![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
3 College of Public Health, and Departments of 4 Physiology, 5 Radiology, 6 Family and Community Medicine, and 7 Nutritional Science, University of Arizona, Tucson, AZ 85724; 8 Columbia University, St. Luke's-Roosevelt Hospital, New York, NY 10025; and 9 University of Washington, Fred Hutchinson Cancer Research Center, Seattle, WA 98109
* To whom correspondence should be addressed. E-mail: zchen{at}u.arizona.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT IntroductionMethodsResultsDiscussionLITERATURE CITED |
|---|
|
Introduction |
|---|
|
TOPABSTRACTIntroductionMethodsResultsDiscussionLITERATURE CITED |
|---|
Computed tomography (CT) and MRI can measure skeletal muscle and adipose tissue volumes. Cadaver validation studies have shown excellent accuracy of CT and MRI in measuring SMM (r = 0.99) (1). Nevertheless, these methods are impractical in clinical settings and for large epidemiologic studies because of the high cost (CT and MRI) and high radiation exposure (CT). Dual-energy X-ray absorptiometry (DXA) has been widely employed in clinical practice for osteoporosis screening and diagnosis. Body composition assessments by DXA are readily available, less expensive, and less invasive compared with MRI and CT. Previous studies have demonstrated good correlations between DXA-derived lean soft tissue mass (LSTM) and SMM for the lower limb region when CT (2) and MRI (3) were used as the criterion. A high correlation between DXA-derived appendicular LSTM and MRI-derived total body SMM was also reported for younger men and women (4). Because changes in the components of body composition in older populations may alter the association between SMM and LSTM, the utility of DXA in assessing total body and regional SMM in the elderly needs further investigation.
With the aim of investigating the utility of DXA for assessing whole body and regional SMM in U.S. elderly women, the relationship between DXA-derived LSTM and MRI-derived SMM was examined in a cross-sectional sample of postmenopausal women. Factors that may affect the association between DXA and MRI measurements were examined and predictive models for assessing SMM from DXA measurements were developed.
|
Methods |
|---|
|
TOPABSTRACTIntroductionMethodsResultsDiscussionLITERATURE CITED |
|---|
We recruited women in this project from participants in the U.S. nationwide Women's Health Initiative (WHI) study who resided in southern Arizona. The WHI study design and recruitment method has been reported previously (5,6). The WHI participants were postmenopausal women aged 50–79 y at the time of enrollment.
Between the spring of 2004 and the spring of 2005, WHI staff in Tucson sent a letter of recruitment for this study to local WHI participants who were recruited at the WHI Tucson Clinic and were scheduled to have their regular WHI visit for DXA scans within 3 mo. To assess the eligibility of the potential participants, a questionnaire was administrated via phone. The exclusion criteria were: 1) weighed >113.6 kg (250 pounds, due to weight restrictions by the DXA machine); 2) inability to undergo an MRI scan due to metal implants, extreme claustrophobia, or recent surgery; 3) unable to lie supine for 30 min; and 4) unable to raise her arms over her head for 15 min.
After the initial screening, eligible women were scheduled for a whole body MRI scan at the University of Arizona Medical Center within 6 wk of the time after the DXA scan. Additional procedures, including phlebotomy and anthropometric measurements, were also conducted. The Institutional Review Board at the University of Arizona approved the study.
DXA measurements
The Hologic QDR 4500w with fan beam features (Hologic) and array mode was used for all DXA scans. The scan measurements and analyses were conducted following standard procedures. Participants were measured wearing only gowns to eliminate possible artifacts due to clothing and fasteners. The WHI Bone Density Center at the University of California, San Francisco (Prevention Sciences, San Francisco, CA) was responsible for quality control, including monitoring DXA operator performance, DXA machine performance, and managing DXA databases from the WHI clinic centers. Review of the scan positioning and analysis by the Bone Density Center at the University of California, San Francisco was done in random samples throughout the WHI study for quality control purposes. Machine performance was evaluated by scanning hip, spine, and total body phantoms.
