Comparison of Anthropometry to DXA in Men: A Validation Study A Thesis Presented to The Faculty of the Graduate School At the University of Missouri In Partial Fulfillment Of the Requirements for the Degree Master of Science By Celsi Cowan Dr. Steve Ball, Thesis Supervisor MAY 2013 The undersigned, appointed by the dean of the Graduate School, have examined the thesis entitled Comparison of Anthropometry to DXA in Men: A Validation Study Presented by Celsi Cowan A candidate for the degree of Master of science And hereby certify that, in their opinion, it is worthy of acceptance. Dr. Steve Ball Dr. John Thyfault Dr. Thomas LaFontaine Acknowledgements
I would like to thank my committee members for agreeing to be on my committee
and helping me through this project. Dr. Steve Ball, you have been a fantastic advisor
over the last three years and have been so wonderful to work with. Dr. John Thyfault,
thank you so much for serving on my committee. Dr. Tom Lafontaine, thank you for
everything I have learned over the last three years. I have learned so much at Optimus
working under you and have a great appreciation for working in clinical exercise
physiology. You are a great mentor!
I also would like to thank my undergraduate assistants. Ashley Friedland, Rachael
Grassmuck, and Melanie Stone, I could have not done it with out you!
And finally, to my husband Dan Leuchtman, I could have not made it with out
you. Your love and support got me through it all! Thanks for always encouraging me to
keep going.
ii
TABLE OF CONTENTS
Acknowledgements………………………………………………………………..ii
List of Tables and Figures………………………………………………………...iv
Abstract…………………………………………………………………………....v
Introduction. ………………………………………………………………………1
Methods.…………………………………………………………………………...3
Results……………………………………………………………………………..7
Discussion………………………………………………………………………...11
Conclusion………………………………………………………………………..16
References………………………………………………………………………..17
Appendix A: Extended Literature Review……………………………………….20
Appendix B: %BF Table for DC Equation………………………………………47
iii
LIST OF TABLES AND FIGURES
Table 1. Skinfold equations for men.…………………………………………………..5
Table 2. Descriptive statistics for subjects …………………………………………….7
Table 3. Mean %BF with mean % difference and correlation related to DXA………..8
Table 4. Mean difference %BF from DXA across fatness quartiles………………….11
Table 5. Mean difference of %BF from DXA across age quartiles ………………….11
Table 6. Lohman percent fat errors and prediction errors…………………………….12
Table 7. Descriptive statistic comparisons among Cowan, Ball, and JP samples…….14
Table 8. Comparison between studies for the mean %BF difference from DXA…….14
Figure 1. Difference against mean for %BF: DXA & DC……………………………..9
Figure 2. Difference against mean for %BF: DXA & JP7……………………………10
iv
Comparison of Anthropometry to DXA in Men: A Validation Study
ABSTRACT
INTRODUCTION. Accurate and valid body composition methods are needed in order
to assess percent body fat (%BF) but the accuracy of anthropometric equations is limited
by the criterion method from which they are created. Current equations by Jackson and
Pollock (JP), recommended by the American College of Sports Medicine (ACSM), have
been shown to underestimate body composition due to the assumptions of the 2C model
(2, 10, 16, 23). The more recent DXA criterion (DC) skinfold equation created by Ball et
al. is based on dual X-ray absorptiometry (DXA), a three-component model (3C), which
has been shown to be more accurate than the previous 2C model (14). The purpose of this
study is to validate the DC skinfold equation. The secondary purpose of this study is to
compare the currently recommended skinfold equations to DXA. METHODS. Two
hundred ninety-seven male subjects, aged 18-65, completed a seven site skinfold
assessment and one DXA scan to determine %BF. Three JP equations and the DC
equation were used to predict %BF (2, 15, 17, 30). RESULTS. Two hundred and seventy
two subjects were included in the analysis. Mean age was 32.4 ± 14.0 y and mean BMI
was 25.6 ± 3.3 kg/m2. The mean DXA %BF was 18.0 ± 5.9. The mean %BF for skinfold
equations DC, JP7, JP3a, and JP3b were 19.1 ± 6.3, 16.1 ± 7.4, 14.8 ± 6.8, 15.6 ± 6.7
respectively. The standard error of the estimate (SEE) of DC was low (2.72%) and was
highly correlated with DXA. CONCLUSION. The DC equation was found to better
predict %BF across a general population than three JP SF equations when compared to
the 3C model, DXA. This study suggests that the DC equation is a more accurate
prediction equation than the current ACSM recommended equations.
v
INTRODUCTION
Excess body fat (BF) has been shown to be associated with many chronic diseases
such as type 2 diabetes mellitus, coronary artery disease, hypertension, metabolic
syndrome, cancers of the breast, prostate, and colon, stroke, hyperlipidemia and all-cause
mortality (4, 8, 12, 18, 25). Individual assessment of excess BF is thus critical for
determining chronic disease risk. Accurate and valid body composition methods that are
inexpensive and practical are needed in order to help design safe, effective, and objective
nutritional and exercise programs for overweight and obese individuals as well as
athletes. Valid, reliable and sensitive methods are also needed to track progress over time
(13).
Unfortunately, there is no direct way to measure body composition in vivo, as
cadaver dissection is the only accurate method. Indirect methods, then, are necessary for
predicting body composition in both the field and the laboratory. Several models are
available to measure body composition. The most common is the two component (2C)
model which separates the body into fat free mass (FFM) and fat mass (FM) (3). The 2C
model assumes that within FFM, water, protein, and mineral proportions are constant
across all individuals (22, 28). Under these assumptions, the density of FFM was
calculated as 1.1000 g/cm3 (7). However, it has been found that the proportions of water,
protein, and mineral are not constant across all individuals (31). Thus, three, four, and
five component models (3C, 4C, and 5C) have proven to be more accurate since they
measure body water, protein, and mineral proportions (13).
The most common anthropometric methods (skinfolds) used in field settings have
been developed from hydrostatic weighing (HW), a 2C model. HW uses densitometry to
1
find %BF and has previously been termed the “gold standard”. The Jackson and Pollock
skinfold (SF) equations were developed based on this model and continue to be
recommended by the American College of Sports Medicine (ACSM) (30). Because the
body is made up of more than two components, anthropometric measurements based on
2C are limited since they cannot accurately predict the density of bone and water within
the body. Thanks to new technologies and advancements in the area of body composition
research, multi-component models, such as the 4C method, have been termed the “new
gold standard” as they do not have as many assumptions as the 2C method. The 4C
method uses a combination of methods to measure fat mass, total body water, bone
mineral mass, and residual protein and other non-bone material.
Dual energy x-ray absorptiometry (DXA), a 3C model, is becoming a more
popular laboratory method. The DXA uses X-rays to scan the body and divide it into
three components: fat mass, lean mass, and bone mineral mass (23). It has been shown
that bone density differs across individuals, leaving the 2C model with another significant
source of error. The DXA has been termed the “practical gold standard” since it is quick,
accurate, valid, and requires little subject compliance (10, 13).
Several studies have shown SF based on the 2C model that underestimate body
composition when compared to DXA (1, 2, 9, 15, 31). However, anthropometrics remain
an important tool for clinicians and health professionals in the field, as these methods are
inexpensive and fast. By developing an anthropometric equation from a 3C model, field
methods have the ability to bring their accuracy closer to that of laboratory models such
as DXA.
2
In 2004, Ball et al. published a new skinfold prediction equation using DXA as
the criterion model (DC) (2). This study used 160 men, aged 18-62 y and found that %BF
calculated with the original Jackson & Pollock prediction equations was significantly
underestimated (3.1-3.3%) when compared to DXA. Ball et al. also found similar results
in a sample of women when comparing the skinfold method to DXA (1). This revealed a
need for a new equation based on the newer technology of the 3C model and thus, the DC
equation was developed. The DC equation was found to predict %BF better than three
generally recommended equations by Jackson & Pollock when compared to %BF
measured by DXA in men (2). The mean difference between DC and DXA was small
(0.5%), the SEE was excellent (2.2%), and the correlation was significant (0.89).
However, no cross validation study has been done since this equation was published. A
larger sample size is needed before the DC equation can become the standard for
measuring body composition in the field.
It is hypothesized that this new DC equation, when cross-validated on a larger,
independent sample, will more accurately predict %BF than previously recommended
anthropometric equations by Jackson and Pollock. The purpose of this study is to crossvalidate the DC SF equation in a sample of young men. The secondary purpose is to
compare the currently recommended Jackson and Pollock skinfold equations to DXA.
METHODS
Subjects. Two hundred ninety-seven subjects were recruited for the study. All
subjects signed an informed consent in accordance with the University of Missouri
3
Human Subjects Institutional Review Board. Subjects were male between the ages 18-65
y and completed the study with 100% compliance rate.
Subjects were skinfolded at 7 sites (subscapular, tricep, mixaxillary, chest,
abdomen, suprailiac, and thigh) and underwent one whole body DXA scan. Subjects were
instructed not to eat two hours or exercise four hours before testing. All measurements
were completed within one hour of each other on the same day.
Anthropometric measurements. Height, weight, waist circumference, and hip
circumference were measured. Height (Seca 216, Seca gmbh & co. kg., Hamburg,
Germany) was taken to the nearest 0.005 meter (m) and weight (Toledo scale, MettlerToledo Inc., Columbus, OH, USA) was taken to the nearest 0.5 kilogram (kg). Waist
circumference (around the narrowest part of the waist between the umbilicus and rib
cage) and hip circumference (the largest protrusion of the buttocks) were measured to the
nearest 0.5 cm using a Medco Tape Measure (Medco Sports Medicine, Tonawanda, NY,
USA). Waist to hip ratio (waist circumference/ hip circumference) and BMI (kg/m2) were
calculated (30).
