Status of anthropometry and body composition data

in elderly subjects3
w Cameron Chumlea and Richard N Baumgartner

ABSTRACT Understanding the normal changes in the body and its composition with
increasing old age and their health implications are important to the health care and nutritional
support of elderly subjects. Distribution statistics for selected body measurements of persons
aged 65-80 y are available from the national health surveys. Recumbent anthropometric tech-
niques and B-mode ultrasound may be applicable to measuring those > 80 y who have diffi-
culty standing or are chair- or bedfast. The problems ofestimating body composition in elderly
subjects could be improved by using a four-compartment model. Noninvasive methods, such
as anthropometry and bioclectric impedance, could be used to predict body composition in
elderly subjects if appropriate equations were available and validated against direct methods.
The most pressing need is for the development of suitable reference data for anthropometry
and body composition in large representative samples of black, white, Hispanic and Oriental
elderly persons in the US. Am J C/in Nuir 1989;50: 1158-66.

KEY WORDS Anthropometry, ultrasound, body composition, bioelectric impedance,

reference data, elderly

Introduction persons, aged 70 y, were children, there were signifi-

cant group and racial differences in their growth and 1ev-
There is little doubt that significant changes occur with els ofnutritional status (4). We would be remiss to think
old age in the body’s composition, its size and shape, but that the effects of these early differences among these
only limited information is available regarding the status groups have decreased in old age; they may have in-
ofthese changes at representative ages for elderly subjects creased. Understanding the normal changes in body size,
(1). Also, the directions and magnitudes of the changes shape and composition with increasing old age, the pa-
that occur in body size and composition with old age are rameters of these changes and their health implications
relatively unknown, and there is no clear understanding is important for the health, nutritional support, and
of the relationship of these changes to health in large pharmacologic treatment of elderly persons in the
numbers of representative elderly persons in the United United States and for the development of appropriate
States (2). The majority of the information that is avail- health guidelines.
able is for whites aged 65-80 y. However, the elderly pop-
ulation aged > 85 y is the fastest growing segment of our
society in terms of numbers and in the consumption of Status of anthropometry
health dollars (3). We are relatively ignorant ofthe nor-
Reference data
mal status and changes that occur in body size and com-
position with old age and their health implications in Mean values and distribution statistics for statures,
persons aged 80 y and especially for black, Hispanic, weights and selected body circumferences, breadths and
or Oriental elderly Americans. skinfold thicknesses of persons aged 65-80 y are avail-
Most individuals experience the normal effects of able from the national health surveys that have been con-
aging on the body personally and from observations of ducted in the United States since the early 1960s (5-1 1).
parents and grandparents. Generally, there is a loss of
stature, decrease in weight, change in skin texture, and
I From the Division of Human Biology, Department of Pediatrics,
loss of muscle tissue. These changes may be universal, Wright State University School ofMedicine, Yellow Springs, OH.
but their expression and incidence may vary consider- 2 Supported by grants HD-12252 and AG-0754l from the National
ably within and between groups of elderly persons of Institutes ofHealth, Bethesda, MD and Ross Laboratories, Columbus,
similar or dissimilar genetic backgrounds. It is important OH.
to remember that when the present cohorts of elderly 3 Reprints not available.

1158 Am J C/in Nuir l989;50:l 158-66. Printed in USA. © 1989 American Society for Clinical Nutrition

Downloaded from https://academic.oup.com/ajcn/article-abstract/50/5/1158/4695343

by Cornell University Library user
on 11 June 2018
 ANTHROPOMETRY AND BODY COMPOSITION IN ELDERLY SUBJECTS 1159