Whole body scans were manually analyzed for the manufacturer-defined regions of interest (ROI) following the standard analysis protocol in the Hologic User Manual. Customized ROI, including total leg, upper leg, and lower leg, were also analyzed using the Hologic whole body and subregion analysis modes (QDR System software ver. 12.1). For the customized ROI, the total leg region was defined by placing a horizontal line at the lowest point of the ischial tuberosity as the upper margin and a horizontal line past the tips of the toes as the lower margin of the leg. The upper leg was defined as the region between the ischial tuberosity and the knee joint. The lower leg was defined as that total leg minus upper leg. The same regions were selected on MRI for comparison.
Fat, LSTM, and percent fat (%fat) from total body (without the head region) and selected ROI were assessed in this study. DXA-derived %fat was calculated from fat mass divided by body mass.
MRI measurements
Obtaining images. A whole body scan was conducted using the 3-T MRI scanner (model GE, General Electric) platform at the University of Arizona Medical Center. Participants were placed on the scanner with arms extended above the head and were scanned for the lower body first and then the upper body. To obtain lower body imaging, a sagittal image was obtained to locate the ischial tuberosity, which was used as the point of origin. The imaging then proceeded from the ischial tuberosity to tips of the toes. The image thickness was set at 10 mm and the distance between images was 40 mm.
Upper body imaging was obtained after the lower body measurement was completed. Participants were repositioned with arms extended above the head. One scout view for the upper body was performed and scanning then proceeded from the ischial tuberosity to the fingertips.
MRI image analysis. The MRI images were archived to disks and sent electronically to the Obesity Research Center at St. Luke's-Roosevelt Hospital, Columbia University for SMM analysis. A multiple-step procedure was used to segment images into skeletal muscle, adipose tissue, and other tissue areas. The analysis was conducted using SliceOmatic image analysis software (TomoVision). This program allows analysis of the images through a "painting" interface whereby readers can apply a specified color assigned to a specific tissue compartment. After the analysis by painting, we created a data file yielding the results of body composition (i.e. skeletal muscle, adipose tissue, etc.) by color in area per slice. The CV for repeated skeletal muscle measurements of the same scan by the same observer at the Obesity Research Center's laboratory was 0.7%.
A special protocol was implemented to enable anatomical segmentation of the data into regional measurements. We calculated SMM using the following equation: SMM = 0.00104 x 
[A x (B1 + B2)/2], where SMM is in kg, A is the distance (cm) between slices, B1 and B2 are the skeletal muscle areas (cm2) in adjacent slices, and 0.00104 is the assumed constant density of skeletal muscle in kg/cm3(7). Validation of the above MRI protocol has been conducted in phantoms, cadavers, animals, and humans (1,8,9).
A broad and rigorous quality assurance program was used. The precision error of the reading was <1%.
Anthropometric measurements
Anthropometric measurements consisted of weight, height, upper arm and thigh circumferences, skin folds, waist circumference and sagittal diameter measures. All measures were obtained following the standard protocols outlined in the Anthropometric Reference Standardization Manual (10).
Statistical analysis
Descriptive analyses were conducted on participants' body composition assessments from anthropometry, DXA, and MRI. Pearson correlation coefficients were calculated to examine the relationships between DXA and MRI measurements. Scatter plots were used to illustrate the distribution and association of body composition measurements between MRI and DXA scans. Both simple linear and multiple linear regressions were performed. Possible confounding or effect modification were considered when including a covariate or an interaction term in the regression model leading to a 
10% change in the regression coefficient of the main effect. To assess the discrepancies between measured and predicted SMM, we conducted Bland-Altman analysis (11), where the mean of measured and predicted values are plotted against the mean difference between the predicted and measured values. In all the analyses, the head region was excluded from the total body measurements. Customized ROI as defined above for the total leg, lower, and upper leg were used in the analyses. The appendicular region was defined as the sum of the manufacturer-defined arm regions and the customized leg regions. Statistical analyses were conducted using 2-tailed tests at 
= 0.05 (STATA version 8.2). Values in the text are means ± SD, unless indicated otherwise.
