Highlights
· Girls with central precocious puberty show enhanced functional connectivity between right lateral hypothalamus and right superior frontal gyrus that borderline correlates with peak LH levels, while connectivity to fusiform/supramarginal gyri decreases.
· Left lateral hypothalamus displays increased coupling with right cuneus, revealing hemisphere-specific hypothalamic-cortical network alterations linked to central premature pubertal activation.
· These lateral hypothalamic connectivity shifts, absent in medial hypothalamus, provide novel neuroendocrine insights.
Introduction
Precocious puberty is clinically defined as the premature development of secondary sexual traits and somatic growth. This condition is categorized into 2 primary subtypes: central precocious puberty (CPP), resulting from premature activation of the hypothalamuspituitary-gonadal (HPG) axis; and peripheral precocious puberty (PPP), which develops without involvement of HPG axis regulation [1]. This disorder clinically presents with accelerated sexual maturation (diagnostic criteria: <8 years in females, <9 years in males). It is marked by abnormally early surges in gonadotropin levels followed by downstream activation of sex steroid pathways (estrogen/androgen synthesis) [2].
In diagnostic protocols, the gonadotropin-releasing hormone (GnRH) provocation test is widely recognized as the most reliable method for confirming CPP. Luteinizing hormone (LH) levels peaking above 5 IU/L during this procedure, owing to their elevated diagnostic accuracy, serve as a robust indicator of HPG axis activation [3]. Recent studies have demonstrated a rising incidence of CPP in girls [4], which coincides with increased reports of mental health concerns, such as anxiety and social adjustment difficulties, in this population.
CPP is associated with adverse outcomes, including risk-taking behaviors in adolescents, short stature and obesity in adulthood, and elevated risks of diabetes and premenopausal breast cancer in later life [5]. Furthermore, adolescents with CPP are more likely to experience psychological challenges, such as anxiety, depression, irritability, aggression, and medicine abuse, which may persist into adulthood [6-9]. The neuroendocrine mechanisms linking CPP to these physical and psychological consequences remain poorly understood.
Recent advances in resting-state functional magnetic resonance imaging (rs-fMRI) have enabled its application in neuroendocrine research by capturing spontaneous low-frequency blood oxygen level-dependent (BOLD) signal fluctuations [10]. Functional connectivity (FC) is typically assessed using Pearson correlation between seed-region time series and whole-brain voxels. Given the central role of the hypothalamus in endocrine and behavioral regulation, it has attracted particular interest in CPP research [11]. However, owing to its small size and complex subnuclear anatomy, careful region of interest (ROI) definition is necessary, such as the bilateral lateral and medial subregions described by Baroncini et al. [12,13].
Studies have revealed distinct FC alterations in CPP. In a CPP group compared to a PPP group, Chen et al. [14] reported reduced FC between the bilateral insulae and the right middle frontal gyrus (MFG), as well as between the left fusiform gyrus (FFG) and the right amygdala. Higher basal and peak estradiol (E2) levels correlated with increased resting-state FC of the bilateral insulae. Qin et al. [15] revealed increased FC in CPP subjects between the right posterior cerebellar lobe and the bilateral medial prefrontal gyri/posterior cingulate cortex. Strengthened connectivity was also observed between the left cerebellum posterior lobe and the bilateral MFG and bilateral precuneus. Concurrently, they observed attenuated FC between the left superior occipital gyrus and right calcarine.
Previous studies have identified functional abnormalities in multiple brain regions of CPP patients, including the frontal lobe, FFG, insula, posterior cerebellum, and parietal lobe. However, these findings were derived from whole-brain FC analyses, which lacked targeted investigation of hypothalamic involvement in CPP. The present study addressed this gap by focusing on the hypothalamus as the primary ROI. We investigated hypothalamic FC patterns in girls with CPP and analyzed their associations with hormone levels to elucidate the neuroendocrine mechanisms underlying the disorder.
