1000
Asia Pacific J Clin Nutr (1995) 4: 19-22
Asia Pacific J Clin Nutr (1995) 4: 19-22
Low-cost appropriate technologies
for body composition assessment: a field researcher's view
Noel W. Solomons MD and Manolo Mazariegos MD
Centre for Studies of Sensory Impairment, Aging and Metabolism
(CeSSlAM), the research branch for the National Committee for the
Blind and Deaf of Guatemala, Hospital de Ojos y Oidos Dr Rodolfo Robles
V', Diagonal 21 y 19 Calle, Zona 11, Guatemala City, 01011, Guatemala.
The field setting, as distinct from the clinical
and laboratory settings, relates to the study of populations and
subpopulations. It can involve either free-living or institutionalized
individuals. The concept of 'body composition' goes beyond the traditional
assumptions of screening anthropometrics, although it includes many
of the same measures. The principal practical considerations for
the selection of measurement techniques are ethics and cost. The
quantitative considerations are the precision and accuracy of the
measures. The biological considerations relate to the interpretation
of the measure in terms of underlying body constituents; do the
values mean what we hope them to mean?
Introduction
The definition of field studies and field research
should be apparent. In contrast to clinical studies, which deal with
individuals in the context of disease and pathological diagnoses,
field research relates to populations as a whole and to subsegments
of the groups. Field applications are in the domains of epidemiology
and public health. The goals are various: (1) to describe the distributions
and prevalences of abnormalities of specific variables; (2) to monitor
spontaneous changes over time (surveillance); (3) to determine the
response to (changes with) interventions; and/or (4) to determine
associations between and among variables.
With respect to any variables, there are a series
of practical considerations and a series of quantitative considerations
in field research, as well as the most fundamental of considerations,
biological meaning.
Practical considerations
The practical considerations in field measurements
relate to ethical considerations and cost considerations.
Ethical considerations
On 4 May 1993, a curious new story broke over the
worldwide CNN network about a scandal related to the dosing of 'radio-active
pills' to Alaskan Indians and Eskimos during US Air Force inspired
research into the adaptation to severe cold during the heart of the
Cold War in the 1950s. 'Were the Eskimos used as guinea-pigs?' is
what the 1990s Alaskan Senator wants to know. Well, we have the issue
of hazard and risk, joining the issue of informed consent, joining
the issue of healthy subjects' research. As the story—embellished
for the lay viewer—was reported, anyone with medical training
realized that what they were talking about was radioiodine thyroid
uptake, possibly with thyroid imaging scans. The Eskimo subjects in
the 1950s w 1000 ere still hunter-gatherers in lifestyle, and generally
were not English-speakers. How 'informed' was any consent they might
have provided, given the linguistic and conceptual barrier? In a clinical
context, one would hardly hesitate to recommend a thyroid-uptake test
and radioscintographic thyroid scan; it is of low risk, and of enormous
diagnostic benefit in clarifying a diagnostic situation. It is of
some risk from long-lasting radiation exposure; of what benefit is
it to the individual healthy Aleutian or Inuit to have the test? In
this particular case, the knowledge gained was to benefit a population,
and since the US Military—not the native Alaskans— was the
favored population, the 40-year-old experience was sufficient to represent
a news headline and a political football.
The exposure to any risk or discomfort in the field
setting can be questioned at any time. All research procedures must
be duly reviewed and assessed by the institutional review board in
the research approval process, but decades from now, a lawyer or news
reporter may reopen the issue. In healthy persons, with no benefits
to gain from a test, an intensive scrutinity must be given and exhaustive
assurance that assent and consent were truly informed must be established.
Related to the ethical considerations are those of
feasibility of application related to the cultural acceptability of
the technique. Populations may accept techniques that are hazardous
to their health, but reject some which have no health implications
because they offend norms of the culture. In anthropometry, the removal
of certain garments of clothing by an observer, even of the same sex,
may impede the obtaining of certain measurements1.
