Computed tomography (CT) is a noninvasive technique that is routinely used as a criterion method in clinical research studies for assessing in vivo skeletal muscle (SM) mass (13,16) and quality (SM lipid content) in the thigh (4) and visceral adiposity at the L4-L5 intervertebral space. Lee et al. (13) have recently demonstrated that abdominal muscle measured from a single image at L4-L5 is also related to whole-body muscle mass in men (R2 = 0.63, SEE = 8%) and women (R2 = 0.58, SEE = 10%). Similarly, Goodpaster et al. (4) demonstrate that the psoas (r = 0.65) and erector spinae (r = 0.77) muscle quality (lipid content) at the level of L4-L5 are correlated to those acquired at the midthigh in a heterogeneous group of individuals ranging from normal-weight glucose-tolerant to obese men and women with type 2 diabetes. This suggests that CT images at L4-L5, normally used to assess visceral adiposity, can also be used to assess SM mass and quality. However, it is unclear is whether changes in abdominal SM quantity and quality at L4-L5 are associated with corresponding changes in thigh SM quantity and quality, which is commonly used as the criterion method.
Whether measures of abdominal SM at locations other than L4-L5 are similarly related to those in the thigh is also unknown. CT is also routinely used to assess liver fat (14,18) by acquiring a single axial image at the level of T12-L1 (3). Recent findings also suggest that images acquired at T12-L1 provide better measures of total visceral fat mass than images acquired at L4-L5 (10). It is currently unknown whether measures at T12-L1 can also be used to assess SM quantity and quality, or changes therein. If CT scans normally acquired to determine visceral and/or liver fat could be simultaneously used to assess muscle composition, then it could present additional insight into the regional body fat distribution without added cost to the researcher or subjecting study participants to additional radiation exposure.
The purpose of this study was to determine the associations between muscle quantity and quality, measured in the abdomen and midthigh region. Furthermore, we examined the ability of abdominal images to reflect changes in thigh SM quantity and quality in response to a 6-month aerobic exercise intervention.
Participants for this analysis are a subsample (n = 125) of a large 6-month clinical exercise trial composed of overweight/obese postmenopausal women aged 45-75 yr, which was conducted at the Cooper Institute in Dallas, TX. Details of the intervention are published elsewhere (2). Briefly, the women were sedentary (not exercising more than 20 min on 3 d·wk−1 or more, and <8000 steps·d−1 assessed during the course of 1 wk), overweight or obese [body mass index (BMI) of 25.0 to 43.0 kg·m−2], and had a systolic blood pressure of 120.0 to 159.9 mm Hg. The women were randomly assigned to a control group or to one of three exercise groups that differed in energy expenditure. The women exercised three to four times per week on a cycle ergometer or treadmill at 50% of V˙O2max, expending 4, 8, or 12 kcal·kg body weight−1·wk−1. All exercise sessions were performed under observation and supervision in an exercise laboratory with complete and strict monitoring of the amount of exercise completed each session. All subjects gave their written informed consent before participation according to the ethical guidelines of The Cooper Institute Institutional Review Board, and the study was reviewed and approved annually.
CT measurement of muscle composition.
Axial images of the abdominal and thigh region were obtained using an electron beam CT (Imatron; General Electric, Milwaukee, WI) using a standard protocol (19). Subjects were examined in a supine position with their arms extended above their heads. Approximately 40 contiguous transverse images (6-mm thickness) were acquired in the abdomen spanning from the midregion of the iliac crest to the caudal region of the heart. Similarly, 50 contiguous images were acquired in the thigh region beginning just distal to the patella. Images were obtained using 130 kV and 630 mA with a 48-cm field of view and a 512 × 512 matrix. The CT data obtained in Dallas, TX, was transferred electronically to the laboratory in Kingston, Ontario, Canada, for analysis using a specially designed image analysis software (Tomovision, Inc, Montreal, Canada).