Methods
1. Patient information acquisition
We recruited 106 girls with suspected precocious puberty from the Pediatric Development Clinic at the Second Affiliated Hospital of Anhui Medical University. This study was approved by the Ethics Committee of the Second Affiliated Hospital of Anhui Medical University (YX2022-107[F1]). Participants were enrolled if they exhibited signs of puberty (e.g., breast development, increased axillary/pubic hair) before age 8. Exclusion criteria included (1) magnetic resonance imaging (MRI) contraindications; (2) left-handedness; (3) menarche; (4) history of traumatic brain injury; (5) intracranial spaceoccupying lesions; and (6) use of neuroactive medications. The research received ethical approval from institutional review boards, with documented parental/guardian consent and participant assent obtained in compliance with pediatric research protocols. Following inclusion/exclusion criteria, 49 girls were excluded, resulting in 57 participants (37 CPP, 20 non-CPP). No participants underwent MRI scanning during or after GnRH agonist treatment. Demographic and clinical data were collected for all participants, including age, height, weight, blood pressure, bone age, parental heights, and pituitary height (PH). PH was defined as the perpendicular distance from the midpoint of the anterior pituitary gland to the hypophysial fossa (expressed in mm). It was measured on midsagittal 3-dimensional (3D) T1-weighted structural images by raters blinded to participants’ clinical information.
2. Acquisition of hormones associated with GnRH stimulation experiments
Following overnight fasting, venous blood samples were collected from all participants. Baseline LH and follicle-stimulating hormone (FSH) levels were obtained using chemiluminescence-based immunoassays. Subsequent measurements of peak LH/FSH concentrations and their proportional ratios were recorded after GnRH stimulation. The GnRH stimulation protocol involved subcutaneous administration of gonadorelin (2.5 μg/kg, maximum dosage 0.1 mg). Serial blood collection was performed preinjection and at 30-, 60-, and 90-minute intervals postadministration for LH/FSH level determination, with maximal hormone concentrations identified from these temporal measurements. Poststimulation LH concentrations reaching ≥5 IU/L are internationally recognized biomarkers of HPG axis activation [3,16]. Subjects underwent classification according to post-provocation LH maxima: the CPP group (n=37) exhibited LH peaks meeting diagnostic thresholds (≥5 IU/L) with confirmed HPG axis activity; the non-CPP group (n=20) showed peak LH levels below the diagnostic threshold (<5 IU/L), consistent with suspected CPP and a quiescent HPG axis.
3. Acquisition of MRI data
Neuroimaging was conducted using a 3.0 T Siemens MRI system (eight-channel head coil), with participants maintaining a supine position, eyes-closed resting state, and head immobilization via foam padding during acquisition. Earplugs were provided to attenuate scanner noise. Postscan image quality and the absence of intracranial abnormalities were assessed independently by 2 radiologists with expertise in neuroimaging. The 3D T1-weighted structural image parameters were the following: repetition time (TR)=1900 msec, echo time (TE)=2.98 msec, slice thickness=1 mm (no gap), 176 slices, voxel size=1 mm×1 mm×1 mm, and field of view (FOV)=256×256 mm². The resting-state functional MRI (fMRI) parameters were as follows: TR=2000 ms, TE=25 ms, flip angle (FA)=90°, slice thickness=4.0 mm, matrix=64×64, and FOV=240 mm×240 mm. For each participant, a total of 240 volumes were obtained.
4. fMRI data preprocessing
Preprocessing of rs-fMRI data for 37 girls with CPP and 20 non-CPP girls was performed using Statistical Parametric Mapping v12 (SPM12) and Data Processing & Analysis for Brain Imaging (http://rfmri.org/dpabi) for voxel-wise statistical analysis and intrinsic connectivity network computation, respectively. The preprocessing steps were as follows: (1) The initial volumes were discarded: The first 10 time points were excluded to account for scanner stabilization and participant acclimation to the scanning environment. (2) The slice timing was corrected: The temporal discrepancies between slice acquisitions were corrected to align all slices to a common time point. (3) Motion was corrected: The exclusion criteria encompassed fMRI datasets demonstrating head movement exceeding 3.0-mm displacement or 3.0° angular rotation across any directional axis. The spatial normalization procedures involved 3 sequential computational steps: (1) The individual neuroimaging datasets underwent rigid-body alignment to the Montreal Neurological Institute stereotaxic template. (2) Tissue segmentation partitioned the brain volumes into distinct gray matter, white matter, and cerebrospinal fluid (CSF) compartments. (3) Nuisance covariates derived from white matter and CSF signals were computationally regressed. For spatial preprocessing, a 6-mm full width at half maximum Gaussian filter was implemented to attenuate high-frequency noise components while enhancing hemodynamic response specificity. Temporal filtering restricted the BOLD signal oscillations to the 0.01- to 0.08-Hz frequency bandwidth to isolate intrinsic neural fluctuations. Linear detrending algorithms were subsequently applied to eliminate scannerinduced signal drift artifacts.
