Abstract
Background
Obesity with low skeletal muscle mass (OLM) is increasingly recognized as a high-risk phenotype in pediatric populations. The skeletal muscle-to-fat mass ratio (MFR) reflects muscle-fat imbalance, but its association with metabolic health in children and adolescents remains incompletely understood. This study examined the associations between MFR and metabolic indicators, with a focus on insulin resistance, in children and adolescents with OLM.
Methods
This cross-sectional study included 258 children and adolescents with obesity and low muscle mass. Body composition was assessed using bioelectrical impedance analysis. Skeletal muscle mass index (SMI) was calculated to define low skeletal muscle mass, and the muscle-to-fat mass ratio (MFR) was categorized into quartiles. Multivariable linear, quantile, and logistic regression models were used to examine the associations between MFR and metabolic parameters, adjusting for age and sex. Insulin resistance was assessed using the homeostasis model assessment of insulin resistance (HOMA-IR), with abnormal insulin resistance defined as HOMA-IR ≥ 3.0.
Results
After adjustment for age and sex, higher MFR was associated with lower levels of LDL cholesterol, fasting insulin, and HOMA-IR. Quantile regression analyses showed predominantly inverse associations between MFR and triglycerides, fasting glucose, and uric acid across multiple quantiles. Compared with participants in the lowest MFR quartile, the adjusted odds ratios for abnormal insulin resistance were 0.41 (95% CI, 0.18–0.92), 0.37 (95% CI, 0.15–0.89), and 0.08 (95% CI, 0.04–0.24) across increasing MFR quartiles (p for trend < 0.001).
Conclusions
In children and adolescents with OLM, a higher MFR was associated with lower insulin resistance and more favorable levels of selected metabolic indicators. These findings suggest the potential relevance of muscle–fat imbalance in pediatric obesity and highlight the need for further research into the role of skeletal muscle mass, alongside adiposity, in metabolic risk assessment.
Clinical trial registration
Not applicable.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12902-026-02234-w.
Keywords: Low muscle mass, Pediatric obesity, Skeletal muscle-to-fat mass ratio, HOMA-IR
Introduction
The prevalence of pediatric obesity has increased substantially over the past few decades. Childhood obesity not only predisposes individuals to obesity later in life, but is also associated with a range of metabolic complications, including type 2 diabetes mellitus (T2DM) [1], elevated blood pressure [2], dyslipidemia [3], non-alcoholic fatty liver disease (NAFLD) [4], and adverse psychosocial outcomes [5]. In recent years, obesity has been increasingly recognized not merely as an excess accumulation of fat mass, but as a heterogeneous condition that may also involve insufficient skeletal muscle mass [6]. This phenotype, characterized by increased fat mass accompanied by low skeletal muscle mass, is referred to in the present study as obesity with low muscle mass (OLM). Although traditionally described as sarcopenic obesity in adults, similar body composition patterns are now increasingly observed in pediatric populations [7]. While low muscle mass is often considered an age-related condition, accumulating evidence suggests that a disproportionate reduction in skeletal muscle mass relative to fat mass is also common among children and adolescents [8]. In China, national surveillance data have shown that increases in body weight among school-aged children and young adults have not been accompanied by proportional gains in muscle strength, which has plateaued or even declined over the past decade, highlighting a growing imbalance between weight gain and muscular development [9].
Skeletal muscle is a critical component of body composition and plays an essential role in normal growth and development, as well as in systemic glucose metabolism during childhood and adolescence [5]. Approximately 40% of postnatal bone growth has been attributed to muscle development [10], and skeletal muscle mass increases in parallel with bone mineral density across childhood and adolescence [11]. Beyond its structural role, skeletal muscle is central to metabolic homeostasis, and reduced or disproportionate muscle mass relative to fat mass has been increasingly linked to metabolic risk. In adults, this imbalance has been described in the context of sarcopenic obesity and is associated with insulin resistance, dyslipidemia, and metabolic syndrome [12]. Similar patterns of mismatch between fat accumulation and muscular development have been documented in large Chinese pediatric cohorts, and may partly explain why obese youth exhibit metabolic abnormalities at relatively young ages [9].
Growing evidence suggests that skeletal muscle plays a pivotal role in glucose and lipid metabolism and can be assessed using indices such as fat-free mass (FFM), lean body mass (LBM), and skeletal muscle mass (SMM) [13]. A systematic review of 15 studies comparing children and adolescents with insulin resistance, impaired glucose tolerance, or metabolic syndrome to metabolically healthy peers found that those with metabolic disturbances consistently exhibited lower proportions of FFM or LBM [14]. Another systematic review further reported that reduced skeletal muscle mass was closely linked to insulin resistance and metabolic syndrome [8, 15]. These findings indicate that assessments focusing solely on body weight or body mass index (BMI) may overlook important aspects of body composition. Consequently, indices that capture both skeletal muscle mass and adiposity may provide additional insight into metabolic risk beyond BMI alone [16, 17].
Despite increasing recognition of OLM, few studies have specifically examined low skeletal muscle mass during childhood and adolescence, particularly in combination with obesity. Moreover, limited data are available regarding how muscle–fat imbalance relates to metabolic disturbances in pediatric populations. To address this gap, the present study aimed to investigate the association between the skeletal muscle-to-fat mass ratio (MFR), an index reflecting the balance between skeletal muscle and fat mass, and metabolic indicators, with a particular focus on insulin resistance, in children and adolescents with obesity and low skeletal muscle mass.
