Abstract
Background
The aim of this study was to address the intramuscular adipose tissue (IMAT) accumulation in the lower extremities and further detect the relationship between adipose tissue (AT) distribution in the muscle and glucose metabolism in subjects with obesity.
Methods
We conducted a cross-sectional study in 120 Chinese obese adults (80 male and 40 female) with BMI ≥ 28 kg/m2. MRI was applied to access the IMAT content in lower extremities. The oral glucose tolerance test was used to evaluate the glucose metabolism and insulin secretion in all individuals. The correlations between glucose metabolism and the fat content of the lower extremities were further assessed.
Results
Among 120 included subjects, 54 were classified as subjects with normal glucose tolerance (NGT) and 66 with impaired glucose regulation (IGR). We presented that those with IGR had higher fat accumulation in semitendinosus, adductor magnus, gracilis and sartorius than those with NGT (all P < 0.05). In sex-specific analyses, females have higher IMAT in adductor magnus than males (P < 0.001). Males with IGR had higher fat fraction of semitendinosus and sartorius than those with NGT (P = 0.020, P = 0.014, respectively). Logistic regression analyses revealed that IMAT content in semitendinosus was the independent factor of IGR in individuals with obesity after adjustment for age, gender, triglycerides, creatinine and albumin (odds ratio: 1.13, 95% CI: 1.02–1.26, P = 0.024).
Conclusions
Increased adipose tissue accumulation in thigh muscles was associated with glucose dysregulation in patients with obesity. IMAT content in semitendinosus may serve as a possible risk factor for impaired glucose metabolism.
Introduction
Obesity has become one of the greatest public health challenges nowadays. It is estimated that more than 50% of adults worldwide will be overweight or obese by 2030 (1). In China, the prevalence of overweight/obesity in adults is 34.3% and 16.4% respectively according to the China Chronic Disease and Nutrition Surveillance 2015–2019 survey (2).
Although BMI has been adopted as a convenient tool to assess overall adiposity in epidemiological studies, it fails to evaluate fat distribution or fat content within the body. Ectopic fat accumulation is regarded as a risk factor for metabolic diseases (3, 4). Our team recently demonstrated that visceral adipose tissue (VAT) and fat content of the liver in overweight/obese subjects is associated with glucose regulation (5). Furthermore, we found that the presence of fat within the liver was an independent risk factor for impaired glucose regulation (IGR) (5).
Other than fat, skeletal muscle also plays a key role in glucose clearance. It is responsible for the majority of glucose uptake in the postprandial state and is also considered as the primary driver of systemic insulin resistance (6, 7). Due to its crucial role in systemic glycometabolism, skeletal muscle has received increasing attention in recent years (8, 9). With the development of obesity, increased lipid levels and ectopic lipid deposition could interfere with insulin signaling in skeletal muscle (10, 11), contributing to insulin resistance and glucose dysregulation (12). What’s more, several adipokines secreted by adipose tissue link to obesity, insulin resistance and systemic glucose regulation (13, 14). Thus, slight changes in lipid accumulation and distribution in skeletal muscle may correlate strongly with impaired glucose homeostasis. As the thigh muscle is one of the largest muscle groups in the human body, it is important to establish the relationship with metabolic diseases by assessing both local and systemic muscle composition.
Although intramuscular adipose tissue (IMAT) represents a small proportion of body fat, it is closely related to muscle insulin sensitivity (15, 16). An earlier study has revealed that excess amount of muscle fat is a crucial risk factor for type 2 diabetes mellitus (T2DM) and adipose tissue (AT) in thigh muscle was positively associated with fasting insulin (FINS) (17). There is also evidence that IMAT in the thigh is inversely related to insulin sensitivity (18). Most previous studies have estimated fat content and distribution by dual-energy X-ray absorptiometry (DXA) and CT (19, 20, 21), which are widely used in clinical practice. However, DXA is unable to differentiate between fat and lean tissue compartments. DXA accuracy can be affected by tissue hydration status (16, 22, 23). Although CT scans can quantitatively measure fat content at the tissue-organ level with high image resolution, its high radiation exposure cannot be ignored (22, 23). In recent decades, magnetic resonance imaging (MRI) has gained increasing attention in the measurement of adipose depots. With excellent image resolution, MRI is considered to be the most accurate method for determining body composition at the tissue-organ level, specifically whole-body and regional AT (22), and for quantifying fat content in different organs (16, 24).
Given the growing prevalence of obesity, the assessment of related body composition, especially the quantitative estimation of ectopic adipose tissue in skeletal muscle, is increasingly warranted. Yet there are few studies focusing on thigh muscle fat among treatment-naïve subjects with obesity through precise quantification of MRI multipoint Dixon technique. The aim of this study was to investigate the accumulation and distribution of IMAT in the lower extremities by MRI and to further evaluate its relationship with glucose metabolism in patients with obesity.
Materials and methods
Subjects
This cross-sectional study recruited 120 consecutive treatment-naïve Chinese outpatients with obesity, including 80 males and 40 females, from Renji Hospital Endocrine Department (Shanghai, China) from September 2019 to December 2020 (ChiCTR2200066123).
All the patients underwent clinical evaluation, physical examination and biochemical tests as well as an MRI scan of the lower extremities. The inclusion criteria were as follows: (i) Aged 18–60 years and (ii) met the criterion for obesity (BMI ≥ 28 kg/m2) (25). The exclusion criteria were (i) secondary obesity, including Cushing’s syndrome, polycystic ovary syndrome (PCOS) and uncontrolled hypothyroidism; (ii) use of anti-diabetic, slimming drugs, glucocorticoids, antidepressant or antianxiety medications; (iii) pregnancy or lactation; (iv) contraindications of MRI; and (v) severe systemic diseases, such as heart failure, hepatic failure, renal insufficiency and cancer. The study was conducted in accordance with the Declaration of Helsinki and approved by the Clinical Research Ethics Committee of Renji Hospital Affiliated with Medical College of Shanghai Jiao Tong University (LY2022-058-B). Informed consent has been obtained from each patient after a full explanation of the purpose and nature of all procedures used. All the individuals with obesity underwent an oral glucose tolerance test (OGTT). Fasting blood samples were taken to measure liver function, kidney function and blood lipids. MRI was performed to assess AT in the lower extremities.
