Abstract
Purpose
Both muscle mass and strength relate to metabolic health, but their separate contributions to metabolic syndrome (MetS) are unclear. Using the 2024 Korea National Health and Nutrition Examination Survey (KNHANES) with concurrent dual-energy X-ray absorptiometry (DXA)–based appendicular skeletal muscle mass (ASM) and handgrip strength (HGS), we examined the independent associations of relative muscle mass and HGS with MetS in Korean adults aged ≥40 years.
Methods
We analyzed 3033 adults from KNHANES. ASM/body weight (ASM/BW) and HGS/body weight (relative HGS) served as primary indicators. Participants were classified into four groups by sex-specific medians of both indicators, and complex-sample logistic regression estimated odds ratios (ORs) for MetS with both indicators entered simultaneously.
Results
In the fully adjusted joint model, ASM/BW was inversely associated with MetS in men (OR 0.79, 95% CI 0.73–0.86) and women (OR 0.86, 0.80–0.92), as was relative HGS (men OR 0.98, 0.96–0.99; women OR 0.97, 0.95–0.98). The odds of MetS were highest with concurrently low muscle mass and strength (men OR 3.04; women OR 2.42). The ASM/BW association persisted when MetS was defined without the waist criterion, whereas the relative HGS association was attenuated; ASM/BMI showed similar inverse associations, whereas ASM/Ht² showed a different pattern.
Conclusion
ASM/BW showed more consistent associations with MetS across the sensitivity analyses performed than relative HGS, although associations differed by normalization method, the optimal choice of which remains uncertain. Because the design is cross-sectional, these findings indicate association only and require prospective validation before clinical application.
Keywords: Metabolic syndrome, Muscle mass, Handgrip strength, Sarcopenia, Body composition, KNHANES
Introduction
Metabolic syndrome (MetS) affects approximately one-quarter to one-third of adults worldwide and substantially increases the risk of cardiovascular disease and type 2 diabetes [1]. As populations age and obesity rates rise, MetS is becoming more common [2], including in South Korea [3].
Skeletal muscle is the principal site of insulin-mediated glucose uptake and is essential for maintaining systemic metabolic homeostasis [4]. Reduced muscle mass is closely linked to greater insulin resistance, lower basal metabolic rate, and increased fat accumulation [5], and several epidemiological studies have shown that decreased muscle mass, or sarcopenia, is significantly associated with MetS [6]. Muscle strength is also an important health indicator [7]. Handgrip strength (HGS) provides a simple, reliable measure of overall muscle strength, and an inverse association exists between HGS and the risk of various chronic diseases and MetS [8, 9]. Because absolute HGS is influenced by body size, normalizing HGS to body weight (relative HGS) is a more suitable measure for metabolic research [10].
Because muscle mass and muscle strength are closely related [11], it is difficult to fully account for their mutual influence when they are evaluated separately. While some previous studies have included both variables in multivariate models [12], these analyses were primarily designed to comprehensively consider various risk factors and thus had limited ability to clearly distinguish the independent contributions of the two measures. In particular, it remains unclear whether a decline in muscle strength increases MetS risk when muscle mass is sufficient, or whether muscle mass is a more critical determinant regardless of strength. Moreover, although appendicular skeletal muscle mass relative to body weight (ASM/BW) has been used in metabolic research as an index of relative muscle mass [13], research directly comparing the independent roles of muscle mass and muscle strength using this metric remains insufficient. Establishing whether each indicator remains associated with MetS after adjustment for the other is important for evaluating their potential complementary value in metabolic risk assessment.
Therefore, we used data from the 2024 Korea National Health and Nutrition Examination Survey (KNHANES)—the first cycle to simultaneously include DXA-based muscle mass and HGS measurements—to examine the independent and complementary metabolic roles of muscle mass and muscle strength in relation to MetS in Korean adults aged 40 years and older. By integrating these two indicators within a single analytical framework, we aimed to examine whether muscle mass and strength are associated with partly distinct metabolic patterns and to evaluate whether these associations varied by sex, age, and body mass index (BMI).
Materials and methods
Study design and participants
This cross-sectional study used data from the 2024 KNHANES, conducted by the Korea Disease Control and Prevention Agency (KDCA). KNHANES is a nationally representative survey employing a stratified cluster sampling design; analyses were weighted using the KNHANES combined health-interview and examination weight (wt_itvex), with the survey variance-estimation strata (kstrata) and primary sampling units (psu); KNHANES 2024 does not provide a DXA-specific sampling weight, as DXA is performed within the health examination component rather than as a separately sampled sub-survey. The 2024 cycle was selected because it is the first KNHANES cycle since 2008–2011 to include body-composition measurement by DXA, enabling accurate assessment of appendicular skeletal muscle mass (ASM). Although HGS has been measured in KNHANES since 2014, the 2024 cycle is the only one in which both DXA-based muscle mass and HGS are simultaneously available; thus, integrated multi-year or longitudinal analyses were not possible. Of 6997 total participants, those aged <40 years (n = 2317), with missing grip strength (n = 498), with missing DXA-based muscle mass, height, or body weight (n = 1046), and with missing MetS components (n = 103) were sequentially excluded, yielding a final analytic sample of 3033 participants (1276 men and 1757 women) (Fig. 1). The KNHANES protocol was approved by the KDCA Institutional Review Board (No. 2022-11-16-R-03), and all participants provided written informed consent. For this secondary analysis, the protocol was exempt from review by the Institutional Review Board of Chung-Ang University Hospital (IRB No. 2604-001-19620).
Fig. 1.