Materials and methods
Participants
A total of 281 children and adolescents with obesity aged 6–17 years, who attended the outpatient clinic of the Department of Clinical Nutrition at Xinhua Hospital (Shanghai, China) between September 2012 and December 2019, were enrolled in this study. Body mass index (BMI) was calculated as weight (kg) divided by height squared (m²). Obesity was defined according to the Chinese screening standard for overweight and obesity among school-age children and adolescents (WS/T 586–2018), using age- and sex-specific BMI cut-offs [18]. Low skeletal muscle mass was defined based on published age- and sex-specific reference values of the skeletal muscle index (SMI) among Chinese children and adolescents [19]. SMI was calculated as whole-body skeletal muscle mass (kg) divided by height squared (m²). Participants were classified as having low skeletal muscle mass if their SMI was below the age- and sex-specific cut-off defined as the reference mean minus 2 standard deviations (mean-2 SD) [19]. These criteria were used to identify participants with OLM (obesity and low skeletal muscle mass). After applying the inclusion criteria, 258 participants (125 boys and 133 girls) were included in the final analysis. Pubertal maturation may influence both body composition and insulin sensitivity; however, Tanner stage information was not routinely assessed in our dataset, which limited our ability to incorporate it into the definition of low skeletal muscle mass or adjust for it in the analyses. Pubertal stage is known to influence both skeletal muscle mass and insulin sensitivity, and this represents a significant limitation. To partially address this, we conducted age-stratified sensitivity analyses (Supplementary Table S3) as a proxy for pubertal development. The study was conducted in accordance with the Declaration of Helsinki (2013 revision) and was approved by the Ethics Committee of Xinhua Hospital, Shanghai Jiao Tong University School of Medicine.
Anthropometry and body composition
Body weight and height were measured by trained nurses at the time of admission. Body composition was assessed using bioelectrical impedance analysis (BIA) with a whole-body impedance analyzer (InBody720; Biospace Inc., South Korea), and all measurements were performed by experienced nurses. The BIA assessment provided estimates of skeletal muscle mass (SMM, kg), fat-free mass (FFM, kg), percent body fat (PBF, %), percent skeletal muscle (PSM, %), and basal metabolic rate (BMR). Skeletal muscle mass index (SMI) was calculated as skeletal muscle mass (SMM, kg) divided by height squared (m²). Fat mass index (FMI) was calculated as fat mass (kg) divided by height squared (m²). The muscle-to-fat mass ratio (MFR) was calculated as SMM (kg) divided by fat mass (kg). These indices were calculated according to commonly used definitions in body composition research [20, 21]. The definitions and grouping criteria used in this study are summarized in Supplementary Table S5.
Laboratory measurements
Venous blood samples were obtained after an overnight fast. Serum levels of high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), triglycerides (TG), total cholesterol (TC), uric acid, alanine aminotransferase (ALT), and aspartate aminotransferase (AST) were analyzed in the Clinical Laboratory Center of Xinhua Hospital using a Hitachi 7060 C automated biochemical analyzer. Fasting glucose, insulin, interleukin-6 (IL-6), and tumor necrosis factor-α (TNF-α) concentrations were determined with monoclonal antibody-based sandwich enzyme-linked immunosorbent assays (ELISA). Insulin resistance was assessed using the homeostasis model assessment of insulin resistance (HOMA-IR), calculated as (fasting insulin [mU/L] × fasting glucose [mmol/L]) / 22.5. Abnormal insulin resistance was defined as HOMA-IR ≥ 3.0, a commonly used cut-off for children and adolescents in previous studies [22].
Statistical analysis
The distribution of continuous variables was evaluated using the Shapiro-Wilk test and visual inspection of histograms. Continuous variables are presented as mean ± standard deviation (SD) when normally distributed and as median (interquartile range, IQR) when non-normally distributed. Categorical variables are summarized as number (percentage). Comparisons across quartiles of the muscle-to-fat mass ratio (MFR) were conducted using one-way analysis of variance (ANOVA) for normally distributed continuous variables and the Kruskal-Wallis test for skewed variables. Categorical variables were compared using the chi-square test or Fisher’s exact test, as appropriate. When an overall difference across quartiles was observed, post hoc pairwise comparisons were performed using Dunn’s test with Bonferroni correction. Differences between boys and girls were assessed using the independent-samples t test for normally distributed continuous variables and the Mann-Whitney U test for non-normally distributed variables. Categorical variables were compared using the chi-square test or Fisher’s exact test, as appropriate. Quantile regression analysis was additionally performed to evaluate the associations between MFR and selected metabolic indicators, including HDL cholesterol, triglycerides, fasting glucose, and uric acid, across different points of their conditional distributions. This method was used to explore potential heterogeneity in associations beyond mean effects and is particularly suitable for skewed outcome variables. Quantile regression models were adjusted for age and sex, and regression coefficients with corresponding 95% confidence intervals were estimated at selected quantiles. BMI was not included in quantile regression models to avoid potential collinearity with MFR, which incorporates both muscle and fat components. Associations between MFR and metabolic outcomes were examined using multivariable linear regression models for continuous outcomes and multivariable logistic regression models for abnormal insulin resistance, defined as HOMA-IR ≥ 3.0. In analyses including the total study population, regression models were adjusted for age and sex. In sex-stratified analyses (boys and girls), models were adjusted for age only. Regression results are presented as regression coefficients (β) or odds ratios (ORs) with corresponding 95% confidence intervals (CIs). All statistical analyses were conducted using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA), and a two-sided p value < 0.05 was considered statistically significant.
Results
General characteristics of the participants
A total of 258 children and adolescents with obesity and low skeletal muscle mass (OLM) were included in the analysis (125 boys and 133 girls). The mean age of the participants was 9.67 years (10.17 years in boys and 9.20 years in girls) (Supplementary Table S1). Participants were categorized into quartiles according to the muscle-to-fat mass ratio (MFR), and body composition characteristics across MFR quartiles are summarized in Table 1. Overall, BMI, fat-free mass (FFM), and percent body fat (PBF) differed significantly across MFR quartiles (all p < 0.001). In addition, the distribution of sex varied across quartiles, with a higher proportion of boys in the highest MFR quartile (Q4) (p < 0.001) (Supplementary Table S1).
