Assessment of Muscle Quality and Quantity for the Diagnosis of Sarcopenia Using Mid-Thigh Computed Tomography : A Cohort Study


 BackgroundPreviously, sarcopenia was diagnosed based on walking speed, grip strength, and muscle mass. However, recent consensus has proposed a concern regarding the use of muscle mass measurement for sarcopenia diagnosis due to lack of accuracy; therefore, new assessment methods in clinical practice are required. We compared computed tomography (CT) with various motor function tests to determine the validity of CT as a potential diagnostic method for sarcopenia.MethodsIn total, 214 patients who were examined at our center during the study period (2016–2017) were included in the study. Single-slice CT scan of the mid-thigh region was performed, from which cross-sectional area (CSA) and CT attenuation value (CTV) of quadriceps femoris were evaluated for each subject. Other assessments included skeletal mass index, muscle strength and physical performance. Furthermore, subjects were classified into four groups: normal, low muscle mass (low mass), poor muscle function/strength (poor function), and sarcopenia.ResultsCSA and muscle strength exhibited the strongest correlation. Some physical performance tests showed the strongest correlation with CTV. CSA was significantly reduced both in men and women with low mass and sarcopenia group (p < 0.01), whereas CTV was significantly lower in men with sarcopenia group and in women with poor function and sarcopenia group (p < 0.01).ConclusionsCSA mostly reflects muscle strength, whereas CTV mostly reflects physical performance. CT with measurements of CSA and CTV enables the evaluation of muscle mass and quality simultaneously. Thus, CT is thought to be useful for the assessment of motor function and diagnosis of sarcopenia.


Background
As the adult population ages, the extension of healthy life expectancy has become a major issue. Sarcopenia is de ned as a reduction in the skeletal muscle mass with aging. The term "sarcopenia" was rst introduced by Rosemberg et al. [1] in 1989, from the Greek terms "sarx" (muscle) and "penia" (loss). In 1998, Baumgartner et al. [2] de ned sarcopenia as a mean skeletal mass index (SMI) of − 2SD or less determined using dual energy X-ray absorption (DXA) in young individuals, and since then, sarcopenia has been determined using SMI. However, it has been reported that compared with muscle mass, muscle strength is more associated with falling and limited mobility [3,4], and therefore, in determining sarcopenia, it is thought that muscle strength and physical performance should be included. The 2010 consensus report from the European Working Group on Sarcopenia in older people (EWGSOP) [5] proposed a diagnostic criteria for determining sarcopenia according to SMI, grip strength, and walking speed. In 2011, a condition "sarcopenia with limited mobility" was described in the United States of America in which muscle strength was lost despite maintaining muscle mass requiring treatment was proposed [6].
In 2014, the Asia Working Group of Sarcopenia (AWGS) 2014 presented diagnostic criteria for sarcopenia based on data obtained from various Asian countries [7]. Furthermore, In the EWGSOP2 revised in 2018, attention was focused on loss of muscle strength as the primary parameter of sarcopenia, and for a de nite diagnosis of sarcopenia, it was reported that the assessment of muscle mass and quality is important [8]. AWGS 2019 updated the consensus of sarcopenia diagnosis in Asia [9].
Muscle mass is conventionally measured using DXA and bioelectrical impedance analysis (BIA). DXA can measure appendicular skeletal muscle mass; however, there are problems with this method, i.e., the values differ depending on the machine used [10], and in Asian women, this measurement does not change with age [11]. BIA is affected by the amount of bodily uids; therefore, precise evaluations cannot be achieved [12]. Therefore, the EWGSOP2 described computed tomography (CT) and magnetic resonance imaging (MRI) as methods to verify the muscle mass and quality; however, there are only few reports of their use in clinical practice [13].
Motor function is evaluated using measurements of walking speed and grip strength; however, the reproducibility is poor depending on the physical and mental condition of the participant and the measurement methods and tool used [14]. CT and MRI are considered the gold standards of minimally invasive tests to measure muscle mass [15]; however, their use in the evaluation of muscle quality remains controversial. Goodpaster et al. [16] noted a good correlation between CT attenuation value (CTV) and intramuscular fat in ltration, whereas Ikemoto-Uezumi et al. [17] reported reduced muscle bers and fatty tissue in ltration in the muscular tissue of the vastus medialis in patients with osteoarthritis. Therefore, it is inferred that CT can evaluate muscle mass and that CTV might be able to evaluate motor function, or in other words, muscle quality.
However, there are few reports of the use of CTV in CT scans of the mid-thigh and their use in evaluating motor function. The purpose of this study was to compare CT with a variety of motor function tests to determine the validity of CT as a potential assessment method for muscle quality as well as for the diagnosis of sarcopenia.

