This study utilized data from the EMSNGS project (Chinese Clinical Trial Registry: ChiCTR2100051722). This large-scale longitudinal study was initiated in 2021 by Shenzhen Children's Hospital and Shenzhen Chronic Disease Control Center. Employing random stratified sampling, the project included one elementary school, one junior high school, and one senior high school from each of Shenzhen's 10 districts. A total of 27 schools were sampled, including three institutions encompassing both elementary and junior high divisions. One class was randomly selected from each school per grade level, excluding classes with fewer than 40 students. The final sample comprised 112 classes with 4,794 participants who completed comprehensive physical examinations and medical assessments.

A schematic representation of the overall study design is shown in Figure 1. The inclusion criteria were: ① participation in the current epidemiological survey with signed informed consent; ② age 6–17 years (under 18 years old on the physical examination date); and ③ availability of complete study-related data.

The exclusion criteria were: ① individuals with severe organ dysfunction, endocrine/metabolic disorders (e.g., growth hormone deficiency, thyroid dysfunction, hematologic/oncologic diseases, chronic hepatic/renal diseases); ② prolonged use of medications or supplements known to significantly affect growth or metabolism (e.g., growth hormone, glucocorticoids, immunosuppressants, weight-altering supplements); ③ physical disabilities or missing anthropometric data (height, weight, and WC); ④ incomplete demographic data (age, sex); ⑤ sexual developmental abnormalities: precocious puberty (secondary sexual characteristics appearing before age 8 in girls or 9 in boys; menarche <10 years), or delayed puberty (absence of secondary sexual characteristics by age 13 in girls or 14 in boys); and, ⑥ psychomotor developmental delay.

Written informed consent was obtained from legal guardians, and additional independent consent was obtained from high school-aged adolescents. Ethical approval was granted by the Shenzhen Ethics Committee (Approval No. 2021105). This study was registered in the Chinese Clinical Trial Registry (registration no. ChiCTR2100051722) and strictly adhered to the Declaration of Helsinki.

Trained project investigators conducted standardized anthropometric and developmental assessments using uniform measuring instruments. Data collected included height, weight, WC, and pubertal development indicators (e.g., testicular volume in boys and breast development stages in girls).

Height and weight measurement: Height and weight were measured using the Shanghe Ultrasonic Height-Weight Meter (SH-200, Zhengzhou Shanghe electronic Technology company, Zhengzhou City, China), with a precision of 0.1 cm and 0.1 kg, respectively.

Waist circumference (WC) measurement: A standardized leather circumference tape (National Institute of Sports Science) with a maximum range of 200 cm and accuracy of 0.1 cm was used. The tape was positioned horizontally at the midpoint between the 12th rib at the mid-axillary line and the highest point of the iliac crest. Measurements were taken at the end of normal expiration and recorded in centimeters (accurate to 0.1 cm). Two measurements were performed; if the difference exceeded 1 cm, a third measurement was performed. The average of the two closest values was calculated. The leather tape was regularly calibrated before and during the survey against a non-stretch steel ruler and was replaced with a new one after every 50 participants to minimize potential stretching.

Hip circumference (HC) measurement: The same leather tape was applied horizontally around the most prominent posterior point on the buttocks. Measurements were recorded at the end of normal expiration (accurate to 0.1 cm), following the same two-measurement protocol as that for WC.

Assessment of pubertal development: All examiners received standardized training from the Endocrinology Department of Shenzhen Children's Hospital. The measurement process was strictly supervised on-site by Pediatric endocrinologists at the same institution. Pubertal stages were evaluated according to the Tanner criteria. In girls, the breast development stage (B1–B5) was determined through visual inspection and palpation. Bilateral assessments were recorded, with the more advanced stages used for the final classification. In boys, testicular volume (mL) was measured by palpation and compared using a Prader orchidometer. The larger volume was selected if asymmetry existed. Additionally, the participants were interviewed regarding pubertal milestones, which comprised menarche status and age at menarche (if applicable) in girls and spermarche status and age at spermarche (if applicable) in boys.

Blood pressure measurement: Blood pressure was measured by using an upper-arm electronic sphygmomanometer (Omron HEM1320, Omron, Dalian, China). Before measurement, the subjects were instructed to rest quietly for 15–30 min in a temperature-controlled, quiet environment; avoid strenuous activity; and empty their bladders. Two consecutive measurements were recorded. If the difference between two readings exceeded 5 mmHg, or above the hypertension cutoff value (P95th for age, sex, height), additional measurements were performed until consistency (within 5 mmHg) was achieved.

