Body Composition Techniques in Health and Disease

(Society for the Study of Human Biology Symposium Series,

Series Number 36)

Visit the link below to download the full version of this book:
https://cheaptodownload.com/product/body-composition-techniques-in-health-and-di
sease-society-for-the-study-of-human-biology-symposium-series-series-number-36-i
llustrated-edition-full-pdf-docx-download/
SOCIETY FOR THE STUDY OF HUMAN BIOLOGY
SYMPOSIUM SERIES: 36

Body composition techniques in health and disease

 PUBLISHED SYMPOSIA OF THE
SOCIETY FOR THE STUDY OF HUMAN BIOLOGY

10 Biological Aspects of Demography ED. w. BRASS

11 Human Evolution ED. M.H. DAY
12 Genetic Variation in Britain ED. D.F. ROBERTS AND E. SUNDERLAND
13 Human Variation and Natural Selection ED. D.F. ROBERTS (Penrose Memorial Volume
reprint)
14 Chromosome Variation in Human Evolution ED. A.J. BOYCE
15 Biology of Human Foetal Growth ED. D.F. ROBERTS
16 Human Ecology in the Tropics ED. J.P. GARLICK AND R.W.J. KEAY
17 Physiological Variation and its Genetic Base ED. J.S. WEINER
18 Human Behaviour and Adaptation ED. N J . BLURTON JONES AND V. REYNOLDS
19 Demographic Patterns in Developed Societies ED. R.W. HIORNS
20 Disease and Urbanisation ED. E.J. CLEGG AND J.P. GARLICK
21 Aspects of Human Evolution ED. C.B. STRINGER
22 Energy and Effort ED. G.A. HARRISON
23 Migration and Mobility ED. A.J. BOYCE
24 Sexual Dimorphism ED. F. NEWCOMBE et al.
25 The Biology of Human Ageing ED. A.H. BITTLES AND K.J. COLLINS
26 Capacity for Work in the Tropics ED. K.J. COLLINS AND D.F. ROBERTS
27 Genetic Variation and its Maintenance ED. D.F. ROBERTS AND G.F. DE STEFANO
28 Human Mating Patterns ED. C.G.N. MASCIE-TAYLOR AND A.J. BOYCE
29 The Physiology of Human Growth ED. J.M. TANNER AND M.A. PREECE
30 Diet and Disease ED. G.A. HARRISON AND J. WATERLOW
31 Fertility and Resources ED. J. LANDERS AND V. REYNOLDS
32 Urban Ecology and Health in the Third World ED. L.M. SCHELL, M.T. SMITH AND
A. BILLSBOROUGH
33 Isolation, Migration and Health ED. D.F. ROBERTS, N. FUJIKI AND K. TORIZUKA
34 Physical Activity and Health ED. N.G. NORGAN
35 Seasonality and Human Ecology ED. S.J. ULIJASZEK AND S.S. STRICKLAND

Numbers 1-9 were published by Pergamon Press, Headington Hill Hall, Headington, Oxford
0X3 OBY. Numbers 10-24 were published by Taylor & Francis Ltd, 10-14 Macklin Street,
London WC2B 5NF. Futher details and prices of back-list numbers are available from the
Secretary of the Society for the Study of Human Biology.
Body composition techniques
in health and disease

EDITED BY

P. S. W. DAVIES AND T. J. COLE

MRC Dunn Nutritional Laboratory, Cambridge

CAMBRIDGE
UNIVERSITY PRESS
Published by the Press Syndicate of the University of Cambridge
The Pitt Building, Trumpington Street, Cambridge CB2 1RP
40 West 20th Street, New York, NY 10011-4211, USA
10 Stamford Road, Oakleigh, Melbourne 3166, Australia

© Cambridge University Press 1995

First published 1995

A catalogue record for this book is available from the British Library

Library of Congress cataloguing in publication data

Body composition techniques in health and disease / edited by P.S.W. Davies and T J . Cole.
p. cm. - (Society for the Study of Human Biology symposium series ; 36)
Includes index.
ISBN 0 521 46179 0 (hardback)
1. Body, Human - Composition - Measurement. 2. Anthropometry - Methodology. I.
Davies, P. S. W. II. Cole, T. J. III. Series.
[DNLM: 1. Body Composition - congresses. 2. Anthropometry - methods - congresses.
Wl SO861 v.36 1995 / QU 100 B668 1995]
QP33.5.B63 1995
612 - dc20
DNLM/DLC
for Library of Congress 94-5327 CIP