Cost considerations
The cost of a test is the result of the expenses for
the instrument (initial outlay, maintenance) and the operational costs
per measurement (subject cost, labor, disposables). The investment
for the instrument can be huge or trivial, but the more measurements
that are made the lower becomes the average cost of each test, eventually
approaching the ongoing operational expenses. The simplest paradigm
of cost considerations would be to choose the techniques for measurement
that have the absolutely lowest cost; in theory one could maximize
the test performed for a given ceiling of resources. Generally, however,
one discards this paradigm in favor of one that looks at the cost-effectiveness
of learning what one wants to learn from the exercise. This begins
with one's factoring in the minimal population sample-size needed
for statistical confidence limits, and the exact biological questions
that need to be addressed. The final cost paradigm thus becomes a
complex interplay of judgements and considerations. If a test with
a higher validity and precision—but higher cost—obligates
a lower number of measurements, one has a favorable trade-off. Finally,
if a low-cost procedure fails to reliably address the biological issues
of interest, its application is no bargain at any price !
Quantitative considerations
The prime considerations in the quantitative domain
are the accuracy and precision of a measurement, per se, and its reliability
when used in screening to classify individuals diagnostically.
Accuracy
Accuracy is the ability of a measurement system to
achieve the 'right answer' (correct value), as arbitrated by a 'gold
standard' method. It may be necessary to take the average of multiple
measurements to achieve maximal accuracy.
Precision
Precision is the ability of a measurement technique
to produce a consistent, stable value. It is the reproducibility or
repeatability of a measure. This can 1000 be effected by the inherent
biological oscillation of the variable measured, by intrinsic variability
of the measuring instrument, by the stability of technique within
an observer, and/or by the degree of standardization and calibration
between or among multiple measures and or instruments.
Sensitivity and specificity
A measurement, per se, is descriptive, but diagnostically
neutral. However, one gives it diagnostic meaning by establishing
a cut-off criterion, above or below which 'abnormality' is defined.
When it comes to a screening test, one can determine how frequently
the measure correctly detects the abnormal-status condition when it
truly exists, as determined by a gold standard; this is known as the
'sensitivity' of the measure. One is also interested in how frequently
the measure correctly detects the normal-status ;situation, when the
person is truly free of the abnormal condition. This is known as the
'specificity' of the measure.
It is also usual to define another quantitative diagnostic
quality, which is the predictive value of a positive test, meaning
how often is abnormal deviance of the measurement truly reflective
of the condition in question. The converse is the predictive value
of a negative test, meaning how often is the failure to detect the
condition truly an assurance of normalcy.
Biological considerations
Suppose we have a measure that is ethical to measure,
inexpensive to measure, precisely and accurately measured. The proof
of the pudding becomes the meaning, that is, the biological interpretation.
Does the measure or the index faithfully address the question one
is proposing to ask? One concern is 'overweight'. This operationally
would be an inappropriate body mass in relation to height. From normative
curves, one could determine the outliers on the heavy side of the
distribution and classify excess weight. However, overweight may have
different classificatory and health implications if the excess is
muscle tissue or fat mass. A more refined concern is overweight due
to excessive fattiness or adiposity.
What are the low-cost anthropometric options in
field assessment of body composition?
In Table 1 are listed the tools and techniques that
could conceivably be used to assess body composition; not all of them,
of course, are appropriate for the field. Recently, Deurenberg reviewed
an even more extensive list of methods with comments on utility and
cost. The reader is referred to that paper2.
With respect to weight, both accuracy and precision
are important. The intrinsic discrimination of the instrument need
not be greater than 0.5 kg, but in the field context it must be portable,
durable and maintain its calibration. If multiple instruments are
used to gather data for a common data-pool, they must be inter-calibrated
frequently to eliminate systematic bias from site to site. If an 'extravagance'
has to be made, investing in balances that maximize portability, durability
and measurement stability, despite a higher purchase price, is a wise
decision for the field setting.
The instruments for measuring height and surface dimensions
are relatively trivial in price. The professional, manufactured knee-height
caliper is available (Medical Express, Beaverton, Oregon, USA) but
a home-made variety can be made for a lower cost. The measurement
of skinfold thicknesses are generally made with spring calipers, either
with the North American variety, Lange calipers (Lange, Cambridge,
Maryland, USA) or with the European brand, Holtain calipers (Holtain,
UK). Infrared (IR) (refractometry) spectrophotometry has been advanced
as a method for assessing subcutaneous fat. It has been called an
'expensive skinfolds caliper'. Nothing 1000 convinces us the IR refractometry
has any advantage over calipers in assessing subcutaneous fat layers.