5. FC analysis
In this study, a ROI analysis approach was employed. In accordance with prior methodologies [13], the hypothalamus was subdivided into lateral and medial regions, resulting in 4 ROIs (left/right hemispheres; detailed in Table 1). For each region, spherical seed ROIs with a 2-mm radius were used for seed-to-voxel whole-brain analysis. Individual FC maps were generated by computing Pearson correlations between the time course of each seed region and the time courses of all whole-brain voxels. These subject-specific FC maps were then Fisher z-transformed to produce z-value maps for subsequent group-level analyses.
6. Statistical analysis
The fMRI data analysis was conducted with SPM12. Two-sample t-tests were conducted to examine the differences in FC between the CPP and non-CPP groups, with age as a covariate. Multiple family-wise error comparison corrections were performed in the FC analysis. The voxel-wise threshold was set to P<0.001 and the cluster threshold was set to >44 voxels, which corresponded to a corrected P<0.05. The FC patterns were mapped using BrainNet Viewer.
Statistical processing of demographic and clinical parameters was conducted with IBM SPSS Statistics ver. 23.0 (IBM Co., USA). GnRH provocation test-derived hormonal datasets underwent logarithmic transformation (base 10) for assessing associations with FC. Analyses of covariance were performed for the FC comparisons and correlation analyses. The Shapiro-Wilk method was employed to evaluate the data distribution normality. Parametric data were processed through independent samples t-tests, while the nonparametric datasets were assessed using the Mann-Whitney U-test. Quantitative parameters demonstrating normal distribution were expressed as arithmetic mean±standard deviation, whereas the nonnormally distributed variables were characterized by median values accompanied by the interquartile range. Bivariate associations between FC indices and endocrine parameters were evaluated using Pearson correlation coefficient for normally distributed data or Spearman rank correlation analysis for skewed distributions. To further enhance data normality, baseten logarithmic transformation was implemented for hormonal concentration values. The statistical significance threshold was defined as 2-tailed P-values below 0.05.
Results
1. Results of demographic analysis of the 2 groups
Compared to the non-CPP group, girls with CPP exhibited a significantly higher chronological age (P=0.008) and bone age (P=0.022), along with a lower PH (P=0.016). No statistically significant differences were observed in the other demographic or clinical characteristics, including height, weight, blood pressure, maternal height, or PH. Table 2 summarizes the demographic characteristics of all participants.
2. Results of analysis of hormone levels associated with GnRH stimulation experiments
Compared to the non-CPP group, the CPP patients exhibited significantly higher baseline FSH (P<0.001), baseline LH (P<0.001), peak FSH (P=0.028), peak LH (P<0.001), and peak LH/FSH ratios (P<0.001). Significant differences persisted after log10 transformation for baseline FSH (P<0.001), baseline LH (P<0.001), peak FSH (P=0.028), peak LH (P<0.001), and peak LH/FSH ratios (P<0.001). Table 3 summarizes the statistical outcomes of the hormonal profiles obtained via the GnRH stimulation test.
3. Results of FC analysis between hypothalamus and brain based on hypothalamic ROI coordinates in girls with CPP
As depicted in Supplementary Table 1 and Fig. 1, the CPP group exhibited greater FC between the right lateral hypothalamus and right superior frontal gyrus (SFG) relative to the non-CPP group. In addition, the CPP group showed diminished FC linking the right lateral hypothalamus to both the right FFG and right supramarginal gyrus (SMG). Meanwhile, the CPP group displayed amplified FC between the left lateral hypothalamus and the right cuneus (CUN). However, FC analyses revealed no significant differences between the medial hypothalamus and other regions.