Flowchart of study participant selection from KNHANES 2024. Of 6997 participants, those aged <40 years (n = 2317), with missing handgrip strength (n = 498), with missing DXA-based muscle mass, height, or body weight (n = 1046), and with missing MetS component data (n = 103) were sequentially excluded, yielding 3033 participants (1276 men, 1757 women)
Muscle mass assessment
ASM was measured using DXA (Hologic Discovery, Hologic, USA). ASM divided by body weight (ASM/BW, kg/kg) was used as the primary muscle-mass indicator. ASM/BW should be understood as relative appendicular muscle mass expressed as a proportion of body mass; body weight is incorporated into the exposure itself rather than adjusted for, and ASM/BW is therefore not independent of adiposity-related constructs. Additional analyses using ASM divided by height squared (ASM/Ht2, kg/m2) and by BMI (ASM/BMI, kg/[kg/m²]) are presented in Supplementary Table S1. DXA-based body-composition estimates can vary across racial and ethnic populations [14]; the present analyses are therefore interpreted within a single, ethnically homogeneous Korean population, and cross-population comparisons are made cautiously.
Muscle strength assessment
Muscle strength was evaluated by HGS using a digital dynamometer (TKK 5401; Takei Scientific Instruments, Tokyo, Japan). Each participant performed two measurements per hand alternately, and the maximum value across trials was used. Relative HGS was calculated as absolute HGS divided by body weight (kg/kg) and used as the primary strength indicator, analogous to ASM/BW.
Classification of muscle status
Participants were classified into four groups by sex-specific median values of ASM/BW and relative HGS: High ASM/BW + High relative HGS (reference), High ASM/BW + Low relative HGS, Low ASM/BW + High relative HGS, and Low ASM/BW + Low relative HGS. The sex-specific medians were ASM/BW 0.277 kg/kg in men and 0.219 kg/kg in women, and relative HGS 0.553 kg/kg in men and 0.423 kg/kg in women. These medians define balanced analytical strata for internal comparison and to preserve statistical power; they are not proposed diagnostic thresholds and should not be interpreted as clinical cutoffs.
Definition of MetS
MetS was defined according to the harmonized Joint Interim Statement criteria [15], adopted for comparability with prior KNHANES-based studies and applying abdominal-obesity thresholds validated for the Korean population [16], requiring ≥3 of: (1) abdominal obesity (waist circumference ≥90 cm in men or ≥85 cm in women); (2) elevated triglycerides (≥150 mg/dL or lipid-lowering medication); (3) reduced HDL cholesterol (<40 mg/dL in men or <50 mg/dL in women); (4) elevated blood pressure (systolic ≥130 or diastolic ≥85 mmHg, or antihypertensive medication); and (5) elevated fasting glucose (≥100 mg/dL or antidiabetic medication).
Covariates
Covariates included age (continuous), education, household income quartile, residential area (urban/rural), BMI ( < 25 or ≥25 kg/m2), smoking (never/former/current), aerobic activity (meeting WHO guidelines), and high-risk alcohol use (sex-specific thresholds). BMI was included in the primary model because it is strongly associated with MetS and to assess whether the ASM/BW association persisted after accounting for adiposity and body size as proxied by BMI. In BMI-subgroup analyses, BMI served as the stratification variable rather than a covariate.
Statistical analysis
All analyses were weighted for the KNHANES complex sampling design. Continuous variables are presented as weighted means ± SD and categorical variables as weighted percentages; differences according to MetS status were assessed using weighted t-tests and chi-square tests. Weighted MetS prevalence with standard errors was calculated for each of the four groups. Associations were assessed by complex-samples logistic regression estimating ORs and 95% CIs across three sequential models: Model 1 (age); Model 2 (plus education, income, area); Model 3 (plus BMI category, smoking, alcohol, physical activity). A linear trend test assigned ordinal scores (1–4) to the four groups. To assess independent contributions, ASM/BW and relative HGS were entered simultaneously as continuous variables in the fully adjusted joint model, and ORs per 1-SD increment were derived for standardized comparison. Subgroup analyses were conducted by age (40–64 vs ≥65 years) and BMI category, with interactions tested using Wald joint F-tests. Because ASM/BW and relative HGS share a body-weight denominator with the abdominal-obesity component of MetS, we performed prespecified sensitivity analyses: (i) redefining MetS after excluding the waist-circumference criterion (≥3 and ≥2 of the remaining four components); (ii) modeling absolute ASM and HGS, additionally adjusted for body weight; and (iii) restricted cubic spline (three-knot) models to assess dose-response and non-linearity. As additional sensitivity analyses, we estimated prevalence ratios using survey-weighted modified Poisson regression with robust variance, given the high MetS prevalence; refitted the fully adjusted models with BMI excluded; and additionally adjusted for total energy and protein intake. To examine whether the findings depended on the method used to normalize muscle mass, we compared ASM/BW, ASM/Ht², and ASM/BMI under the same covariate sets, with and without BMI adjustment; for ASM/BMI, the model without BMI adjustment was the primary specification because BMI is already incorporated into the exposure (Supplementary Table S1). Because ASM/BW incorporates body weight and BMI is also derived from body weight, potential collinearity was assessed by examining design-based correlations among the muscle indices and body-size measures, variance inflation factors, and the condition number of the design matrix. We also compared the standard error of the ASM/BW coefficient before and after inclusion of BMI (Supplementary Table S9). Because the four muscle mass–strength groups were defined using sex-specific medians, all analytical models were prespecified as sex-stratified; the sex-by-exposure interaction was tested in the same manner, and a pooled model of the total sample adjusting for sex was additionally fitted as a sensitivity analysis. Analyses used R 4.5.3 and the survey package, with significance at p < 0.05.
Results
Baseline characteristics
The 3033 participants (1276 men, 1757 women) had a mean age of 57.9 ± 11.1 years. The weighted MetS prevalence was 38.1% (46.9% in men, 29.8% in women). ASM/BW was 0.28 ± 0.02 kg/kg in men and 0.22 ± 0.02 kg/kg in women, and relative HGS was 0.56 ± 0.09 kg/kg in men and 0.43 ± 0.08 kg/kg in women. Compared with participants without MetS, those with MetS were older and had higher BMI and waist circumference, lower ASM/BW and relative HGS, and a higher proportion of men (Table 1).