All study participants fasted for 10 h (no food or water intake) before venous blood collection in the morning. Serum samples were separated by centrifugation and analyzed using a fully automated biochemical analyser (Beckman Coulter AU5811). Fasting blood glucose (FBG), high-density lipoprotein cholesterol (HDL-C), and triglyceride (TG) levels were measured.

Assessment of weight status: BMI was calculated as weight in kilograms divided by height in meters squared (kg/m²). Weight status was categorized into four groups based on sex- and age-specific BMI percentiles (14, 15): wasting: BMI <5th percentile for sex and age; normal weight: BMI ≥5th percentile but <85th percentile for sex and age; overweight: BMI ≥85th percentile but <95th percentile for sex and age; and, obesity: BMI ≥95th percentile for sex and age.

Pubertal Staging: Pubertal development was assessed using the Tanner staging system. Breast development stage was documented in girls, while in boys, testicular volume (mL) was measured via palpation using a Prader orchidometer. Based on secondary sexual characteristics, participants were categorized into three groups: prepubertal stage (Tanner stage I), early-middle pubertal stage (Tanner stages II-III), or late pubertal stage (Tanner stages IV-V). The prepubertal stage (Tanner I) was defined as the absence of signs of sexual maturation in both sexes, while the pubertal stage (Tanner II-V) was defined as: testicular volume ≥4 mL or spermarche in boys, and breast development ≥ stage B2 or post-menarche in girls (16, 17).

Definition of cardiometabolic risk: Cardiometabolic risk was defined using the CMRI which defined as follows: CMRI z-score = WC z-score + SBP z-score + DBP z-score + GLU z-score + TG z-score – HDL-c z-score. In this study, sex- and age-specific means and standard deviations (SD) for each CMRI component were established using a reference population with BMI between the 5th and 95th percentiles. Age- and sex-adjusted CMRI values were computed based on these standardized parameters. Higher CMRI values indicated an elevated cardiometabolic risk. Increasing CMRI values indicated heightened cardiometabolic risk, using z-score ≥1 as the threshold for high risk. The threshold of CMRI z-score ≥1 was chosen to define “high cardiometabolic risk” (18–20) as it represents a value one standard deviation above the mean of the reference population (children with normal BMI). This cutoff is statistically conventional for identifying individuals at increased risk within a normally distributed continuous variable and has been employed in prior studies to define elevated cardiometabolic risk clusters.

The calculation formulas of these anthropometric indices:

BMI: Weight (kg)/Height (m²).

Waist-to-Height Ratio (WHtR): WC (cm)/height (cm).

Waist-to-Hip Ratio (WHR): waist circumference (cm)/hip circumference (cm).

ABSI (7): WC (m)/(BMI (kg/m²)^2/3 × height (m)^1/2).

BRI (6): 364.2 – 365.5 × (1– (WC (m)/2π)^2/(0.5 × Height (m)²)^1/2.

All measurements were conducted by a unified survey team following standardized training. All instruments were calibrated prior to testing, and random quality checks were performed during data collection. Post-survey data verification was conducted daily with immediate retesting for outliers. A centralized database was developed by the Shenzhen Chronic Disease Control Center (SCCDC). District-level teams completed the data entry, followed by cross-validation and feedback from the SCCDC. Final data reconciliation was performed by the district-level chronic disease agencies. All data underwent an independent double-entry verification protocol, with dual researchers crosschecking and inputting the records, followed by database construction using EpiData software (version 3.0; EpiData Association, Odense, Denmark).

