Assessment of body composition typically refers to the quantification of body fat and muscle mass, and it is most commonly assessed by medical imaging. In the past few decades, body composition imaging has mostly focused on evaluation of body fat, such as visceral fat or subcutaneous fat, because obesity and its clinical implications has been extensively investigated (12). However, in the past decade, the importance of muscle mass has been emphasized and has become a focal point for clinical research (3). Indeed, in 2016, sarcopenia itself was classified by the International Classification of Diseases (ICD-10-CM), with the code M62.84.

Sarcopenia, a term first introduced in 1984 by Rosenberg, refers to age-related loss of muscle mass, and is thus a type of geriatric syndrome (4). Many studies have proven its relationship with physical impairment and poorer quality of life, and increased morbidity, mortality, and health care costs. Recently, the concept of sarcopenia has been extended to various diseases, beyond merely being considered as a geriatric syndrome (5). In particular, the influence of sarcopenia on morbidity/mortality in cancer patients treated with chemotherapy or major surgery has been extensively investigated.

Currently, there are diverse methods for evaluating muscle mass by imaging, and these methods must be standardized. Such standardization requires a massive accumulation of evidence and efforts to achieve a worldwide consensus. Evidence and consensus are only beginning to mount at present, indicating that the imaging modalities and methods used to diagnose sarcopenia will remain varied for a while to come. The role of imaging techniques has rapidly increased in the field of sarcopenia. Accordingly, radiologists should be aware of sarcopenia and its clinical implications.

In this paper, we aim to present a comprehensive review of the role of the different imaging modalities currently available for the non-invasive assessment of skeletal muscle. We also discuss several controversial issues in sarcopenia and their clinical relevance.

Sarcopenia, derived from the Greek words for flesh (sarx) and loss (penia), is a condition of decreased skeletal muscle mass that can lead to a decline in physical ability (4). There is a lack of worldwide agreement on the definition of sarcopenia. There are three major consensus groups (4). Currently, the most commonly accepted and used definition is from the European Working Group on sarcopenia in older people (EWGSOP, the Sarcopenia Working Group), created in 2010. The EWGSOP recommends using the presence of both low muscle mass and low muscle function (strength or performance) for the diagnosis of sarcopenia (4). Regarding the severity of sarcopenia, EWGSOP has suggested a staging concept that can be helpful in clinical management: “presarcopenia,” “sarcopenia,” and “severe sarcopenia” (5).

Sarcopenia is often confused with the terms frailty and cachexia. These conditions show marked overlap because sarcopenia (i.e., decreased muscle mass and muscle function) is a major component of frailty and cachexia. Frailty is another geriatric syndrome characterized by reduced homeostatic reserves (resilience) that carries an increased risk of poor health outcomes, including falls, incident disability, hospitalization, and mortality (6). Fried et al. have defined frailty as meeting three of five of the following criteria: low grip strength, low energy, slowed walking speed, low physical activity, and unintentional weight loss (7). Frailty is not exclusively related to muscle itself, but is more complex and is often preceded by sarcopenia (78). Cachexia is a progressive and involuntary loss of skeletal muscle mass and fat mass due to severe disease, such as cancer, chronic obstructive pulmonary disease, chronic heart failure, and chronic kidney disease. Cachexia is a form of secondary sarcopenia caused by severe disease, but it is characterized by both weight loss and malnutrition, due to systemic metabolic imbalances and inflammation, with prominent overexpression of proinflammatory cytokines, which can cause proteolysis and lipolysis (9).

Sarcopenia has a high prevalence (29–33%) in community-dwelling populations and long-term care populations (10). In general, aging leads to changes in muscle fibers and α-motor neurons. Decreased production of growth hormone and sex-steroid hormones with aging also can aggravate muscle changes. Sarcopenia can be primary (resulting from aging) or secondary (resulting from early life developmental influences, malnutrition, bed rest or little physical activity, and certain treatments), although in most cases, as sarcopenia is a multi-faceted syndrome, it may not be possible to categorize it as primary or secondary (11).

The concept of muscle function is being revolutionized. In the past, skeletal muscle was considered an organ related to mobility, with a mechanical function. It is now known that muscle has various functions, such as metabolic, endocrine, neurological, and even psychological functions (Fig. 1). The concept of “muscle as a secretory organ” has been proposed, as studies have shown the multi-systemic effects of muscle via control of myokine release (12).

Myokines are cytokines and other peptides, such as IL-6 and brain-derived neurotrophic factor (BDNF), which perform actions by communicating with other organs (12). For example, IL-6 promotes glucose production in the liver, lipolysis in adipose tissue, increased pancreatic beta-cell viability, and insulin secretion, suggesting that it has an important role in maintaining glucose homeostasis. BDNF plays a crucial role in regulating the growth and maintenance of neurons; therefore, it influences learning, memory, and even mood (13).

