients (Group A), 25 were normal-weight (Group A2), and 18 were overweight/obese (Group A1). Twelve CPP patients underwent a 6-month follow-up after GnRHa treatment. The 48 controls (Group B) included 21 overweight/obese boys (Group B1), and 27 normal-weight boys (Group B2). The mean ages of Groups A and B were 9.85 ± 1.70 and 9.53 ± 1.54 years, respectively (P = 0.358), with no significant age differences among subgroups (P > 0.05). Compared to Group B, Group A showed significantly greater height, weight, baseline LH and FSH, testosterone, testicular volume, bone age, insulin-like growth factor-1 (IGF-1), and bone age advancement (BA-CA) (P < 0.05). Specifically, body weight in Group A was significantly higher than that in Group B (38.57 ± 10.57 kg vs 34.06 ± 10.25 kg, P = 0.044), whereas no significant difference was observed in BMI between the two groups (19.09 ± 3.34 vs 18.57 ± 4.29, P = 0.537). In addition, serum osteocalcin levels were significantly elevated in Group A compared to Group B (124.5 (113, 182) ng/mL vs 94.95 (62.55, 102) ng/mL, P < 0.001). Serum levels of PNX-20 and PNX-14 were significantly higher in Group A than in Group B (P = 0.001 for both). Further subgroup analysis revealed that Group A1 had higher serum levels of both PNX-20 and PNX-14 compared to their counterparts in Group B1 (PNX-20: P = 0.002; PNX-14: P < 0.001). Among normal-weight boys, Group A2 also exhibited higher levels of PNX-20 and PNX-14 than Group B2 (PNX-20: P = 0.008; PNX-14: P = 0.001). Within Group A, boys in Group A1 had significantly higher PNX-20 and PNX-14 levels than those in Group A2 (PNX-20: P < 0.001; PNX-14: P < 0.001). Similarly, in the Group B, Group B1 had higher PNX levels than Group B2 (PNX-20: P = 0.005; PNX-14: P < 0.001) (Table 1 and Fig. 1).

Table 1.

Comparison of clinical and laboratory parameters between children with CPP and normal controls.

	CPP (n = 43)		NC (n = 48)
Variables	Overweight/obese	Normal weight	Overweight/obese	Normal weight	P value
	A1 (n = 18)	A2 (n = 25)	B1 (n = 21)	B2 (n = 27)
CA (years)	9.84 ± 2.07	9.85 ± 1.42	9.46 ± 1.78	9.59 ± 1.37	0.358
BA (years)	11.55 ± 2.40c	10.74 ± 2.40d	8.92 ± 2.96	8.13 ± 1.77	<0.001‡
BA-CA (years)	1.70 ± 1.91c	0.89 ± 1.68d	−0.54 ± 1.90	−1.47 ± 1.38	<0.001‡
Testicular volume (mL)	13.25 (7.38, 16.25)c	12 (5.5, 15)d	3 (2, 3.25)	2.5 (2, 3)	<0.001‡
Weight (kg)	44.60 ± 12.45a	33.85 ± 5.52d	42.28 ± 9.56b	27.66 ± 4.72	0.044*
BMI (kg/m2)	21.96 ± 2.92a	16.84 ± 1.30d	22.74 ± 2.61b	15.33 ± 1.75	0.537
Height (cm)	140.75 ± 13.47	139.62 ± 10.85	135.43 ± 10.89	133.99 ± 7.10	0.014*
B-LH (mIU/mL)	2.35 ± 1.13c	2.42 ± 1.36d	0.23 ± 0.17	0.25 ± 0.18	<0.001‡
B-FSH (mIU/mL)	3.29 ± 1.41c	3.77 ± 2.14d	1.95 ± 1.32	1.95 ± 1.31	<0.001‡
P-LH (mIU/mL)	22.71 ± 13.96	17.68 ± 14.71	-	-	-
P-FSH (mIU/mL)	7.79 ± 4.55	5.79 ± 5.06	-	-	-
P-LH/P-FSH	3.45 ± 1.50	3.56 ± 2.50	-	-	-
IGF-1 (ng/mL)	278.03 ± 129.42c	275.06 ± 122.69d	151.38 ± 46.44	128.72 ± 44.25	<0.001‡
Testosterone (ng/mL)	1.81 (0.39, 2.55)c	1.89 (0.34, 4.10)d	0.06 (0.04, 0.19)	0.05 (0.02, 0.06)	<0.001‡
Osteocalcin (ng/mL)	128.5 (112.03, 196.75)d	120 (112, 174)c	98.1 (70.4, 112)	87 (62.2, 99.4)	<0.001‡
PNX-14 (pg/mL)	556.2 (467.22, 837.47)a,c	87.35 (39.82, 152.78)d	330.27 (93.65, 463.33)b	31.19 (21.51, 65.77)	0.001†
PNX-20 (pg/mL)	656.48 (569.34, 739.83)a,c	305.83 (160.96, 366.68)d	519.82 (166.35, 637.71)b	135.88 (105.57, 269.07)	0.001†

