National Academies Press: OpenBook

Designing Foods: Animal Product Options in the Marketplace (1988)

Chapter: Assessing Body Composition

« Previous: Hormonal Regulation of Growth
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 242
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 243
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 244
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 245
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 246
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 247
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 248
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 249
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 250
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 251
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 252
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 253
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 254
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 255
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 256
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 257
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 258
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 259
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 260
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 261
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 262
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 263
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 264
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 265
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 266
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 267
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 268
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 269
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 270
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 271
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 272
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 273
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 274
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 275
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 276
Suggested Citation:"Assessing Body Composition." National Research Council. 1988. Designing Foods: Animal Product Options in the Marketplace. Washington, DC: The National Academies Press. doi: 10.17226/1036.
×
Page 277

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Methodologies for Measuring Body Composition in Humans HWAI-PING SHENG "Nothing is measured with greater error than the human body." Beneke, 1878 Studies of the composition of the human body are relatively new in human biology. Although interest in the effects of malnu- trition on body tissues, morbidity, and mor- tality clates back to the time of Hippocrates, the primary research interest in body com- position began in the 1940s. Since that time, many procedures ant! techniques have been developed to assess indirectly the various components of the body. These techniques have been discusser] and evaluated at a number of symposia and in review articles (Brozek, 1963, 1965; Brozek and Henschel, 1961; Garrow, 1982; Lohman, 1984; Na- tional Academy of Sciences, 1968; Siri, 1961~. This paper briefly reviews the indirect methods that are currently available to measure adipose tissue (fat) in the body. These procedures are directed either to whole-body measurement or to specific sites and regions, in which case they extrapolate data to whole-bocly fat content using pre- viously determined relationships. Methods range from the simple to the complex; most make use of constants and assumptions de- rivec3 from either the guinea pig data of Pace and Rathbun (1945) or cadaver analy- ses, which form the basis for the "reference man" (Brozek et al., 1963~. Most indirect 242 methods were validated against another in- direct method; few studies were validatecl by direct cadaver analysis (Knight et al., 1986) or carcass analysis of an animal (Lewis et al., 1986; Sheng and Huggins, 1979~. The terms lean body mass (LBM) andfat- free mass (FFM) are used interchangeably and sometimes cause confusion among in- vestigators. Body mass can be considered as the sum of adipose tissue and LBM or, alternatively, of ether-extractable fat and FFM. The terms would be synonymous if adipose tissue were composed of pure fat instead of approximately 80 to 85 percent fat, 2 percent protein, en cl 13 to 18 percent water. Thus, the distinction between LBM and FFM is not critical in a fairly lean incliviclual, but it is important in an obese individual in whom the contribution of the nonfat component to the adipose tissue can be large. The terms fat en c! FFM will be used in this paper. WHOLE-BODY MEASUREMENTS Fat-Soluble Gases Because most anesthetic gases are rare inert gases (for example, cyclopropane, xe

MEASURING BODY COMPOSITION IN HUMANS none, and krypton) and are highly soluble in fat but not in water, it is theoretically possible to use the dilution principle to calculate total body fat by measuring the absorption of these gases. Measurements of fat by this technique have been reported by several investigators who used either the absorption phase (Lesser and Zak, 1963) or both the absorption and Resorption phases of the gas (Mettau et al., 1977) to calculate body fat. Results for small animals agreed with those obtainer] by carcass analysis; results for humans were in the range of published data for fat (Mettau et al., 19771. The disadvantages of this method include the necessity of a closed respiratory system and the length of time required to attain equilibrium conditions, both of which in- convenience the subject. Attempts to re- duce the duration of the experiment by extrapolation of the early phase of measure- ments have been relatively unsuccessful. Body Compartmentalization into Two Components Most indirect methods compartmentalize the body in numerous ways, depending on the purpose of the study and the require- ments of the investigators. In its simplest form, body mass can be considered to consist of two phases: fat and FFM. In the research setting, body fat content is often determined by deriving FFM from a set of measure- ments and then calculating body fat content as part of the body mass not accounted for by FFM. The concept of a fat-free body was originally suggested in 1915 by Dubois and Benedict, who proposed that the FFM was metabolically important and had a constant chemical composition. Research on FFM was accelerated markedly during the 1940s by Behnke et al. (1942), who attempted to measure the amount of"primary energy- exchanging mass" of tissues in the body, which they called the LBM. The LBM or FFM has been estimated by several meth- ods, all of which, as summarized by Wedg 243 wood (1963), assume that LBM has a con- stant density, LBM has a constant proportion of water, bone is a constant proportion of LBM, and cell water is a constant proportion of cell mass. It is also frequently assumed that LBM has a constant percentage of potassium. Determination of Fat by Densitometry Densitometric determination of body fat is considered by many investigators as the "reference method" or the "standard" against which other indirect methods are compared. Equations have also been developed to predict body fat from anthropometric meas- urements using fat data obtained by densi- tometry. The estimation of body fat from densitometry was pioneered by Behnke et al. (1942), who reasoned that if the densities of the two body components (fat ant] LBM) were known and if the density of the whole body could be measured, then the propor- tional masses of fat and LBM could be calculated. Although the concept of densi- tometry is theoretically sound, an accurate measurement of body density and, for the two-compartment approach, the known densities for body fat and FFM are required. Theoretically, body density can be meas- ured with an accuracy of + 0.001 to + 0.025 g/ml (Sir), 1961), but in practice this is difficult to achieve. Body density is calcu- lated using an Archimedean principle: Body density = Body mass/body volume. Many different methods have been devel- oped to measure body volume, but as yet, none appears to yield a satisfactory level of accuracy. The original, and still most widely used, physical method to measure body volume uses either underwater weighing (Gnaedin- ger et al., 1963) or water displacement (Garn and Nolan, 1963~. This measurement can be made with a relatively simple apparatus, but it supers from two practical problems: (1) subject cooperation is required because

244 whole body submersion is essential, and (2) residual volumes of air in the lungs and the gut have to be measured separately. A1- though the residual volume of air in the lungs can be measured easily, no adequate method is available for measuring air in the gut. Photogrammetry has been suggested as a tool for the measurement of body volume (Pierson, 1963), but thus far has not proved successful because photomapping requires complicated mathematics and highly skilled personnel. An adclitional drawback in this method is the inclusion of the residual volumes of air in the lungs and gut. Diethelm et al. (1977) and Garrow et al. (1979) have reported the successful use of a combination of water displacement (to meas- ure a partially submerges! body) and air displacement (to measure the nonsub- merged head region). Certain technical dif- ficulties, such as volume of air in the gut and thermodynamic problems with the air- displacement method, have yet to be re- solved. Bocly volume measured by air dis- placement is theoretically simple but tech- nically difficult. In theory, the volume of air clisplaced by an infant placer! in a rigid chamber can be measured by either the helium dilution method or by measuring the pressure cli~erence as described by Boyle's law (Faulkner, 1963; Fomon et al., 1963; Gnaedinger et al., 1963; Lim, 1963; Taylor et al., 1985~. If the chamber volume is 30 liters and if a piston changes the volume by 0. 3 liters, a 2-liter premature infant would only change the incremental pressure over that of the empty chamber by 0. 76 cm of water (or 0.073 percent). To measure such small pressure changes accurately is difficult, since a change of temperature from 36 to 37°C at a constant volume and an ambient pressure of 760 mm Hg would cause a pressure rise of 3.34 cm of water. This technical difficulty can be resolved by the development of a differential dynamic system where identical volume changes in two identical chambers are induced by two APPENDIX yokel! pistons (Taylor et al., 1985~. Any differential pressure, as measured by a ma- nometer between the two chambers, would! be due entirely to the difference in air volume between the chambers. This system would require a resolution of the differential pressure of 1 percent (instead of 0.073 percent) for a 1 percent change in body volume (Taylor et al., 1985~. Body volume measurements obtained by this system are generally reasonable, although widely di- vergent values are producer! occasionally, probably because of pressure fluctuations from respiratory movement and tempera- ture changes (Taylor et al., 1985~. When the technical difficulties are resolved, this method may be particularly suited for in- fants, because corrections for residual vol- umes of air in the lungs and gut are not necessary. The acoustic plethysmograph is another method being explored to measure body volume (Deskins et al., 1985~. It makes use of the Helmholtz principle that resonant frequency is inversely proportional to the volume of the resonating chamber; that is, the volume of an object placed inside the resonating chambers can be calculated from the difference in resonant frequencies. The acoustic plethysmograph can be constructed and operated relatively inexpensively and can be easily used to measure holly volume in infants. Its disadvantages include a lack of ability to measure the residual volume of air in the lungs and gut. Even with the assurance that body (lensity can be measured with great "precision" for a given in~liviclual, and perhaps with great "accuracy," the application of the densito- metric approach to measure FFM and fat is not without error. The values of 0.9 for the density of body fat and 1.10 or 1.095 for the density of mixed tissues of the FFM are used in the calculations (Brozek et al., 1963~. Although the density of body fat varies at different body sites and from con- sumption of different diets, the variations reported are less than 2 percent (Pearson

MEASURING BODY COMPOSITION IN HUMANS et al., 1968~. Therefore, their contribution to the error in the estimation of fat is small. However, there is an increasing realization that it is invalid to assume the chemical constancy of FFM (WedgwoocI, 1963~; thus, the value of 1.095 for the density of FFM (derived from cadaver analysis) must be used cautiously. The greatest change in the chem- ical composition of FFM occurs during the growth of the infant, resulting in an increase of FFM density from 1.064 in infants (Fomon et al., 1982) to 1.095 for the reference man (Brozek et al., 1963~. The extent of error in fat estimation can be calculated for an infant: Fat content was estimated as 11 percent of body weight when a density value of 1.064 was used for FFM and 23 percent when 1.095 was used. Thus, reported percentages of body fat must be viewer] with caution. As Brozek et al. (1963) concluded after a detailed review of the method, "It appears that no universally valic] formulas for clen- sitometric estimation of the fat content can be offered." Determination of FFM by Hydrometry A value for body fat may be derived simply from the total body water (TBW) measure- ment baser] on the assumption that FFM has a constant water content of 73.2 percent: FFM = TBW/0.732. Measurement of TBW is theoretically sim- ple, requiring the estimation of dilution spaces of small-molecular-weight substances or tracer doses of isotopically labeled water (Schoeller et al., 1980; Sheng and Huggins, 1979~. However, increasing evidence sug- gests that tritiated water overestimates TBW to varying degrees in animals in various nutritional ant! physiological states, espe- cially in rapidly growing young animals (McManus et al., 1969; Sheng and Huggins, 1979~. The degree of overestimation of TBW would affect the degree by which body fat was underestimatecI. Use of a "constant" for the hydration of 245 FFM has been questioned. In the derivation of this constant from eviscerated guinea pigs and several other species of animals, Pace and Rathbun (1945) recognized that the constant-73.2 percent can be applied only to adult animals, a provision that has occa- sionally been overlooked. Even in adult animals, fatter animals tend to have a higher FFM water content. Moulton (1923) rec- ognized in 1923 that relative water content was reduced during early growth in a num- ber of animal species. The animal reaches chemical maturity only when its relative water content stabilizes, ant] the age at which stabilization occurs depends on the species. This concept has been challenged by various investigators and lately by Shie! :Is et al. (1983), who could! finch no evidence of a constant chemical composition in the fat- free body portion of pigs. The pigs in the growth study reached a body weight of 150 kg. Consequently, care must be exercised when the value 73.2 percent is applied in the young; otherwise, underestimation of the fat will result. Determination of Fat from Potassium FFM can be estimated from potassium (K) by the following equation: FFM = Total body K/68.1. In this equation, FFM is assumed to have a constant proportion of K throughout life: 68.1 mmol/kg of FFM, a value derived from cadaver analysis (Kirton and Pearson, 1963). This assumption, however, does not apply in all circumstances; evidence has shown that infants have a lower concentration of K (Forbes and Hursh, 1963) and that adult K concentrations may differ between pop- ulations and ethnic groups (Meneely et al., 19631. Total body K has been measured with the dilution of OK, a radioisotope of K (Corsa et al., 1950). Alternatively, body K can be estimated by measuring the naturally oc- curring radioisotope 40K which constitutes

246 approximately 0.012 percent of the natural K in humans (Forbes, 1962~. The high- energy gamma ray emitted from 40K can be measured with highly sensitive, but expen- sive, whole-body counters. Proper calibra- tion of this system has permitted quantifi- cation of the K concentration in the human body from which FFM, and hence fat, can be estimated. Total Body Electrical Conductivity and Impedance Total body electrical conductivity (TO- BEC) and bioelectrical impedance analysis (BIA) have recently been used to assess adiposity (Cochran et al., 1986; Harrison and Van Itallie, 1982; Lukaski et al., 1985; Segal et al., 1985~. These methods are cliscussed in greater detail in the paper by Boileau in this volume. Briefly, these two techniques use the basic principle that lean tissue conducts an electrical current better than fat tissue. Values for FFM obtained by these techniques compare favorably with those obtained by other indirect methods, such as anthropometry, clensitometry, hy- c3rometry, and total body K (Cochran et al., 1986; Harrison ant] Van Itallie, 1982; Lu- kaski et al., 1985; Segal et al., 1985), and with direct carcass analysis of animals (Fior- otto et al., 1987~. However, as discussed by Cohn (1985), data that validate these tech- niques are incomplete and additional studies are needed. Multicompartmentalization of the Body The above methods to estimate body fat from FFM assume that the chemical com- position of FFM is constant, an assumption that undoubtedly introduces an error whose boundaries are not well defined`. In contrast, the recent development of more sophisti- cated acid complex techniques for the ele- mental analysis of the belly allows a more accurate estimate of body fat without such an assumption. This multicompartmental APPENDIX approach (Anderson, 1963; Cohn et al., 1984, 1985) was first used by Moore et al. (1963), who objected to using FFM as a reference standard because it contained a significant amount of extracellular tissues, primarily skeleton and extracellular fluid. Moore suggested that the term "body cell mass" (BCM), which is a more homogeneous mass responsible for basal metabolism, re- place the term FFM. The calculation of BCM is based on the assumptions that nearly all K is in the cells, the ratio of K to nitrogen (N) is constant (3 mmol of K/g of N), and N is a constant proportion of BCM. Thus, BCM can be calculated from meas- ured K multiplied by a coefficient factor of 8.33. The three-compartment approach, as conceived by Moore and colleagues, clivides the body mass into fat, BCM, and extracel- lular tissue (ECT) compartments. The concept of BCM only recently re- ceived the attention it deserves. The de- velopment of total body neutron activation analysis allows the estimation of extracel- lular tissues-the solid phase estimated from body calcium (Ca) and the aqueous phase from body chlorine (Cat. Bo(ly cell mass can be estimated from body K by measuring 40K (Cohn et al., 1984, 1985~. Body fat then can be calculated after estimation of the BCM and ECT compartments (Cohn et al., 1984, 1985~. This approach, although theo- retically superior to the two-compartment approach, also uses assumptions derived from cadaver analysis-that is, a constant proportion of body Ca in the extracellular solids, a constant ratio of K/N, and a constant proportion of N in the BCM. Body Ca can be measured accurately by a well-calibrated neutron activation system; a relatively small error is introduced into the final estimation of fat by assuming a constant proportion of body Ca in extracellular solids. The potential error resulting from the use of the K/N ratio of 3 mmol/g may be substantial (Sheng and Huggins, 1973~. Expansion of the three-compartment ap- proach to the four-compartment approach

