Anthropometric and ultrasound methods quantify anatomic skeletal muscle and are thus influenced by all muscle components. Observed temporal changes can reflect, for example, alterations in intramuscular adipose tissue at the tissue system level or in water and glycogen at the molecular level. A change in anthropometrically- or ultrasound-derived muscle mass may therefore not always reflect parallel changes in the main component of interest, muscle proteins.
A second concern related to anthropometry and ultrasound is measurement error, which is relatively large for both methods (e.g., within- and between-observer technical errors in the range of ˜5-8%) (Nelson et al., 1996). Both anthropometric and ultrasonic measurements are useful mainly in evaluating large skeletal muscle mass changes over long time periods. The cost of these methods is relatively low and the measurements can be made in field settings.
Urinary metabolites are markers of muscle cell mass, but as already mentioned, there are non muscle sources of both creatinine and 3-methylhistidine. These methods require careful urine collection protocols and subject compliance on a meat-free diet. Even under well-controlled conditions, the between-day coefficient of variation is relatively large (i.e., 4-8%). These methods may therefore be useful in evaluating large muscle mass changes over time or changes in groups of subjects. As creatinine is easy to measure, this method may be applicable in settings without sophisticated analytical equipment.
Bioelectric impedance analysis requires strict attention to measurement protocol (Nelson et al., 1996). Under carefully controlled conditions, the between-measurement technical error of BIA is very small (˜1-3%). Measured impedance may, however, reflect changes in electrolyte concentrations, fluid distribution, and adiposity that occur in patients followed over time. Because BIA methods are inexpensive and simple, they have value in field settings. More studies of the ability of BIA methods to quantify small changes in skeletal muscle mass are needed.
Imaging methods have the most potential for quantifying small skeletal muscle mass and composition changes. At present, the CT method is well suited to measure adipose tissue-free muscle (Sjöström et al., 1991; Heymsfield et al., 1997b), and the same applies to MRI although lingering concerns remain. Specifically, validated procedures for estimating adipose tissue content of muscle by MRI have not yet been published. Both of these imaging methods will record a change in muscle mass related to any one or more components including protein, water, glycogen, and various fluid spaces. With this proviso, the between-measurement technical errors for skeletal muscle areas and volumes by CT and MRI are small (< 2%) (Sjöström et al., 1991; Heymsfield et al., 1997b; Wang et al., 1996a; Ross et al., 1995), particularly if repeated measurements are made by the same observer. Hence, the likelihood of detecting small muscle volume changes by CT and MRI is good. The disadvantage of these methods is their relatively high cost and limited access.