Abstract
Background and Objectives
Type 2 diabetes (T2D) stems from metabolic dysregulation, and body cell mass (BCM) comprises the principal physiological substrate of basal metabolic rate (BMR); however, the association between BCM and diabetes is still unclear. This study examined the relationship of percentage body cell mass (PBCM) with insulin resistance (IR) and T2D, and evaluated the mediators of this association.
Methods and Study Design
This study utilized data from U.S. adults in the National Health and Nutrition Survey (1999–2004). Weighted multivariate regression, restricted cubic spline analysis, subgroup analyses, and sensitivity analysis were employed to investigate the relationship of PBCM with T2D risk biomarkers, IR and T2D. The potential mediating role of body mass index (BMI) was explored.
Results
A total of 2,923 participants aged 18 years or older were included. Preliminary analysis demonstrated an inverse association between PBCM and biomarkers of T2D risk. Further analyses revealed that higher PBCM was associated with lower odds of IR (OR = 0.91, 95% CI: 0.88–0.94, p <0.001) and T2D (OR = 0.89, 95% CI: 0.84–0.94, p <0.001) in the fully adjusted model. Mediation analysis indicated that BMI mediated 70.6% (95% CI: 55.1%–88.9%, p <0.001) of the total effect of PBCM on IR, and 29.9% (95% CI: 14.2%–65.9%, p <0.001) of that on T2D, respectively.
Conclusions
PBCM was negatively associated with the prevalence of IR and T2D. Mediation analyses demonstrated a mediating effect of BMI, suggesting that PBCM might be related to IR and T2D, potentially mediated by BMI.
Key Words: body mass index, insulin resistance, mediation analysis, percentage body cell mass, type 2 diabetes
Introduction
Type 2 diabetes (T2D), a chronic metabolic disorder, has evolved into an increasingly severe global health crisis. Insulin resistance (IR) is one of the most important pathophysiological factors driving disease progression and complications in T2D.1,2 Identifying modifiable factors associated with IR and T2D is therefore crucial for prevention strategies. The most prominent risk factor for IR and T2D is obesity.3,4 Body mass index (BMI) serves as a commonly employed diagnostic tool for overweight and obesity, with relevant research demonstrating its significant correlation with T2D onset and progression.5,6 However, BMI is a crude proxy that fails to capture detailed body composition or fat distribution.7,8
Imaging techniques, such as computed tomography (CT) and dual-energy X-ray absorptiometry (DXA), can provide precise body composition assessment. However, their high cost, risk of ionizing radiation exposure, and lack of feasibility limit their use in large-scale epidemiological studies.9 In recent years, bioelectrical impedance analysis (BIA) has gained increasing attention due to its simple operation, high safety, non-invasiveness, and low cost, making it applicable to both clinical practice and large-scale epidemiological investigations.10 BIA employs various parameters, including body cell mass (BCM), total body water (TBW), extracellular water (ECW), intracellular fluid (ICF), fat-free mass (FFM), and fat mass (FM) (Figure 1). In BIA parameters, both FFM and FM were associated with T2D, with low FFM and excess FM, particularly visceral adiposity, most frequently regarded as risk factors for IR and T2D.4,8,11,12 BCM, comprising all metabolically active tissues within the organism, the sum of intracellular water and protein, has been proposed as a comprehensive marker of muscle quantity, quality, and nutritional status. BCM was a strong and independent risk factor for mortality, BCM assessment might provide more useful prognostic in-formation for clinicians.13 Percentage body cell mass (PBCM), defined as BCM relative to body weight, could more accurately reflect the bioactive components in the body.14 However, there are relatively few studies on the association between PBCM and IR or T2D.
Figure 1.
Bioelectrical impedance analysis (BIA) for measuring body composition
Therefore, utilizing the nationally representative, large-scale National Health and Nutrition Examination Survey (NHANES) dataset, the objective of this study was to investigate the associations of PBCM with IR and T2D. We hypothesized that PBCM may be inversely associated with IR and T2D. Furthermore, the study examined the mediating role of BMI in these associations.
Methods
Data sources
Data were collected from publicly available datasets spanning three consecutive NHANES cycles between 1999 and 2004. The NHANES is a continuous cross-sectional observational study designed to gather health data from a nationally representative sample of the non-institutionalized U.S. population. The survey employs a complex, multi-stage probability cluster sampling design to ensure comprehensive and reliable data collection. The NHANES study protocol (#98-12) was approved by the National Center for Health Statistics (NCHS) Institutional Review Board (IRB), with written informed consent obtained from all participants. Our secondary analysis adhered to STROBE guidelines and did not require additional IRB approval. Detailed methodological and ethical documentation regarding NHANES is accessible through the Centers for Disease Control and Prevention (CDC) and NCHS websites (https://www.cdc.gov/nchs/nhanes/about/erb.html). All participant data were accessed for research purposes from 1 December 2025 to 27 February 2026. We had access to information that did not identify individual participants during or after data collection.
Study design and population
Among 31,126 participants, we excluded minors aged <18 years (n=14,065), pregnant individuals (n=927), participants with missing data on T2D diagnosis (n=2), fasting plasma glucose (FPG) (n=9,197), glycohemoglobin (HbA1c) (n=6), fasting serum insulin (FSI) (n=120), body composition (n=3,843), body weight (n=1), or dietary sample weights data (n=42). Finally, 2,923 participants were included in this study (Figure 2).
Figure 2.
Flowchart of participants selection
Exposure, outcome, and mediator
PBCM (%) was calculated as BCM (kg) divided by body weight (kg) and multiplied by 100, representing the percentage of body cell mass relative to total body weight.14 BCM was calculated as ICF (L) / 0.70 (L/kg), where 0.70 is the mean hydration of BCM.15 ICF was measured using the Hydra ECF/ICF Bioimpedance Spectroscopy Analyzer (Model 4200; Xitron Technologies, San Diego, CA, USA). Measurements were performed by trained medical technicians in accordance with the operational guidelines stipulated in the NHANES Body Composition Procedures Manual. This multi-frequency analyzer employs 12-bit digital signal processing to measure impedance at 50 logarithmically spaced frequency points ranging from 5 kHz to 1 MHz. The raw spectral data were fitted to the Cole biophysical model using non-linear least-squares curve fitting. Subsequently, ICF volume was predicted using the RE and RI parameters derived from the Cole model, in conjunction with equations based on the Hanai mixture theory. Body weight was measured at the Mobile Examination Center (MEC).