MEASURING BODY COMPOSITION IN HUMANS reduces the body into its four elemental phases: fat, water, protein, and minerals. An accurate estimate of fat is possible if water (measurer! by dilution technique), protein (calculated from body nitrogen measured by prompt gamma neutron acti- vation analysis), and body minerals (calcu- lated from body Ca measurer] by delayed neutron activation analysis) are accurately measured (Cohn et al., 1984, 1985~. The only assumption made is that Ca is a constant proportion of body minerals. Any error introduced into the body fat estimation by this assumption is small because of the small proportion of minerals in the whole body (4 percent). The predominant clisadvantages of neutron activation analysis are its com- plexity, cost, and the radiation exposure, however minimal, to growing infants and adults of childbearing age. The clensitometric method has been ap- plied recently to the pediatric population using the four-compartment approach in which total body water was measured with a tracer, and the mineral content and the densities of fat, water, protein, ant! minerals were obtainer! Tom relevant literature (Sheng et al., 1984~. The use of literature values for mineral content and the various densities to estimate fat appeared to introduce only a small error. Body volume was measured using either the pressure-differential method] (Dell et al., 1987; Taylor et al., 1985) or the acoustic plethysmograph (Deskins et al., 1985~. As discusser! earlier, the overestimation of TBW by tritium, particularly in the infant, may introduce error to the four-compart- ment approach of estimating fat. Recently, Lewis et al. (1986) reported that TOW in the infant baboon can be measured accu- rately by nuclear magnetic resonance (NMR). NMR's potential for the analysis of body composition appears promising; also with NMR imaging, regional distribution of body fat can be analyzed (Fuller et al., 19851. Further developments of this technique may result in the measurement of total body 247 fat without the use of assumptions as in the compartmental approaches. REGIONAL FAT MEASUREMENT Progress has been made in the develop- ment of methods to measure composition at various regions of the body. Body fat is calculated from these measurements using equations that establish a relationship be- tween these measurements and body fat estimated by another indirect method. Such methods, which will not be discussed in detail, range from simple anthropometric measures used primarily for population studies to sophisticated computerized meth- ocis in a research setting. For all these methods, validation has been primarily with another indirect method; that is, the values obtained were compared with reported body fat values in the literature or compared against values obtained by other indirect methods performer] on the same individual. The least expensive and most frequently used method uses calipers to measure skin- fold thicknesses at specific sites. Other methods include soft-tissue radiography (Garn, 1957) and ultrasonography (Borkan et al., 1982), both of which use expensive and nonportable instruments. More re- cently, infrared interactance has been pro- posed as a rapid, safe, and noninvasive method to measure subcutaneous fat in both research and field settings (Conway et al., 1984~. Numerous studies have attempted to validate the extrapolation of subcutaneous fat thickness measured at a number of sites on the body to total body fat and to establish subcutaneous fat thickness as a "standard" for the assessment of total body fat (Durnin and Rahaman, 1967~. Although the thickness of subcutaneous fat is roughly proportional to the total weight of body fat, body fat calculated by this method may be inaccurate and misleading because of the variation among population norms. Equations are being developed to overcome this difficulty; specific formulas for body fat estimation are

248 suggested for specific population groups (Lohman, 1981). Interest in adapting complex diagnostic tools to estimate body fat is increasing. Images depicting fat and muscle of body regions can be obtained with computerized axial tomography (Borkan et al., 1983; Heymsfield en c] Noel, 1981; Sjostrom et al., 1986), dual-photon absorptiometry (Got- fredsen et al., 1986; Mazess et al., 1984), and nuclear magnetic resonance (Fuller et al., 1985~. Sophisticated software allows to- tal body fat to be computed from a series of cross-sectional fat areas along the length of the body. Although all these techniques show great potential in the estimation of body fat, they are expensive and relatively unavailable for routine measurements. Fur- thermore, a degree of radiation exposure is involved with both computerized axial to- mography and dual-photon absorptiometry methods. SUMMARY Many indirect methods of varying degrees of complexity are available for estimation of body fat. Most of the methods have been validated for predictability and precision using other indirect methods. The reference method most commonly used is that based on densitometry to estimate body fat from a two-compartment approach (Sheng et al., 1984~. The accuracy of most of these meth- ods has been validated only in a few in- stances by direct carcass analysis. The final choice of an indirect method ultimately depends on its cost, the objective of the experiment, and the physical conditions under which it is to be used. ACKNOWLEDGMENTS This work is a publication of the U.S. Department of Agriculture/Agricultural Re- search Service Children's Nutrition Re- search Center, Department of Pediatrics, Baylor College of Medicine and Texas Chil APPENDIX dren's Hospital, Houston, Texas. This proj- ect has been partially funded by the U.S. Department of Agriculture, Agricultural Re search Service, under Cooperative Agree ment 58-7MNl-6-100. REFERENCES Anderson, E. C. 1963. Three-component body com- position on analysis based upon potassium and water determinations. Ann. N. Y. Acad. Sci. 110: 189-212. Behnke, A. R., B. G. Feen, and W. C. Welham. 1942. The specific gravity of healthy men: Body weight and volume as an index to obesity. J. Am. Med. Assoc. 118:495-498. Borkan, G. A., D. E. Hults, J. Cardarelli, and B. A. Burrows. 1982. Comparison of ultrasound and skin- fold measurements in assessment of subcutaneous and total fatness. Am. J. Phys. Anthropol. 58:307- 313. Borkan, G. A., D. E. Hulls, S. G. Gerzof, B. A. Burrows, and A. H. Robbins. 1983. Relationships between computed tomography tissue areas, thick- nesses and total body composition. Ann. Human Biol. 10:537-546. Brozek, J. 1963. Body composition, Parts I and II. Ann. N. Y. Acad. Sci. 110: 1-1018. Brozek, J., ed. 1965. Human Body Composition: Approaches and Applications. New York: Pergamon. Brozek, J., and A. Henschel, eds. 1961. Techniques for Measuring Body Composition. Washington, D. C.: National Academy of Sciences. Brozek, J., F. Grande, J. T. Anderson, and A. Keys. 1963. Densitometric analysis of body composition: Revision of some quantitative assumptions. Ann. N.Y. Acad. Sci. 110:11~140. Cochran, W. J., W. J. Klish, W. W. Wong, and P. D. Klein. 1986. Total body electrical conductivity used to determine body composition in infants. Pediatr. Res. 20:561-564. Cohn, S. H. 1985. How valid are bioelectric impedance measurements in body composition studies? Am. J. Clin. Nutr. 44:30~308. Cohn, S. H., A. N. Vaswani, S. Yasumura, K. Yuen, and K. J. Ellis. 1984. Improved models for deter- mination of body fat by in viva neutron activation. Am. J. Clin. Nutr. 40:25~259. Cohn, S. H., A. N. Vaswani, S. Yasumura, K. Yuen, and K. J. Ellis. 1985. Assessment of cellular mass and lean body mass by noninvasive nuclear tech- niques. J. Lab. Clin. Med. 105:305-311. Conwav. T. M.. K. H. Norris and C. E. Bodwell. 1984. A new approach for the estimation of body composition: Infrared interactance. Am. J. Clin. Nutr. 40:1123-1130. Corsa, J., Jr., J. M. Olney, Jr., R. W. Steenburg, M.

MEASURING BODY COMPOSITION IN HUMANS R. Ball, and F. D. Moore. 1950. The measurement of exchangeable potassium in many by isotope di- lution. J. Clin. Invest. 29: 128~1295. Dell, R., Y. Aksoy, S. Kashyap, M. Forsythe, R. Ramakrishnan, C. Zucker, and W. C. Heird. 1987. Relationship between density and body weight in prematurely born infants receiving different diets. Pp. 91-97 in In Vivo Body Composition Studies, K. J. Ellis, S. Yasamura, and W. D. Morgan, eds. London: The Institute of Physical Medicine. Deskins, W. G., D. Winter, H.-P. Sheng, and C. Garza. 1985. Use of a resonating cavity to measure body volume. J. Acoust. Soc. Am. 77:-756-758. Diethelm, R., J. S. Garrow, and S. F. Stalley. 1977. An apparatus for measuring the density of obese subjects. J. Physiol. 267:14P. Durnin, J. V. G. A., and M. M. Rahaman. 1967. The assessment of the amount of fat in the human body from measurements of skinfold thickness. Br. J. Nutr. 21:681-689. Faulkner, F. 1963. An air displacement method for measuring body volume in babies: A preliminary communication. Ann. N. Y. Acad. Sci. 110:75-79. Fiorotto, M. L., W. J. Cochran, R. C. Funk, H.-P. Sheng, and W. J. Klish. 1987. Total body electrical conductivity measurements: Effects of body com- position and geometry. Am. J. Physiol. 252:R794- R800. Fomon, S. J., R. L. Jensen, and G. M. Owen. 1963. Determination of body volume of infants by a method of helium displacement. Ann. N.Y. Acad. Sci. 110:80- 90. Fomon, S. J., F. Haschke, E. E. Ziegler, and S. E. Nelson. 1982. Body composition of reference chil- dren from birth to age 10 years. Am. J. Clin. Nutr. 35:1169-1175. Forbes, G. B. 1962. Methods for determining com- position of the human body. Pediatrics 29:477-494. Forbes, G. B., and J. B. lIursh. 1963. Age and sex trends in lean body mass calculated from K40 meas- urements: With a note on the theoretical basis for the procedure. Ann. N.Y. Acad. Sci. 110:255-263. Fuller, M. F., M. A. Foster, and J. M. S. Hutchison. 1985. Estimation of body fat by nuclear magnetic resonance imaging. Proc. Nutr. Soc. 44:108A. Garn, S. M. 1957. Roenteeno~rammetric determina 0 c, tions of body compositions. Human Biol. 29:337- 353. Garn, S. M., and P. Nolan, Jr. 1963. A tank to measure body volume by water displacement (BOVOTA). Ann. N.Y. Acad. Sci. 110:91-95. Garrow, J. S. 1982. New approaches to body compo- sition. Am. J. Clin. Nutr. 35:1152-1158. Garrow, J. S., S. Stalley, R. Diethelm, P. H. Pittet, R. Hesp, and D. A. Halliday. 1979. A new method for measuring the body density of obese adults. Br. J. Nutr. 42:17~183. 249 Gnaedinger, R. H., E. P. Reineke, A. M. Pearson, W. D. Van Huss, J. A. Wessel, and H. J. Montoye. 1963. Determination of body density by air displace- ment, helium dilution, and underwater weighing. Ann. N.Y. Acad. Sci. 110:96-108. Gotfredsen, A., J. Jensen, J. Borg, and C. Christiansen. 1986. Measurement of lean body mass and total body fat using dual photon absorptiometry. Metab- olism 35:38-93. Harrison, G. G., and T. B. Van Itallie. 1982. Estimation of body composition: A new approach based on electromagnetic principles. Am. J. Clin. Nutr. 35:1176-1179. Heymsfield, S. B., and R. A. Noel. 1981. Radiographic analysis of body composition by computerized axial tomography. Pp. 161-172 in Nutrition and Cancer: Etiology and Treatment, G. R. Newell and N. M. Ellison, eds. New York: Raven. Kirton, A. H., and A. M. Pearson. 1963. Relationships between potassium content and body composition. Ann. N.Y. Acad. Sci. 110:221-228. Knight, G. S., A. H. Beddoe, S. J. Streat, and G. L. Hill. 1986. Body composition of two human cadavers by neutron activation and chemical analysis. Am. J. Physiol. 250:E179-E185. Lesser, G. T., and G. Zak. 1963. Measurement of total body fat in man by the simultaneous absorption of two inert gases. Ann. N. Y. Acad. Sci. 110:40-54. Lewis, D. S., W. L. Rollwitz, H. A. Bertrand, and E. J. Masoro. 1986. Use of NMR for measurement of total body water and estimation of body fat. I. Appl. Physiol. 60:836-840. Lim, T. P. 1963. Critical evaluation of the pneumatic method for determining body volume: Its history and technique. Ann. N.Y. Acad. Sci. 110:72-74. Lohman, T. G. 1981. Skinfolds and body density and their relation to body fatness: A review. Human Biol. 53:181-225. Lohman, T. G. 1984. Research in progress in validation of laboratory methods of assessing body composition. Med. Sci. Sports Exerc. 16:59~603. Lukaski, H. C., P. E. Johnson, W. W. Bolonchuk, and G. I. Lykken. 1985. Assessment of fat-free mass using bioelectrical impedance measurements of the human body. Am. J. Clin. Nutr. 41:810-817. Mazess, R. B., W. W. Peppler, and M. Gibbons. 1984. Total body composition by dual-photon (~53Gd) absorptiometry. Am. J. Clin. Nutr. 40:834-839. McManus, W. R., R. K. Prichard, C. Baker, and M. V. Petruchenia. 1969. Estimation of water content by tritium dilution of animals subjected to rapid live- weight changes. J. Agric. Sci. Cambridge 72:31-40. Meneely, G. R., R. M. Heyssel, C. O. T. Ball, R. L. Weiland, A. R. Lorimer, C. Constantinides, and E. U. Meneely. 1963. Analysis of factors affecting body composition determined from potassium content in 915 normal subjects. Ann. N.Y. Acad. Sci. 110:271- 281.

250 Mettau, J. W., H. J. Degenhart, and H. K. A. Visser. 1977. Measurement of total body fat in newborns and infants by absorption and Resorption of nonra- dioactive xenon. Pediatr. Res. 11:1097-1101. Moore, F. D., K. H. Olesen, J. D. McMurrey, H. V. Parker, M. R. Ball, and C. M. Boyden. 1963. The Body Cell Mass and Its Supporting Environment. Philadelphia: W. B. Saunders. Moulton, C. R. 1923. Age and chemical development in mammals. J. Biol. Chem. 57:7~97. National Academy of Sciences. 1968. Body Composi- tion in Animals and Man. Washington, D.C.: Na- tional Academy of Sciences. Pace, N., and E. N. Rathbun. 1945. Studies on body composition, III. Water and chemically contained nitrogen content in relation to fat content. J. Biol. Chem. 158:685-691. Pearson, A. M., R. W. Purchas, and E. P. Reineke. 1968. Theory and potential usefulness of body den- sity as a predictor of body composition. Pp. 15~169 in Body Composition in Animals and Man. Wash- ington, D.C.: National Academy of Sciences. Pierson, W. R. 1963. A photogrammetric technique for the estimation of surface area and volume. Ann. N. Y. Acad. Sci. 110: 109-122. Schoeller, D. A., E. van Santen, D. W. Peterson, W. Dietz, J. Jaspan, and P. D. Klein. 1980. Total body water measurement in humans with i80 and 2H labeled water. Am. J. Clin. Nutr. 33:2686-2693. Segal, K. R., B. Gutin, E. Presta, J. Wang, and T. B. APPENDIX Van Itallie. 1985. Estimation of human body com- position by electrical impedance methods: A com- parative study. J. Appl. Physiol. 58:1565-1571. Sheng, H.-P., and R. A. Huggins. 1973. Body cell mass and lean body mass in the growing beagle. Proc. Soc. Exp. Biol. Med. 142:175-180. Sheng, H.-P., and R. A. Huggins. 1979. A review of body composition studies with emphasis on total body water and fat. Am. J. Clin. Nutr. 32:630 647. Sheng, H.-P., W. G. Deskins, D. Winter, and C. Garza. 1984. Estimation of total body fat and protein by densitometry. Pediatr. Res. 18:212A. Shields, R. G., Jr., D. C. Mahan, and P. L. Graham. 1983. Changes in swine body composition from birth to 145 KG. J. Anim. Sci. 57:4~54. Siri, W. E. 1961. Body composition from fluid spaces and density: Analysis of methods. Pp. 22~244 in Techniques for Measuring Body Composition, J. Brozek and A. Henschel, eds. Washington, D.C.: National Academy of Sciences. Sjostrom, L., H. Kvist, A. Cederblad, and U. Tylen. 1986. Determination of total adipose tissue and body fat in women by computed tomography, 40K, and tritium. Am. J. Physiol. 250:E736-E745. Taylor, A., Y. Aksoy, J. W. Scopest, G. du Mont, and B. A. Taylor. 1985. Development of an air displace- ment method for whole body volume measurement of infants. J. Biomed. Eng. 7:9-17. Wedgwood, R. J. 1963. Inconstancy of the lean body mass. Ann. N.Y. Acad. Sci. 110:141-152.