Anthropometric measurements
The body weight and height of each subject were measured to the nearest 0.1 kg and 0.1 cm, respectively. BMI was calculated as weight divided by height squared (kg/m2). Abdominal waist circumference (WC) was measured to the nearest 0.1 cm midway between the lower rib and the iliac crest at the end of expiration. Hip circumference (HC) was measured in the standing position at the widest point over the greater trochanters. Anthropometric data were collected by a trained clinician.
Laboratory assays
OGTT was performed after 12 h of fasting in all subjects. For glucose and insulin assessment, blood samples were taken at 0, 30, 60, 120 and 180 min after glucose administration. Insulin and C-peptide levels were measured using chemiluminescence assays (Roche Diagnostics). Plasma glucose levels, liver enzymes, total cholesterol (TC), triglycerides (TG), HDL-c and LDL-c levels were determined using Roche reagents (D 2400 and E 170 Modular Analytics modules with Roche/Hitachi analyzers; Roche Diagnostics). According to the World Health Organization (WHO) OGTT definition (26), subjects were classified as normal (fasting plasma glucose (FPG) <5.6 mmol/L) and impaired glucose metabolism either with established T2DM (FPG ≥7.0 and/or 2-h postprandial plasma glucose (2h-PBG) ≥11.1 mmol/L) or prediabetes (impaired fasting glucose (IFG): 5.6–6.9 mmol/L and/or impaired glucose tolerance (IGT): 2h-PBG 7.8–11.0 mmol/L). Insulin resistance was estimated by the use of the homeostasis model assessment of insulin resistance (HOMA-IR). HOMA-IR = (FPG (mmol/L) × FINS (μU/mL))/22.5 (27).
MRI image analysis
All IMAT measurements by MRI were performed using a 3.0 T wide bore scanner (Ingenia; Philips Healthcare, Best, The Netherlands) as described previously (5). The proton density fat fraction (PDFF) pulse sequences are multipoint Dixon techniques that use a low flip angle of 3°–5° to limit T1 bias, acquire six echoes to correct for T2* effects and use a multipeak fat model. 3D axial images were acquired from the diaphragmatic dome to the sole of the foot. Water, fat, fat fraction (FF%), R2* and T2* maps were automatically generated after MRI acquisition. Measurements for each participant were obtained by two independent readers (SYH and WZ). Through a vendor post-processing platform (Philips IntelliSpace, Philips Healthcare), the regions of interest (ROIs) on FF% and fat map datasets were drawn semi-automatically to calculate skeletal muscle PDFF. Different combinations of images in the multichannel data could be applied to identify specific ROIs.
The reviewers were blinded to the results of the other reviewers. The average of two measurements from the two reviewers was defined as the mean value and was used for further analyses.
Statistical analysis
Statistical analysis was performed with IBM SPSS software (version 26.0). G*Power software (version 3.1.9.7; Heinrich-Heine-Universität Düsseldorf, Düsseldorf, Germany) was used to calculate the power (1 − β) of the study. The type of power analysis was post hoc with an alpha level of 0.05 and the statistical tests were two tailed. The effect size was set at |ρ| = 0.3 and the total sample size was 120. The output power (1 − β) was 0.927. Data were presented as the mean ± s.d. for continuous variables. Differences between the subjects in NGT and IGR groups were assessed by either an independent sample t-test (numerical variables of normal distribution) or Mann–Whitney U test (numerical variables of skewed distribution). The chi-square test was used for comparing categorical variables. Associations of the glycemic status and thigh muscle FF% were determined by binary logistic regression adjusted for demographics and clinical laboratory factors in Table 1.
Demographics, clinical laboratory characteristics of all the patients with obesity based on glucose metabolism.
| Characteristics | NGT (n = 54) | IGR (n = 66) | P |
|---|---|---|---|
| Age (years) | 35.07 ± 10.34 | 37.6 1 ± 9.27 | 0.071 |
| Sex (male/female, n) | 41/13 | 39/27 | 0.052 |
| BMI (kg/m2) | 34.73 ± 5.45 | 35.06 ± 5.91 | 0.895 |
| WC (cm) | 111.56 ± 12.91 | 117.78 ± 18.38 | 0.232 |
| HC (cm) | 115.10 ± 10.90 | 121.31 ± 16.14 | 0.032 |
| WHR | 0.97 ± 0.07 | 0.97 ± 0.08 | 0.916 |
| ALT (IU/L) | 53.31 ± 58.87 | 56.76 ± 38.04 | 0.077 |
| AST (IU/L) | 29.42 ± 26.06 | 35.76 ± 20.16 | 0.004 |
| Alb (g/L) | 47.53 ± 2.64 | 46.21 ± 3.16 | 0.019 |
| Cre (μmol/L) | 75.40 ± 14.65 | 68.02 ± 14.73 | 0.009 |
| UA (mmol/L) | 464.02 ± 123.89 | 440.78 ± 107.38 | 0.288 |
| TG (mmol/L) | 1.92 ± 1.07 | 2.71 ± 1.99 | 0.002 |
| TC (mmol/L) | 4.97 ± 1.02 | 5.34 ± 0.97 | 0.054 |
| HDL (mmol/L) | 1.11 ± 0.24 | 1.12 ± 0.26 | 0.909 |
| LDL (mmol/L) | 3.01 ± 0.82 | 3.27 ± 0.82 | 0.100 |
| NEFA (mmol/L) | 0.55 ± 0.20 | 0.65 ± 0.20 | 0.015 |
| FBG (mmol/L) | 5.00 ± 0.44 | 6.58 ± 2.23 | <0.001 |
| 2h-PBG (mmol/L) | 6.44 ± 0.88 | 10.96 ± 3.77 | <0.001 |
| FINS (μU/mL) | 16.75 ± 14.20 | 19.67 ± 10.39 | 0.028 |
| 2h-INS (μU/mL) | 79.66 ± 47.02 | 130.18 ± 87.14 | 0.002 |
| HbA1c (%) | 5.36 ± 0.37 | 6.71 ± 1.96 | <0.001 |
| HOMA-IR | 3.80 ± 3.46 | 5.78 ± 3.65 | <0.001 |
| IL-1β (pg/mL) | 6.10 ± 2.51 | 6.31 ± 3.61 | 0.959 |
| IL-2R (U/mL) | 351.11 ± 116.47 | 347.60 ± 111.80 | 0.878 |
| IL-6 (pg/mL) | 3.94 ± 2.23 | 3.53 ± 1.42 | 0.647 |
| IL-8 (pg/mL) | 117.34 ± 115.86 | 92.73 ± 91.27 | 0.420 |
| IL-10 (pg/mL) | 5.07 ± 0.33 | 5.13 ± 0.60 | 0.796 |
| TNF-α (pg/mL) | 7.25 ± 1.85 | 7.34 ± 1.77 | 0.832 |
Data are presented as mean ± s.d.