Outcome variables comprised the prevalence of IR and T2D, along with T2D risk markers including FPG (mmol/L), HbA1c (%), FSI (pmol/L), and homeostasis model assessment of insulin resistance (HOMA-IR). FPG, HbA1c, and FSI data were obtained from laboratory test results. HOMA-IR was calculated as (FPG [mmol/L] × FSI [µU/mL]) / 22.5.16,17 Based on prior research, a HOMA-IR value exceeding 2.6 defines IR in the US general population, which was adopted as the diagnostic criterion in our study.17,18 Individuals were considered to have T2D if they met one or more of the following criteria: (1) FPG ≥7.0 mmol/L; (2) HbA1c ≥6.5%; (3) taking diabetic pills or insulin; and (4) self-reported physician diagnosis of T2D.19-21
BMI (kg/m²) was calculated as weight (kg) divided by the square of height (m). Body weight and height were measured at MECs.
Covariates
The selection of covariates was based on established clinical relevance and prior research findings.17,20,20 Age, gender, race, educational level, marital status, family poverty-income ratio (PIR), smoking status, drinking status, physical activity, hypertension, coronary heart disease, stroke, and kidney disease were obtained from demographic data and questionnaires in the NHANES survey.23,24 Race was categorized as non-Hispanic white, non-Hispanic black, Mexican American, and others. Education level was categorized as <9 years, 9-12 years, and >12 years. Marital status was classified as married or cohabiting, and solitary living. Family PIR was categorized as <1.3, 1.3-3.5, and >3.5. Smoking status was categorized as never (< 100 cigarettes smoked in a lifetime), former (quit smoking after smoking ≥100 cigarettes), and current (smoking ≥100 cigarettes and smoking some days or every day now). Drinking status was defined by consuming ≥12 alcoholic drinks in any given year. Physical activity was categorized as sedentary, moderate (≥10 minutes of light to moderate exercise in the past 30 days), and vigorous (≥10 minutes of heavy exercise in the past 30 days). Waist circumference was measured at MECs. ECF was measured by BIA, TBW was calculated as ECW + ICF, and percent body fat (PFAT) was calculated using the hybrid model. Energy, carbohydrate, protein, and total fat intakes were obtained through 24-hour dietary recall interviews.23 Hypertension, coronary heart disease, stroke, and kidney disease were identified based on self-reported prior clinical diagnoses. Triglycerides (TG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-c), low-density lipoprotein cholesterol (LDL-c), alanine aminotransferase (ALT), creatinine (Cr), and uric acid (UA) were obtained from laboratory test results.24 Dyslipidemia was defined as having any one of the following: TG ≥3.89 mmol/L, TC ≥5.18 mmol/L, LDL-c ≥3.37 mmol/L, or HDL-c <1.04 mmol/L (males) or 1.30 mmol/L (females).25
Statistical analysis
Dietary sampling weights were applied to account for the complex survey design, as recommended in NHANES guidelines, to generate nationally representative estimates. The weighting procedure was as follows: 1999–2002 data were weighted at two-thirds of the four-year dietary sampling weights, and 2003–2004 data at one-third of the two-year dietary sampling weights. Baseline characteristics for categorical variables are presented as unweighted counts and weighted percentages, while continuous variables are summarized as means with standard deviations (SD) or medians with interquartile ranges (IQR). To analyze differences in continuous variables, weighted regression models were employed, whereas differences in categorical variables were assessed using weighted chi-square tests. Potential multicollinearity among covariates was evaluated using the variance inflation factor (VIF); a VIF ≥ 5 was considered indicative of significant multicollinearity. Missing values in covariates were handled using multiple imputation by chained equations (MICE) using the R mice package, generating five multiply imputed datasets to maximize statistical efficiency and minimize bias.
According to the STROBE statement,26 unadjusted, minimally adjusted, and fully adjusted models were employed in this study. Model 1 was unadjusted for covariates; model 2 was adjusted for age, gender, race, and NHANES cycle; model 3 was additionally adjusted for physical activity, smoking status, drinking status, energy intakes, hypertension, and dyslipidemia. To evaluate the association of PBCM with T2D risk markers, IR, and T2D, we applied weighted multivariable linear or logistic regression models and computed β values or odds ratios (ORs) with corresponding 95% confidence intervals (CIs) across three models. Additionally, we employed complex sampling-weighted restricted cubic spline (RCS) analysis to examine potential non-linear associations between them. Then, sensitivity analyses were conducted to verify the stability of the results after excluding participants with missing covariate data. Furthermore, mediation analyses were performed using the Sobel test, bootstrap resampling, and the quasi-Bayesian Monte Carlo method with 1,000 simulations based on the normal approximation. Subsequently, stratified analyses were conducted to test for interaction while accounting for potential confounders across predefined subgroups, including age (<35 and ≥35), gender, physical activity, smoking status, drinking status, hypertension, and dyslipidemia. Weighted multivariable logistic regression was employed in the stratified analyses. If the p-value for interaction was not statistically significant, the results across different strata were considered robust; otherwise, potential heterogeneity may exist in specific subpopulations.
Data analysis was performed using R software (version 4.2.1; https://www.r-project.org/) and Free Statistics software (version 2.3; Beijing Free Clinical Medical Technology Co., Ltd.). Statistical significance was defined as a two-tailed p-value < 0.05.
Results
Characteristics of the population
The baseline characteristics of enrolled participants grouped by PBCM are presented in Table 1. Among 2,923 participants (weighted sample size of 42,885,440), the mean (SD) age was 31.5 (10.1) years; 1,544 (52.4%) were men; 1,462 (44.8%) had insulin resistance; and 115 (3.4%) had diabetes. Compared with the individuals in the lower PBCM group, those in the higher PBCM group tended to be younger, men, solitary living, drinkers, and taller, engage in vigorous physical activity, have higher TBW, ICF, and ECW, have higher nutrient intakes (energy, carbohydrate, protein, and total fat), and have higher levels of Cr and UA, whereas PFAT, HbA1c, FSI, and HOMA-IR were lower (all p <0.05). Notably, participants with higher PBCM levels exhibited significantly lower BMI, prevalence of IR and T2D (all p <0.001).
Table 1.
