Puberty is a complex process occurring between childhood and adulthood, producing internal and external physical changes that promote the development of primary and secondary sexual characteristics important for sexual reproduction (1). Sex hormones are responsible for the physical manifestations of female puberty, including thelarche, pubarche, and menarche (2). Menarche is the first menstrual bleeding and one of the most important events of female puberty. Over the last several decades, the mean age at menarche (AAM) has declined, which is concerning (3). Epidemiological evidence suggests that the onset of puberty is advancing in humans through undetermined mechanisms (4).

Environmental endocrine disruptors (EEDs) are chemicals that interfere with the endocrine system and evoke adverse developmental, immunological, neurological, and reproductive effects in humans (5). EEDs are common in human living environments and include pesticides, plasticizers, industrial by-products, drugs, and some naturally occurring phytochemicals (6). These exogenous chemicals can interfere with the complex endocrine system, causing adverse health effects, such as reproductive disorders, metabolic diseases, and various cancers (7). Chronic exposure to EEDs may play a role in accelerating or delaying the onset of menarche, and extensive research has shown that pesticides, phenols, polycyclic aromatic hydrocarbons, phthalates, and some heavy metals are responsible for hormone metabolism disorders occurring during puberty (8). Consequently, the effects of EEDs on fertility and reproduction are a growing concern in modern societies (9).

Recent genome-wide association studies (GWASs) have identified thousands of genetic variants associated with complex phenotypes and have provided insights into their genetic architectures. GWASs are also extremely well-suited for identifying common single-nucleotide polymorphism (SNP)-based variants (10). GWAS analysis has been conducted to research early puberty and identify the genetic characteristics of idiopathic central precocious puberty and validate the polygenic risk for early puberty (11). Felix R Day et al. have identified 389 independent, genome-wide significant signals for AAM, which explained ~7.4% of the population variance in AAM and corresponded to ~25% of the estimated heritability (11).

Genetic loci cause trait variations, ranging from growth and fitness in simple organisms to disease in humans. Determining the genetics of gene-expression differences has emerged as a key approach for linking DNA-sequence variations to phenotypes (12). Transcriptome-wide association study (TWAS) analysis has been used to identify significant expression-trait associations by integrating genotypes, gene-expression levels, and phenotypes in order to gain insights into the genetic basis of complex traits (13). In a recent study, a TWAS was performed to discover transcriptome differences that affect the age of natural menopause (ANM), and 34 ANM-associated genes were reported (14). The hypothalamic–pituitary–gonadal axis controls puberty and reproduction (15). Therefore, investigating the associated gene-regulation relationships may help identify important genes that are co-expressed in all tissue types during AAM.

Here, we aimed to determine the influences of genetic and environmental factors on AAM by performing a large-scale TWAS for AAM based on a GWAS data set. We investigated gene-expression levels in the hypothalamus, pituitary gland, ovaries, uterus, and whole blood. We also reevaluated the expression of TWAS-identified genes, functionally explored the genes, and identified AAM-associated EEDs.

Puberty is a developmental period affecting the body and behavior of children developing secondary sexual characteristics (29). Pubertal hormones play important roles in the adrenal, gonadal, and growth axes (30). Pubertal timing is determined by the activation of pulsatile hypothalamic GnRH secretion when pituitary gonadotropin secretion and downstream gonadal maturation are also initiated (31). Thus, various tissue types are affected by AAM, including the hypothalamus, pituitary gland, ovaries, uterus, and whole blood.