|
Results |
|---|
|
TOPABSTRACTIntroductionMethodsResultsDiscussionLITERATURE CITED |
|---|
|
For the leg region, the SMM from MRI was 8.5 ± 1.3 kg and adipose tissue from MRI was 11.8 ± 3.6 kg. In the customized total leg region from DXA, the LSTM was 12.9 ± 2.0 kg and the fat mass was 9.8 ± 3.0 kg.
Correlation coefficients. MRI-derived SMM and DXA-derived LSTM were inter-correlated (r = 0.91–0.95; P < 0.001) for both total body and the leg region (Fig. 1).
|
|
For the total leg region, r = 0.91 (P < 0.001) between SMM and LSTM, but it was 0.98 (P < 0.001) between adipose tissue and fat mass (data not shown).
The correlations were r = 0.98 and 0.92 (P < 0.001) between adipose tissue and fat mass for the upper and lower leg region, respectively, and r = 0.86 and 0.82 (P < 0.001) between SMM and LSTM for the upper and lower leg region, respectively (data not shown).
Predictive models. The results from linear regression analyses for the univariate and multivariate models are presented in Table 2. Age, ethnicity, height, weight, BMI, and %fat were not significant confounders or effect modifiers in this study. However, age and %fat were included in the multivariate models to allow comparisons of our model with previously published models that contained the same set of variables. The R2 indicated that DXA-derived LSTM was a good proxy of SMM from MRI. Total body LSTM (model 1) explained 88% of the variance [root measured squared error (RMSE) = 0.91 kg] in the total body SMM and leg LSTM (model 4) explained 82% of the variance (RMSE = 0.82 kg) for the leg SMM. Using appendicular LSTM (model 2) improved the prediction of total body SMM, whereas using leg LSTM had a poorer prediction for total body SMM (model 3). Adjusting for age and %fat did not result in any appreciable changes in the predictions for total SMM and leg SMM.
We used Bland-Altman plots to illustrate the relationship between predicted and measured SMM (Fig. 2). The results indicated that the discrepancies in the predictions were independent of the absolute values of SMM. The ranges of the difference between the predicted and measured value for total body SMM were –3.01 to 2.21 kg (when total body LSTM was used), –2.32 to 2.06 kg (when appendicular LSTM was used), and –2.98 to 3.12 kg (when total leg LSTM was used). For the leg region, the difference between predicted and measured SMM ranged from –1.29 to 1.30 kg.
|
|
Discussion |
|---|
|
TOPABSTRACTIntroductionMethodsResultsDiscussionLITERATURE CITED |
|---|
The accuracy of CT and MRI for assessing skeletal muscle is comparable, but CT exposes subjects to a considerable amount of radiation. Considering participants' safety and their acceptance of methods, we used MRI in this validation study. DXA-derived fat mass was lower than MRI-derived adipose tissue mass, because DXA fat mass, a chemical compartment, is closer to adipose tissue minus its protein, mineral, and water content. LSTM by DXA was higher than SMM by MRI in our study. LSTM included total body protein, carbohydrates, soft tissue mineral, and water. In other words, in addition to skeletal muscle, LSTM also includes skin, connective tissues, lean portions of adipose tissue, and organ tissues (13) and is thus expected to be larger than SMM. Despite the differences in absolute values between the corresponding DXA and MRI measurements, a strong correlation between MRI-measured adipose tissue and DXA-measured fat mass was found, as well as a strong correlation between MRI-measured SMM and DXA-measured LSTM. The association between fat mass and adipose tissue was stronger than the association between SMM and LSTM. This can be explained by the fact that DXA-derived LSTM includes organ lean tissue in the trunk and a large amount of skin, nonfat components in the adipose tissue. In this study, the ratio of LSTM:SMM was only 1.5 for the leg region, but the ratio increased to 2.1 for the total body. Thus, in studies of sarcopenia, appendicular LSTM was often chosen over total body LSTM to minimize the difference between LSTM and SMM assessments. Indeed, in our study, although total body, appendicular, and leg LSTM were all highly correlated with total body SMM, appendicular LSTM appeared to be the best and leg LSTM the worst predictor for predicting SMM according to the amount of the variance in SMM explained by the LSTM (R2) and the precision of the model [mean squared error (MSE)]. Hence, in future studies, appendicular LSTM is recommended for predicting SMM. The correlation between MRI and DXA measurements was lower for the upper and lower leg regions, which may be due to the difficulties in matching the anatomic regions on the DXA and MRI images for these smaller regions.