Characteristics of the study population†
| Characteristics | PBCM (%) | |||||
|---|---|---|---|---|---|---|
|
| ||||||
| Total | Q1 (<36) | Q2 (36–41) | Q3 (41–48) | Q4 (=48) | p value | |
| Unweighted sample size, n | 2923 | 730 | 730 | 731 | 732 | |
| Weighted sample size, n (%) | 42,885,440 | 10,738,476 | 11,046,147 | 11,569,195 | 9,531,622 | |
| Age (years), Mean (SD) | 31.5 (10.1) | 33.0 (10.2) | 32.2 (10.2) | 31.9 (9.9) | 28.8 (9.6) | <0.001 |
| Gender, n (%) | <0.001 | |||||
| Men | 1544 (52.4) | 27 (3.2) | 217 (31.9) | 592 (81.1) | 708 (96.9) | |
| Women | 1379 (47.6) | 730 (96.8) | 513 (68.1) | 139 (18.9) | 23 (3.1) | |
| Race | 0.011 | |||||
| Non-Hispanic white | 1193 (68.7) | 278 (67.2) | 324 (71.0) | 313 (69.8) | 278 (66.2) | |
| Non-Hispanic black | 678 (12.6) | 197 (16.9) | 156 (11.1) | 144 (10.2) | 181 (12.2) | |
| Mexican American | 802 (9.1) | 205 (9.0) | 187 (7.9) | 211 (9.7) | 199 (10.0) | |
| Others | 250 (9.6) | 51 (6.9) | 63 (9.9) | 63 (10.3) | 73 (11.6) | |
| Education level (years), n (%) | 0.034 | |||||
| <9 | 227 (4.8) | 50 (3.2) | 61 (5.5) | 63 (4.8) | 53 (5.7) | |
| 9-12 | 971 (39.7) | 239 (40.3) | 234 (35.6) | 250 (39.4) | 248 (44.2) | |
| >12 | 1146 (55.6) | 326 (56.5) | 302 (58.9) | 289 (55.8) | 229 (50.1) | |
| Marital status, n (%) | <0.001 | |||||
| Married or cohabiting | 1420 (57.2) | 377 (63.8) | 388 (62.3) | 370 (53.8) | 285 (48.0) | |
| Solitary living | 1416 (42.8) | 327 (36.2) | 318 (37.7) | 341 (46.2) | 430 (52.0) | |
| Family PIR, n (%) | 0.489 | |||||
| <1.3 | 834 (23.2) | 219 (26.0) | 205 (21.8) | 221 (23.6) | 189 (21.3) | |
| 1.3-3.5 | 1011 (34.6) | 248 (36.5) | 251 (32.8) | 243 (35.2) | 269 (33.6) | |
| >3.5 | 859 (42.2) | 205 (37.5) | 222 (45.3) | 213 (41.2) | 219 (45.1) | |
| Physical activity, n (%) | <0.001 | |||||
| Sedentary | 972 (30.4) | 283 (36.2) | 256 (30.7) | 238 (30.2) | 195 (23.6) | |
| Moderate | 640 (24.7) | 196 (29.1) | 178 (28.6) | 157 (24.6) | 109 (15.3) | |
| Vigorous | 1308 (44.9) | 252 (34.7) | 294 (40.7) | 335 (45.2) | 427 (61.0) | |
| Smoking status, n (%) | 0.001 | |||||
| Never | 1316 (54.6) | 384 (61.1) | 343 (53.3) | 309 (51.0) | 280 (52.8) | |
| Former | 359 (16.8) | 89 (17.5) | 84 (17.0) | 107 (16.5) | 79 (16.0) | |
| Current | 672 (28.6) | 143 (21.3) | 169 (29.7) | 186 (32.4) | 174 (31.2) | |
| Drinking status, n (%) | 1666 (76.4) | 353 (64.6) | 390 (72.9) | 479 (82.7) | 444 (86.2) | <0.001 |
| BMI (kg/m2), Mean (SD) | 27.3 (6.0) | 30.5 (7.3) | 27.0 (5.7) | 26.6 (4.9) | 25.0 (4.2) | <0.001 |
| Weight (kg), Mean (SD) | 78.2 (18.5) | 81.5 (20.7) | 75.4 (19.5) | 79.7 (17.8) | 76.2 (14.7) | <0.001 |
| Height (cm), Mean (SD) | 169.2 (10.0) | 163.3 (7.3) | 166.6 (10.0) | 172.7 (9.6) | 174.3 (8.4) | <0.001 |
| Waist circumference (cm), Mean (SD) | 91.9 (14.8) | 97.2 (16.4) | 91.3 (15.0) | 92.5 (13.7) | 86.6 (11.6) | <0.001 |
| TBW (L), Mean (SD) | 39.7 (9.9) | 33.4 (7.4) | 36.0 (8.9) | 42.9 (9.1) | 46.5 (8.3) | <0.001 |
| ECW (L), Mean (SD) | 16.7 (3.6) | 15.1 (3.1) | 15.6 (3.6) | 17.8 (3.6) | 18.2 (3.0) | <0.001 |
| ICF (L), Mean (SD) | 23.0 (6.6) | 18.3 (4.4) | 20.4 (5.4) | 25.1 (5.7) | 28.3 (5.6) | <0.001 |
| ICW (L), Mean (SD) | 23.0 (6.6) | 18.3 (4.4) | 20.4 (5.4) | 25.1 (5.7) | 28.3 (5.6) | <0.001 |
| BCM (kg), Mean (SD) | 32.9 (9.4) | 26.2 (6.4) | 29.1 (7.7) | 35.8 (8.1) | 40.4 (8.0) | <0.001 |
| PFAT (%), Mean (SD) | 31.2 (11.1) | 44.8 (4.2) | 35.8 (3.0) | 27.3 (3.3) | 16.9 (5.1) | <0.001 |
| Energy intakes (kcal/d), Mean (SD) | 2419 (1103) | 1990 (856) | 2228 (1007) | 2646 (1105) | 2812 (1216) | <0.001 |
| Carbohydrate intakes (g/d), Mean (SD) | 304 (146) | 254 (116) | 282 (135) | 327 (147) | 352 (160) | <0.001 |
| Protein intakes (g/d), Mean (SD) | 87.4 (46.7) | 71.6 (36.0) | 78.5 (39.9) | 97.0 (46.5) | 103 (54.8) | <0.001 |
| Total fat intakes (g/d), Mean (SD) | 88.5 (49.3) | 75.1 (41.0) | 83.0 (46.7) | 95.8 (49.1) | 100 (55.2) | <0.001 |
| Hypertension, n (%) | 261 (10.6) | 85 (12.9) | 77 (13.2) | 70 (9.7) | 29 (6.2) | <0.001 |
| Coronary heart disease, n (%) | 8 (0.5) | 2 (0.2) | 4 (1.3) | 2 (0.3) | 0 (0.0) | 0.32 |
| Stroke, n (%) | 10 (0.5) | 4 (1.2) | 4 (0.7) | 2 (0.2) | 0 (0.0) | 0.239 |
| Kidney disease, n (%) | 33 (1.1) | 14 (2.1) | 9 (1.1) | 7 (0.5) | 3 (0.8) | 0.095 |
| Dyslipidemia, n (%) | 1781 (63.3) | 495 (64.1) | 477 (64.8) | 459 (66.4) | 350 (56.6) | <0.001 |
| FPG (mmol/L), Mean (SD) | 5.3 (1.5) | 5.4 (1.5) | 5.3 (1.5) | 5.4 (1.7) | 5.3 (1.2) | 0.207 |
| HbA1c (%), Mean (SD) | 5.3 (0.7) | 5.3 (0.8) | 5.3 (0.8) | 5.3 (0.7) | 5.2 (0.5) | 0.012 |
| FSI (pmol/L), Median (IQR) | 55.4 (37.6, 85.2) |