Many studies have reported that myokine production is related to muscle contraction. Accordingly, physical inactivity may cause imbalances between these substances, which leads to a pro-inflammatory status and metabolic dysfunction that promotes the critical cascade of sarcopenia and fat accumulation (14). In connection with this new pathophysiological concept, the impact of sarcopenia on chronic disease is also being investigated. Skeletal muscle is also strongly related to energy metabolism, since it is the largest organ that consumes glucose as an energy source. Thus, sarcopenia is related to metabolic syndromes and various cardiovascular diseases (151617). In addition to the metabolic effects, sarcopenia can lead to physical disability, falling, fracture, hospitalization, depression, poor quality of life, increased health care costs, and notably increased mortality (12). Currently, the evidence of the impact of sarcopenia in various disease settings and on health status is rapidly accumulating (Fig. 2). First, the impact of sarcopenia on the outcomes of various treatments, such as major surgery and cancer chemotherapy, has been extensively investigated. Systematic reviews of patients operated for gastrointestinal and hepato-pancreato-biliary malignancies have reported impaired overall survival and increased postoperative morbidity in patients with sarcopenia (1819). Sarcopenia is also associated with increased mortality in patients who were treated with liver transplantation (20), selective and non-abdominal aortic aneurysm repair (2122), and pancreatic adenocarcinoma treatment (23). In addition, in the setting of an intensive care unit and emergency care, sarcopenia has been reported to be a significant predictor of mortality (32425).

One of the critical issues in sarcopenia is the discrepancy between muscle mass and muscle function, which might be mainly attributed to the fatty degeneration or fatty infiltration of the muscle; i.e., myosteatosis. Myosteatosis is regarded as a part of the aging process or muscle disuse that is independent of obesity (26). Myosteatosis has been proven to be associated with decreased muscle strength and mobility, and increased disability (272829).

Recent studies have found evidence indicating that disuse leads to fat infiltration, which influences muscle fiber transition from type II to type I, thus resulting in impaired contractile capacity with decreased muscle power (30). During this process, metabolic pathways, such as leptin signaling, inflammatory adipokine, sex steroid, and glucocorticoid signaling, and glucose metabolic pathways, are involved. Thus, myosteatosis is not merely a problem of muscle, but is closely related to metabolic and systemic dysfunction (26). Several studies have reported that accumulation of muscle fat decreases insulin sensitivity, leading to type 2 diabetes, impairing the capacity for normal protein synthesis in skeletal muscle, leading to loss of muscle mass and strength, which results in increased disability and mortality among older adults (313233).

There are many methods to evaluate muscle mass, including anthropometry, bioelectrical impedance analysis (BIA), and medical imaging (Table 1). Anthropometry assessments, such as body-mass index (BMI), skin-fold thickness, and body circumference (e.g., waist, thigh, and calf), are simple and readily available in any clinical setting, but may not be sufficiently accurate to evaluate muscle mass. BIA, which calculates the muscle mass and fat mass based on different conduction of current through tissues, is low cost and simple to use. However, measurement errors may be present depending on differences in hydration level, exercise status, or even food intake (3). In a study aiming to validate the accuracy of BIA in body composition prediction, dual-energy X-ray absorptiometry (DXA) and MRI showed that BIA devices tend to overestimate at lower levels and underestimate at higher levels, while being relatively accurate at predicting fat mass and skeletal muscle mass percentages.

Medical imaging is considered an accurate and reliable method for quantifying muscle and fat mass and it discerns the spatial distribution of fat and muscle, enabling diagnosis and grading of sarcopenia, obesity, or both (i.e., sarcopenic obesity). Medical imaging is useful for tracking longitudinal changes in muscle and fat; thus, it can be used in clinical trials as a treatment assessment tool (23). The common imaging modalities for evaluating muscle mass include whole-body DXA, CT, and MRI. Use of the ultrasonography (US) is limited in some clinical settings (3).

Currently, there are many issues regarding the utilization of medical imaging for body composition assessment. There is no consensus on which imaging modality to choose, which body region to measure, which guideline for image acquisition to follow, or what cutoff points should be used to evaluate skeletal muscle quality and quantity. Standardization of the diagnostic index and its diagnostic cutoffs, based on sufficient evidence and international consensus, remains necessary. One important consideration is ethnicity. The cutoff points appear to differ between Asian and Caucasian populations, due to differences in ethnicity, adiposity, size, and lifestyle, which may also affect the prevalence of sarcopenia.

In CT and MRI, there is no standardized protocol for image acquisition of body fat and muscle mass quantification. In particular, the measurement field (i.e., whole body scanning versus single-slice selection) and measurement area/level (i.e., which area or level should be used when using a single-slice selection for sarcopenia assessment) have not been standardized to date. Although the assessment of sarcopenia based on whole body imaging is the most accurate, it is very time consuming and costly; thus, it may not be practical in most clinical settings. Therefore, identifying which anatomical level or muscles best represents the total lean body mass is a very important issue. To date, there have been three main proposals: TAMA at the lumbar spine level; psoas muscle area in the lumbar spine level; and thigh muscles at the mid-thigh level, as illustrated in Figure 5 (35).