ISBN 0 521 46179 0 hardback

Transferred to digital printing 2004

VN
 Contents

List of contributors page ix

1 Application of dual-energy X-ray absorptiometry and

related techniques to the assessment of bone and body
composition 1
A. PRENTICE

2 In vivo neutron activation analysis: past, present and future 14

S.J.S. RYDE

3 Magnetic resonance imaging for the assessment of body

composition 38
P. BRAMBILLA, P. MANZONI, P. SIMONE AND G. CHIUMELLO

4 Multi-frequency impedance as a measure of body water

compartments 45
P. DEURENBERG

5 Body composition assessed by electrical conductivity

methods 57
J.F. SUTCLIFFE AND M.A. SMITH

6 Body composition in malnutrition 71

P.S. SHETTY

7 Influence of body composition on protein and energy

requirements: some new insights 85
C.J.K. HENRY

8 Prediction of adult body composition from infant and

child measurements 100
M.F. ROLLAND-CACHERA

9 Assessment of body composition in the obese 146

E.M.E. POSKITT

Vll
viii Contents

10 The role of body physique assessment in sports science 166

M.H. SLAUGHTER AND C.B. CHRIST

11 The assessment of the body composition of populations 195

N.G. NORGAN

12 Changes in approach to the measurement of body

composition 222
j . PARIZKOVA

13 Multi-compartment models for the assessment of body

composition in health and disease 240
S.A. JEBB and M. ELIA

14 The future of body composition research 255

S.B. HEYMSFIELD a n d ZI-MIAN WANG

Index 271
Contributors

P. Brambilla
Clinica Pediatrica IV, Ospedale San Raffaele, Via Olgettina 60,1-20132
Milan, Italy
G. Chiumello
Clinica Pediatrica IV, Ospedale San Raffaele, Via Olgettina 60,1-20132
Milan, Italy
C.B. Christ
Department of Kinesiology, University of Illinois, 125 Freer Hall, 906
South Goodwin Avenue, Urbana, IL 61801, USA
P. Duerenberg
Department of Human Nutrition, Wageningen Agricultural University,
Bromenweg 2, NL-6703 HD Wageningen, The Netherlands
M. Elia
Dunn Clinical Nutrition Centre, 100 Tennis Court Road, Cambridge
CB2 1QL, UK
CJ.K. Henry
School of Biological & Molecular Science, Oxford Brookes University,
Gipsy Lane, Headington, Oxford OX3 0BP, UK
S.B. Heymsfield
Obesity Research Centre, Weight Control Unit, 411 West 114th Street,
New York, NY 10025, USA
S.A. Jebb
Dunn Clinical Nutrition Centre, 100 Tennis Court Road, Cambridge
CB2 1QL, UK
P. Manzoni
Clinica Pediatrica IV, Ospedale San Raffaele, Via Olgetinna 60,1-20132
Milan, Italy
N.G. Norgan
Department of Human Sciences, Loughborough University, Loughborough,
Leicestershire LEU 3TU, UK
J. Parizkova
IVth Clinic of Internal Medicine, U Nemocnice 2, Prague 2,12000 Czech
Republic

IX
Contributors

E.M.E. Poskitt
Dunn Nutrition Group, MRC Laboratories, Fujara, Banjul PO Box 273,
The Gambia, West Africa
A. Prentice
MRC Dunn Nutritional Laboratory, Downham's Lane, Milton Road,
Cambridge CB4 1XJ, UK
M.F. Rolland-Cachera
ISTNA-CNAM, 2 Rue Conte, F-75141 Paris Cedex 03, France
SJ.S. Ryde
Swansea In Vivo Analysis Research Group, Department of Medical
Physics and Clinical Engineering, Singleton Hospital, Swansea SA2 8QA,
UK
P.S. Shetty
Centre for Human Nutrition, London School of Hygiene and Tropical
Medicine, Keppel Street, London WC1E 7HT, UK
P. Simone
Clinica Pediatrica IV, Ospedale San Raffaele, Via Olgetinna 60,1-20132
Milan, Italy
M.H. Slaughter
Department of Kinesiology, University of Illinois, 125 Freer Hall, 906
South Goodwin Avenue, Urbana, IL 61801, USA
M.A. Smith
Centre for Bone and Body Composition Research, Academic Unit of
Medical Physics, Leeds General Infirmary, Great George Street, Leeds
LSI 3EX, UK
J.F. Sutcliffe
Centre for Bone and Body Composition Research, Academic Unit of
Medical Physics, Leeds General Infirmary, Great George Street, Leeds
LSI 3EX, UK
Zi-Mian Wang
Obesity Research Centre, Weight Control Unit, 411 West 114th Street,
New York, NY 10025, USA
1 Application of dual-energy X-ray
absorptiometry and related techniques
to the assessment of bone and body
composition
ANN PRENTICE

Introduction

The development of absorptiometric techniques has revolutionised our

ability to measure the mineral content of the human skeleton in living
individuals. The first technique to be introduced was single-photon
absorptiometry (SPA), capable of the precise measurement of bone mineral
in the appendicular skeleton (Cameron et al.91962; Cameron & Sorenson,
1963; Sorenson & Cameron, 1967). Some years later, dual-photon
absorptiometry (DPA) was developed for the measurement of bone in the
axial skeleton (Mazess, 1971; Roos & Skoldborn, 1974). Versions of these
instruments are capable of scanning the entire body and these provide a
measure of whole-body fat and lean masses in addition to total-body bone
mineral content. The photon beams required for SPA and DPA are
produced by gamma-emitting radioisotopes. In recent years, X-ray technology
has replaced the use of gamma-rays and the instruments are referred to as
single- or dual-energy X-ray absorptiometers (SXA, DXA).
Absorptiometric techniques are known by a variety of names. For
example, bone densitometry refers to any absorptiometric method which
measures bone mineral content; X-ray absorptiometry (XRA), quantitative
digital radiography (QDR) and X-ray spectrophotometry are synonyms
for DXA, and there is a continuing debate as to whether the accepted
abbreviation for dual-energy X-ray absorptiometry should be DXA or
DEXA.
Currently, the principal application of absorptiometry is for the
measurement of bone mineral content in clinical practice (Johnston et a/.,
1991). This includes screening for osteoporosis, either as a primary
condition or secondary to a number of disease states, and long-term
monitoring of therapies and drugs known to affect bone. Imaging of
prostheses and stones is possible with the latest DXA instruments. In recent
years there has been an increasing use of absorptiometry as a research tool.
2 A. Prentice