64.6 (44.6, 103) |
58.3 (38.9, 86.3) |
54.5 (37.2, 83.6) |
46.3 (32.5, 69.5) |
<0.001 |
| HOMA-IR, Median (IQR) | 2.1 (1.4, 3.4) | 2.5 (1.6, 4.1) | 2.2 (1.4, 3.5) | 2.1 (1.4, 3.4) | 1.8 (1.2, 2.8) | <0.001 |
| TG (mmol/L), Median (IQR) | 1.0 (0.7, 1.6) | 1.0 (0.7, 1.5) | 1.0 (0.7, 1.6) | 1.0 (0.7, 1.6) | 1.0 (0.7, 1.5) | 0.062 |
| TC (mmol/L), Mean (SD) | 4.9 (1.0) | 4.9 (1.0) | 4.9 (1.0) | 5.0 (1.1) | 4.7 (1.1) | <0.001 |
| HDL-c (mmol/L), Mean (SD) | 1.3 (0.4) | 1.4 (0.4) | 1.3 (0.4) | 1.3 (0.4) | 1.3 (0.3) | <0.001 |
| LDL-c (mmol/L), Mean (SD) | 2.9 (0.9) | 2.9 (0.8) | 2.9 (0.9) | 3.0 (0.9) | 2.9 (1.0) | 0.005 |
| ALT (U/L), Median (IQR) | 21.0 (16.0, 30.0) |
18.0 (14.0, 24.0) |
19.0 (14.0, 28.0) |
24.0 (17.0, 34.0) |
23.0 (18.0, 32.0) |
<0.001 |
| Cr (umol/L), Mean (SD) | 71.5 (24.2) | 58.6 (12.9) | 66.7 (19.7) | 77.7 (34.5) | 83.0 (15.6) | <0.001 |
| UA (umol/L), Mean (SD) | 316 (81.1) | 279 (69.8) | 292 (81.2) | 341 (80.1) | 351 (68.4) | <0.001 |
| IR, n (%) | 1462 (44.8) | 446 (54.2) | 378 (44.9) | 360 (42.8) | 278 (36.4) | <0.001 |
| T2D, n (%) | 115 (3.4) | 42 (5.4) | 34 (3.8) | 28 (2.7) | 11 (1.7) | <0.001 |
ALT, alanine aminotransferase; BCM, body cell mass; Cr, creatinine; ECW, extracellular water; FPG, fasting plasma glucose; FSI, fasting serum insulin; HbA1c, glycated hemoglobin; HDL-c, high-density lipoprotein cholesterol; HOMA-IR, homeostasis model assessment of insulin resistance; ICF, intracellular fluid; IR, insulin resistance; LDL-c, low-density lipoprotein cholesterol; PBCM, percentage body cell mass; PFAT, percent body fat; T2D, type 2 diabetes; TBW, total body water; TC, total cholesterol; TG, triglycerides; UA, uric acid.
Data were presented as unweighted number (weighted percentage) for categorical variables, mean (SD, standard deviation) or median (IQR, interquartile range) for continuous variables. Differences in baseline characteristics were compared using the χ2 test for categorical variables and the t-test or Wilcoxon rank-sum test for continuous variables.
Association of PBCM with T2D risk markers
In fully adjusted multivariate linear regression models, statistically significant inverse associations were observed between PBCM and FPG (β = -0.02, 95% CI: -0.03 – -0.01, p <0.001), HbA1c (β = -0.01, 95% CI: -0.02 – -0.00, p = 0.002), FSI (β = -1.87, 95% CI: -2.62 – -1.13, p <0.001), and HOMA-IR (β = -0.09, 95% CI: -0.13 – -0.06, p <0.001). A statistically significant trend across PBCM quartiles persisted. Compared with the lowest quartile, participants in the second, third, and fourth PBCM quartiles demonstrated progressively lower levels of FPG, HbA1c, FSI, and HOMA-IR (p for trend <0.001, = 0.009, <0.001, and <0.001, respectively) (Supplementary Table 1).
Association between PBCM and IR
As shown in Table 2, an elevated PBCM level was associated with decreased odds of IR across all multivariate logistic regression models (model 1: OR = 0.96, 95% CI: 0.95 – 0.98, p <0.001; model 2: OR = 0.90, 95% CI: 0.87 – 0.93, p <0.001; model 3: OR = 0.91, 95% CI: 0.88 – 0.94, p <0.001). Full adjustment for potential confounders (model 3) revealed that every one-unit increase in PBCM was associated with a 9% decrease in the odds of IR. Furthermore, this trend maintained statistical significance when analyzed as a categorical variable (quartiles): higher PBCM quartiles exhibited progressively decreased odds of IR compared with the lowest quartile (p for trend <0.001). This inverse association was found in both men and women (Supplementary Table 2).
Table 2.
Association between PBCM and IR
| Variable | Total | Event (%) | Model 1 | Model 2 | Model 3 | |||
|---|---|---|---|---|---|---|---|---|
|
| ||||||||
| OR (95%CI) | p value | OR (95%CI) | p value | OR (95%CI) | p value | |||
| PBCM (%) | 2923 | 1462 (44.8) | 0.96 (0.95, 0.98) | <0.001 | 0.90 (0.87, 0.93) | <0.001 | 0.91 (0.88, 0.94) | <0.001 |
| PBCM (Quartiles) | ||||||||
| Q1 (<36) | 731 | 446 (54.2) | 1 (Ref) | 1 (Ref) | 1 (Ref) | |||
| Q2 (36-41) | 730 | 378 (44.9) | 0.69 (0.51, 0.94) | 0.019 | 0.48 (0.34, 0.68) | <0.001 | 0.48 (0.34, 0.69) | <0.001 |
| Q3 (41-48) | 731 | 360 (42.8) | 0.63 (0.48, 0.83) | 0.001 | 0.23 (0.15, 0.34) | <0.001 | 0.25 (0.16, 0.38) | <0.001 |
| Q4 (≥48) | 731 | 278 (36.4) | 0.48 (0.35, 0.67) | <0.001 | 0.14 (0.09, 0.23) | <0.001 | 0.17 (0.11, 0.28) | <0.001 |
| p for trend | <0.001 | <0.001 | <0.001 | |||||
IR, insulin resistance; NHANES, National Health and Nutrition Examination Survey; PBCM, percentage body cell mass.
†Model 1 was unadjusted for covariates.