Table 1.
General characteristics of the MFR quartiles
| Q1 | Q2 | Q3 | Q4 | p value | |
|---|---|---|---|---|---|
| n = 65 | n = 64 | n = 64 | n = 65 | ||
| MFR | 0.62 ± 0.01 | 0.75 ± 0.004 | 0.87 ± 0.005 | 1.09 ± 0.02 | ‒ |
| Age (year) | 8.00 (7.00, 9.00) | 9.00 (8.00, 10.00) | 9.00 (8.00, 10.00) | 11.00 (10.00, 13.00) | 0.5691 |
| Weight (kg) | 41.27 (35.90, 44.40) a | 53.19 (50.55, 56.33) ab | 49.20 (46.80, 51.07) ab | 69.10 (59.90, 77.10) b | < 0.0001 |
| Height (cm) | 134.25 (130.00, 138.10) | 142.45 (137.30, 146.15) | 141.50 (136.80, 145.00) | 155.50 (147.50, 161.10) | 0.7712 |
| BMI (kg/m2) | 22.60 (20.80, 23.87) a | 26.43 (25.11, 27.27) a | 24.82 (24.08, 26.14) ab | 28.47 (27.18, 30.96) b | < 0.0001 |
| FFM (kg) | 25.40 (23.34, 27.72) | 30.17 (27.80, 31.50) | 31.73 (29.80, 33.63) | 39.30 (34.96, 44.05) | 0.4435 |
| FM (kg) | 27.60 (23.50, 33.40) a | 17.88 (16.20, 19.70) ab | 23.20 (22.30, 25.43) b | 14.94 (12.53, 17.66) b | < 0.0001 |
| SMM (kg) | 13.11 (11.81, 14.46) b | 15.63 (14.08, 16.36) ab | 16.70 (15.60, 18.20) ab | 21.40 (19.10, 24.30) a | < 0.0001 |
| SMI (kg/m2) | 7.27 (6.88, 7.68) | 7.63 (7.39, 7.84) | 8.49 (8.17, 8.60) | 8.92 (8.53, 9.52) | 0.7121 |
| PBF (%) | 37.10 (33.40, 39.40) a | 44.0 (42.80, 46.10) a | 35.90 (33.40, 37.90) ab | 41.60 (39.20, 44.80) b | < 0.0001 |
| PSM (%) | 32.30 (30.90, 33.80) b | 29.30 (28.10, 30.00) ab | 31.50 (29.60, 33.10) ab | 33.80 (32.80, 35.50) a | < 0.0001 |
| BMR (kcal) | 920.0 (874.1, 970.0) | 1019.5 (963.5, 1047.0) | 1055.6 (1014.0, 1096.5) | 1224.0 (1139.0, 1340.0) | 0.4349 |
Data are presented as median (interquartile range, IQR) for continuous variables and number (percentage) for categorical variables. p values were not calculated for MFR, as it was used to define the quartile groups. p values were calculated using the Kruskal–Wallis test for continuous variables and the chi-square test (or Fisher’s exact test, when appropriate) for categorical variables. When overall group differences were statistically significant, post hoc pairwise comparisons were performed using Dunn’s test with Bonferroni correction. Superscript letters are shown only for variables with significant overall group differences
MFR, muscle-to-fat mass ratio; BMI, body mass index; FFM, fat-free mass; FM, fat mass; SMM, skeletal muscle mass; SMI, skeletal muscle mass index; PBF, percent body fat; PSM, percent skeletal muscle; BMR, basal metabolic rate
Laboratory measures in the MFR quartile groups
Laboratory measures across MFR quartiles are presented in Table 2. Significant differences were observed for LDL cholesterol (p < 0.05), total cholesterol (p < 0.05), triglycerides (p < 0.01), fasting insulin (p < 0.001), HOMA-IR (p < 0.001), and alanine aminotransferase (ALT) (p < 0.01). Post hoc pairwise comparisons using Dunn’s test with Bonferroni correction revealed significant differences between MFR quartiles for several lipid and insulin-related markers. No statistically significant differences were detected across MFR quartiles for HDL cholesterol, fasting glucose, uric acid, AST, IL-6, or TNF-α. Sex-specific comparisons are shown in Supplementary Table S2. Girls had higher levels of total cholesterol, fasting insulin, HOMA-IR, uric acid, and ALT than boys (all p < 0.05), while other laboratory measures did not differ significantly between sexes.
Table 2.