The results reported for this study in older women are largely in agreement with previously published results in younger men and women showing a high correlation between DXA-derived appendicular LSTM and MRI-derived SMM for the total body (4), although the R2 was slightly lower in the elderly adults compared with the younger adults. In the analysis between SMM and LSTM at the leg region, a customized cutoff point (lower edge of the ischial tuberosity to the tips of the toes) was used for the leg region. We found good agreement between DXA and MRI measurements for the leg region. The R2 (0.81) from the regression analysis in this study was also lower than what was previously reported (3) for younger adults where R2 = 0.895 for the leg. It should be noted that in our study, the RMSE was 0.81 kg for predicting total body SMM from appendicular LSTM and 0.82 kg for predicting total leg mass from leg LSTM. These values are similar if not smaller than the previously reported RMSE in Kim's study (4) (RMSE = 1.63 kg) and in Shih's study (3) (RMSE = 1.07 kg), suggesting good precision in our models. The discrepancies in R2 between our models and those previously published may be due to a number of reasons. First, the participants in this study were much older than the participants in the previous studies. The mean age was 71 y in our sample, whereas the mean age was 49 (4) and 45 (3) y in the other studies. SMM decreased with aging with a smaller variation in the older sample. The total body SMM in this study was 16.7 ± 2.0 kg, much lower than the mean (20.3 ± 3.4 kg) in Kim's study (4). The lower SMM in older women may increase the variation in the percentage of skin and organ lean tissues counted in the DXA-derived LSTM; hence, the relative amount of the variance explained by the models in this study was slightly lower than the models in younger adults. Second, this study included a more diverse sample of multiethnic groups of women. There are well-known differences in body composition and fat distribution by ethnicity (14). Due to the small sample size, the possibility that the association between LSTM and SMM was altered by ethnicity cannot be excluded.
Age, ethnicity, fat, BMI, and %fat did not modify or confound the association between SMM and LSTM at any region of interest in this study. Although adding covariates such as age and %fat into the models did increase the amount of variance in SMM that can be explained by the DXA measurements, it is not recommended for use in the multivariate models, because the increased R2 was not appreciably large and including additional variables may introduce more measurement error and reduce the generalizability of the model.
The correct positioning of elderly women on the DXA and MRI machines is sometime difficult, which might have introduced measurement errors into the DXA and MRI scans. The applicability of the predictive models from this study should be tested in other samples of elderly women. Nevertheless, these results have provided direct and strong evidence supporting the use of DXA-derived LSTM to assess SMM in older women.
The results of the Bland-Altman analysis suggest that the models are not biased by the magnitude of the measurements. However, because anyone >250 lb (120 kg) was excluded from this study, these models should not be applied to women beyond this weight limit. A higher precision error may be expected when using these DXA models to predict an individual woman's SMM, because the MSE reported here reflects the average performance of the model to all the women in this study.
In conclusion, LSTM measured by DXA scans is highly correlated with SMM in older women. This study result suggests that DXA is a reliable and feasible technique for assessing SMM in large cross-sectional epidemiologic studies with older women. The validity of DXA in assessing changes in SMM remains to be studied.