‡Model 2 was adjusted for age, gender, race, and NHANES cycle.
§Model 3 was additionally adjusted for physical activity, smoking status, drinking status, energy intakes, hypertension, and dyslipidemia.
To investigate potential non-linear association between PBCM and IR, we performed RCS analysis. The statistical analysis revealed a significant overall relationship (p <0.001) and a non-linear trend (p = 0.017), suggesting a non-linear association between PBCM and IR (Figure 3a). In subsequent threshold (two-piecewise) analyses, when PBCM was <41%, the OR for IR was 0.87 (95% CI: 0.82 – 0.92, p <0.001), whereas when PBCM was To investigate potential non-linear association between PBCM and IR, we performed RCS analysis. The statistical analysis revealed a significant overall relationship (p <0.001) and a non-linear trend (p = 0.017), suggesting a non-linear association between PBCM and IR (Figure 3a). In subsequent threshold (two-piecewise) analyses, when PBCM was <41%, the OR for IR was 0.87 (95% CI: 0.82 – 0.92, p <0.001), whereas when PBCM was ≥41%, the OR was 0.95 (95% CI: 0.91 – 0.99, p = 0.010) (Supplementary Table 4). In gender-stratified analyses, a non-linear association between PBCM and IR was observed in men (p for non-linearity = 0.022; p for overall <0.001). When PBCM was <48%, the OR for IR was 0.84 (95% CI: 0.76–0.92, p <0.001), indicating that the odds of IR were 16% lower for every one-unit increase in PBCM. However, no significant association was observed between PBCM and IR in participants with PBCM values ≥48% (Supplementary Figure 1a and Table 5). No evidence of a non-linear association between PBCM and IR was found in women (Supplementary Figure 1b).
Figure 3.
The complex sampling-weighted RCS analysis of PBCM with (A) IR and (B) T2D. The solid curved line represents the estimates for the association of PBCM with IR and T2D, and shading, the 95% CI. IR, insulin resistance; PBCM, percentage body cell mass; T2D, type 2 diabetes
Across multiple subgroups, stratified analyses were performed to assess potential effect modification of the relationship between PBCM and IR. After stratifying by age, gender, physical activity, smoking status, drinking status, hypertension, and dyslipidemia, no significant interactions were observed across all subgroups (all p >0.05), suggesting that the inverse association between PBCM and IR remained consistent across these factors (Figure 4a).
Figure 4.
Association of PBCM with (A) IR and (B) T2D according to basic features. Note: Except for the stratification component itself, each stratification factor was adjusted for all other variables, including age, gender, race, physical activity, smoking status, drinking status, energy intakes, hypertension, dyslipidemia, and NHANES cycle. IR, insulin resistance; PBCM, percentage body cell mass; T2D, type 2 diabetes
Association between PBCM and T2D
As indicated in Table 3, regardless of adjustment for confounding variables, PBCM remained negatively associated with T2D prevalence in all multivariate logistic regression models (model 1: OR = 0.93, 95% CI: 0.91 – 0.96, p <0.001; model 2: OR = 0.88, 95% CI: 0.84 – 0.93, p <0.001; model 3: OR = 0.89, 95% CI: 0.84 – 0.94, p <0.001). Notably, after fully adjusting for potential confounders (model 3), every one-unit increase in PBCM was associated with an 11% decrease in the odds of T2D. Furthermore, when PBCM was treated as a categorical variable, this trend remained statistically significant. In the fully adjusted multivariable logistic regression model, higher quartiles of PBCM were associated with progressively lower odds of T2D compared with the lowest quartile group (p for trend = 0.006). The gender-stratified analyses showed that the association between PBCM and T2D remained statistically significant in men but not in women (Supplementary Table 3). Moreover, RCS analysis was employed to examine potential non-linear association between PBCM and T2D, and no evidence of non-linearity was observed (p for non-linearity = 0.444) (Figure 3b). In the gender-stratified analysis, no evidence of non-linearity was found either in men or women (Supplementary Figure 2).
Table 3.
Association between PBCM and T2D
| Variable | Total | Event (%) | Model 1 | Model 2 | Model 3 | |||
|---|---|---|---|---|---|---|---|---|
|
| ||||||||
| OR (95%CI) | p value | OR (95%CI) | p value | OR (95%CI) | p value | |||
| PBCM (%) | 2923 | 115 (3.4) | 0.93 (0.91, 0.96) | <0.001 | 0.88 (0.84, 0.93) | <0.001 | 0.89 (0.84, 0.94) | <0.001 |
| PBCM (Quartiles) | ||||||||
| Q1 (<36) | 731 | 42 (5.4) | 1 (Ref) | 1 (Ref) | 1 (Ref) | |||
| Q2 (36-41) | 730 | 34 (3.8) | 0.69 (0.36, 1.32) | 0.260 | 0.54 (0.23, 1.26) | 0.150 | 0.53 (0.21, 1.33) | 0.169 |
| Q3 (41-48) | 731 | 28 (2.7) | 0.48 (0.23, 0.98) | 0.045 | 0.28 (0.09, 0.92) | 0.037 | 0.28 (0.08, 0.96) | 0.043 |
| Q4 (≥48) | 731 | 11 (1.7) | 0.29 (0.14, 0.60) | 0.001 | 0.18 (0.06, 0.56) | 0.004 | 0.18 (0.05, 0.58) | 0.006 |
| p for trend | <0.001 | 0.005 | 0.006 | |||||
NHANES, National Health and Nutrition Examination Survey; PBCM, percentage body cell mass; T2D, type 2 diabetes.
†Model 1 was unadjusted for covariates.
‡Model 2 was adjusted for age, gender, race, and NHANES cycle.
§Model 3 was additionally adjusted for physical activity, smoking status, drinking status, energy intakes, hypertension, and dyslipidemia.
In subgroup analyses stratified by age, gender, physical activity, smoking status, drinking status, hypertension, and dyslipidemia, no significant interactions were observed in any subgroups. Although p-values for interaction with physical activity and hypertension were < 0.05, the findings may not have reached clinical significance given the consistency in the direction of association and the potential influence of multiple testing (Figure 4b)
Mediation analysis
Mediation analysis indicated the potential mediating role of BMI in the relationship of PBCM with IR and T2D (Figure 5). PBCM was inversely related to BMI (β = -0.33, 95% CI: -0.38 – -0.27, p <0.001), and BMI was positively associated with odds of IR (OR = 1.24, 95%CI: 1.21–1.28, p <0.001) and T2D (OR = 1.13, 95%CI: 1.08–1.18, p <0.001). Significant mediating effects of BMI in the relationships of PBCM with IR (proportion of mediated: 70.6%, 95%CI: 55.1%–88.9%, p <0.001) and T2D (proportion of mediated: 29.9%, 95%CI: 14.2%–65.9%, p <0.001) were observed.