Data were analyzed using SPSS version 26.0 (IBM, Armonk, NY, USA). Normally distributed continuous variables were expressed as mean ± standard deviation (SD), whereas non-normally distributed variables were presented as median and interquartile range. Independent sample t-tests were used for normally distributed data, non-parametric tests for skewed data, and chi-square tests for categorical comparisons. Sex-specific determinants of ABSI and BRI were analyzed using quantile regression. Quantile regression models were applied at the 5th, 25th, 50th, 75th, and 95th percentiles to capture the full distribution of BRI and ABSI. Covariates including pubertal stage and weight status were adjusted based on prior evidence of their confounding effects. Sex-specific percentile references for the BRI and ABSI were established based on the quantile regression models. Participants were categorized into four groups according to BRI, ABSI, WHtR, WHR, and BMI quartiles (Q1 < 25th, 25th ≤ Q2 < 50th, 50th ≤ Q3 < 75th, and Q4 ≥ 75th). Multivariate binary logistic regression was employed to calculate odds ratios (ORs) for the association of BRI, ABSI, WHtR, WHR, and BMI with high cardiometabolic risk (CMRI ≥1). The second quantile (Q2) was chosen as the reference group for logistic regression analyses because the first quantile (Q1) may include individuals with underweight or atypical body composition, who are not an ideal “low-risk” reference group for cardiometabolic outcomes. The diagnostic performance of novel obesity indices (BRI, ABSI) in identifying high cardiometabolic risk (CMRI ≥1) was evaluated using receiver operating characteristic (ROC) curve analysis. The optimal cutoff values were determined by maximizing the area under the curve (AUC), sensitivity, and specificity (i.e. the highest Youden index).

The study initially sampled 5,348 individuals, of whom 5,184 participated in the survey (consent rate: 96.9%). Valid responses were obtained from 5,153 participants (response rate: 96.4%); ultimately, 4,794 participants were included in the analysis (2,656 boys and 2,138 girls). Of these, 1,971 participants (1,131 boys and 840 girls) had a CMRI ≥1 (Table 1). The proportion of boys and girls with high CMRI scores gradually increased with the progression of BMI and pubertal development (Table 2).

Multivariate quantile regression analyses revealed significant associations between BRI and pubertal staging/weight status in girls and boys, whereas the ABSI showed correlations exclusively with weight status (Tables 3–6).

Quintile regression analysis (BRI): For girls, the regression coefficients for normal weight (P5th–P95th), overweight (P5th–P95th), obesity (P5th–P95th), and puberty (P5th–P50th) were statistically significant (P < 0.05; Table 3). For boys, the regression coefficients for normal weight (P5th–P95th), overweight (P5th–P95th), obesity (P5th–P95th), and puberty (P5th–P75th) were significant (Table 4).

Quintile regression analysis (ABSI): In girls, significant associations were observed across multiple categories: normal weight (P5th–P95th), overweight (P5th–P95th), obesity (P5th), and puberty (P5th–P25th, P95th) (Table 5). For boys, the regression coefficients for normal-weight (P5th–P75th), overweight (P5th–P25th), and obesity (P50th–P75th) were statistically significant (Table 6).

Pediatric reference percentiles for BRI/ABSI were derived from quantile regression analyses of local normal-weight cohorts segregated by sex.

Sex-differentiated BRI percentile norms were established by segmenting the male and female cohorts by pubertal developmental phase (prepubertal/pubertal) (Tables 7, 8) based on a normal-weight cohort.

Sex-differentiated ABSI percentile norms were established by analyzing male and female cohorts separately. The ABSI standard was constructed based only on the normal-weight cohort (Tables 7, 8).

Using the presence of high cardiometabolic risk (defined as CMRI ≥1) as the dependent variable, novel obesity indices (BRI and ABSI) were analyzed as independent variables, and the participants categorized into quantile groups, with the second quantile group (Q2) serving as the reference. Girls and boys in the fourth BRI quantile group (Q4) exhibited significantly elevated cardiometabolic risk, with adjusted ORs of 6.540 (95% CI: 4.829–8.856) and 10.165 (95% CI: 7.621–13.561) compared to Q2, respectively. In contrast, ABSI showed no statistically significant association with CMRI ≥1 in either sex (P > 0.05). With the increase in the WHtR, WHR, and BMI, the presence of high cardiometabolic risk gradually increased (Figure 2).

The ROC curve analysis comparing obesity indices for detecting high cardiometabolic risk (CMRI ≥1) revealed that the 75th percentile of BRI demonstrated optimal diagnostic performance in both sexes. Compared to sex- and age-matched thresholds of WHR (WHR: P75th/P85th/P95th), WHtR (WHtR: P75th/P85th/P95th), and BMI (BMI: P75th/P85th/P95th), BRI-P75th achieved the highest Youden index (girls: AUC = 0.693 [95% CI: 0.670–0.716], sensitivity = 61.80%, specificity = 76.80%, Youden = 0.386, Table 9 and Figure 3; boys: AUC = 0.752 [95% CI: 0.732–0.771], sensitivity = 71.3%, specificity = 79.1%, Youden = 0.504, Table 10 and Figure 4).