Table 1.
Baseline characteristics of study participants according to the presence of metabolic syndrome (N = 3033, Weighted, KNHANES 2024)
| Variable | Total (N = 3033) | No MetS (n = 1869) | MetS (n = 1164) | P |
|---|---|---|---|---|
| Age (years) | 57.9 ± 11.1 | 56.1 ± 10.8 | 60.7 ± 10.9 | <0.001 |
| Sex, % | <0.001 | |||
| Male | 48.6 | 41.7 | 59.9 | |
| Female | 51.4 | 58.3 | 40.1 | |
| BMI (kg/m²) | 24.4 ± 3.6 | 23.0 ± 3.1 | 26.5 ± 3.2 | <0.001 |
| Waist circumference (cm) | 85.0 ± 10.4 | 80.5 ± 9.0 | 92.2 ± 8.2 | <0.001 |
| ASM/BW (kg/kg) | 0.249 ± 0.036 | 0.250 ± 0.037 | 0.246 ± 0.035 | 0.007 |
| ASM/Ht² (kg/m²) | 6.04 ± 1.19 | 5.76 ± 1.12 | 6.51 ± 1.16 | <0.001 |
| Relative HGS (kg/kg) | 0.491 ± 0.108 | 0.504 ± 0.108 | 0.470 ± 0.103 | <0.001 |
| Smoking status, % | <0.001 | |||
| Never | 57.6 | 63.2 | 48.6 | |
| Former | 26.1 | 22.7 | 31.7 | |
| Current | 16.3 | 14.1 | 19.7 | |
| High-risk drinking, % | 13.1 | 10.6 | 17.1 | <0.001 |
| Household income, % | 0.615 | |||
| Q1 (lowest) | 23.2 | 22.4 | 24.4 | |
| Q2 | 25.5 | 26.0 | 24.6 | |
| Q3 | 25.5 | 25.4 | 25.7 | |
| Q4 (highest) | 25.8 | 26.1 | 25.3 | |
| Education, % | <0.001 | |||
| ≤ Elementary | 13.6 | 9.7 | 19.9 | |
| Middle school | 11.2 | 9.4 | 14.1 | |
| High school | 32.9 | 33.8 | 31.4 | |
| ≥ University | 42.3 | 47.1 | 34.6 |
Data are presented as weighted mean ± standard deviation for continuous variables or weighted percentage (%) for categorical variables. All estimates account for the KNHANES complex sampling design. P-values were obtained using the design-based weighted t-test for continuous variables and the Rao–Scott chi-square test for categorical variables. Percentages for categorical variables are column-wise within each metabolic syndrome group
ASM, appendicular skeletal muscle mass; BW, body weight; Ht, height; HGS, handgrip strength; Relative HGS, handgrip strength divided by body weight; BMI, body mass index; MetS, metabolic syndrome
Association between muscle mass–strength groups and MetS
In both sexes, weighted MetS prevalence was lowest in the High ASM/BW + High relative HGS group and highest in the Low ASM/BW + Low relative HGS group (Fig. 2). This pattern persisted after multivariable adjustment, with the Low–Low group showing the highest odds of MetS (OR 3.04 in men and 2.42 in women; Table 2).
Fig. 2.
Weighted prevalence of metabolic syndrome by muscle mass–strength group in men (A) and women (B). Bars show weighted prevalence (%) with 95% CI; the horizontal dashed line indicates the weighted prevalence of the reference group (HM-HS). Groups were defined by sex-specific medians of ASM/BW and relative HGS. No covariate adjustment was applied. HM-HS, high ASM/BW + high relative HGS; HM-LS, high + low; LM-HS, low + high; LM-LS, low + low
Table 2.
Association between muscle mass–strength groups and metabolic syndrome
| Group | N (weighted N) | MetS % (SE) | Crude OR (95% CI) | Model 1 OR (95% CI) | Model 2 OR (95% CI) | Model 3 OR (95% CI) |
|---|---|---|---|---|---|---|
| Male (n = 1276) | ||||||
| HM-HS (ref)a | 423 (3,460,388) | 27.3 (2.3) | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) |
| HM-LS | 215 (1,666,088) | 47.6 (4.0) | 2.42 (1.62–3.62) | 2.40 (1.60–3.60) | 2.53 (1.68–3.81) | 1.60 (1.03–2.49) |
| LM-HS | 215 (1,748,552) | 49.5 (4.0) | 2.61 (1.76–3.90) | 2.61 (1.75–3.89) | 2.62 (1.71–4.00) | 2.11 (1.30–3.41) |
| LM-LS | 423 (3,066,509) | 67.2 (2.6) | 5.47 (4.00–7.50) | 5.36 (3.89–7.37) | 5.73 (4.14–7.93) | 3.04 (2.08–4.45) |
| p for trend < 0.001 | ||||||
| Female (n = 1757) | ||||||
| HM-HS (ref)a | 595 (3,759,419) | 13.1 (1.4) | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) |
| HM-LS | 284 (1,638,477) | 30.9 (2.9) | 2.97 (2.12–4.17) | 2.41 (1.67–3.47) | 2.40 (1.66–3.47) | 1.73 (1.20–2.48) |
| LM-HS | 284 (1,712,681) | 25.7 (2.9) | 2.30 (1.60–3.29) | 2.09 (1.46–3.00) | 2.18 (1.49–3.20) | 1.92 (1.30–2.85) |
| LM-LS | 594 (3,384,547) | 50.0 (2.2) | 6.64 (5.08–8.66) | 5.08 (3.84–6.71) | 5.14 (3.86–6.84) | 2.42 (1.78–3.30) |
| p for trend < 0.001 | ||||||
n, unweighted sample size; weighted N, estimated population size. MetS %, weighted prevalence (SE, standard error).