Table 1. Classification by categories of techniques
available for the assessment of human body composition.
| ANTHROPOMETRY |
| |
Surface Measurements |
| |
Standing height (stature) |
| |
Armspan |
| |
Knee-height |
| |
Circumferences (diverse
sites) |
| |
Thickness measurements |
| |
Skinfold thicknesses
(diverse sites) |
| DILUTION METHODS |
| |
Deuterium dilution |
| |
Tritium dilution |
| |
Bromide dilution |
| |
Sulfate dilution |
| ELECTROMAGNETIC |
| |
Bioelectrical impedance
analysis (monofrequency) |
| |
Bioelectrical impedance
analysis (multifrequency) |
| |
Total body electroconductivity
(TOBEC) |
| IMAGING |
| |
Single-photon |
| |
Dual-photon |
| |
Dual-X-ray |
| |
Computerized tomography
(CT) |
| |
Infrared refractometry |
| |
Magnetic resonance imaging
(MRI) |
1000
| ENDOGENOUS RADIATION |
| |
Potassium 40 content |
| |
Whole body neutron activation
analysis |
Dilution techniques for water spaces—total body
water; extracellular water; intracellular water—exist. If only
total body water is of interest, the test can be very non-invasive
with collection of urine, saliva or even tears. For the different
dated water spaces, however, in most laboratories a sample of blood
must be obtained. Bioelectrical impedance analysis, especially the
emerging multifrequency BIA, is inexpensive after the initial equipment
is purchased.
Applying low-cost anthropometric measures in field
studies: some experiences from Guatemala
In Guatemala, as in other developing countries, the
first approach has been standard measurements of mass, surface dimensions
and thicknesses. These require only a balance, metric tape and skinfold
calipers, all non-invasive and eminently portable. Over the past eight
years, CeSSIAM has obtained the basic measures in over 1000 elderly
and perhaps 5000 children. In Guatemala, we generate the Quetelet
body mass index (BMI) as the index of comparison for adults and elderly,
given the cut-off criteria for undernutrition4,5 and for
overweight and obesity4. In children, we have generally
related the height and weight data to the internationally accepted
reference standards for the US National Centre for Health Statistics
(NCHS), in terms of weight-for-age, height for-age and weight-for-height.
We have, however, had to temper our acceptance of
the interpretive standards with critical skepticism. Do the same criteria
for critical underweight (chronic energy deficiency) apply all the
way up to the third age? Until functional and/or mortality correlations
are made with BMI in a longitudinal fashion in Guatemala and elsewhere
in the very elderly, we remain skeptical of the <20 kg/m4
or <18.5 kg/m5 cut-off criterion for undernutrition
in the older population.
Armspan is the distance from fingertip to fingertip
with the arms extended laterally, parallel to the ground, to the maximal
extent. As long bones do not change in length with aging, and since
the height/armspan ratio is approximately 1.0 in early adult life,
Dequeker et al.6 recommend this index for determining the
loss of height with advancing age. In two rural Guatemalan communities,
Vasquez et al.7 of CeSSIAM measured this index in 1475
persons across the spectrum of age.
Bioelectrical impedance analysis (BIA) has been advanced
as a method for assessing total body water, and hence the size of
lean body mass. In preschool children of a poor, peri-urban community
in Guatemala, CeSSIAM staff members measured basic anthropometry and
conducted BIA determinations in over 2700 individuals9.
Of interest was the close correlation of the BIA index (height squared/resistance)
with other anthropometric measures. More recently, Manolo Mazariegos
and Micheal McCormick have begun to explore the use of multifrequency
BIA in young children. This is a technique that portends the possibility
of distinguishing intra- and extracellular water and the integrity
of cell membranes with the same, simple non-invasive approach.