These reference data include information for whites, metric indices of body composition to provide a better
blacks, and in the near future for Hispanics, but the data understanding of the underlying parameters of changes
for 75-80-y-old white and black persons in these surveys in weight, such as alterations in the relative amounts and
are up to 20 y out of date. In the third National Health anatomical distributions of adipose and muscle tissues
and Nutrition Examination Survey, old age is not a re- with old age.
striction on participation (1 1), but additional efforts may Adipose tissue thicknesses decrease on the arm and the
be necessary to provide adequate reference data for those leg with age (30, 35), but the thickness of subcutaneous
persons aged > 80 y. The available reference data for and internal adipose tissues are reported to increase on
stature and weight ofelderly persons aged 80 y is sparse the trunk with age (35-37). Therefore, skinfolds and cir-
or is available from groups that may not be representa- cumference measurements from the limbs decrease, but
tive of the sample of elderly persons living today. For abdominal circumference increases (1 8, 32). These
example, the data of Damon and Stoudt (12) for men, changes could be associated with deterioration of limb
aged 72-91 y, were collected from Spanish-American and abdominal muscle structure or tone as well as
War Veterans. The data of Masters et al, (13) included changes in fat patterning (38). Postmenopausai women
individuals up to age 90 y, but these data were collected are reported to have a more upper-body fat pattern than
in the late l950s, so that the oldest individuals were born premenopausal women so that some changes may have
at about the end of the War Between the States. Also, endocrinologic significance (39). Changes in the elastic-
these data were not collected in a standardized manner ity, hydration, and compressibility of subcutaneous adi-
with appropriate consideration for measurement or tech- pose and connective tissues in elderly subjects can alter
nical errors. For elderly persons in the United States cur- the relationship of skinfold thickness measurements to
rently older than age 80 y, reference data for stature, body composition (40) and the interpretation of indices
weight, and other body measurements are limited to ofadipose tissue distribution.
small samples primarily ofwhite ambulatory individuals
(14-23). Measuring techniques
Changes in stature and weight after about age 50-60 y The collection of anthropometry in elderly subjects
have been well documented in whites, and large changes presents special problems. Standard anthropometric
in these measurements have been reported to be associ- measurements are taken with the participant standing
ated with increased morbidity and mortality in elderly (41), but many elderly persons have difficulty standing
subjects (24-29). The rate ofloss in stature has been re- or maintaining an erect posture and some are chair- or
ported to be between 0.5 and 1 .5 cm per decade, but bedfast due to injury or disease. Stature cannot be mea-
these rates have been estimated from trends or the slopes sured accurately in nonambulatory elderly individuals,
of regressions on age using cross-sectional data and may and it may be difficult to measure in some who are ambu-
be confounded by the secular trends towards increased latory because of increased kyphosis, thinning of carti-
stature (16, 19, 24, 28, 30). Serial studies of anthropome- laginous disks between the vertebrae, bowing ofthe legs,
try in older persons have been limited by the ages of the or a bent-knee gait due to muscular contraction (42). The
subjects or by the restriction ofthe sample to a single sex. measurement of weight in nonambulatory persons re-
In a recent serial study of white healthy, middle-class, quires special, expensive equipment, such as portable
elderly men and women, aged 60-80 y, the rate of de- bed or wheelchair scales. It is true that some elderly per-
crease in stature was estimated to be 0.5 cm/y (3 1). This sons are health conscious and can accurately recall a re-
rate ofloss in stature is much greater than previously re- cent stature or weight measurement (43). For many
ported estimates and is also approximately constant other elderly persons, especially those who are nonam-
across cohorts. The direction ofthe changes in small an- bulatory or in long-term care facilities, recall may be in-
nualized increments in weight varied with age in this accurate or impossible, and suitable measuring equip-
study which agrees with the findings from other studies ment may be absent. For these elderly persons, the esti-
(32-34). The general trend was a decrease in body weight mation of stature and weight with known errors from
with age at a mean rate of 1 .0 kg/decade. anthropometry is the best method at present for provid-
The screening ofthose elderly persons at risk for osteo- ing this information for medical and nutritional assess-
porosis, obesity, or emaciation would be improved by ments (44, 45).
knowledge ofthe normal distributions ofchanges in stat- Recumbent anthropometric techniques have been de-
ure and weight that occur in old age. Elderly persons veloped that are applicable to an elderly individual re-
could be identified who have values of change that cx- gardiess of his or her ability to stand or assume an erect
ceed the expected normal limits for their age and sex. posture (15, 16). Recumbent knee height can be used to
However, investigators studying changes in stature in ci- estimate stature (44), and weight can be estimated using
derly subjects need to be aware ofand account for alter- recumbent measurements of arm and calf circumfer-
ations in spinal curvature. Body weight is easily affected ences, subscapular skinfold thickness and knee height
by short-term environmental aspects of life so that the (45). Recumbent measurements ofarm and calf circum-
detection of significant trends in weight can require a ference, and triceps and subscapular skinfold thicknesses
longer passage of time than for stature. Studies of are reliable and accurate. Values from recumbent mea-
changes in body weight should also include anthropo- surements are not systematically different from those us-

Downloaded from https://academic.oup.com/ajcn/article-abstract/50/5/1158/4695343

by Cornell University Library user
on 11 June 2018
 1160 CHUMLEA AND BAUMGARTNER

ing corresponding standing techniques (46, 47). Limited terol, glucose intolerance and decreased high-density-li-
reference data for recumbent anthropometric measure- poprotein cholesterol (14). Changes in weight in elderly
ments for elderly persons up to age 90 y have been pub- men are associated significantly with changes in fasting
lished (1 5, 16). Triceps and subscapular skinfold thick- glucose, uric acid, and forced vital capacity (57). There
ness are useful indices of subcutaneous adipose tissue is increasing evidence that the anatomical distribution of
thickness and are significantly correlated with total and adipose tissues confounds associations between mea-
percent body fat in children and young adults (48). In sures of obesity and disease and is associated indepen-
elderly subjects, however, other body measurements dently with risk factors for chronic diseases, including
such as circumferences of the trunk may provide more glucose intolerance, hyperlipidemia, and hypertension
important information regarding stores of body fat than (58). Changes in fat-free mass (FFM), TBF, and %BF
the skinfolds since correlations of skinfold thicknesses may also be due in part to decreased levels of physical
with total body fat (TBF) and percent body fat (%BF) activity reported in elderly persons (59). In the present
are reported to be less in older than in younger adults cohorts of elderly persons in the United States, there is
(1 7), and fat redistributes to the trunk with aging (35- possibly a greater change in physical activity levels with
37). Studies ofanthropometry in elderly subjects need to aging in men than in women. Most elderly women have
consider a variety ofthese possible measures, due in part been homemakers, and their life styles have remained
to a need for measures of subcutaneous adipose tissue relatively constant. With retirement, men are frequently
thickness in elderly subjects that are not confounded by reported to spend more time in sedentary activities than
age-related changes that affect measurement. they did before retirement (59).