Model 1: adjusted for age (continuous)
Model 2: additionally adjusted for education level, household income, and residential area
Model 3: additionally adjusted for BMI category (<25 vs ≥25 kg/m²), smoking status, alcohol consumption, and physical activity
p for trend was assessed by entering the group variable as an ordinal continuous variable (1–4) in Model 3.
Survey-weighted logistic regression using complex sampling design
aHM-HS, high ASM/BW + high relative HGS; HM-LS, high ASM/BW + low relative HGS; LM-HS, low ASM/BW + high relative HGS; LM-LS, low ASM/BW + low relative HGS. Groups defined by sex-specific median values of ASM/BW and relative HGS. HM-HS served as the reference group
ASM, appendicular skeletal muscle mass; BW, body weight; HGS, handgrip strength; OR, odds ratio; CI, confidence interval
Independent contributions of ASM/BW and relative HGS
When analyzed individually, both ASM/BW and relative HGS were inversely associated with MetS in men and women (Table 3). In the joint model including both measures simultaneously, both associations remained significant. Per 1-SD increment, ASM/BW was associated with lower odds of MetS in men (OR 0.59, 95% CI 0.49–0.71) and women (OR 0.72, 95% CI 0.61–0.84), with corresponding ORs for relative HGS of 0.80 (0.67–0.94) and 0.75 (0.64–0.87), respectively.
Table 3.
Independent and joint associations of ASM/BW and relative HGS with metabolic syndrome (continuous models)
| Model | Variable | Male OR (95% CI) | p | Female OR (95% CI) | p |
|---|---|---|---|---|---|
| Model A: ASM/BW alone | |||||
| ASM/BW (per 0.01 kg/kg ↑) | 0.77 (0.71–0.83) | <0.001 | 0.82 (0.76–0.88) | <0.001 | |
| Model B: Relative HGS alone | |||||
| Relative HGS (per 0.01 kg/kg ↑) | 0.95 (0.94–0.97) | <0.001 | 0.95 (0.94–0.97) | <0.001 | |
| Model C: Joint model — both entered simultaneously | |||||
| ASM/BW (per 0.01 kg/kg ↑) | 0.79 (0.73–0.86) | <0.001 | 0.86 (0.80–0.92) | <0.001 | |
| ASM/BW (per 1-SD ↑) | 0.59 (0.49–0.71) | <0.001 | 0.72 (0.61–0.84) | <0.001 | |
| Relative HGS (per 0.01 kg/kg ↑) | 0.98 (0.96–0.99) | 0.009 | 0.97 (0.95–0.98) | <0.001 | |
| Relative HGS (per 1-SD ↑) | 0.80 (0.67–0.94) | 0.009 | 0.75 (0.64–0.87) | <0.001 | |
All models adjusted for age (continuous), education level, household income, residential area, BMI category (<25 vs ≥25 kg/m²), smoking status, alcohol consumption, and physical activity. Models A and B each entered ASM/BW and relative HGS separately into independent models. Model C (joint model) entered both indices simultaneously to assess their independent contributions after mutual adjustment. ORs per 0.01 kg/kg increment and per 1-SD increment are both derived from Model C; the per 1-SD metric enables standardized comparison of effect sizes between the two indices
All analyses used survey-weighted logistic regression accounting for the complex sampling design of KNHANES
ASM, appendicular skeletal muscle mass; BW, body weight; HGS, handgrip strength; OR, odds ratio; CI, confidence interval; SD, standard deviation; Relative HGS, absolute handgrip strength divided by body weight (kg/kg)
Subgroup analyses by age and BMI
In age- and BMI-stratified analyses, the overall pattern of higher MetS odds with lower muscle mass and strength was generally observed across strata (Fig. 3; Supplementary Table S2). The Low–Low group showed elevated odds of MetS across all age and BMI strata. However, there was no evidence of effect modification by age or BMI, as none of the group-by-age or group-by-BMI interaction tests was statistically significant (all p for interaction >0.05).
Fig. 3.
Subgroup analyses of the association between muscle mass–strength groups and metabolic syndrome in men (A) and women (B). Odds ratios (95% CI) for HM-LS, LM-HS, and LM-LS relative to HM-HS, stratified by age (40–64 vs ≥65 years) and BMI ( < 25 vs ≥25 kg/m²). No significant interaction was detected (all p > 0.05). *p < 0.05, **p < 0.01, ***p < 0.001
Associations with individual MetS components
ASM/BW was inversely associated with abdominal obesity, hypertriglyceridemia, elevated blood pressure, and elevated fasting glucose in both sexes, but not with low HDL cholesterol (Supplementary Table S3). Relative HGS was inversely associated with abdominal obesity in both sexes; an inverse association with elevated fasting glucose was also observed in women.
Sensitivity analyses
Several prespecified sensitivity analyses examined whether the associations were influenced by the shared body-weight denominator of the two indices or by the abdominal-obesity component of MetS (Supplementary Table S4). When MetS was redefined after excluding the waist-circumference criterion, ASM/BW remained inversely associated with MetS in both sexes, whereas the association with relative HGS was attenuated in men. Findings for ASM/BW were similar using a ≥-of-four threshold, where the association with relative HGS was attenuated in both sexes. When absolute ASM and HGS were modeled with additional 2 adjustment for body weight, absolute ASM was inversely associated with MetS once body weight was included, whereas absolute HGS did not show an independent association.