Utilization of Total Body Electrical Conductivity in Determining Body Composition R. A. BOIL~EAU Assessment of body composition is an important part of evaluating nutritional sta- tus, health, and physical fitness. In general, body composition analysis uses concepts and measurement techniques that permit par- titioning of bocly weight into two or more components. The simplest conceptual mode! partitions body weight into a fat weight component ant] a fat-free or lean body weight component. These are of particular interest in relation to human nutrition and health, since obesity is a major health prob- lem in Western societies, both among chil- dren (Coates et al., 1982; Ylitalo, 1981) and adults (Buskirk, 1971; McArdle et al. ,1981~. Furthermore, body composition analysis has many applications for the animal scientist, including nondestructive monitoring of meat production. Measurement of human body composi- tion has been a somewhat perplexing prob- lem because of the necessity to use nonin- vasive techniques and the lack of a substantial data base characterizing the chemical com- position of the body for validation purposes. Hence, the status of our understanding of human body composition has developed from use of indirect measurement tech . 251 piques, the conceptual framework of which is baser] on the chemical analysis of only a few cadavers. The reference techniques judged to be most accurate, precise, and conceptually sound include clensitometrY. hydrometry, and body potassium (40K) spec- troscopy. The methodology of these tech- niques has been reviewed in a number of reports (Behnke and Wilmore, 1974; Boi- leau and Lohman, 1977; Boileau et al., 1985; Keys and Brozek, 1953; Lohman et al., 1984a). Other techniques, consiclered to be less precise but applicable in large population studies, involve skinfold thickness ant] other anthropometric measurements. More re- cent technologies have spawned develop- ment of several new techniques including total body neutron activation analysis (Cohn et al., 1974), computerizes! axial tomography (Borkan and Hults, 1983), nuclear magnetic resonance imaging (Lohman, 1984), whole- body impedance (Nyboer, 1972), and total bocly electrical conductivity analysis (Har- rison and Tan Itallie, 19824. The focus of this report is on total body electrical con- rluctivity (TOBEC) as a technique for body composition analysis. , ~

252 BACKGROUND AND MEASUREMENT PRINCIPLE Electrical conductivity analysis is a method of compositional analysis that uses an in- strument (U.S. Patent 3,735,247, 1973 - called electronic meat measuring equip- ment (EMME - to measure the fat and lean content of live swine (mode} SA-1~. EMME SA-1 was later mollifier! for measurement of packaged meat and in vivo measurements of humans (EMME/TOBEC HA-1. Data presented in the literature are primarily based on the EMME/TOBEC HA-1, which is the prototype of the new TOBEC HA-2. Application of the electrical conductivity method to the two-component body com- position mode} is based on the concept that the fat-free body (FFB) component conducts electrical current more readily than the fat component. This is due to the higher water and electrolyte content found in the tissues ant] extracellular water making up the FFB. Electrical conductivity of various biological materials indicates that constituents asso- ciated with the FFB (for example, muscle, liver, and blood) have conductivity values of about 4 mmho-cm versus conductivity values for fat of about 0.3 mmho-cm in the 2.5- to 5.0-MHz range, an FFB/fat ratio of about 13 (GecIdes and Baker, 1968; Pethig, 1979~. Van Itallie et al. (1985) have sug- gested that the FFB/fat conductivity ratio may be as high as 20 to 1. Current flow inclucec] in a biological sys- tem is a Unction of conductive and dielectric properties. The conductive properties are related to the intra- and extracellular ionic content, and the dielectric effect is associ- ated primarily with capacitance relater} to cell membranes. Impedance to current flow in the system results in an irreversible loss of energy as heat. This energy loss is related to the conductive mass. The dielectric or capacitance properties of current flow in a biological system must also be considered; these represent the reactive part of imped- ance in which energy transfer is reversible APPENDIX due to temporary storage of electrical en- ergy. Capacitance is partly determiners by the geometry of the conductor, which may produce an effect whereby capacitance in- creases as cross-sectional area, length, or both increase. While theoretically both elec- trical properties define the flow of current in a conductive mass, the conductive prop- erties appear to exert a more dominant effect in estimating FFB mass. A detailed! treatment of the electrical properties of biological tissues can be fount] in Pethig (1979~. There are two basic bioe~ectrical tech ~ . ~. . . . . niques used to measure whole-bo(ly con- ~luctivity for body composition assessment: (1) direct injection of current and (2) non- contact total holly electrical conductivity (TOBEC). This discussion focuses on TO- BEC. In this technique current applied to a coil induces an electromagnetic field in which the body is statically situated (HA-1) or scanned (HA-2~. The conducting mass (subject) passing through the electromag- netic field of the cod! absorbs heat energy, thereby perturbing the electrical field of the coil. The loss of energy cletected in the cod! is an index of the concluctive mass of the body. The power dissipated in the subject at any one time is less than 1 ,uW/ cm2-less than 1/lOOth of the standard set by the American National Standards Insti- tute for human exposure. The oscillating current frequency applied to the coil is an important aspect of the measurement, since the degree of separation in the conductivi- ties of FFB and fat has been shown to be frequency dependent (Pethig, 1979~. The first TOBEC model (HA-1) used 5-MHz oscillating cod] current and required a 0.5- second measurement on the statically situ- atec3 subject (Harrison and Van Itallie, 19821. The new TOBEC instrument (HA-2) is a scanning crevice in which the subject moves on a motor-driven sled through a 2.5-MHz cod] electromagnetic field at a constant rate. It requires about 40 seconds for one meas- urement, during which conductivity is

TOTAL BODY ELECTRICAL CONDUCTIVllrY measurer! at 64 equidistant intervals. The change in cod] energy as the body moves through the length of the coil is detected as change in conductance ant! capacitance relative to an empty coil. The measurer! conductance en c] capacitance of the concluc- tor (subject) is reflected in a phase angle/ distance curve. The area under the curve is an index of total body conductivity. The phase angle/distance curve generates] from the HA-2 mode! is transformed by a Fourier series analysis, which partitions components of the average phase curve into discrete terms of phase (PC) and amplitucle (AM) coefficients. Van Itallie et al. (1985) have reported a high correlation (r = 0.98) be- tween the phase average readings of HA-1 and HA-2 in 40 men and women. MEASUREMENT PRECISION AND VALIDITY OF TOBEC Most available information on the relia- bility and validity of TOBEC analysis has been reported on the SA-1 or HA-1 instru- ment. Measurement precision was reported for two studies on animals. Domermuth et al. (1976), using the SA-1 model, measured 12 pigs 14 times a clay for 2 days and fount] an average coefficient of variation (CV) among the animals of 4 percent. Bracco et al. (1983) evaluated measurement reliability on 30 lightly anesthetized rats using a DMe 100 Ground Meat Fat Tester and reported an average intraday reliability coefficient of 0.99 for three consecutive trials. High reliability coefficients have also been reported on humans measured with the HA- 1 instrument. Presta et al. (1983) measured 32 subjects 10 times consecutively over a period of 3 minutes with a reporter] intra- ciass correlation of 0.99 among the trials. Segal et al. (1985), using the same reliability assessment method on 75 subjects, also reported an intracIass correlation of 0.99 and fount] the CV to be less than 2 percent for each subject. Using the HA-2 instru- ment, Van Loan ant! Mayclin (1987) meas 253 urea 14 subjects five times a day for 5 days. No significant cli~erences were found either within subjects or between test days. Pre- cision of the HA-2 measurement has also been evaluated in 12 subjects, measured five consecutive times a day for 3 consecu- tive days, with the University of Illinois HA-2 instrument. No significant variability was cletectec! for the within-day trials or the between-day trials for either the HA-2 reacI- ings or body density. The relative errors were lower for the HA-2 readings than for body density. With the exception of one subject who had one within-day CV of 3.6 percent, all indiviclual daily CVs were less than 2 percent. On the other hand, 17 of 36 CVs exceeded 2 percent for the body density measurement. Therefore, the meas- urement precision of TOBEC appears to be excellent relative to that of other techniques for the assessment of body composition. This can be partly attributed to minimal requirements for subject learning and par- ticipation in the measurement process. Validation of the TOBEC method has been stuclied in both animals and humans. In animal studies, two approaches have been usecI: (1) comparison of TOBEC values to 40K total, body water, and FFB components derived from carcass analysis, and (2) FFB clerivec] from densitometry, total body water, and/or 40K spectroscopy. In humans, vali- dation stu(lies have compared TOBEC-de- rivecl FFB values with FFB derived from several indirect reference methods. Domermuth et al. (1976) were the first to report the relationship between TOBEC and other body composition methods in- clucling total potassium (40K) and carcass analysis in pigs. Two experiments were conducted, one with 42 pigs ant] the other with 35 pigs, in which the animals were faster! for 16 to 18 hours to obtain a shrunk body weight before being measured by TOBEC anal 40K. The animals were then killecl and their carcasses analyzed for spe- cific gravity and fat, water, and protein content. The linear correlations between

254 the live animal TOBEC readings and 40K measurements were 0.75 and 0.81 for ex- periments 1 and 2, respectively. The cor- relations of carcass analysis-derived total body water weight and protein weight to the TOBEC reading in experiment 1 were higher (r = 0.87 and 0.83, respectively) than for the respective live animal 40K data (r = 0.78 and 0.69~. The same trend was observed in experiment 2, but because of the homogeneity of this sample, the coef- ficients were somewhat depressed. Bracco et al. (1983) observed a high as- sociation between TOBEC values in 30 live rats and FFB estimated by clensitometry (r = 0.97; the standard] error of estimate tSEE] was 13.6 g) and by chemical analysis of the carcass (r = 0.97; SEE = 14.2 g). High linear correlations were also observed between TOBEC and total protein (r = 0.95) and total body water (r = 0.98~. Klish et al. (1984) reported a correlation of 0.99 between the natural log of the TOBEC (Infant Moclel, HI-1) reading determined on live rabbits (mean weight = 2.8 kg) and the FFB chemically determined from car- cass analysis. Similarly, Cochran et al. (1985) using five infant miniature pigs (weight range, 2.3 to 4.7 kg) found a high linear correlation (r = 0.99) between the natural log of the TOBEC (HI-1) signal ant] total body water measured by desiccation. The data on animals indicate a strong relation- ship between TOBEC and FFB estimated directly by chemical ant! carcass analyses as well as indirectly by various reference meth- ods. The basic approach to both calibration and validation of TOBEC in the human has been to investigate the association of the TOBEC reacting to FFB and fat estimated by one or more of the reference body composition methods. Four of the five stud- ies reported used the HA-1 or HI-1 instru- ments, and all but one studied adult samples (age range, 18 to 63 years). In the infant study, Cochran et al. (1986) measured sub- jects (aged 2 days to 9.7 months) and re APPENDIX ported the relationships between the TO- BEC (HI-1 model) reading and total body water (deuterium dilution). FFB was esti- mated from total body water (TBW) and the sum of the triceps and subscapular skinfolds. A high correlation (r = 0.96) was obtained for the observed TOBEC reading and FFB was estimated from TBW (using 0.82 as the fraction of water in FFB), with a somewhat lower correlation (r = 0.82) fount] for TO- BEC and the sum of skinfolds. The adult HA-1 data, all of which have been reported by the Columbia University St. Luke's- Roosevelt Hospital group, indicate a strong relationship between TOBEC and the var- ious reference methods. On this basis, it is possible to calibrate the TOBEC number with FFB estimated from body density or another method. While the high correla- tions reported for the HA-1 crevice and FFB derived from the reference methods are impressive, it is noteworthy that the SEE represents CVs of 7.3 percent (Presta et al., 1983) and 5.9 percent (Segal et al., 1985~. When sex was considered as a categorical variable, the correlations increased (0.95 to 0.97) and the SEE was reducer] (3.4 to 2.5 kg) (Segal et al., 19851. The error associates] with estimating fat content clerivecl from body density (DB) for the various indirect methods is lowest for the TOBEC method (SEE = 3.5 percent fat) (Van Itallie et al., 1985~. At present, five TOBEC HA-2 instru- ments are in use for pilot testing. In one study (Van Loan ant! Mayclin, 1987), FFB from DB ant] also DB and TBW were pre- dicted from a series of TOBEC variables (Fourier coefficients) in a group of young adult males and females (aged 18 to 35 years). High multiple r values (0.98) were obtained from regression analysis, with the SEE ranging from 1.4 to 1.7 kg. The re- markably low FFB SEE observed from the DB and Ds-TBw methods of 2.6 and 3.2 percent, respectively, may in part be sample specific, but also may indicate that the HA- 2 predicts FFB better than the HA-1, where

TOTAL BODY ELECTRICAL CONDUCTIVITY the best prediction of FFB was 5.9 percent (Segal et al., 1985). UNIVERSITY OF ILLINOIS TOBEC STUDIES The TOBEC H A-2 was recently evaluated in relation to clensitometry and body potas- sium to assess the relationship of TOBEC to independent measures of body compo- sition. A diverse sample of 190 children and adults were classified by the following levels of maturation: (1) prepubescent and pubes- cent children (aged 8 to 12 years), (2) post- pubescent youths (aged 13 to 18 years), (3) young aclults (aged 19 to 34.9 years), and (4) mature adults (aged 35 years and older). Classification was deemed important, since evidence suggests that the composition of the FFB is unstable cluring growth, devel- opment, en c] aging (Boileau et al., 1985~. Each subject was measured densitomet- rically and with the TOBEC H A-2. Body density was measured weighing the subject underwater; pulmonary resiclual volume was measured at the time of weighing (Boileau et al., 1984~. Fat and FFB expressed both in absolute and relative terms were esti- matec! from density (Sir), 1961~. The TO- BEC technique was describer! above. From a TOBEC scan, several orders of phase angle coefficients (PC-I, PC-II, PC-III, and PC-AYE) and amplitude coefficients (A M 1, A M-II, A M-III, and A M-AYE) were generated by Fourier series analysis. Each subject underwent at least three TOBEC scans. In addition to the DB and TOBEC measurements, 44 subjects were measurer! for body potassium ant] 126 for bone mineral content. Bocly potassium was measurer! by 40K spectroscopy in a at liquid scintillation counter (Boileau et al., 1973~. Bone mineral was measured by single-photon absorptio- metry (Lohman et al., 1984b). The bone mineral content was user! to estimate the percentage of mineral FFB, which in turn was used to calculate a mineral-free FFB (MF-FFB) that could be relatecl to TOBEC variables (Boileau et al., 1985~. 255 The results of this analysis suggest an excellent empirical relationship between TOBEC and FFB estimated by either the DB or 40K methods. Correlations ofthe phase angle coefficients (PC) with FFB DB reflect- ing conductivity appear to be slightly higher than the amplitude (AM) coefficients reflect- ing capacitance. Regression analysis was then conducted to predict FFB DB using linear combinations of the individual TO- BEC variables and the conductivity indexes (for example, PC-I' x Ht. AM-III' x Ht) with the best prediction equations for each selected on the basis of the highest r2 and the lowest SEE. Overall, the SEEs for predicting FFBDB from the conductivity indexes were slightly lower than were those from a combination of individual TOBEC variables, except in the postpubescent group. The next step in the analysis was to evaluate the applicability of the prediction equations across maturation levels. Since the clensitometric method assumes that the FFB is stable ant! chemically mature, and since this assumption may not be valid in growing children, youths, and aging adults, the equation for young aclults was used to predict FFB from TOBEC in the other three groups. * Analysis indicates that while the slopes of the regression of actual (FFBDB) and predicted (FFBToBEc) were somewhat similar to the line of identity (1.0), the intercepts varied consiclerably and were significantly different from zero. Further- more, significant mean differences between actual and predicted FFB were observed for the prepubescent/pubescent and mature aclult groups, but not for the postpubescent group. This suggests that specific equations may need] to be considered, at least for the youngest and oldest groups. The lack of precise estimation of FFB from TOBEC analysis in these groups may partly be flue to assumptions implicit in the DB method * . . . This equation is: FFBToBEc = 1.545 + 0.28 (pC_~/2 X Ht) _ 0,OIO(pC_~/2 X Ht).

