The pathogenesis of central precocious puberty (CPP) involves the combined effects of genetic susceptibility, exposure to environmental endocrine disruptors and nutritional/metabolic imbalances.1 Recent studies have found that vitamin D deficiency (VDD) and insulin resistance (IR) play significant roles in the onset of CPP.2,3 However, existing studies predominantly analyze the effects of vitamin D or insulin on CPP in isolation, overlooking the potential interactions and synergistic effects between the two. Previous studies have found that vitamin D can enhance insulin sensitivity by upregulating insulin receptor expression.4,5 Inhibited vitamin D-binding protein (VDBP) may also increase vitamin D bioavailability under IR conditions.6 This bidirectional regulation may be a source of bias when assessments are made using a single indicator.

Based on this, we conducted a study proposing the novel use of the VDIR as an integrated biomarker with the aim of overcoming the limitations of using a single indicator. The choice of the vitamin D/insulin ratio (VDIR) is based on the following theoretical foundations: First, the synergistic effects of vitamin D and insulin on glucose and lipid metabolism may impact the pace of gonadal development together.7 Second, the use of a ratio can minimize the effect of metabolic differences between individuals. This is particularly important in cases such as obese children, in whom VDD and compensatory hyperinsulinemia can distort results, leading to a risk of “false negatives”.8 Our hope was that establishing a dose-response relationship between the VDIR and the risk of CPP would allow for more accurate identification of high-risk groups.

After obtaining approval from the institutional ethics committee, we selected a total of 568 girls who underwent growth and development evaluations from the hospital’s electronic health record system. Based on the predefined inclusion and exclusion criteria, 232 participants were ultimately included, of which 139 had a diagnosis of CPP and 93 had normal growth and development. After matching controls for age, the final sample included 85 girls with CPP and 85 controls (Fig. 1).

Figure 1.

Flowchart of sample selection.

Table 1 presents the physical development and metabolic indicators for both groups. The physical development indicators (height, weight, BMI, BA, etc.) were higher in the CPP group compared to the control group. In the CPP group, the 25(OH)D level was lower (40.96 vs 51.41 nmol/L in the control group), and the proportion of VDD increased to 22.4% (compared to 8.2% in the control group). In addition, FIns levels (10.90 vs 8.72 nmol/L) and the HOMA-IR (2.61 vs 2.11) were significantly higher in the CPP group. However, there was not a statistically significant difference in the proportion of IR between groups (54.1% vs 41.2%; P = .091). Notably, the VDIR was significantly lower in the CPP group (3.45 vs 6.18).

The sex hormone levels and gonadal development indicators were significantly higher in the CPP group compared to the control group (Table 2). The CPP group had higher levels of E2, T, baseline LH, baseline FSH and peak LH and a higher peak LH/peak FSH ratio, while the difference in peak FSH values was not statistically significant. It is worth noting that the group of girls with CPP in our study included two subtypes: those who developed breast tissue before age 7.5 years and those who experienced menarche before age 10 years. In some cases, baseline LH levels had already reached pubertal levels, allowing for a clinical diagnosis of CPP without the need for a GnRH agonist stimulation test. Regarding gonadal development, uterine volume and bilateral ovarian volume were significantly greater in the CPP group compared to the control group (uterine volume, 4125.18 vs 1281.41; ovarian volume: left, 2.80 vs 2.00 and right, 3.11 vs 2.00). The proportion of girls with a non-linear endometrial midline pattern was higher in the CPP group (55.3% vs 24.7%), but there was no significant difference in endometrial thickness or the number of large follicles (>4 mm). Regarding secondary sexual characteristics, we found a significantly higher proportion in the CPP group of girls with breast development at Tanner stage B4 (43.5% vs 12.9%) and of menarche (44.7% vs 0%).

After adjusting for CA, BA, height, weight and BMI, we found that the VDIR was independently and inversely associated with the risk of CPP (OR = 0.814; 95% CI, 0.715−0.926) (Table 3). When we analyzed the association with gonadal development indicators, we found that each unit increment in VDIR was associated with a reduction in uterine volume of 233.54 mm3 (95% CI, −445.79 to −21.29). However, we did not find significant associations with the proportion of non-linear endometrial midline patterns, endometrial thickness, mean ovarian volume or the number of follicles with a diameter greater than 4 mm.

The ROC curve analysis (Table 4, Fig. 2) demonstrated that the VDIR offered the best diagnostic performance for the prediction of CPP, with an AUC of 0.731 (95% CI, 0.657−0.806), a sensitivity of 70.59% and a specificity of 67.06% (Youden index, 0.38). Although the combined indicator (Vitamin D combined with FIns) improved specificity to 78.82% and achieved a positive predictive value of 73.53%, its sensitivity was significantly reduced (58.82%). In contrast, vitamin D alone exhibited a higher sensitivity (76.47%), while the insulin level exhibited a higher specificity (78.82%), showing an imbalance between sensitivity and specificity.

Figure 2.

ROC curves of different indicators for diagnosis of CPP.

This case-control study systematically explored the correlation between the VDIR and CPP and the diagnostic value of the former. We found that VDIR levels were significantly lower in the CPP group compared to healthy controls, with a diagnostic AUC of 0.731 (sensitivity, 70.59%; specificity, 67.06%), outperforming vitamin D or FIns alone. Multivariate logistic regression further confirmed that VDIR was independently and inversely associated with the risk of CPP. These findings suggest that, by integrating the metabolic interactions between vitamin D and insulin, the VDIR may serve as a potential biomarker for CPP.