Up to now, previous studies have identified several genes associated with AAM, these results are similar to our studies (Table 3). Here, we performed a comprehensive TWAS to evaluate the relationship between AAM and predicted genes found in the hypothalamus, pituitary gland, ovaries, uterus, and whole blood. We identified 11 genes whose genetically predicted expression was associated with AAM (P _TWAS ≤ 0.05), including 10 novel genes (RBM6, PILRB, CPSF1, INPP5E, MRPL43, TIPIN, FLYWCH1, EXOSC6, ADORA2B, and SPATA20) and one gene (HSD17B12) linked to AAM in a previous GWAS. The 11 protein-coding genes identified in our study have been implicated in estrogen metabolism. We also identified enriched fatty acid-metabolism pathways, similar to the findings of a recent study (36), indicating that different types of fatty acids may influence puberty timing. Additionally, individual fatty acids might have different physiological and metabolic effects, including the progression of pubertal development. Therefore, our study provides new information that improves our understanding of the genetics and etiology of AAM.

HSD17B12 encodes 17β-hydroxysteroid dehydrogenase (17β-HSD), which is crucial for converting estrone into estradiol and fatty acid elongation. Our results support those of Kemiläinen et al. (37), who found that HSD17B12 played a vital role in female fertility through arachidonic acid metabolism. Specifically, an investigation of the 17β-HSD enzyme expression in human and mouse ovaries revealed that female HSD17B^+/- mice more often had frequent dysfunctional oogenesis and ovulation, leading to less frequent births. Under extreme circumstances, haploinsufficiency of the HSD17B12 gene in female mice resulted in subfertility. HSD17B12 can affect fatty acid elongation and ceramide accumulation in the serum (38). Previous data showed that central ceramides participate in the timing of female puberty (39), suggesting that HSD17B12 could be a candidate gene for female puberty. Notably, HSD17B12 was also related to endometrial (40), ovarian (41), and breast (42) cancers.

Previous results have shown that INPP5E is concentrated at the cilia base, where it helps control phosphoinositide metabolism (43). Primary cilia are present on virtually all cell types. Cilia appear to exert a crucial modulatory role in appropriate axonal wiring due to INPP5E-dependent activation of the PI3K–AKT signaling pathway, triggering an axonal Ca²⁺ wave (44). GnRH neurons in adults are multi-ciliated, and the percentage of GnRH neurons possessing multiple Kiss1r-positive cilia increases during puberty, correlating with sexual maturation (45). ADORA2B encodes an adenosine receptor and is involved in axon elongation. A previous report showed that ADORA2B transcripts were significantly downregulated in GnRH neurons during proestrus (46), while it has not been studied in pituitary. Therefore, these newly defined genes may influence the electrophysiology of puberty in hypothalamus and pituitary gland.

TIPIN is part of the replisome complex and binds the replication fork-protection complex TIMELESS, which is involved in circadian rhythm regulation (47). Circadian rhythms are well known to play key roles in animal reproduction (48). The existing body of research suggests that during the chronotype and the circadian timekeeping system change during the puberty, including endogenous rhythm period and sensitivity to environmental time cues (49). The results of several studies have established that melatonin promotes follicle-stimulating hormone in the pituitary gland and increases serum estrogen levels, thereby accelerating the onset of puberty (50). Those studies indirectly support a potential role of TIPIN in AAM.

CPSF recognizes the AAUAAA signal in pre-mRNA and interacts with other factors to facilitate RNA cleavage and the poly(A) addition. CPSF1 is the largest subunit of the CPSF complex. Previous data established that CPSF1 may promote ovarian cancer (51), cell proliferation, and triple-negative breast cancer (52). However, women have a higher risk for developing autoimmune diseases than men, which is attributed to sweeping endocrinological changes during puberty that considerably affect the immune system (53). CPSF1 was previously found to have a high affinity for HLA molecules (54). Thus, CPSF1 may be linked to the immune response in adolescent girls.

