EXERCISE AND QUALITY OF LIFE
Volume 3, No. 2, 2011, 43-48
BODY MASS INDEX AND BODY FAT CONTENT
IN ELITE ATHLETES
Jelena Popadi„ Ga„eöa
, Otto Barak, Dea Karaba Jakovljevic, Aleksandar Klaönja,
Vladimir Gali„, Miodrag Drapöin, Damir LukaË and Nikola Gruji„
Department of Physiology, Faculty of Medicine
University of Novi Sad, Serbia
The aim of this study was to evaluate body fat content (BF) of elite athletes obtained by two
different field methods for body composition measurements and to compare it with body mass
index (BMI) values. The research was conducted on 40 male athletes (20 runners and 20
handball players) and 30 non athletes. BF was calculated from the skinfold values (BFsft) and
estimated using a hand-held impedance analyzer (BFbia%). Body mass index, waist to hip ratio
(WHR) and waist to stature ratio (WSR) were calculated from adequate anthropometric values.
Comparing the BF content between non athletes and two different sport groups, significant
difference was found in all parameters between runners and non athletes (p < 0.05). Significant
difference was found between BF values of runners and handball players (p < 0.05). Runners
have had significantly lower BF, estimated by both methods. They also have had significantly
lower WHR and WSR (p < 0.05). In the group of athletes and non athletes with BMI higher than
, or lower than 20 kg/m
, comparing with others, no significant difference was found in
BFsft and WHR. BMI is not a good predictor of BF, because it does not provide specific
information about body fatness, but rather body heaviness. Bioimpedance and anthropometry
methods could be used to monitor non obese subjects in clinical routine and population based
studies. For BF estimation in athletes, we recommend anthropometry, rather than bioimpedance
because of inter individual and inter sports variations in arms length and regional masculinity.
body fat, BMI, body composition, anthropometry, athletes
Athletic performance is partially influenced by body composition characteristics of an
athlete. Measurements of total or whole body fat content range from indirect estimates based on
compartmental modeling approaches, to more direct measures of adipose tissue volume, such as
in-vivo magnetic resonance imaging ñ MRI (Goodpaster, 2002).
Corresponding author. Department of Physiology, Faculty of Medicine, University of Novi Sad, Hajduk Veljko St.
3, 21000 Novi Sad, Serbia, e-mail:
© 2010 Faculty of Sport and Physical Education, University of Novi Sad, Serbia
J. Popadi„ Ga„eöa et al.
Bioelectrical impedance (BIA) and skinfold thickness (SFT) techniques, so-called field
methods, are simple, readily available and noninvasive ones, along with the advantage of being
radiation free. BIA and SFT method could be used to monitor non obese subjects in clinical
routine and population based studies. The prediction equations are derived from lean subjects
(Erselcan et al. 2000), and the prediction was generally good at a population level (Deurenberg et
Body mass index (BMI), like all anthropometric measurements, is only a surrogate
measure of body fatness. In specific groups such as athletes and women, military personnel,
police, firemen, etc., they are considerably leaner than indicated by their BMI. This relationship
holds for most sport personnel, even down to the very low BMI associated with elite runners
The aim of this study was to evaluate body fat content of elite athletes obtained by two
different field methods for body composition measurements and to compare it with BMI values.
The research was conducted on 40 male athletes, with average age 20.8 years, who have
been active athletes for 7.8 years. They have been divided into subgroups, according to they
sport specialties: 20 runners (active for 6.0 ± 3 years) and 20 handball players (active for 9.5 ± 4
years). They were all healthy, and voluntarily participated in the study. Control group was
consisted of 30 non athletes, medical students, and they were not participated in any sport
activity in the past 6 months.
The study was approved by the Ethics Committee of the Medical School of the
University of Novi Sad and the investigation was performed according to the principles outlined
in the Declaration of Helsinki.
Anthropometric measures, used in this research, were: Body weight (W), body height
(H), 7 skinfolds (chest, midaxillary, triceps, subscapular, abdomen, anterior suprailiac and mid
thigh skinfold thicknesses) and 5 circumferences (arm relaxed, arm flexed and tensed, waist,
chest and hip). All anthropometric measurements were conducted under the recommendations
and rules given by Heywarth and Stolarczyk (1996). At each skinfold site, measurements were
conducted with Holtain-Koln caliper, to the nearest 0.1mm. Girths were measured with a flexible
steel tape and all circumferences were recorded to the nearest millimeter.
After that, values of body mass index (BMI), waist to hip ratio (WHR) and waist to
(WSR) were calculated. Body mass index
(BMI ñ kg/m
) was derived as
, waist to hip ratio ñ WHR ñ was calculated as waist circumference divided by hip
circumference, and waist to stature ratio ñ WSR ñ as waist circumference divided by height. For
normal values of BMI, the references from World Health Organization were used (WHO, 1997)
ñ BMI 18.5 to 24.9 kg/m
for lean individuals, border values from 25 to 29.9 kg/m
, and obesity
BMI higher than 30 kg/m
For body composition analysis, total body fat was calculated from the skinfold and
(BFsft), according to the recommendations given by Heywarth and
Stolarczyk (1996). Other method for estimation of body fat was using a hand-held impedance
analyzer - Omron BF300. During the measurement the instrument recorded impedance from
hand to hand and consequently calculated body fat percentage from the impedance value and the
pre-entered personal particulars
(weight, height, age and sex). Two values were obtained:
percentage of total body fat (BFbia%) and total body fat mass (BFbia).
BMI and body fats in elite athletes
All data were presented as mean values ± standard deviation (SD). In statistical analysis,
variance analysis ñ one way ANOVA was used, as well as Pearsonís correlation coefficient.