Seven skinfold sites (subscapular, midaxillary, tricep, chest, abdomen, suprailiac,
and thigh) were measured using a Lange caliper (Cambridge Scientific Industries,
Cambridge MD, USA). Each site was measured twice and averaged. If the first two
measurements were greater than 2 mm apart, a third measurement was taken and the
average of the two closest measurements was calculated (30). Four skinfold equations
were used (Table 1). The first three equations were developed by Jackson and Pollock
(JP7, JP3a, and JP3b) and were used to find body density (Db) (13, 15). The Siri equation
was then used to calculate percent body fat from Db (29). The DC equation was the forth
4
skinfold equation used in which %BF was directly calculated (2).
DXA. Subjects completed one whole body DXA scan on one of the two different
models (QDR 4500A or QDR Discovery-A, Hologic, Inc., Bedford, MA, USA) per
manufacturer directions and wore minimal clothing with no metal. The subjects were
placed supine on the DXA table and were positioned according to manufacturer
instructions (17). Body composition was found using computer software (QDR Software
for windows XP, Hologic, Inc., Bedford, MA) and results for fat, lean, and bone mineral
mass were recorded in grams.
Table 1. Skinfold equations for men.
Equation
JP7
(chest, midaxillary, triceps, thigh,
Formula
Body density = 1.112 - 0.00043499 (Σ7SF) + 0.00000055
(S7SF)2 - 0.00028826 (age)
subscapular, suprailiac, abdomen)
JP3a
(chest, abdomen, thigh)
JP3b
(chest, triceps, subscapular)
DC
(chest, midaxillary, triceps, thigh,
Body density = 1.10938 - 0.0008267 (Σ3SF) + 0.0000016
(S3SF)2 - 0.0002574 (age)
Body density = 1.1125025 -0.0013125 (Σ3SF) + 0.0000055
(S3SF)2 - 0.000244 (age)
%BF=0.465 + 0.180(Σ7SF) - 0.0002406(Σ7SF)2 +
.0.06619(age)
subscapular, suprailiac, abdomen)
Σ7SF= sum of seven SF sites; Σ3SF= sum of three SF sites; JP7= 7 site equation by
Jackson and Pollock (16); JP3a= 3 site SF equation by Jackson and Pollock (16); JP3b= 3 site SF
equation by Jackson and Pollock (14); DC= DXA criterion equation (2).
5
Statistical analysis. All statistical analyses were done using SPSS (version 20.0).
Subjects included in these analyses were drawn from the current study, Company & Ball
(10) and Ryder & Ball (26). All data were collected in the same laboratory.
DXA was used as the criterion measure against which all other body composition
methods were compared. Descriptive statistics were used to describe the subjects.
Differences between the means of each group compared to DXA were assessed using
ANOVA. A post hoc analysis using a Dunnetts T test was done to compare each mean of
the equations DC, JP7, JP3a, and JP3b to DXA. Bland Altman plots were also used to
compare each skinfold method to the DXA (6). The standard estimation of error (SEE)
was found to examine the accuracy of the DC equation.
The reliability of DXA and SF’s were tested by measuring ten subjects twice.
These duplicate tests were repeated in the same day within one hour of each other.
Between DXA scans, each subject was removed from the table and repositioned
according to manufacturer instructions. A paired samples T test was used to find any
significant differences.
Since intertester variability can contribute to a majority of the error in SF
measurements, all three technicians (Cowan, Ryder, and Company) were compared. The
sum of 7 site skinfolds was taken in 10 randomly chosen subjects (30 total) by the three
researchers.
6
RESULTS
Participants. Descriptive statistics for 272 subjects that completed the study and
included in the analyses are presented in Table 2. Mean age was 32.4±14.0 y and mean
BMI was 25.6± 3.3 kg/m2. A majority of the subjects were non-Hispanic white and
included 1% Hispanic, 2% African American, and 2% Asian. Twenty-five subjects
completed the study but were excluded from analysis. Exclusion criteria were: BMI ≥ 35
kg/m2 (13 subjects), ≥ 3 standard deviations from the mean (3 subjects), or missing data
(9 subjects).
Table. 2 Descriptive statistics for subjects
n=272
Variable
Mean ± SD
Age (y)
32.4±14.0
Body height (m)
1.8±0.1
Body mass (kg)
82.1±11.9
2
Body Mass Index (kg/m )
25.6±3.3
Waist (cm)
85.4±9.0
Hip (cm)
97.5±6.3
WHR
0.87±0.41
Body Fat % (DXA)
18.0±5.9
Σ7 Skinfolds
Σ3a Skinfolds
Σ3b Skinfolds
111.9±50.9
48.9±21.5
39.0±17.3
Range
18-65
1.5-2.0
58.5-120.0
16.9-33.9
62.0-114.0
83.0-119.0
0.78-1.06
8.3-36.4
38.0-289.0
17.0-122.5
14.5-105.0
Body fat %. The mean %BF for each method is described in Table 3. The mean
DXA %BF was 18.0±5.9. The mean %BF for skinfold equations DC, JP7, JP3a, and JP3b
were 19.1 ± 6.3, 16.1 ± 7 .4, 14.8 ± 6.8, 15.6 ± 6.7, respectively. All prediction equations
were significantly correlated with DXA, as shown in Table 3.
7
Table 3. Mean %BF with mean % difference and correlation related to DXA.
Variable
Mean %BF ±
Mean % difference
SD
from DXA
Sig
Correlation with
DXA
DXA
18.0 ± 5.9
DC
19.1 ± 6.3
+ 1.1
.145
0.934**
JP7
16.1 ± 7.4
-1.9*
.004
0.937**
JP3a
14.8 ± 6.8
-3.2*
.000
0.926**
JP3b
15.6 ± 6.7
-2.4*
.000
0.901**
*Significantly different from DXA. p<.05
**Significantly correlated to DXA. p<.05
The ANOVA showed that Ho was rejected, meaning there was a significant
difference among groups. An a posteriori test (Dunnett’s T test) was used to find which
groups were different. JP7, JP3a, and JP3b were found to be significantly different from
DXA (Table 3). The DC equation was not found to be significantly different from DXA.
Using the same predictors as Ball’s sample, the SEE was found to be 2.72%.
Reliability. The mean %BF of two DXA trials for 10 subjects was 23.4 ± 2.4 and
23.5 ± 2.4. No difference between trials was found (p<.05). The mean difference between
trials was 0.09 ± 0.08 %BF with the trials highly correlated (r=0.99). Using the same 10
subjects, the sum of the seven SF site for each trial had means at 117.3 ± 6.6 mm and
117.5 ± 7.4 mm. No significant difference was found between trials and the two trials
were significantly correlated (r=0.99). Finally, no statistically significant difference in the
sum of a seven site skinfold occurred among the three skinfold testers (p<.05).
Differences between methods. Bland Altman plots (Figures 1 and 2) were used
8
to compare the JP7 and DC equations to DXA. Each data point represents the difference
between each method for each subject. The vertical axis represents the difference of the
two measurements and the horizontal axis represents the average of the two
measurements. If two methods are comparable, more data points will be closer to zero
meaning the differences are small (6). In Figure 2, the data points span across the vertical
axis much farther than Figure 1 meaning the difference between DXA and JP7 are larger
than the difference between DXA and DC. Since the JP3a and JP3b means differed more
from DXA than from JP7, similar results would be expected, therefore, no Bland Altman
plots are depicted for these two equations.
Figure 1- Difference against mean for %BF: DXA & DC. Negative values represent an
overestimation compared to DXA.
9
Figure 2- Difference against mean for %BF: DXA & JP7. Positive values represent an
underestimation in relation to DXA.
%BF across quartiles. The difference of mean %BF for JP7 and DC was
compared to DXA across different levels of fatness. Table 4 shows the mean differences
and significance from DXA. The DC equation differed significantly from DXA in the
first two quartiles but was not significant in the 3rd and 4th quartiles. JP7 was significantly
different from DXA in all quartiles except the 4th.
10
Table 4. Mean difference %BF from DXA across fatness quartiles.
Quartiles %BF of
DXA_DC mean
Sig.
DXA_JP7 mean
Sig.
n=68
quartile
difference
1
8.3-13.2
1.08*
.004
-2.86*
.000
2
13.2-16.3
1.25*
.001
-2.51*
.000
3
16.3-22.5
1.01
.069
-2.06*
.000
4
*p<.05
22.5-36.4
.122
.109
-.015
.964
difference
In addition, the mean difference in %BF for JP7 and DC was compared to DXA
across age groups. The DC equation was not significantly different from DXA
throughout all four quartiles. The JP7 equation was significant from DXA in the first two
age groups but was not in the last two.
Table 5. Mean difference of %BF from DXA across age quartiles.
Quartiles Age range
DXA_DC mean diff
Sig.
DXA_JP7 mean diff
Sig.
n=68
of quartile
1
18-21
1.01
.227
-3.05*
.000
2
21-25
1.16
.379
-2.51*
.019
3
25-43
1.19
.407
-1.96
.105
4
*p<.05
43-65
1.21
.401
-0.07
.997
DISCUSSION
When Jackson and Pollock created their equations in the 1970’s, the U.S.
population had a much lower BMI and lower %BF than it currently does. Updated,
generalizable equations are needed in the field with the changing body composition
11
among the population. In addition, the equations created in the 1970’s were based on 2C
models, which are now outdated due to improvements in technology and the use of multicompartmental models. However, many of these techniques are limited to their use in the
laboratory while skinfolds are a fast and simple method to use in the field. This drives the
need for reliable and valid skinfold equations.
Validation for the DC equation. The DC equation meets standards presented by
Lohman et al (20). Ideally, a SF prediction equation will have a low SEE, high
correlation to the criterion method, and a small mean difference.
The SEE for the DC equation was 2.72% and helps explain the accuracy of a
prediction equation. According to Lohman (Table 6), this SEE has a rating of “very
good” (20). The DXA and DC were highly correlated (r=0.934 and R2= 0.87) and the
mean difference between the two was minimal (1.1%).