Conclusion
Arguably, the diversity of applications and volume
of measurements would be much greater in the field setting, ie for
epidemiological and public health purp 1000 oses, than in the clinical
context in the interaction of patient and practitioner. Even if it
were not for the differentially greater risk of undernutrition in
less developed and underprivileged nations, the fact that the third
world population represents three-quarters of the earth's inhabitants
would incline the opportunities to the developing and transitional
nations. The cost constraints are dual in the third world field setting.
First, it is more difficult to justify the use of scarce resources
on diagnosis—as opposed to intervention and action—in confronting
public health problems. Secondly, when a course of study is elected,
the resources are unlikely to be abundant, and the less expensive
the technology the more justifiable is the investigation. It can honestly
be said that, were it not for low-cost, acceptable body composition
techniques, there would be little opportunity for work in the field
in developing countries.
In planning the field application of techniques for
body composition assessment, the virtues of precision and accuracy
must be weighed. If the purpose is descriptive and comparative, then
the accuracy of a method has primacy. If the study, is longitudinal
with serial measurements, standardization and precision are at a premium.
Many low-cost techniques arise in their biological interpretation.
We attempt to use the measures both to define body composition and
to classify nutritional status. However, as we see emerging through
inter-method comparative calibration exercises, such as the 'Rosetta
Stone' project at Columbia University in New York10, the
interpretations given to the measures and indices may not represent
their true biological reality. Take the following examples: short
stature is not prima facie evidence for 'chronic undernutrition;'
waist/hip circumference ratios may not reflect the intra-abdominal
distribution of fat; body mass index cut-off criteria may not be generalizable
across ethnic groups. Thus, a field worker sees an urgent need to
co-correlate low-cost techniques with their more direct and probing
'gold standards.'
Field workers have long been wedded to anthropometric
techniques. The 'brave new world' before us is to move the paradigm
from nutritional status classification, to an assessment of the composition
of the body at all of the five levels suggested recently by Wang10,
and to do it within budgetary constraints. We must link biology and
engineering and harness their effort to permitting maximum 'resolution'
of body composition with minimum invasiveness and cost.
References
- Solomons NW, Mazariegos M, Mendoza 1. Uses of anthropometry
in the elderly in the field setting with notes on screening in developing
countries. Asia Pacific J Clin Nutr 1993; 2: 15-23.
- Deurenberg R Assessment of body composition. Uses
and misuses. Annual Report 1992. Nestle Foundation for the Study
of the Problems of Nutrition in the World. Nestle Foundation, Laussanne,
Switzerland, 1993: 35-72.
- Conway JM, Norris KH, Bodwell CE A new approach
for the estimation of body composition: infrared interactance. Am
J Clin Nutr 1984;40:1123-1130.
- Garrow JS, Webster J. Quetelet's index (W/H2) as
a measure of fatness. Int J Obesity 1985;9:147-153.
- Ferro-Luzzi A, Sette S, Franklin M, James WPT.
A simplified approach to assessing adult chronic energy deficiency.
Eur J Clin Nutr 1991;
- Dequeker JV, Baeyens JP, Claessen J. The significance
of stature as a clinical measurement of ageing. J Am Geriatr Soc
1969; 15:169-179.
- Vasquex A, Guerrero A-M, L 5dd opez CY, Solomons
NW. Relacion entre el indice talla/brazada y edad en poblaciones
rurales guatemaltecas. Gerontologia (Costa Rica) 1989; 3 (suppl)
:27-36.
- Lukaski HC, Johnson PE, Bolonchuk WW, Lykken Cl.
Assessment of the fat-free mass using bioelectrical impedance measurements
of the human body. Am J Clin Nutr 1985; 41:810-817.
- Mazariegos M, Romero-Abal IC, Valdez C, Graxioso
CF, Solomons NW. Bioelectrical impedance used in association with
anthropometry: experience in a field survey of underprivileged Guatemalan
children. Age and Nutrition 1994; (suppl): 5:90-96.
- Wang Z, Pierson RB Jr, Heymsfield SB. The five
level model: a new approach to organizing body composition research.
Am J Clin Nutr 1992; 56:19-28.
Copyright © 1995 [Asia Pacific Journal of Clinical
Nutrition]. All rights reserved.
Revised:
January 19, 1999
.
to the top
0
HEC
PRESS Publisher
of the 