Methods
Of 300 patients who were examined at the outpatient services of the National Center for Geriatrics and Gerontology, Center for Frailty and Locomotive Syndrome between 2016 and 2017, 214 (78 male and 136 female; average age, 78.4 years) were included in the study after excluding those with gait impairment caused by severe osteoarthritis of the knee and hip and those with missing data. To assess the risk of falls, we investigated fall scores, as proposed by Toba et al. [18] To assess motor function, we measured walking speed (Walkway MW-1000, ANIMA, Tokyo Japan); grip strength using a new grip dynamometer developed at our institution with ZP-500N (Imada Co., Ltd., Toyohashi, Aichi, Japan) [19]; knee extension strength (ZP-500N, Imada Co., Ltd., Toyohashi, Aichi, Japan) that is the measure of the isometric muscle extension strength in the sitting position with knee at 90° of exion, developed at our institution [20]; single-leg standing; timed up and go (TUG); short physical performance battery (SPPB) using stand-up test and two-step test. For single-leg standing, the participants were instructed to stand on a single leg with their eyes open as long as possible and the time was recorded. For TUG, we measured the time required to stand up from the chair and go back 3 m and to sit on the chair again. The stand-up and two-step tests were performed to assess the degree of locomotive syndrome. For the stand-up test, we measured whether the participants could stand up on both the legs from the sitting position on a platform of 10, 20, 30, and 40 cm [21]. For the two-step test, the participants stood at the starting line and took two strides as long as possible; then, the two legs were aligned and the distance of the two steps was measured. For SPPB, the test evaluated the balance, gait, strength, and endurance by examining an individual's ability to stand on their feet together in side-by-side semi-tandem and tandem positions, time to walk 8 ft, and time to rise from a chair and return to the seated position ve times [22]. Furthermore, the muscle mass of the four limbs from the whole-body scan was measured by DXA (Lunar iDXA, GE Healthcare, Chicago, Illinois, USA) and BIA (Inbody720, BIOSPACE, Iowa, USA). BIA provides an estimate of the muscle mass based on whole-body electrical conductivity [23].
For CT (SOMATOM Sensation 64; Siemens, Munich, Germany), the subject was placed in the supine position, and a single slice of the right midthigh. We de ned the mid-thigh as the midpoint of the inguinal ligament and the superior pole of the patella manually. The scanning conditions were as follows: 120 kV, 120 mA, rotation time of 1 s, and eld of view of 233 mm. From this single slice, the cross-sectional surface area (CSA) and CTV of the quadriceps femoris muscle were measured using SliceOmatic ver.5.0 (Tomovision, Canada) by a single examiner. CTV was de ned as − 1,000 HU for air and 0 HU for water and indicated the degree of X-ray absorption of a substance. The average CTV is 40-100 HU for a muscle and approximately − 50 to − 100 HU for fat. For CSA and CTV, the imaging data of 21 subjects were measured by three examiners, and the associated interclass correlation was determined. The interclass correlation of CSA and CTV of the examiners was > 0.998, and the interclass correlation among the three examiners was 0.999.
The subjects were classi ed in accordance with the AWGS 2014 diagnostic criteria [7] into four groups. Among the subjects without reduced walking speed and grip strength, the group with normal muscle mass was de ned as the "normal" group and that with reduced muscle mass as the "low muscle mass (low-mass)" group, that with reduced walking speed and/or grip strength but with normal muscle mass was de ned as the "poor muscle function/strength (poor function)" group and that with reduced muscle mass as the "sarcopenia" group. According to AWGS 2014 criteria, the cutoff values [7] were as follows: walking speed, 0.8 m/s; grip strength, 26 kg for men and 16 kg for women; SMI with DXA 7.0 kg/m 2 for men and 5.4 kg/m 2 for women; and SMI with BIA 7.0 kg/m 2 for men and 5.7 kg/m 2 for women. For the four groups, we examined the results of muscle function measurements and CT.
Statistical analyses were performed via SPSS (IBM SPSS Statistics ver.24, Tokyo, Japan) using chi-square test for fall scores, analysis of variance for CSA of the quadriceps femoris muscle, and CTV. Tukey's test was applied for post-hoc multiple testing, and normality test was not applied. Pearson's correlation coe cient was used for the muscle strength test and physical performance tests, CSA, and CTV. The deference of correlation coe cient among the subjects were calculated, and then Bonferroni correction was performed. P values of < 0.05 were considered statistically signi cant. Power analysis was performed and the sample size was su cient for the study.