We extended the well-established GSEA approach to detect associations between environmental chemicals and AAM using published GWAS summary data sets and identified 120 chemicals, including drugs, organic compounds, inorganic compounds, plant extracts, nutrients, phenols, plasticizers, pesticides, herbicides, pollutants, and heavy metals. Aromatase is a rate-limiting enzyme in the conversion of androgens to estrogens, and letrozole is a selective aromatase inhibitor (55). Letrozole has been used to treat McCune–Albright syndrome, which has been associated with precocious puberty in girls (56) and delayed growth and puberty in boys (57). Further, fluoxetine is drug that is commonly used to treat adults with depressive disorders (58). The effect of fluoxetine on puberty has been a controversial issue. The results of one study demonstrated that fluoxetine exposure in utero delays puberty onset in female rats (59), whereas other data suggested that fluoxetine exposure during gestation did not alter plasma estrogen concentrations in peripubertal offspring (60).

Some plant extracts act as EEDs, as they are among a group of secondary metabolites with chemical structures similar to those of endogenous hormones. Isoflavones are phytoestrogens that are mainly produced in soybeans and can promote advancement of the vaginal opening in female rats after exogenous supplementation (61). Galangin is a naturally occurring flavonoid that inhibits the effects of flavonoids on human cytochrome P450 (62). Galangin was found to inhibit the aryl hydrocarbon receptor and is considered a potential drug for treating breast cancer (63, 64). Notable health benefits have also been reported for galangin (65), and previous data suggest that galangin is useful for treating precocious puberty. Zearalenone (ZEN) is a non-steroidal mycoestrogen that can exert adverse endocrine effects in mammals (66). To date, there is little agreement regarding the effect of ZEN on puberty. One group found that ZEN might trigger central precocious puberty (CPP) development in girls (67), whereas another group found an association between ZEN and normal pubertal development in adolescent girls (68).

A growing body of literature has associated being overweight or obese with early puberty (69); overweight and obese girls undergo menarche earlier than normal-weight girls (70). Several cross-sectional data have suggested that a substantial proportion of girls with CPP have hypertriglyceridemia (71). Cholesterol is an endogenous ligand of estrogen-related receptor alpha (72) that plays a role in activating estrogen receptors (73). Based on these data, we can infer that a high-cholesterol diet may also be a risk factor for precocious puberty. Plasticizers are additives used to produce or promote plasticity and flexibility in plastics, which are commonly used in everyday life. Further, parabens have been added to personal care products as antimicrobial preservatives (74), and methylparaben exposure increases glandular tissue sizes during critical developmental windows (75). A large survey showed that elevated phthalate metabolite levels pose potentially high health risks to children (76). In addition, monobutyl phthalate exposure might be associated with a risk for sexual precocity in girls (77). Animal experiments have demonstrated that two alkylphenols (4-nonylphenol and 4-tert-octylphenol) modify hormone biosynthesis and delay the onset of puberty (78).

This study has some limitations. First, the pooled GWAS data were obtained from the United Kingdom Biobank, and the study subjects were predominantly from European populations; these aspects affect the extrapolation of the results. Therefore, our results should be used cautiously when studying AAM in other populations. Second, although some of the genes we screened have been confirmed in other studies, our enrichment analysis used genes that did not pass correction for multiple testing (P <0.05), which may introduce some false positives so it requires caution when using our results. And some genes related to AAM susceptibility identified here have not been verified via molecular biology experiments, which should be performed in future studies. Further, some EEDs identified in this study were previously demonstrated to play a role in AAM, whereas others are not yet validated, which will require more clinical observations and cohort studies. However, to the best of our knowledge, this is the first large study using CGSEA to identify candidate EEDs related to AAM. Our TWAS analysis detected AAM-associated genes at the DNA level, and our CGSEA extended the classic GSEA approach to detect associations between environmental chemicals and AAM.

In this study, we aimed to determine the influences of genetic and environmental factors on AAM. Therefore, we performed a TWAS and CGSEA related to AAM and identified multiple AAM-associated genes and EEDs. The results of this study expand our understanding of the genetic and environmental factors affecting the timing of female puberty.

The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

Author ML and RF collected and processed the data, as well as wrote this article. YQ and HD provided language help and writing assistance. BL proofreaded the article. CY and YX designed the study. All authors contributed to the article and approved the submitted version.

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

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.