Restricted cubic spline analyses showed generally inverse associations of ASM/BW and relative HGS with predicted MetS probability, although the shapes of the associations varied by sex and index (Fig. 4). In analyses comparing alternative muscle-mass normalization methods, ASM/BW and ASM/BMI were inversely associated with MetS in both sexes, whereas ASM/Ht² showed a different pattern (Supplementary Table S1). Results were also directionally consistent when prevalence ratios were estimated using modified Poisson regression, when BMI was excluded from the fully adjusted model, and after additional adjustment for total energy and protein intake (Supplementary Tables S5–S7). Assessment of collinearity between ASM/BW and BMI showed no evidence of severe multicollinearity based on variance inflation factors, condition numbers, and changes in the standard error of the ASM/BW coefficient (Supplementary Table S9). The group-by-sex interaction was not statistically significant, and a similar pattern was observed in the pooled analysis adjusting for sex (Supplementary Table S8). Estimates were also similar across alternative sampling-weight specifications and in unweighted analyses (Supplementary Table S10).
Fig. 4.
Restricted cubic spline analysis of the association of ASM/BW (A, C) and relative HGS (B, D) with the predicted probability of metabolic syndrome in men (A, B) and women (C, D). Solid lines show adjusted predicted probabilities with 95% CI (shaded), modeled with three knots and adjusted for age, education, household income, residential area, BMI category, smoking, alcohol use, and physical activity at a fixed covariate profile; p-values denote tests for non-linearity. ASM/BW, appendicular skeletal muscle mass-to-body weight ratio; HGS, handgrip strength
Discussion
Using 2024 KNHANES national data, we found that muscle mass and strength are distinct but complementary factors associated with MetS. In joint models, DXA-derived ASM/BW and relative HGS were each independently associated with MetS in both sexes, and individuals with concurrently low muscle mass and strength had the highest odds of MetS. Because the study is cross-sectional, all associations reported here are correlational; the data cannot establish whether low muscle mass or strength precedes metabolic syndrome, results from it, or shares common antecedents.
Our results are consistent with previous studies reporting associations of ASM/BW and relative HGS with MetS [8, 17–19], which typically analyzed the two indicators separately or within broader risk-factor frameworks, often in sex-specific analyses [12], constraining direct comparison of their independent contributions. In community-dwelling older Chinese adults, Song et al. reported a graded inverse association of weight-normalized ASM with MetS, whereas weight-normalized grip strength showed no such graded association with overall MetS [20]. Our contribution is methodological rather than one of scale: concurrent DXA and grip-strength measurements in the same nationwide sample allowed the two constructs to be directly compared and mutually adjusted within a single analytical framework—an approach not feasible in studies relying on a single modality.
The associations of the two indices differed across sensitivity analyses. Both ASM/BW and relative HGS are normalized to body weight, which also informs the abdominal-obesity component of MetS, so neither index is fully independent of adiposity-related constructs. Our sensitivity analyses (see Sensitivity analyses; Supplementary Table S4) address this issue directly: the ASM/BW association persisted when MetS was defined without the waist criterion, and an inverse association between absolute ASM and MetS emerged after adjustment for body weight, indicating that the association persisted when ASM and body weight were modeled as separate terms rather than as a single ratio, whereas the relative-HGS association was attenuated in some analyses. ASM/BMI showed inverse associations with MetS similar to those for ASM/BW, whereas ASM/Ht² showed a different pattern (Supplementary Table S1). In our data, ASM/BW showed more consistent associations with MetS across the sensitivity analyses performed; however, the optimal method for normalizing muscle mass in relation to metabolic risk remains uncertain.
ASM/BW, rather than ASM/Ht², was selected as the primary indicator. Although ASM/Ht² is the international standard for diagnosing sarcopenia [21], previous Korean data showed that ASM/Ht² was positively correlated with BMI, visceral fat area, and HOMA-IR, whereas ASM normalized to body weight was inversely correlated with these measures [17]. A separate DXA-based comparison in Japanese adults similarly found that weight-adjusted ASM was more closely related to cardiometabolic risk factors, whereas height-adjusted ASM better identified bone-related characteristics [13]. Longitudinal Korean data have also shown that higher weight-normalized ASM and increases in this measure over time were associated with a lower incidence of MetS [22]. We acknowledge that current sarcopenia consensus guidelines, including the Asian Working Group for Sarcopenia [23] and the Korean Working Group on Sarcopenia [24], recommend ASM adjusted for height² or BMI for diagnosing sarcopenia; our aim, however, was not to diagnose sarcopenia but to examine muscle mass in relation to metabolic risk. ASM/BW should not be interpreted as a measure that adjusts for or offsets adiposity; its association with MetS is best understood as an association of a relative body-composition phenotype with metabolic risk rather than as the isolated contribution of muscle mass. In our analysis, ASM/Ht² showed no association in men and a positive association in women, while ASM/BW and ASM/BMI showed consistent inverse associations in both sexes (Supplementary Table S1). This divergence likely reflects that the indices capture different body-size and body-composition constructs, rather than indicating that one index is inherently more valid than another. In this sample, ASM/Ht² was strongly correlated with BMI (r = 0.80 in men and r = 0.78 in women; Supplementary Table S9), suggesting that height-normalized ASM may be more closely related to overall body size. Because DXA-based body composition can differ across racial and ethnic groups [14], the interpretation, performance, and any proposed cutoffs of these indices should be evaluated before extrapolation beyond Korean populations.
Retaining BMI in the primary model requires explicit justification because BMI and ASM/BW are both functions of body weight. BMI was included because it is strongly associated with MetS in this sample (Table 1) and because we sought to examine whether the association between ASM/BW and MetS persisted after accounting for overall adiposity and body size as proxied by BMI. However, because the two measures share body weight, BMI adjustment cannot fully disentangle relative muscle mass from adiposity or body-size normalization. The BMI-adjusted and BMI-unadjusted models should therefore be interpreted as addressing related but distinct estimands: the former compares participants within similar BMI categories but with different ASM/BW, whereas the latter describes the association between the overall relative muscle-mass/body-size phenotype and MetS. Associations were directionally consistent but larger without BMI adjustment (Low ASM/BW + Low relative HGS group: men, OR 5.58, 95% CI 3.98–7.80; women, OR 5.25, 95% CI 3.94–6.99; Supplementary Table S6). This consistency does not establish whether BMI functions as a confounder, mediator, or component of the body-size-normalization construct. Likewise, the absence of severe numerical multicollinearity in formal diagnostics (Supplementary Table S9) does not resolve this epidemiological and interpretative issue.