Laboratory measures according to MFR quartiles
| Q1 | Q2 | Q3 | Q4 | p value | |
|---|---|---|---|---|---|
| n = 65 | n = 64 | n = 64 | n = 65 | ||
| MFR | 0.62 ± 0.01 | 0.75 ± 0.004 | 0.87 ± 0.005 | 1.09 ± 0.02 | ‒ |
| HDL-C (mmol/L) | 1.44 (1.23, 1.54) | 1.40 (1.29, 1.50) | 1.38 (1.24, 1.51) | 1.29 (1.16, 1.49) | 0.293 |
| LDL-C (mmol/L) | 2.45 (2.18, 2.78) a | 2.84 (2.38, 3.28) ab | 2.29 (2.19, 2.41) b | 2.40 (2.11, 2.85) b | 0.021 |
| TC (mmol/L) | 4.25 (3.95, 4.76) | 4.41 (4.18, 4.81) | 3.86 (3.65, 4.41) | 4.16 (3.74, 4.66) | 0.029 |
| TG (mmol/L) | 0.89 (0.59, 1.39) b | 1.41 (0.96, 1.84) a | 1.13 (0.77, 1.35) a | 0.93 (0.89, 1.66) ab | 0.003 |
| Fasting glucose (mmol/L) | 5.01 (4.82, 5.30) | 4.98 (4.85, 5.16) | 5.09 (4.90, 5.33) | 5.10 (4.89, 5.32) | 0.279 |
| Fasting insulin (pmol/L) | 39.2 (25.8, 62.1) b | 85.7 (49.0, 104.6) a | 76.0 (51.9, 129.4) a | 57.4 (47.1, 77.0) ab | < 0.001 |
| HOMA-IR | 2.52 (1.70, 4.11) a | 1.74 (1.53, 2.36) b | 2.92 (1.61, 3.39) b | 1.22 (0.79, 2.05) b | < 0.001 |
| Uric acid (µmol/L) | 312 (275, 348) | 359 (307, 414) | 315 (254, 399) | 329 (289, 408) | 0.065 |
| ALT (U/L) | 19.00 (14.00, 36.5) a | 25.00 (16.5, 39.0) a | 19.0 (15.0, 36.0) ab | 34.0 (19.0, 55.0) b | 0.002 |
| AST (U/L) | 25.0 (22.5, 34.0) | 27.5 (20.5, 31.5) | 23.0 (22.0, 27.5) | 26.0 (19.0, 41.0) | 0.629 |
| IL-6 (pg/mL) | 3.23 (2.01, 4.39) | 2.41 (1.00, 4.15) | 2.74 (1.00, 3.62) | 2.58 (1.00, 3.82) | 0.2860 |
| TNF-α (pg/mL) | 8.51 (6.89, 10.60) | 9.60 (8.03, 11.90) | 8.89 (7.43, 10.10) | 9.40 (7.17, 11.50) | 0.761 |
Data are presented as median (interquartile range, IQR) for continuous variables and number (percentage) for categorical variables. p values were not calculated for MFR, as it was used to define the quartile groups. p values were calculated using the Kruskal–Wallis test for continuous variables and the chi-square test (or Fisher’s exact test, when appropriate) for categorical variables. When overall group differences were statistically significant, post hoc pairwise comparisons were performed using Dunn’s test with Bonferroni correction. Superscript letters are shown only for variables with significant overall group differences
HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; TC, total cholesterol; TG, triglycerides; ALT, alanine aminotransferase; AST, aspartate aminotransferase; HOMA-IR, homeostasis model assessment of insulin resistance; IL-6, interleukin-6; TNF-α, tumor necrosis factor-alpha
Associations between laboratory measures and MFR
Multivariable linear regression analyses examining the associations between laboratory measures and MFR are presented in Table 3. In the total sample, after adjustment for age and sex, higher MFR was significantly associated with lower levels of LDL cholesterol (p < 0.01), total cholesterol (p < 0.01), fasting insulin (p < 0.001), HOMA-IR (p < 0.001), and alanine aminotransferase (ALT) (p < 0.05). In sex-stratified analyses adjusted for age, higher MFR was significantly associated with lower fasting insulin levels (p < 0.05) and HOMA-IR (p < 0.05) in boys. In girls, higher MFR was significantly associated with lower levels of LDL cholesterol (p < 0.01), total cholesterol (p < 0.05), fasting insulin (p < 0.01), and HOMA-IR (p < 0.01). No statistically significant associations were observed between MFR and inflammatory markers in either sex.
Table 3.
Multivariable linear regression analyses of the associations between muscle-to-fat mass ratio (MFR) and laboratory measures in children and adolescents with obesity and low skeletal muscle mass
| Total (n = 258) | Boys (n = 125) | Girls (n = 133) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| β | Standardized β | 95%CI | p value | β | Standardized β | 95%CI | p value | β | Standardized β | 95%CI | p value | |||
| HDL-C (mmol/L) | -0.637 | -0.210 | (-0.993, -0.281) | 0.001 | -0.306 | -0.106 | (-0.8681, 0.256) | 0.288 | -0.991 | -0.313 | (-1.715, -0.267) | 0.008 | ||
| LDL-C (mmol/L) | -0.633 | -0.167 | (-1.097, -0.170) | 0.008 | -0.308 | -0.083 | (-1.1016, 0.401) | 0.392 | -0.896 | -0.239 | (-1.685, -0.107) | 0.027 | ||
| TC (mmol/L) | -0.168 | -0.044 | (-0.630, 0.294) | 0.475 | -0.180 | 0.040 | (-0.015, 0.656) | 0.671 | -0.191 | -0.063 | (-0.847, 0.465) | 0.564 | ||
| TG (mmol/L) | 0.058 | 0.033 | (-0.155, 0.271) | 0.593 | -0.233 | -0.118 | (-0.589, 0.123) | 0.197 | 0.212 | 0.130 | (-0.097, 0.522) | 0.178 | ||
| Fasting glucose (mmol/L) | -0.074 | -0.234 | (-0.106, -0.043) | < 0.001 | -0.065 | -0.201 | (-0.125, -0.005) | 0.033 | -0.079 | -0.256 | (-0.134, -0.023) | 0.006 | ||
| Fasting insulin (pmol/L) | -2.458 | -0.237 | (-3.509, -1.406) | < 0.001 | -2.210 | -0.206 | (-4.197, -0.222) | 0.030 | -2.585 | -0.259 | (-4.431, -0.739) | 0.007 | ||
| HOMA-IR | -0.022 | -0.062 | (-0.084, 0.039) | 0.469 | -0.087 | -0.163 | (-0.202, 0.028) | 0.135 | -0.009 | -0.028 | (-0.072, 0.055) | 0.790 | ||
| Uric acid (µmol/L) | -18.491 | -0.116 | (-33.366, -3.616) | 0.015 | -33.081 | -0.145 | (-76.715, 10.552) | 0.136 | -11.523 | -0.132 | (-29.686, 6.640) | 0.211 | ||