CPP, central precocious puberty; NC, normal control; CA, chronological age; BA, bone age; BA-CA, bone age advancement (difference between bone age and chronological age); B, base; LH, luteinizing hormone; FSH, follicle-stimulating hormone; P, peak; IGF-1, insulin-like growth factor-1.

Group comparisons: a = A1 vs A2; b = B1 vs B2; c = A1 vs B1; d = A2 vs B2.

^*P < 0.05.

^† P < 0.01.

^‡ P < 0.001.

Figure 1.

The serum (A) phoenixin-20 and (B) phoenixin-14 levels in different subgroups. CPP: central precocious puberty; NC: normal controls; Group A1: overweight/obese children with CPP; Group A2: normal-weight children with CPP; Group B1: overweight/obese controls; Group B2: normal-weight controls. **P < 0.01, ***P < 0.001.

Comparison of clinical and laboratory parameters before and after GnRHa treatment

To evaluate the changes in serum PNX levels before and after GnRHa treatment in the CPP group, we measured clinical and laboratory parameters both pre- and post-treatment. A total of 12 patients completed 6 months of GnRHa treatment. Serum baseline levels of PNX-20, LH, FSH, and testosterone were significantly reduced compared to pre-treatment levels (P < 0.05). However, no significant changes were observed in BMI or serum PNX-14 levels before and after treatment (P > 0.05). Specifically, PNX-20 levels showed significant change: 407.41 ± 212.31 pg/mL before treatment vs 293.54 ± 160.77 pg/mL after treatment (P = 0.032). Similarly, PNX-14 levels did not show a significant change: 355.99 ± 117.33 pg/mL before treatment vs 236.42 ± 61.95 pg/mL after treatment (P = 0.339) (Table 2).

Table 2.

Changes in clinical and laboratory indicators after 6 months of GnRHa treatment in children with CPP.

Parameters	Before treatment	After treatment	P value
Height (cm)	140.08 ± 13.61	145.19 ± 13.66	0.019*
Weight (kg)	36.89 ± 10.16	39.13 ± 9.11	0.310
BMI (kg/m2)	18.50 ± 3.39	18.32 ± 2.76	0.900
Testicular volume (mL)	13.42 ± 3.28	9.46 ± 2.80	<0.001†
Testosterone (ng/mL)	3.05 ± 2.01	0.19 ± 0.17	<0.001†
B-LH (mIU/mL)	2.93 ± 1.19	0.60 ± 0.39	<0.001†
B-FSH (mIU/mL)	3.51 ± 2.11	0.75 ± 0.74	<0.001†
BA-CA (years)	2.04 ± 1.26	1.50 ± 1.67	0.039*
IGF-1 (ng/mL)	343.39 ± 148.56	324.55 ± 143.34	0.552
PNX-20 (pg/mL)	407.41 ± 212.31	293.54 ± 160.77	0.032*
PNX-14 (pg/mL)	355.99 ± 117.33	236.42 ± 61.95	0.339

P value, comparison between before and after GnRHa treatment; BMI, body mass index; B-LH, baseline luteinizing hormone; B-FSH, baseline follicle-stimulating hormone; BA-CA, bone age advancement; IGF-1, insulin-like growth factor-1.