256 that may not be valid. The Siri (1961) equation used to compute percent fat and FFB from DB assumes an FFB density of 1.100 g/cm3. However, this assumption ap- pears to be invalid in children and older adults, since changes in water ant! mineral content of the FFB have been shown to lower the FFB density (Layman and Bo- ileau, 1986~. Thus, part of the variability associated with the predictors of FFB from the equation for the young adult group may be related to biological error in the FFB estimated densitometrically in the other three groups. To further study the relationship between FFBDB and FFBToBEc' FFBDB was ex- pressed as MF-FFB. The MF-FFB was calculated by subtracting the estimated bone mineral weight, measured by photon ab- sorptiometry, from the FFBDB. Since bone mineral contains little water and electro- lytes, its conductivity is relatively low com- pared with other components of the FFB; therefore, theoretically the MF-FFB better represents the conductive component of the FFB. Although the SEE is reduced from 2.60 to 2.41 kg, the CV is similar for FFBD" (6.1) and MF-FFBDB (6.0), suggesting that a correction for bone mineral content does relatively little to reduce the error in pre- dicting the actual FFB. Body potassium was also found to be highly related to TOBEC (r = 0.99~. The best predictors of body potassium were PC- I1/2 X Ht and PC-II172 x Ht. yielding an SEE of 6.6 g (CV = 5.7 percent). These inde- pendent variables represent the conduct- ance function of the TOBEC measurement. Potassium is associated primarily with the mineral-free FFB, since only a negligible amount of potassium is found in either fat or bone. Therefore, a high correlation be- tween body potassium and TOBEC further supports the validity ofthe TOBEC method. Total body electrical conductivity analysis appears to be a promising method for as- sessing body composition on both theoret- ical and empirical grounds. The association APPENDIX demonstrated between TOBEC and FFB determined by both densitometry and 40K spectroscopy statistically confirms the theo- retical basis of the method. The reported error in estimating FFB from TOBEC, ranging from 2.6 percent in a homogeneous group (Van Loan and Mayclin, 1987) to 5.0 percent in a heterogeneous group of chil- dren, youth, and adults (University of Illi- nois study), indicates that the method can be applied to a variety of subjects with good prediction precision. This study further sug- gests that population-specific calibration equations may improve the precision of estimating body composition. Although the ability of the TOBEC method to detect changes in body composition during weight reduction appears excellent (M. Van Loan, personal communication, 1986), further val- idation is needed in terms of the technique's capacity for detecting change in FFB, body water, and body potassium as a consequence of dietary, physical training, and dehydra- tion treatments. REFERENCES Behnke, A. R., and J. H. Wilmore. 1974. Evaluation and Regulation of Body Build and Composition. Englewood Cliffs, N.J.: Prentice-Hall. Boileau, R. A., and T. G. Lohman. 1977. The meas- urement of human physique and its effect on physical performance. Orthopedic Clin. N. Am. 8:56~581. Boileau, A., B. H. Massey, and J. E. Misner. 1973. Body composition changes in adult men during selected weight training and jogging programs. Res. Q. 44:158-168. Boileau, R. A., T. G. Lohman, M. H. Slaughter, T. E. Ball, S. B. Going, and M. K. Hendrix. 1984. Hydration of the fat-free body in children during maturation. Human Biol. 56:651-666. Boileau, R. A., T. G. Lohman, and M. H. Slaughter. 1985. Exercise and body composition of children and youth. Scan. J. Sports Sci. 7:17-27. Borkan, G. A., and D. E. Hults. 1983. Change in body fat content and distribution with aging. Am. J. Phys. Anthropol. 60:175. Bracco, E. F., M. U. Yang, K. Segal, S. A. Hashim, and T. B. Van Itallie. 1983. A new method for estimation of body composition in the live rat. Proc. Soc. Exp. Biol. 174:143-146.

TOTAL BODY ELECTRICAL CONDUCTlVll~Y Buskirk, E. R. 1971. Obesity. Pp. 22~242 in Phys- iological Basis of Rehabilitation Medicine, J. A. Downey and R. C. Darling, eds. Philadelphia: W. B. Saunders. Coates, T., J. Killen, and L. Slinkard. 1982. Parent participation in a treatment program for overweight adolescents. Int. J. Eating Disorders 1:37~8. Cochran, W. J., W. J. Klish, W. W. Wong, M. L. Fiorott, P. D. Klein, and B. L. Nichols. 1985. The use of total body impedance to determine body composition in infants. Pediatr. Res. 4:216 (Abstain. Cochran, W. J., W. J. Klish, W. W. Wong, and P. D. Klein. 1986. Total body electrical conductivity used to determine body composition in infants. Pediatr. Res. 20:561~86. Cohn, S. H., K. J. Ellis, and S. Wallach. 1974. In viva neutron activation analysis: Clinical potential in body composition studies. Am. J. Med. 57:683 686. Domermuth, W., T. L. Veum, M. A. Alexander, H. B. Hedrick, J. Clark, and D. Eklund. 1976. Predic- tion of lean body composition of live market weight swine by indirect methods. J. Anim. Sci. 43:96 976. Geddes, L. A., and L. E. Baker. 1968. P. 155 in Principles of Applied Biomedical Instrumentation. New York: John Wiley & Sons. Harrison, G. G., and T. B. Van Itallie. 1982. Estimation of body composition: A new approach based on electromagnetic principles. Am. J. Clin. Nutr. 35:117~1179. Keys, A., and J. Brozek. 1953. Body fat in adult man. Physiol. Rev. 33:245-325. Klish, W. J., G. B. Forbes, A. Gordon, and W. J. Cochran. 1984. New method for the estimation of lean body mass in infants (EMME instrument): Validation in nonhuman models. J. Pediatr. Gas- troenterol. Nutr. 3: 199-204. Layman, D. K., and R. A. Boileau. 1986. Aerobic exercise and body composition. Pp. 12~141 in Nutrition and Aerobic Exercise, D. K. Layman, ed. Washington, D.C.: American Chemical Society. Lohman, T. G. 1984. Research progress in validation of laboratory methods of assessing body composition. Med. Sci. Sports Exercise 16:596 603. Lohman, T. G., R. A. Boileau, and M. H. Slaughter. 1984a. Body composition in children and youth. Pp. 29-59 in Advances in Pediatric Sports Sciences, R. A. Boileau, ed. Champaign, Ill.: Human Kinetics Publishers. 257 Lohman, T. G., M. H. Slaughter, R. A. Boileau, J. Bunt, and L. Lussier. 1984b. Bone mineral meas- urements and their relation to body density in children, youth and adults. Human Biol. 56:667- 697. McArdle, W. D., F. I. Katch, and V. L. Katch. 1981. Exercise Physiology: Energy, Nutrition and Per- formance. Philadelphia: Lea & Febiger. Nyboer, J. 1972. Workable volume and flow concepts of big-segments by electrical impedance plethys- mography. T-I-T J. Life Sci. 2: 1-13. Pethig, R. 1979. Pp. 207-243 in Dielectric and Elec- tronic Properties of Biological Materials. New York: John Wiley & Sons. Presta, E., K. R. Segal, B. Gutin, G. G. Harrison, and T. B. Van Itallie. 1983. Comparison in man of total body electrical conductivity and lean body mass derived from body density: Validation of a new body composition method. Metabolism 32:524 527. Segal, K. R., B. Gutin, E. Presta, J. Wang, and T. B. Van Itallie. 1985. Estimation of human body com- position by electrical impedance methods: A com- parative study. J. Appl. Physiol. 58:1565-1571. Siri, W. E. 1961. Body composition from fluid spaces and density: Analysis of methods. Pp. 223-244 in Techniques for Measuring Body Composition, J. Brozek and A. Henschel, eds. Washington, D.C.: National Academy of Sciences. U.S. Patent 3,735,247. May 22, 1973. Method and Apparatus for Measuring Fat Content in Animal Tissue Either In Vivo or in Slaughtered and Prepared Form, W. H. Harker (inventor). The EMME Com- pany, Assiguee. Washington, D. C.: U. S. Patent Office. Van Itallie, G. G., K. Segal, M. U. Yang, and R. C. Funk. 1985. Clinical assessment of body fat content in adults: Potential role of electrical impedance methods. Pp. 5-8 in Body Composition Assessment in Youth and Adults: Report of the Sixth Ross Conference on Medical Research, A. Roche, ed. Columbus, Ohio: Ross Laboratories. Van Loan, M., and P. Mayclin. 1987. A new TOBEC instrument and procedure for the assessment of body composition: Use of Fourier coefficients to predict lean body mass and total body water. Am. J. Clin. Nutr. 45:131-137. Ylitalo, V. 1981. Treatment of obese school children. Acta Paediatr. Scand. (Suppl. 290~:1-108.

Live Animal and Carcass Composition Measurement DAVID G. TOPEL and ROBERT KAUFFMAN Animals of all species vary considerably in composition (Reicl et al., 1968) as a result of their stage of growth, nutritional history, and genetic base. This is of concern to livestock producers, the meat industry, and consumers because the economic value of a meat-producing animal depends greatly on its composition. During the last 15 to 20 years, the meat industry has made dramatic progress in reducing the fat content of domestic animals in response to consumer demand for more lean meat and the eco- nomic pressure to produce animals more efficiently. Nevertheless, the average pork carcass and U. S. Department of Agriculture Choice beef carcass are still about 30 to 35 percent fat (Topel, 1986~. The proportion of muscle in an animal's body varies from less than 35 percent to nearly 50 percent of the body weight (Web- ster, 1986~. In addition to stage of growth, nutrition, and genetics, several other factors contribute to variation in body composition, such as contents of the alimentary canal, pregnancy, and presence of abnormalities. All these traits collectively complicate the accurate measurement of body composition. Nevertheless, it is important to seek meth 258 oafs that will reliably approximate body composition because of its contribution to the total worth of meat-producing animals. Therefore, this paper identifies the various techniques that have been used to estimate body composition, as well as new techniques that have potential for future application. LINEAR MEASUREMENTS OF LIVE ANIMALS Some years ago, some livestock evaluators believed that measurements of the length, width, height, and circumference of live animals could be used to predict various carcass characteristics inclucling composi- tion (Busch et al., 1969; Cook et al., 1951; Green et al., 1969; Kidwell, 1955~. They used tape measures, centrex curves, and several types of rather elaborate calipers that were designee! to make almost any linear measurement imaginable on the non- symmetrical surfaces of live animals. Ref- erence points were usually confined to an- atomically defined locations of the skeleton. Animals (primarily cattle) were led onto a flat surface and constrained in a neck stan- chion. If the animal was not gentle, it was

LIVE ANIMAL AND CARCAS S COMPOSITION nearly impossible to palpate the body to locate the skeletal reference points, let alone make the measurements. From the numer- ous dimensions measured, the evaluator could very accurately reconstruct the to- pography of the animal's body; however, most if not all of the data were of limited value in determining compositional pat- terns. Many of the measurements were obviously related to weights of carcass wholesale cuts, but not to percent fat or muscle. This approach may have provided some insight into the variations in frame and skeletal size and their relationship to live weight, breed, and stage of maturity; but it was not useful in predicting the muscle, fat, and bone content of animals. LINEAR MEASUREMENTS OF THE CARCASS A simple, inexpensive ruler to measure fat thickness and length and width of the longissimus dorsi muscle has been used by animal scientists for many years. Palsson (1939) evaluated lamb carcasses, Hirzel (1939) measured beef carcasses, and McMeekan (1941) reported strong relationships be- tween linear measurements of back fat thick- ness and carcass fat percentage. These work- ers also reported good estimates of carcass muscle when carcass length was combined with the depth and length of the longissimus dorsi. From these early studies have come more than 200 papers relating linear meas- urements of body fat thickness and longis- simus dorsi area to the muscle, fat, and bone percentage of the carcass as well as to its chemical fat, protein, and water content (see, for example, Berg and Butterfield, 1976; Breidenstein et al., 1968; Brozek, 1961; Cross, 1982; Doornenbal, 1968; Kauff- man, 1971; Kempster, 1986; Zobrisky, 1963~. The relationships of back fat thickness and size of the longissimus dorsi are considered to be good, but not excellent, predictors of body composition. 259 TEIE BACK FAT PROBE The back fat probe was first described scientifically by Hazel and Kline (1952) and has been used often to determine quanti- tatively the actual thickness of subcutaneous fat on live animals. It has been used more extensively on pigs than on cattle and sheep because a greater proportion of the pig's fat is deposited subcutaneously and there is greater variation in the measurement ob- tained when comparing lean and fat pigs. With cattle there is the problem of hide thickness, and with sheep there are minimal variations of fat depth. To measure the actual fat depth, a small incision is made in the skin with a scalpel and a narrow metal ruler is forced through the fat layers, or a ruler containing a needle point is forced directly through the skin and fat layers. (Since the nerve and vascular supplies in the skin and subcutaneous fat are minimal, pain and bleeding are not much of a problem.) The measuring device must penetrate the false lean or aponeurosis (sheet of fascial connective tissue separating the outer and middle layers of subcutaneous fat) and continue until there is a second resistance due to the epimysial connective tissue covering the muscle (usually the lon- gissimus dorsi when probing the thoracic and lumbar portions of the back). The depth should be visually verified and recorded after slipping a flat object over the ruler, firmly sliding it to the skin surface (avoiding undue pressure), and then removing the ruler, needle, or both. Once the fat depth is known, it can be used in a previously developed regression equation with other variables such as live weight and muscling score to estimate com- position. Fat depth alone usually accounts for most of the variation in composition, but live weight and degree of muscling should improve the accuracy of the measurement (Fahey et al., 1977~. The major advantages of this method are that it affords a reasonably accurate predic