| ALT (U/L) | -3.407 | -0.048 | (-10.661, 3.847) | 0.356 | -12.878 | -0.127 | (-32.429, 6.674) | 0.199 | 0.779 | 0.019 | (-7.905, 9.464) | 0.859 | ||
| AST (U/L) | -0.875 | -0.050 | (-3.842, 2.092) | 0.562 | -2.175 | -0.111 | (-5.561, 1.210) | 0.206 | 0.600 | 0.046 | (-1.730, 2.930) | 0.611 | ||
| IL-6 (pg/mL) | -1.597 | -0.092 | (-4.530, 1.336) | 0.285 | -0.018 | -0.001 | (-2.391, 2.355) | 0.988 | -2.393 | -0.124 | (-5.727, 0.942) | 0.158 | ||
Models in the total sample were adjusted for age and sex. Sex-stratified models (boys and girls) were adjusted for age only. Regression coefficients (β) are presented with 95% confidence intervals (CIs)
CI, confidence interval; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; TC, total cholesterol; TG, triglycerides; HOMA-IR, homeostasis model assessment of insulin resistance; ALT, alanine aminotransferase; AST, aspartate aminotransferase; IL-6, interleukin-6; TNF-α, tumor necrosis factor-alpha
Quantile regression analysis of MFR and metabolic indicators
Table 4 summarizes the quantile regression analyses examining the associations between the muscle-to-fat mass ratio (MFR) and selected metabolic indicators, including HDL cholesterol, triglycerides (TG), fasting glucose, and uric acid, across different points of their conditional distributions. The estimated regression coefficients varied in both magnitude and precision across quantiles, indicating heterogeneity in the associations along the outcome distributions. For HDL cholesterol, the regression coefficients tended to be negative across the lower-to-middle quantiles (10th -70th percentiles); however, the corresponding 95% confidence intervals included the null value, indicating limited statistical precision. For TG, inverse associations with MFR were mainly observed at the median and upper-middle quantiles (50th -90th percentiles). Similarly, fasting glucose showed negative regression coefficients across several lower-to-middle quantiles (20th -70th percentiles). For uric acid, inverse associations with MFR were observed across most quantiles, with weaker or no evidence of association at higher quantiles.
Table 4.
Coefficient of association between HDL-C, TG, fasting glucose and uric acid values at quantile levels of MFR (Coefficients and 95% CIs)
| Quantile | HDL-C (mmol/L) | TG (mmol/L) | Fasting glucose (mmol/L) | Uric acid (µmol/L) | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Coefficients | 95% CI | Coefficients | 95% CI | Coefficients | 95% CI | Coefficients | 95% CI | ||||
| 10th | -0.047 | (-0.176, 0.082) | 0.014 | (-0.038, 0.066) | 0.001 | (-0.094, 0.096) | -0.0002 | (-0.0008, 0.0004) | |||
| 20th | -0.036 | (-0.123, 0.052) | 0.012 | (-0.024, 0.048) | -0.023 | (-0.103, 0.057) | -0.0002 | (-0.0007, 0.0003) | |||
| 25th | -0.066 | (-0.169, 0.037) | 0.015 | (-0.023, 0.053) | -0.011 | (-0.085, 0.064) | -0.0004 | (-0.0008, 0.0001) | |||
| 30th | -0.057 | (-0.161, 0.047) | 0.003 | (-0.034, 0.041) | -0.040 | (-0.108, 0.027) | -0.0004 | (-0.0008, 0.0000) | |||
| 40th | -0.059 | (-0.157, 0.040) | 0.006 | (-0.034, 0.046) | -0.048 | (-0.127, 0.031) | -0.0002 | (-0.0007, 0.0004) | |||
| 50th | -0.021 | (-0.128, 0.085) | -0.006 | (-0.045, 0.033) | -0.029 | (-0.099, 0.041) | -0.0004 | (-0.0008, 0.0001) | |||
| 60th | -0.001 | (-0.117, 0.114) | -0.010 | (-0.053, 0.032) | -0.018 | (-0.098, 0.062) | -0.0004 | (-0.0008, 0.0001) | |||
| 70th | -0.011 | (-0.124, 0.102) | -0.011 | (-0.056, 0.034) | -0.034 | (-0.122, 0.054) | -0.0002 | (-0.0007, 0.0004) | |||
| 75th | 0.016 | (-0.110, 0.143) | 0.006 | (-0.042, 0.054) | 0.002 | (-0.086, 0.090) | -0.0002 | (-0.0008, 0.0003) | |||
| 80th | 0.031 | (-0.103, 0.165) | 0.016 | (-0.034, 0.066) | -0.006 | (-0.105, 0.094) | 0.0000 | (-0.0006, 0.0006) | |||
| 90th | 0.011 | (-0.179, 0.202) | -0.028 | (-0.090, 0.034) | 0.082 | (-0.093, 0.257) | 0.0002 | (-0.0011, 0.0007) | |||
Adjusted for age and sex. Quantile levels ranged from 0.1 to 0.9, with corresponding MFR values of 0.61, 0.67, 0.69, 0.72, 0.77, 0.80, 0.85, 0.90, 0.93, 0.96, and 1.07, respectively. CI, confidence interval
The quantile-specific regression coefficients and their 95% confidence intervals are illustrated in Figs. 1, 2, 3 and 4 for HDL cholesterol, TG, fasting glucose, and uric acid, respectively. In each figure, the solid line represents the estimated regression coefficient, the shaded area indicates the corresponding 95% confidence interval, and the horizontal dashed line denotes the null value (coefficient = 0). Overall, inverse associations between MFR and these metabolic indicators were more apparent in the lower-to-middle quantile ranges, although the strength and precision of the estimates differed across outcomes and quantiles.
Fig. 1.
Quantile regression coefficients for the association between muscle-to-fat mass ratio (MFR) and HDL across quantiles of HDL-C. The solid line represents the estimated regression coefficients, and the shaded area indicates the 95% confidence intervals. The horizontal dashed line denotes the null value (coefficient = 0). Models were adjusted for age and sex
Fig. 2.