Associations of ASM/BW and relative HGS with MetS were observed in both sexes. Among women, the weighted MetS prevalence was higher in the High ASM/BW + Low relative HGS group (30.9%) than in the Low ASM/BW + High relative HGS group (25.7%), a pattern not observed in men. However, the group-by-sex interaction was not statistically significant, providing no statistical evidence that the associations differed by sex; apparent differences between sex-specific estimates should therefore be interpreted cautiously.
These associations are biologically plausible and showed different patterns across individual MetS components. ASM/BW was inversely associated with four MetS components in both sexes, whereas relative HGS was inversely associated with abdominal obesity in both sexes and with fasting glucose in women. Skeletal muscle is the dominant site of insulin-stimulated glucose disposal, and low muscle mass has been linked to impaired glucose and lipid homeostasis through insulin resistance, reduced basal metabolic rate, and altered myokine secretion [25–28]. Muscle strength, tied to habitual physical activity, may act through energy expenditure, fat distribution, and contraction-stimulated glucose uptake [29, 30]. The absence of an association with low HDL cholesterol, which is influenced by multiple behavioral and genetic factors [31], is consistent with this heterogeneity.
The BMI-stratified results showed generally similar patterns across BMI categories. The Low–Low group had elevated MetS odds in both BMI strata and in both sexes, including among participants with BMI < 25 kg/m², consistent with the concept that metabolic abnormalities may occur despite a BMI below the obesity threshold [32, 33]. Consistent with the 2022 ESPEN–EASO consensus, which emphasizes combined assessment of body composition and muscle function in obesity [34], our analysis incorporated both relative muscle mass and strength. Age-stratified analyses showed a similar picture, with the Low–Low group having elevated MetS odds in both age strata, although fewer associations reached statistical significance among adults aged ≥65 years. However, none of the group-by-age or group-by-BMI interactions was statistically significant, and differences in statistical significance between strata should not be interpreted as evidence that the associations differed by age or BMI category.
These findings may have practical implications for metabolic risk assessment beyond BMI alone, a consideration of particular relevance in Asian populations [18, 35]. Where DXA is unavailable, relative HGS could, in principle, serve as a low-cost preliminary measure within a stepwise approach. However, we did not develop or validate a prediction model and did not assess discrimination, calibration, or external validity, so any screening application is hypothesis-generating and would require dedicated, model-based evaluation before routine clinical use. Integrating measures of muscle mass and strength into metabolic risk assessment may help identify individuals whose risk is not captured by BMI alone, but this requires prospective evaluation.
This study has several strengths: a large sample from a nationally representative survey, analyzed with complex sampling weights; concurrent assessment of DXA-derived muscle mass and grip strength within the same sample, enabling direct comparison and mutual adjustment of both constructs; analysis of both independent and combined associations; and multi-perspective evaluation through age- and BMI-stratified analyses with interaction testing.
Limitations should be acknowledged. First, because ASM/BW and relative HGS are normalized to body weight, their associations are not fully independent of adiposity-related constructs. Different muscle-mass normalization methods also yielded different patterns, and the optimal method for normalizing muscle mass in relation to metabolic outcomes remains uncertain. Because ASM/BW and BMI are both functions of body weight, no single model specification fully separates relative muscle mass from body-size normalization. Second, the cross-sectional design precludes causal inference. Third, DXA was performed in only a subset of participants (71.4% of those aged ≥40 years) and KNHANES 2024 provides no DXA-specific sampling weight; participants who underwent DXA were younger and had a lower prevalence of MetS than those who did not, although BMI and waist circumference did not differ (Supplementary Table S10), so selection bias cannot be excluded. Fourth, insulin was not measured in KNHANES 2024, precluding HOMA-IR and limiting mechanistic interpretation. Fifth, dichotomizing by sex-specific medians aids interpretation but does not represent clinical cutoffs, although analyses using continuous measures showed similar patterns. Sixth, lifestyle variables were self-reported; although additional adjustment for total energy and protein intake did not materially alter the associations (Supplementary Table S7), residual confounding from other unmeasured dietary or inflammatory factors cannot be excluded. Finally, strength assessment was limited to grip strength, and DXA-based body composition may vary across ethnicities [14], so cross-population extrapolation warrants caution.
Conclusion
In Korean adults aged ≥40 years, lower relative muscle mass and grip strength were associated with higher odds of MetS, with the highest odds observed among individuals with concurrently low levels of both measures. ASM/BW showed more consistent associations with MetS across the sensitivity analyses performed in this study, although associations differed according to the method used to normalize muscle mass and the optimal normalization method remains uncertain. These findings suggest that the two measures may provide complementary information regarding MetS. Because of the cross-sectional design, these findings do not establish temporality or causation. Whether combined assessment of muscle mass and strength improves metabolic risk identification beyond conventional measures such as BMI requires prospective, model-based evaluation.
Supplementary information
Acknowledgements
We thank the KDCA for providing access to KNHANES data.
Abbreviations
- ASM
Appendicular skeletal muscle mass
- ASM/BW
ASM-to-body weight ratio
- ASM/Ht2
ASM divided by height squared
- BMI
Body mass index
- CI
Confidence interval
- DXA
Dual-energy X-ray absorptiometry
- HDL
High-density lipoprotein
- HGS
Handgrip strength
- HOMA-IR
Homeostatic model assessment of insulin resistance
- KDCA
Korea Disease Control and Prevention Agency
- KNHANES
Korea National Health and Nutrition Examination Survey
- MetS
Metabolic syndrome
- OR
Odds ratio
- SD
Standard deviation
Author contributions
E.Y.L. and H.M.K. conceptualized and designed the study, curated the data, and conducted the investigation. E.Y.L. performed the formal analysis, developed the methodology, prepared the figures, and wrote the original draft. H.M.K. supervised the study, validated the results, and reviewed and edited the manuscript. Both authors read and approved the final manuscript.