Results
Overall, 58 subjects were included in the normal group, 30 in the poor-function group, 65 in the low-mass group, and 61 in the sarcopenia group. Compared with the normal group, age and the fall risk index were signi cantly higher in the poor-function and sarcopenia groups, height was lower in poor-function and sarcopenia group and body weight was signi cantly lower in the low-mass and sarcopenia groups than in the normal group (Table 1). Age and height were signi cantly higher in the poor-function and sarcopenia groups than in the normal group, whereas body weight was signi cantly lower in the low-mass and sarcopenia groups than in the normal group. (BMI : body mass index, SD : standard deviation) Knee extension strength and grip strength were correlated with CSA, BIA, DXA and CTV. CSA had signi cantly stronger correlation with knee extension and grip strength than DXA and CTV, whereas CTV did not have signi cantly stronger correlation with others ( Table 2). *or † means that CSA or CTV has signi cantly stronger correlation than the value. CSA showed the most association with muscle power, whereas CTV showed the most association with physical performance.
In contrast, for the physical performance measurements, the correlation coe cients for CSA and CTV showed that these were correlated with single-leg standing, walking speed, standing up test, TUG, SPPB, and two-step test. In some instances of BIA and DXA, CTV showed signi cantly stronger correlations, whereas CSA did not have signi cantly stronger correlation (Table 2).
Furthermore, when all groups (normal, low-mass, poor-function, and sarcopenia groups) were compared, CSAs of the quadriceps femoris muscle were signi cantly smaller in the low-mass and sarcopenia groups (both P < 0.001) for men and signi cantly smaller in the poor-function (P = 0.11), low-mass (P < 0.001), and sarcopenia (P < 0.001) groups for women (Fig. 1). Conversely, CTVs were signi cantly smaller in the sarcopenia group (P = 0.044) for men and signi cantly lower in the poor-function (P = 0.007) and sarcopenia (P < 0.001) groups for women ( Fig. 1).