Examples of research studies that have involved absorptiometry include

investigations of the determinants of peak bone mass and bone loss in later
life, assessments of risk factors for osteoporosis, and estimations of
nutritional requirements for growth and reproduction. The ability of
whole-body instruments to estimate total fat and lean tissue masses has also
been exploited in studies of obesity, ageing, sports physiology, anorexia,
and wasting due to cancer and AIDS.
This review will describe and discuss the principles of single- and
dual-energy absorptiometry, the advantages and limitations of the
instruments, and some issues concerning the interpretation of bone mineral
data.

Methodology
Single-energy absorptiometry
The principle of absorptiometry is based on the exponential attenuation of
penetrating photons as they pass through tissue. When a monochromatic
beam passes through a region of the body consisting of two tissues, the
intensity of the emerging beam is related to the initial intensity as follows:

ln(/// 0 ) = - / / a m a - / i b m b (1.1)

where / is the final intensity, / 0 the initial intensity, /xa the mass attenuation
coefficient of tissue a, ma the mass of tissue a, fih the mass attenuation
coefficient of tissue b, and mb the mass of tissue b. The mass attenuation
coefficients are constants that are characteristic of the tissues at the energy
of the photon beam.
In single-energy absorptiometry, as applied to the measurement of bone
mineral, the amount of soft tissue in the beam has to be known in order to
quantify bone mass. This is achieved by packing the scan region with
tissue-equivalent material, by immersing the limb in a water bath, using a
tissue-equivalent gel, or wrapping a supple bag of tissue-equivalent
material around the limb and using moveable plates to flatten the upper
and lower bag surfaces. This ensures a constant overall thickness of the scan
region. The baseline intensity is measured as the intensity of the beam after
passing through the soft tissue region immediately adjacent to the bone.
This allows for differences in soft tissue thickness and composition from one
scan to another. The composition of the soft tissue over- or underlying bone
has to be assumed to be homogeneous and of identical composition to the
adjacent soft tissue. A consequence of these requirements is that single-energy
absorptiometry cannot be used for regions of the body where the skeleton is
surrounded by soft tissue of variable thickness and composition, such as in
 DXA and assessment of body composition 3

the trunk, and, in practice, it is used only for the bones of the arms and legs,
particularly the radius, ulna, humerus and femoral shaft.
The mineral content in a transverse cross-section of the bone of interest, 1
cm wide in the axial direction, is obtained by measuring the emerging
intensity of the finely collimated photon beam at frequent intervals along
the scan path. The final equation becomes:

BMC = lLPh/(fihPh - fisPs)-] Zln(I 0 /I) (1.2)

where BMC is the bone mineral content (g/cm), L the data acquisition
interval (cm), fib the mass attenuation coefficient (cm2/g) of bone, ph the
density (g/cm3) of bone, fis the mass attenuation coefficient of soft tissue, ps
the density of soft tissue, IQ the baseline intensity after passage through soft
tissue, and / the emerging intensity after passage through bone and soft
tissue (Sorenson and Cameron, 1967; Wahner et al., 1984; Wahner &
Riggs, 1986; Tothill, 1989).
In general, the monochromatic photon beam in SPA is provided by the
radioisotopes 125 I or 241 Am, and detection is with scintillation crystals,
usually sodium iodide (Nal). 125 I produces gamma-rays of energy 31 and
35 keV, which after passing through a tin filter provide a monochromatic
beam of energy 27.4 keV. At this energy there is good contrast between
bone and soft tissues while the attenuation characteristics of muscle, water,
tendons and blood vessels are similar. However, major changes in fat
content within bone and in the adjacent tissues can produce difficulties
(Wahner & Riggs, 1986). The half-life of 125 I is relatively short (60 days),
and such sources require replacement every 5-6 months. 241 Am produces a
gamma-ray of higher energy, 59.6 keV, which has greater penetrating
power but lower contrast between bone and soft tissues and is more prone
to problems associated with intra- and extra-osseous fat. The greater
penetrating power of 241 Am makes this isotope useful for scans of the
humerus and femur and its long half-life (438 years) means that frequent
source changes are not necessary. Single-energy machines based on X-ray
technology are now becoming available.