Ultrasound Measuring techniques

Ultrasound has been explored as an alternative to In discussing measures ofbody composition, it is help-
skinfold calipers for measuring subcutaneous adipose tis- ful to distinguish direct from indirect methods. Direct
sue thicknesses because it should be unaffected by factors methods measure physical properties, chemical or ana-
such as compressibility or thickness that limit caliper tomical constituents that can be used to calculate various
measurements (49). To date, most studies have used A- components ofbody composition. Direct methods of es-
mode ultrasound, but the measurements have been timating body composition and adipose tissue distribu-
found to be generally less reliable than caliper measure- tion can be difficult to apply to elderly subjects for tech-
ments of skinfold thicknesses (49-52). In a study of ci- nical, conceptual, or financial reasons. Underwater
derly men and women, the best interobserver reliability weighing continues to be considered the gold standard
of A-mode ultrasound measurements of subcutaneous among the direct methods. Indirect methods, which in-
adipose tissue thickness was 68%, whereas the worst reli- dude anthropometry and bioelectric impedance are
ability for caliper measurements was 88% (52). A recent nomnvasive and provide less specific measurements that
study, however, suggests that B-mode imaging ultra- can be used to predict body composition if appropriate
sound may be as reliable as skinfold calipers (53). The B- equations are available.
mode ultrasound provides a real-time two-dimensional A possible reason for the absence of large amounts of
image in which the fat-muscle interface can be observed body composition data is that the techniques and equip-
visually on a screen and confounding factors such as skin ment may not be functionally accessible for many elderly
compression, fibrous tissue interfaces, and muscle con- subjects. Underwater weighing can be stressful and
tractions can be considered. The screen image can be fro- difficult to perform for many elderly persons. There is a
zen for measurement with electronic calipers or a hard decrease in the body cell mass resulting from reductions
copy obtained and measured using a digitizer. The use in total body water, but there are conificting reports re-
of B-mode ultrasound for measuring adipose tissue garding changes with age in extracellular water (19, 30,
thicknesses in elderly subjects has not been tested. How- 33, 34, 60, 6 1). FFM estimated from total body water
ever, this method may be more informative for describ- (TBW) may be inaccurate in elderly subjects due to these
ing subcutaneous adipose tissue thickness in elderly sub- changes in TBW and extraceliular fluid volume with age
jects than present caliper measurements of skinfold that are not clearly understood (1 8, 32-34, 62). In addi-
thicknesses. tion, loss ofbone mineral with aging (63) requires differ-
ent assumptions regarding the density of fat-free tissues.
Bone mineral can account for as much as 6% ofthe FFM
Status of body composition in young adult men (64), but it is lower and decreases
substantially following menopause in women (63).
Changes in body composition and fat patterning with
age may be associated with changes in various physio-
Invasive methods
logic functions that affect metabolism, nutrient intake,
physical activity, and risk for chronic diseases (54-56). Methods, such as neutron activation, computed to-
Obesity is a significant health problem among the youn- mography, dual photon absorptiometry, and magnetic
ger elderly and is associated with increased blood pres- resonance imaging (MRI), are invasive and require cum-
sure, plasma lipids and low-density-lipoprotein choles- bersome, expensive equipment and highly trained per-

Downloaded from https://academic.oup.com/ajcn/article-abstract/50/5/1158/4695343

by Cornell University Library user
on 11 June 2018
 ANTHROPOMETRY AND BODY COMPOSITION IN ELDERLY SUBJECTS 1161

FIG 1 Magnetic
. resonance image ofthe abdomen at the umbilicus ofan elderly man.

sonnel. As a partial result ofthese problems, reported ref- whites. Dual photon absorptiometry also has the capac-
erence data for these measures ofbody composition has ity for measuring fat and FFM directly (64).
been limited to those elderly persons who are supposedly
healthy and young enough to participate, primarily those Four-compartment model
persons aged 85 y. Improvements in the design of
equipment and technology may help to increase the col- A major concern is the validity ofthe assumptions un-
lection ofbody composition data from large numbers of derlying estimates of body composition in elderly sub-
elderly persons. jects. To date, most studies of body composition have
Ofthese newer direct methods, MRI and dual photon used the simple two-compartment model or Siri’s equa-
absorptiometry have considerable potential for use in ci- tion (66) which divides the body into fat and fat-free
deny subjects. MRI is associated with little risk and can components on the basis ofbody density from underwa-
provide important information regarding the anatomical ter weighing. This equation is based upon the assump-
changes that occur in muscle and fat. As can be seen in tions that the densities offat and FFM are 0.9 g/mL and
the MRI image ofan elderly person in the Longitudinal 1 10 g/mL,
. respectively (67). The density of fat is sup-
Nutrition in a Healthy Elderly Cohort Study at the Urn- posed to vary little among individuals, but the density of
versity of New Mexico (65), there appears to be large FFM can vary substantially among individuals depend-
amounts of internal adipose tissue in the abdomen and ing upon the relative proportions of its constituents,
fatty infiltration of muscle tissue (Fig 1). Dual photon mainly water, protein, and osseous and nonosseous mm-
absorptiometry is also an important method of quantify- eral. As can be seen in Figure 2, a variation of ±0.02 g/
ing the amounts and changes in bone tissue with age. mL away from the assumed value of 1. 10 g/mL for the
Measures of bone tissues are important because density of FFM can translate into an error of ±5% body
equations for determining body density in elderly sub- fat for an individual with a body density of 1 .05 g/mL.
jects are presently inaccurate due to a lack of knowledge These errors can be compounded in elderly subjects due
of bone density in elderly subjects, especially for non- to reported greater variations among individuals and