Thigh and abdominal SM was defined as the mean attenuation value of all pixels within the range of 0-100 Hounsfield units (HU) and −30 to 130 HU, respectively. Abdominal muscle quality (attenuation value) was assessed at the level of L4-L5 and T12-L1. Abdominal muscle included the rectus abdominis, obliques, multifidus, and erector spinae at both sites; the quadratus lumborum, iliacus, gluteal, and psoas muscles at L4-L5; and the intercostals, latissimus dorsi, and serratus anterior and posterior muscles at T12-L1. Thigh muscle quality was determined using the mean attenuation values for 11 contiguous images spanning from 12 to 18 cm proximal to the top of the patella and included the quadriceps, hamstrings, and thigh adductor muscles.
Descriptive variables for this subsample are presented as mean ± SD. All variables were normally distributed. Change scores were determined for only the cohort of 86 women that had both baseline and follow-up measurements. Associations between abdominal and thigh muscle quantity (mass or surface area), muscle quality (attenuation) at baseline (n = 125), follow-up (n = 86), and the change scores (n = 86) were determined with linear regression. P values of <0.05 were accepted to indicate statistical significance. All statistical analyses were conducted using SAS v9.
Subject characteristics are shown in Table 1. In general, the women were abdominally obese with a low cardiorespiratory fitness values. A wide range was observed in muscle quality and quantity changes in response to 6 months of aerobic exercise.
Cross-sectional associations between abdominal and thigh muscle quality.
Abdominal muscle quality at L4-L5 and T12-L1 was significantly (P < 0.01) associated with thigh muscle quality (HU) at baseline (L4-L5: R2 = 0.22, T12-L1: R2 = 0.37; Figs. 1A, B) and follow-up (L4-L5: R2 = 0.17, T12-L1: R2 = 0.43). The associations remained significant when thigh muscle quality points more than 70 HU were excluded from the analyses (L4-L5: R2 = 0.20, T12-L1: R2 = 0.31) and with control for age (baseline L4-L5: R2 = 0.21, T12-L1: R2 = 0.32; follow-up L4-L5: R2 = 0.16, T12-L1: R2 = 0.43). BMI, body weight, and waist circumference were not significant predictors of thigh muscle quality either at baseline or at follow-up (P > 0.10).
Cross-sectional associations between abdominal and thigh muscle quantity.
Abdominal muscle quantity at L4-L5 and T12-L1 was also significantly (P < 0.01) associated with thigh muscle quantity (mass) at baseline (L4-L5: R2 = 0.37, T12-L1: R2 = 0.48; Figs. 1C, D) and at follow-up (L4-L5: R2 = 0.28, T12-L1: R2 = 0.47). Results were similar with control for age (baseline L4-L5: R2 = 0.36, T12-L1: R2 = 0.41; follow-up L4-L5: R2 = 0.29, T12-L1: R2 = 0.38). BMI (R2 = 0.11 and R2 = 0.09), body weight (R2 = 0.17 and R2 = 0.14), and waist circumference (R2 = 0.05 and R2 = 0.08) were also significant (P < 0.05) predictors of thigh muscle mass at baseline and at follow-up, respectively. However, abdominal muscle measures tended to be a stronger predictor of thigh muscle mass than the anthropometric measures.
Associations between changes in abdominal and thigh muscle quality and quantity.
Changes in abdominal muscle quality at both sites were associated with changes in thigh muscle quality in response to 6 months of aerobic exercise (P < 0.01; Figs. 2A, B). However, changes in abdominal muscle at L4-L5 (Fig. 2C) were not significantly associated with changes in thigh muscle mass, whereas changes in abdominal muscle at T12-L1 (P = 0.01; Fig. 2D) were associated with changes in thigh muscle mass. Results were similar with control for age (quality L4-L5: R2 = 0.44, T12-L1: R2 = 0.73, P < 0.05; quantity T12-L1: R2 = 0.08, P = 01). BMI, body weight, and waist circumference were not significant predictors of changes in muscle quantity or quality (P > 0.10).
The novel observation of this study is that abdominal muscle quality can be used to assess thigh muscle quality and changes therein. However, abdominal muscle does not seem to be as useful for assessing muscle mass. These findings suggest that future investigations can use a single image in the abdomen for the simultaneous assessment of visceral fat, liver fat, and muscle quality without added cost or subjecting the participant to further radiation.