Table 6. Lohman percent fat errors and prediction errors (20).
SEE%
Actual Error
Subjective Rating
2.0
2.5
3.0
3.5
4.0
4.5
5.0
0-2.0
1.5
2.2
2.9
3.5
4.0
4.6
Ideal
Excellent
Very good
Good
Fairly good
Fairly good
Not recommended
The DC equation was created using variables highly correlated to %BF, as found
by Jackson and Pollock (16). Thus, it is likely this equation will accurately predict %BF.
Predictors used while creating the equation included the sum of skinfolds, and the sum of
skinfolds squared since the relationship between %BF and skinfolds is curvilinear. The
12
DC equation also included age as a final predictor. Other predictors highly correlated to
%BF were used (waist and hip circumference, weight, and height) but were not included
in the DC equation because they did not account for any additional variance.
Comparison among studies. Descriptive data of the anthropometrics of Cowan,
Ball et al, and the original Jackson and Pollock sample are shown in Table 7. Age, height,
and body mass were similar between the three samples. The age ranges for Cowan, Ball,
and Jackson and Pollock were 18-65, 18-62, and 18-61 y, respectively. Body mass was
increased in Cowan (82.1 ± 11.9 kg) and Ball (82.3 ± 14.1 kg) compared to Jackson and
Pollock (74.8 ± 11.8 kg). BMI was slightly higher in this sample (25.6 ± 3.3 kg/m2)
compared to Ball (23.0 ± 4.1 kg/m2) and Jackson and Pollock (23.3 kg/m2) samples. Like
the Jackson and Pollock sample, this sample was mostly white. The validation and Ball
samples had very similar race data as well. Because Ball and Cowan had higher body
mass but slightly lower %BF than Jackson & Pollock, our samples may have included
more athletic and fit men. Due to the similarities between samples, it is unlikely the
differences between the equations are due to a difference in fatness. The differences are
more likely due to criterion methods the equations are based on.
13
Table 7. Descriptive statistic comparisons among Cowan, Ball, and JP samples.
Variable
Age (y)
Body height (m)
Body mass (kg)
Body Mass Index (kg/m2)
Σ7 Skinfolds
Σ3a Skinfolds
JP7 %BF
Hispanic
Black
Asian
Cowan N=272
Mean ± SD
32.4 ± 14.0
1.79 ± 0.1
82.1 ± 11.9
25.6 ± 3.3
111.9 ± 50.9
48.9 ± 21.5
16.1 ± 7.4
1%
2%
2%
Ball N=160
Mean ± SD
32.1 ± 11.0
178.8 ± 6.9 (cm)
82.3 ± 14.1
23.0 ± 4.1
114.0 ± 51.9
45.3 ± 20.1
16.3 ± 7.3
2%
8%
2%
JP N=308
Mean ± SD
32.6 ± 10.8
1.8 ± 0.1
74.8 ± 11.8
23.3
122.6 ± 52.0
59.4 ± 24.3
17.7 ± 8.0
mostly white
This study found all three JP equations to underestimate % BF, which is
concurrent with Ball et al’s findings (2). The mean difference for %BF was comparable
between this study and the original study in which the equation was created. Both the
original Ball et al study and the current study found similar results for all three Jackson
and Pollock equations (Table 8).
Table 8. Comparison between studies for the mean %BF difference from DXA.
JP7
JP3a
JP3b
Mean %BF difference from DXA
Ball
Cowan
-3.1*
-1.9*
-3.4*
-3.2*
-3.2*
-2.4*
JP7= 7 site equation by Jackson and Pollock; JP3a= 3 site SF equation by Jackson and Pollock;
JP3b= 3 site SF equation by Jackson and Pollock.
*p<.05
%BF across age and fatness. The DC equation differed significantly from DXA
in the first two quartiles but was not significant in the 3rd and 4th quartiles. JP7 was
14
significantly different from DXA in all quartiles except the 4th. This shows that the DC
equation may work better in predicting %BF in men of higher BF (16.3-36.4%). While
this may seem as a limitation, many clients in the field are within this %BF and is
applicable for clinical use.
The DC equation was not significantly different from DXA through out all four
age quartiles. The JP7 equation was significant from DXA in the first two age groups but
was not in the last two. This suggests that DC can accurately predict %BF within 18-65 y
subjects.
DXA as a criterion. Even though 4C models have been shown to provide more
accurate estimations of body composition, DXA was used as the criterion method for this
study. 4C models are often elaborate, time consuming, and costly for large population
studies, and thus are not practical in this type of research setting. The next best option is a
3C model. Research has shown a strong correlation between DXA and 4C models. Prior
et al found no significant difference in %BF between DXA and the 4C model regardless
of gender, race, athletic status, or body fatness (24). Other studies found similar results
(27, 31).
Data for this analysis was collected from two different DXA machines (Hologic
QDR 4500A and Hologic QDR Discovery-A) in the same laboratory. Studies have shown
only very slight, non-significant differences between two DXA machines of the same
model (5, 11, 19, 21). Data collected by a single DXA machine increases the likelihood
that a false difference between the currently employed SF equations may occur and an
inaccurate, faulty DXA machine is being used. By using two DXA machines and
merging the data, it is unlikely an overestimation of %BF would occur by the currently
15
recommended SF equations. Both DXA’s were calibrated and followed the manufactures
directions when used.
CONCLUSION
The purpose of this study was to validate the DC skinfold equation developed by
Ball et al. in 2004. The data confirms that the DC equation more accurately predicts %BF
with less error across a general population of men than three ACSM recommended SF
equations (2C model) compared to the DXA (3C model). Practitioners limited to field
methods of body composition should consider using the DC skinfold equation to predict
body fatness of men. This is the first study to validate the DC skinfold prediction
equation from DXA. While the data of this study is promising for the use of this new
equation, further validation studies may be necessary to convince practitioners and
researchers to replace the current ACSM recommended %BF equations.
16
REFERENCES
1. Ball S, Swan PD, and DeSimone R. Comparison of anthropometry to dual energy Xray absorptiometry: a new prediction equation for women. Res Q Exerc Sport.
2004;75(3):248-58.
2. Ball SD, Altena TS, and Swan PD. Comparison of anthropometry to DXA: a new
prediction equation for men. Eur J Clin Nutr. 2004;58(11):1525-31.
3. Behnke A, Feen, BG., Welham, EC. . The specific gravity of healthy men. J Am Med
Assoc. 1942;118:495-8.
4. Bellanger TM, and Bray GA. Obesity related morbidity and mortality. J La State Med
Soc. 2005;157 Spec No 1:S42-9; quiz 9.
5. Blake GM, Tong CM, and Fogelman I. Intersite comparison of the Hologic QDR-1000
dual energy X-ray bone densitometer. Br J Radiol. 1991;64(761):440-6.
6. Bland JM, and Altman DG. Statistical methods for assessing agreement between two
methods of clinical measurement. Lancet. 1986;1(8476):307-10.
7. Brozek J, Grande F, Anderson JT, and Keys A. Densitometric Analysis of Body
Composition: Revision of Some Quantitative Assumptions. Ann N Y Acad Sci.
1963;110:113-40.
8. Calle EE, Thun MJ, Petrelli JM, Rodriguez C, and Heath CW, Jr. Body-mass index
and mortality in a prospective cohort of U.S. adults. N Engl J Med. 1999;341(15):1097105.
9. Clark RR, Kuta JM, and Sullivan JC. Prediction of percent body fat in adult males
using dual energy x-ray absorptiometry, skinfolds, and hydrostatic weighing. Med Sci
Sports Exerc. 1993;25(4):528-35.
10. Company J BS. Body composition comparison: Bioelectric impedance analysis with
dual-energy X-ray absorptiometry in adult athletes. Measurement in physical education
and exercise science. 2010;14:186-201.
11. Economos CD, Nelson ME, Fiatarone MA, Dallal GE, Heymsfield SB, Wang J,
Russell-Aulet M, Yasumura S, Ma R, Vaswani AN, and Pierson RN. A multicenter
comparison of dual-energy X-ray absorptiometers: in vivo and in vitro measurements of
bone mineral content and density. J Bone Miner Res. 1996;11(2):275-85.
12. Haskell WL, Lee IM, Pate RR, Powell KE, Blair SN, Franklin BA, Macera CA,
Heath GW, Thompson PD, and Bauman A. Physical activity and public health: updated
recommendation for adults from the American College of Sports Medicine and the
American Heart Association. Med Sci Sports Exerc. 2007;39(8):1423-34.
17
13. Heymsfield SB, Lohman, TG., Wang, Z., Going, SB. Human body composition.
Champaign, IL: Human Kinetics; 2005.
14. Jackson A, Pollock ML. Practical assessment of body composition. Phys Sport Med.
1985;13(3):76-90.
15. Jackson AS, Ellis KJ, McFarlin BK, Sailors MH, and Bray MS. Cross-validation of
generalised body composition equations with diverse young men and women: the
Training Intervention and Genetics of Exercise Response (TIGER) Study. Br J Nutr.
2009;101(6):871-8.
16. Jackson AS, and Pollock ML. Generalized equations for predicting body density of
men. Br J Nutr. 1978;40(3):497-504.
17. Kelly TL, Berger N, and Richardson TL. DXA body composition: theory and
practice. Appl Radiat Isot. 1998;49(5-6):511-3.
18. Klein S, Burke LE, Bray GA, Blair S, Allison DB, Pi-Sunyer X, Hong Y, and Eckel
RH. Clinical implications of obesity with specific focus on cardiovascular disease: a
statement for professionals from the American Heart Association Council on Nutrition,
Physical Activity, and Metabolism: endorsed by the American College of Cardiology
Foundation. Circulation. 2004;110(18):2952-67.
