The entire cohort comprised 174 prepubertal children (87 PA and 87 healthy controls) with a mean age at BIA measurement of 8.1 ± 1.2 years. Their collective SEP was similar and above average (median SEP cluster of 8, range 3–10). The PA group was significantly more likely to have a family history of cardiometabolic disease (obesity, hypertension, type 2 diabetes, dyslipidemia, cardiovascular disease) compared to the healthy controls (68.2% vs 45.9%, respectively, P = 0.003).

The pregnancy and perinatal characteristics of both groups are presented in Supplementary Table 1. The overall perinatal characteristics were similar for both groups, 26.4% of the PA children were conceived following assisted reproduction, 9.6% of their mothers had PCOS, and 8.2% PA children had in utero exposure to gestational diabetes mellitus. Around 80% of the entire cohort were born at term with an AGA birthweight, while in the PA group 15.7% were born SGA compared to 11.9% of the controls (P = 0.701).

The anthropometric characteristics and body composition parameters of the study group and their controls are presented in Table 1. The children in the PA group had a higher median height z-score compared to the controls (0.59 [IQR −0.08, 1.35] vs. −0.13 [IQR −0.79, 0.77], P < 0.001, respectively), greater mean delta height z-score (the subjects’ height z-scores minus their MPHt z-scores, 0.56 ± 0.94 vs. −0.14 ± 0.94, P < 0.001), and higher BMI z-scores (0.52 [IQR −0.22, 1.44] vs. −0.07 [IQR −0.93, 0.83], P = 0.001).

Table 1 Anthropometric characteristics and body composition parameters of the premature adrenarche group and their healthy controls.

The body composition of the children in the PA group was characterized by higher total fat percentage compared to the healthy controls (24.9 [IQR 21.7, 27.3] vs. 21.9 [IQR 20.2, 26.5], respectively, P = 0.006), higher truncal fat percentage (18.4 [IQR 15.8, 21.4] vs. 16.4 [IQR 14.3, 21.0], P = 0.027), higher truncal fat mass (3.1 [IQR 2.4, 4.1] vs. 2.5 [IQR 2.0, 3.7], P = 0.006), higher appendicular fat mass (3.8 [IQR 3.3, 5.4] vs.3.4 [IQR 2.7, 4.5], P = 0.003), and lower MFR z-scores (−0.82 [IQR −1.19, −0.28] vs. −0.38 [IQR −1.08, 0.06], P = 0.050). There were no group differences in the trunk-to-appendicular fat ratio or parent-reported physical activity levels, and both the study and control groups had negative correlations between their MFR z-scores and their trunk-to-appendicular fat ratios (r = −0.669, P < 0.001 and r = −0.468, P < 0.001, respectively).

The median basal androgen levels of prepubertal children with PA were as follows: DHEAS 1.55 [IQR 0.87, 2.51] µmol/L, androstenedione 0.52 [IQR 0.50, 1.50] nmol/L, testosterone 0.21 [IQR 0.20, 0.33] nmol/L, 17OHP 1.40 [IQR 0.90, 2.30] nmol/L and cortisol 244 [IQR 178, 317] nmol/L.

The MetS components of the study and control groups are listed in Table 2. There were no significant group differences in the rates of obesity, elevated BP, or elevated fasting glucose and dyslipidemia (elevated TG and low HDL concentrations). The mean HbA1c level of the PA group was 5.4 ± 0.3% (35.5 mmol/mol) and the mean HOMA-IR level was 2.4 ± 1.3; 66% of the PA children had elevated HbA1c and/or HOMA-IR levels.

Table 2 Metabolic syndrome components of the premature adrenarche group and their healthy controls.