Downloaded from https://academic.oup.com/ajcn/article-abstract/50/5/1158/4695343

by Cornell University Library user
on 11 June 2018
 1162 CHUMLEA AND BAUMGARTNER
80

I 20
1.10

D#{231}m i#{176}8 #{149}d’Dffm 1.12

0
1.00 1.05 1.10

Whole Body Density (gfmI)

FIG 2. Changes in the values ofpercent body fat as a function ofchanges in the density ofthe fat-free mass.

changes in amounts of body water and bone mineral an increase in connective tissue this assumption may be
content that affect body density (32-34). questionable in elderly subjects.
The problems of estimating body composition in ci- The use of a four-component model is considered
derly subjects can be improved by using a four-compart- necessary for studies determining body composition in
ment model. The equation for this model is as follows: elderly subjects because of the known and suspected
changes that occur in the body with age. Before this
1/D = F/df + TBW/dw + B/db + P/dp model can be applied, however, it is necessary to estab-
iish the amounts and directions of change with age in
In this equation, l/D equals the sum of the volumes
the densities ofthe body components. Until this infor-
(fractions ofweight/density) for fat (F), total body water
mation is available, estimates or predictions of body
(TBW), total bone mineral (B), and protein plus nonos-
composition in elderly subjects are going to be subject
seous mineral and glycogen (P). In comparison to the
to significant errors of unknown magnitude. In addi-
two-compartment model, the volume of FFM is broken
tion, certain measurements that require active partici-
into three constituents, water, bone mineral, and pro-
pation by the subject may lose accuracy due to deterio-
tein. Other nonosseous minerals and carbohydrates
ration in physical performance with aging. For exam-
which are only a small fraction of the fat-free body
plc, in young adults, the gold standard of underwater
N 1 .5% in young adults) are lumped together with the
weighing is estimated to have at best, a minimum re-
protein fraction. Water comprises the largest fraction of
sidual error of 2.5% for estimates of %BF (62). For ci-
the fat-free body and is assumed to be ‘‘-‘73% ofthe fat-
free volume in young adults. However, studies show that deny persons, this error could be as high as 10-15%
this percentage is somewhat higher in women and in- due to poor performance alone, especially with the use
ofa spring scale instead ofload cells. Conventional ap-
creases with levels ofadiposity (32-34). Reference values
proaches to measuring body composition in young
in adults, aged > 65 y are unknown although some stud-
adults need to be modified and new techniques made
ies suggest decreases with old age (1 9, 32-34, 60, 6 1). A
available or developed for accurately estimating body
decrease in the amount ofwater will increase the overall
composition in elderly subjects.
density of the fat-free body, but a decrease in bone mm-
eral content will also decrease the density. The fraction of
Noninvasive methods
the fat-free body that is composed ofprotein, nonosseous
mineral, and carbohydrate is assumed to be relatively Indirect methods of estimating body composition,
constant, but because ofthe changes in body tissues and such as anthropometry and bioelectric impedance, are

Downloaded from https://academic.oup.com/ajcn/article-abstract/50/5/1158/4695343

by Cornell University Library user
on 11 June 2018
 ANTHROPOMETRY AND BODY COMPOSITION IN ELDERLY SUBJECTS 1163

2 Phase Angle

I
increasing
frequency (F)

LF

Resistance (R)
FIG 3. Bioelectric impedance (Z) as a function ofresistance (R), reactance (Xc), and frequency (F).