19. Kolta S, Ravaud P, Fechtenbaum J, Dougados M, and Roux C. Accuracy and
precision of 62 bone densitometers using a European Spine Phantom. Osteoporos Int.
1999;10(1):14-9.
20. Lohman T. Advances in body composition assessment. Champaign, IL: Human
Kinetics; 1992.
21. Orwoll ES, and Oviatt SK. Longitudinal precision of dual-energy x-ray
absorptiometry in a multicenter study. The Nafarelin/Bone Study Group. J Bone Miner
Res. 1991;6(2):191-7.
22. Pace N, Rathbun, EN. Studies on body composition, III: The body water and
chemically combined nitrogen content in relation to fat content. Journal of Biological
Chemistry. 1945;158:685-91.
23. Pietrobelli A, Formica C, Wang Z, and Heymsfield SB. Dual-energy X-ray
absorptiometry body composition model: review of physical concepts. Am J Physiol.
1996;271(6 Pt 1):E941-51.
24. Prior BM, Cureton KJ, Modlesky CM, Evans EM, Sloniger MA, Saunders M, and
Lewis RD. In vivo validation of whole body composition estimates from dual-energy Xray absorptiometry. J Appl Physiol. 1997;83(2):623-30.
18
25. Rimm EB, Stampfer MJ, Giovannucci E, Ascherio A, Spiegelman D, Colditz GA,
and Willett WC. Body size and fat distribution as predictors of coronary heart disease
among middle-aged and older US men. Am J Epidemiol. 1995;141(12):1117-27.
26. Ryder J, Ball, SD. Three-dimensional body scanning as a novel technique for body
composition assessment: a preliminary investigation. Journal of Exercise Physiology
online. 2012;15:1-14.
27. Salamone LM, Fuerst T, Visser M, Kern M, Lang T, Dockrell M, Cauley JA, Nevitt
M, Tylavsky F, and Lohman TG. Measurement of fat mass using DEXA: a validation
study in elderly adults. J Appl Physiol. 2000;89(1):345-52.
28. Segal KR, Gutin B, Presta E, Wang J, and Van Itallie TB. Estimation of human body
composition by electrical impedance methods: a comparative study. J Appl Physiol.
1985;58(5):1565-71.
29. Siri WE. The gross composition of the body. Adv Biol Med Phys. 1956;4:239-80.
30. Wilkins LW. ACSM's Guidelines for Exercise Testing and Prescription. Philidelphia;
2006.
31. Withers RT, LaForgia J, Pillans RK, Shipp NJ, Chatterton BE, Schultz CG, and
Leaney F. Comparisons of two-, three-, and four-compartment models of body
composition analysis in men and women. J Appl Physiol. 1998;85(1):238-45.
19
APPENDIX A: AN EXTENDED LITERATURE REVIEW
Obesity is a growing epidemic in the United States with currently over two-thirds
of individuals classified as overweight, and one-third as obese (28). Excess body fat (BF)
is associated with chronic diseases such as type 2 diabetes mellitus, coronary artery
disease, hypertension, metabolic syndrome, cancers of the breast, prostate, and colon,
stroke, hyperlipidemia and all-cause mortality (10, 15, 29, 40, 58, 88). There are no
universal standards for healthy BF percentages but ranges of 10-25% for men and 2035% for women are considered acceptable by some researchers (35, 64). Means of
measuring an individual’s excess BF is essential due to the disease risk. Accurate and
valid body composition methods that are both inexpensive and practical are needed in
order to better design nutritional and exercise programs. These methods are also
necessary to track progress over time once these programs have been implemented (42).
Unfortunately, there is no direct way to measure body composition in vivo. .
Cadaver dissection is the only way to obtain a precise measurement of body composition.
In addition, there are a very limited number of complete cadaver dissection studies (4, 18,
70, 71). This makes indirect methods crucial for predicting body composition in both the
field and laboratory. Indirect methods are composted of many different compartment
models in which the body is divided. The basic theoretical model is the two-compartment
(2C) model, in which the body is separated into fat mass (FM) and fat free mass (FFM).
In the 2C model, directly measuring FM has been challenging. However, FM can be
indirectly calculated by measuring FFM using a variety of methods and assumptions and
then subtracting FFM from the total mass of the subject. The 2C model assumes that the
20
water and mineral proportions are constant across all individuals and that total body
water makes up 73.2% of FFM (79, 97). The density assumed for FFM in the 2C model is
1.1000g/cm3 (14). However, it has been cited that these proportions of water, protein,
and mineral are not constant among individuals of different age, gender, or fatness (17,
87, 110, 116). Because of these assumptions, multi-compartment models (3C, 4C, and
5C) have been developed which measure different components of the FFM (42).
There are many methods and models currently being used in the area of body
composition (Figure 1), and each includes its own assumptions and limitations. In order
to predict body fat accurately, proper equations must be developed and thoroughly
validated. Some equations are more generalizable while some are population specific.
When developing new equations, the standard error of the estimate (SEE) for predicting
the criterion should be between ±2.5-3.5 percent body fat (65). These equations should
also be applicable to specific populations, use the appropriate compartment model, have a
representative sample, have the variables be measured in the same way as the original
investigators, be cross-validated in additional samples, and give an accurate estimation of
percent body fat (51, 65).
21
Figure 1- Levels of body composition (111).
Field Methods.
BMI
Anthropometric data such as height and weight can be used to access risk for
chronic diseases (29). The body mass index (BMI) is calculated by weight (kg) divided
by height squared (m2). The American College of Sports Medicine (ACSM) recommends
a BMI between 18.5-24.9 kg/m2 and a BMI outside this range increases the risk for
chronic diseases (115). A BMI below 18.5 kg/m2 is categorized as underweight, 25-29.9
kg/m2 is considered overweight and over 30 kg/m2 is considered obese (80).
BMI may also be used to predict body fat percentage. However, while it is a great
field tool for measuring large population numbers, it fails to distinguish between lean and
fat mass and has a large standard of error (±5%) (35). Gallagher et al developed equations
to predict %BF from BMI in an international study in 2000 (35). Sex, age, and ethnic
specific equations were created and consisted of 1626 subjects. Additionally, a metaanalysis by Flegal et al found significantly higher all cause mortality in obese subjects
22
when defined by BMI (29). This shows that BMI may be a useful tool in determining risk
factors related to obesity.
Waist-to-hip ratio
Certain patterns of fat distribution can increase the risk for chronic diseases (54,
61). Android obesity is characteristic of fat distribution around the waist while gynoid
obesity is characterized by lower fat distribution, around the hips and thighs (115). The
waist to hip ratio (WHR) is an assessment of the fat distribution for an individual and
therefore determine risks associated with high abdominal adipose tissue (13). A WHR of
more than 0.86 for women and more than .95 for men increases health risks (80).
Correlations between abdominal circumferences and body density have been
shown to be about 0.7 (50, 95). This correlation is lower than most correlations for
skinfolds (SF) and body density (33, 66). Like BMI, WHR does not consider the
proportions of lean and fat mass and is not the most acceptable way to measure body
composition.
Skinfolds
SF measure the amount of subcutaneous fat on an individual and assume that the
amount of subcutaneous fat is proportional to overall fat percentage. This varies with age,
sex, and ethnicity (84, 85). Skinfolds are determined by using pinch calipers to measure
the subcutaneous adipose tissue. Assuming a skilled technician uses proper technique,
skinfolds have about an ±3.5% error (115). However, errors less than 3% and up to 5%
23
have been reported (1, 27). The error may also increase if the skinfold technician is not
skilled (47)
A skinfold method assumes that overall adipose tissue is related to the amount of
subcutaneous adipose tissue and as subcutaneous skinfold measurements increase, so
does the relative body fat content (7, 53). Typically, a three or seven site formula is used
and the sum of the skinfolds is used to find body density (Db) (44, 49). Once Db is found,
the Siri or Brozek equations (based on 2C model) can be used to calculate body fat
percentage (14, 99).
Although skinfolds have a larger error than other methods, they remain a very
practical way to measure %BF in a large number of individuals in the field. Measuring
%BF with skinfolds is fast and relatively easy. In addition, calipers are inexpensive and
portable, there is minimal health risk (bruising of the skin), and the procedure requires
little subject compliance.
Bioelectrical impedance analysis
Bioelectrical impedance analysis (BIA) is a simple and efficient field method that
uses electrical current to estimate %BF by relying on the relationship between body
composition and total body water content (46, 52, 105). Water is an excellent electrical
conductor. Since lean tissue contains more water and electrolytes than fat, the electrical
current is quick. When the current runs through fat tissue, the current is impeded
compared to lean tissue. Using BIA, reactance (Xc) and resistance (R) are usually
24
measured by attaching two current electrodes and two detection electrodes to the body
(42).
There are many methods of BIA. Two of the most common methods include the
single frequency (SF) BIA and multi frequency (MF) BIA. The SF-BIA usually operates
at 50kHz and electrodes are placed on the hands and feet. SF-BIA estimates extracellular
water (ECW) and total body water (TBW), but not intracellular water (ICW). This leaves
proper hydration as a limitation for this method (38). The BIA assumes a constant ratio
between ECW and ICW and if this ratio is altered, accuracy for this method is decreased.
MF-BIA uses different frequencies to measure FFM, TBW, ECW, and ICW. However, it
has been cited that this method has poor reproducibility at very low (5 kHz) and high
(200 kHz) frequencies (39). MF-BIA has also been shown to be more accurate than SFBIA in predicting ECW, but SF-BIA is more accurate in predicting TBW, leaving
advantages for both methods (81). The SEE for BIA predicting %BF is typically ±3.5-5%
(42).
As stated above, one of the main limitations of BIA is the impact of hydration
status on measuring the different body compartments (77). Due to the difference in
electrical properties of ECW and ICW, hydration must remain constant in order to
achieve accurate body composition results (31). Electrolyte and fluid changes impact the
ICW and ECW ratio (41). Hydration also changes with age and the prediction equations
used need to account for this change (3, 86). Hydration status and fluid balance need to
be tightly controlled in subjects for accurate results.