The correlations between HOMA-IR and the following parameters reached levels of significance: BMI z-score (r = 0.604, P < 0.001), FATP (r = 0.564, P = 0.002), MFR z-score (r = -0.456, P = 0.017), and DHEAS concentrations (r = 0.664, P = 0.001). The correlation between the DHEAS concentrations and FATP was also significant (r = 0.251, P = 0.036). Birthweight z-scores, however, were not significantly correlated with BMI z-scores, FATP, MFR z-scores, or HOMA-IR levels and DHEAS concentrations. Laboratory characteristics are shown in Supplementary Table 2.

Subgroup analyses of children with pa stratified by weight status (Table 3)

Children with PA and overweight/obesity had higher height z-scores (P = 0.007), higher delta height z-scores (P = 0.022), higher fat percentages (P < 0.001), higher truncal fat mass (P < 0.001), higher truncal fat percentages (P < 0.001), higher appendicular fat mass (P < 0.001) and lower MFR z-scores (P < 0.001) compared to children with PA and healthy weight status. Comparative analysis of MetS components among children with PA and overweight/obesity revealed higher rates of elevated systolic BP (P < 0.001), dyslipidemia (low HDL, P = 0.006) and higher TG:HDL ratios (P < 0.001). No differences were found between PA subgroups in family history of cardiometabolic disease.

Table 3 Subgroup analyses within the premature adrenarche group according to weight status.Comparative analyses of children with healthy weight status (PA vs. control, Table 4)

The comparison between children with healthy weight in the PA group and those in the control group revealed that the PA group had higher height z-scores (P < 0.001), delta height z-scores (P < 0.001), higher fat percentages (P = 0.007), truncal fat mass (P = 0.009), truncal fat percentages (P = 0.031), and higher appendicular fat mass (P = 0.010). MFR z-scores tended to be lower for the healthy weight subgroup of children with PA compared to the control group, with no differences observed in the rates of MetS components.

Table 4 Subgroup analyses of children with healthy weight status: premature adrenarche vs. healthy controls.Comparative analyses of children with overweight/obesity (PA vs. control, Supplementary Table 3)

The comparison between children with overweight/obesity in the PA group and those in the control group revealed similar anthropometric and body composition parameters. Children with PA and overweight/obesity exhibited higher systolic and diastolic blood pressure levels and percentiles and greater rates of elevated blood pressure (57.1% vs. 21.1%, P = 0.014) compared to the control group, as well as higher TG:HDL ratios (P = 0.026).

Associations between body composition and MetS components (Table 5)In the PA group

FATP was positively correlated with the BMI z-score (r = 0.858, P < 0.001), systolic BP percentile (r = 0.486, P < 0.001), diastolic BP percentile (r = 0.244, P = 0.023), TG z-score (r = 0.358, P = 0.002), and TG:HDL ratio (r = 0.545, P < 0.001). The trunk-to-appendicular fat ratio was positively correlated with the BMI z-score (r = 0.245, P = 0.022), TG z-score (r = 0.302, P = 0.010) and TG:HDL ratio (r = 0.325, P = 0.006). The MFR z-score was negatively correlated with the BMI z-score (r = −0.816, P < 0.001), systolic BP percentile (r = −0.446, P < 0.001), diastolic BP percentile (r = −0.226, P = 0.035), TG z-score (r = −0.300, P = 0.010), and TG:HDL ratio (r = −0.476, P < 0.001).

Table 5 Associations between body composition and metabolic syndrome components of the premature adrenarche group and their healthy controls.In the healthy control group

FATP was positively correlated with the BMI z-score (r = 0.774, P < 0.001), and the trunk-to-appendicular fat ratio was positively correlated with the HDL z-score (r = 0.421, P = 0.004), while the MFR z-score was negatively correlated with the BMI z-score (r = −0.576, P < 0.001).

There were significant correlations between the parents’ and offsprings’ MFR findings in the control group (father-child: r = 0.528, P = 0.020; mother-child: r = 0.359, P = 0.031), no significant correlations were found in the PA group (father-child: r = 0.266, P = 0.0243; mother-child: r = 0.280, P = 0.080).