Discussion
In this study, we investigated which inspection methods were superior for the assessment of motor function including muscle strength and physical performance, in other words, the muscle quality. The result indicated that CSA showed the most correlation with muscle strength, whereas CTV showed the most correlation with physical performance.
The concept of sarcopenia has considerably changed over time, and recently, the EWGSOP2 has stated that the importance of measuring muscle quality is expected to grow as a de ning feature of sarcopenia; however, there is no universal consensus on the assessment method for routine clinical practice [8]. Patients with dynapenia and sarcopenia are considered to be at a high risk of falling [3]. Further, in the present study, patients with poor function and sarcopenia had a signi cantly high risk of falling, which we believe re ected their reduced motor function.
There is a physical state in which muscle mass is maintained but with reduced functioning and muscle quality. McGregor et al. [24] referred that not only changes in muscle mass but also other factors underpinning muscle quality including composition, metabolism, aerobic capacity, insulin resistance, fat in ltration, brosis, and neural activation may play a role in the decline in muscle function and impaired mobility associated with aging.
To date, muscle quality is determined by evaluating intramuscular fat in ltration using highly sensitive measurement devices such as CT and MRI [25]. Moreover, in the recent years, muscle quality has been evaluated using ultrasound; however, there is no consensus for any of these assessments. There are few reports wherein CT of the mid-thigh has been used to evaluate the relationship of CSA and CTV using motor function. The present study was performed to evaluate the muscle quality by CTV and its relationship with motor function.
In the evaluation of simple muscle strength such as grip and quadriceps femoris muscle strength, we found the strongest correlation with CSA in all subjects. Furthermore, in the evaluation of physical performance test that combined physical movement (single-leg standing, walking speed, stand-up test, TUG, SPPB, and two-step test), the highest correlation was observed with CTV, and some test showed signi cant stronger correlation with CTV suggesting that physical performance could be evaluated using CTV.
Lang et al. [26] reported that a low CTV of the thigh indicated a higher risk of fall and hip fracture; the results of the present study are consistent with these results. In normal and low-mass group, there was no decrease in CTV, whereas the poor-function and sarcopenia groups showed a signi cant decline in CTV. Therefore, we believe that this indicated that reduced motor function was caused by the reduced muscle quality in patients with poor function and sarcopenia. In this study, we classi ed patients into four groups according to muscle mass and motor function; this classi cation method including an evaluation of muscle quality may be useful in assessing functional status in the aging population.
Ikemoto-Uezumi et al. [17] compared tissue in the vastus medialis muscle of patients with osteoarthritis and reported that intramuscular adipose tissue (IMAT) and an increased proportion of interstitial tissue were observed. Thus, a decline in CTV with an increased proportion of IMAT was associated with intramuscular fat in ltration [27]. Similarly, in sarcopenia, increased IMAT is observed [16], and owing to changes in muscle tissue, CTV is decreased. Reportedly, increased IMAT measured by MRI is a prognostic factor of gait ability [28], and IMAT is increased in young adults with reduced physical activity [29]. Therefore, reduced physical performance and activity could lead to increased IMAT.
As a minimally invasive tool, CT is considered as a gold standard to measure muscle mass; however, it is not used in general practice owing to the high costs, poor portability, and requirement of an experienced operator [15]. In patients with cancer, reduced muscle mass is an independent predictor of immobility and mortality [30]. Kasai et al. [31] reported that among the muscles in the mid-thigh CT, CSA decreased with age mainly in quadriceps femoris. Schweitzer et al. [32] reported that on MRI, the best estimates for skeletal muscle were shown in the thigh. CT of the femur was a useful method to evaluate muscle mass for the entire body [33].
Compared with other test equipment, DXA showed no change with age among Asian women [11]. Recently, we reported signi cant associations of thigh CT CSA with muscle strength [34].
For BIA, changes associated with the physical condition such as uid balance and body temperature uctuated during the day, resulting in reduced accuracy of BIA [35]. CT could be a better form of assessment because of its accuracy, reproducibility, and objectivity; it also enables the simultaneous measurement of the muscle mass and quality [36]. CT is useful in diagnosing the severity of sarcopenia by simultaneously evaluating CSA and CTV. Regarding the amount of radiation exposure, as only a single slice is scanned, we consider that there is no problem in terms of the radiation levels [10]. In addition single slice leg CT scanning takes only about ve minutes.
There were limitations to the present study. First, the sample size was relatively small. Second, this was a cross-sectional study and changes caused by natural progress or therapeutic intervention could not be evaluated. Further study is required to determine whether or not the changes occurred as a result of exercise therapy, nutritional guidance, and pharmacotherapy. Third, this study did not include large-scale data of CSA and CTV from each age group of the general population without sarcopenia, and therefore, the healthy control information was weak. Thus, analysis using a large-scale epidemiological study is required.
As noted in the EWGSOP2, our results showed that CT may be used to diagnose sarcopenia more accurately and understand the pathology in greater detail [8].

Conclusions
This study was the rst to report a relationship between mid-thigh CT and motor function in clinical setting. CT allows CSA and CTV to be measured simultaneously; CSA showed the most association with muscle strength, whereas CTV showed the most association with physical performance. Thereby enabling the precise evaluation of muscle mass and muscle quality, which is useful for diagnosing and considering severity of sarcopenia, low mass, and poor function. These results could be useful for the con rmation of the diagnosis of sarcopenia. The study was conducted with the approval of the Ethical Review Board of the National Center for Geriatrics and Gerontology, Japan and with the written consent of all the participants.

Consent for publication
Not applicable Availability of data and materials The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests Funding This work was supported by a grant from Expenses for Research and Development of Geriatrics and Gerontology (26-12, 29-12) and Furuno Electronics; the funding agencies played no role in the study.
Authors' contributions HO analyzed and interpreted the participants date. YM suggested this study's concept and design. HA supervised this study. YS analyzed midthigh CT date and calculated CSA and CTV. SY supervised about statistical analysis. TW, HI, TM, SI, HH and SI advised study design and results.
All authors read and approved the nal manuscript.