Figure 5.
Mediation analysis of BMI in the associations of PBCM with IR (A) and T2D (B). ORs in logistic regression are non-collapsible. Therefore, the total effect (c) does not necessarily equal the sum of direct (c’) and mediation effects (a×b) when mediators are added to the model. This is a well-established property of mediation analysis with binary outcomes (VanderWeele, 2010). BMI, body mass index; IR, insulin resistance; PBCM, percentage body cell mass; T2D, type 2 diabetes
Sensitivity and additional analysis
After excluding participants with missing covariate data, 2,172 individuals remained. The associations of PBCM with IR and T2D remained robust. Specifically, PBCM was inversely associated with the odds of IR (fully adjusted OR = 0.91, 95% CI: 0.89 – 0.94, p <0.001), with higher quartiles showing progressively lower odds (p for trend <0.001). Likewise, PBCM was inversely associated with T2D (fully adjusted OR = 0.89, 95% CI: 0.84 – 0.94, p <0.001), and higher quartiles were associated with lower prevalence compared with the lowest quartile (p for trend = 0.021) (Supplementary Table 6).
Discussion
In this study, we found that PBCM was significantly and negatively associated with the prevalence of IR and T2D. Furthermore, mediating effects of BMI in the relationships between PBCM and both IR and T2D was observed. Sensitivity analyses supported the robustness of these associations.
Our results corroborate and extend previous research. A systematic review and meta-analysis by Gupta et al. included 20 observational studies, indicating that increasing FFM may contribute to reduced T2D risk.27 In a 6-year prospective cohort study among Hispanic/Latino adults in the U.S., LeCroy et al. found that attenuating FFM loss during adulthood may lower prediabetes risk primarily through improving insulin resistance, suggesting that maintenance of muscle mass holds significant implications for preventing glucose metabolism disorders.11 Kahn et al. proposed in their review article that obesity increases the risk of IR and T2D, providing a comprehensive mechanistic framework linking obesity to IR and T2D.4 As BCM is the metabolically active compartment of FFM, reflecting the cellular components (living cells) of the body involved in resting metabolism, oxygen consumption, and carbon dioxide production,28,29 we provide a more physiologically precise measure. In this study, PBCM was inversely associated with IR and T2D, a finding consistent with its relevance to metabolic health. Notably, our study further indicated that BMI mediated the associations between PBCM and both IR and T2D.
The observed inverse associations of PBCM with IR and T2D are biologically plausible. Physiological modeling of body composition at the cellular level can be categorized into functionally distinct compartments: energy storage in FM, energy expenditure and metabolism by BCM, exchanges in ECW, and support by mineral mass (Figure 1).28 FFM, approximately 50% composed of which is skeletal muscle (SM), serves as the primary site for insulin-mediated glucose disposal.30,31 SM, as a critical insulin-sensitive tissue, is involved in glucose metabolism;32,33 its mass reduction may directly contribute to the functional impairment of glucose metabolic capacity.34 While excess FM, particularly visceral adiposity, is strongly associated with IR and T2D,8,35 lower FFM has been linked to impaired glucose metabolism and heightened T2D risk.11,36 BCM, a key indicator of cellular mass and metabolic activity, is the major determinant of basal metabolic rate (BMR) and energy expenditure, contains all the metabolically active tissues (living cell protein mass) of the body and ICF (water inside living cells). ICF constitutes a principal component of BCM and is widely used to estimate BCM.37-39 BCM represents a more precise physiological measure than FFM, as it excludes ECW and mineral mass.
While preserving BCM, reduction of FM and ECW may confer favorable effects on glucose metabolism; enhancing the proportion of metabolically active tissue may be beneficial even in the presence of elevated FM. This is particularly relevant to conditions such as sarcopenic obesity, wherein diminished muscle mass (and presumably reduced BCM) coexists with excess adiposity, constituting a high-risk metabolic phenotype.36,40 Higher or preserved BCM, such as through resistance exercise training combined with adequate protein intake, may be associated with greater insulin sensitivity and lower odds of T2D.41-43
Additionally, in the gender-stratified analysis, we observed an inverse association between PBCM and IR in both men and women; a non-linear association was observed in men but not in women. The inverse association between PBCM and T2D was significant in men but not in women; no evidence of non-linearity was found in either gender. These disparities may be attributable to differences in sex hormones, adipose tissue distribution, or skeletal muscle metabolism.
This study has several strengths, including its nationally representative sample, the analysis of associations with multiple T2D risk markers, mediation analysis, and rigorous adjustment for various confounders. Nevertheless, it should be acknowledged that certain limitations exist. First, although we employed multivariable regression and conducted extensive subgroup and sensitivity analyses, residual confounding cannot be entirely ruled out. Second, the cross-sectional design precludes determination of a causal relationship between PBCM and IR or T2D. Our conclusions describe observed associations rather than definitive causal relationships. Mediation analyses using cross-sectional data can yield unstable or misleading results. Any detected mediation should be interpreted as an association consistent with a proposed causal framework, not as proof of causality. The observed patterns may reflect statistical mediation, but only within the constraints of the study design. Prospective longitudinal studies or randomized controlled trials are required to establish causality. Third, PBCM and BMI both incorporate body weight, and that some degree of structural dependency between these variables cannot be completely excluded. Although no significant multicollinearity was detected, the observed associations should be interpreted with caution. Lastly, this study was conducted in a U.S. population. Given established variations in diabetes phenotypes across ethnic groups and regional healthcare disparities, extrapolation to non-U.S. populations requires caution. Validation in other populations is warranted to assess the generalizability of these findings.
Conclusion
This study observed inverse associations of PBCM with IR and T2D among U.S. adults. Furthermore, BMI mediated the associations between PBCM and both IR and T2D. These findings provide novel insights for the prevention and management of insulin resistance and type 2 diabetes. Prospective longitudinal studies are needed to clarify causal relationships and underlying mechanisms.
Ethics Approval and Informed Consent
The survey protocol was approved by National Center for Health Statistics Ethics Review Board of the Centers for Disease Control and Prevention (Protocol #98-12, https://www.cdc.gov/nchs/nhanes/about/erb.html). Written informed consent was obtained from all participants.
Acknowledgements
The author would like to thank the National Center for Health Statistics (NCHS) and the Centers for Disease Control and Prevention (CDC) for making NHANES data publicly available. The authors also acknowledge the efforts of NHANES participants and data collection staff.
Disclosure on The use of AI and AI-Assisted Technologies
The authors declare that no artificial intelligence or AI-assisted technologies were used in this study. All data collection, data analysis, and figure preparation were conducted independently by the authors.