applicable in elderly subjects but must be validated small relative to resistance in most biological conductors
against criterion methods. Present prediction equations (71, 72).
using anthropometric and bioelectric variables were de- To date, most body composition studies using bioelec-
veloped for young or middle-aged adults, and they are tric impedance have estimated the volume of TBW or
not likely to be valid when applied to elderly subjects. FFM using resistance measured from the wrist to the an-
Weight and stature are used in indices ofbody mass, such ide to an 8#{174}-A, 50-kHz current and validated against
as weight-for-stature or W/S2 (48), in estimating basal underwater weighing (73-76). The present use of bio-
metabolic rate, and for adjusting various indicators of electric impedance tends to overestimate fatness in lean
nutritional status such as creatinine excretion rate (68). and underestimate it in overweight young and middle-
Circumferences and skinfold thicknesses are accepted as aged adults (77). Recently, the formula ZW/S2 was
important predictors ofmuscle and fat mass and fat pat- shown to provide a less biased index of %BF (78). %BF
tenting (62). Decreases with age in indices of FFM and has also been predicted from anthropometric variables
TBF based on circumferences of the arms and the legs and S2/R using muitivariate regression equations (79).
and skinfold thicknesses have been reported, but primar- The addition of S2/R improved the prediction of %BF
ily for white men and women (17, 18, 32-34, 69, 70). from anthropometry marginally in women, but it did not
provide a significant improvement in men. The validity
Bioelectric impedance ofthe present use ofbioelectric impedance has not been
Bioelectric impedance is a new method for determin- established for elderly subjects (80).
ing body composition based on the resistance of FFM Alternative approaches to estimating body composi-
to a high-frequency, low-amplitude alternating electric tion in elderly subjects from bioeiectric impedance
current that is inexpensive, safe, quick, and highly rei- should be explored thoroughly. For example, the estima-
able. It has the potential of being an ideal method for tion of body composition from the bioelectric imped-
use in measuring body composition in elderly subjects ance of body segments may be particularly useful in ci-
because it makes few physical demands of the subjects. derly subjects. Studies of young adults have shown that
Impedance (Z) is the frequency-dependent opposition of most ofthc whole-body resistance measured convention-
a conductor to the flow ofan alternating electric current ally from the hand to the foot is determined primarily by
and is described in Figure 3 by two vectors, resistance the resistances ofthe arm and the leg(81-83). As a result,
(R) and reactance (Xc), according to the formula Z2 measures of the length and resistance of the arm or the
= R2 + Xc2. Resistance is the pure or real opposition of leg should correlate highly with FFM and could be used
the conductor to the current, whereas reactance is a to estimate body composition when a measurement of
small, additional imaginary resistance produced by ca- stature is difficult to obtain (82). Also, most current stud-
pacitance or the storage ofvoltage by a condenser. In the ies using bioeiectric impedance to estimate body compo-
human body, the current is conducted by electrolytes in sition have ignored reactance. Early studies (84, 85) and
body fluids, most of which are contained in the FFM. recent clinical and experimental studies (86) suggest that
The amount ofresistance (R) is a function ofthe geome- the phase angle, the ratio of reactance (Xc) to resistance
try ofthe body and its electrolytic composition. Accord- (R) at a single frequency of 50 kHz, is associated with
ing to electrical theory, R is proportional to the length of the degree of hydration, fluid distribution, and level of
the conductor squared (L2) divided by its volume (V), or fatness. Bioelectric theory indicates that at low frequen-
R = pL2/V, where p is the specific resistivity and is in- cies (< 5 kHz) the current does not penetrate cell mem-
versely proportional to the number of free electrolytic branes and is conducted only through extracellular fluid
ions (N1) per unit volume, or p = kV/N1 Reactance
. is volume (87). Thus, S2/R1, where R1 is whole-body resis-
produced by cell membranes and tissue interfaces and is tance measured at 5 kHz, should be directly proportional