260 tion of composition, it is relatively easy to standardize, it makes a rapic! measurement, and it is inexpensive. On the other hand, it requires that the animal be restrained, and it is too slow if large numbers of animals are involvecI. For cattle, a restraining chute is needed, and for lambs, the variations in fat depth may be so small that the method might not be sophisticated enough to yield an accurate measurement. REFLECTANCE PROBE The reflectance probe was developed by researchers at the Danish Meat Research Institute for use on pork carcasses. It is widely used in Europe, but not in the United States. The instrument measures reflectance of the muscle and fat compo- nents when inserted into the loin section of the pork carcass. Desmoulin (1984) sum- marized its value for estimating fat and muscle thickness. Fat content was estimates] with greater accuracy than lean content when weight was not included in the pre- diction equation (r2 = 0.82~. The combi- nation of weight and carcass length with the reflectance probe measurement correlates well fry = 0.77) with muscle percent. The cost of the equipment is relatively high but not excessive if it is used on a ciaily basis for grading pork carcasses in large slaughter plants. It is simple to use, and refilings can be obtained in less than a minute. The optical probe is used on cold carcasses only, since these yield the largest clifferences in light reflection between mus- cle ant] fat tissue. The reflection probe can also indicate the meat quality characteristics of the carcass (Barton, 1983~. LIVE WEIGHT The development of growth curves from the animal's live weight provides a practical and simple method for estimating body composition if the genetic history for bocly composition is known. As animals grow, APPENDIX their carcass composition changes and the proportion of fat increases at the expense of muscle and bone (Rouse et al., 1970~. When comparing animals of similar type grown in the same environment, live weight will normally show a high positive correlation with the percentage of fat in the carcass and, because of the close relationship be- tween muscle and bone, a high negative correlation with the proportion of muscle in the carcass (Busch et al., 1969). Even in mixed-breed populations with animals obtained from different production systems, there is often a strong association between weight and fatness, although the degree of correlation is more variable. Used sensibly, live weight can be a guide to carcass composition. It can be made more effective when a sample of cattle, swine, or sheep representative of the population or production group is slaughtered to establish the relationship between weight ant] com- position (Cianzio et al., 1982; Hammack ant! Shrode, 1986~. The relationship between live weight and fatness is such that it wit! be influenced by the way animals are fed, the environment in which they are grown, en c] any subclinical disease that may alter their growth rate. Live weight and its relationship to compo- sition are also dependent on the contents of the digestive tract, which can vary from 10 to 20 percent of live weight depencling on diet. Any relationship developed to pre- dict fatness from live weight in one set of circumstances is therefore unlikely to apply with acceptable accuracy in other circum- stances. For these reasons, it is difficult to establish guidelines that can apply to the general population of domestic animals. The farmer or livestock feeder shouIcI, however, be monitoring the performance, weight, and carcass traits of the animals used for his or her production system (Kempster, 19821. If a farmer has these records, live weight can be used to predict composition. Live weight of cattle, swine, anal sheep is still the major anal, in many markets, the

LIVE ANIMAL AND CARCASS COMPOSITION only factor used in cleciding when animals wfl] be soil] for slaughter. Therefore, live- stock breeders and farmers must select for cattle, swine, and sheep that have the ge- netic traits for low fat and high muscle percentage at live weights that will provide conveniently sizecl cuts of meat for the consumer and meet the standards estab- lished by the packing industry to operate economically. VISUAL ASSESSMENT AND SUBJECTIVE EVALUATION Visual assessment and subjective evalu- ation methods are the most commonly used techniques to estimate body muscle and fat characteristics in meat-proclucing animals. A major problem with visual evaluation is distinguishing between muscling and fat- ness. Visual assessments of muscling are therefore likely to be more elective as indicators of muscle deposition within a narrow range of fatness and particularly when fat levels are low. Gregory et al. (1962, 1964) and Wilson et al. (1964) reported on the extent to which carcass traits can be predicted! from live characteristics in beef cattle. Their studies included the use of subjective techniques of appraisal in selecting breeding stock from relatively homogeneous populations under similar feeding regimes. They concluded that subjective live scores can account for only 20 to 40 percent of the variation in carcass traits and are of moderate value in ranking individual animals for selection from a breeding population. Lewis et al. (1969) iclentified the variation in carcass cutabflity and grade characteris- tics that could be accounted for by visual appraisal. After live estimates were statis- tically compared with carcass measure- ments, the coefficients of determination of either a weight-adjusted or a weight-unad- juste(1 basis indicatecl that trained personnel could account for more than half of the variation in carcass traits ant] that their estimates account for, on the average, over 26] twice the variation accounted for by un- trained personnel. Three-fourths of the var- iation in fat thickness could be detected by experienced evaluators. ULTRASONICS Recent developments in ultrasonics have led to new interest in the use of ultrasonic techniques for estimating body composition in meat-producing animals (Recio et al., 1986~. Ultrasonics is based on the principle of high-frequency sound signals passing through tissues, but that when an interface between two tissues is encountered, some sound is reflected back. A pulse generator sends electrical pulses that are converted into sound signals in the transmitter. These signals are then passed through the tissues until they are reflected at an interface. The reflected] signals are picked up by the re- ceiver and can be amplified and shown in a visual form by an oscflIoscope. Variations in the time taken for the reflected signals to return to the transmitter-receiver are used to measure variations in the distances of the boundaries between tissues. These concepts are outliner] by Miles (19781. The "A" mode ultrasonic machines dis play echo amplitude against time which is shown on the screen as peaks superimposed on a time baseline. The distance between the peaks represents the thickness of the tissues being measure:!. For"B" mocle ma- chines, the signals are shown on a cathode ray tube as a series of bright spots. The thickness of the tissues is represented by the distance between successive bright spots. Generally, these machines have a single transducer that moves across the body of the animal on a track. As the transducer moves, a picture is built up either on Polaroid film or on a cathode ray screen. An example is the Scanogram machine. Real-time machines produce a practically instantaneous picture by rapid electronic switching from element to element. The principle involved is similar to that aIreacly

262 described except that movements of the tissues can be seen because of the contin- uous nature of the picture. The Danscanner is an example. The interpretation of results for B mode machines and real-time scan- ning usually requires the tracing of depths and areas from pictures. This can now be done by using planimeters linked to micro- processors or computers (Alliston, 1983~. Of all the nondestructive evaluation tech- niques used to evaluate the body composi- tion of living animals, the new ultrasonic techniques appear to have the greatest po- tential for practical application at this time. A large cooperative research study between scientists in Denmark and the United King- dom on ultrasonic methods was reported by Andersen et al. (19821. The application of ultrasonics to the meas- urement of carcass traits of meat-producing animals was first reported in the United States by Stouffer et al. (19591. Since then, ultrasonic techniques have improved con- siderably, and Andersen et al. (1982) pro- vide an excellent comparative report on five ultrasonic machines: the Scanogram, the Danscanner, the Philips, the Ohio, and the Bruel and Kjaer. A summary follows. Ultrasonic scanning predicts body com- position with a degree of accuracy similar to that of the corresponding cut surface measurements on the carcass. The ultra- sound prediction was somewhat better than would be expected from the relationship between ultrasonic measurements and the corresponding cut surface measurements such as fat thickness over the loin. When the five ultrasonic units were eval- uated, the Bruel and Kjaer scanner was less accurate than the other four. No clear dif- ferences were detected for the other four units between the machine-operator com- binations in terms of predicting body com- position traits. Although the Philips and Ohio machines were only able to scan a section of the longissimus dorsi muscle and its overlying subcutaneous fat, they pro- vided an acceptable description of the car- anatomy and carcass composition; APPENDIX cass composition traits. Among the opera- tors, however, there was a preference for the Danscanner and Scanogram, which are specially constructed for use with farm an- imals. Because the four ultrasonic machines had similar technical capabilities, other fac- tors to consider include capital investment, ease of use, operating costs, and quality of service. A review of the recent literature (Miller et al., 1986; Recio et al., 1986) indicates that real-time ultrasound measurements ob- tained by very experienced operators can accurately predict carcass composition traits. The real-time ultrasound live measures of longissimus dorsi area, 12th rib, and shoul- der fat thickness were significantly corre- lated (P < 0.05) to the comparable carcass measurements (r2 = 0.98, 0.88, and 0.79, respectively). For predicting percentage of carcass fat, the 9th-lOth-llth rib fat per- centa~e (coefficient of determination fCD] = 85. 4), the U. S. Department of Agricul- ture (USDA) yield grade (CD = 74.0), the real-time ultrasound, the 12th rib fat thick- ness (CD = 55.8), and the carcass specific gravity (CD = 48.8) were all significantly correlated (P < 0. 05~. Adjusted fat thickness was the single most useful carcass measure- ment for predicting percentage of carcass fat (CD = 69.3). VIDEO IMAGE ANALYSIS Video image analysis (VIA) has been stud- ied as a replacement for or supplement to the subjective visual assessment of grading carcasses. The concept is based on the use of a video camera to obtain a video image (a numerical array of gray values) through an analog/digital converter. The values can then be manipulated by computer. In prac- tice, the application of VIA to carcass grad- ing is not simple. Some of the potential difficulties are as follows: 1. Development of image analysis pro- cedures with optimal correlation to carcass

LIVE ANIMAL AND CARCA S S COMPOSITION 2. Technical problems associated with daily use of electronic equipment in the harsh slaughterhouse environment; 3. Standardization ofthe video inspection process (relative position of the object, back- groun<1 contrast, lighting); and 4. Development of software with the abil- ity to handle carcasses with large variations . . In size. In the Uniter] Kingdom, the VIA principle is now commercially user] to measure visible lean in fabricated beef. In Sweden, a version of the VIA system called Electronic Scan- ning Planimetry has been developed and evaluated for prediction of carcass compo . . . . sltlon In pigs. In 1978, the U.S. Department of Agri- culture, in cooperation with the National Aeronautics en c! Space Administration and the let Propulsion Laboratory, began a project to develop an instrument for objec- tive evaluation of carcass quality ant! yield grade. Video image analysis was iclentifiec! as having the greatest potential. The best combination of VIA-measured traits for pre- dicting kilograms of lean was total lean area at the 12th rib, rib weight (kilograms), total fat area at the 12th-13th rib, and fat thickness (centimeters). This equation hac] a CD of 93.6. The results ofthe study clearly indicate that strong potential exists for VIA as a yield-grading device. Wassenberg et al. (1986) reported that VIA is as reliable as an expert three-member committee using USDA traits to evaluate the percent beef carcass yield of primal lean. In abolition, it might be less subject to human error. It also shows potential for use as a predictor of total production weight yields which could facilitate the sale of boxed beef products. This method could! be used in the U. S. commercial beef industry in the near future. WlIOLE-BODY 40K COUNTING Estimating the body composition of a living subject by whole-body 40K counting 263 is feasible because of the direct relation of potassium to lean body mass and its indirect relation to fat. Potassium is the only single element fount! in body tissue in significant amounts that has any predictive value of body composition. It is found mostly in the intracellular space, and thus, total body potassium is indicative of total body cell mass. Potassium is not found in fat in any significant amount, and therefore, if potas- sium is present in the fat-free tissue as a constant percentage, then a value for total body potassium can be converter] to a weight of lean body mass (Ward, 1968~. Approximately 0.012 percent of all natu- rally occurring potassium is made up of the radioactive isotope 40K. The rest is com- posec! of the stable isotopes 39K and 4iK. Because 40K emits gamma radiation at 1.46 MeV, the intensity of 1.46-MeV gamma emission from the body can be user! to estimate total potassium content. From these values, total belly protein and lean mass can be estimated, assuming that the mass of protein or muscle versus potassium is constant (Schmidt et al., 1974~. Lohman et al. (1968) reporter! that the standard] error for the estimation of the total mass of potassium in beef carcass was 3.4 percent ant] that the corresponding figure for carcass lean mass was 4.2 percent. Brei- clenstein et al. (1968) used 103 steers rep- resenting four breed types ant] slaughtered at four different weights to compare several alternative methods of determining carcass lean muscle mass (CLMM). Fifty-four steers were subjected] to whole-body counting after consuming a low-counting diet for a week. Constants for breec! type were included in all regressions except those using whole- body potassium and physical measurements of the live animal. The dependent variable was weight or percentage of CLMM. Results were similar to those obtained by other scientists working in the 40K area. The inclusion of either live weight or carcass weight in a regression model results in an appreciable decrease in the coefficient

264 of variation. Fat thicknesses measured at one-half and three-fourths of the medial to lateral axis of the longissimus dorsi were useful criteria, but little or no recluction in the coefficient of variation resulted from inclucling fat thickness measurer! at one- fourth of the medial to lateral axis. Inclusion of the longissimus dorsi area generally re- cluced the coefficient of variation. Carcass specific gravity was a useful indicator of CLMM, and the coefficient of variation was further reduced by including carcass weight and fat thickness. Measurement of trimmer! hindquarter was no better than other more easily acquirer! measures. Whole-body 40K counting resulted in the lowest coefficient of variation of any regression mode} except the mode! including stanclard trimmed lean, which had a coefficient of variation of only 1.4 percent (which was not reduced by including fat thicknesses). The 40K method has proved useful in research projects where the research center has a 40K Whole Body Counter, which only a few agricultural experiment stations in the Unitecl States have. Furthermore, facilities must be specially shielded from background radiation because of the low levels of 40K being measured. Uncertainties in the meas- urement of total-body potassium arise from various sources including random error due to counting statistics, instability of the counting apparatus, and variation in sensi- tivity clue to differences in body geometry and position of the animal. These factors restrict the use of 40K in the commercial inclustry. The topic of 40K estimation of bo(ly composition was reviewer] in detail in Body Composition in Animals and Man in 1968 by the National Academy of Sciences (see Reid et al., 1968; Breidenstein et al., 1968; Lohman et al., 1968; Ward, 1968~. BODY DENSITY Discovery of the principle of density is credited to Archimedes, around 200 B.C. The Archimeclean principle is based on the APPENDIX fact that a body clisplaces a volume equal to its own. Density is expressed in relation to the density of a reference standard, usually water at 20°C. In the case of gas, however, the standard is generally air. The rationale for estimating fatness or muscling from density is based on the as- sumption that the body can be considered a two-component system, with the compo- nents being of different but constant den- sities (Keys and Brozek, 19531. If this is the case and the densities of the components are known, the proportions of the two com- ponents can be estimates] from the density of the whole body. The two components in meat-producing animals are usually consi(l- ered to be the fatty tissue and the fat-free body. There is considerable evidence that the fat-free body is fairly constant in com- position in mature animals (Elsley et al., 1964; Messinger and Steele, 1949; Morales et al., 1945; Murray, 1922~. The water content in the fat-free body is not constant in young, growing animals, however; and therefore, the density measurements are not likely to be as accurate for predicting composition (Pearson et al., 1968~. The commonly reported density value for lean is 1.10; for fat it is 0.90. The major problem in determining den . . Slty IS the measurement of volume. A1- though this would appear to be a simple procedure, it is not, especially in live ani- mals. Even water displacement, which is the simplest method, has numerous pitfalls. Air displacement procedures, such as he- lium dilution, are even more complicated. Timon and Bichard (1965) reported an inverse relationship between fat and carcass specific gravity in lambs. The correlation coefficients ranged from - 0.56 to - 0. 88. Working with 83 lambs, they found that carcass specific gravity accounted for 86.1 and 78.1 percent of the respective variances in carcass fat and muscle percentages. Kraybill et al. (1951) determined the specific gravity of the beef carcass and the empty body (without blood, hide, and lungs)