Dual-energy absorptiometry
The simultaneous use of two photon beams with different energies makes
possible the quantification of bone mineral in regions of the body where it is
impossible to achieve constant thickness and where soft tissue composition
may be variable.
The basic equations for the attenuation of each beam as they are
transmitted through two tissues are identical to those in equation (1.1)
(Peppier & Mazess, 1981):
4 A. Prentice

Energy 1: ln(J'/J' o ) = - // a m a - fifhmh (1.3)

Energy 2: ln(J"//" 0 ) = - / / > a - fi\mb (1.4)

The abbreviations are the same as for equation (1.1) with the superscripts
' and " denoting the values at energy 1 and 2 respectively. The mass
attenuation coefficients for the two tissues are constants which depend on
the beam energy. Measurement of the initial and emerging intensities of
both beams passing through the same volume enables the simultaneous
equations (1.3) and (1.4) to be solved and the masses of both tissues to be
calculated.
Dual-energy absorptiometry makes use of this principle to determine the
bone mineral content in anatomical regions of the axial skeleton (Roos &
Skoldborn, 1974; Peppier & Mazess, 1981; Wahner et a/., 1984; Wahner &
Riggs, 1986; Tothill, 1989). Attenuation at two energies in a region of soft
tissue gives fat and lean masses; in regions containing bone it provides bone
and soft tissue masses. The composition, and hence the mass attenuation
coefficients, of soft tissue in regions containing bone is assumed to be
identical to that of adjacent soft tissue in these calculations. The bone
mineral, fat and lean masses along one scan are built up by repeated data
acquisition along the length of the scan, as in single-energy absorptiometry.
Rectilinear scanning over the entire region allows the masses within a
defined anatomical region to be determined. Some instruments are dedicated
to the measurement of bone mineral in the lumbar spine and proximal
femur. Others are capable of scanning the entire body, and provide
information about total and regional distributions of bone mineral, fat and
lean tissues.
In dual-photon absorptiometry the beams of photons at two energies are
provided either by one radioisotope with two gamma-emissions or by a
combination of two radioisotopes. The detection systems are based on
scintillation crystals, usually Nal. Theoretically, the optimum energies for
the low and high energy beams are 35-45 keV and 200-800 keV respectively
(Roos & Skoldborn, 1974). The lower energy provides high contrast
between bone and soft tissue while mass attenuation at the higher energy is
related to tissue density. The most common source used for DPA is 153 Gd,
which emits gamma-rays with energies 44 keV and 100 keV. Combinations
of radioisotopes that have been used include 125 I/ 241 Am (27.4 keV, 59.6
keV) and 241 Arn/ 137 Cs (59.6 keV, 662 keV).
In recent years the use of radioisotopes has been superseded by X-ray
technology. In DXA instruments the two collimated energy beams are
produced in one of two ways, depending on the manufacturer (Sartorius &
Resnick, 1988; Arai et al.91990). In the first method, X-rays with a spectrum
of energies are produced by a constant-voltage X-ray generator and the
 DXA and assessment of body composition 5

beam is separated into high- and low-energy regions using k-edge filters of
cerium or samarium. Current examples of this type of instrument include
the Norland XR-26 (tube voltage 100 kV; peak beam energies 46.8 keV, 80
keV; dual Nal scintillation detection) and Lunar DPX (tube voltage 76 kV;
peak beam energies 38 keV, 70 keV; single Nal scintillation detection). The
second method involves a switching-pulse system that rapidly alternates
the voltage of the X-ray generator producing two beams of high and low
energy virtually simultaneously. An internal calibration wheel corrects for
any small fluctuations caused by this method of beam generation and
compensates for beam hardening effects. An example of this type of
instrument is the Hologic QDR-1000 (tube voltages 70 kV, 140 kV; peak
energies 43 keV, 110 keV; detection by CdWO 4 scintillation). DXA
instrumentation is an improvement over DPA in that scan times are faster,
imaging is better, precision is improved and the purchase, replacement and
disposal of radioactive sources is not required.

Advantages and limitations of absorptiometry

Radiation exposure
One of the major advantages of absorptiometry compared with other
techniques for assessing bone mineral content is that radiation exposure is
low. The effective dose equivalent per measurement (EDE, a measure
which takes into account surface exposure and penetrating power of the
energy beam together with the depth and vulnerability of exposed tissues)
depends on both the scan site and the instrument used, but values are
generally within natural background radiation levels (Table 1.1; Kalendar,
1992). As a consequence, absorptiometry can be used to study healthy
children and pre-menopausal women, as well as other, less vulnerable
population groups, whereas methods such as neutron activation analysis
and computed tomography are unsuitable. An exception is the measurement
of the lumbar spine using DPA, as the EDE for pre-menopausal women is
comparatively high because of the presence of the ovaries in the scan volume.
The surface exposure per measurement varies considerably between
instruments and it is important for separate dosimetric evaluations to be
made on each machine. Table 1.1 gives EDE estimates for pre-menopausal
women measured on a number of instruments currently in use at the Dunn
Nutrition Unit, Cambridge, together with some examples of EDE from
exposure to other radiation sources. In general, the EDE values for men
and for post-menopausal women are lower than those for pre-menopausal
women (Kalendar, 1992). Although the radiation risks are low, it is prudent
not to scan women during pregnancy.
6 A. Prentice