Downloaded from https://academic.oup.com/ajcn/article-abstract/50/5/1158/4695343

by Cornell University Library user
on 11 June 2018
 1164 CHUMLEA AND BAUMGARTNER

to the volume ofextracellular fluid volume. Some studies representative reference data for anthropometry and
have demonstrated that the ratio of whole-body resis- body composition in elderly subjects are needed before
tance at 5 kHz to resistance at 100 kHz is correlated reliable and valid prediction equations can be developed
highly with the ratio of extracellular fluid volume to for epidemiologic and clinical use. 13
TBW (88, 89).
Finally, bioclectric impedance has the potential of be- References
coming a criterion method ofestimating body composi-
tion. Adipose tissue and bone have very high specific re- 1. Exton-Smith AN. Epidemiological studies in the elderly:
sistivities at the frequencies and currents used in bioelec- methodological considerations. Am J Clin Nutr 1982; 35:1273-9.
tric impedance methods (72), so that the path of 2. Bowman BB, Rosenberg IH. Assessment of the nutritional status
conductance is primarily through extraceliular fluid vol- oftheelderly. Am J Clin Nutr l982;35:l 142-51.
3. Butler RN. Introduction. In: Haynes SG, Fienlieb M, eds.
ume, blood, and intracellular fluid in muscle and organ
Epidemiology ofaging. Washington, DC: US Department Health
tissues (87). In electrical theory, the specific resistivity,
Human Services, 1980. (NIH publication 80-969.)
p, is considered to be a constant physical property for a
4. Boyd E. Origins of the study of human growth. Corvalis, OR:
homogeneous material (72). For the arm or the leg, the University ofOregon Health Sciences Foundation, 1980.
conductive volume closely approximates the muscle vol- 5. Abraham S, Johnson CL, Najjar MF. Weight and height of adults
ume, and for these body segments a fat-free specific resis- 18-74 years of age. Rockville, MD: National Center for Health
tivity can be estimated (90). Although the derivation is Statistics, 1979. (Vital and health statistics series 1 1 [DHEW
more complex, a similar estimate can be obtained for the publication (PHS) 79-1659].)
trunk. Estimates of fat-free specific resistivities can be 6. Johnson CL, Fulwood R, Abraham S. Byrner JD. Basic data on
used together with anthropometric and bioelectric resis- anthropometric measurements and angular measurements of the
hip and knee joints for selected age groups 1-74 years of age.
tance measurements of body segments to calculate total
Rockville, MD: National Center for Health Statistics, 1981. (Vital
FFM (90) and total subcutaneous adipose tissue (91).
and health statistics series 1 1 [DHEW publication (PHS) 81-
1169].)
Conclusions
7. Stoudt HW, Damon A, McFarland R, Roberts J. Weight, height
Anthropometry, recumbent anthropometry, B-mode and selected body dimensions ofadults. Rockville, MD: National
imaging ultrasound, and bioelectric impedance arc re- Center for Health Statistics, 1969. (Vital and health statistics series
cently introduced methods that have considerable poten- 1 1 [DHEW publication (PHS) 8].)
tial for describing the status and changes in body size and 8. Stoudt HW, Damon A, McFarland R, Roberts J. Skinfold, body
composition in elderly subjects, including paraplegic, girths, biacromial diameter and selected anthropometric indices of
chair-, or bedfast individuals. These methods are also ap- adults. Rockville, MD: National Center for Health Statistics, 1970.
(Vital and health statistics series 1 1 [DHEW publication (PHS)
propriate for epidemiologic surveys. For these methods
35].)
to be useful in predicting body composition, however,
9. Engel A. Osteoarthritis and body measurements. Rockville, MD:
they must be validated against body composition esti- National Center for Health Statistics, 1968. (Vital and health
mates from other criterion methods. Because estimation statistics series 1 1 [DHEW publication (PHS)].)
equations are commonly sample-specific, it is necessary 10. Najjar MF, Rowland M. Anthropometric reference data and
to derive them using appropriate statistical methods and prevelance of overweight. Rockville, MD: National Center for
to cross-validate them using large independent samples. Health Statistics, 1987. (Vital and health statistics series 11
Ifthese methods are to achieve widespread use, it is desir- [DHEW publication (PHS) 87-1688].)
able to demonstrate their applicability to nonambulatory 1 1. Woteki CE, Briefel RR, Kuczmarski RM. Contributions of the
elderly persons ofvarious levels ofage and impairment. National Center for Health Statistics. Am J Clin Nutr I988;47:
The development of accurate, inexpensive, noninva- 320-8.
12. Damon A, Stoudt HW. The functional anthropometry ofold men.
sive, portable methods of estimating body composition
Hum Factors 1963; 5:485-91.
and fat patterning in elderly subjects will facilitate greatly
13. Masters AM, Lasser RP, Beckman G. Tables ofaverage weight and
future investigations of the complex relationships be- height ofAmericansaged 65 to 94 years. JAMA l960;20l:658-62.
tween changes with age in body composition and nutri- 14. Yearick ES. Nutritional status ofthe elderly: anthropometric and
tional status, physiological, and psychological health, re- clinical findings. J Gerontol 1978; 33:657-62.
sistance to acute and chronic diseases, and longevity. Fu- 15. Chumlea WC, Roche AF, Mukherjee D. Nutritional assessment in
ture work should emphasize comparisons between the elderly through anthropometry. 2nd ed. Columbus, OH: Ross
elderly men and women for biological, social and behav- Laboratories, 1984.
ioral factors affecting changes with age in body composi- 16. Chumlea WC, Steinbaugh ML, Roche AF, Mukherjee D,
tion. However, the most pressing need is for the develop- Gopalaswamy N. Nutritional anthropometric assessment in
elderly persons 65 to 90 years ofage. J Nutr Elder l985;4:39-5 1.
ment of reference data for anthropometry and body
17. Chumlea WC, Roche AF, Webb P. Body size, subcutaneous
composition in large representative samples of black,
fatness and total body fat in older adults. Int J Obes 1984;
white, Hispanic, and Oriental elderly persons. These
8:311-7.
data should be sex-specific and by levels ofmobility. The 18. FriedlanderJS, Costa PT, Bosse R, Ellis E, RhoadsiG, Stoudt HW.
body composition data should be developed using a four- Longitudinal physique changes among health white veterans at
compartment model and divided into cohorts of elderly Boston. Hum Biol 1977;49:54l-58.
persons so as to remove survivor effects. Reliable and 19. Norris AH, Lundy 1, Shock NW. Trends in selected indices of

Downloaded from https://academic.oup.com/ajcn/article-abstract/50/5/1158/4695343

by Cornell University Library user
on 11 June 2018
 ANTHROPOMETRY AND BODY COMPOSITION IN ELDERLY SUBJECTS 1165