^*P < 0.05.

^† P < 0.001.

Correlations

Using Spearman correlation analysis, we examined the associations between serum PNX levels and a range of clinical and biochemical parameters. The correlation analysis revealed that both PNX-20 and PNX-14 levels were significantly positively correlated with BA, BA-CA, BMI, weight, testicular volume, baseline LH, osteocalcin (OC), and testosterone (P < 0.05). However, no significant correlations were observed between PNX-20 and PNX-14 levels and chronological age (CA), height, baseline FSH, or IGF-1 (P > 0.05) (Table 3).

Table 3.

Correlation between serum phoenixin levels and anthropometric and biochemical parameters.

Variables	n	Phoenixin-14		Phoenixin-20
		R	P-value	R	P-value
CA	91	0.050	0.637	0.004	0.967
BA	91	0.221	0.036*	0.225	0.032*
BA-CA	91	0.278	0.008†	0.313	0.002†
BMI	91	0.603	<0.001‡	0.554	<0.001‡
Weight	91	0.489	<0.001‡	0.445	<0.001‡
Height	91	0.154	0.146	0.125	0.238
Testicular volume	91	0.307	0.003†	0.268	0.010*
Testosterone	91	0.268	0.010*	0.210	0.046*
B-LH	91	0.272	0.009†	0.212	0.044*
B-FSH	91	0.171	0.106	0.036	0.734
IGF-1	91	0.190	0.071	0.163	0.124
Osteocalcin	91	0.225	0.032	0.238	0.023*

Correlations were assessed using Spearman’s method.

^*P < 0.05.

^† P < 0.01.

^‡ P < 0.001.

CA, chronological age; BA, bone age; BA-CA, bone age advancement; BMI, body mass index; B-, base-; LH, luteinizing hormone; IGF-1, insulin-like growth factor-1; FSH, follicle-stimulating hormone.

Multivariate linear regression analysis

To further explore the factors influencing serum PNX levels, we conducted two multivariate linear regression models using PNX-20 and PNX-14 as dependent variables. Independent variables included age, height, BA–CA, BMI, testicular volume, basal LH, basal FSH, testosterone, and IGF-1. Analysis revealed that BMI, basal LH, and testosterone were the major contributors to PNX levels. Among them, BMI was most strongly associated with PNX-20, while basal LH had the most pronounced effect on PNX-14 (Tables 4 and 5).

Table 4.

Multiple linear regression analysis of factors influencing serum phoenixin-20 levels.

Variables	Unstandardized coefficients		Standardize coefficients	t	P value
	B	Std. Error	Beta
(Constant)	188.312	376.079	-	0.501	0.618
Age	−6.664	23.167	−0.044	−0.288	0.774
Height	−2.840	4.046	−0.124	−0.702	0.485
BA-CA	13.436	13.810	0.114	0.973	0.334
BMI	34.425	5.779	0.544	5.957	<0.001*
Testicular volume	13.064	7.526	0.306	1.736	0.086
Basal LH	88.377	25.342	0.497	3.487	<0.001*
Basal FSH	−25.087	14.170	−0.182	−1.770	0.080
Testosterone	−46.583	19.060	−0.352	−2.444	0.017*
IGF-1	−0.415	0.273	−0.193	−1.518	0.133

Dependent variable: phoenixin-20; BA-CA, bone age advancement; BMI, body mass index; LH, luteinizing hormone; FSH, follicle-stimulating hormone; IGF-1, insulin-like growth factor-1.