Short list of WHI investigators: Program Office: (National Heart, Lung, and Blood Institute, Bethesda, Maryland) Elizabeth Nabel, Jacques Rossouw, Shari Ludlam, Linda Pottern, Joan McGowan, Leslie Ford, and Nancy Geller. Clinical Coordinating Center: (Fred Hutchinson Cancer Research Center, Seattle, WA) Ross Prentice, Garnet Anderson, Andrea LaCroix, Charles L. Kooperberg, Ruth E. Patterson, Anne McTiernan; (Wake Forest University School of Medicine, Winston-Salem, NC) Sally Shumaker; (Medical Research Labs, Highland Heights, KY) Evan Stein; (University of California at San Francisco, San Francisco, CA) Steven Cummings. Clinical Centers: (Albert Einstein College of Medicine, Bronx, NY) Sylvia Wassertheil-Smoller; (Baylor College of Medicine, Houston, TX) Jennifer Hays; (Brigham and Women's Hospital, Harvard Medical School, Boston, MA) JoAnn Manson; (Brown University, Providence, RI) Annlouise R. Assaf; (Emory University, Atlanta, GA) Lawrence Phillips; (Fred Hutchinson Cancer Research Center, Seattle, WA) Shirley Beresford; (George Washington University Medical Center, Washington, DC) Judith Hsia; (Los Angeles Biomedical Research Institute at Harbor- UCLA Medical Center, Torrance, CA) Rowan Chlebowski; (Kaiser Permanente Center for Health Research, Portland, OR) Evelyn Whitlock; (Kaiser Permanente Division of Research, Oakland, CA) Bette Caan; (Medical College of Wisconsin, Milwaukee, WI) Jane Morley Kotchen; (MedStar Research Institute/Howard University, Washington, DC) Barbara V. Howard; (Northwestern University, Chicago/Evanston, IL) Linda Van Horn; (Rush Medical Center, Chicago, IL) Henry Black; (Stanford Prevention Research Center, Stanford, CA) Marcia L. Stefanick; (State University of New York at Stony Brook, Stony Brook, NY) Dorothy Lane; (The Ohio State University, Columbus, OH) Rebecca Jackson; (University of Alabama at Birmingham, Birmingham, AL) Cora E. Lewis; (University of Arizona, Tucson/Phoenix, AZ) Tamsen Bassford; (University at Buffalo, Buffalo, NY) Jean Wactawski-Wende; (University of California at Davis, Sacramento, CA) John Robbins; (University of California at Irvine, CA) F. Allan Hubbell; (University of California at Los Angeles, Los Angeles, CA) Howard Judd; (University of California at San Diego, LaJolla/Chula Vista, CA) Robert D. Langer; (University of Cincinnati, Cincinnati, OH) Margery Gass; (University of Florida, Gainesville/Jacksonville, FL) Marian Limacher; (University of Hawaii, Honolulu, HI) David Curb; (University of Iowa, Iowa City/Davenport, IA) Robert Wallace; (University of Massachusetts/Fallon Clinic, Worcester, MA) Judith Ockene; (University of Medicine and Dentistry of New Jersey, Newark, NJ) Norman Lasser; (University of Miami, Miami, FL) Mary Jo O'Sullivan; (University of Minnesota, Minneapolis, MN) Karen Margolis; (University of Nevada, Reno, NV) Robert Brunner; (University of North Carolina, Chapel Hill, NC) Gerardo Heiss; (University of Pittsburgh, Pittsburgh, PA) Lewis Kuller; (University of Tennessee, Memphis, TN) Karen C. Johnson; (University of Texas Health Science Center, San Antonio, TX) Robert Brzyski; (University of Wisconsin, Madison, WI) Gloria E. Sarto; (Wake Forest University School of Medicine, Winston-Salem, NC) Denise Bonds; (Wayne State University School of Medicine/Hutzel Hospital, Detroit, MI) Susan Hendrix.
|
FOOTNOTES |
|---|
2 Author disclosures: Z. Chen, Z. Wang, T. Lohman, S. B. Heymsfield, E. Outwater, J. S. Nicholas, T. Bassford, A. LaCroix, D. Sherrill, M. Punyanitya, G. Wu, and S. Going, no conflicts of interest.