LIVE ANIMAL AND CARCASS COMPOSITION of 30 beef animals. The correlation coeffi- cient between carcass and empty-bocly spe- cific gravities was 0.98. Empty-bocly specific gravity was correlated ~ - 0.95) with the percentage of carcass fat. Zinn et al. (1966) and Albin et al. (1967) found that the crude protein content of beefcattle estimated from carcass specific gravity using the relation- ship of Kraybill et al. (1951) is higher than values obtained by actual laboratory analy- sis. The accuracy of specific gravity for estimating body composition traits varies from average to good. At the 1986 meeting of the American Society of Animal Science, Miller et al. (1986) reported an average relationship (CD = 48.8) for specific gravity in predicting carcass fat. The variation in predicting carcass com- position in pork carcasses is similar to that for cattle ant! sheep. Garrett (1968) proviclec] a summary of the variation. Density techniques are slow and have no commercial use, but the specific gravity technique is still used to a limiter] degree in the research field. ELECTRONIC MEAT-MEASURING EQUIPMENT The electronic meat-measuring equip- ment (EMME) principle works on ecicly currents being induced in the animal by an alternating magnetic field produced by a current passing through a coil surrounding the animal. The eddy currents generate a magnetic field that can be picked up by a change of impedance in the coil. The con- cept is based on the method that the con- cluctivity of muscle is much higher than that of fat. Domermuth et al. (1976) investigated the predictive value of EMME for body com- nosition of pigs ant] found that EMME values in combination with fasted weight could predict carcass protein (r2 = 0.78) and lean cuts (r2 = 0.80~. Fredeen et al. (1979) fount! correlations of-0.79 between EMME values and the ,: 265 percentage of total fat and 0.79 and 0.40 for percentage of muscle for 130 and 228 pigs, respectively. The results obtaine(l with the EMME method in general were too variable to be acceptable. Each EMME machine has to be cali- brated and a formula established for the specific machine. Temperature and humid- ity can influence the results. At present, the EMME machine is not used to estimate body composition of swine on a commercial basis. Some units have been user] to esti- mate the fat content of boneless meat pack- aged in boxes for interstate shipment. ANYL-RAY The Anyl-ray technique is based on x-ray attenuation as an index of tissue fatness. It is user! on a regular basis by the commercial meat industry in the United States to de- termine the fat content in ground meat used for meat processing. The method is fast, requires a small sample (2 to 3 kg), and has a high degree of accuracy. TISSUE SAWDUST TECHNIQUE Vance et al. (1970) reported a correlation (P < 0.01) between the chemical compo- nents of beef carcass sicles an/1 the meat sawdust from sawing through the frozen round, loin, rib, ant! chuck at 2.54-cm intervals. This technique was further eval- uate(l by Williams et al. (1974), who eval- uated 20 bull carcasses averaging 282 kg. Correlations between the chemical com- position ofthe carcass an(l the tissue sawdust were 0.82, 0.94, 0.64, and 0.68 for moisture, fat, protein, and ash, respectively. Also, 12 carcass sides from six Holstein calves aver- aging 138 kg were used to evaluate storage methods (chilling versus freezing of the carcass) and two types of sawdust (cross- section every 2.54 cm versus retail cut) for estimating carcass composition. Chilled car- casses yiel(led only 23 percent as much sawdust as frozen carcasses. The reliability

266 of the sawdust procedure for predicting carcass composition was greatest from frozen carcasses sawer] every 2.54 cm, followed by the meat sawdust from cutting frozen car- casses into retail cuts. Chemical composition of meat sawdust can provide a good estimate of the chemical composition of cattle when samples are collected from frozen carcasses. The method is simple but time-consuming. Slight devaluation of the carcass occurs. This method has value as a research tool, but is not feasible for commercial application (Williams et al., 1974~. DILUTION TECHNIQUES The dilution technique involves the in- troduction of a known amount of tracer, which will become uniformly distributed throughout a compartment in the animal body (Cuthbertson, 1975; Odwongo et al., 1985~. A sample of the compartment is then taken ant! the concentration of the tracer measured. A tracer should not be toxic, must be metabolized, should be easily meas- urable, and mu/st cliffuse homogeneously into all the volume to be measurer] (Robelin, 1982~. Some of the different tracers used in animals include antipyrine ant! N-acetyl- 1,4-aminoantipyrine, urea, tritiated water (TOH), and deuterium oxide (D2O). Deu- terium oxide and urea appear to be the most suitable tracers because they are more accurate and not radioactive. Results from research with antipyrine generally demon- strate that antipyrine dilution is too variable to give good estimation of total belly water (Panaretto and Till, 1963~. UREA DILUTION Urea dilution has been used to estimate body composition in cattle (Bartle et al., 1983; Kock and Preston, 1979; Meissner et al., 1980; Preston and Kock, 1973) and in lambs (Bartle et al., 1985~. Because urea is inexpensive and the technical requirements of plasma urea N analysis are minimal, the APPENDIX urea dilution technique could be used for both research and industrial purposes where measurement of body composition during growth is necessary. The correlation be- tween urea space and fatness range(l from 0.71 to 0.82. The technique works best with heavy cattle having a relatively large degree of fatness, compared to lighter cattle with a small degree of fatness. Rule et al. (1986) reported that some equations developed for urea dilution estimates of body water over- estimated empty holly water in 6-month- old steers by.7.59 percent, but that for 12- and 18-month-old steers the calculated and percentage empty body water did not diner (P > 0.05~. Bartle and Preston (1986) further evalu- ate(1 the amount of urea diEusing from the blood into the rumen and urine of cattle after urea infusion ant! found that urea did not diffuse into the reticulo-ruminal water. They indicated that the urea dilution method overestimates empty body water by the urine volume produced in the 12-minute collection period. To ensure more accurate data, it is sug- gestec] that before using any prediction equation to calculate body composition by dilution techniques, the equation should be tested with a subsample of cattle from the population for which its use is intended (Rule et al., 1986~. Cuthbertson (1975) and Robelin (1982) examined the problems associated with the estimation of body water in ruminants caused by the variation in the water content of the alimentary tract. Robelin used D2O to study 340 beef cattle for whole-body composition and found that the weight of water and protein was fairly closely related to fat-free mass. He reported that for a single beef animal, an accuracy of 13 percent for lipid and 7 percent for protein deposition is obtained for a total body weight gain of 300 kg. Foot and Greenhalgh (1970) used the D2O procedure to estimate body fat content in sheep. The values they obtained differed

LIVE ANIMAL AND CARCASS COMPOSITION from those obtained by analysis of the slaughtered animals by 0.8 to 1.7 kg in seven ewes containing 5.2 to 21.4 kg of fat. The standard! deviation was + 1.2 percent- age units. The D2O dilution method has no com- mercial application and is used on a limited basis as a research tool to estimate body composition. It is relatively simple for sci- entists to use but too complex for industrial application. It is a good way to estimate total body water but is limited in its level of accuracy for total body fat. Application of a kinetic technique to solve an anatomic problem with no clocumenta- tion of the congruity of the kinetic and anatomic models is clearly limited. Simu- lation analysis indicates that the kinetic mode! is very insensitive to changes in anatomic pool sizes, but very sensitive to changes in exchange rates of water among pools (R. W. Russell and R. B. Reed, personal communication, 1986~. COMPUTERIZED TOMOGRAPHY The Nobel Prize was awarder! to A. M. Cormack and G. N. Houndsfield for the development of the computerized tomog- raphy (CT) technique. The concept is based on presentation of anatomic areas of the body by computer] synthesis of an image from x-ray transmission data obtained in many different directions through the plane under consideration (Cormack, 1980; Houndsfield, 1980~. An x-ray tube rotates around] an object, and the computer recon- structs from a series of pictures a slide through the object. By this technique, the density (CT number) of different body tis- sues at different distances from the x-ray tube can be calculated. One of the first applications of this tech- nique for estimating composition of meat- producing animals was reported by Skjer- vold (1982) from the Agricultural University of Norway. Their study of 23 pigs indicated that it was possible to obtain a good predic 267 lion of the body composition on the basis of the relative CT distribution from one tomographic plane. Skjervold also reported the CT numbers of different body tissues. Lung tissue had values of-200 to -100; fat tissue, - 100 to 0; muscle tissue, +30 to + 100; and bone, + 400 to + 500. Allen and Vangen (1984) used comput- erizec! tomography to estimate the body composition of 207 pigs ranging in weight from 59 to 120 kg. The values they obtained are similar to those reported by Skjervold (1982). European researchers have been active in evaluating CT for use in estimating body composition of meat-proclucing animals, but only limited research is being clone in the United States. Researchers at the Meat and Animal Science Department of the Univer- sity of Wisconsin are cooperating with med- ical college faculty ant! are currently col- lecting data from pigs. The main drawbacks to computerized tomography are expense, the time required to obtain an estimate, and the necessity to anesthetize the animal before scanning. Even with these limitations, however, improved techniques are expected that will make computerized tomography acceptable for scanning animals for genetic selection of breeding stock. This method therefore has great potential for future use in the livestock industry. NUCLEAR MAGNETIC RESONANCE IMAGING The nuclear magnetic resonance (NMR) method for estimating belly composition is based on a strong static magnetic field and pulsed radio waves that in(luce resonance of protons in the measured body. The signals emitted are a reaction of the body to the high-frequency disturbance. Therefore, they are a product of the matter itself, with intensities (lepending on the proton spin densities and molecular structures. The NMR signal does not continue indefinitely. En

268 vironmental influences cause the individual flipped magnetic moments to get out of phase and return to the orientation they had before the radio frequency pulse was applied. The time requires] to reestablish original conditions has been defined as spin lattice relaxation time T1 ant] spin-spin relaxation time T2. Procedures to determine T1 are known as inversion recovery, and for T2 as spin-echo methods. Both systems produce a data matrix of the size 12~128 or 25~256 that contains in x-ray CT the normalizer] Houncisfield units ranging from -1,000 (air) to more than 1,400 (compact bone). There are several ways to produce images. On a cievice with seven colors, the total data space in a matrix is subdivided into seven regions, with each region rep resenting a different color. Fat, muscle, bone, and connective tissue are always pre sented if the total data space is mapped onto seven colors (Groenevelc] et al., 1984~. Fuller et al. (1984) user] the Aberdeen SOLUBLE SHORT-LIVED r RADIOACTIVE GAS TRACERS NMR Imaging machine to evaluate pigs for body composition. Only three pigs were evaluated. Images were obtained at nine sites along the body, three each of the shoulder, midback, ant! rump. Good images reportedly were obtained of the muscle, fat, ant! bone portions of the sites scanned. Nuclear magnetic resonance imaging has great potential, but very limited data are available on its usefulness for predicting body composition traits of meat-producing animals. The equipment is very expensive, and the method is very complex; its future will depend on the amount of resources available for its development as an agricul tural tool. APPENDIX the NIR method. Reaclings were taken at specific sites on the ham, shoulder, and side of the pig. Carcass composition was deter- mined by analysis ofthe soft tissue dissected from the eviscerates] carcass for lipid, pro- tein, and water content. Multiterm regres- sion correlations were generates] for carcass fat as a percentage of live body weight. For the carcass, percent fat correlated best with NIR readings taken on the ham. The meas- urements taken from the carcass accounted for about 50 to 60 percent of the variation. The values for the live pig were lower. These relationships indicate that with re- finement in instrumentation and technique. this method may be useful in predicting body composition. It is simple, ant] the equipment is not Drohibitivelv expensive. More research is needed, however, before NIR can be consiclered for commercial use. , . , ,¢ . ~. ~ NEAR-INFRARED REFLECTANCE Near-infrared reflectance (NIR) is widely used to predict the composition of various plant materials and may have potential ap- plication for estimating carcass composition. Mitchell et al. (1986) used 20 pigs for each weight group of SO, 60, and 90 kg to evaluate A range of halogenated gases with a par- ticular affinity for adipose tissue could be considered for predictors of body composi- tion. The commonly used anesthetic halo- thane (2-bromo-2-chloro-1, 1,1-trifluoro- ethans) is an example. The label can be 11C, OF, 77Br, or 38C1. This idea was reported by Ettinger et al. (1984), who suggested that an animal can be given labeled halo- genated gases in concentrations small enough to have no noticeable anesthetic effect but large enough that the gases are taken up by the adipose tissue. The amount taken up could be measured by a conventional whole- body counter or a whole-body scanner. The hypothesis has not been tested, but a good theoretical basis exists for the concept. SUMMARY OF THE PRACTICABILITY/ COST-BENEFIT COMPARISON OF BODY COMPOSITION MEASURES More than 30 techniques for estimating live animal or carcass composition were

LIVE ANIMAL AND CARCASS COMPOSITION reported in this review. The cost of the equipment to measure belly composition can range from 1 dollar to over 1 million dollars. Accuracy, precision, ant] practicality are also considerations. Many promising techniques have been rejected for commer- cial use, not because of costs but because of practicability. One of the least costly techniques avail- able for estimating fat thickness in cattle and swine is the ruler back fat probe. Its accuracy is as high as the best ultrasonic techniques ant] almost as high as the com- puterized tomography methods recently cle- veloped for meat-proclucing animals. The cost of a ruler probe can range from 1 to 50 dollars, but still, the device is not used extensively in the meat industry because personnel are concerned about its practi- cality. (This concern is not really valid; a trained person can probe an incliviclual pig or steer in less than a minute when the animal is restrainecI.) The scientific community must under- stanc] that most producers and buyers of livestock in the United States prefer the use of live weight ant! visual assessment methods for estimating belly composition because of their practicality, low cost, and rapidity in making the measurements. This must reflect the limited interest of the U. S. livestock industry in reducing fat in meat- proclucing animals by objective methods. One reason for this is the small margin paid by the packing industry for trim, well- musclecT animals versus fat, less muscular ones. We need] an improved marketing program that will pay farmers for producing trim, muscular animals. A system of this type will encourage the use of more objec- tive methods for the selection of breeding animals and the marketing of animals for meat production. From the research standpoint, many tech- niques are available to estimate body com- position, but their accuracies are not out- stancling. Most can account for 60 to 80 percent of the variation in muscle, fat, or 269 bone of the carcass. Thus, more accurate methods are needed for researchers working in the body composition field. Based on recent literature, it may be possible to improve accuracy with such new methods as computerized tomography and nuclear magnetic resonance imaging. The cost of the equipment currently prevents their widespread use, but with further research on new methods, we may, in the near future, develop the ultimate technique- one that is cost-effective, simple, and accurate. REFERENCES Albin, R. C., D. W. Zinn, S. E. Curl, and G. H. Tatsch. 1967. Growth and fattening of the bovine. III. Effect of energy intake upon carcass composition. J. Anim. Sci. 26:209. Allen, P., and O. Vangen. 1984. x-ray tomography of pigs some preliminary results. P. 52. in In Vivo Measurements of Body Composition in Meat Ani- mals, D. Lister, ed. London and New York: Elsevier Applied Science Publications. Alliston, J. C. 1983. Evaluation of carcass quality in live animals. P. 79 in Sheep Production, W. Hare- sign, ed. Boston: Butterworth. Andersen, B. B., H. Busk, J. P. Chadwick, A. Cuth- ertson, G. A. J. Fursey, D. W. Jones, P. Lewin, C. A. Miles, and M. G. Owen. 1982. CEC supported ultrasonic trial in U.K. and Denmark. Pp. 13~1 in In Vivo Estimation of Body Composition in Beef, CEC Workshop Report, B. B. Andersen, ed. Co- penhagen: National Institute of Animal Sciences. Bartle, S. J., and R. L. Preston. 1986. Plasma, rumen and urine pools in urea dilution determination of body composition in cattle. J. Anim. Sci. 63:77. Bartle, S. J., J. R. Males, and R. L. Preston. 1983. Evaluation of urea dilution as an estimator of body composition in mature cows. J. Anim. Sci. 56:410. Bartle, S. J., R. L. Preston, M. A. McCann, and F. B. Craddock. 1985. Evaluation of urea dilution as an estimator of body composition in finishing lambs. J. Anim. Sci. 61(Suppl. 1~:265. Barton, P. 1983. Quality traits of pork carcasses. P. 15 in Annual Report of the Danish Meat Research Institute. Copenhagen: Danish Meat Research In- stitute. Berg, R. T., and R. M. Butterfield. 1976. New Con- cepts of Cattle Growth. New York: John Wiley & Sons. Breidenstein, B. C., T. G. Lohman, and H. W. Norton. 1968. Comparison of potassium-40 method with other methods of determining carcass lean muscle _ _ 7 ~_ 7 ~_ _ _ 7 ~