CT is a noninvasive technique that is considered one of the criterion measures for assessing SM mass (16). Lee et al. (13) have previously demonstrated that thigh muscle mass is a strong predictor of whole-body muscle mass (R2 = 90%), and although not reported, in that study, abdominal muscle had a similar association with total muscle mass compared with thigh muscle mass (both R2 = 0.58; not reported) (13). However, it remains to be determined whether our observations would remain the same had we measures of whole body muscle. Nevertheless, the results from this study indicate that, within the abdomen, measurements at L4-L5 may be a weaker measure of muscle quantity compared with measurements at T12-L1. This may be in part due to the technical difficulty of landmarking L4-L5 consistently in relation to the iliac crest. In many individuals, and especially in women, the top of the iliac crest lies at the approximate level of the L4-L5 intervertebral space. As such, slight deviations in landmarking or inconsistent placement of the subject in the CT scanner (i.e., slanted hips) cause large deviations in the muscles visible in a single image. Notwithstanding, these technical issues should not influence muscle measures at T12-L1 to the same degree, and yet the association with muscle quantity changes at T12-L1 and changes in thigh muscle quantity is still quite weak. However, this may be expected as muscle hypertrophy is muscle-specific, and as such, the greatest gains would occur in the exercising muscles.
Although CT-measured muscle lipid is not considered a criterion measure, many studies report a significant association between measures of muscle quality using CT compared with criterion measures of intramyocellular lipid (4,11) and metabolic risk (5,8,20). Further, CT is associated with better repeatability (coefficient of variation <2% vs 6-10%) compared with criterion measures (1,9,11), thus increasing the likelihood of detecting significant changes in muscle lipid using CT. Indeed, several studies have detected significant reductions in muscle lipid using CT in response to exercise and weight loss interventions (12,15). Although muscle quality is normally assessed in the thigh, it is important to note that there is no clear evidence to suggest that measures of thigh muscle quality or quantity provide any added advantage in predicting morbidity or mortality over abdominal muscle, regardless of how it was measured.
It is interesting to note that changes in thigh and abdominal muscle attenuation were more strongly associated than the corresponding baseline values. Lipid content is higher in Type I slow twitch muscle fibers (6), and although the percentage of Type I fibers can vary significantly between individuals and muscles within the thigh and abdomen (7,17,21), abdominal muscles are reported on average to have a greater percentage (50% vs 40%) of Type I fibers compared with the thigh (7,17). Nevertheless, the wide variations in muscle fiber composition (i.e., two- to threefold difference in % Type I fibers) between muscles even within a given individual (7) are reflected by the relatively weak association between thigh and abdominal muscle composition at baseline. In other words, individuals who have a high percentage of slow twitch muscle in the thigh may not necessarily have a high percentage of slow twitch muscle in the abdomen or vice versa. Conversely, longitudinal measures of the muscle fiber-type composition should remain constant assuming proper landmarking. As such, changes in thigh or abdominal muscle attenuation would likely be more a function of differences in whole-body muscle lipid accretion as opposed to differences in muscle fiber type.
That our study included a relatively homogeneous sample of predominantly younger obese white postmenopausal women that participated in an exercise intervention at the Cooper Institute limits the generalizability of the results but should not affect the internal validity. Future research is needed to verify the applicability of these results in other populations, such as men or frail elderly, as there may be differences in the SM mass distribution (13). Additional study is also needed to confirm these observations using other interventions, such as diet or other types of exercise perturbations. Further, whether abdominal muscle can be used to assess health risk, similar to thigh muscle is unclear and warrants investigation.
In conclusion, results from this study suggest that a single abdominal image at L4-L5 or T12-L1 can be used to assess muscle quality, in addition to measures of visceral fat and/or liver fat, but is not a good surrogate of muscle quantity. As such, at least two images (T12-L1 and midthigh) are required to assess muscle quality and quantity, in addition to visceral fat and liver fat with reasonable accuracy. These findings have important implications, and these will enable researchers to minimize the cost of data acquisition and limit subject radiation exposure while maintaining the ability to explore novel research questions.