The authors reviewed and edited the manuscript and bear full responsibility for the content of the publication.
Conflict of Interest and Funding Disclosure
The authors declare no conflicts of interest.
No financial support was received from any institution or person for this study.
Supplementary Material
Supplementary data
References
- 1.Sun H, Saeedi P, Karuranga S, Pinkepank M, Ogurtsova K, Duncan BB, et al. IDF Diabetes Atlas: Global, regional and country-level diabetes prevalence estimates for 2021 and projections for 2045. Diabetes Res Clin Pract. 2022;183:109–119. doi: 10.1016/j.diabres.2021.109119. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Marunaka Y. Roles of interstitial fluid pH in diabetes mellitus: Glycolysis and mitochondrial function. World J Diabetes. 2015;6:125–135. doi: 10.4239/wjd.v6.i1.125. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kautzky-Willer A, Harreiter J, Pacini G. Sex and Gender Differences in Risk, Pathophysiology and Complications of Type 2 Diabetes Mellitus. Endocr Rev. 2016;37:278–316. doi: 10.1210/er.2015-1137. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Kahn SE, Hull RL, Utzschneider KM. Mechanisms linking obesity to insulin resistance and type 2 diabetes. Nature. 2006;444:840–846. doi: 10.1038/nature05482. [DOI] [PubMed] [Google Scholar]
- 5.Leung MY, Carlsson NP, Colditz GA, Chang SH. The Burden of Obesity on Diabetes in the United States: Medical Expenditure Panel Survey, 2008 to 2012. Value Health. 2017 Jan;20:77–84. doi: 10.1016/j.jval.2016.08.735. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Xu G, Liu B, Sun Y, Du Y, Snetselaar LG, Hu FB, et al. Prevalence of diagnosed type 1 and type 2 diabetes among US adults in 2016 and 2017: population based study. BMJ. 2018;362:k1497. doi: 10.1136/bmj.k1497. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Zhang TY, Zhang ZM, Wang XN, Kuang HY, Xu Q, Li HX, et al. Relationship between weight-adjusted-waist index and all-cause and cardiovascular mortality in individuals with type 2 diabetes. Diabetes Obes Metab. 2024;26:5621–5629. doi: 10.1111/dom.15929. [DOI] [PubMed] [Google Scholar]
- 8.Liu D, Lei YL, Zhang L, Wang W, Shao C, Zhou Q, et al. Associations of the fat-free mass index and the fat mass index with the risk of developing diabetes and prediabetes in US adults: a nationally representative cross-sectional study. Lipids Health Dis. 2024;23:383. doi: 10.1186/s12944-024-02370-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Lee SY, Ahn S, Kim YJ, Ji MJ, Kim KM, Choi SH, et al. Comparison between Dual-Energy X-ray Absorptiometry and Bioelectrical Impedance Analyses for Accuracy in Measuring Whole Body Muscle Mass and Appendicular Skeletal Muscle Mass. Nutrients. 2018;10:738. doi: 10.3390/nu10060738. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Kyle UG, Bosaeus I, De Lorenzo AD, Deurenberg P, Elia M, Manuel Gómez J, et al. Bioelectrical impedance analysis-part II: utilization in clinical practice. Clin Nutr. 2004;23:1430–1453. doi: 10.1016/j.clnu.2004.09.012. [DOI] [PubMed] [Google Scholar]
- 11.LeCroy MN, Hua S, Kaplan RC, Sotres-Alvarez D, Qi Q, Thyagarajan B, et al. Associations of changes in fat free mass with risk for type 2 diabetes: Hispanic Community Health Study/Study of Latinos. Diabetes Res Clin Pract. 2021;171:108557. doi: 10.1016/j.diabres.2020.108557. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Malone JI, Hansen BC. Does obesity cause type 2 diabetes mellitus (T2DM)? Or is it the opposite? Pediatr Diabetes. 2019;20:5–9. doi: 10.1111/pedi.12787. [DOI] [PubMed] [Google Scholar]
- 13.Volpato S, Romagnoni F, Soattin L, Blè A, Leoci V, Bollini C, et al. Body mass index, body cell mass, and 4-year all-cause mortality risk in older nursing home residents. J Am Geriatr Soc. 2004;52:886–891. doi: 10.1111/j.1532-5415.2004.52254.x. [DOI] [PubMed] [Google Scholar]
- 14.Meireles MS, Wazlawik E, Bastos JL, Garcia MF. Comparison between nutritional risk tools and parameters derived from bioelectrical impedance analysis with subjective global assessment. J Acad Nutr Diet. 2012;112:1543–1549. doi: 10.1016/j.jand.2012.07.005. [DOI] [PubMed] [Google Scholar]
- 15.Wang Z, St-Onge MP, Lecumberri B, Pi-Sunyer FX, Heshka S, Wang J, et al. Body cell mass: model development and validation at the cellular level of body composition. Am J Physiol Endocrinol Metab. 2004;286:E123–E128. doi: 10.1152/ajpendo.00227.2003. [DOI] [PubMed] [Google Scholar]
- 16.Wallace TM, Levy JC, Matthews DR. Use and abuse of HOMA modeling. Diabetes Care. 2004;27:1487–1495. doi: 10.2337/diacare.27.6.1487. [DOI] [PubMed] [Google Scholar]
- 17.Yin B, Wu Z, Xia Y, Xiao S, Chen L, Li Y. Non-linear association of atherogenic index of plasma with insulin resistance and type 2 diabetes: a cross-sectional study. Cardiovasc Diabetol. 2023;22:157. doi: 10.1186/s12933-023-01886-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Liu Y, Gong R, Luo G, Li J, Li Q, Yang L, et al. Associations of Triglycerides/High-Density Lipoprotein Cholesterol Ratio With Insulin Resistance, Impaired Glucose Tolerance, and Diabetes in American Adults at Different Vitamin D3 Levels. Front Endocrinol (Lausanne) 2021;12:735736. doi: 10.3389/fendo.2021.735736. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Wan Z, Guo J, Pan A, Chen C, Liu L, Liu G. Association of Serum 25-Hydroxyvitamin D Concentrations With All-Cause and Cause-Specific Mortality Among Individuals With Diabetes. Diabetes Care. 2021;44:350–357. doi: 10.2337/dc20-1485. [DOI] [PubMed] [Google Scholar]