Table 1.1. Examples of effective dose equivalents (EDE) received during

absorptiometry and from other sources in premenopausal women

Radiation source Instrument EDE (jiSv)

SPA midshaft radius Norland 2780 0.04

SPA distal radius Lunar SP2 0.18
DPA lumbar spine Novolab 22A 20
DPA hip Novolab 22A 2
DXA lumbar spine Hologic QDR 1000/w 4
DXA hip Hologic QDR 1000/w 4
DXA whole-body Hologic QDR 1000/w 5
Daily background in Cambridge 6
Transatlantic return flight 80
Spinal radiograph 700
Cardangiographic examination 10000

Data are from independent dosimetry of instruments used at the MRC Dunn Nutrition Unit
(A. Prentice, unpublished data) and from Kalendar (1992). The effective dose equivalent
(EDE) takes into account the surface radiation exposure, the penetrating power of the energy
beam, and the depth and vulnerability of tissues irradiated.
SPA, single-photon absorptiometry; DPA, dual-photon absorptiometry; DXA, dual-energy
X-ray absorptiometry.

Convenience
Absorptiometric techniques are very convenient as the patient/volunteer is
required only to sit or lie still for the duration of the scan. Scan times on the
latest instruments are much faster than with the older machines, with a
whole-body scan typically taking 10-20 min and a lumbar spine scan taking
5-10 min. The nature of the technique means that absorptiometry can be
used when other methods would be impossible. An example is the difficulty
of using hydrodensitometry for body fat determinations when dealing with
sick or disabled individuals. The necessity to keep still, however, can be a
problem for some individuals, especially children and infirm people.

Precision and accuracy

The precision of absorptiometry is high and is superior to earlier methods
for measuring bone mineral. The coefficients of variation of repeated bone
scans across phantoms are about 1-2%, or less with the latest DXA
instruments. The precision of repeated measurements is poorer when
people are scanned, particularly if they are re-positioned each time or if they
are osteoporotic, but generally within-day reproducibility is 3-5%. In
addition, long-term reproducibility is good, especially with X-ray absorp-
tiometry, which does not have the problems of source strength changes that
 DXA and assessment of body composition 1

inevitably effect long-term precision in SPA and DPA (Dunn et al, 1987).
Such high precision facilitates the longitudinal monitoring of individuals,
although, even in osteoporosis, changes in bone mass can be sufficiently
slow that measurements have to be spaced by several months or years for
significant increments to be detectable (Davis et a/., 1991). The precision of
fat and lean body mass determinations is also good and generally superior
to other body composition methods such as hydrodensitometry and
skinfold thickness measurements (e. g. Pritchard et a/., 1993). The precision
of all absorptiometric measurements, however, is affected by poor positioning
and excessive movement, and this can be a particular problem with
children, disabled or anxious people.
The accuracy of absorptiometry is more problematical and, currently,
there are a number of issues which give rise to concern and which remain
unresolved. Accuracy can be affected by choice of calibrating materials,
assumptions made in the computer programming about bone-edge detection
and intra-osseous fat, the depths of tissues in the scan path, the non-uniformity
of soft tissues overlying bone, and differences in clothing and bedding
(Farrell & Webber 1989; Nord & Payne, 1990; Tothill & Pye 1990; Mazess
et al.9 1991; Jonson, 1993). As a consequence, both absolute and relative
values can differ substantially between manufacturers and between
instruments. Use of different versions of computer software can affect
results and updates of pre-existing programs necessitate re-analysis of data.
This is a particular problem in longitudinal assessments or when comparisons
are made between individuals measured on different machines. As a result,
it is important that both the instrument and the software version used for
analysis are specified in published accounts of absorptiometric studies.
Possibly the greatest problems with accuracy are found with fat and lean
tissue mass estimates, especially at the extremes of tissue depth (equivalent
to > 10 cm or <25 cm water), such as are encountered in children,
anorexics, lean male athletes and the obese. At present there have been
comparatively few validation experiments of the accuracy of either bone or
body composition measurements by cadaver analysis and more are
necessary before we can be confident about the accuracy of absorptiometry.

Bone measurements
One of the major limitations of bone measurements made by absorptiometry
is that the result represents an integration of absorption over all elements
within the bone envelope, such as the medullary cavity, and is not confined
to osseous tissue per se. Thus a thin annulus of heavily mineralised bone in
the forearm potentially could have the same measured bone mineral
content as a wider annulus of osteoporotic bone with identical external
8 A. Prentice

dimensions. Similarly, absorptiometry cannot differentiate between trabecular

and cortical bone, nor exclude abnormalities, such as osteophytes, crush
fractures or calcifications, which interfere with the measurement of bone
mineral content. Other methods, such as computed tomography and
X-radiography are necessary if structural information about bone is required.