This work was supported by the National Institutes Health grant nos. HL66262 and HL071900. The authors thank Life Fitness for providing exercise equipment. The results of the present study do not constitute endorsement by ACSM.
1. Boesch C, Kreis R. Observation of intramyocellular lipids by 1
H-magnetic resonance spectroscopy. Ann N Y Acad Sci
2. Church TS, Earnest CP, Skinner JS, Blair SN. Effects of different doses of physical activity on cardiorespiratory fitness among sedentary, overweight or obese postmenopausal women with elevated blood pressure: a randomized controlled trial. JAMA
3. Davidson LE, Kuk JL, Church TS, Ross R. Protocol for measurement of liver fat by computed tomography
. J Appl Physiol
4. Goodpaster BH, Kelley DE, Thaete FL, He J, Ross R. Skeletal muscle attenuation determined by computed tomography
is associated with skeletal muscle lipid content. J Appl Physiol
5. Goodpaster BH, Thaete FL, Simoneau JA, Kelley DE. Subcutaneous abdominal fat and thigh muscle composition predict insulin sensitivity independently of visceral fat. Diabetes
6. Gorski J, Oscai LB, Palmer WK. Hepatic lipid metabolism in exercise
and training. Med Sci Sports Exerc
7. Johnson MA, Polgar J, Weightman D, Appleton D. Data on the distribution of fibre types in thirty-six human muscles. An autopsy study. J Neurol Sci
8. Kelley DE, McKolanis TM, Hegazi RA, Kuller LH, Kalhan SC. Fatty liver in type 2 diabetes mellitus: relation to regional adiposity, fatty acids, and insulin resistance. Am J Physiol Endocrinol Metab
9. Kiens B, Essen-Gustavsson B, Christensen NJ, Saltin B. Skeletal muscle substrate utilization during submaximal exercise
in man: effect of endurance training. J Physiol
10. Kuk JL, Church TS, Blair SN, Ross R. Does measurement site for visceral and abdominal subcutaneous adipose tissue alter associations with the metabolic syndrome? Diabetes Care
11. Larson-Meyer DE, Smith SR, Heilbronn LK, et al. Muscle-associated triglyceride measured by computed tomography
and magnetic resonance spectroscopy. Obesity
12. Lee S, Kuk JL, Davidson LE, et al. Exercise
without weight loss is an effective strategy for obesity reduction in obese individuals with and without type 2 diabetes. J Appl Physiol
13. Lee SJ, Janssen I, Heymsfield SB, Ross R. Relation between whole-body and regional measures of human skeletal muscle. Am J Clin Nutr
14. Longo R, Ricci C, Masutti F, et al. Fatty infiltration of the liver. Quantification by 1
H localized magnetic resonance spectroscopy and comparison with computed tomography
. Invest Radiol
15. Mazzali G, Di Francesco V, Zoico E, et al. Interrelations between fat distribution, muscle lipid content, adipocytokines, and insulin resistance: effect of moderate weight loss in older women. Am J Clin Nutr
16. Mitsiopoulos N, Baumgartner RN, Heymsfield SB, Lyons W, Gallagher D, Ross R. Cadaver validation of skeletal muscle measurement by magnetic resonance imaging and computerized tomography. J Appl Physiol
17. Polgar J, Johnson MA, Weightman D, Appleton D. Data on fibre size in thirty-six human muscles. An autopsy study. J Neurol Sci
18. Ricci C, Longo R, Gioulis E, et al. Noninvasive in vivo
quantitative assessment of fat content in human liver. J Hepatol
19. Rich S, McLaughlin VV. Detection of subclinical cardiovascular disease: the emerging role of electron beam computed tomography
. Prev Med
20. Ryan AS, Nicklas BJ. Age-related changes in fat deposition in mid-thigh muscle in women: relationships with metabolic cardiovascular disease risk factors. Int J Obes Relat Metab Disord
21. Staron RS, Hagerman FC, Hikida RS, et al. Fiber type composition of the vastus lateralis muscle of young men and women. J Histochem Cytochem