ALT, alanine aminotransferase; Alb, albumin; AST, aspartate aminotransferase; BMI, body mass index; Cre: creatinine; FBG, fasting blood glucose; FINS, fasting insulin; 2h-INS, 2-h insulin; 2h-PBG, 2-h postprandial blood glucose; HbA1c, glycosylated hemoglobin; HC, hip circumference; HDL, high-density lipoprotein; HOMA-IR, homeostasis model assessment of insulin resistance; IL-1β, interleukin 1β; IL-2R, interleukin-2 receptor; IL-6, interleukin 6; IL-8, interleukin 8; IL-10, interleukin 10; LDL, low-density lipoprotein; NEFA, non-esterified fatty acid; NGT, normal glucose tolerance; TC, total cholesterol; TG, triglycerides; TNF-α, tumor necrosis factor alpha; UA, uric acid; WC, waist circumference.
In addition, areas under the receiver operating characteristic (ROC) curve (AUC) analyses were used to calculate the optimal cutoff value for the thigh AT content or combination of biochemical factors in predicting impaired glucose metabolism. All hypotheses were two-tailed. P < 0.05 was considered statistically significant.
Results
Demographic and clinical characteristics of the patients with obesity
One hundred and twenty individuals with obesity were recruited for this study. The characteristics of all the subjects were as follows: age (years) 36.47 ± 9.81; BMI (kg/m2) 34.89 ± 5.70. The clinical characteristics of this study population are shown in Table 1. The study population was divided into two groups according to OGTT: normal glucose tolerance (NGT) group and IGR group, including prediabetes and T2DM. Fasting blood glucose (FBG), 2-h postprandial blood glucose (2h-PBG), FINS, 2-h insulin (2h-INS), glycosylated hemoglobin (HbA1c) and HOMA-IR levels were significantly higher in IGR group than in the NGT group (all P < 0.05, Table 1). Patients with obesity in the IGR group had a more pronounced metabolic risk profile than those in NGT group, for example, higher HC and higher levels of aspartate aminotransferase (AST), TG and non-esterified fatty acid (NEFA) (all P < 0.05, Table 1).
No significant differences were observed between the NGT group and IGR group in terms of age, gender, BMI, waist circumference (WC), renal function, other fasting lipid parameters and serum inflammatory factors (all P > 0.05, Table 1). However, there was a trend towards sex preponderance between the two groups (P = 0.052).
Fat accumulation in the muscles of lower extremities of all the patients with obesity based on glucose metabolism
The results of the FF% in the lower extremities measured by MRI were provided in Table 2, Fig. 1 and Supplementary Table 1 (see section on supplementary materials given at the end of this article). In general, subjects in the NGT group had lower thigh fat content than those in the IGR group. The mean thigh muscle FF% was lower in the NGT group than IGR group (8.62 ± 2.74 and 9.62 ± 2.94, P = 0.037, Table 2). There was a positive correlation between thigh muscle FF% and HbA1c (Supplementary Table 2). No correlation was found between thigh muscle FF% and FBG or between thigh muscle FF% and HOMA-IR (Supplementary Table 2).
Color-coded fat fractional mapping calculated from mDIXON MRI. Representative thigh MRI images of IMAT between NGT group and IGR group. Representative thigh MRI images in a subject with obesity with NGT (A) and an subject with obesity with IGR (B). Two patients’ information is available in Supplementary Table 1. The dark blue color corresponds to the lowest fat content and the red to the highest fat content. IMAT, intramuscular adipose tissue; FF%, fat fraction.
Citation: Endocrine Connections 12, 11; 10.1530/EC-23-0248
Fat accumulation in muscles of lower extremity of all the patients with obesity based on glycemic status.
| Muscle FF% | NGT (n = 54) | IGR (n = 66) | P |
|---|---|---|---|
| Anterior tibialis | 4.54 ± 2.30 | 4.80 ± 3.23 | 0.904 |
| Peroneus longus | 9.25 ± 4.90 | 9.58 ± 5.32 | 0.865 |
| Soleus | 6.56 ± 2.12 | 6.89 ± 2.24 | 0.497 |
| Gastrocnemius | 8.46 ± 4.16 | 8.51 ± 3.75 | 0.823 |
| Mean calf muscle | 7.20 ± 2.90 | 7.45 ± 3.30 | 0.812 |
| Vastus lateralis | 6.72 ± 2.40 | 7.43 ± 2.49 | 0.059 |
| Rectus femoris | 6.15 ± 2.07 | 6.35 ± 2.17 | 0.810 |
| Vastus medialis | 5.82 ± 3.40 | 5.81 ± 2.07 | 0.423 |
| Vastus internus | 5.14 ± 1.58 | 5.69 ± 2.05 | 0.109 |
| Biceps femoris | 10.30 ± 3.76 | 11.55 ± 4.24 | 0.102 |
| Semitendinosus | 10.25 ± 4.50 | 12.42 ± 5.17 | 0.012 |
| Semimembranosus | 11.32 ± 4.62 | 11.60 ± 4.09 | 0.376 |
| Adductor magnus | 5.40 ± 2.02 | 6.32 ± 3.13 | 0.031 |
| Gracilis | 12.12 ± 5.19 | 13.89 ± 5.41 | 0.031 |
| Sartorius | 12.94 ± 5.15 | 15.17 ± 4.85 | 0.006 |
| Mean thigh muscle | 8.62 ± 2.74 | 9.62 ± 2.94 | 0.037 |
Data are presented as mean ± s.d.
FF%, fat fraction; IGR, impaired glucose regulation (including prediabetes and T2DM); NGT, normal glucose tolerance.
Notably, patients with obesity and impaired glucose metabolism had higher FF% values in semitendinosus (12.42 ± 5.17 and 10.25 ± 4.50, P = 0.012, Table 2), adductor magnus (6.32 ± 3.13 and 5.40 ± 2.02, P = 0.031, Table 2), gracilis (13.89 ± 5.41 and 12.12 ± 5.19, P = 0.031, Table 2) and sartorius (15.17 ± 4.85 and 12.94 ± 5.15, P = 0.006, Table 2) than those with normal glucose tolerance (Table 2). On the other hand, there were no significant differences in terms of fat content in calf muscles between the two groups (P > 0.05, Table 2).
Fat accumulation in the muscles of lower extremities of male and female individuals with obesity based on glucose metabolism
There was a trend toward a difference in gender distribution between NGT and IGR group (P = 0.052, Table 1). Female subjects have higher FF% values in adductor magnus than that in male subjects (7.14 ± 3.78 and 5.29 ± 1.70, P < 0.001, Supplementary Table 5). In male individuals, FF% of semitendinosus and gracilis were higher in IGR group than in NGT group (P = 0.020, P = 0.014, respectively, Supplementary Table 6). There was a positive correlation between HbA1c and FF% of sartorius in males (P < 0.01, Supplementary Table 7). In female individuals, no significant difference was observed between NGT and IGR groups in terms of IMAT in lower extremities (all P > 0.05, Supplementary Table 8). There was no correlation between thigh muscle FF% and glycemic status indicated by HbA1c, FBG and HOMA-IR in female individuals with obesity (all P > 0.05, Supplementary Table 9).
Univariate and multivariate analyses for potential factors associated with impaired glucose metabolism in all the patients with obesity
Potential explanatory variables associated with IGR are listed in Table 3. In the univariate analysis, TG and NEFA were significantly and positively correlated with abnormal glycometabolism (odds ratio: 1.57, 95% CI:1.08–2.28, P = 0.018; odds ratio: 16.05, 95% CI: 1.54–167.46, P = 0.020, Table 3). Negative correlations of biochemical indicators with impaired glucose metabolism were observed in the patients with obesity (odds ratio: 0.86, 95% CI: 0.75–0.98, P = 0.023; odds ratio: 0.97, 95% CI: 0.94–0.99, P = 0.012, Table 3). Interestingly, fat accumulation in the thigh muscle especially in semitendinosus and sartorius was shown to be significantly correlated with impaired glucose regulation (odds ratio: 1.10, 95% CI: 1.02–1.20, P = 0.021; odds ratio: 1.10, 95% CI: 1.02–1.19, P = 0.019; Table 3).
Univariate analysis for factors associated with impaired glucose metabolism in all the patients with obesity.
| OR | 95% CI | P | |
|---|---|---|---|
| Age (years) | 1.03 | 0.99–1.07 | 0.161 |
| Gender (female) | 2.18 | 0.99–4.83 | 0.054 |
| HC (cm) | 1.04 | 1.00–1.08 | 0.059 |
| AST (IU/L) | 1.01 | 1.00–1.03 | 0.155 |
| Alb (g/L) | 0.86 | 0.75–0.98 | 0.023 |
| Cre (μmol/L) | 0.97 | 0.94–0.99 | 0.012 |
| TG (mmol/L) | 1.57 | 1.08–2.28 | 0.018 |
| NEFA (mmol/L) | 16.05 | 1.54–167.46 | 0.020 |
| Muscle FF% | |||
| Semitendinosus | 1.10 | 1.02–1.20 | 0.021 |
| Adductor magnus | 1.18 | 0.98–1.42 | 0.075 |
| Gracilis | 1.07 | 0.99–1.15 | 0.077 |
| Sartorius | 1.10 | 1.02–1.19 | 0.019 |
| Mean thigh muscle | 1.14 | 0.99–1.31 | 0.061 |
Alb, albumin; AST, aspartate aminotransferase; Cre, creatinine; FF%, fat fraction; HC, hip circumference; NEFA, non-esterified fatty acid; OR, odds ratio; TG, triglycerides.
Multivariable logistic regression analyses further revealed that higher FF% values in semitendinosus and sartorius were the independent explanatory variables associated with impaired glucose metabolism after adjustment for age and gender (odds ratio: 1.09, 95% CI: 1.00–1.19, P = 0.040; odds ratio:1.10, 95% CI: 1.02–1.19, P = 0.016, respectively, Table 4). In addition, the association of FF% values in semitendinosus with impaired glucose metabolism remained significant after adjusting age, gender, BMI, TG, Cre and Alb (odds ratio: 1.13, 95% CI: 1.02–1.26, P = 0.024, Table 4), while this relationship was attenuated in sartorius after adjustment (odds ratio: 1.08, 95% CI: 0.99–1.19, P = 0.091, Table 4).
Multivariate analysis between the thigh muscle fat fraction and impaired glucose metabolism in all the patients with obesity.
| Muscle FF% | Model 1 | Model 2 | Model 3 | ||||||
|---|---|---|---|---|---|---|---|---|---|
| OR | 95% CI | P | OR | 95% CI | P | OR | 95% CI | P | |
| Semitendinosus | 1.10 | 1.02–1.20 | 0.021 | 1.09 | 1.00–1.19 | 0.040 | 1.13 | 1.02–1.26 | 0.024 |
| Adductor magnus | 1.18 | 0.98–1.42 | 0.075 | 1.13 | 0.94–1.37 | 0.193 | 1.23 | 0.97–1.55 | 0.089 |
| Gracilis | 1.07 | 0.99–1.15 | 0.077 | 1.06 | 0.99–1.15 | 0.110 | 1.05 | 0.96–1.15 | 0.298 |
| Sartorius | 1.10 | 1.02–1.19 | 0.019 | 1.10 | 1.02–1.19 | 0.016 | 1.08 | 0.99–1.19 | 0.091 |
| Mean thigh muscle | 1.14 | 0.99–1.31 | 0.061 | 1.13 | 0.98–1.29 | 0.097 | 1.15 | 0.97–1.37 | 0.110 |
Model 1: Unadjusted.
Model 2: Adjusted for age and gender, respectively.
Model 3: Adjusted for age, gender, BMI, TG, Cre and Alb, respectively.
FF%, fat fraction; OR, odds ratio.
ROC analyses showed that at an optimal cutoff expression level of 10.32%, the sensitivity and specificity for FF% of semitendinosus as a predictor of impaired glucose regulation were 66.7% and 59.3%, respectively, with an area under the ROC (AUC) of 0.633 (Fig. 2A). ROC analysis also showed that for FF% of sartorius, the cutoff value was 10.97%, the sensitivity was 87.9% and its specificity was 51.9% with an AUC of 0.647 (Fig. 2B).
Receiver operating characteristic curve (ROC) plot of indicators for diagnosis of impaired glucose regulation. (A) ROC curve analyses of semitendinosus FF% (blue line, AUC = 0.633) and sartorius FF% (black line, AUC = 0.647). (B) ROC curve analyses of three parameters (green line) and five parameters (red line). Three parameters (AUC = 0.771): TG, NEFA and HOMA-IR. Five parameters (AUC = 0.805): semitendinosus FF%, sartorius FF%, TG, NEFA and HOMA-IR.
Citation: Endocrine Connections 12, 11; 10.1530/EC-23-0248
In addition, the AUC of three-component prediction model (including TG, NEFA and HOMA-IR) was 0.771. The sensitivity and specificity of impaired glucose regulation were 82.4% and 62.8% respectively (Fig. 2B). The AUC of five-component prediction model (including TG, NEFA, HOMA-IR, FF% of semitendinosus and FF% of sartorius) was 0.805. The sensitivity and specificity of impaired glucose regulation were 78.4% and 69.8%, respectively (Fig. 2B).
Discussion
In the present study, we observed that there were significant differences in IMAT in semitendinosus, adductor magnus, gracilis and sartorius between the NGT and IGR groups in subjects with obesity. In gender-specific analyses, we showed that female subjects have higher IMAT in adductor magnus than male subjects. Male subjects with IGR had higher IMAT in semitendinosus and sartorius than those with NGT. Among all the patients, those with IGR had higher muscle fat content in semitendinosus and sartorius than those with NGT, independent of age and gender. Semitendinosus fat faction was an independent risk factor of impaired glucose metabolism in patients with obesity.
Up to now, the exact mechanisms underlying the relationship between skeletal muscle fat and impaired glucose regulation remain unclear. Previous studies have provided evidence for the cross talk between human adipocytes and skeletal muscle cells. IMAT is characterized by a highly immunogenic and inflammatory secretome, associated with inflammatory cytokines, extracellular matrix proteins and increasing interstitial free fatty acid (FFA) levels in the muscle (15). It has been reported that IMAT secretes significantly more eicosanoids, resistin and hepatocyte growth factor than subcutaneous adipose tissue (SAT) and VAT, which are associated with inflammatory state and insulin sensitivity in skeletal muscle (28). The secretome of obese adipocytes may affect the muscle function by downregulating the expression of genes related to the skeletal muscle contractility complex and myogenesis (29). Another group found that the adipose tissue secretions could also suppress the insulin signaling in human skeletal myotubes (30). Given its anatomical location, IMAT secretions are likely to influence glucose regulation and insulin sensitivity in adjacent muscle cells. This proximity suggests that therapeutic interventions targeting the IMAT may be promising in the prevention and treatment of diabetes (15, 28). Nevertheless, no differences in serum inflammatory cytokine levels were observed between NGT and IGR patients with obesity in our study, which may be related to our small sample size. Further studies are needed to elucidate the relationship between adipose tissue in skeletal muscle and chronic inflammation.
Earlier data have consistently shown that ectopic fat accumulation in thigh muscle is correlated with insulin resistance and glucose dysregulation. Goodpaster et al. reported that IMAT was higher in the thigh of patients with obesity and T2DM than that in patients with obesity and in lean glucose-tolerant subjects (31). The same group conducted another cross-sectional study of 2964 elderly men and women by CT scans. Similarly, they found the intermuscular fat was higher in individuals with T2DM and IGT than in normoglycemic controls (17). In line with these studies, our findings also showed that IMAT in semitendinosus and sartorius was higher in IGR subjects than in the NGT group among the adults with obesity.
Kim et al. (32) reported that IMAT was related to higher fasting serum glucose concentration and HOMA-IR in middle-aged and older American subjects with BMI between 25 and 38 kg/m2. Yu et al. (33) assessed the adipose tissue in 24 young Chinese male adults of normal weight using MRI and showed that IMAT was associated with higher HOMA-IR, HbA1c and 2-hPBG. In our present study, we did not find an association between IMAT and HOMA-IR (Supplementary Table 2) in the whole population. Differences in the study population may explain this discrepancy. Previous studies have included healthy lean subjects as a control group. The correlations in healthy controls between IMAT and impaired glucose regulation were analyzed. In our study, all recruited subjects were Chinese subjects with a BMI over 28 kg/m2. Baseline glucose profiles such as FBG, insulin levels and HOMA-IR may differ between healthy lean controls and obese patients.
Further analyses showed that in the IGR group, there was a positive correlation between thigh muscle FF% and HOMA-IR (Supplementary Table 3). In addition, our results indicated a positive correlation between thigh muscle IMAT and 30-min and 60-min post-load glucose levels (Supplementary Table 4). Taken together, these findings suggest a negative impact of increasing fat accumulation in skeletal muscle on glucose homeostasis. We presented that fat accumulation in semitendinosus and sartorius was correlated with glucose metabolic status after adjusting for age and gender in individuals with obesity. After further adjustment for other clinical parameters, the association between IMAT in sartorius and IGR was attenuated, whereas the associations remained for IMAT in semitendinosus. This finding may provide new evidence on the role of IMAT in glucose metabolism.
In our study, female subjects had a higher FF% of adductor magnus than males. A positive correlation between the FF% of sartorius and HbA1c was found in male subjects but not in females. In addition, the FF% of semitendinosus and sartorius were significantly higher in the IGR group than in the NGT group among the male subjects, which were not observed in female subjects. It is possible that the sample size has limited the statistical significance of the correlation as there are only 40 female individuals in the study. However, there was little research on the relationship between glucose metabolism and specific thigh muscles, especially the semitendinosus. Further studies are warranted to address this issue.
Nowadays, with an increasing number of studies focusing on the role of muscle fat in obesity, appropriate methods to assess the adipose tissue in skeletal muscle are important. MRI is considered as the non-invasive gold standard technique for quantifying adipose tissue in recent decades (34, 35). Currently, the most widely used methods for quantifying IMAT include magnetic resonance spectroscopy (MRS) and MRI Dixon. With excellent image resolution, Dixon technique can access fat from surrounding muscle tissue and provide accurate results of fat accumulation and distribution. Additionally, the absence of radiation and high repeatability made it superior to other methods (36, 37) and safer to use in all age groups.
Our study was mainly based on the PDFF measured by the mDIXON-Quant quantitative technique. Kiefer et al. found significant differences in the amount and distribution of fat in psoas major and autochthonous back muscles among subjects with T2DM, prediabetes and healthy controls (38), indicating that IMAT of specific abdominal skeletal muscles was associated with glycemic status. Similarly, fat content in specific thigh muscles (semitendinosus and sartorius) was related to impaired glucose regulation in our findings. These results further supported that the fat distribution in skeletal muscle may be a potential imaging marker of abnormal glycometabolism.
Several limitations of the present study should be acknowledged. Firstly, our study failed to include a healthy lean group. Although we observed a marked difference in the FF% of the thigh muscle between NGT and IGR patients with obesity, a comparison between subjects with obesity and subjects with normal weight would be desirable. Secondly, the sample size is relatively small. It is possible that the sample size has limited the statistical significance of the correlation between the mean thigh muscle FF% and IGR, rather than a true differential association between the FF% in a specific group of thigh muscle and IGR. Besides, a small sample size may limit the statistical significance of the difference in AUC (AUC difference: 0.034, 95% CI −0.020–0.088, P = 0.342). However, net reclassification improvement (NRI) and integrated discrimination improvement (IDI) are more sensitive to changes in discrimination performance than Harrell’s C-statistic (NRI: 0.252, 95% CI (0.055–0.449), P = 0.012) (IDI: 0.051, 95% CI 0.003–0.098, P = 0.037). The results of NRI and IDI may demonstrate the additive value of thigh muscle FF% to the prediction of IGR. As this is a cross-sectional study, causality cannot be established. Future prospective and large-scale studies with longer follow-up, especially after treatment, would be warranted. Thirdly, the research population in this study is Chinese. Therefore, the results cannot be generalized to other races. It requires more multicentric studies in the future to draw any significant correlation of thigh muscle fat and IGR in individuals with obesity. Fourthly, lifestyle factors such as diet, physical activity, smoking and alcohol consumption or other confounding risk factors such as family history of coronary heart disease, hypertension and diabetes, which could affect intramuscular fat storage, were not collected in this study. The effects of these factors may be studied and correlated with the fat accumulation in the body (subcutaneous, visceral and ectopic fats). Lastly, although the adipose tissue segmentation method we used has high reproducibility and inter- and intra-reliability, future studies should consider using the new method to minimize user dependence.
Conclusions
There are significant differences in IMAT of thigh muscles between NGT and IGR subjects with obesity. Our findings indicate that higher fat content in semitendinosus is associated with glucose dysregulation in individuals with obesity.
Supplementary materials
This is linked to the online version of the paper at https://doi.org/10.1530/EC-23-0248.
Declaration of interest
The authors have no disclosures or conflicts of interest to report.
Funding
This work was supported by Shanghai Medicine and Health Development Foundation (DMRFP_I_06), 2019 management construction project of hospital (CHDI-2019-A-01), Ministry of Education, Science and Technology Development Center-New Generation of Information Technology Innovation Program (2019ITA01004) and Science and Technology Commission of Shanghai Municipality-Science and Technology Program(20DZ2201500).
Author contribution statement
XL analyzed data and drafted the manuscript. JY and QL collated clinical data of these patients. SH, YQ and MZ collected the clinical data of the patients. MY, YD, HX and WZ provided clinical support for the study. SH and QL collected the MRI data of the patients. JM conceived, designed and supervised the study. JM and QL reviewed and edited the manuscript. All authors contributed to the article and approved the submitted version.
Acknowledgements
The authors would like to appreciate the guidance and help on the statistical analysis of this article to the statistician Jieying Wang from the Clinical Research Center of Renji Hospital, Shanghai Jiao Tong University School of Medicine. The authors would like to thank all patients and investigators for their cooperation and contributions.
References
- 1↑
Kelly T, Yang W, Chen CS, Reynolds K, & He J. Global burden of obesity in 2005 and projections to 2030. International Journal of Obesity 2008 32 1431–1437. (https://doi.org/10.1038/ijo.2008.102)
- 2↑
Pan XF, Wang L, & Pan A. Epidemiology and determinants of obesity in China. Lancet Diabetes and Endocrinology 2021 9 373–392. (https://doi.org/10.1016/S2213-8587(2100045-0)
- 3↑
Neeland IJ, Ross R, Despres JP, Matsuzawa Y, Yamashita S, Shai I, Seidell J, Magni P, Santos RD, Arsenault B, et al.Visceral and ectopic fat, atherosclerosis, and cardiometabolic disease: a position statement. Lancet Diabetes and Endocrinology 2019 7 715–725. (https://doi.org/10.1016/S2213-8587(1930084-1)
- 4↑
Piche ME, Tchernof A, & Despres JP. Obesity phenotypes, diabetes, and cardiovascular diseases. Circulation Research 2020 126 1477–1500. (https://doi.org/10.1161/CIRCRESAHA.120.316101)
- 5↑
Yang M, Chen J, Yue J, He S, Fu J, Qi Y, Liu W, Xu H, Li S, Lu Q, et al.Liver fat is superior to visceral and pancreatic fat as a risk biomarker of impaired glucose regulation in overweight/obese subjects. Diabetes, Obesity and Metabolism 2023 25 716–725. (https://doi.org/10.1111/dom.14918)
- 6↑
Merz KE, & Thurmond DC. Role of skeletal muscle in insulin resistance and glucose uptake. Comprehensive Physiology 2020 10 785–809. (https://doi.org/10.1002/cphy.c190029)
- 7↑
DeFronzo RA, & Tripathy D. Skeletal muscle insulin resistance is the primary defect in type 2 diabetes. Diabetes Care 2009 32(Supplement 2) S157–S163. (https://doi.org/10.2337/dc09-S302)
- 8↑
Bianchi L, & Volpato S. Muscle dysfunction in type 2 diabetes: a major threat to patient's mobility and independence. Acta Diabetologica 2016 53 879–889. (https://doi.org/10.1007/s00592-016-0880-y)
- 9↑
Miljkovic-Gacic I, Wang X, Kammerer CM, Gordon CL, Bunker CH, Kuller LH, Patrick AL, Wheeler VW, Evans RW, & Zmuda JM. Fat infiltration in muscle: new evidence for familial clustering and associations with diabetes. Obesity 2008 16 1854–1860. (https://doi.org/10.1038/oby.2008.280)
- 10↑
McGarry JD. Banting lecture 2001: dysregulation of fatty acid metabolism in the etiology of type 2 diabetes. Diabetes 2002 51 7–18. (https://doi.org/10.2337/diabetes.51.1.7)
- 11↑
Brehm A, Krssak M, Schmid AI, Nowotny P, Waldhausl W, & Roden M. Increased lipid availability impairs insulin-stimulated ATP synthesis in human skeletal muscle. Diabetes 2006 55 136–140. (https://doi.org/10.2337/diabetes.55.01.06.db05-1286)
- 12↑
Perseghin G, Scifo P, De Cobelli F, Pagliato E, Battezzati A, Arcelloni C, Vanzulli A, Testolin G, Pozza G, Del Maschio A, et al.Intramyocellular triglyceride content is a determinant of in vivo insulin resistance in humans: a 1H-13C nuclear magnetic resonance spectroscopy assessment in offspring of type 2 diabetic parents. Diabetes 1999 48 1600–1606. (https://doi.org/10.2337/diabetes.48.8.1600)
- 13↑
Wu H, & Ballantyne CM. Skeletal muscle inflammation and insulin resistance in obesity. Journal of Clinical Investigation 2017 127 43–54. (https://doi.org/10.1172/JCI88880)
- 14↑
Kahn SE, Hull RL, & Utzschneider KM. Mechanisms linking obesity to insulin resistance and type 2 diabetes. Nature 2006 444 840–846. (https://doi.org/10.1038/nature05482)
- 15↑
Sachs S, Zarini S, Kahn DE, Harrison KA, Perreault L, Phang T, Newsom SA, Strauss A, Kerege A, Schoen JA, et al.Intermuscular adipose tissue directly modulates skeletal muscle insulin sensitivity in humans. American Journal of Physiology. Endocrinology and Metabolism 2019 316 E866–E879. (https://doi.org/10.1152/ajpendo.00243.2018)
- 16↑
Yu F, Fan Y, Sun H, Li T, Dong Y, & Pan S. Intermuscular adipose tissue in type 2 diabetes mellitus: non-invasive quantitative imaging and clinical implications. Diabetes Research and Clinical Practice 2022 187 109881. (https://doi.org/10.1016/j.diabres.2022.109881)
- 17↑
Goodpaster BH, Krishnaswami S, Resnick H, Kelley DE, Haggerty C, Harris TB, Schwartz AV, Kritchevsky S, & Newman AB. Association between regional adipose tissue distribution and both type 2 diabetes and impaired glucose tolerance in elderly men and women. Diabetes Care 2003 26 372–379. (https://doi.org/10.2337/diacare.26.2.372)
- 18↑
Goss AM, & Gower BA. Insulin sensitivity is associated with thigh adipose tissue distribution in healthy postmenopausal women. Metabolism: Clinical and Experimental 2012 61 1817–1823. (https://doi.org/10.1016/j.metabol.2012.05.016)
- 19↑
Yoshizumi T, Nakamura T, Yamane M, Islam AH, Menju M, Yamasaki K, Arai T, Kotani K, Funahashi T, Yamashita S, et al.Abdominal fat: standardized technique for measurement at CT. Radiology 1999 211 283–286. (https://doi.org/10.1148/radiology.211.1.r99ap15283)
- 20↑
Iyengar NM, Arthur R, Manson JE, Chlebowski RT, Kroenke CH, Peterson L, Cheng TD, Feliciano EC, Lane D, Luo J, et al.Association of body fat and risk of breast cancer in postmenopausal women with normal body mass index: a secondary analysis of a randomized clinical trial and observational study. JAMA Oncology 2019 5 155–163. (https://doi.org/10.1001/jamaoncol.2018.5327)
- 21↑
Anjana M, Sandeep S, Deepa R, Vimaleswaran KS, Farooq S, & Mohan V. Visceral and central abdominal fat and anthropometry in relation to diabetes in Asian Indians. Diabetes Care 2004 27 2948–2953. (https://doi.org/10.2337/diacare.27.12.2948)
- 22↑
Prado CM, & Heymsfield SB. Lean tissue imaging: a new era for nutritional assessment and intervention. Journal of Parenteral and Enteral Nutrition 2014 38 940–953. (https://doi.org/10.1177/0148607114550189)
- 23↑
Andreoli A, Garaci F, Cafarelli FP, & Guglielmi G. Body composition in clinical practice. European Journal of Radiology 2016 85 1461–1468. (https://doi.org/10.1016/j.ejrad.2016.02.005)
- 24↑
Kramer H, Pickhardt PJ, Kliewer MA, Hernando D, Chen GH, Zagzebski JA, & Reeder SB. Accuracy of liver fat quantification with advanced CT, MRI, and ultrasound techniques: prospective comparison with MR spectroscopy. American Journal of Roentgenology 2017 208 92–100. (https://doi.org/10.2214/AJR.16.16565)
- 25↑
Zhou BF & Cooperative Meta-Analysis Group of the Working Group on Obesity in China. Predictive values of body mass index and waist circumference for risk factors of certain related diseases in Chinese adults--study on optimal cut-off points of body mass index and waist circumference in Chinese adults. Biomedical and Environmental Sciences 2002 15 83–96.
- 26↑
Alberti KG, & Zimmet PZ. Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus provisional report of a WHO consultation. Diabetic Medicine 1998 15 539–553. (https://doi.org/10.1002/(SICI)1096-9136(199807)15:7<539::AID-DIA668>3.0.CO;2-S)
- 27↑
Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, & Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985 28 412–419. (https://doi.org/10.1007/BF00280883)
- 28↑
Kahn D, Macias E, Zarini S, Garfield A, Zemski Berry K, Gerszten R, Schoen J, Cree-Green M, & Bergman BC. Quantifying the inflammatory secretome of human intermuscular adipose tissue. Physiological Reports 2022 10 e15424. (https://doi.org/10.14814/phy2.15424)
- 29↑
Pellegrinelli V, Rouault C, Rodriguez-Cuenca S, Albert V, Edom-Vovard F, Vidal-Puig A, Clement K, Butler-Browne GS, & Lacasa D. Human adipocytes induce inflammation and atrophy in muscle cells during obesity. Diabetes 2015 64 3121–3134. (https://doi.org/10.2337/db14-0796)
- 30↑
Sarr O, Strohm RJ, MacDonald TL, Gaudio N, Reed JK, Foute-Nelong J, Dyck DJ, & Mutch DM. Subcutaneous and visceral adipose tissue secretions from extremely obese men and women both acutely suppress muscle insulin signaling. International Journal of Molecular Sciences 2017 18. (https://doi.org/10.3390/ijms18050959)
- 31↑
Goodpaster BH, Thaete FL, & Kelley DE. Thigh adipose tissue distribution is associated with insulin resistance in obesity and in type 2 diabetes mellitus. American Journal of Clinical Nutrition 2000 71 885–892. (https://doi.org/10.1093/ajcn/71.4.885)
- 32↑
Kim JE, Dunville K, Li J, Cheng JX, Conley TB, Couture CS, & Campbell WW. Intermuscular adipose tissue content and intramyocellular lipid fatty acid saturation are associated with glucose homeostasis in middle-aged and older adults. Endocrinology and Metabolism 2017 32 257–264. (https://doi.org/10.3803/EnM.2017.32.2.257)
- 33↑
Yu F, He B, Chen L, Wang F, Zhu H, Dong Y, & Pan S. Intermuscular fat content in young Chinese men with newly diagnosed Type 2 diabetes: based on MR mDIXON-quant quantitative technique. Frontiers in Endocrinology (Lausanne) 2021 12 536018. (https://doi.org/10.3389/fendo.2021.536018)
- 34↑
Thomas EL, Fitzpatrick JA, Malik SJ, Taylor-Robinson SD, & Bell JD. Whole body fat: content and distribution. Progress in Nuclear Magnetic Resonance Spectroscopy 2013 73 56–80. (https://doi.org/10.1016/j.pnmrs.2013.04.001)
- 35↑
Ross R, Leger L, Morris D, de Guise J, & Guardo R. Quantification of adipose tissue by MRI: relationship with anthropometric variables. Journal of Applied Physiology 1992 72 787–795. (https://doi.org/10.1152/jappl.1992.72.2.787)
- 36↑
Grimm A, Meyer H, Nickel MD, Nittka M, Raithel E, Chaudry O, Friedberger A, Uder M, Kemmler W, Engelke K, et al.Repeatability of Dixon magnetic resonance imaging and magnetic resonance spectroscopy for quantitative muscle fat assessments in the thigh. Journal of Cachexia, Sarcopenia and Muscle 2018 9 1093–1100. (https://doi.org/10.1002/jcsm.12343)
- 37↑
Kiefer LS, Fabian J, Lorbeer R, Machann J, Storz C, Kraus MS, Wintermeyer E, Schlett C, Roemer F, Nikolaou K, et al.Inter- and intra-observer variability of an anatomical landmark-based, manual segmentation method by MRI for the assessment of skeletal muscle fat content and area in subjects from the general population. British Journal of Radiology 2018 91 20180019. (https://doi.org/10.1259/bjr.20180019)
- 38↑
Kiefer LS, Fabian J, Rospleszcz S, Lorbeer R, Machann J, Storz C, Kraus MS, Schlett CL, Roemer F, Wintermeyer E, et al.Assessment of the degree of abdominal myosteatosis by magnetic resonance imaging in subjects with diabetes, prediabetes and healthy controls from the general population. European Journal of Radiology 2018 105 261–268. (https://doi.org/10.1016/j.ejrad.2018.06.023)