body composition in men between the ages 30 and 80 years. Ann 43. Heaney RP, Ryan R. Relation between measured and recalled
NY Acad Sci 1963; 1 10:623-39. body height. N EngI J Med l988;3l9:795-96.
20. McEvoy AW, James OFW. Anthropometric indices in normal 44. Chumlea WC, Roche AF, Steinbaugh ML. Estimatingstature from
elderly subjects. Age Ageing 1982; 1 1:97-100. knee height for persons 60 to 90 years of age. J Am Geriat Soc
21. Vir SC, Love AHG. Anthropometric measurements in the elderly. 1985;33:l16-20.
Gerontology 1980;26: 1-8. 45. Chumlea WC, Guo 5, Roche AF, Steinbaugh ML. Prediction of
22. Stoudt HW. The anthropometry of the elderly. Hum Factors body weight for the nonambulatory elderly from anthropometry.
198 l;23:29-36. J Am Diet Assoc l988;88:564-8.
23. Damon A, Seltzer CC, Stoudt HW, Bell B. Age and physique in 46. Chumlea WC, Roche AF, Rogers E. Replicability for anthro-
healthy white veterans at Boston. Gerontology 1972;27:202-8. pometry in the elderly. Hum Biol 1984; 56:329-37.
24. Dequeker JV, Baeyens JP, Claessens J. The significance of stature 47. Chumlea WC, Roche AF, Steinbaugh ML, Mukheijee D. Errors
as a clinical measurement of ageing. J Am Geriatr Soc 1969; 17: of measurement for methods of recumbent nutritional anthro-
169-79. pometry in the elderly. J Nutr Elder 1985;5 1:3-1 1.
25. Masoro EJ. A critical review: nutrition and aging-a current 48. Roche AF, Siervogel RM, Chumlea WC, Webb P. Grading body
assessment. J Nutr 1985; 1 15:842-8. fatness from limited anthropometric data. Am J Clin Nutr
26. Mial WE, Ashcroft MT, Lovell HG, Moore F. A longitudinal study l98l;34:283l-8.
ofthe decline ofadult height with age in two Welsh communities. 49. Kuczmarski RK, Fanelli MT. Koch GG. Ultrasonic assessment of
Hum Biol l967;39:445-54. body composition in obese adults: overcoming the limitations of
27. Munro HN. Nutrition and aging. Br Med Bull 198 1;37:83-99. the caliper. Am J Clin Nutr l987;45:7l7-24.
28. Hertzog KP, Gain SM, Hempy HO. Partitioning the effects of 50. Borkan GA, Hults DE, Cardarelli J, Burrows BA. Comparison
secular trend and aging on adult stature. Am J Phys Anthropol of ultrasound and skinfold measurements in assessment of
l969;3l:1 11-6. subcutaneous and total fatness. Am J Phys Anthropol l982;58:
29. Forbes GB. The adult decline in lean body mass. Hum Biol 307- 13.
I976;48: 161-73. 51. Fanelli MT. Kuczmarski Ri. Ultrasound as an approach to
30. Young CM, Blondin J, Tensuan R, Fryer JH. Body composition assessing body composition. Am J Clin Nutr 1984; 39:703-9.
studies of “older” women, thirty to seventy years of age. Ann NY 52. Chumlea WC, Roche AF. Ultrasonic and skinfold caliper
Acad Sci 1963; 1 10:598-607. measures ofsubcutaneous adipose tissue thickness in elderly men
31. Chumlea WC, Garry PJ, Hunt WC, Rhyne RL. Serial changes in and women. Am J Phys Anthropol l986;7l:35l-7.
stature and weight in a healthy elderly population. Hum Biol 53. Weits T, Van Der Beck ES, Wedel M. Comparison of ultrasound
l988;60:9l8-25. and skinfold caliper measurement ofsubcutaneous fat tissue. mt J
32. Noppa H, Andersson M, Gengtsson C, Bengtsson C, Bruce A, Obes 1986; 10:161-8.
Isaksson B. Longitudinal studies ofanthropometric data and body 54. Munro HN. Nutrition and aging. Brit Med Bull 198 1;37:83-8.
composition: the population ofwomen in G#{246}teborg,Sweden. Am 55. Young EA. Nutrition, aging and the aged. Med Clin North Am
J Clin Nutr 1980; 33:155-62. l983;67:295-3l 1.
33. Steen B, Bruce A, Isaksson B, Lewin T, Svanborg A. Body 56. Durnin JVGA. Energy intake, energy expenditure, and body
composition in 70-year-old males and females in Gothenburg, composition in the elderly. In: Chandra RK, ed. Nutrition,
Sweden: a population study. Acta Med Scand l977;6 I l(suppl):87- immunity and illness in the elderly. New York: Pergamon Press,
112. 1985: 19-33.
34. Steen GB, Isaksson B, Svanberg A. Body composition at 70 and 75 57. Borkan GA, Sparrow D, Wisniewski C, Vokonas PS. Body weight
years ofage: a longitudinal population study. J Clin Exp Gerontol and coronary disease risk: pauerns of risk factor change associated
1979; 1:185-200. with long-term weight change: the Normative Aging Study. Am J
35. Enzi G, Gasparo M, Biondetti PR, Fiore D, Semisa M, Zurlo F. Epidemiol 1986; 124:410-9.
Subcutaneous and visceral fat distribution according to sex, age, 58. Stern M, Haffner S. Body fat distribution and hyperinsulinemia as
and overweight, evaluated by computed tomography. Am J Clin risk factors for diabetes and cardiovascular disease. Arteriosclerosis
Nutr l987;45:7-l3. l986;6: 123-30.
36. Borkan GA, Hults DE, Gerzof SG, Robbins AH. Comparison of 59. Patrick JM, Bassey El, Fentem PH. Changes in body fat and
body composition in middle-aged and elderly males using muscle in manual workers at and after retirement. Eur J AppI
computed tomography. Am J Phys Anthropol l985;66:289-95. Physiol l982;49:187-96.
37. Baumgartner RN, Heymsfield SB, Roche AF, Bernardino M. 60. Fryer JH. Studies of body composition in men aged 60 and over.
Quantification of abdominal composition by computed tomog- In: Shock NW, ed. Biological aspects of aging. New York:
raphy. Am J Clin Nutr l989;50:22 1-6. Columbia Univ Press, 1962:59-79.
38. Borkan GA, Norris AH. Fat redistribution and the changing body 61. Friis-Hansen B. Hygrometry ofgrowth and aging. Symp Soc Hum
dimensions ofthe adult male. Hum Biol l977;49:495-5l4. Biol l965;7: 19 1-209.
39. Raison J, Basdevant A, Guy-Grand B. Relation entre repartition 62. Behnke AR, Wilmore JH. Evaluation and regulation ofbody build
du tissu adipeux et lipide circulation chez Ia femme obese. La and composition. Englewood Cliffs, NJ: Prentice Hall, 1974.
Presse M#{233}dicale 1987; 16:43 1-3. 63. Hansson T, Roos B. Age changes in the bone mineral ofthe lumbar
40. Grahame R. A method for measuring human skin elasticity in vivo spine in normal women. CalcifTissue mt 1986; 38:249-51.
with observations on the effects ofage, sex and pregnancy. Din Sci 64. Mazess RB, Peppler WW, Gibbons M. Total body composition by
1970;39:223-38. dual-photon (1 53Gd) absorptiometry. Am J Clin Nutr l984;40:
41. Lohman TG, Roche AF, Martorell R, eds. Anthropometric 834-9.
standardization reference manual. Champaign, IL: Human 65. Garr PJ, Goodwin JS, Hunt WC, Hooper EM, Leonard AG.
Kinetics Books, 1988. Nutrition status in a healthy elderly population: dietary and
42. Trotter M, Gleser G. The effect of aging on stature. Am J Phys supplemental intakes. Am J Clin Nutr 1982; 36:319-31.
Anthropol 195l;9:3l 1-24. 66. Siri WE. Body composition from fluid spaces and density analysis