4. Correlation results of hypothalamic FC and GnRH excitation test in CPP girls
Fig. 2 shows that elevated peak LH concentrations exhibited borderline significant correlation with enhanced FC between the right lateral hypothalamic and the right SFG (correlation coefficient [r]=0.362, P=0.050) in the CPP group. The results of correlation analyses of analysis of covariance were the following: age (r=0.224, P=0.122), bone age (r=0.158, P=0.216), and father’s height (r=0.216, P=0.139). Under this analytical approach, no significant correlations were observed between FC and the other parameters.
Discussion
The experimental data revealed that CPP-diagnosed girls demonstrated elevated baseline FSH and LH concentrations, along with heightened FSH peaks, LH peaks, and LH/FSH ratio peaks compared with agematched controls of girls without CPP. Enhanced FC was detected between the right lateral hypothalamic area and the right SFG in the CPP subjects, exhibiting borderline significant correlation with peak LH concentrations. In contrast, diminished connectivity patterns emerged between the right lateral hypothalamus and the right FFG/SMG. However, greater connectivity was evident between the left lateral hypothalamic region and the right CUN. Comparative analysis of medial hypothalamic connections demonstrated no statistically significant alterations in FC across examined cerebral areas.
Our findings highlight the critical association between CPP and the lateral hypothalamus, a region integral to the regulation of motivation, metabolism, motor function, and social behavior. The left lateral hypothalamus is implicated in learning, sustained motivation, reward processing, and aggression suppression [17-19], while the right lateral hypothalamus modulates motor control, interference inhibition, and social interaction [17-20]. In the CPP subjects, peak LH levels were borderline significantly correlated with FC in the right lateral hypothalamus. According to one report, androgen exposure reduces mean diffusivity in the lateral hypothalamus [21]. Furthermore, females with depression exhibit elevated lateral hypothalamic hypocretin-1 levels, with estrogen receptors being colocalized alongside hypocretin neurons in this region [22]. Additionally, estrogen-regulated kisspeptin neurons in the lateral septum project extensively to GnRH neurons [23], which could reflect a potential interplay between the lateral hypothalamus, sex hormones, and GnRH signaling in CPP pathogenesis.
Our studies found FC between the right lateral hypothalamic region and right SFG among the CPP subjects. The right SFG is involved in cognitive regulation, emotional processing, and behavioral functioning [24,25]. In the study of Yi et al. [24] the right SFG showed increased FC with the left middle occipital cortex. The right SFG also exhibited heightened FC with the bilateral superior occipital gyri. No similar associations were found for the left SFG. Investigations by Xie et al. [26] found increased amplitude of low-frequency fluctuations within the right SFG among individuals exhibiting HPG axis activation, with these neural oscillations demonstrating positive correlations with the prolactin concentrations. Yu et al. [27] observed reduced regional homogeneity (ReHo) values in the bilateral orbital and medial SFG of CPP girls. In their study, the ReHo values of right orbital pars of SFG positively correlated with baseline/peak E2 levels; however, it was inversely associated with perceptual reasoning and Wechsler Intelligence Scale scores. Recent studies have indicated that the lateral hypothalamus connects to the medial aspect of the SFG [28,29]. In CPP patients, FC between the right lateral hypothalamus and right SFG showed a trend-level correlation with peak LH levels (r=0.362, P=0.050). This association should be interpreted cautiously, as it suggests a potential, but not robust, link.
Meanwhile, experimental data revealed altered FC between the right lateral hypothalamic region and right FFG in CPP-affected females. Neuroimaging investigations have demonstrated the critical involvement of the FFG in facial recognition processes, semantic decoding, and nonverbal cognitive functions [30,31]. Neuroimaging comparisons demonstrated diminished FC between the left ventral anterior insula and ipsilateral FFG in somatic depression patients versus non-somatic depression cohorts [32]. Parallel investigations by Chen et al. [14] identified reduced left FFG gray matter density in CPP populations compared to PPP controls, coupled with attenuated connectivity between the left FFG and contralateral amygdala. Our results contrast with those of earlier studies by revealing decreased connectivity between the right FFG and right lateral hypothalamic zone in CPP subjects. This divergence from prior findings involving the left FFG raises the possibility that hemisphere-specific involvement of fusiform regions may contribute to CPP-related neural alterations.