270 mass in steers. P. 393 in Body Composition in Animals and Man. Washington, D. C.: National Academy of Sciences. Brozek, J. 1961. Body measurements including skinfold thickness as indicators of body composition. P. 3 in Techniques for Measuring Body Composition. Washington, D.C.: National Academy of Sciences. Busch, D. A., C. A. Dinkel, and J. A. Minyard. 1969. Body measurements, subjective scores and estimates of certain carcass traits as predictors of edible portion in beef cattle. J. Anim. Sci. 29:557. Cianzio, D. S., D. G. Topel, G. B. Whitehurst, D. C. Beitz, and H. L. Self. 1982. Adipose tissue growth in cattle representing two frame sizes: Dis- tribution among depots. J. Anim. Sci. 55:305. Cook, A. C., M. L. Kohli, and W. M. Dawson. 1951. Relationship of five body measurements to slaughter grade, carcass grade and dressing percentage in milking Shorthorn steers. J. Anim. Sci. 10:386. Cormack, A. M. 1980. A presentation of anatomical information by computed synthesis of an image from x-ray transmission data obtained in many different directions through the plan under consideration. Nobel Lecture. J. Computer Assisted Tomography 4:658. Cross, H. R. 1982. In viva and in vitro measurements of composition. Proc. Recip. Meat Conf. 35:1. Cuthbertson, A. 1975. Carcass quality. P. 147 in Meat, D. J. A. Cole and R. A. Lawrie, eds. London: Butterworth. Desmoulin, B. 1984. Pig carcass evaluation by linear measurement and the fat-o-meater (reflectance probe). P. 167 in In Vivo Measurements of Body Compo- sition in Meat Animals, D. Lister, ed. London and New York: Elsevier Applied Science Publications. Domermuth, W., T. L. Veum, M. A. Alexander, H. B. lIedrick, J. Clark, and D. Eklund. 1976. Predic- tion of mean body composition of live market weight swine by indirect methods. J. Anim. Sci. 43:966. Doornenbal, H. 1968. Relationship to body composi- tion of subcutaneous backfat, blood volume, and total red-cell mass. Pp. 218-230 in Body Composition in Animals and Man. Washington, D.C.: National Academy of Sciences. Elsley, F. W. H., I. McDonald, and V. R. Fowler. 1964. The effect of plane of nutrition on the carcass of pigs and lambs when variations in fat content are excluded. Anim. Prod. 6:141. Ettinger, K. V., M. A. Foster, and U. J. Miola. 1984. Future developments in the in vivo measurements of body composition of pigs. P. 207 in In Vivo Measurements of Body Composition in Meat Ani- mals, D. Lister, ed. London and New York: Elsevier Applied Science Publications. Fahey, T. J., D. M. Schaefer, R. G. Kauffman, B. J. Epley, P. F. Gould, J. R. Romans, G. C. Smith, and D. G. Topel. 1977. A comparison of practical APPENDIX methods to estimate pork carcass composition. J. Anim. Sci. 44:8. Foot, J. Z., and J. F. D. Greenhalgh. 1970. The use of deuterium oxide space to determine the amount of body fat in pregnant blackface ewes. Br. J. Nutr. 24:815. Fredeen, H. T., A. lI. Martin, and A. P. Sather. 1979. Evaluation of an electronic technique for measuring lean content of the live pig. J. Anim. Sci. 48:536. Fuller, M. F., M. A. Foster, and J. M. S. Hutchison. 1984. Nuclear magnetic resonance imaging of pigs. P. 123 in In Vivo Measurements of Body Compo- sition in Meat Animals, D. Lister, ed. London and New York: Elsevier Applied Science Publications. Garrett, W. N. 1968. Expenses in the use of body density as an estimator of body composition of animals. P. 170 in Body Composition in Animals and Man. Washington, D.C.: National Academy of r~ . ~clences. Green, W. W., W. R. Stevens, and M. B. Gauch. 1969. Use of body measurements to predict the weights of wholesale cuts of beef carcasses: Whole- sale round of 900 pound steers. Pp. 1-18 in Agri- cultural Experiment Station Bulletin A-165. College Park: University of Maryland. Gregory, K. E., L. A. Swiger, V. H. Arthaud, R. B. Warren, D. K. Hallet, and R. M. Koch. 1962. Relationships among certain live and carcass char- acteristics of beef cattle. J. Anim. Sci. 21:720. Gregory, K. E., L. A. Swiger, B. C. Breidenstein, V. H. Arthaud, R. B. Warren, and R. M. Koch. 1964. Subjective live appraisal of beef carcass traits. J. Anim. Sci. 23:1176. Groeneveld, E., E. Kollweitt, M. lIenning, and A. Pfau. 1984. Evaluation of body composition of live animals by x-ray and nuclear magnetic resonance computed tomography. P. 52 in In Vivo Measure- ments of Body Composition in Meat Animals, D. Lister, ed. London and New York: Elsevier Applied Science Publications. Hammack, S. P., and R. R. Shrode. 1986. Calfl~ood weights, body measurements and measures offatness vs. criteria of overall size and shape for predicting yearling performance in beef cattle. J. Anim. Sci. 63:447. Hazel, L. N., and E. A. Kline. 1952. Mechanical measurement of fatness and carcass value of live hogs. J. Anim. Sci. 11:313. Hirzel, R. 1939. Factors affecting quality in mutton and beef with special reference to the proportions of muscle, fat and bone. Onderstepoort J. Vet. Sci. 12:379. Houndsfield, G. M. 1980. Computer medical imaging. Nobel Lecture. J. Computer Assisted Tomography 4:665. Kauffman, R. G. 1971. Variation in gross composition of meat animals. Proc. Recip. Meat Conf. 24:292.

LIVE ANIMAL AND CARCASS COMPOSITION Kempster, A. J. 1982. Management and selection of cattle for slaughter. Pp. 127-133 in In Vivo Esti- mation of Body Composition in Beef, CEC Workshop Report, B. B. Andersen, ed. Copenhagen: National Institute of Animal Sciences. Kempster, A. J. 1986. Correlations between indirect and direct measurements of body composition. Proc. Nutr. Soc. 45:55. Keys, A., and J. Brozek. 1953. Body fat in adult man. Physiol. Rev. 33:245. Kidwell, J. F. 1955. A study of the relationship between body composition and carcass quality in fat calves. J. Anim. Sci. 14:233. Kook, S. W., and R. L. Preston. 1979. Estimation of bovine carcass composition by urea dilution tech- nique. J. Anim. Sci. 48:319. Kraybill, H. F., H. L. Bitter, and O. G. Hankins. 1951. Body composition of cattle. II. Determination of fat and water content from measurement of body specific gravity. J. Appl. Physiol. 4:575. Lewis, T. R., G. G. Suess, and R. G. Kauffman. 1969. Estimation of carcass traits by visual appraisal of market livestock. J. Anim. Sci. 28:601. Lohman, T. G., W. J. Coffman, A. R. Twardock, B. C. Breidenstein, and H. W. Norton. 1968. Factors affecting potassium-40 measurement in the whole body and body components. P. 291 in Body Com- position in Animals and Man. Washington, D.C.: National Academy of Sciences. McMeekan, C. P. 1941. Growth and development in the pig with special references to carcass quality characteristics. J. Agric. Sci. 31:1. Meissner, H. H., J. H. van Staden, and E. Pretorius. 1980. In vivo estimation of body composition in cattle with tritium and urea dilution. I. Accuracy of prediction equations for the whole body. South African J. Anim. Sci. 10: 165. Messinger, W. J., and J. M. Steele. 1949. Relationship of body specific gravity to body fat and water content. Proc. Soc. Exp. Biol. Med. 70:316. Miles, C. A. 1978. A note on recent advances in ultrasonic scanning of animals. P. W133 in Proceed- ings of the 24th European Meat Research Workers Conference. Kulmbach, West Germany: European Meat Research Workers. Miller, M. F., H. R. Cross, G. C. Smith, J. F. Baker, F. M. Byers, and H. A. Recio. 1986. Evaluation of live and carcass techniques for predicting beefcarcass composition. J. Anim. Sci. 63(Suppl. 1):234 (Abstr.). Mitchell, A. D., K. H. Norris, H. H. Klueter, N. C. Steele, and M. B. Soloman. 1986. Estimation of live body and carcass composition of pigs by near-infrared reflectance. J. Anim. Sci. 63(Suppl. 1):234 (Abstr.). Morales, M. F., E. N. Rathbun, R. E. Smith, and N. Pace. 1945. Studies on body composition. II. Theo- retical consideration regarding the major body tissue 271 components with suggestions for application to men. J. Biol. Chem. 158:677. Murray, J. A. 1922. The chemical composition of animal bodies. J. Agric. Sci. 12:103. Odwongo, W. O., H. R. Conrad, A. E. Staubus, and J. H. Harrison. 1985. Measurement of body water kinetics with deuterium oxide in lactating dairy cows. J. Dairy Sci. 68:1155. Palsson, H. 1939. Meat quality in sheep with special reference to Scottish breeds and crosses. J. Agric. Sci. 29:544. Panaretto, B. A., and A. R. Till. 1963. Body compo- sition in vivo. II. The composition of mature goats and its relationship to antipyrine, tritiated water and N-acetyl-4-amine antipyrine spaces. Aust. J. Agric. Res. 14:926. Pearson, A. M., R. W. Purchas, and E. P. Reineke. 1968. Theory and potential usefulness of body den- sity as a predictor of body composition. P. 153 in Body Composition in Animals and Man. Washington, D.C.: National Academy of Sciences. Preston, R. L., and S. W. Kock. 1973. In vivo prediction of body composition in cattle from urea space measurements. Proc. Soc. Exp. Biol. Med. 143:1057. Recio, H. A., J. W. Savell, H. R. Cross, and J. M. Harris. 1986. Use of real-time ultrasound for pre- dictingbeefcutability. J. Anim. Sci. 63(Suppl. 1):260 (Abstr. ). Reid, J. T., A. Bensadaum, L. S. Bull, J. H. Burton, P. A. Gleeson, I. K. Han, Y. D. Joo, D. G. Johnson, W. R. McManus, O. L. Paladines, J. W. Straud, H. F. Tyrrell, B. D. H. Van Nickerk, and G. W. Wellington. 1968. Some peculiarities in the body composition of animals. P. 19 in Body Composition in Animals and Man. Washington, D.C.: National Academy of Sciences. Robelin, J. 1982. Measurement of body water in cattle by dilution technique. P. 107 in In Vivo Estimation of Body Composition in Beef, CEC Workshop Re- port, B. B. Andersen, ed. Copenhagen: National Institute of Animal Sciences. Rouse, G. H., D. G. Topel, R. L. Vetter, R. E. Rust, and T. W. Wickersham. 1970. Carcass composition of lambs at different stages of development. J. Anim. Sci. 31:846. Rule, D. C., R. N. Arnold, E. J. Hentges, and D. C. Beitz. 1986. Evaluation of urea dilution as a tech- nique for estimating body composition in beef steers in vivo: Validation of published equations and com- parison with chemical composition. J. Anim. Sci. 63:1935. Schmidt, M. K., J. L. Clark, T. L. Veum, and G. E. Krause. 1974. Prediction of composition of crossbred swine from birth to 136 kg live weight via the liquid scintillation whole body counter. J. Anim. Sci. 39:855.

272 Skjervold, H. 1982. Estimation of body composition in live animals by the use of computerized tomog- raphy. P. 148 in In Vivo Estimates of Body Com- position in Beef, CEC Workshop Report, B. B. Andersen, ed. Copenhagen: National Institute of Animal Sciences. Stouffer, J. R., M. V. Vallentine, and G. H. Wellington. 1959. Ultrasonic measurements of fat thickness and loin eye area on live cattle and hogs. J. Anim. Sci. 18:1483. Timon, V. M., and M. Bichard. 1965. Quantitative estimates of lamb carcass composition. Anim. Prod. 7:183. Topel, D. G. 1986. Future meat animal composition: Industry adaptation of new technologies. J. Anim. Sci. 63:633. Vance, R. D., H. W. Ockerman, V. R. Cahill, and R. F. Plimpton, Jr. 1970. Carcass composition as related to meat sawdust and analysis. J. Anim. Sci. 31:192 (Abstr.~. Ward, G. M. 1968. Introduction to whole-body count- ing. P. 263 in Body Composition in Animals and APPENDIX Man. Washington, D.C.: National Academy of Sci- ences. Wassenberg, R. L., D. M. Allen, and K. E. Kemp. 1986. Video image analysis prediction of total kilo- grams and percent primal lean and fat yield in beef carcasses. J. Anim. Sci. 62:1609. Webster, H. J. F. 1986. Factors affecting the body composition of growing and adult animals. Proc. Nutr. Soc. 45:45. Williams, D. B., D. G. Topel, and R. L. Vetter. 1974. Evaluation of a tissue-sawdust technique for pre- dicting beef carcass composition. J. Anim. Sci. 39:849. Wilson, L. L., C. A. Dinkel, H. J. Turna, and J. A. Minyard. 1964. Live animal prediction of curability and other beef carcass characteristics by several judges. J. Anim. Sci. 23:1102. Zinn, D. W., R. C. Albin, S. E. Curl, and C. T. Gaskins. 1966. Growth and fattening of the bovine. II. Postweaning protein and gross energy composi- tion. Proc. W. Sec. Am. Soc. Anim. Sci. 17:151. Zobrisky, S. E. 1963. Status of methods in pork carcass evaluation. Proc. Recip. Meat Conf. 16:266.