Since exercise can influence hydration status, BIA measurements should be taken
at appropriate time points around exercise. Liang et al found that after 30 minutes of
25
aerobic exercise, R decreased by 3% and Xc decreased by 8% (63). However, these
measurements were found to return to pre-exercise baseline after one-hour post exercise.
The predictability in each BIA equation also creates another limitation with this
method. A majority of the studies for creating and validating equations have been in adult
white males. However, body build, leg length, and frame size have been shown to be
different across ethnic populations leaving these assumptions to be a limiting factor (24,
25).
Research data shows that the BIA is reliable in healthy human subjects when the
correct prediction equation is used (41). This method is very portable, does not require a
lot of technical skill, is relatively inexpensive, and requires little subject compliance (59).
Acceptable reliability has been demonstrated in unhealthy subjects that have stable water
and electrolyte balance (60).
Laboratory Methods.
Hydrostatic weighing
Hydrostatic (under water) weighing (HW) is another valid and reliable technique
that was used as the original gold standard for measuring body composition. HW uses
Archimedes’s principle to find body volume, which states that a submerged object is of
equal weight to the weight of water it displaces. Since body mass and body volume can
be measured, Db can be found (Db=m/v). Since fat is less dense than muscle or bone,
individuals with higher fat content will have a lower body density (42). Other variables
such as residual lung volume, density of the water (calculated by known water
temperature), gastrointestinal gas, dry body weight, and under water weight must be
26
known. The equations most commonly used to convert Db to percent body fat are the Siri
and Brozek formulas (14, 99).
Reliability of this method has been reported as exceptional (r=0.95) (65).
However, this method is not without error. FFM has been shown to have different
densities among individuals (65). Certain populations, such as African Americans, youth,
and the elderly, have been shown to have different lean tissue densities. HW assumes all
lean tissue has a density of 1.1000 g/cc which creates measurement errors for populations
with different lean tissue density (42). The error in HW has been shown to be 3-4% (8,
43, 99). HW limitations include increase cost, time requirements, the need for specialized
equipment, a trained technician, and subject compliance (55). Many subjects find it
difficult to stay under water long enough to obtain measurements. One major limitation
of HW is the determination of residual lung volume (55). Residual lung volume is
difficult to measure, forcing researchers to estimate this value. In addition, accuracy is
limited since it is still not a direct measurement. Newer, practical, and more valid
methods are now commonly used to predict %BF.
Air displacement plethysmography
Air displacement plethysmography (ADP), commercially known as the BOD
POD (Life Measurements, Inc, Concord, CA) was made popular in the mid-nineties as an
alternative to HW (23). This method is similar to HW but instead of using water
displacement, the BOD POD uses air displacement to find body volume by using the
relationship of pressure and volume. Boyle’s law states that when temperature is
constant, pressure and volume are inversely related.
27
This dual chambered (test chamber and reference chamber), pod shaped vesicle
(Figure 2) can accommodate various subjects, is portable, emits no radiation, and requires
little technical skill. The first step in using the Bod Pod to complete body composition
measurements is to measure the volume inside the chamber without the subject. The
subject is then placed inside the chamber and body volume is then measured twice. If the
two measurements are not within 2% or 150ml, a third measurement is taken and the
closest two measurements are averaged and used (23).
Figure 2- Setup of BOD POD (23).
Once volume is found, normal densitometry principles can be applied (9, 100).
Similar to HW, Db is calculated and %BF can then be determined by using the Siri or
Brozek formula based on the 2C model (14, 99). Race specific equations can also be used
instead of Siri or Brozek (96, 107).
The Bod Pod requires the subject to sit very still for 2-3 minutes inside the
chamber while measurements are being made, creating the need for some subject
compliance. However, when finding total body volume, thoracic lung volume should be
considered. While the technician is able to use a prediction equation, directly measuring
thoracic lung volume is preferred. This requires the subject to follow precise directions
28
during which they exhale into a tube and may be uncomfortable. Using ADP to determine
body composition relies on the 2C model similar to skinfolds and HW, and therefore has
the same limitations and assumptions described above. The Bod Pod also takes longer to
calibrate and complete the test (compared to DXA) and is inadequate for subjects that
suffer from claustrophobia. Also, the results may be affected by pressure changes in the
room or abnormal subject breathing,
The reliability of the Bod Pod has been tested over the years. When tested with
inanimate objects (such as the aluminum cylinder, reliability has been excellent (23, 76).
Research studies have shown that when measuring %BF within subjects on the same day,
the mean %BF between two trials ranges from 1.7-4.5% (11, 48, 73, 74, 76, 94). Also,
Miyatake et al reported that intertester reliability was 4.5% (74). The reliability between
two Bod Pod machines has also been evaluated. Collins et al measured body volume on
30 subjects on two different Bod Pod devices in separate laboratories (19). The mean
body volume for the two laboratories were 66.86 ± 0.08 L and 66.92 ± 0.17 L and were
not significantly different.
DXA
Dual energy x-ray absorptiometry (DXA) is considered the new “practical gold
standard” of body composition methods. This machine uses dual energy X-rays to scan
the body and has been shown to be an accurate method in measuring body composition
(56, 67, 82, 108, 109, 112). Three compartments are measured; fat mass, lean mass, and
bone mineral mass. The DXA is best known for its usefulness in diagnosis of osteopenia
29
and osteoporosis (62, 82). DXA is a 3C model, but is also used as an aid in providing the
density of bone mineral mass in 4C models.
This method is also able to break up the body into compartments and compare the
%BF of the limbs, trunk, and the head. It is also a quick, accurate, and valid method,
which requires little subject compliance and causes little discomfort. The radiation
exposure is small, about equal to a transcontinental flight in the United States. Therefore,
pregnant women are not recommended to undergo a DXA scan. DXA has shown to
predict skeletal muscle mass close to that of other multi-compartmental methods such as
CT scans and neutron activation analysis (56, 108, 109, 112). These studies have shown
that DXA can predict %BF within 2-3%.
While the DXA is an accurate laboratory reference method, limitations do exist.
The DXA is more expensive than other methods, has an upper weight and height limit,
and involves a small amount of radiation. Like other techniques, DXA assumes that lean
tissue has a hydration status of 73% when in reality, hydration status can fluctuate among
individuals (83).
Imaging Methods
Imaging methods such as Computed Tomography (CT) and Magnetic Resonance
Imaging (MRI) are considered some of the most accurate methods in analyzing body
composition. These methods are the only ones that can measure internal organs and
tissues (42). These imaging methods are also beneficial because they can separate
visceral adipose tissue from subcutaneous adipose tissue. Being able to measure body fat
distribution is important due to the accumulation of visceral adipose tissue being
30
associated with increased risk for chronic diseases (6, 16, 32, 57, 72, 78, 114, 117). The
use of imaging has been very beneficial in the intervention of decreasing visceral adipose
tissue in subjects (69, 90, 91, 103). It has also been beneficial for measuring changes in
both visceral and subcutaneous fat throughout an intervention
CT scanning uses X-rays to provide imaging of the soft tissue. With the
attenuation of the X-rays, tissue density can be recorded. Scanning the density of bone,
skin, adipose tissue, and muscle shown to be incredibly accurate (less than 1% error) (91,
101). CT scanning is preferred over DXA because of its ability to determine body fat
distribution, making it an excellent multi-component model (12, 102, 106). CT has also
been shown to be more accurate in measuring visceral adipose tissue than an MRI (42,
98).
Unlike CT, MRI’s (introduced widely in the 1980’s) do not use radiation, making
it a good alternative to CT in certain populations. An MRI measures the interaction of
hydrogen protons, which behave like magnets. A pulsed radio frequency is applied when
a subject is scanned and the protons absorb the energy, creating a signal that generates
cross sectional images of a subject’s body. An MRI of the whole body can take up to 30
minutes or longer (89). Subjects are required to hold their breath when doing abdominal
slices as movement can reduce the clarity of the image (34).
While CT and MRI are excellent models, they are not without limitations. These
procedures are exceedingly costly, time consuming, and complex. Therefore, these
techniques are not suitable for large population studies. CT also emits some radiation and
it is not advisable to due multiple scans in a short period of time, or scan pregnant
31
women. However, if the pixels of the image can be decreased, the radiation exposure
could be similar to a DXA scan (104).
3D body scanning
The 3D body scanner has been used in the textile industry to obtain
circumferences and body shapes for clothing, but little research has been done as a
method in body composition. The scanner uses fan beam technology, which requires light
to capture a digital 3D body image of a subject. One important aspect of the scanner is
that it can measure body volume and therefore, density can be found and computed into
percent body fat. The 3D scanner has shown promise in being a method to measure body
composition as it is very quick (~5 sec), requires little technical skill and calibration, is
non invasive for the subject, and output anthropometric data is more accurate than hand
circumference methods (93).
Currently, only one other study has compared the body scanner to the DXA and
Bod Pod (93). Percent body fat was determined using the scanner via three methods: Siri
equation from bulk body volume, an equation developed by the Department of Defense
(DoD) (36), and a new prediction equation (3D MU). The percent body fat found from
the 3D MU prediction was significantly correlated to DXA (P<.01) and predicted body
composition similar to other 2C models. The 3D body scanner shows promise as a body
composition technique as it is fast and very non-invasive. According to Lohman (65), the
SEE of the 3D scan (3.03) falls in the “very good” rating, making it an acceptable
equation for measuring body composition with the scanner.
32
The methods used in 3D body scanning are based on the 2C model and therefore,
are limited by its assumptions. The DXA is able to calculate bone density separate from
the other fat free components, while the 3D body scanner assumes that bone density is the
same across all individuals. This could contribute to the error in the method. Because of
the many advantages the scanner has at measuring body composition, however, it is
worth continuing research in this area and validating the equations.