Furthermore, experimental observations revealed impaired FC between the right lateral hypothalamic region and right SMG in CPP-affected female subjects. Anatomically situated within the inferior parietal lobular complex, the SMG is critically engaged in linguistic operations and affective information processing, as established in neuroimaging literature [33,34]. Complementary neurocognitive investigations by Steinbeis et al. [35] demonstrated heightened emotional egocentric bias tendencies in pediatric populations relative to adults. These tendencies were associated with diminished neural activation in the right SMG and compromised connectivity patterns with the left dorsolateral prefrontal cortex. Park et al. [36] observed increased gray matter volume along the SMG in patients with idiopathic CPP (ICPP). The diminished FC between the right SMG and right lateral hypothalamus in CPP girls could reflect right lateral hypothalamic dysregulation impacting the right SMG.
The present study identified increased FC between the left lateral hypothalamus and right CUN in girls with CPP. The right CUN is a key region for visual processing [37]. Adolescents with borderline personality disorder exhibit elevated ReHo in the right CUN [24], while ICPP patients show increased CUN matter volume [36]. The CUN functionally connects with the precuneus. Zhao et al. [38] identified genetic underpinnings of default mode network connectivity involving the precuneus and adjacent cortices. Girls with HPG axis activation exhibit reduced cortical thickness in the right precuneus, which inversely correlates with LH and E2 levels [39]. Conversely, CPP patients demonstrate greater precuneus cortical thickness than non-CPP peers, although both groups have shown age-related declines in precuneus volume and cortical thickness [40]. Enhanced function between the left lateral hypothalamus and the right CUN in CPP subjects may suggest left lateral hypothalamic modulation of right CUN activity. However, a significant limitation of this study and similar CPP neuroimaging studies is the absence of concurrent cognitive or behavioral outcome measures.
CPP is associated with distinct psychological and behavioral challenges. One study demonstrated that girls with precocious puberty often develop distorted perceptions of their body image and breast development, which correlate with elevated depression scores [41]. Adolescents with CPP exhibit greater anxiety and poorer body image [8]. CPP is also linked to increased emotional reactivity, higher heart-rate variability, and slower information processing speeds in adolescents [42]. Intriguingly, these mental health challenges relate to our findings of altered FC in the lateral hypothalamus, right SFG, right FFG, right SMG, and right CUN. While the altered FC patterns observed in this study may contribute to emotional or behavioral outcomes previously reported in CPP, such interpretations remain speculative. Future studies incorporating neuropsychological testing are warranted.
Several limitations warrant consideration when interpreting our findings. First, the cross-sectional design precluded longitudinal assessment of FC changes in CPP patients over time. Second, given the higher prevalence of CPP in girls, our study consisted exclusively of female participants, which limits the generalizability of our results to males. Third, the modest sample size of the CPP group (n=37) may have introduced bias and limited the statistical power of our analyses. We plan to increase the sample size in future research to further validate these findings. Fourth, some of the variables (bone age and father’s height) were not controlled in the FC comparisons; thus, our FC results must be used cautiously when extrapolating the data. Finally, while we linked alterations in FC to emotional, cognitive, and behavioral characteristics, the absence of reported neuropsychological assessments means these interpretations remain speculative in nature and lack direct empirical support. Consequently, these relationships should be considered hypotheses requiring future investigation rather than established conclusions. Despite these limitations, this study integrated hypothalamic ROI mapping and hormonal profiling to demonstrate altered hypothalamic FC in CPP subjects, providing new insights that hormonal dysregulation directly modulates hypothalamus–brain interactions.
In conclusion, abnormal lateral hypothalamus FC patterns in CPP girls, particularly linked to peak LH levels, offer new insights into the neuroendocrine mechanisms underlying CPP.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Full text via PMC
Download Citation
Supplement
Print