Funding
Open Access funding enabled and organized by Chung-Ang University Hospital.
Data availability
The data analyzed in this study are derived from the Korea National Health and Nutrition Examination Survey (KNHANES), a publicly available dataset provided by the Korea Disease Control and Prevention Agency (KDCA). The data can be accessed from the KNHANES website (https://knhanes.kdca.go.kr) upon registration and submission of a data-use request.
Declarations
Ethics approval
The KNHANES protocol was approved by the KDCA Institutional Review Board (No. 2022-11-16-R-03), and all participants provided written informed consent. This secondary analysis was exempt from review by the Institutional Review Board of Chung-Ang University Hospital (IRB No. 2604-001-19620).
Declaration of generative AI
During preparation of this work, the authors used Claude (Anthropic) and ChatGPT (OpenAI) to assist in manuscript drafting and language editing. The authors reviewed and edited all content and take full responsibility for the published article.
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.
Supplementary information
The online version contains supplementary material available at https://doi.org/10.1007/s12020-026-04804-3.
References
- 1.S. Mottillo, K.B. Filion, J. Genest et al. The metabolic syndrome and cardiovascular risk: a systematic review and meta-analysis. J. Am. Coll. Cardiol. 56(14), 1113–1132 (2010). 10.1016/j.jacc.2010.05.034 [DOI] [PubMed] [Google Scholar]
- 2.J.J. Noubiap, J.R. Nansseu, U.F. Nyaga et al. Worldwide trends in metabolic syndrome from 2000 to 2023: a systematic review and modelling analysis. Nat. Commun. 17, 573 (2026). 10.1038/s41467-025-67268-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.D. Park, M.-J. Shin, J.-P. Després, R.H. Eckel, J. Tuomilehto, S. Lim, 20-year trends in metabolic syndrome among Korean adults from 2001 to 2020. JACC Asia 3(3), 491–502 (2023). 10.1016/j.jacasi.2023.02.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.R.A. DeFronzo, D. Tripathy, Skeletal muscle insulin resistance is the primary defect in type 2 diabetes. Diabetes Care 32(Suppl 2), S157–S163 (2009). 10.2337/dc09-s302 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.P. Srikanthan, A.S. Karlamangla, Relative muscle mass is inversely associated with insulin resistance and prediabetes: findings from the Third National Health and Nutrition Examination Survey. J. Clin. Endocrinol. Metab. 96(9), 2898–2903 (2011). 10.1210/jc.2011-0435 [DOI] [PubMed] [Google Scholar]
- 6.H. Zhang, S. Lin, T. Gao et al. Association between sarcopenia and metabolic syndrome in middle-aged and older non-obese adults: a systematic review and meta-analysis. Nutrients 10(3), 364 (2018). 10.3390/nu10030364 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.D.P. Leong, K.K. Teo, S. Rangarajan et al. Prognostic value of grip strength: findings from the Prospective Urban Rural Epidemiology (PURE) study. Lancet 386(9990), 266–273 (2015). 10.1016/S0140-6736(14)62000-6 [DOI] [PubMed] [Google Scholar]
- 8.D. Yi, A.R. Khang, H.W. Lee, S.M. Son, Y.H. Kang, Relative handgrip strength as a marker of metabolic syndrome: the Korea National Health and Nutrition Examination Survey (KNHANES) VI (2014–2015). Diabetes Metab Syndr Obes 11, 227–240 (2018). 10.2147/dmso.s166875 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.C. Ji, Y. Xia, S. Tong, Q. Wu, Y. Zhao, Association of handgrip strength with the prevalence of metabolic syndrome in US adults: the National Health and Nutrition Examination Survey. Aging (Albany NY) 12(9), 7818–7829 (2020). 10.18632/aging.103097 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.J.Y. Byeon, M.K. Lee, M.-S. Yu et al. Lower relative handgrip strength is significantly associated with a higher prevalence of the metabolic syndrome in adults. Metab. Syndr. Relat. Disord. 17(5), 280–288 (2019). 10.1089/met.2018.0111 [DOI] [PubMed] [Google Scholar]
- 11.A.J. Cruz-Jentoft, G. Bahat, J. Bauer et al. Sarcopenia: revised European consensus on definition and diagnosis. Age Ageing 48(1), 16–31 (2019). 10.1093/ageing/afy169 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.E. Atlantis, S.A. Martin, M.T. Haren, A.W. Taylor, G.A. Wittert, Inverse associations between muscle mass, strength, and the metabolic syndrome. Metabolism 58(7), 1013–1022 (2009). 10.1016/j.metabol.2009.02.027 [DOI] [PubMed] [Google Scholar]
- 13.T. Furushima, M. Miyachi, M. Iemitsu et al. Comparison between clinical significance of height-adjusted and weight-adjusted appendicular skeletal muscle mass. J. Physiol. Anthropol. 36(1), 15 (2017). 10.1186/s40101-017-0130-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.M.N.M. Blue, G.M. Tinsley, E.D. Ryan, A.E. Smith-Ryan, Validity of body-composition methods across racial and ethnic populations. Adv. Nutr. 12(5), 1854–1862 (2021). 10.1093/advances/nmab016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.K.G.M.M. Alberti, R.H. Eckel, S.M. Grundy et al. Harmonizing the metabolic syndrome: a joint interim statement of the International Diabetes Federation Task Force on Epidemiology and Prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; and International Association for the Study of Obesity. Circulation 120(16), 1640–1645 (2009). 10.1161/CIRCULATIONAHA.109.192644 [DOI] [PubMed] [Google Scholar]