Downloaded from https://academic.oup.com/ajcn/article-abstract/50/5/1158/4695343

by Cornell University Library user
on 11 June 2018
 1166 CHUMLEA AND BAUMGARTNER

ofmethods in techniques for measuringbody composition. Brozek composition predictions from bioclectric impedance. Hum Biol
J, Henschel A, eds. Washington, DC: National Academy of l987;59:22 1-34.
Sciences, 1961. 80. Chumlea WC, Roche AF, Guo S, Woynarowska B. The influence
67. Pace N, Rathbun EN. Studies on body composition. III. The body of physiologic variables and oral contraceptives on bioclectric
water and chemically combined nitrogen content in relation to fat impedance. Hum Biol l987;59:257-70.
content. J Biol Chem 1945; 158:685-91. 81. Settle RG, Foster KR, Epstein BR, Mullen JL Nutritional
68. Blackburn GL, Bistrian BR, Maini BS, Schlamm HT, Smith MF. assessment: whole body impedance and body fluid compartments.
Nutritional and metabolic assessment ofthe hospitalized patient Nutr Cancer I980;2:72-80.
JParenterEntcralNutr 1977;l:l-9. 82. Baumgartner RN, Chumlea WC, Roche AF. Estimation of body
69. Chumlea WC, Roche AF, Mukbeijce D. Some anthropometri composition from bioclectric impedance ofbody segment& Am J
indicesofbodycomposition forekierlyadults. J Gerontol 1986;41: Clin Nutr l989;50:22l-6.
36-9. 83. Rangathan C, Engel R, Berkseth R. Measurement of body fluid
70. Masoro El. Biology ofaging. Arch list Med 1987; 147:166-9. volume changes using multisite impedance measurements. Med
71 . Nyboer J. Electrical impedance plethysmography. Springfield, IL: Biol EngComput l988;26:33-7.
CharlesCThomas, 1959. 84. Brazier MAB. The impedance angle test for thyrotoxicosis.
72. Pethig R. Dielectric and electronic properties of biological Western J Surg Obstet Gynecol 1935;43:5 14-27.
materials. New York: J Wiley and Sons, 1979. 85. Barnet A, Bagno S. The physiological mechanism involved in the
73. Hoffer EC, Meador CK, Simpson DC. A relationship between clinical measure ofphase angle. Am J Physiol 1936; 114:366-82.
whole body impedance and total body water volume. Ann NY 86. Baumgartner RN, Chumlea WC, Roche AF. Bioelcctric
AcadSCI l970;IlO:452-6l. impedance, phase angle, and body composition. Am J Clin Nutr
74. Kushner RF, Schoeller DA. Estimation of total body water by l988;48: 16-23.
bioelectrical impedanceanalysis. Am J Gin Nutr 1986;44:417-24. 87. Ackmann JJ, Seitz MA. Methods of complex impedance
75. Segal KR, Gutin B, Presta E, Wang J, Vanltallie TB. Estimation measurements in biologic tissue. Crit Rev Biomed Eng 1984; 11:
of human body composition by electrical impedance methods a 281-3 11.
comparative study. J AppI Physiol 1985;58:1565-7l. 88. Jenin P, Lenoir J, Roullet C, Thomasset A, Ducrot H.
76. Lukaski HC, Johnson PE, Bolonchuk WW, Lykken GI. Determination of body fluid compartments by electrical imped-
Assessment of fat-free mass using bioelectrical impedance ance measurements. Aviat Space Environ Med 1975;46:l52-5.
measurements of the human body. Am J Clin Nutr 1985; 89. Thomasset A. Bioelectrical properties of tissue impedance
41:810-7. measurements. Lyon Med l962;207: 107-18.
77. Hodgdon JA, Fitzgerald P1. Validity of impedance predictions at 90. Chumlea WC, Baumgartner RN, Roche AF. The use of specific
variouslevels offatness. Hum Biol l987;59:281-98. resistivity to estimate fat-free mass from segmental body measures
78. Khaled MA, McCutcheon MJ, Reddy S. Pearman PL, HunterGR, ofbioelectric impedance. Am J Clin Nutr l988;48:7-15.
Weinsier RL Electrical impedance in assessing human body 91. Chumlea WC, BaumgartnerRN, Roche AF. Fat volumesand total
composition: the BIA method. Am J am Nutr 1988;47:789-92. body fat from impedance. Med 8th Sports Exeic l988;20S82
79. Guo S, Roche AF, Chumlea WC, Miles DS, Pohlman RL Body (abstr).

Downloaded from https://academic.oup.com/ajcn/article-abstract/50/5/1158/4695343

by Cornell University Library user
on 11 June 2018