Quantile regression coefficients for the association between MFR and triglycerides (TG) across quantiles of TG. The solid line represents the estimated regression coefficients, and the shaded area indicates the 95% confidence intervals. The horizontal dashed line denotes the null value (coefficient = 0). Models were adjusted for age and sex
Fig. 3.
Quantile regression coefficients for the association between MFR and fasting glucose across quantiles of fasting glucose. The solid line represents the estimated regression coefficients, and the shaded area indicates the 95% confidence intervals. The horizontal dashed line denotes the null value (coefficient = 0). Models were adjusted for age and sex
Fig. 4.
Quantile regression coefficients for the association between MFR and uric acid across quantiles of uric acid. The solid line represents the estimated regression coefficients, and the shaded area indicates the 95% confidence intervals. The horizontal dashed line denotes the null value (coefficient = 0). Models were adjusted for age and sex
Association between MFR and risk of abnormal HOMA-IR
Based on clinical cut-off values, no participants met the criteria for abnormal HDL-C, triglycerides, fasting glucose, or uric acid. In contrast, multivariable logistic regression analyses (Table 5) showed a significant inverse association between MFR quartiles and abnormal insulin resistance, defined as HOMA-IR ≥ 3.0. Using the lowest quartile (Q1) as the reference, the odds of abnormal insulin resistance were lower in Q2 (OR = 0.41, 95% CI: 0.18–0.92; p < 0.05), Q3 (OR = 0.37, 95% CI: 0.15–0.89; p < 0.05), and Q4 (OR = 0.08, 95% CI: 0.04–0.24; p < 0.001). A significant linear trend was observed across increasing MFR quartiles (p for trend < 0.001). To address potential residual confounding related to biological maturation, we performed age-stratified sensitivity analyses using chronological age categories (6–9, 10–13, and 14–17 years) as a proxy for pubertal development. As shown in Supplementary Table S3, higher MFR (per 0.1 increase) was consistently associated with lower odds of abnormal insulin resistance across age strata, although the associations were not statistically significant after adjustment for sex and BMI.
Table 5.
Odds ratios for abnormal insulin resistance according to MFR quartiles
| Q1 | Q2 | Q3 | Q4 | p for trend | |
|---|---|---|---|---|---|
|
Abnormal HOMA-IR (n = 59) |
1.00 | 0.410 (0.183, 0.918)* | 0.371 (0.154, 0.893)* | 0.081 (0.026, 0.241)*** | < 0.001 |
Abnormal insulin resistance was defined as HOMA-IR ≥ 3.0. Odds ratios (ORs) and 95% confidence intervals (CIs) were estimated using multivariable logistic regression, with Quartile 1 (Q1) as the reference group. Models were adjusted for age, sex, and BMI. p for trend was calculated by modelling MFR quartiles as an ordinal variable
MFR, muscle-to-fat mass ratio; HOMA-IR, homeostasis model assessment of insulin resistance; OR, odds ratio; CI, confidence interval
Discussion
Obesity with low skeletal muscle mass (OLM), characterized by an unfavorable balance between skeletal muscle and fat mass, is increasingly recognized as a high-risk phenotype in pediatric populations. In this study, we investigated the associations between the muscle-to-fat mass ratio (MFR) and metabolic indicators among children and adolescents with OLM. Higher MFR was associated with lower levels of LDL cholesterol, fasting insulin, and HOMA-IR, and with lower odds of abnormal insulin resistance as defined by HOMA-IR ≥ 3.0. These findings are in line with national surveillance data from China showing that recent increases in body weight among school-aged children and young adults have not been accompanied by proportional improvements in muscle strength, suggesting a growing mismatch between weight gain and muscular development [9]. Thus, MFR may help characterize muscle-fat imbalance within obesity phenotypes that are not fully captured by BMI alone.
Our findings support the concept that metabolic risk in pediatric obesity may relate not only to excess adiposity but also to relatively insufficient skeletal muscle mass. BMI reflects overall body size and does not distinguish between fat and lean compartments. In contrast, MFR incorporates both skeletal muscle mass and fat mass and may capture muscle–fat imbalance more directly. Importantly, MFR is not intended to replace established adiposity indicators but may provide complementary information on body composition balance, particularly in children and adolescents with obesity, in whom impaired muscular development may coexist with fat accumulation and be linked to adverse metabolic profiles.
Skeletal muscle plays a central role in systemic energy metabolism and is the primary site of insulin-stimulated glucose uptake [23]. Therefore, lower muscle mass combined with greater fat accumulation may contribute to impaired glucose handling and dyslipidemia. A large population-based study from Korea reported that lower skeletal muscle mass was associated with less favorable lipid profiles, including higher LDL cholesterol, particularly in individuals with excess adiposity [24]. In line with this evidence, our supplementary analyses suggested that higher adiposity burden, as reflected by fat mass–related indices, was associated with less favorable lipid and glucose metabolism profiles (see Supplementary Table S4).
Several biological mechanisms may explain these associations. Fat infiltration within skeletal muscle and ectopic lipid deposition have been linked to insulin resistance, dyslipidemia, and low-grade inflammation [25, 26]. Prior studies have reported associations between skeletal muscle fat infiltration and unfavorable lipid profiles, including higher LDL cholesterol [26, 27]. In addition, reduced lipid oxidation, accumulation of intramyocellular lipids, and impaired GLUT4-mediated glucose transport have been implicated in disrupted insulin signaling [28–30]. It should be noted that much of the mechanistic evidence comes from adult populations and experimental studies, and the relevance of these pathways in pediatric obesity warrants further investigation.