^*P < 0.05 indicates statistical significance.

Table 5.

Multiple linear regression analysis of factors influencing serum phoenixin-14 levels.

Variables	Unstandardized coefficients		Standardize coefficients	t	P value
	B	Std. Error	Beta
(Constant)	58.546	534.415		0.110	0.913
Age	2.951	32.920	0.015	0.09	0.929
Height	−3.784	5.750	−0.127	−0.658	0.512
BA-CA	28.543	19.625	0.185	1.454	0.150
BMI	35.164	8.212	0.427	4.282	<0.001*
Testicular volume	3.433	10.695	0.062	0.321	0.749
Basal LH	130.887	36.011	0.565	3.635	<0.001*
Basal FSH	−20.628	20.136	−0.115	−1.024	0.309
Testosterone	−66.259	27.085	−0.385	−2.446	0.017*
IGF-1	−0.134	0.388	−0.048	−0.346	0.730

Dependent variable: phoenixin-14.

BA-CA, bone age advancement; IGF-1, insulin-like growth factor-1; LH, luteinizing hormone; BMI, body mass index; FSH, follicle-stimulating hormone.

^*P < 0.05 indicates statistical significance.

ROC analysis

ROC analysis demonstrated that the AUC for serum PNX-14 and PNX-20 levels in distinguishing CPP from the control group was 0.710 (cut-off value: 74.28 pg/mL; sensitivity: 79.1%; specificity: 56.3%) and 0.696 (cut-off value: 269.68 pg/mL; sensitivity: 76.7%; specificity: 62.5%), respectively. When these two indicators were combined, the AUC increased slightly to 0.718 (sensitivity: 81.4%; specificity: 56.3%) (Fig. 2 and Table 6).

Figure 2.

The results of ROC curve. ROC curves were used to assess the ability of PNX to differentiate children with CPP from normal controls. AUC: area under the curve.

Table 6.

ROC curve analysis for differentiating CPP from normal controls.

	Variables	AUC (95%CI)	P value	Cut-off	Youden	Specificity, %	Sensitivity, %
CPP vs NC	Phoenixin-14	0.710 (0.605, 0.815)	0.001	74.28	0.353	56.3	79.1
	Phoenixin-20	0.696 (0.588, 0.804)	0.001	269.68	0.392	62.5	76.7
	Combined	0.718 (0.613, 0.822)	<0.001	-	-	56.3	81.4

CPP, central precocious puberty; NC, normal controls; AUC, area under the curve.

Discussion

This study is the first to investigate the relationship between serum levels of PNX-14 and PNX-20 and CPP in boys. Our findings revealed that serum levels of both PNX-14 and PNX-20 were significantly higher in boys with CPP compared to healthy prepubertal boys. Multivariate linear regression analysis showed that PNX-20 and PNX-14 levels were positively correlated with baseline LH and testosterone levels. This result is consistent with findings from studies on female CPP, where PNX-20 was significantly associated with markers of HPGA activation, such as LH and estradiol (25). Animal studies have demonstrated that intracerebroventricular injection of PNX can activate GnRH neurons, promoting increases in LH, GnRH, and testosterone levels (18, 21, 22). Notably, Guvenc et al. (22) showed that central co-administration of PNX and nesfatin-1 in male rats synergistically elevated plasma LH and testosterone levels compared to single treatments, indicating their collaborative role in HPGA regulation. Given the elevated PNX levels observed in boys with CPP in our study, future research could further investigate concurrent changes in nesfatin-1, as their interplay may contribute to the premature activation of the HPG axis during early puberty. In vitro studies have indicated that PNX can upregulate the expression of GnRH receptors in anterior pituitary cells, thereby inducing an increase in FSH and LH secretion (17). These findings collectively suggest that PNX plays a role in modulating HPGA function and may be involved in regulating the initiation of puberty. Therefore, our results further support the notion that the promoting effect of PNX on the HPGA is associated with the onset of CPP. However, the precise mechanisms by which PNX exerts its effects require further elucidation.