Body Composition and Obesity
Obesity provides a challenge for many researchers in measuring body
composition (2, 22, 30). Most of the assessment techniques used today were developed
from non-obese subjects and the obesity rate has greatly amplified in recent years.
Research on body composition assessment in the obese is more limited than in normal
weight subjects. Certain factors have to be taken into consideration such as weight limits
of the measuring equipment, location of equipment (for those less mobile), wider chairs
and clothing, larger weight scales, and overall consideration for accommodating larger
subjects. In addition, body compartments differ between reference subjects and that in
obese subjects (Figure).
33
Figure- Compartmental breakdown comparison in a reference female and severe
obese subject (21).
Anthropometrics may be difficult to measure in obese populations. There should
be some caution in measuring skinfolds for these groups. Calipers limit measuring heavy
subjects since the maximum jaw openings are between 50-65 mm wide (45). It has also
been cited that it is difficult to find the appropriate landmarks for skinfolds in the obese
(113). Because obesity rates have increased since Jackson and Pollock created their
skinfold equations in the 1970’s, new reference equations may be needed. Nevill et al
found that predicting %BF in subjects with a sum of skinfolds greater than 120 mm was
difficult, and using Jackson and Pollock 3 site equation did not accurately predict %BF in
these subjects when compared to DXA (75).
34
For waist and hip circumferences, two people and two tape measures may be
needed for certain subjects. Also, determining the waist circumference can be challenging
in severe obesity as waist circumference is defined by the narrowest part of the waist (5).
Single frequency BIA may also have some challenges since electrical properties
in the obese are different from normal reference methods (26). BIA has been shown to
overestimate FFM in obese subjects due to an increase in ECW/ICW ratios and an overall
increase in total body water (20, 26). Like other methods, population specific equations
should be used. However, obese specific equations for BIA are limited and the accuracy
in predicting %BF has shown to be highly dependent on the equation used (22). When
compared to hydrodensiometry, one of the most accurate equations has been from
Lukaski et al with a 2.7% error (22, 68).
The BOD POD has had success in the obese population and does not seem to
have larger error than with normal weight subjects. Ginde et al found that the BOD POD
was accurate in overweight and obese subjects when compared to HW (37). The BOD
POD can accommodate large subjects and requires little compliance from the subject. It
has been known for fitting large professional athletes such as Shaquille O'Neal.
Measuring obese subjects with a DXA is difficult since most severely obese
people exceed the weight limit for the table and do not fit within the scanning zone (21).
Previously, MRI’s had similar problems (limited to 300 lbs) but with new advancements,
obese subjects can now be measured. MRI’s are now built with an open magnet and
larger table weight limit, and also have a new low field MR imager. The MRI is well
suited for obese subjects (studies have cited up to 490 lbs) because the radio frequency
does not have any problems penetrating a large amount of adipose tissue (92).
35
Accurately measuring excess BF within an individual is now essential, especially
with the high US obesity rates. There are advantages and disadvantages to all body
composition methods and the goal as researchers is to choose the one that will best
provide the most accurate results for each situation and purpose. Measuring body fat is
helpful in assessing disease risk, creating nutritional and exercise programs, increasing
athletic performance for athletes, and tracking progress of individuals over time. It is
critical to validate new and old methods as technology and knowledge expand in the area
of body composition.
36
References
1. Ackland TR, Lohman TG, Sundgot-Borgen J, Maughan RJ, Meyer NL, Stewart AD,
and Muller W. Current status of body composition assessment in sport: review and
position statement on behalf of the ad hoc research working group on body composition
health and performance, under the auspices of the I.O.C. Medical Commission. Sports
Med. 2012;42(3):227-49.
2. Albu J, Smolowitz J, Lichtman S, Heymsfield SB, Wang J, Pierson RN, Jr., and PiSunyer FX. Composition of weight loss in severely obese women: a new look at old
methods. Metabolism. 1992;41(10):1068-74.
3. Alexy U, Cheng G, Libuda L, Hilbig A, and Kersting M. 24 h-Sodium excretion and
hydration status in children and adolescents--results of the DONALD Study. Clin Nutr.
2012;31(1):78-84.
4. Allen TH, Welch BE, Trujillo TT, and Roberts JE. Fat, water and tissue solids of the
whole body less its bone mineral. J Appl Physiol. 1959;14:1009-12.
5. Allison DB, Nathan JS, Albu JB, Heymsfield SB, Duprat LJ, and Pi-Sunyer FX.
Measurement challenges and other practical concerns when studying massively obese
individuals. Int J Eat Disord. 1998;24(3):275-84.
6. Antuna-Puente B, Feve B, Fellahi S, and Bastard JP. Adipokines: the missing link
between insulin resistance and obesity. Diabetes Metab. 2008;34(1):2-11.
7. Baker GL. Human adipose tissue composition and age. Am J Clin Nutr.
1969;22(7):829-35.
8. Bakker HK, and Struikenkamp RS. Biological variability and lean body mass
estimates. Hum Biol. 1977;49(2):187-202.
9. Behnke A, Feen, BG., Welham, EC. . The specific gravity of healthy men. J Am Med
Assoc. 1942;118:495-8.
10. Bellanger TM, and Bray GA. Obesity related morbidity and mortality. J La State Med
Soc. 2005;157 Spec No 1:S42-9; quiz 9.
11. Biaggi RR, Vollman MW, Nies MA, Brener CE, Flakoll PJ, Levenhagen DK, Sun M,
Karabulut Z, and Chen KY. Comparison of air-displacement plethysmography with
hydrostatic weighing and bioelectrical impedance analysis for the assessment of body
composition in healthy adults. Am J Clin Nutr. 1999;69(5):898-903.
12. Borkan GA, Gerzof SG, Robbins AH, Hults DE, Silbert CK, and Silbert JE.
Assessment of abdominal fat content by computed tomography. Am J Clin Nutr.
1982;36(1):172-7.
37
13. Bray GA, and Gray DS. Obesity. Part I--Pathogenesis. West J Med. 1988;149(4):42941.
14. Brozek J, Grande F, Anderson JT, and Keys A. Densitometric Analysis of Body
Composition: Revision of Some Quantitative Assumptions. Ann N Y Acad Sci.
1963;110:113-40.
15. Calle EE, Thun MJ, Petrelli JM, Rodriguez C, and Heath CW, Jr. Body-mass index
and mortality in a prospective cohort of U.S. adults. N Engl J Med. 1999;341(15):1097105.
16. Chiba Y, Saitoh S, Takagi S, Ohnishi H, Katoh N, Ohata J, Nakagawa M, and
Shimamoto K. Relationship between visceral fat and cardiovascular disease risk factors:
the Tanno and Sobetsu study. Hypertens Res. 2007;30(3):229-36.
17. Chumlea WC, Schubert CM, Sun SS, Demerath E, Towne B, and Siervogel RM. A
review of body water status and the effects of age and body fatness in children and adults.
J Nutr Health Aging. 2007;11(2):111-8.
18. Clarys JP, Martin AD, and Drinkwater DT. Gross tissue weights in the human body
by cadaver dissection. Hum Biol. 1984;56(3):459-73.
19. Collins AL, Saunders S, McCarthy HD, Williams JE, and Fuller NJ. Within- and
between-laboratory precision in the measurement of body volume using air displacement
plethysmography and its effect on body composition assessment. Int J Obes Relat Metab
Disord. 2004;28(1):80-90.
20. Coppini LZ, Waitzberg DL, and Campos AC. Limitations and validation of
bioelectrical impedance analysis in morbidly obese patients. Curr Opin Clin Nutr Metab
Care. 2005;8(3):329-32.
21. Das SK. Body composition measurement in severe obesity. Curr Opin Clin Nutr
Metab Care. 2005;8(6):602-6.
22. Das SK, Roberts SB, Kehayias JJ, Wang J, Hsu LK, Shikora SA, Saltzman E, and
McCrory MA. Body composition assessment in extreme obesity and after massive weight
loss induced by gastric bypass surgery. Am J Physiol Endocrinol Metab.
2003;284(6):E1080-8.
23. Dempster P, and Aitkens S. A new air displacement method for the determination of
human body composition. Med Sci Sports Exerc. 1995;27(12):1692-7.
24. Deurenberg P, and Deurenberg-Yap M. Validity of body composition methods across
ethnic population groups. Acta Diabetol. 2003;40 Suppl 1:S246-9.
25. Deurenberg P, Deurenberg-Yap M, and Schouten FJ. Validity of total and segmental
impedance measurements for prediction of body composition across ethnic population
groups. Eur J Clin Nutr. 2002;56(3):214-20.
38
26. Durenberg P. Limitations of the bio-electrical impedance method for the assessment
of body fat in severe obesity. American journal of clinical Nutrition. 1996;64 (suppl
3):449S-52S.
27. Durnin JV, and Womersley J. Body fat assessed from total body density and its
estimation from skinfold thickness: measurements on 481 men and women aged from 16
to 72 years. Br J Nutr. 1974;32(1):77-97.
28. Flegal KM, Carroll MD, Ogden CL, and Curtin LR. Prevalence and trends in obesity
among US adults, 1999-2008. Jama. 2010;303(3):235-41.
29. Flegal KM, Kit BK, Orpana H, and Graubard BI. Association of all-cause mortality
with overweight and obesity using standard body mass index categories: a systematic
review and meta-analysis. Jama. 2013;309(1):71-82.
30. Fogelholm GM, Sievanen HT, van Marken Lichtenbelt WD, and Westerterp KR.
Assessment of fat-mass loss during weight reduction in obese women. Metabolism.
1997;46(8):968-75.
31. Foster KR, and Lukaski HC. Whole-body impedance--what does it measure? Am J
Clin Nutr. 1996;64(3 Suppl):388S-96S.