Notably, overt abnormalities in HDL cholesterol, triglycerides, fasting glucose, and uric acid were uncommon in this relatively young cohort when applying standard clinical cut-offs. However, quantile regression analyses suggested that MFR was associated with these metabolic indicators across several quantiles, indicating that muscle–fat imbalance may relate to subtle metabolic variation even before clinical thresholds are reached. This distribution-based approach may complement mean-based models and help characterize early metabolic changes in pediatric obesity [31, 32]. The inverse association observed for HDL-C in the quantile regression analysis was unexpected and should be interpreted cautiously. This finding may reflect residual confounding (including age- and sex-related differences in lipid metabolism), the cross-sectional design, or measurement variability in body composition estimates, and warrants confirmation in future studies. Importantly, this finding does not detract from the consistent associations observed for insulin-related markers, which represent the primary metabolic outcomes of interest in this study.
Several limitations should be considered. First, the cross-sectional design precludes causal inference, and the single-center setting may limit generalizability. Second, residual confounding cannot be ruled out. Pubertal maturation is closely related to both body composition and insulin sensitivity, yet Tanner staging was not routinely assessed and therefore could not be directly accounted for. As a partial approach, we performed age-stratified sensitivity analyses using chronological age categories (6–9, 10–13, and 14–17 years) as a proxy for pubertal maturation, and the inverse association between MFR and abnormal insulin resistance remained directionally consistent but did not reach statistical significance after adjustment (Supplementary Table S3). In addition, several important confounders were unavailable for adjustment, including objectively measured physical activity, detailed dietary intake, and socioeconomic status. Third, skeletal muscle mass was estimated using bioelectrical impedance analysis (BIA). Although BIA is feasible in clinical practice, it is less precise than imaging-based methods such as DXA or MRI, and its accuracy may be affected by hydration status and body fat distribution in obese pediatric populations, which may introduce measurement error and misclassification. Finally, functional muscle outcomes and mechanistic biomarkers (e.g., grip strength, IL-6, IGF-1, and TNF-α) were not comprehensively assessed. Moreover, universally accepted pediatric thresholds for low skeletal muscle mass are still lacking, which may limit comparability across studies.
Despite these limitations, this study provides data from a clinically relevant pediatric population with obesity and low skeletal muscle mass. Rather than proposing diagnostic cut-offs for MFR, our findings highlight the relevance of considering skeletal muscle mass, alongside adiposity, in research settings aimed at understanding metabolic heterogeneity in pediatric obesity. Future longitudinal studies in larger, multi-center cohorts using more precise body composition assessments (e.g., DXA or MRI) and incorporating pubertal staging, lifestyle factors, and mechanistic biomarkers are needed to validate these associations and clarify their clinical implications.
Conclusion
In children and adolescents with obesity and low skeletal muscle mass, a higher muscle-to-fat mass ratio (MFR) was associated with a lower risk of abnormal insulin resistance. These findings suggest that muscle–fat imbalance may be a relevant factor in metabolic heterogeneity in pediatric obesity. Further longitudinal studies are needed to confirm these associations and explore their potential implications for metabolic health.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
This study was approved by the Ethics Committee of Xinhua Hospital, Shanghai Jiao Tong University School of Medicine. Written informed consent was obtained from all participants’ legal guardians/parents. Clinical trial registration: Not applicable.
Author contributions
Xiuhua Shen and Yi Feng contributed to the study design. Fangfang Song, Luyao Xie and Yang Niu analyzed the data. Fangfang Song wrote the manuscript. All authors provided input into the interpretation of data and have read, reviewed and approved the final manuscript.
Funding
National Natural Science Foundation of China (81773407).
Data availability
The datasets generated and/or analyzed during the current study are not publicly available due to institutional and ethical restrictions involving patient privacy, but are available from the corresponding author upon reasonable request.
Declarations
Ethics approval and consent to participate
The study was conducted in accordance with the Declaration of Helsinki (2013 revision) and approved by the Ethics Committee of Xinhua Hospital, Shanghai Jiao Tong University School of Medicine. Written informed consent was obtained from all participants’ legal guardians/parents, and assent was obtained from children as appropriate.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Fangfang Song and Luyao Xie contributed equally to this work.
References
- 1.Healthy Study Group, Kaufman FR, Hirst K, Linder B, Baranowski T, Cooper DM, et al. Risk factors for type 2 diabetes in a sixth- grade multiracial cohort: the Healthy study. Diabetes Care. 2009;32(5):953–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Sorof J, Daniels S. Obesity Hypertension in Children: A Problem of Epidemic Proportions. Hypertension. 2002;40:441–7. [DOI] [PubMed] [Google Scholar]
- 3.Kumar S, Kelly AS. Review of childhood obesity: from epidemiology, etiology, and comorbidities to clinical assessment and treatment. Mayo Clin Proc. 2017; 92(2): 251–65. [DOI] [PubMed]
- 4.Feldstein AE, Charatcharoenwitthaya P, Treeprasertsuk S, Benson JT, Enders FB, Angulo P. The Natural History of non-Alcoholic Fatty Liver Disease in Children: A Follow-Up Study for Up to 20 Years. Gut. 2009;58:1538–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Rolland-Cachera MF. Childhood obesity: current definitions and recommendations for their use. Int J Pediatr Obes. 2011;6(5–6):325–31. [DOI] [PubMed] [Google Scholar]