Altering Carcass Measurements and Composition of the Pig V. C. SPEER GENETICS AND SELECTION The pig and fat are closely linked in the mind of the consumer, much to the detri- ment of the pig. But in reality the amount ant] type of fat in a pig carcass is quite similar to that in other red meat animals. A dramatic change in carcass measurements and composition has come about with the development of the modern lean-type pig. The lean-type pig utilizes and deposits pro- tein more efficiently than a fat-type pig, yielding a carcass with more lean tissue. This change came about through selection (genetics) during the period 195S to 1970. In the late 1960s, the incidence of sudden pig death, or Porcine Stress Syndrome (PSS), became an acute problem among heavily muscles! pigs developed through genetic improvement (Cassens et al., 1972~. The improvement in muscling since 1970 has cleclinec3 for the barrows submitted to the Iowa Swine Testing Station (Evans, 1986), largely because of the association of PSS with heavy muscling. SEX At typical slaughter weights for pigs, the intact male (boar) yields a carcass with the least fat and most lean, followed by the female (gilt). The castrated male (barrow) yields a carcass with the most fat ant] least lean. During the growth phase and until male aggressiveness (ranting) develops, the boar will gain weight the most rapidly and most efficiently. There is the potential prob- lem of strong odor or flavor in the meat from boars slaughterer] at typical market weights in the Uniter! States. Carcasses from boars are readily accepted! in some other countries (for example, Australia and Eng- land) but are slaughtered at light weights to reduce the possibility of boar taint in the carcass. W EIGHT Beyonc! a live weight of about 90 kg, the rate of lean tissue deposition reaches a plateau and, in many pigs, actually declines as fat deposition increases. Furthermore, (laity gain seems to decline slightly, although the daily feed requirement increases. NUTRITION Protein Ashton et al. (1955) and Jensen et al. (1955) reporter] that an increase in protein 273

274 levels in corn-soybean meal diets produced only minimal carcass responses in fat-type pigs. Genetically improved pigs fed similar corn-soybean meal diets 10 years later (Iohn son, 1965) were more responsive to protein level: An increase in dietary protein level yielded a greater reduction in back fat depth and a greater increase in ham and loin percent. Responsiveness to protein level as related to type (fat versus lean) is evident in the U.S. Department of Agriculture's selection study, reported by Davey and Morgan (1969~. Amino Acids In the typical protein level study, the levels and ratios of the essential amino acids change, so it is difficult to determine whether responses are due to protein level or to one or more amino acids. In the typical corn soybean meal diet, lysine is the first limiting amino acid as protein level is reduced. Carcass measurements improved in re sponse to increases in lysine levels when all other diet components were held constant (Asche et al., 1985~. Energy Diet density (energy level) will affect carcass measurements of pigs that are fed corn. act libitum. Feecling pigs a diet with acldecl fat (3,600 keel of metabolizable energy tME]/ kg) versus a corn-soybean meal diet (3,100 keel of ME/kg) reduced the feed require ment but increased the back fat measure ment (Wagner et al., 1963~. Calorie/Protein Ratio Increasing the energy content of a corn soybean meal diet by adding fat may depress daily gain, and feet! efficiency may not improve as much as expected. Carcass meas urements are also adversely affected. To counteract these adverse performance and APPENDIX carcass criteria, the diet can be formulated to contain a constant calorie/protein ratio. Daily gain and feed efficiency were shown by Allee et al. (1976) to improve markedly when the protein level in the diet was adjuster! proportionately to the energy level. Carcass back fat, however, increased com- parecl with those pigs given a control diet (1.35 versus 1.22 inches). Generally, diets are formulated to constant calorie/protein ratios using metabolizable energy values for the ingredients. Because fat has a propor- tionately lower heat increment than normal energy sources such as grain, its energy value is underestimated. Perhaps if diets were formulated to contain constant calorie/ protein ratios using net energy values for the ingredients when fat is included in the flirt, the adverse effects on carcass meas- urements would be corrected. Grain Source The two most commonly fed grain sources for pigs are corn and barley. Corn is better than barley in terms of performance criteria, but barley is superior to corn with regard to carcass measurements (Greer et al., 1965~. Much, if not all, of the positive carcass response to barley is related to its lower energy composition compared with that of RESTRICTED FEED INTAKE Reducing the feed intake of growing- finishing pigs will improve carcass measure- meets (Braude, 1972; Greer et al., 1965; Speer, 1966~. Restricted or controlled feed- ing is commonly practiced in pig production in Europe, but because daily gain is reduced it has not been adopted by U. S. producers. The improved feed efficiency reported by Braude (1972) in response to restricted feed- ing compared to ad libitum feeding was not evident from the studies of Speer (1966) and Greer et al. (1965~.

ALTERING CARCASS MEASUREMENTS TEMPERATURE At environmental temperatures higher than ideal, the pig reduces feed intake and expends energy in an attempt to stay cool. The result is an adverse effect on production criteria but an improvement in carcass meas- urements (Stahly and Cromwell, 1979~. At lower environmental temperatures, the pig increases feed intake and once again ex- pends energy to maintain body tempera- ture. With respect to carcass measurements, the increased energy expenditure to main- tain body temperature may counteract the effect of increased feed intake. HORMONES AND RELATED COMPOUNDS Diethyistilbestrol Plimpton and Teague (1972) implanted diethylstilbestrol in boars weighing 70 kg and then slaughtered them at about 110 kg live weight. This procedure retained the positive carcass attributes of the young boar, while the effects of objectionable odor and flavor of boar meat were reduced. Diethylstilbestrol and Methy~testosterone A combination of diethyistilbestro! (2.2 mg/kg of diet) and methy~testosterone (2.2 mg/kg of diet) added to the feed improved the feed efficiency and carcass measure- ments of growing-finishing pigs (Baker et al., 1967~. This product was never approved by the Food and Drug Administration for use in the United States, but it was approved and marketed in Great Britain. Epinephrine and Epinephrine-Like Stimulators Cunningham et al. (1963) user] epineph- rine to increase fat mobilization, lipolysis, 275 and nitrogen deposition in the pig. Results were encouraging, but the required daily injection was a distinct disadvantage. In subsequent studies, Cunningham and Friend (1964) and Cunningham (1968) added ni- cotine or caffeine to the feed in an attempt to stimulate epinephrine-like responses. Both compounds seemed to improve carcass measurements of growing-finishing pigs. Similarly, the action of the beta-adrenergic agonists clenbuterol and cimaterol Jones et al., 1985; Moser et al., 1984) improved carcass measurements in growing-finishing pigs. The beta-adrenergic agonists are orally active, making them easier to use than eplnep. crane. Growth Hormone Daily injections of porcine growth hor- mone by Machlin (1972) have been shown to improve daily gain, feed efficiency, and carcass measurements. Chung et al. (1985) used a porcine preparation that was more highly purified than Machlin's and found similar responses in growing pigs, but at a much lower dosage rate. Bacterially synthe- sized human growth hormone is also active in stimulating growth rate and carcass im- provement (Baile et al., 1983). Both the natural and bacterially synthesized hor- mones must be administered by daily in- jections, which is a distinct disadvantage for IMMUNOLOGY Immunization of growing boars against androstene steroids the compounds re- sponsible for boar taint and odor-controls these undesirable characteristics without significantly affecting other characteristics such as weight gain and feed efficiency (Brooks et al., 1986; Williamson et al., 1985~. A reduction in an(lrostene steroids might also be attained through selection, since Booth et al. (1986) detected positive cor

276 relations between the bulbourethral and submaxillary gland weights and concentra- tions of 3-cx-androstenol and 5-~-androsten- one in market weight boars. Encouraging results with immunology have been obtained in lambs by autoimmunizing the lambs against somatostatin (Spencer and Garssen, 1983~. Somatomedin concentra- tion increased (nonsignificant) and growth rate improved compared with control lambs. A similar approach has been reported by Flint and Futter (1986), in which rats im- munized against their fat cells were found upon postmortem examination to have about 30 percent less carcass fat than untreated rats. TISSUE COMPOSITION The type of dietary fat fed to the pig will influence the fat composition of the carcass. The percentage of unsaturated fat in back fat samples reflects the type of oil fed (Ellis and Isbell, 1926~. Changing carcass fat com- position can be accomplished more readily in the pig than in any other large farm animal. The amount and type of fat found in the lean tissue of the pig longissimus dorsi muscle will respond to differences in diet and management. Restricted feeding re- duces the fat content and the level of un- saturation in the muscle, as does feeding barley instead of corn (Greer et al., 1965~. And increasing the protein level or reducing the energy concentration of the diet reduces the fat content of the longissimus dorsi lean tissue (Wagner et al., 1963~. From these examples, it seems that carcass back fat and the fat content of lean tissue are positively correlated. If this is true, then as producers in the United States strive for leaner ani- mals, they could encounter some of the problems that have surfaced in England. According to a technical report of the Meat and Livestock Commission of the United Kingdom (Phelps, 1985), the marked re- duction in back fat that has occurred in APPENDIX England's pig population has been accom- panied by an increase in retailer and con sumer complaints that the very lean car casses produce meat that looks unattractive, lacks succulence and flavor, and has a ten dency to be tough. REFERENCES Allee, G. L., B. A. Koch, and R. H. Hines. 1976. Effect of fat level and calorie:protein ratio on per- formance of finishing pigs. J. Anim. Sci. 42:1349. Asche, G. L., A. J. Lewis, E. R. Peo, Jr., and J. D. Crenshaw. 1985. The nutritional value of normal high lysine corns for weanling and growing-finishing swine fed at four lysine levels. J. Anim. Sci. 60:1412. Ashton, G. C., J. Kastelic, D. C. Acker, A. H. Jensen, H. M. Maddock, E. A. Kline, and D. V. Catron. 1955. Different protein levels with and without antibiotics for growing-finishing swine: Effect on carcass leanness. J. Anim. Sci. 14:82. Baile, C. A., M. A. Della-Fera, and C. L. McLaughlin. 1983. Performance and carcass quality of swine injected daily with bacterially synthesized human growth hormone. Growth 17:225. Baker, D. H., C. E. Jordan, W. P. Waitt, and D. W. Gouwens. 1967. Effect of a combination of diethyl- stilbestrol and methyltestosterone, sex and dietary protein level on performance and carcass character- istics of finishing swine. J. Anim. Sci. 26:1059. Booth, W. D., E. D. Williamson, and R. L. S. Patterson. 1986. 16-Androstene steroids in the sub- maxillary salivary gland of the boar in relation to measures of boar taint in carcasses. Anim. Prod. 42:145. Braude, R. 1972. Feeding methods. P. 279 in Pig Production, D. J. A. Cole, ed. London: Butterworth. Brooks, R. I., A. M. Pearson, M. G. Hogberg, J. J. Pestka, and J. I. Gray. 1986. An immunological approach for prevention of boar odor in pork. J. Anim. Sci. 62:1279. Cassens, R., F. Giesler, and Q. Kolb. 1972. Proceed- ings of the Pork Quality Symposium. Madison: Cooperative Extension Service, University of Wis- cons~n. Chung, C. S., T. D. Etherton, and J. P. Wiggins. 1985. Stimulation of swine growth by porcine growth hormone. J. Anim. Sci. 60:118. Cunningham, H. M. 1968. Effect of caffeine on nitro- gen retention, carcass composition, fat mobilization and oxidation of Ci4-labeled body fat in pigs. J. Anim. Sci. 27:424. Cunningham, H. M., and D. W. Friend. 1964. Effect of nicotine on nitrogen retention and fat deposition in pigs. J. Anim. Sci. 23:717. Cunningham, H. M., D. W. Friend, and J. W. G.

ALTERING CARCASS MEASUREMENTS Nicholson. 1963. Effect of epinephrine on nitrogen and fat deposition of pigs. J. Anim. Sci. 22:632. Davey, R. J., and D. P. Morgan. 1969. Protein effect on growth and carcass composition of swine selected for high and low fatness. J. Anim. Sci. 28:831. Ellis, N. R., and H. S. Isbell. 1926. Soft pork studies. II. The influence of the character of the ration upon the composition of the body fat of hogs. J. Biol. Chem. 69:219. Evans, R. 1986. Fall 1985 summary. Iowa Swine Testing Station. Ames, Iowa: Iowa Swine Testing Station. Flint, D. J., and C. E. Futter. 1986. Immunological manipulation of body fat. P. 123 in Hannah Research 1985. Ayr, Scotland: Hannah Research Institute, University of Glasgow. Greer, S. A. N., V. W. Hays, V. C. Speer, J. T. McCall, and E. G. Hammond. 1965. Effects of level of corn- and barley-base diets on performance and body composition of swine. J. Anim. Sci. 24:1008. Jensen, A. H., D. C. Acker, H. M. Maddock, G. C. Ashton, P. G. Homeyer, E. O. Heady, and D. V. Catron. 1955. Different protein levels with and without antibiotics for growing-finishing swine: Ef- fect on growth rate and feed efficiency. J. Anim. Sci. 14:69 Johnson, R. A. 1965. Substitution Rates and Economic Optima in Corn-Soybean Rations for Growing-Fin- ishing Swine. Ph.D. dissertation. Iowa State Uni- versity, Ames. Jones, R. W., R. A. Easter, F. K. McKeith, R. H. Dalrymple, H. M. Maddock, and P. J. Bechtel. 1985. Effect of the B-adrenergic agonist cimaterol 277 (CL 263,780) on the growth and carcass character- istics of finishing swine. J. Anim. Sci. 61:905. Machlin, L. J. 1972. Effect of porcine growth hormone on growth and carcass composition of the pig. J. Anim. Sci. 35:794. Moser, R L., R. H. Dalrymple, S. G. Cornelius, J. E. Pettigrew, and C. E. Allen. 1984. Evaluation of a repartitioning agent on the performance and carcass traits of finishing pigs. J. Anim. Sci. 59(Suppl. 1~:255. Phelps, A. 1985. Consumer backlash against U.K.'s drive for leaner pork. Feedstuffs 57(40~:S-6. Plimpton, R. F., Jr., and H. S. Teague. 1972. Influence of sex and hormone treatment on performance and carcass composition of swine. J. Anim. Sci. 35:1166. Speer, V. C. 1966. Floor feed or self feed? Hog Farm Management 3~7):13. Spencer, G. S. G., and G. J. Garssen. 1983. A novel approach to growth promotion using auto-immuni- zation against somatostatin. I. Effects on growth and hormone levels in lambs. Livestock Prod. Sci. 10:25. Stahly, T. S., and G. L. Cromwell. 1979. Effect of environmental temperature and dietary fat supple- mentation on the performance and carcass charac- teristics of growing and finishing swine. J. Anim. Sci. 49:1478. Wagner, G. R., A. J. Clark, V. W. Hays, and V. C. Speer. 1963. Effect of protein-energy relationships on the performance and carcass quality of growing swine. J. Anim. Sci. 22:202. Williamson, E. D., R. L. C. Patterson, E. R. Buxton, K. G. Mitchell, I. G. Partridge, and N. Walker. 1985. Immunization against 5-~-androstenone in boars. Livestock Prod. Sci. 12:251.

Next: Production and Processing Options to Alter Composition »
Designing Foods: Animal Product Options in the Marketplace Get This Book
×
Buy Paperback | $65.00 Buy Hardback | $65.00
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

This lively book examines recent trends in animal product consumption and diet; reviews industry efforts, policies, and programs aimed at improving the nutritional attributes of animal products; and offers suggestions for further research. In addition, the volume reviews dietary and health recommendations from major health organizations and notes specific target levels for nutrients.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!