- 20.Dong G, Gan M, Xu S, Xie Y, Zhou M, Wu L. The neutrophil-lymphocyte ratio as a risk factor for all-cause and cardiovascular mortality among individuals with diabetes: evidence from the NHANES 2003-2016. Cardiovasc Diabetol. 2023;22:267. doi: 10.1186/s12933-023-01998-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Huang W, Ma X, Liang H, Li H, Chen J, Fang L, et al. Dietary Magnesium Intake Affects the Association Between Serum Vitamin D and Type 2 Diabetes: A Cross-Sectional Study. Front Nutr. 2021;8:763076. doi: 10.3389/fnut.2021.763076. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Nie Y, Zhou H, Wang J, Kan H. Association between systemic immune-inflammation index and diabetes: a population-based study from the NHANES. Front Endocrinol (Lausanne) 2023;14:1245199. doi: 10.3389/fendo.2023.1245199. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Wang Y, Hao H, Lu Q, Zhang L. Serum albumin and hypertension: The mediating roles of BMI, C-reactive protein and extracellular fluid. Asia Pac J Clin Nutr. 2026;35(1):60–70. doi: 10.6133/apjcn.202602_35(1).0006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Wang Y, Lu Q, Zhang L, Wei G, Shen X, Peng Y. Association of serum albumin and obesity: Mediation effect of extracellular fluid, percent body fat and C-reactive protein. Medicine (Baltimore) 2026;105:e47802. doi: 10.1097/MD.0000000000047802. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Kammerlander AA, Mayrhofer T, Ferencik M, Pagidipati NJ, Karady J, Ginsburg GS, et al. Association of Metabolic Phenotypes With Coronary Artery Disease and Cardiovascular Events in Patients With Stable Chest Pain. Diabetes Care. 2021;44:1038–1045. doi: 10.2337/dc20-1760. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Vandenbroucke JP, von Elm E, Altman DG, Gøtzsche PC, Mulrow CD, Pocock SJ, et al. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE): explanation and elaboration. PLoS Med. 2007;4:e297. doi: 10.1371/journal.pmed.0040297. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Gupta P, Lanca C, Gan ATL, Soh P, Thakur S, Tao Y, et al. The Association between Body Composition using Dual energy X-ray Absorptiometry and Type-2 Diabetes: A Systematic Review and Meta-Analysis of Observational studies. Sci Rep. 2019;9:12634. doi: 10.1038/s41598-019-49162-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Savalle M, Gillaizeau F, Maruani G, Puymirat E, Bellenfant F, Houillier P, et al. Assessment of body cell mass at bedside in critically ill patients. Am J Physiol Endocrinol Metab. 2012;303:E389–E396. doi: 10.1152/ajpendo.00502.2011. [DOI] [PubMed] [Google Scholar]
- 29.Fiaccadori E, Morabito S, Cabassi A, Regolisti G. Body cell mass evaluation in critically ill patients: killing two birds with one stone. Crit Care. 2014;18:139. doi: 10.1186/cc13852. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Oliver CJ, Climstein M, Rosic N, Bosy-Westphal A, Tinsley G, Myers S. Fat-Free Mass: Friend or Foe to Metabolic Health? J Cachexia Sarcopenia Muscle. 2025;16:e13714. doi: 10.1002/jcsm.13714. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.DeFronzo RA, Tripathy D. Skeletal muscle insulin resistance is the primary defect in type 2 diabetes. Diabetes Care. 2009;32(Suppl 2):S157–S163. doi: 10.2337/dc09-S302. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Sukackiene D, Laucyte-Cibulskiene A, Vickiene A, Rimsevicius L, Miglinas M. Risk stratification for patients awaiting kidney transplantation: Role of bioimpedance derived edema index and nutrition status. Clin Nutr. 2020;39(9):2759–2763. doi: 10.1016/j.clnu.2019.12.001. [DOI] [PubMed] [Google Scholar]
- 33.Hioka A, Akazawa N, Okawa N, Nagahiro S. Increased total body extracellular-to-intracellular water ratio in community-dwelling elderly women is associated with decreased handgrip strength and gait speed. Nutrition. 2021;86:111175. doi: 10.1016/j.nut.2021.111175. [DOI] [PubMed] [Google Scholar]
- 34.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;178:523–531. doi: 10.1530/EJE-17-0882. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Adeva-Andany MM, Dominguez-Montero A, Adeva-Contreras L, Fernandez-Fernandez C, Carneiro-Freire N, Gonzalez-Lucan M. Body Fat Distribution Contributes to Defining the Relationship between Insulin Resistance and Obesity in Human Diseases. Curr Diabetes Rev. 2024;20:e160823219824. doi: 10.2174/1573399820666230816111624. [DOI] [PubMed] [Google Scholar]
- 36.Armandi A, Rosso C, Caviglia GP, Ribaldone DG, Bugianesi E. The Impact of Dysmetabolic Sarcopenia Among Insulin Sensitive Tissues: A Narrative Review. Front Endocrinol (Lausanne) 2021;12:716533. doi: 10.3389/fendo.2021.716533. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Yajima T, Yajima K. Ratio of extracellular water to intracellular water and simplified creatinine index as predictors of all-cause mortality for patients receiving hemodialysis. PLoS One. 2023;18:e0282864. doi: 10.1371/journal.pone.0282864. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Gracia-Iguacel C, Gonzalez-Parra E, Mahillo I, Ortiz A. Low Intracellular Water, Overhydration, and Mortality in Hemodialysis Patients. J Clin Med. 2020;9:3616. doi: 10.3390/jcm9113616. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Szeluga DJ, Stuart RK, Utermohlen V, Santos GW. Nutritional assessment by isotope dilution analysis of body composition. Am J Clin Nutr. 1984;40:847–854. doi: 10.1093/ajcn/40.4.847. [DOI] [PubMed] [Google Scholar]
- 40.Kim TN, Park MS, Lim KI, Choi HY, Yang SJ, Yoo HJ, et al. Relationships between sarcopenic obesity and insulin resistance, inflammation, and vitamin D status: the Korean Sarcopenic Obesity Study. Clin Endocrinol (Oxf) 2013;78:525–532. doi: 10.1111/j.1365-2265.2012.04433.x. [DOI] [PubMed] [Google Scholar]
- 41.Gillen JB, Estafanos S, Govette A. Exercise-nutrient interactions for improved postprandial glycemic control and insulin sensitivity. Appl Physiol Nutr Metab. 2021;46:856–865. doi: 10.1139/apnm-2021-0168. [DOI] [PubMed] [Google Scholar]
- 42.Strasser B, Siebert U, Schobersberger W. Resistance training in the treatment of the metabolic syndrome: a systematic review and meta-analysis of the effect of resistance training on metabolic clustering in patients with abnormal glucose metabolism. Sports Med. 2010;40:397–415. doi: 10.2165/11531380-000000000-00000. [DOI] [PubMed] [Google Scholar]
- 43.Paddon-Jones D, Rasmussen BB. Dietary protein recommendations and the prevention of sarcopenia. Curr Opin Clin Nutr Metab Care. 2009;12:86–90. doi: 10.1097/MCO.0b013e32831cef8b. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary data