Introduction

The 2019 Asian Working Group for Sarcopenia (AWGS2019) originally defined sarcopenia as “age-related loss of skeletal muscle mass plus loss of muscle strength and/or reduced physical performance” without reference to comorbidities, and the age cutoffs were set at 60 or 65 years of age [1]. Although sarcopenic characteristics may occur in younger adults, the underlying pathophysiology should be explored, rather than simply pursuing a sarcopenia diagnosis, to identify people at risk as early as possible for timely intervention and prevention of possible deterioration. The AWGS2019 recommended necessary interventions for sarcopenia, with or without specific clinical contributing conditions. For younger patients with uncontrolled or acute clinical conditions that may result in sarcopenia, the AWGS2019 suggests proper interventions to prevent sarcopenia from developing or deteriorating, in addition to proper treatment of the underlying clinical conditions. The AWGS2019 diagnostic criteria are used not only for older patients, but also for younger patients with heart failure, chronic obstructive pulmonary disease, diabetes mellitus, and chronic kidney disease [1]. Stroke-related sarcopenia (SRS) refers to a decrease in overall muscle mass, muscle function, or physical function in patients who have suffered a cerebral stroke [2,3]. At present, the diagnostic criteria of SRS are mainly based on the consensus of the AWGS2019, including measured values of grip strength (<28 kg in males and <18 kg in females) and skeletal muscle index (<7.0 kg/m² in males and <5.7 kg/m² in females) [4]. A simultaneous decrease in these parameters is indicative of SRS. In younger or older patients with SRS, stroke is the basic clinical condition that causes sarcopenia. Proper interventions should be applied to prevent sarcopenia from developing or deteriorating, in addition to the appropriate management of the underlying stroke condition. Using bioelectrical impedance analysis (BIA), the skeletal muscle index of the limbs is calculated as limb muscle mass (kg)/height² (m²). Nonetheless, most clinical departments do not routinely have BIA diagnostic equipment, and routine SRS imaging may be urgently needed considering the hydration status of the patient, patient/room temperature, and time of day, which may influence BIA. The correct diagnosis of SRS is closely related to the prognosis of patients who had a stroke, and immediate accurate diagnosis can lead to early treatment, thus reducing the occurrence and complications of stroke and improving the prognosis. Although medical imaging modalities can be used to diagnose SRS, there is a lack of clear research on the quantitative and qualitative changes in fat inside and outside skeletal muscle and subcutaneous tissue, and in skeletal muscle fiber cells. Computed tomography (CT), quantitative CT, and magnetic resonance imaging (MRI) have been used for skeletal muscle measurements [5-7], and dual-energy CT has also been used for quantification of skeletal muscle fat in experiments and patient validation [8-10]. However, few studies have been performed on changes in skeletal muscle fiber cells and their interstitial components using quantitative MRI.

Concerning MRI fat quantification techniques that are applicable to skeletal muscle, the mDIXON-Quant sequence (Philips Healthcare, Best, the Netherlands) is mainly based on chemical-shift-encoded water-fat separation imaging, using multi-echo acquisition and a peak fat model to accurately and quantitatively analyze muscle fat content [5]. The water-fat separation image is obtained by collecting signals of the hydrogen protons in water and fat in an inverse phase state through different echo times (TEs) [6]. The average muscle fat fraction (FF) is obtained using the formula: FF=fat signal intensity/(fat signal intensity+water image signal intensity)×100% [11]. Currently, FF is mainly used to measure fatty liver in clinical practice. Proton magnetic resonance spectroscopy (¹H-MRS) can be used to distinguish and quantitatively analyze intra- and extracellular lipids in skeletal muscle cells [6,12-14]. It has been found that ¹H-MRS has significant clinical value in the assessment and dynamic monitoring of skeletal muscle injury [12]. There is a significant correlation between the total lipid content of the skeletal muscles and the degree and duration of muscle damage [13]. SRS may aggravate the prognosis of patients who had a stroke, resulting in a longer total hospital stay, higher incidence of complications, higher hospitalization costs, poor rehabilitation outcomes, slower recovery, and higher mortality rates than in patients without SRS who had a stroke (non-SRS) [2,15-18]. Therefore, early diagnosis and intervention are critical. It was hypothesized that mDIXON-Quant and ¹H-MRS could be effectively used to evaluate the muscle and fat content in the skeletal muscles of patients with SRS. Therefore, this study aimed to characterize the muscle cross-sectional area (CSA) and FF of patients with SRS compared with healthy controls using mDIXON-Quant and ¹H-MRS.

Methods

Ethics statement: This prospective, single-center study was approved by the Ethics Committee of The Second Hospital of Kunming Medical University (No: FEY-BG-39-2.0), and all participants provided signed informed consent.

1. Subjects

Patients with stroke-related hemiplegia were prospectively enrolled between January 2020 and December 2022 in our hospital and divided into SRS and non-SRS groups (patients without sarcopenia who had a stroke) for comparative analysis. Patients with ischemic or hemorrhagic strokes were enrolled if they were aged 30 to 50 years, experienced their first stroke within 6 months before admission, had body mass index (BMI) >18.5 kg/m² and <24 kg/m², had no significant intellectual or cognitive impairment (Mini-Mental State Examination [MMSE] score ≥27) [19], and had muscle strength of at least Brunnstrom grade 4 on one side. Ischemic and hemorrhagic strokes were defined as focal cerebral infarcts or hemorrhages with neurological deficits as confirmed by diffusion-weighted MRI and/or CT. SRS was defined as “muscle failure due to stroke” characterized by any loss of muscle mass and decrease in muscle strength or physical function. According to the AWGS2019, the diagnostic criteria were a grip strength <28 kg in males and <18 kg in females and a skeletal muscle index <7.0 kg/m² in males and <5.7 kg/m² in females in BIA [4]. Patients who met these sarcopenia criteria were assigned to the SRS group, and those who did not meet these criteria were assigned to the non-SRS group. The exclusion criteria were (1) patients with serious infectious diseases (e.g., sepsis), malignant tumor, mental disease, chronic bronchitis, emphysema, hereditary muscle disease, pre-stroke myasthenia gravis, mitochondrial disease, lipid storage myopathy, muscular hypertrophy, pseudohypertrophy, other neuromuscular diseases (i.e., pre-stroke symptoms such as myasthenia, fatigue, myalgia and tenderness, and muscular atrophy), osteoarthritis, transient ischemic attacks, pregnancy, kidney dialysis, severe limb spasms, and concurrent heart, liver, kidney, and lung dysfunction or failure; (2) patients who failed to perform the BIA because of hand and foot fractures, amputation, presence of metal foreign bodies (e.g., pacemakers, stents, and steel plates), or oral diuretics affecting body composition; (3) patients with severe coma, intellectual and mental disorders or cognitive impairments who were unable to perform grip strength and body composition analysis tests; (4) patients with MRI contraindications; and (5) severe upper limb spasms or pain affecting measurement of grip strength. Healthy individuals without any diseases were enrolled as the control group.

2. Magnetic resonance imaging examination methods and image post-processing

Scanning equipment and parameters: Philips Achieva 3.0T superconducting MRI equipment (Philips Healthcare) was used with abdominal torso coils. Scanning was performed in the supine position using an advanced head-scanning method. The flat scanning range was from the femoral head to the tibial plateau and the mDIXON scanning position was from the middle layer of the femoral head to the tibial plateau. The designed scanning parameters were packaged for routine MRI T1-weighted imaging (T1WI)/ T2-weighted imaging (T2WI)/¹H-MRS scanning, using the following sequences: T1WI-turbo spin echo (TSE). (repetition time [TR], 260 milliseconds; TE, 15 milliseconds; field-of-view [FOV], 280×379×99 mm; voxel, 0.75×0.85×4 mm; matrix, 372×445×20 slices; slice thickness, 4 mm; and slice gap, 4 mm), T2WI-TSE-breath-hold (BH) (TR, 950 milliseconds; TE, 80 milliseconds; FOV, 400×442×149 mm; voxel, 113×16×5 mm; matrix, 308×276×25 slices; slice thickness, 1 mm; and slice gap, 4 mm), mDIXON-Quant-BH (ATR, 9.1 milliseconds; TE, 1.33 milliseconds; FOV, 400×421×180 mm; voxel, 2.5×2.5×6 mm; matrix, 160×168×60 slices; slice thickness, 3 mm; and slice gap, 3 mm), and SV-PRESS-35 (TR/TE, 2,000/34 milliseconds; MinTR/TE, 729/34 milliseconds; spectral resolution, 1.95 Hz/p; readout duration, 512 milliseconds; local torso specific absorption rate [SAR], <16%; whole body SAR/level, <0.2 W/kg/normal; specific energy dose, <0.1 kJ/kg; MaxB1+rms, 0.64 μT; PNS/level, 33%/normal; volume of interest [VOI], 39.94×128.51×61.85 mm; dB/dt, 29.8 T/sec; sound pressure level, –7.2 dB; FOV, 400×421×180 mm; VOI, 12×12×12 mm; spectral bandwidth, 2,000 Hz; flip angle, 90°; window, 140 Hz; second pulse angle, 300°; startup acquisitions, 2; number of signal averages, 128; and phase cycle, 16). The mDIXON images were then scanned from the middle layer of the thigh, with a total scanning time of 21 minutes. The image data obtained from the MRI scans were uploaded to the PACS (picture archiving and communication system) workstation for post-processing using the function software. Two senior MRI attending physicians used a double-blind method to select the region of interest (ROI) with a diameter of 10 mm at the center of each muscle in the T1WI sequence for the automatic measurement of CSA using software. Fat, blood vessels, bone tissue, fascia, and artifacts were avoided throughout the ROI. Simultaneously, saturation zones were added around the ROI to eliminate the influence of surrounding tissues and respiratory movements. The mDIXON post-processing analysis mainly included the measurement of the FF, spectral lip 1 and lip 2 peak heights of the rectus femoris muscle, and the area under the peak. Height 1 of the lip 1 peak represented the highest intracellular fat content in skeletal muscle cells, while area 1 under the lip 1 peak represented the intracellular fat concentration in skeletal muscle cells. Height 2 of the lip 2 peak represented the highest extracellular fat content in skeletal muscle cells, whereas area 2 under the lip 2 peak represented the concentration of extracellular fat in skeletal muscle cells. The fat at a displacement of 1.5 ppm (parts per million) represents the methylene extramyocellular lipids (EMCL) stored in the extracellular lipids in the subcutaneous and muscle spaces and in the adipocytes within the connective tissue between the muscle fibers. The lipid at a displacement of 1.2 ppm represents the lipid stored in the muscle cells in the form of triglycerides in lipid droplets.

3. Clinical parameters

After the participants who met the selection criteria fasted for 8 hours on the evening of admission, their body weights and heights were measured three times, and the average value was calculated. A Jamar Plus digital grip meter was used to measure grip strength three times, and the average value was calculated. The human body composition analyzer InBodyS10 was used with multi-frequency BIA to detect the skeletal muscle and body fat contents, body FF, and limb skeletal muscle mass index of the patients. Imaging evaluation using a plain MRI scan, mDIXON-Quant, and ¹H-MRS was conducted by two radiologists with 5 years of experience. If there was a disagreement, a third senior radiologist was involved to reach a consensus.

4. Statistical analysis

Statistical analyses were performed using IBM SPSS ver. 26.0 (IBM Corp., Armonk, NY, USA). Continuous variables were tested for normality; normally distributed data are shown as mean±standard deviation and non-normal data as median and interquartile range. One-way analysis of variance was used to compare the FF and muscle CSA among the three groups (SRS, non-SRS, and healthy controls). Pairwise comparisons between groups were conducted using the least significant difference method, and the rank-sum test was used when the variances were uneven. Receiver operating characteristic curve analysis was conducted on the peaks and areas under the curves of CSA, FF, lip 1, and lip 2 to obtain the optimal cutoff values. Statistical significance was set at p<0.05.

Results

1. Participants

Eighty patients with cerebral stroke were enrolled and divided into the SRS (n=40) and non-SRS (n=40) groups according to the standards for SRS. All patients underwent a plain MRI scan, mDIXON-Quant, and ¹H-MRS examination (Table 1). No significant (p>0.05) difference was detected in the male sex percentage (95.0% vs. 97.5%); age (46.2±7.1 years vs. 42.6±6.5 years); BMI (21.6±4.6 kg/m² vs. 24.0±2.3 kg/m²); disease course (53 days vs. 49 days); percentage of patients with cerebral hemorrhage (45.0% vs. 47.5%), infarction (55.0% vs. 52.5%), or left paralysis (82.5% vs. 72.5%); and MMSE score (27.8±3.5 vs. 26.9±5.8) between the SRS and non-SRS groups (Table 1). Forty healthy individuals were enrolled in the control group, and none of these values were significantly different (p>0.05).

2. Comparison of skeletal fat fraction in three groups

There was a significant difference (p<0.05) in the skeletal muscle FF values among the three groups (SRS, non-SRS, and healthy controls). The skeletal muscle FF value was significantly higher (p<0.05) in the SRS group than in the non-SRS group or normal controls, but was not significantly different (p>0.05) between the normal controls and non-SRS group (Tables 2–4, Figs. 1, 2). The skeletal muscle FF was significantly higher (p<0.05) on the affected side than on the healthy side.

3. Height and area of lip 1 and 2 peaks on proton magnetic resonance spectroscopy

The extracellular fat content (height 2 and area 2) of the rectus femoris muscle in the control group was 4× to 10× the intracellular fat content (height 1 and area 1). A significant increase (p<0.05) in intracellular (height 1 and area 1) and extracellular (height 2 and area 2) fat content was detected in the SRS group compared to that in the non-SRS and normal control groups. There was no significant difference (p>0.05) in intracellular or extracellular fat content between the non-SRS and normal control groups or between the SRS and non-SRS groups. The degree of increase in extracellular fat content (height 2 and area 2) was significantly smaller (p<0.05) than that in the intracellular fat content (height 1 and area 1) in the SRS group, with an extracellular/intracellular fat content ratio of <4 (Table 6). The degree of increase in intracellular fat content was significantly greater (p<0.05) in the SRS group than in the non-SRS group (mean, 7.521±0.834; 95% confidence interval [CI], 0.70–10.84; p=0.022) or in healthy controls (Table 6), and the degree of increase in extracellular fat content was also significantly greater (p<0.05) in the non-SRS group than in the control group (mean, 6.901±0.724; 95% CI, 0.24–9.39; p=0.041). The ratio of extracellular to intracellular fat content gradually decreased among the normal control, non-SRS, and SRS groups. The ratio gradually decreased from 11× in the normal control group to 2× in the SRS group, and the absolute value of the intracellular/extracellular fat content gradually increased. The extracellular fat content was approximately 4× to 10× the intracellular fat content. When sarcopenia occurred, the ratio of extracellular to intracellular fat content was <4. In the SRS group, the increase in intracellular fat content was more significant than that in extracellular fat content.

Correlation analysis between the fat percentage of each muscle and the intra- and extracellular fat content of the rectus femoris muscle indicated a moderate-to-strong positive correlation between the intracellular fat content (area 1) and muscle fat percentage, suggesting that the muscle FF represented by mDIXON mainly reflected the intracellular fat content (Table 5).

4. Comparison of muscle and fat cross-sectional areas in thigh skeletal muscles among the three groups

There was no significant difference (p>0.05) in muscle CSA between the SRS and non-SRS groups or between the non-SRS and control groups; however, the CSA of the thigh muscles was significantly smaller (p<0.05) in the SRS group than in the control group. In the SRS group, the degree of decrease in CSA was significantly greater (p<0.05) in the posterior muscle group than in the anterolateral or medial muscle group (Fig. 4, Table 4). The thigh muscle CSA was significantly smaller in patients with SRS (4,656.8 cm²) than in healthy controls (8,530.5 cm²), and the thigh fat CSA was significantly greater in patients with SRS (3,189.6 cm²) than in healthy controls (1,963.2 cm²) (Fig. 5). In patients with SRS, the thigh muscle CSA was significantly smaller on the SRS-affected side than on the healthy side, and the thigh fat CSA was significantly greater on the SRS-affected side than on the healthy side.

Discussion

1. Major findings

In this study, the imaging features of mDIXON-Quant and ¹H-MRS of the thigh muscles in patients with SRS were investigated, and it was found that the thigh muscle CSA in SRS was significantly decreased, while the FF value was increased. The intra- and extracellular fat contents of the skeletal muscle were significantly increased, especially the intracellular fat content. SRS was confirmed when the ratio of extracellular to intracellular fat content was <4.

2. Clinical value and advantages of various quantitative methods for measuring muscle mass

Skeletal muscle mass can be assessed via adjusted BIA validated by dual-energy X-ray absorptiometry, enabling precise calculation of total body muscle mass and differentiation of muscular distribution in the limbs (upper and lower extremities) and trunk. This helps identify site-specific sarcopenia but cannot visually reflect multilayered changes in fat distribution and muscle volume [20]. BIA estimates muscle mass by applying a micro-alternating current to the human body and using the differences in resistance and reactance of the current in different tissues, such as fat and muscle. BIA can reflect limb muscle mass and total body fat percentage but cannot be used for cross-sectional analysis or visual characterization of muscle changes [21]. CT generates tomographic images of the muscle tissue, allowing precise measurement of muscle CSA (such as the thigh and paraspinal muscles). CT values can assess muscle density, but the specific CSA of each muscle and intermuscular fat within the cross-section cannot be accurately assessed [22]. Currently, MRI is the optimal method for soft tissue imaging. With multiparameter and multiplanar imaging, high soft tissue resolution, and clear visualization of anatomical relationships, MRI can determine biochemical characteristics, including lesion size, scope, blood supply, and tissue metabolic status, thus playing crucial roles in disease diagnosis, differential diagnosis, and prognosis assessment. Magnetic resonance spectroscopy (MRS) is a histochemical analysis method that uses nuclear magnetic resonance and chemical shifts to non-invasively detect tissue metabolism and biochemistry, quantitatively analyze compounds in living tissues, and obtain spectral information on tissue metabolites for early disease diagnosis and treatment efficacy assessment. Research on MRS has mainly focused on the brain and prostate, with relatively few studies on sarcopenia of the musculoskeletal system [12]. ¹H-MRS can distinguish and quantitatively analyze intra- and extracellular lipids in skeletal muscle. The total lipid content of the skeletal muscles is significantly correlated with the degree and duration of muscle injury. ¹H-MRS has important clinical applications in the assessment and dynamic monitoring of skeletal muscle injuries. However, no relevant quantitative studies have indicated changes in the percentage of fat in the skeletal muscle of patients with SRS as well as changes in signal intensity and fat content between the affected and healthy sides of the skeletal muscle. As a noninvasive fat quantitative imaging biomarker technique, mDIXON technology is based on chemical shifts. The hydrogen protons in water and fat molecules exist in different molecular environments, leading to inconsistent resonance frequencies. Using this difference to apply the in-phase and out-of-phase echo signals, pure water and pure fat images are obtained. Currently, the percentage of liver fat content is used to evaluate intrahepatic lipid content, enabling large-scale screening and long-term follow-up monitoring of disease conditions [23,24]. The magnetic resonance mDIXON-Quant technique allows precise measurement of liver fat content, is convenient and time-saving, and provides important references for early diagnosis, dynamic monitoring of the degree of fat infiltration, and prognostic assessment, demonstrating high clinical value. Muscle fatty degeneration is a common pathological feature in muscle diseases. Typically, the more severe the condition, the higher the fat content is in the muscle. Therefore, using the most precise magnetic resonance, mDIXON-Quant, and H-MRS to measure the fat content in the middle segment of the thigh skeletal muscle can predict and evaluate the fat content of the skeletal muscle in the entire body.

3. Correlation between reduced cross-sectional area of thigh skeletal muscle in stroke-related sarcopenia and increased fat infiltration

Our study found that the CSA of the thigh muscles gradually decreased from the normal controls to the non-SRS and SRS groups, which has clinical significance in assessing sarcopenia in patients with stroke. Thigh muscle CSA was significantly lower in the SRS group than in the control group. The CSA of the muscles on both the paralyzed and healthy sides decreased in the SRS group, and the paralyzed side showed a more significant reduction than the healthy side, with a reduction of approximately 20% to 30%. The non-SRS group also had a smaller muscle CSA on the paralyzed side than on the healthy side; however, the reduction in muscle size on the paralyzed side was smaller. After a stroke, the volume and mass of the lower limb muscles decreased, and the degree of limb atrophy was more pronounced on the affected side than on the healthy side. At the same time, subcutaneous fat and intermuscular fat infiltration were more pronounced on the paralyzed side than on the healthy side. These changes were more pronounced in the SRS group than in the non-SRS group. Related studies have found that patients who had a stroke 6 months ago have a 24% reduction in thigh muscle CSA on the paralyzed side compared to the healthy side, a 5% increase in subcutaneous fat, and a 17% increase in intermuscular fat [15,25]. A retrospective study also showed that the average CSA of the mid-thigh muscles on the paralyzed side decreased by 15.4 cm² compared to that on the healthy side after stroke [26,27].

Muscle reduction in SRS is not merely caused by simple muscle disuse, reduced exercise, or malnutrition. Loss of central nervous system inhibition and reduction of α motor neurons caused by post-stroke injury lead to changes in synapses and interruption of cortical spinal cord nutrient input in the ascending system, resulting in degeneration of motor neurons [28-30]. This subsequently results in loss of muscle innervation and muscle fiber atrophy [26,31] after stroke, in addition to the role played by decreases in muscle function, metabolism, and mitochondrial function.

The posterior group of muscles, including the medial heads of the biceps femoris, semitendinosus, and semimembranosus muscles, showed the most significant fat infiltration, followed by the medial group of muscles (i.e., the greater adductor, longer adductor, and gracilis muscles). The anterolateral muscle group showed less obvious fat infiltration, including the rectus femoris, lateral femoral muscle, and medial femoral muscle. Simultaneously, we found that the muscles with the most obvious fat infiltration had the most significant CSA atrophy. The degree of muscle fat infiltration gradually increased from the normal control group to the non-SRS and SRS groups. After a stroke, central activation of the flexor muscle decreases, and the activation ability of the motor unit decreases, leading to an increase in intramuscular fat (fat in the fascia and abdominal muscles). The reasons for fat infiltration include muscle disuse, sex hormone deficiency, and changes in leptin signaling, which can lead to a decrease in insulin sensitivity and protein synthesis, possibly promoting skeletal muscle atrophy and muscle strength decline [27].

4. Proton magnetic resonance spectroscopy characteristics of stroke-related sarcopenia thigh skeletal muscle

¹H-MRS is mainly based on the different precession frequencies and displacements of hydrogen protons in water and fats. According to the different intra- and extracellular displacement lipid peaks, the fat content can be quantitatively obtained by measuring the relevant peak and area under the peak [32]. The area under the peak is directly proportional to the resonance number of the hydrogen nuclei representing the concentration of the metabolites and can be used for quantitative analysis. The peak positions of the spectra of lipids in different compounds are located at 1.2 and 1.5 ppm, which correspond exactly to the normal intra- and extracellular lipid peaks, respectively. These peaks mainly represent the free fatty acids contained in the skeletal muscle tissue, namely tiny triglycerides [33]. Because of its high accuracy and repeatability, this technique is widely used in measurements of muscle fat content; however, the only drawback is the assessment of fat changes in a small area of 10 to 12 mm² of the delineated ROI. Among the three groups, the peak with the highest signal intensity from skeletal muscle cells (mostly single peaks) was the extracellular lipid peak, lip 2, located at 1.5 ppm. The intracellular lipid peak, lip 1, had a lower peak beside the lip 2 peak, was located at 1.2 ppm, and could be considered a double peak with lip 2. Creatine was located at 3.03 ppm, and the positions of these three metabolites were consistent with those of previous reports. The difference in chemical shift between the intra- and extracellular lipid peaks (0.3 ppm) was consistent with relevant studies reported in the literature [34]. Our study found that the intra- and extracellular fat content (EMCL) under the spectral peak curve gradually increased from healthy controls to the non-SRS group, and finally to the SRS group, with a gradual increase in the intra- and extracellular fat content of the rectus femoris muscle cells. The extracellular fat content of the rectus femoris muscle was significantly higher than the intracellular fat content in all three groups.

Our study detected a significant increase (p<0.05) in intracellular (height 1 and area 1) and extracellular (height 2 and area 2) fat content in the SRS group compared to that in the non-SRS and normal control groups. Nonetheless, there was no significant difference (p>0.05) in the intracellular or extracellular fat content between the non-SRS and normal control groups or between the SRS and non-SRS groups. One of the reasons for this finding is that some patients develop sarcopenia and others do not within the same time after a stroke, which is related to the location of the stroke and brain functional areas. For example, strokes that occur in the motor and brainstem areas damage motor and eating functions, leading to adaptive changes in skeletal muscle cell metabolism with quickly increasing muscle catabolism and fat metabolism. For patients without SRS after stroke, the impact of the location of cerebral infarction or bleeding on muscles is not very rapid or intense; consequently, muscle catabolism and fat anabolism are not significantly increased. Therefore, patients with or without SRS after stroke may have different degrees of muscle catabolism and fat anabolism, which has certain clinical significance for the evaluation of patients with SRS.

We conducted a consistency analysis between ¹H-MRS and mDIXON MRI quantitative analyses of thigh skeletal muscle fat content and that muscle FF had the highest consistency with intracellular fat content. Skeletal muscle FF reflects intracellular fat content in skeletal muscle cells. The more severe the degree of sarcopenia, the more significant the increase in intra- and extracellular fat content. The intra- and extracellular fat content is positively correlated with stroke time to a certain extent and negatively correlated with rehabilitation time, with the increase in extracellular lipid content being directly proportional to disease duration. These findings can be used clinically to evaluate SRS.

5. mDIXON characteristics of thigh skeletal muscles

The FF of mDIXON represents the average value of the ROI, including both intra- and extracellular fat. Using the mDIXON-Quant technique, this study compared the percentage of muscle fat in each muscle mass of the thigh skeletal muscles among the three groups: normal control, non-SRS, and SRS groups. Extracellular fat content gradually increased in each group, and the visual fat score gradually increased significantly. This study found that lipid changes in the thigh skeletal muscles of patients with sarcopenia after stroke were consistent with the overall lipid content changes in limb skeletal muscles after denervation [3,32-34]. Due to the lack of an inhibitory effect of the superior central nervous system on the subordinate central nervous system in patients who had a stroke, muscle denervation changes can occur, which can also lead to the accumulation of intra- and extracellular fat. As the thigh is an active muscle, the sudden absence of explosive movements after a stroke leads to a faster accumulation of intracellular fat than extracellular fat. Thus, the detection of intracellular fat content is of great importance for the diagnosis of skeletal muscle denervation injury and the severity of the lesion, where the accumulation of extracellular lipids suggests a correlation between motor deficits and prolonged lesion duration. After SRS, the increase in fat was mainly intracellular in the posterior muscle group and extracellular in the lateral muscle group.

6. Limitations

This study has some limitations, including the single-center study design, a small cohort of patients, enrollment of only Chinese patients, manually measured CSAs or ROIs, and no longitudinal follow-up investigation, which may affect the generalization of the outcomes. Future prospective, multicenter, randomized, controlled clinical trials with multiple races and ethnicities should be conducted to overcome all these issues for better outcomes.

In conclusion, the CSA of the thigh skeletal muscle in patients with SRS significantly decreased, and the interstitial FF significantly increased, especially in the latter muscle group, which was probably related to passive movement. The subcutaneous, interfascial, and subfascial fat contents of the SRS thigh muscle significantly increased, especially the intracellular fat content compared to the extracellular fat content, resulting in significant increases in the FF value of the skeletal muscle, which is consistent with changes in skeletal muscle denervation.

Article information

Conflicts of interest

No potential conflict of interest relevant to this article was reported.

Funding

None.

Author contributions

Conceptualization, Formal analysis, Funding acquisition, Investigation, and Validation: all authors; Data curation, Methodology, Software, and Supervision: RY, JH, WZ, YT, LY, YJ; Project administration: JH, LY; Resources: WZ, YT, LY, YJ; Visualization: RY, BLG, JH, WZ, YT, LY; Writing-original draft: RY; Writing-review & editing: all authors.

Fig. 1.

Comparison of muscle fat fraction in corresponding skeletal muscles among the three groups. SRS, stroke-related sarcopenia.

Fig. 2.

mDIXON-Quant sequence MRI images of the thigh in a patient with stroke-related sarcopenia, with increased fat fraction values on the paralyzed side (B) compared to the healthy side (A). MRI, magnetic resonance imaging. mDiIXON-Quant sequence: Philips Healthcare, Best, the Netherlands.

Fig. 3.

Intracellular fat content in the normal and stroke-related sarcopenia (SRS) group. (A) The normal control group shows no peak of intracellular fat content, lip 1. (B) In the SRS group, there is an increase in the intracellular fat content with a lip 1 peak (arrow) and a larger area under the curve. The intracellular fat content is significantly greater in the SRS group than in the normal control group.

Fig. 4.

Images of three groups of thigh skeletal muscles from a patient with stroke-related sarcopenia (SRS). (A) Healthy side. (B) Affected SRS side. The cross-sectional areas (CSAs) are all significantly decreased on the affected side (B) compared to the healthy side (A), with the greatest reduction in CSA in the posterior muscle group (43.9%, green), followed by the medial group (29.6%, red) and the anterior lateral group (22.5%, yellow).

Fig. 5.

Comparison of the cross-sectional area (CSA) of thigh muscles and fat between patients with stroke-related sarcopenia (SRS) and healthy controls. (A) CSA of thigh muscles and fat in a patient with SRS. (B) CSA of thigh muscles and fat in a healthy control. The thigh muscle CSA is significantly smaller in patients with SRS (4,656.8 cm²) than in healthy controls (8,530.5 cm²), but the thigh fat CSA is significantly greater in patients with SRS (3,189.6 cm²) than in healthy controls (1,963.2 cm²). In a patient with SRS, the thigh muscle CSA is significantly smaller on the SRS-affected side (right) than on the healthy side (left), but the thigh fat CSA is significantly greater on the SRS side than on the healthy side.

References

• 1. Chen LK, Woo J, Assantachai P, Auyeung TW, Chou MY, Iijima K, et al. Asian Working Group for Sarcopenia: 2019 consensus update on sarcopenia diagnosis and treatment. J Am Med Dir Assoc 2020;21:300–7.ArticlePubMed
• 2. Nozoe M, Kubo H, Yamamoto M, Ikeji R, Seike H, Majima K, et al. Muscle weakness is more strongly associated with functional outcomes in patients with stroke than sarcopenia or muscle wasting: an observational study. Aging Clin Exp Res 2024;36:4.ArticlePubMedPMCPDF
• 3. Li W, Yue T, Liu Y. New understanding of the pathogenesis and treatment of stroke-related sarcopenia. Biomed Pharmacother 2020;131:110721.ArticlePubMed
• 4. Wu WY, Dong JJ, Huang XC, Chen ZJ, Chen XL, Dong QT, et al. AWGS2019 vs EWGSOP2 for diagnosing sarcopenia to predict long-term prognosis in Chinese patients with gastric cancer after radical gastrectomy. World J Clin Cases 2021;9:4668–80.ArticlePubMedPMC
• 5. Yu F, He B, Chen L, Wang F, Zhu H, Dong Y, et al. Intermuscular fat content in young Chinese men with newly diagnosed type 2 diabetes: based on MR mDIXON-Quant quantitative technique. Front Endocrinol (Lausanne) 2021;12:536018.ArticlePubMedPMC
• 6. Pacicco T, Ratner S, Xi Y, Yokoo T, Fetzer D, Oz OK, et al. Pelvic muscle size and myosteatosis: relationship with age, gender, and obesity. Indian J Radiol Imaging 2019;29:155–62.ArticlePubMedPMC
• 7. Chiapparelli E, Okano I, Adl Amini D, Zhu J, Salzmann SN, Tan ET, et al. The association between lumbar paraspinal muscle functional cross-sectional area on MRI and regional volumetric bone mineral density measured by quantitative computed tomography. Osteoporos Int 2022;33:2537–45.ArticlePubMed
• 8. Erley J, Roedl K, Ozga AK, de Heer G, Schubert N, Breckow J, et al. Dual-energy CT muscle fat fraction as a new imaging biomarker of body composition and survival predictor in critically ill patients. Eur Radiol 2024;34:7408–18.ArticlePubMedPMCPDF
• 9. Molwitz I, Campbell GM, Yamamura J, Knopp T, Toedter K, Fischer R, et al. Fat quantification in dual-layer detector spectral computed tomography: experimental development and first in-patient validation. Invest Radiol 2022;57:463–9.ArticlePubMedPMC
• 10. Molwitz I, Leiderer M, McDonough R, Fischer R, Ozga AK, Ozden C, et al. Skeletal muscle fat quantification by dual-energy computed tomography in comparison with 3T MR imaging. Eur Radiol 2021;31:7529–39.ArticlePubMedPMCPDF
• 11. Zhu S, Tian A, Guo L, Xu H, Li X, Wang Z, et al. Investigation of diagnostic biomarkers for osteoporosis based on differentially expressed gene profile with QCT and mDixon-Quant techniques. Orthop Surg 2021;13:2137–44.ArticlePubMedPMC
• 12. Nanga RP, Elliott MA, Swain A, Wilson NE, Swago S, Witschey WR, et al. Identification of new resonances in downfield 1 H MRS of human calf muscle in vivo: potentially metabolite precursors for skeletal muscle NAD. Magn Reson Med 2023;90:1166–71.ArticlePubMedPMC
• 13. Vints WA, Kušleikienė S, Sheoran S, Valatkevičienė K, Gleiznienė R, Himmelreich U, et al. Body fat and components of sarcopenia relate to inflammation, brain volume, and neurometabolism in older adults. Neurobiol Aging 2023;127:1–11.ArticlePubMed
• 14. Yang C, Wang Z, Zhang J, Wang Y, Wang Z, Wang H, et al. MRI assessment of renal lipid deposition and abnormal oxygen metabolism of type 2 diabetes mellitus based on mDixon-Quant. J Magn Reson Imaging 2023;58:1408–17.ArticlePubMed
• 15. Jang Y, Im S, Han Y, Koo H, Sohn D, Park GY. Can initial sarcopenia affect poststroke rehabilitation outcome? J Clin Neurosci 2020;71:113–8.ArticlePubMed
• 16. Konishi M, Kagiyama N, Kamiya K, Saito H, Saito K, Ogasahara Y, et al. Impact of sarcopenia on prognosis in patients with heart failure with reduced and preserved ejection fraction. Eur J Prev Cardiol 2021;28:1022–9.ArticlePubMedPDF
• 17. Lee H, Lee IH, Heo J, Baik M, Park H, Lee HS, et al. Impact of sarcopenia on functional outcomes among patients with mild acute ischemic stroke and transient ischemic attack: a retrospective study. Front Neurol 2022;13:841945.ArticlePubMedPMC
• 18. Lee SH, Choi H, Kim KY, Lee HS, Jung JM. Appendicular skeletal muscle mass associated with sarcopenia as a predictor of poor functional outcomes in ischemic stroke. Clin Interv Aging 2023;18:1009–20.ArticlePubMedPMCPDF
• 19. Jia X, Wang Z, Huang F, Su C, Du W, Jiang H, et al. A comparison of the Mini-Mental State Examination (MMSE) with the Montreal Cognitive Assessment (MoCA) for mild cognitive impairment screening in Chinese middle-aged and older population: a cross-sectional study. BMC Psychiatry 2021;21:485.ArticlePubMedPMCPDF
• 20. Cheng KY, Chow SK, Hung VW, Wong CH, Wong RM, Tsang CS, et al. Diagnosis of sarcopenia by evaluating skeletal muscle mass by adjusted bioimpedance analysis validated with dual-energy X-ray absorptiometry. J Cachexia Sarcopenia Muscle 2021;12:2163–73.ArticlePubMedPMCPDF
• 21. Luengpradidgun L, Chamroonkul N, Sripongpun P, Kaewdech A, Tanutit P, Ina N, et al. Utility of handgrip strength (HGS) and bioelectrical impedance analysis (BIA) in the diagnosis of sarcopenia in cirrhotic patients. BMC Gastroenterol 2022;22:159.ArticlePubMedPMCPDF
• 22. Christensen EW, Drake AR, Lenchik L, Boutin RD. Sarcopenia diagnosis trends and opportunistic use of abdominal CT among medicare beneficiaries. J Am Coll Radiol 2025;22:821–30.ArticlePubMed
• 23. Grimm A, Meyer H, Nickel MD, Nittka M, Raithel E, Chaudry O, et al. Repeatability of Dixon magnetic resonance imaging and magnetic resonance spectroscopy for quantitative muscle fat assessments in the thigh. J Cachexia Sarcopenia Muscle 2018;9:1093–100.ArticlePubMedPMCPDF
• 24. Fan Z, Wang T, Wang Y, Zhou Z, Wu T, Liu D. Risk factors in patients with low back pain under 40 years old: quantitative analysis based on computed tomography and magnetic resonance imaging mDIXON-Quant. J Pain Res 2023;16:3417–31.ArticlePubMedPMCPDF
• 25. Ryan AS, Ivey FM, Serra MC, Hartstein J, Hafer-Macko CE. Sarcopenia and physical function in middle-aged and older stroke survivors. Arch Phys Med Rehabil 2017;98:495–9.ArticlePubMedPMC
• 26. Chang KV, Wu WT, Huang KC, Han DS. Segmental body composition transitions in stroke patients: trunks are different from extremities and strokes are as important as hemiparesis. Clin Nutr 2020;39:1968–73.ArticlePubMed
• 27. Endo S, Takahashi T, Sato M, Noya Y, Obana M. Effects of l-carnitine supplementation, botulinum neurotoxin injection, and rehabilitation for a chronic stroke patient. J Stroke Cerebrovasc Dis 2018;27:3342–4.ArticlePubMed
• 28. Verstraeten LM, Sacchi F, van Wijngaarden JP, Meskers CG, Maier AB. Sarcopenia, malnutrition and cognition affect physiotherapy frequency during geriatric rehabilitation: RESORT cohort. Ann Phys Rehabil Med 2023;66:101735.ArticlePubMed
• 29. Sato PH, Ferreira AA, Rosado EL. The prevalence and risk factors for sarcopenia in older adults and long-living older adults. Arch Gerontol Geriatr 2020;89:104089.ArticlePubMed
• 30. Chan HC, Fei X, Leung EL, Langston K, Marshall S, van der Meij BS. Post-discharge consequences of protein-energy malnutrition, sarcopenia, and frailty in older adults admitted to rehabilitation: a systematic review. Clin Nutr ESPEN 2023;54:382–97.ArticlePubMed
• 31. Kim WJ, Kim KJ, Song DG, Lee JS, Park KY, Lee JW, et al. Response to: “Sarcopenia and Back Muscle Degeneration as Risk Factors for Back Pain: A Comparative Study”. Asian Spine J 2020;14:583–5.ArticlePubMedPMC
• 32. Stokie JR, Abbott G, Howlett KF, Hamilton DL, Shaw CS. Intramuscular lipid utilization during exercise: a systematic review, meta-analysis, and meta-regression. J Appl Physiol (1985) 2023;134:581–92.ArticlePubMed
• 33. Vega G, Ricaurte G, Estrada-Castrillón M, Reyngoudt H, Cardona OM, Gallo-Villegas JA, et al. In vivo absolute quantification of carnosine in the vastus lateralis muscle with 1H MRS using a surface coil and water as internal reference. Skeletal Radiol 2023;52:157–65.ArticlePubMedPDF
• 34. Klepochová R, Leutner M, Bastian M, Krebs M, Weber M, Trattnig S, et al. Muscle-specific relation of acetylcarnitine and intramyocellular lipids to chronic hyperglycemia: a pilot 3-T 1H MRS Study. Obesity (Silver Spring) 2020;28:1405–11.ArticlePubMedPMCPDF

Figure & Data

References

Citations

Citations to this article as recorded by