Interpretation of bone measurements by absorptiometry

Expression of bone data

The results from SPA are expressed as bone mineral content (BMC), which
represents the mass of mineral in a slice of bone, 1 cm deep in the axial
direction, and has the unit grams per centimetre. Results from DPA and
DXA are expressed as bone mass per anatomical feature, such as spinal
vertebrae or neck of femur, and the unit is grams. Confusingly, these values
are also referred to as bone mineral content (BMC), although recently it has
been suggested that the term bone mineral mass (BMM) should be adopted
(Jonson, 1993).
To add to the confusion, BMC values from both single- and dual-energy
absorptiometry are frequently divided by the width or area of the bone in
the scan to give a figure referred to as the bone mineral density (BMD), with
units grams per square centimetre. This value is not a true bone density,
partly because it is only an areal rather than a volumetric measure, and
partly because absorptiometry cannot distinguish between osseous and
non-osseous tissues within the area designated as bone.
The interpretation of BMD is further complicated by the fact that BMC
is not necessarily directly proportional to bone width (BW) or bone area
(BA). In many population groups, a significant correlation remains
between BMD values and BW (or BA), demonstrating that calculation of
BMD does not completely adjust the variation in BMC due to differences in
bone size. In a recent SPA study, for example, BMC at the radial shaft of
young children was shown to be proportional, after age adjustment, to BW
raised to the power 0.77 (Prentice et al, 1990). This shows that, in this case,
the use of BMD to adjust for differences in bone size would artificially
underestimate apparent bone density in those children with the widest
bones at each age.
The purpose of calculating BMD is to facilitate the comparison of bone
mineral measurements between individuals of different bone and body
sizes. In practice, correlations often remain not only between BMD and
BW (or BA) but also with indices of body size, such as body weight and
height. A number of different approaches have been proposed for adjusting
BMC or BMD data for body size. Many involve use of simple indices such
 DXA and assessment of body composition 9

as ideal weight-for-height (weight/height) or body mass index (BMI,

weight/height 2). However, in most data sets, BMC and BMD are both
positively and independently correlated with body weight and height, and
the use of indices in which weight and height act in opposite directions is
inappropriate. A better solution is to adjust BMC values for BW (or BA),
body weight and height separately using multiple regression analysis (Cole
& Prentice, 1992). A particularly informative procedure is to transform
these variables to natural logarithms prior to analysis as this allows power
and proportional relationships to be examined (Prentice et a/., 1990; Cole
& Prentice, 1992).
This begs the question as to why it is felt necessary to express the data as
either BMD or BMC adjusted for bone and body size. For technical
reasons, BMD is a more precise measurement than BMC, as any movement
during the scan will affect both BMC and bone width or area. For
long-term monitoring of individual patients, use of BMD may be preferable
as it removes some of the uncertainties due to measurement error.
However, in osteoporosis screening the relevant variable is probably BMC
as the strength of bone is strongly correlated to mineral mass (Dequeker,
1988; Geusens & Dequeker, 1988). Adjustments for bone and body size can
be helpful, however, in assessing the relative BMC of individuals or groups
of different size. Examples include comparing the bone mineral content of
Blacks and Whites, or athletes and non-athletes, or assessing the importance
of certain hormones in maintaining bone mineral. In these instances
incomplete correction of BMC for size, for example by using either BMD
alone or corrected for BMI, could lead to spurious relationships emerging,
as the residual influence of size on BMC could result in apparent
correlations with other size-related variables (Cole & Prentice, 1992). The
use of multiple regression analysis with BMC as the dependent variable, as
described above, avoids the complexities which are introduced by using
derived indices such as BMD and BMI to adjust for size differences.

Reference data
In general, osteoporosis screening and diagnosis are undertaken by
comparing an individual's BMC or BMD value with data from a reference
population. The patient's results are expressed relative to the sex-specific
reference population either at the subject's age or for young adults. Clinical
interpretation is subjective but typically individuals are considered at
increased fracture risk if their values are less than 80% or 2 standard
deviations (SD) below the reference mean.
At present there are difficulties with the use of reference data. Firstly, the
reference distribution is provided by the manufacturers and may not be
10 A. Prentice

pertinent to local population groups. In addition, differences between

reference databases and instrumentation can lead to anomalies where a
patient appears normal on one machine and osteoporotic on another
(Laskey et al, 1992), thereby affecting the diagnosis and the management of
the patient.

Reversible calcium space

Bone remodelling is a dynamic process whereby resorption at specific sites
is followed, after a period of time, by bone formation. In healthy young
adults these two processes are linked so that, ultimately, total bone mass is
unaffected. When the rate of bone turnover decreases, bone formation
continues at pre-existing resorption sites, while fewer new resorption sites
are formed. Eventually bone formation decreases to match resorption, but
the process can take several months (Melsen & Mosekilde, 1988).
Absorptiometry is a static measurement which provides a 'snapshot' of
BMC at any one moment. If bone turnover decreases with no overall
change in bone volume, BMC values will rise by a few per cent because of
the smaller number of resorption cavities. BMC values continue to rise for
some time until a new steady state is achieved because of the time lag
between resorption and formation (Parfitt, 1980; Kanis, 1991). Similarly, a
decrease in reversible calcium space caused by an increase in bone turnover
will lead to a decrease in measured BMC.
Such artefacts caused by changes in reversible calcium space potentially
could confound interpretation of absorptiometric studies. In a patient with
osteoporosis, an observed rise in BMC after drug treatment might be due to
reduced bone turnover rather than to increased bone mass. In this instance,
the decrease in turnover might of itself be beneficial, as bone loss would be
diminished, and the exact interpretation of the BMC change could be
regarded as academic. In other cases, however, a proper understanding of
the mechanism could alter the conclusions reached. An example is provided
by a recent study of calcium supplementation during childhood which
involved pre-pubertal twins (Johnston et a/., 1992). Those twins who
received a calcium supplement of about 700 mg Ca/day on top of a dietary
intake of about 900 mg Ca/day were shown to have significantly higher
BMC (BMD) values than their co-twins after 3 years. This has been
interpreted as showing that the calcium-supplemented twins had a greater
rate of bone acquisition and that possibly the calcium requirements of
pre-pubertal children are higher than previously thought. However, the
bone turnover of the supplemented twin, as shown by circulating osteocalcin
levels, was significantly lower than that of the co-twin, suggesting that the
observed differences in BMC were, at least in part, due to effects on
 DXA and assessment of body composition 11