32. Fox CS, Massaro JM, Hoffmann U, Pou KM, Maurovich-Horvat P, Liu CY, Vasan
RS, Murabito JM, Meigs JB, Cupples LA, D'Agostino RB, Sr., and O'Donnell CJ.
Abdominal visceral and subcutaneous adipose tissue compartments: association with
metabolic risk factors in the Framingham Heart Study. Circulation. 2007;116(1):39-48.
33. Frerichs RR, Harsha DW, and Berenson GS. Equations for estimating percentage of
body fat in children 10--14 years old. Pediatr Res. 1979;13(3):170-4.
34. Fuller MF, Fowler PA, McNeill G, and Foster MA. Imaging techniques for the
assessment of body composition. J Nutr. 1994;124(8 Suppl):1546S-50S.
35. Gallagher D, Heymsfield SB, Heo M, Jebb SA, Murgatroyd PR, and Sakamoto Y.
Healthy percentage body fat ranges: an approach for developing guidelines based on
body mass index. Am J Clin Nutr. 2000;72(3):694-701.
36. Garlie TN OJ, Corner B, Zabanski E. Comparison of body fat estimates using 3D
digital laser scans, direct
manual anthropometry, and DXA in men. American journal of human biology.
2010;22:695-701.
37. Ginde SR, Geliebter A, Rubiano F, Silva AM, Wang J, Heshka S, and Heymsfield
SB. Air displacement plethysmography: validation in overweight and obese subjects.
Obes Res. 2005;13(7):1232-7.
39
38. Gudivaka R, Schoeller DA, Kushner RF, and Bolt MJ. Single- and multifrequency
models for bioelectrical impedance analysis of body water compartments. J Appl Physiol.
1999;87(3):1087-96.
39. Hannan WJ, Cowen SJ, Fearon KC, Plester CE, Falconer JS, and Richardson RA.
Evaluation of multi-frequency bio-impedance analysis for the assessment of extracellular
and total body water in surgical patients. Clin Sci (Lond). 1994;86(4):479-85.
40. Haskell WL, Lee IM, Pate RR, Powell KE, Blair SN, Franklin BA, Macera CA,
Heath GW, Thompson PD, and Bauman A. Physical activity and public health: updated
recommendation for adults from the American College of Sports Medicine and the
American Heart Association. Med Sci Sports Exerc. 2007;39(8):1423-34.
41. Heitmann BL. Impedance: a valid method in assessment of body composition? Eur J
Clin Nutr. 1994;48(4):228-40.
42. Heymsfield SB, Lohman, TG., Wang, Z., Going, SB. Human body composition.
Champaign, IL: Human Kinetics; 2005.
43. Heymsfield SB, Wang J, Kehayias J, Heshka S, Lichtman S, and Pierson RN, Jr.
Chemical determination of human body density in vivo: relevance to hydrodensitometry.
Am J Clin Nutr. 1989;50(6):1282-9.
44. Heyward VH. Practical body composition assessment for children, adults, and older
adults. Int J Sport Nutr. 1998;8(3):285-307.
45. Himes JH. Prevalence of individuals with skinfolds too large to measure. Am J Public
Health. 2001;91(1):154-5.
46. Hoffer EC, Meador CK, and Simpson DC. Correlation of whole-body impedance
with total body water volume. J Appl Physiol. 1969;27(4):531-4.
47. Housh TJ, Johnson GO, Thorland WG, Cisar CJ, Hughes RA, Kenney KB,
McDowell SL, and Lundvall P. Validity and intertester error of anthropometric
estimations of body density. J Sports Med Phys Fitness. 1989;29(2):149-56.
48. Iwaoka H, Yokoyama T, Nakayama T, Matsumura Y, Yoshitake Y, Fuchi T,
Yoshiike N, and Tanaka H. Determination of percent body fat by the newly developed
sulfur hexafluoride dilution method and air displacement plethysmography. J Nutr Sci
Vitaminol (Tokyo). 1998;44(4):561-8.
49. Jackson A, Pollock ML. Practical assessment of body composition. Phys Sport Med.
1985;13(3):76-90.
50. Jackson AS, and Pollock ML. Factor analysis and multivariate scaling of
anthropometric variables for the assessment of body composition. Med Sci Sports.
1976;8(3):196-203.
40
51. Jackson AS, and Pollock ML. Generalized equations for predicting body density of
men. Br J Nutr. 1978;40(3):497-504.
52. Jenin P, Lenoir J, Roullet C, Thomasset AL, and Ducrot H. Determination of body
fluid compartments by electrical impedance measurements. Aviat Space Environ Med.
1975;46(2):152-5.
53. Kabir N, and Forsum E. Estimation of total body fat and subcutaneous adipose tissue
in full-term infants less than 3 months old. Pediatr Res. 1993;34(4):448-54.
54. Kannel WB, Cupples LA, Ramaswami R, Stokes J, 3rd, Kreger BE, and Higgins M.
Regional obesity and risk of cardiovascular disease; the Framingham Study. J Clin
Epidemiol. 1991;44(2):183-90.
55. Katch F. Practice curves and errors of measurement in estimating underwater weight
by hydrostatic weighing. Med Sci Sports Exerc. 1969;1:212-6.
56. Kim J, Wang Z, Heymsfield SB, Baumgartner RN, and Gallagher D. Total-body
skeletal muscle mass: estimation by a new dual-energy X-ray absorptiometry method. Am
J Clin Nutr. 2002;76(2):378-83.
57. Kim TN, Park MS, Lim KI, Yang SJ, Yoo HJ, Kang HJ, Song W, Seo JA, Kim SG,
Kim NH, Baik SH, Choi DS, and Choi KM. Skeletal muscle mass to visceral fat area
ratio is associated with metabolic syndrome and arterial stiffness: The Korean Sarcopenic
Obesity Study (KSOS). Diabetes Res Clin Pract. 2011;93(2):285-91.
58. Klein S, Burke LE, Bray GA, Blair S, Allison DB, Pi-Sunyer X, Hong Y, and Eckel
RH. Clinical implications of obesity with specific focus on cardiovascular disease: a
statement for professionals from the American Heart Association Council on Nutrition,
Physical Activity, and Metabolism: endorsed by the American College of Cardiology
Foundation. Circulation. 2004;110(18):2952-67.
59. Kyle UG, Bosaeus I, De Lorenzo AD, Deurenberg P, Elia M, Gomez JM, Heitmann
BL, Kent-Smith L, Melchior JC, Pirlich M, Scharfetter H, Schols AM, and Pichard C.
Bioelectrical impedance analysis--part I: review of principles and methods. Clin Nutr.
2004;23(5):1226-43.
60. Kyle UG, Bosaeus I, De Lorenzo AD, Deurenberg P, Elia M, Manuel Gomez J,
Lilienthal Heitmann B, Kent-Smith L, Melchior JC, Pirlich M, Scharfetter H, A MWJS,
and Pichard C. Bioelectrical impedance analysis-part II: utilization in clinical practice.
Clin Nutr. 2004;23(6):1430-53.
61. Larsson B, Svardsudd K, Welin L, Wilhelmsen L, Bjorntorp P, and Tibblin G.
Abdominal adipose tissue distribution, obesity, and risk of cardiovascular disease and
death: 13 year follow up of participants in the study of men born in 1913. Br Med J (Clin
Res Ed). 1984;288(6428):1401-4.
41
62. Laskey MA. Dual-energy X-ray absorptiometry and body composition. Nutrition.
1996;12(1):45-51.
63. Liang MT, and Norris S. Effects of skin blood flow and temperature on bioelectric
impedance after exercise. Med Sci Sports Exerc. 1993;25(11):1231-9.
64. Lohman T. Body composition methodology in sports medicine. Phys Sportsmed.
1982;10:47-58.
65. Lohman T. Advances in body composition assessment. Champaign, IL: Human
Kinetics; 1992.
66. Lohman TG, Boileau RA, and Massey BH. Prediction of lean body mass in young
boys from skinfold thickness and body weight. Hum Biol. 1975;47(3):245-62.
67. Lohman TG, Harris M, Teixeira PJ, and Weiss L. Assessing body composition and
changes in body composition. Another look at dual-energy X-ray absorptiometry. Ann N
Y Acad Sci. 2000;904:45-54.
68. Lukaski HC, Bolonchuk WW, Hall CB, and Siders WA. Validation of tetrapolar
bioelectrical impedance method to assess human body composition. J Appl Physiol.
1986;60(4):1327-32.
69. Marks SJ, Moore NR, Clark ML, Strauss BJ, and Hockaday TD. Reduction of
visceral adipose tissue and improvement of metabolic indices: effect of dexfenfluramine
in NIDDM. Obes Res. 1996;4(1):1-7.
70. Martin AD, Daniel MZ, Drinkwater DT, and Clarys JP. Adipose tissue density,
estimated adipose lipid fraction and whole body adiposity in male cadavers. Int J Obes
Relat Metab Disord. 1994;18(2):79-83.
71. Martin AD, Drinkwater DT, and Clarys JP. Human body surface area: validation of
formulae based on a cadaver study. Hum Biol. 1984;56(3):475-88.
72. Matsuzawa Y, Funahashi T, and Nakamura T. The concept of metabolic syndrome:
contribution of visceral fat accumulation and its molecular mechanism. J Atheroscler
Thromb. 2011;18(8):629-39.
73. McCrory MA, Gomez TD, Bernauer EM, and Mole PA. Evaluation of a new air
displacement plethysmograph for measuring human body composition. Med Sci Sports
Exerc. 1995;27(12):1686-91.
74. Miyatake N, Nonaka K, and Fujii M. A new air displacement plethysmograph for the
determination of Japanese body composition. Diabetes Obes Metab. 1999;1(6):347-51.