- 16.S.Y. Lee, H.S. Park, D.J. Kim et al. Appropriate waist circumference cutoff points for central obesity in Korean adults. Diabetes Res. Clin. Pract. 75(1), 72–80 (2007). 10.1016/j.diabres.2006.04.013 [DOI] [PubMed] [Google Scholar]
- 17.S. Lim, J.H. Kim, J.W. Yoon et al. Sarcopenic obesity: prevalence and association with metabolic syndrome in the Korean Longitudinal Study on Health and Aging (KLoSHA). Diabetes Care 33(7), 1652–1654 (2010). 10.2337/dc10-0107 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.J.H. Moon, S.R. Choo, J.S. Kim, Relationship between low muscle mass and metabolic syndrome in elderly people with normal body mass index. J. Bone Metab. 22(3), 99–106 (2015). 10.11005/jbm.2015.22.3.99 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Y. Wen, T. Liu, C. Ma et al. Association between handgrip strength and metabolic syndrome: a meta-analysis and systematic review. Front. Nutr. 9, 996645 (2022). 10.3389/fnut.2022.996645 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.P. Song, P. Han, Y. Zhao et al. Muscle mass rather than muscle strength or physical performance is associated with metabolic syndrome in community-dwelling older Chinese adults. BMC Geriatr 21(1), 191 (2021). 10.1186/s12877-021-02143-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.L.K. Chen, J. Woo, P. Assantachai et al. Asian Working Group for Sarcopenia: 2019 consensus update on sarcopenia diagnosis and treatment. J. Am. Med. Dir. Assoc. 21(3), 300–307.e2 (2020). 10.1016/j.jamda.2019.12.012 [DOI] [PubMed] [Google Scholar]
- 22.G. Kim, S.E. Lee, J.E. Jun et al. Increase in relative skeletal muscle mass over time and its inverse association with metabolic syndrome development: a 7-year retrospective cohort study. Cardiovasc. Diabetol. 17(1), 23 (2018). 10.1186/s12933-018-0659-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.L.K. Chen, J. Woo, P. Assantachai et al. A focus shift from sarcopenia to muscle health in the Asian Working Group for Sarcopenia 2025 Consensus Update. Nat. Aging 5(11), 2164–2175 (2025). 10.1038/s43587-025-01004-y [DOI] [PubMed] [Google Scholar]
- 24.J.Y. Baek, H.W. Jung, K.M. Kim et al. Korean Working Group on Sarcopenia Guideline: Expert Consensus on Sarcopenia Screening and Diagnosis by the Korean Society of Sarcopenia, the Korean Society for Bone and Mineral Research, and the Korean Geriatrics Society. Ann. Geriatr. Med. Res. 27(1), 9–21 (2023). 10.4235/agmr.23.0009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.G. Kim, J.H. Kim, Impact of skeletal muscle mass on metabolic health. Endocrinol. Metab. (Seoul) 35(1), 1–6 (2020). 10.3803/enm.2020.35.1.1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.H.S. Chung, K.M. Choi, Exercise, exerkines, and sarcopenia. Diabetes Metab. J. 50(1), 19–29 (2026). 10.4093/dmj.2025.0972 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.C.W. Li, K. Yu, N. Shyh-Chang et al. Pathogenesis of sarcopenia and the relationship with fat mass: descriptive review. J. Cachexia Sarcopenia Muscle 13(2), 781–794 (2022). 10.1002/jcsm.12901 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.R. Balakrishnan, D.C. Thurmond, Mechanisms by which skeletal muscle myokines ameliorate insulin resistance. Int. J. Mol. Sci. 23(9), 4636 (2022). 10.3390/ijms23094636 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Z.P. Rostron, R.A. Green, M. Kingsley, A. Zacharias, Associations between measures of physical activity and muscle size and strength: a systematic review. Arch. Rehabil. Res. Clin. Transl. 3(3), 100124 (2021). 10.1016/j.arrct.2021.100124 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.E.A. Richter, M. Hargreaves, Exercise, GLUT4, and skeletal muscle glucose uptake. Physiol. Rev. 93(3), 993–1017 (2013). 10.1152/physrev.00038.2012 [DOI] [PubMed] [Google Scholar]
- 31.R.C. Ellison, Y. Zhang, M.M. Qureshi et al. Lifestyle determinants of high-density lipoprotein cholesterol: the National Heart, Lung, and Blood Institute Family Heart Study. Am. Heart J. 147(3), 529–535 (2004). 10.1016/j.ahj.2003.10.033 [DOI] [PubMed] [Google Scholar]
- 32.N. Ruderman, D. Chisholm, X. Pi-Sunyer, S. Schneider, The metabolically obese, normal-weight individual revisited. Diabetes 47(5), 699–713 (1998). 10.2337/diabetes.47.5.699 [DOI] [PubMed] [Google Scholar]
- 33.R. Williams, M. Periasamy, Genetic and environmental factors contributing to visceral adiposity in Asian populations. Endocrinol. Metab. (Seoul) 35(4), 681–695 (2020). 10.3803/enm.2020.772 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.L.M. Donini, L. Busetto, S.C. Bischoff et al. Definition and diagnostic criteria for sarcopenic obesity: ESPEN and EASO consensus statement. Obes. Facts 15(3), 321–335 (2022). 10.1159/000521241 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Y.C. Hwang, I.J. Cho, I.K. Jeong, K.J. Ahn, H.Y. Chung, Differential association between sarcopenia and metabolic phenotype in Korean young and older adults with and without obesity. Obesity (Silver Spring) 25(1), 244–251 (2017). 10.1002/oby.21694 [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 data analyzed in this study are derived from the Korea National Health and Nutrition Examination Survey (KNHANES), a publicly available dataset provided by the Korea Disease Control and Prevention Agency (KDCA). The data can be accessed from the KNHANES website (https://knhanes.kdca.go.kr) upon registration and submission of a data-use request.