Our study found that in both the CPP and NC groups, the PNX levels in overweight/obese boys were significantly higher than those in normal-weight boys (P = 0.001). Serum PNX levels were positively correlated with BMI, aligning with previous studies (24). After adjusting for BMI using analysis of covariance, PNX-20 and PNX-14 levels remained significantly higher in the CPP group than in the control group (P = 0.001 and P = 0.002, respectively), suggesting that PNX may independently influence pubertal development. PNX may also impact energy metabolism. It has been shown to enhance adipocyte differentiation and proliferation via the cAMP/Epac pathway, while regulating appetite and energy homeostasis (16). Fatty acids can increase PNX mRNA expression in hypothalamic neurons, and adipocyte proliferation in obesity may further enhance PNX secretion (16, 28). In addition, the increased expression of PNX induced by a high-fat diet may aggravate obesity and insulin resistance (29). Given that insulin resistance is a known risk factor for early puberty (30), PNX likely contributes to the obesity-puberty link. The Kiss1 system plays a central role in integrating metabolic cues and pubertal timing (31). In vitro, high PNX concentrations significantly upregulate Kiss1 gene expression, increasing kisspeptin production. Kisspeptin then stimulates GnRH release, raising GnRH mRNA levels (18). Childhood obesity has also been associated with elevated Kiss1 activity (19, 20). Together, these findings support a link between PNX, the Kiss1 system, obesity, metabolism, and the initiation of puberty. In studies of CPP in girls, no correlation was observed between PNX and BMI (25). In contrast, our study found a positive correlation between BMI and serum PNX levels in males, suggesting that the effect of BMI on PNX levels may show gender dimorphism. Therefore, further large-scale population studies are needed to verify the relationship between BMI and PNX in different sexes.

Our correlation analysis showed that serum levels of PNX-20 and PNX-14 were positively correlated with BA, OC, and BA-CA. Specifically, serum OC levels in the CPP group boys were significantly higher at 124.5 (113, 182) ng/mL compared to healthy prepubertal controls at 94.95 (62.55, 102) ng/mL (P < 0.001). This result is consistent with the previous literature and our preliminary findings on elevated OC levels in CPP girls, suggesting that OC may serve as a biomarker for CPP. Further correlation analysis confirmed that both bone age and LH levels are significantly associated with serum OC levels, indicating that OC not only reflects bone metabolism status but may also participate in the regulation of the HPGA (32, 33). In vitro studies have shown that PNX-20 promotes osteogenic differentiation and proliferation of MC3T3-E1 cells through the activation of p38/RUNX2 pathways via GPR173 (34). These results indicate that PNX might be involved in the formation and maturation of bones during puberty. However, whether PNX can serve as an effective biomarker for predicting bone age progression in children with CPP requires further in-depth verification.

PNX-14 and PNX-20 showed moderate accuracy in distinguishing CPP from controls, with AUCs of 0.710 (cut-off value: 74.28 pg/mL; sensitivity: 79.1%; specificity: 56.3%) and 0.696 (cut-off value: 269.68 pg/mL; sensitivity: 76.7%; specificity: 62.5%), respectively. When both markers were combined, diagnostic performance improved (AUC = 0.718), with a sensitivity of 81.4% and specificity of 56.3%. These findings suggest that PNX may serve as a useful serological auxiliary marker in the diagnosis of CPP. However, due to its limited sensitivity and specificity, it should not be used as an independent diagnostic marker.