In recent years, multiple studies have indicated a close association between VDD, IR and the onset of CPP.15,16 Reduced vitamin D levels may upregulate leptin expression, activating the hypothalamic-pituitary gonadal (HPG) axis and accelerating pubertal onset.17,18 At the same time, insulin can stimulate the synthesis and release of GnRH in hypothalamic neurons and enhance the sensitivity of ovarian granulosa cells to gonadotropins, further promoting gonadal development.19,20 Our study found that the proportion of VDD was significantly higher in the CPP group compared to the control group, with concurrently elevated FIns levels and IR. This suggests that the imbalance between vitamin D and insulin may be a key mechanism driving CPP. Specifically, VDD exacerbates IR by downregulating insulin receptor expression, while IR inhibits VDBP, leading to a transient increase in free active vitamin D but also accelerating renal excretion.5,6,21 Ultimately, this forms a vicious cycle of VDD and IR. VDD and IR synergistically activate the HPG axis and ovarian function, jointly contributing to the abnormal advancement of puberty. Therefore, based on the bidirectional regulatory mechanism of vitamin D and insulin, and considering that metabolic interactions can easily influence a single indicator, we propose the VDIR as an integrated marker reflecting their synergistic imbalance.

Although there is a correlation between the VDIR and precocious puberty, it is important to note that children in the CPP group generally had higher body weights and BMIs. Obesity has been widely recognized as an independent risk factor for early pubertal onset.22 Obese children frequently also have metabolic alterations such as IR and VDD, which may serve as confounding factors underlying the association between VDIR and precocious puberty.23,24 In our study, we adjusted the regression analysis for potential confounders, including the BMI, and the VDIR still emerged as an independent factor associated with CPP. This suggests that the association may not be entirely driven by obesity, and the ratio-based nature of the VDIR may help offset individual differences in baseline metabolism (e.g., BMI and prepubertal hormonal fluctuations). However, the BMI may still influence the VDIR indirectly by altering levels of insulin and vitamin D substrates, which warrants further investigation into the underlying mechanisms in this association.

At the level of gonadal development, the negative correlation between the VDIR and uterine volume suggests that metabolic imbalance may directly impact gonadal maturation. In the CPP group, the peak LH/peak FSH ratio was elevated, and levels of sex hormones such as E2 and T were significantly increased, indicating excessive activation of the HPG axis. Although the VDIR was not significantly associated with ovarian volume or endometrial thickness, its relationship with uterine volume may reflect the interaction between metabolic signals (e.g., insulin-mediated ovarian IGF-1R pathways) and local sex hormone synthesis.25–28 Previous studies have found that decreased vitamin D levels may further increase uterine volume in girls with CPP by affecting circulating E2 levels.29 Similarly, animal model studies have shown that insulin signaling is crucial for the development of preovulatory follicles and endometrial proliferation.30 The synergistic effect of the imbalance between vitamin D and insulin may accelerate gonadal development, but the specific mechanisms underlying this need to be further elucidated.

The findings of our study suggest that the VDIR may have significant translational value in screening for CPP in clinical practice. Its high AUC and better balance of sensitivity and specificity make it significantly superior to single indicators, so it could serve as an ancillary screening tool for CPP. Based on the cut-off value, we would recommend integrating the VDIR into routine pediatric health checkups for early screening of children with suspected CPP (e.g., those with early breast development or advanced BA). In those with positive results, the evaluation can continue with a pelvic ultrasound. The VDIR holds promise as a clinical screening tool to help identify individuals that may be at high risk, offering a more comprehensive metabolic marker for determining the need of the GnRH agonist stimulation test. This may help optimize resource allocation and reduce unnecessary examinations. Since some CPP patients already meet the diagnostic threshold based on baseline LH levels, this study does not recommend using the VDIR as a substitute for the stimulation test, but rather as a preliminary risk assessment or stratification tool. In the case of high-risk groups (e.g., positive family history or rapidly progressing precocious puberty), the combined testing of vitamin D and FIns can increase the specificity to 78.82%, guiding clinical decision-making to reduce false positives.

However, there are certain limitations to this study. First, the sample was limited to a single-center cohort of girls, and further studies are required before generalizing its findings to the male population and other ethnic groups. Second, the cross-sectional design makes it difficult to establish causality or the temporal sequence in the relationship between the VDIR and CPP. Additionally, the sensitivity and specificity of the VDIR are still below the ideal threshold (>80% for both). Prospective multicenter studies are needed to refine the clinical translation pathway, establish a dynamic monitoring model based on the VDIR and explore the specific mechanisms by which the VDIR modulates the HPG axis using various techniques.

This study, which combines the dual pathological mechanisms of VDD and IR, reveals the potential impact of vitamin D-insulin interaction on CPP. It provides a more comprehensive risk assessment and screening tool for clinical practice while offering a theoretical basis for developing personalized intervention plans (such as vitamin D supplementation combined with insulin sensitivity regulation).

JS and HRZ wrote the manuscript, WHS and JBL participated in the literature search, JS and SQJ participated in the study design, YX and WW participated in the diagnosis and treatment of the patients.

The study was a retrospective analysis based on the participants’ historical health records and related information, and did not involve any prospective interventional treatments. It does not affect participants’ personal privacy or involve any commercial interests, and therefore received an ethical waiver for informed consent from the ethics committee.

The study is a single-center retrospective case-control study, approved by the Ethics Committee of the Third Affiliated Hospital of Zhengzhou University (approval no: 2024-119-01), and conducted in accordance with the principles of the Declaration of Helsinki.

The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.

The data supporting the results of this study can be obtained from the author, but the availability of these data is limited. However, the author can provide data with a reasonable request.