Currently, the most frequently used landmark among sectional body composition studies is the L3 level of the lumbar vertebra, used for the measurement of TAMA. At this level, the field of view includes the major large muscles and main functional muscles of the human body, which are the psoas, paraspinal muscles (erector spinae, quadratus lumborum) and abdominal wall muscles (transversus abdominus, external and internal obliques, and rectus abdominus), making it the optimal level for skeletal muscle analysis. In several studies, a single scan at the level of L3 was the best compromise site for assessing the total tissue volumes of skeletal muscle, visceral adipose tissue, and subcutaneous adipose tissue (36515253). Schweitzer et al. (5154), investigated level L1 through L5 to determine the best estimates of skeletal muscle and visceral fat using single slice image of MRI, and confirmed that L3 level showed highest correlation with whole-body skeletal muscle and visceral fat volume.

However, many studies have also used the L4 level and the mid-thigh level. A single-slice MR image at the mid femur level showed good estimation of skeletal muscle and fat volume in the thighs and correlations with clinical criteria for sarcopenia in older adults (55). Availability of multiple protocols for image acquisition limits the standardization of muscle mass and quality evaluation; thus, further large-scale studies and standardization of the measurement field/level are warranted.

In most recent studies that used CT/MRI for sarcopenia assessment, clinically acquired CT/MRI were used by applying additional imaging segmentation techniques to measure the muscle and fat mass. The imaging segmentation techniques vary according to the level of automation, such as a manual or semi-automatic, based on a threshold level, or an artificial intelligence-driven fully-automated technique (47). Recently, the AsanJ-Morphometry™ software (Asan Image Metrics, Seoul, Korea) was developed for abdominal muscle and fat area measurements based on ImageJ (NIH, Bethesda, MD, USA) and was made publicly available for non-profit research (available at http://datasharing.aim-aicro.com/morphometry).

Recently, the use of new CT techniques, such as quantitative computed tomography (QCT), have been being investigated. Peripheral QCT (pQCT) is a novel imaging modality that has been primarily used to investigate bone mineral content. Like standard CT scans, pQCT produces a cross-sectional image that enables quantification of 3-dimensional tissue structure and skeletal muscle evaluation. But unlike general CT, pQCT uses a smaller scanner to study peripheral tissue. In the past, pQCT use was limited to the peripheral limb, usually the tibia and mid forearm (47). However, with technological advances, it became possible to analyze larger areas, up to the mid-thigh level. pQCT has the important advantages of extremely low radiation exposure, shorter scan time, and relatively low cost compared to whole body CT (3747).

In one study, Xu et al. (56) compared QCT and biochemical measurement of geese hepatic steatosis. QCT measurements of goose liver fat were shown to be accurate and reliable, and QCT was suggested to have potential in clinical settings (56).

One limitation of pQCT is the lack of a standardize analysis protocol. Since determination of the scanning region is performed by calculating the percentage of limb length from a reference line, comparisons between studies are difficult (47). Furthermore, pQCT scanning is limited to peripheral anatomical sites, which have lower accuracy compared to the sites used for MRI, making it difficult to distinguish individual muscles (4347). With these limitations, current pQCT use may be limited to situations where portability is necessary (55).

Although US has limitations for use in evaluating skeletal muscle, it can be a good option in some clinical settings for an initial evaluation of the quality and quantity of skeletal muscle mass (Fig. 6). Its major advantages as compared to other modalities are its low cost, portability, and lack of radiation exposure. In particular, its portability is particularly advantageous: unlike other modalities, whose lack of portability limit their use in large epidemiological studies, the portability of US yields a marked advantage in clinical settings, which explains its growing importance in studying skeletal muscle (5758). As no radiation is required, US can be used for all patients, including children and pregnant women. Another important advantage of US is that it allows real-time visualization of the target structure, and through echogenicity, it can provide information about the presence of inflammation, fibrosis, and adipose infiltration (57).

The main limitation of US is its lack of a standardized protocol and examiner-dependent factors, which can lead to evaluation errors and thus interfere with the reproducibility of results. Additionally, the numerous devices available commercially show great technological diversity (larger than that commonly found with DXA) and employ different methods for calibration, thus making it difficult to compare results from different studies directly (37).

Body composition imaging has become increasingly common as the importance of sarcopenia has been recognized in clinical practice and research. Currently, DXA and CT/MRI are most commonly used imaging modalities for sarcopenia assessment. In general, DXA is utilized in the clinical setting of primary care for health surveillance/screening, geriatric medicine, rehabilitation medicine, and endocrinology/metabolic disease. In contrast, CT/MRI is used when CT/MRI is clinically acquired for disease diagnosis/treatment/follow-up, particularly in cancer patients or patients treated with major surgery. Because of their rapid and major technological advances, many tools are now available to diagnose and evaluate sarcopenia. However, there is a lack of consensus and standardization in the definition of sarcopenia, imaging modalities, measurement methods, and diagnostic cutoff values. Radiologists should be aware of the current updates on body composition imaging and the contentious issues, and should guide clinicians to utilize the imaging tool best suited for muscle mass assessment and patient management.