reversible calcium space, and may not have been due to differences in bone
mass. Consequently, the likely impact of the calcium supplementation on
the development of peak bone mass is uncertain and must await long-term
follow-up. Indeed, it is not too fanciful to hypothesise that the twin with the
lower bone turnover might have slower bone development and that this
might result in a lower peak bone mass. It is obviously essential that more
research is undertaken to understand the relevance and importance of
reversible calcium space in the interpretation of absorptiometric data, and
in future it may be worth while including measures of bone turnover in
research protocols involving absorptiometry.

Summary
It is now possible to measure the bone mineral content and body
composition of a wide range of individuals, in health and during illness,
using a variety of instruments based on absorptiometry. Single- and
dual-energy absorptiometric measurements are quick and easy to make,
are accompanied by only a very small radiation exposure and are highly
precise. However, there are still considerable concerns about the accuracy
of these techniques, especially at extremes of tissue depth, and interpretation
of the data is complex. Despite substantial advances in instrumentation and
understanding in recent years, there are still many pitfalls for the unwary
and absorptiometers cannot be treated as 'black-box' machines that will
produce meaningful results in unskilled hands and in all circumstances. The
applications of absorptiometry, especially DXA, are likely to increase in the
future as the instruments become more widely available in research
laboratories.

References
Arai, H., Ito, K., Nagao, K. & Furutachi, M. (1990). The evaluation of three
different bone densitometry systems: XR-26, QDR-1000 and DPX. Image
Technology and Information Display, 22, 1-6.
Cameron, J.R. & Sorenson, J. (1963). Measurement of bone mineral in vivo: an
improved method. Science, 142, 230-2.
Cameron, J.R., Grant, R. & MacGregor, R. (1962). An improved technique for the
measurement of bone mineral content in vivo. Radiology, 78, 117.
Cole, T.J. & Prentice, A. (1992). Bone mineral measurements. British Medical
Journal, 305, 1223^.
Davis, J.W., Ross, P.D., Wasnich, R.D., MacLean, C.J. & Vogel, J.M. (1991).
Long-term precision of bone loss rate measurements among postmenopausal
12 A. Prentice

women. Calcified Tissues International, 48, 311-18.

Dequeker, J. (1988). Calcified tissues: structure-function relationships. In Calcium
in Human Biology, ed. B.E.C. Nordin, pp. 209-40. London: Springer-Verlag.
Dunn, W.L., Kan, S.H. & Wahner, H.W. (1987). Errors in longitudinal measurements
of bone mineral: effect of source strength in single and dual photon
absorptiometry. Journal of Nuclear Medicine, 28, 1751-7.
Farrell, T.J. & Webber, C.E. (1989). The error due to fat inhomogeneity in lumbar
spine bone mineral measurements. Clinical Physics and Physiological Measure-
ment, 10, 57-64.
Geusens, P. & Dequeker, J. (1988). Should bone mineral content measurements be
corrected for skeletal size? In Bone Mineral Measurement by Photon Absor-
ptiometry, ed. J. Dequeker, P. Geusens & H.W. Wahner, pp. 96-104. Leuven:
Leuven University Press.
Johnston, C.C., Slemenda, C.W. & Melton, L J . (1991). Clinical use of bone
densitometry. New England Journal of Medicine, 324, 1105-9.
Johnston, C.C., Miller, J.Z., Slemenda, C.W., Reister, T.K., Hui, S., Christian, J.C.
& Peacock, M. (1992). Calcium supplementation and increases in bone
mineral density in children. New England Journal of Medicine, 327, 82-7.
Jonson, R. (1993). Mass attenuation coefficients, quantities and units for use in
bone mineral determinations. Osteoporosis International, 3, 103-5.
Kalendar, W. A. (1992). Effective dose values in bone mineral measurements by
photon absorptiometry and computed tomography. Osteoporosis International,
2, 82-7.
Kanis, J.A. (1991). Calcium requirements for optimal skeletal health in women and
osteoporosis. Calcified Tissues International (Supplement) 49, S33-41.
Laskey, M.A., Crisp, A.J., Cole, T.J. & Compston, J.E. (1992). Comparison of the
effect of different reference data on Lunar DPX and Hologic QDR-1000
dual-energy X-ray absorptiometers. British Journal of Radiology, 65, 1124-9.
Mazess, R.B. (1971). Estimation of bone and skeletal weight by direct photon
absorptiometry. Investigative Radiology, 6, 52-60.
Mazess, R.B., Trempe, J.A., Bisek, J.P., Hanson, J.A. & Hans, D. (1991).
Calibration of dual-energy X-ray absorptiometry for bone density. Journal of
Bone and Mineral Research, 6, 799-806.
Melsen, F. & Mosekilde, L. (1988). Calcified tissues: cellular dynamics. In Calcium
in Human Biology, ed. B.E.C. Nordin, pp. 187-208. London: Springer-Verlag.
Nord, R.H. & Payne, R.K. (1990). Standards for body composition calibration in
DEXA. In Current Research in Osteoporosis and Bone Mineral Measurement,
ed. E.F.J. Ring, pp. 27-8. London: The British Institute of Radiology.
Parfitt, A.M. (1980). Morphologic basis of bone mineral measurements: transient
and steady state effects of treatment in osteoporosis. Mineral and Electrolyte
Metabolism, 4, 273-87.
Peppier, W.W. & Mazess, R.B. (1981). Total body bone mineral and lean body
mass by dual-photon absorptiometry. I. Theory and measurement procedure.
Calcified Tissues International, 33, 353-9.
Prentice, A., Laskey, M.A., Shaw, J., Cole, T.J. & Fraser, D.R. (1990). Bone
mineral content of Gambian and British children aged 0-36 months. Bone and
Mineral, 10,211-24.
 DXA and assessment of body composition 13