75. Nevill AM, Metsios GS, Jackson AS, Wang J, Thornton J, and Gallagher D. Can we
use the Jackson and Pollock equations to predict body density/fat of obese individuals in
the 21st century? Int J Body Compos Res. 2008;6(3):114-21.
42
76. Nunez C, Kovera AJ, Pietrobelli A, Heshka S, Horlick M, Kehayias JJ, Wang Z, and
Heymsfield SB. Body composition in children and adults by air displacement
plethysmography. Eur J Clin Nutr. 1999;53(5):382-7.
77. O'Brien C, Young AJ, and Sawka MN. Bioelectrical impedance to estimate changes
in hydration status. Int J Sports Med. 2002;23(5):361-6.
78. Ormseth MJ, Lipson A, Alexopoulos N, Hartlage GR, Oeser AM, Bian A,
Gebretsadik T, Shintani A, Raggi P, and Stein CM. Epicardial adipose tissue is associated
with cardiometabolic risk and the metabolic syndrome in patients with rheumatoid
arthritis. Arthritis Care Res (Hoboken). 2013.
79. Pace N, Rathbun, EN. Studies on body composition, III: The body water and
chemically combined nitrogen content in relation to fat content. Journal of Biological
Chemistry. 1945;158:685-91.
80. Panel E. Clinical guidelines on the identification, evaluation, and treatment of
overweight and obesity in adults: executive summary. Expert Panel on the Identification,
Evaluation, and Treatment of Overweight in Adults. Am J Clin Nutr. 1998;68(4):899917.
81. Patel RV, Peterson EL, Silverman N, and Zarowitz BJ. Estimation of total body and
extracellular water in post-coronary artery bypass graft surgical patients using single and
multiple frequency bioimpedance. Crit Care Med. 1996;24(11):1824-8.
82. Pietrobelli A, Formica C, Wang Z, and Heymsfield SB. Dual-energy X-ray
absorptiometry body composition model: review of physical concepts. Am J Physiol.
1996;271(6 Pt 1):E941-51.
83. Pietrobelli A, Wang Z, Formica C, and Heymsfield SB. Dual-energy X-ray
absorptiometry: fat estimation errors due to variation in soft tissue hydration. Am J
Physiol. 1998;274(5 Pt 1):E808-16.
84. Pollock ML, Hickman T, Kendrick Z, Jackson A, Linnerud AC, and Dawson G.
Prediction of body density in young and middle-aged men. J Appl Physiol.
1976;40(3):300-4.
85. Pollock ML, Laughridge EE, Coleman B, Linnerud AC, and Jackson A. Prediction of
body density in young and middle-aged women. J Appl Physiol. 1975;38(4):745-9.
86. Powers JS, Buchowski M, Wang L, and Otoo-Boameh A. Total body water in elderly
adults--assessing hydration status by bioelectrical impedance analysis vs urine
osmolality. J Am Geriatr Soc. 2012;60(2):388-90.
87. Prior BM, Modlesky CM, Evans EM, Sloniger MA, Saunders MJ, Lewis RD, and
Cureton KJ. Muscularity and the density of the fat-free mass in athletes. J Appl Physiol.
2001;90(4):1523-31.
43
88. Rimm EB, Stampfer MJ, Giovannucci E, Ascherio A, Spiegelman D, Colditz GA,
and Willett WC. Body size and fat distribution as predictors of coronary heart disease
among middle-aged and older US men. Am J Epidemiol. 1995;141(12):1117-27.
89. Ross R. Magnetic resonance imaging provides new insights into the characterization
of adipose and lean tissue distribution. Can J Physiol Pharmacol. 1996;74(6):778-85.
90. Ross R, Pedwell H, and Rissanen J. Response of total and regional lean tissue and
skeletal muscle to a program of energy restriction and resistance exercise. Int J Obes
Relat Metab Disord. 1995;19(11):781-7.
91. Rossner S, Bo WJ, Hiltbrandt E, Hinson W, Karstaedt N, Santago P, Sobol WT, and
Crouse JR. Adipose tissue determinations in cadavers--a comparison between crosssectional planimetry and computed tomography. Int J Obes. 1990;14(10):893-902.
92. Rothschild PA, Domesek JM, Eastham ME, and Kaufman L. MR imaging of
excessively obese patients: the use of an open permanent magnet. Magn Reson Imaging.
1991;9(2):151-4.
93. Ryder J, Ball, SD. Three-dimensional body scanning as a novel technique for body
composition assessment: a preliminary investigation. Journal of Exercise Physiology
online. 2012;15:1-14.
94. Sardinha LB, Lohman TG, Teixeira PJ, Guedes DP, and Going SB. Comparison of air
displacement plethysmography with dual-energy X-ray absorptiometry and 3 field
methods for estimating body composition in middle-aged men. Am J Clin Nutr.
1998;68(4):786-93.
95. Schlemmer A, Hassager C, Haarbo J, and Christiansen C. Direct measurement of
abdominal fat by dual photon absorptiometry. Int J Obes. 1990;14(7):603-11.
96. Schutte JE, Townsend EJ, Hugg J, Shoup RF, Malina RM, and Blomqvist CG.
Density of lean body mass is greater in blacks than in whites. Journal of applied
physiology: respiratory, environmental and exercise physiology. 1984;56(6):1647-9.
97. Segal KR, Gutin B, Presta E, Wang J, and Van Itallie TB. Estimation of human body
composition by electrical impedance methods: a comparative study. J Appl Physiol.
1985;58(5):1565-71.
98. Seidell JC, Bakker CJ, and van der Kooy K. Imaging techniques for measuring
adipose-tissue distribution--a comparison between computed tomography and 1.5-T
magnetic resonance. Am J Clin Nutr. 1990;51(6):953-7.
99. Siri WE. The gross composition of the body. Adv Biol Med Phys. 1956;4:239-80.
100. Siri WE. Body Composition from fluid spaces and density: Analysis of methods.
Techniques for measuring body composition. Washington, DC: National Acadamy of
Sciences; 1961.
44
101. Sjostrom L. A computer-tomography based multicompartment body composition
technique and anthropometric predictions of lean body mass, total and subcutaneous
adipose tissue. Int J Obes. 1991;15 Suppl 2:19-30.
102. Sjostrom L, Kvist H, Cederblad A, and Tylen U. Determination of total adipose
tissue and body fat in women by computed tomography, 40K, and tritium. Am J Physiol.
1986;250(6 Pt 1):E736-45.
103. Smith SR, and Zachwieja JJ. Visceral adipose tissue: a critical review of
intervention strategies. Int J Obes Relat Metab Disord. 1999;23(4):329-35.
104. Starck G, Lonn L, Cederblad A, Alpsten M, Sjostrom L, and Ekholm S. Dose
reduction for body composition measurements with CT. Appl Radiat Isot. 1998;49(56):561-3.
105. Thomasset MA. [Bioelectric properties of tissue. Impedance measurement in clinical
medicine. Significance of curves obtained]. Lyon Med. 1962;94:107-18.
106. Tokunaga K, Matsuzawa Y, Ishikawa K, and Tarui S. A novel technique for the
determination of body fat by computed tomography. Int J Obes. 1983;7(5):437-45.
107. Wagner DR, and Heyward VH. Validity of two-component models for estimating
body fat of black men. J Appl Physiol. 2001;90(2):649-56.
108. Wang W, Wang Z, Faith MS, Kotler D, Shih R, and Heymsfield SB. Regional
skeletal muscle measurement: evaluation of new dual-energy X-ray absorptiometry
model. J Appl Physiol. 1999;87(3):1163-71.
109. Wang Z, Heymsfield SB, Chen Z, Zhu S, and Pierson RN. Estimation of percentage
body fat by dual-energy x-ray absorptiometry: evaluation by in vivo human elemental
composition. Phys Med Biol. 2010;55(9):2619-35.
110. Wang Z, Heymsfield SB, Pi-Sunyer FX, Gallagher D, and Pierson RN, Jr. Body
composition analysis: Cellular level modeling of body component ratios. Int J Body
Compos Res. 2008;6(4):173-84.
111. Wang ZM, Pierson RN, Jr., and Heymsfield SB. The five-level model: a new
approach to organizing body-composition research. Am J Clin Nutr. 1992;56(1):19-28.
112. Wang ZM, Visser M, Ma R, Baumgartner RN, Kotler D, Gallagher D, and
Heymsfield SB. Skeletal muscle mass: evaluation of neutron activation and dual-energy
X-ray absorptiometry methods. J Appl Physiol. 1996;80(3):824-31.
113. Watts K, Naylor LH, Davis EA, Jones TW, Beeson B, Bettenay F, Siafarikas A, Bell
L, Ackland T, and Green DJ. Do skinfolds accurately assess changes in body fat in obese
children and adolescents? Med Sci Sports Exerc. 2006;38(3):439-44.
45
114. Wiklund P, Toss F, Weinehall L, Hallmans G, Franks PW, Nordstrom A, and
Nordstrom P. Abdominal and gynoid fat mass are associated with cardiovascular risk
factors in men and women. J Clin Endocrinol Metab. 2008;93(11):4360-6.
115. Wilkins LW. ACSM's Guidelines for Exercise Testing and Prescription.
Philidelphia; 2006.
116. Withers RT, LaForgia J, Pillans RK, Shipp NJ, Chatterton BE, Schultz CG, and
Leaney F. Comparisons of two-, three-, and four-compartment models of body
composition analysis in men and women. J Appl Physiol. 1998;85(1):238-45.
117. Yusuf S, Hawken S, Ounpuu S, Bautista L, Franzosi MG, Commerford P, Lang CC,
Rumboldt Z, Onen CL, Lisheng L, Tanomsup S, Wangai P, Jr., Razak F, Sharma AM,
and Anand SS. Obesity and the risk of myocardial infarction in 27,000 participants from
52 countries: a case-control study. Lancet. 2005;366(9497):1640-9.
46
APPENDIX B: %BF TABLE FOR DC EQUATION
47