This study analyzed changes in serum PNX levels in 12 boys with CPP after GnRHa treatment for the first time. After 6 months of treatment, serum PNX-20, baseline LH, FSH, and testosterone levels significantly decreased (P < 0.05), while BMI and serum PNX-14 showed no significant changes before and after treatment. In treated CPP children, the secretion of gonadotropins and sex hormones was significantly suppressed, thereby delaying pubertal progression. However, their body fat levels did not appear to decrease as a result of this suppression. These results are consistent with previous studies showing that BMI does not significantly change in CPP children after GnRHa treatment (35). This study showed that serum PNX-20 levels decreased significantly with the inhibition of the gonadal axis in children with CPP after 6 months of GnRHa treatment (P < 0.05). However, serum PNX-14 levels remained largely unchanged (P > 0.05). In contrast, Suszka-Świtek et al. (36) reported that administration of GnRH analogs (buserelin and cetrorelix) to female rats primarily increased serum PNX levels (significantly in some groups, non-significantly in others). Notably, their study focused on normal adult female rats, whereas the present study focused on male children with CPP. Moreover, they identified the ovary as the primary source of circulating PNX in female rats, while males lack this ovarian-derived PNX secretion mechanism. In addition, this difference could be due to the different expression regions and activation pathways of PNX-20 and PNX-14. PNX-20 is mainly expressed in the hypothalamus, while PNX-14 is mainly expressed in the spinal cord, heart, and other peripheral tissues (16). Previous in vitro studies have shown that PNX-20 activates the cAMP/PKA pathway by acting on GPR173 and stimulates the expression of GnRH and Kiss1 genes through cAMP response element binding protein (18). The level of PNX-20 decreases with the inhibition of the gonadal axis, which can be used as a potential serum marker to evaluate the response of CPP treatment, while the clinical value of PNX-14 remains to be further verified.

However, our study has two limitations. First, the sample size is relatively small, and it is a single-center design. In the future, we plan to expand the sample size to validate these findings and conduct animal and cell experiments to further explore the role of PNX in CPP. Second, due to the cross-sectional design of this study, we were unable to verify a causal relationship between PNX and puberty onset. In addition, our goal is to evaluate the feasibility of using serum PNX-20 levels as a predictor after GnRHa treatment.

Conclusion

This study is the first to investigate serum PNX levels in boys with CPP and healthy prepubertal boys. We found that serum PNX levels were significantly elevated in boys with CPP and positively correlated with BMI. We speculate that PNX may promote LH and testosterone secretion by activating the kisspeptin-mediated GnRH signaling pathway. The effects of PNX-20 may depend on the activation of p38/RUNX, which subsequently activates the GPR173 receptor, promoting skeletal maturation. Meanwhile, PNX-14, via a cAMP/Epac-dependent metabolic pathway, is associated with adiposity and may participate in the initiation of puberty. Due to its low sensitivity and specificity, it should be considered only as an auxiliary diagnostic marker for CPP rather than a standalone diagnostic marker. In addition, serum PNX-20 may serve as a serological marker for evaluating the efficacy of CPP treatment. Our study shows that PNX is associated with puberty onset; however, the causal relationship between the two needs to be verified through a large number of experiments.

Declaration of interest

The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.

Funding

This work was supported by the First Affiliated Hospital of Guangxi Medical University starting fund for study abroad returnees (Grant No. 2010219), and the Guangxi Clinical Research Center of Pediatric Disease (No. AD22035219).

Author contribution statement

T Xie, W Qin, and D Zeng analyzed the results and drafted the manuscript. T Xie, W Qin, and R Wang completed the experimental part of this study and summarized all the information. D Zeng and Y Chen performed the specimen collection. D Lan participated in study design and data analysis, and modified the manuscript. All authors approved the final version.

Ethical approval

This study was conducted in strict accordance with the Declaration of Helsinki and was approved by the Ethics Committee of the First Affiliated Hospital of Guangxi Medical University (Approval No. 2025-E0503). Written informed consent was obtained from the parents or legal guardians of all participating children before enrollment, in compliance with international ethical standards for research involving human subjects.