10 Abbreviations used: CT, computed tomography; DXA, dual-energy X-ray absorptiometry; %fat, percent fat; LSTM, lean soft tissue mass; MSE, mean squared error; RMSE, root measured squared error; ROI, regions of interest; SMM, skeletal muscle mass; WHI, Women's Health Initiative.
Manuscript received 15 June 2007. Initial review completed 30 July 2007. Revision accepted 10 September 2007.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTIntroductionMethodsResultsDiscussionLITERATURE CITED |
|---|
1. 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. 1998;85:115–22.
2. Visser M, Fuerst T, Lang T, Salamone L, Harris TB. Validity of fan-beam dual-energy X-ray absorptiometry for measuring fat-free mass and leg muscle mass. Health, Aging, and Body Composition Study–Dual-Energy X-ray Absorptiometry and Body Composition Working Group. J Appl Physiol. 1999;87:1513–20.
3. Shih R, Wang Z, Heo M, Wang W, Heymsfield SB. Lower limb skeletal muscle mass: development of dual-energy X-ray absorptiometry prediction model. J Appl Physiol. 2000;89:1380–6.
4. Kim J, Wang Z, Heymsfield SB, Baumgartner RN, Gallagher D. Total-body skeletal muscle mass: estimation by a new dual-energy X-ray absorptiometry method. Am J Clin Nutr. 2002;76:378–83.
5. The Women's Health Initiative Study Group. Design of the Women's Health Initiative clinical trial and observational study. Control Clin Trials. 1998;19:61–109.[Medline]
6. Langer RD, White E, Lewis CE, Kotchen JM, Hendrix SL, Trevisan M. The Women's Health Initiative Observational Study: baseline characteristics of participants and reliability of baseline measures. Ann Epidemiol. 2003;13:S107–21.[Medline]
7. Snyder. Report of the Task Group on Reference Man. Oxford: Pergamon; 1975.
8. Heymsfield SB. Imaging techniques of body composition: advantages of measurement and new uses. In: Carlson-Newberry CR, editor. Emerging technologies for nutritional research. Washington DC: National Academy Press; 1997. p. 127–50.
9. Tang H, Vasselli JR, Wu EX, Boozer CN, Gallagher D. High-resolution magnetic resonance imaging tracks changes in organ and tissue mass in obese and aging rats. Am J Physiol Regul Integr Comp Physiol. 2002;282:R890–9.
10. Lohman TG, Roche AF, Martorell R. Anthropometric standardization reference manual. Champaign (IL): Human Kinetics Books; 1988.
11. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;1:307–10.[Medline]
12. Levine JA, Abboud L, Barry M, Reed JE, Sheedy PF, Jensen MD. Measuring leg muscle and fat mass in humans: comparison of CT and dual-energy X-ray absorptiometry. J Appl Physiol. 2000;88:452–6.
13. Wang J, Pierson RN Jr. Disparate hydration of adipose and lean tissue require a new model for body water distribution in man. J Nutr. 1976;106:1687–93.
14. Deurenberg P, Deurenberg-Yap M. Differences in body-composition assumptions across ethnic groups: practical consequences. Curr Opin Clin Nutr Metab Care. 2001;4:377–83.[Medline]
This article has been cited by other articles:
|
|
 |
|
  O. Bouillanne, C. Dupont-Belmont, P. Hay, B. Hamon-Vilcot, L. Cynober, and C. Aussel Fat mass protects hospitalized elderly persons against morbidity and mortality Am. J. Clinical Nutrition, September 1, 2009; 90(3): 505 - 510. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