Pritchard, J.E., Nowson, C.A., Strauss, B.J., Carlson, J.S., Kaymakci, B. & Wark,
J.D. (1993). Evaluation of dual energy X-ray absorptiometry as a method of
measurement of body fat. European Journal of Clinical Nutrition, 47, 216-28.
Roos, B.O. & Skoldborn, H. (1974). Dual photon absorptiometry in lumbar
vertebrae. I. Theory and method. Ada Radiologica: Therapeutics, 13,266-80.
Sartoris, D.J. & Resnick, D. (1988). Digital radiography may spark renewal of bone
densitometry. Diagnostic Imaging, January, 145-51.
Sorenson, J.A. & Cameron, J.R. (1967). A reliable in vivo measurement of
bone-mineral content. Journal of Bone and Joint Surgery [Am], 49A, 481-97.
Tothill, P. (1989). Methods of bone mineral measurement. Physics in Medicine and
Biology, 34, 543-72.
Tothill, P. & Pye, D.W. (1990). Errors due to fat in lateral spine DPA. In Current
Research in Osteoporosis and Bone Mineral Measurement, ed. E.F.J. Ring, p.
19. London: The British Institute of Radiology.
Wahner, H.W. & Riggs, B.L. (1986). Methods and application of bone densitometry
in clinical diagnosis. CRC Critical Reviews in Clinical Laboratory Sciences, 24,
217-33.
Wahner, H.W., Dunn, W.L. & Riggs, B.L. (1984). Assessment of bone mineral. Part
2. Journal of Nuclear Medicine, 25, 1241-53.
2 In vivo neutron activation analysis:
past, present and future
S.J.S. RYDE

Introduction

It is 30 years since the first measurements of the elemental composition of

the human body were undertaken by Anderson et al. (1964) using the
technique of in vivo neutron activation analysis (IVNAA). Their work
followed a report (Hoffman & Hempelmann, 1957) of two nuclear reactor
accidents in 1945-6 during which ten persons were exposed to bursts of
radiation (fast neutrons and gamma-rays). A measure of the serum 24 Na
activity induced in the body was used to estimate the neutron intensity to
which the subjects had been exposed. It was subsequently realised that a
controlled irradiation with neutrons of known intensity could be used as an
investigative tool.
The IVNAA technique involves the irradiation of the total or partial
body by a beam of neutrons and the detection of characteristic gamma-rays
arising from neutron interaction with the nuclei of the element or elements
of interest. Although the principle of the technique is simple, the factors that
determine the successful measurement of a particular element in vivo can be
complex. Nevertheless, IVNAA has been successfully developed at several
centres over the last three decades and gained increasing acceptance as a
clinically useful, although specialised measurement technique. During this
time considerable research has been undertaken to optimise the measurement
of both bulk (e.g. Ca, C, Cl, H, N, Na, O and P) and trace (e.g. Al, Cd, Cu,
Fe and Si) elements. However, the measurements of calcium, cadmium and
nitrogen have been the most widely applied.
A number of review articles on the techniques and clinical applications of
IVNAA have been published (Cohn, 1980, 1981; Chettle & Fremlin, 1984;
Beddoe & Hill, 1985; Cohn & Parr, 1985) and these provide excellent
sources of information, including extensive reference lists. More recently,
proceedings have been published from the 7n Vivo Body Composition
Studies' series of symposia (Ellis et al., 1987; Yasumura et al., 1990; Ellis &
Eastman, 1993) and these too provide a valuable source of information on
developments in IVNAA (and other techniques of body composition).
This chapter, not intended for experienced practitioners of IVNAA, will

14