- 6.Xie L, Mo PKH, Tang Q, Zhao X, Zhao X, Cai W, et al. Skeletal Muscle Mass Has Stronger Association With the Risk of Hyperuricemia Than Body Fat Mass in Obese Children and Adolescents. Front Nutr. 2022;9:792234. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Zembura M, Matusik P. Sarcopenic Obesity in Children and Adolescents: A Systematic Review. Front Endocrinol (Lausanne). 2022;13:914740. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Steene-Johannessen J, Anderssen SA, Kolle E, Andersen LB. Low muscle fitness is associated with metabolic risk in youth. Med Sci Sports Exerc. 2009;41(7):1361–7. [DOI] [PubMed] [Google Scholar]
- 9.Wang X, Wang H, Yuan X, Cai S, Huang Y, Song Y, et al. Imbalance between muscle strength development and weight gain in children and young adults in China: serial cross-sectional evidence from 1.33 million students from five successive national surveys between 2000 and 2019. Lancet Reg Health West Pac. 2025;61:101640. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Frost HM, Schönau E. The muscle-bone unit in children and adolescents. Calcif Tissue Int. 2002;70:405–7. [DOI] [PubMed] [Google Scholar]
- 11.Liu J, Wang L, Sun J, Liu G, Yan W, Xi B, Xiong F, et al. Bone mineral density reference standards for Chinese children aged 3–18: crosssectional results of the 2013–2015 China Child and Adolescent Cardiovascular Health (CCACH) Study. BMJ Open. 2017;7:e014542. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Park BS, Yoon JS. Relative skeletal muscle mass is associated with development of metabolic syndrome. Diabetes Metab J. 2013;37:458–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Gao L, Cheng H, Yan Y, Liu J, Shan X, Wang X, et al. The associations of muscle mass with glucose and lipid metabolism are influenced by body fat accumulation in children and adolescents. Front Endocrinol (Lausanne). 2022;13:976998. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Córdoba-Rodríguez DP, Iglesia I, Gomez-Bruton A, Rodríguez G, Casajús JA, Morales-Devia H, et al. Fat-free/lean body mass in children with insulin resistance or metabolic syndrome: a systematic review and meta-analysis. BMC Pediatr. 2022;22:58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Nishikawa H, Asai A, Fukunishi S, Nishiguchi S, Higuchi K. Metabolic Syndrome and Sarcopenia. Nutrients. 2021;13(10):3519. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Chen F, Liu J, Hou D, Li T, Chen Y, Liao Z, et al. The Relationship between Fat Mass Percentage and Glucose Metabolism in Children and Adolescents: A Systematic Review and Meta-Analysis. Nutrients. 2022;14(11):2272. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Park H, Choi JE, Jun S, Lee H, Kim HS, Lee HA, et al. Metabolic complications of obesity in children and adolescents. Clin Exp Pediatr. 2024;67(7):347–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.National Health Commission of the People’s Republic of China. Screening for overweight and obesity among school-age children and adolescents (WS/T 586–2018). 2018.
- 19.Liu J, Yan Y, Xi B, Huang G, Mi J. China Child and Adolescent Cardiovascular Health (CCACH) Study Group. Skeletal muscle reference for Chinese children and adolescents. J Cachexia Sarcopenia Muscle. 2019;10(1):155–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.VanItallie TB, Yang MU, Heymsfield SB, Funk RC, Boileau RA. Height-normalized indices of the body’s fat-free mass and fat mass: potentially useful indicators of nutritional status. Am J Clin Nutr. 1990;52(6):953–9. [DOI] [PubMed] [Google Scholar]
- 21.Kim TN, Park MS, Yang SJ, Yoo HJ, Kang HJ, Song W, et al. Prevalence and determinant factors of sarcopenia in patients with type 2 diabetes: the Korean Sarcopenic Obesity Study (KSOS). Diabetes Care. 2010;33(7):1497–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Yin J, Li M, Xu L, Wang Y, Cheng H, Zhao X, et al. Insulin resistance determined by Homeostasis Model Assessment (HOMA) and associations with metabolic syndrome among Chinese children and teenagers. Diabetol Metab Syndr. 2013;5(1):71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Honka MJ, Latva-Rasku A, Bucci M, Virtanen KA, Hannukainen JC, Kalliokoski KK, et al. Insulin-stimulated glucose uptake in skeletal muscle, adipose tissue and liver: a positron emission tomography study. Eur J Endocrinol. 2018;178(5):523–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Lee JH, Lee HS, Cho AR, Lee YJ, Kwon YJ. Relationship between muscle mass index and LDL cholesterol target levels: Analysis of two studies of the Korean population. Atherosclerosis. 2021;325:1–7. [DOI] [PubMed] [Google Scholar]
- 25.Shulman GI. Ectopic fat in insulin resistance, dyslipidemia, and cardiometabolic disease. N Engl J Med. 2014;371(23):2237–8. [DOI] [PubMed] [Google Scholar]
- 26.Hamrick MW, McGee-Lawrence ME, Frechette DM. Fatty Infiltration of Skeletal Muscle: Mechanisms and Comparisons with Bone Marrow Adiposity. Front Endocrinol (Lausanne). 2016;7:69. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Vella CA, Nelson MC, Unkart JT, Miljkovic I, Allison MA. Skeletal muscle area and density are associated with lipid and lipoprotein cholesterol levels: The Multi-Ethnic Study of Atherosclerosis. J Clin Lipidol. 2020;14(1):143–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Samuel VT, Shulman GI. Mechanisms for insulin resistance: common threads and missing links. Cell. 2012;148(5):852–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Li Y, Xu S, Zhang X, Yi Z, Cichello S. Skeletal intramyocellular lipid metabolism and insulin resistance. Biophys Rep. 2015;1:90–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Van Gerwen J, Shun-Shion AS, Fazakerley DJ. Insulin signalling and GLUT4 trafficking in insulin resistance. Biochem Soc Trans. 2023;51(3):1057–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Koenker R, Hallock KF. Quantile regression[J]. J economic Perspect. 2001;15(4):143–56. [Google Scholar]
- 32.Wei Y, Pere A, Koenker R, He X. Quantile regression methods for reference growth charts. Stat Med. 2006;25(8):1369–82. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets generated and/or analyzed during the current study are not publicly available due to institutional and ethical restrictions involving patient privacy, but are available from the corresponding author upon reasonable request.