Descriptive characteristics for athletes and non athletes are shown in Table 1.
Characteristics of participants (non-athletes n = 30; runners n = 20; handball players n = 20)
Waist C (cm)
Hip C (cm)
The average values for body fat content in athletes, calculated from skinfold thickness ñ
SFT (BFsft) were 10.2 ± 3.2% and 11.5 ± 3.9% using bioimpedance (BFbia). Average BMI in
athletes was 23.9 ± 5.1 kg/m
. No significant difference was found between BMI and body fat
values of athletes and non athletes. Non athletes were lean, with average BMI
23.8 ± 2.2 kg/m
and average body fat content 12.7 ± 4.1% measured by BIA.
Comparing the body fat content (BMI, BFbia, WHR, WSR) between non athletes and
two different sport groups, significant difference was found in all parameters between runners
and non athletes (p< 0.05). On the other hand, between handball players and non athletes no
significant difference was found in parameters of BFbia and WHR (p> 0.05).
Significant difference was found between body fat values of runners and handball players
(p< 0.05). Runners have had significantly lower body fat, estimated by skinfold thickness and
bioimpedance measurements. They also have had significantly lower WHR and WSR (p< 0.05).
All participants were divided into two subgroups, those with BMI lower than 24,9 kg/m
and those with BMI higher than 25 kg/m
. Comparing body fat content between these two
subgroups, there were no significant difference found in BFsft and WHR in runners and handball
players. In the group of non athletes WHR was not significantly different (p> 0.05). Body fat
content values measured by BIA were significantly different between these subgroups (p< 0.05).
Comparing the body fat content (BFsft and BFbia) in the group of athletes with BMI lower than
and those with BMI higher than 20.0 kg/m
, there were no significant differences
found (p> 0.05). There was no correlation between BMI and BF% in athletes, because BMI is a
surrogate of body mass. Since athletes are heavier (especially handball players) due to their
muscle mass, it directly leads to higher values of BMI. Mismatch between BMI and BF in
athletes is shown in Figure 1.
J. Popadi„ Ga„eöa et al.
Mismatch between body mass index (BMI) and body fat in athletes
Values of BF% calculated from SFT correlate with values measured from BIA (r = 0.87)
in all athletes.
Since athletic performance is partially influenced by the ratio of oneís fat mass (FM) to
fat-free mass (FFM), most athletes are concerned with their body composition.
The proportions of water, protein and mineral in the fat-free body, and thus the overall
density of the fat-free body (FFBd) vary with age, gender, ethnicity, level of body fatness and
physical activity level (Bottaro et al. 2002).
A variety of methods has been developed to determine body composition. Each of these
methods has individual limitations in its assumption, calibration, accuracy and precision. In the
present study, the most widely used two methods of body composition analysis, which are
readily available in clinical routine, were evaluated in athletes and non athletes.
One alternative field technique that is simple, inexpensive and noninvasive is the
measurement of subcutaneous fat thickness, or skinfolds, at selected sites. Since Matiega (1921)
developed the first equations to estimate body fat based on skinfold thickness, several
subsequent equations have been developed to estimate fat free mass or fat mass based on a two-
compartment model from body density (Heyward & Stolarczyk 1996). The error in body fat
estimates from skinfold thickness ranges from +- 3 to +- 11%, and is influenced by sex, race and
age. The reliability of body fat estimates from either skin folds or circumferences is largely
dependent upon the skill of the examiner (Goodpaster, 2002). Bioimpedance (BIA) is based on
the principle that the electrical conductivity of the fat-free mass is greater than that of fat. The
method assumes a cylindrical model of the body and a normal hydration of the fat-free mass.
In our study, average body fat percentages for runners were 8.4 ± 3.0 %, calculated from
SFT and 9.2 ± 2.9 % measured from BIA. Average body fat percentages for handball players
were 12.1 ± 4.4 %, calculated from SFT and 13.8 ± 3.6 % measured from BIA. According to
Heyward and Stolarczyk (1996), average body fat percentages of runners are
handball players are 10-12%. Average body fat percent in non athletes was 12.7 ± 4.1 %
measured by BIA.
BMI and body fats in elite athletes
It was shown in earlier studies (Baumgarthner et al. 1989; Fuller & Elia, 1989; Loy et al.
1998) that segmental impedance measurements (measuring only segments of the body as the legs
or arms) also allow fairly accurate assessments of body composition. These segmental
impedance instruments are easy to use and have the advantage that they are relatively
inexpensive (Deurenberg et al. 2001).
Comparing the results of body fat percentage in our participants, it was shown that BMI
is not a good predictor of BF%. It does not provide specific information about body fat, but
rather body size. In the group of athletes and non athletes with BMI higher than 25 kg/m
lower than 20 kg/m
, comparing with others, no significant difference was found in BFsft and
On the other hand, BF measured by BIA showed significant difference between these
Estimation of body fat from BIA showed different distribution of body fat content than
those calculated from SFT and WHR parameter in the group of athletes and lean non athletes.
This observation might be explained by characteristics of BIA analyzer. In subjects with
relatively larger arm muscles, the total amount of fat-free mass will be overestimated with BIA
technique and hence BF% will be underestimated. Also, in subjects with relatively long arms, the
measured impedance will be high and hence fat-free mass low and thus calculated BF% high
(Snijder et al. 1999). There are reported differences in relative arm length among populations
and it is known that within population groups the variability in arm length is high (Eveleth &
WHR can be used as a good predictor of body fat content in lean subjects (Srdic et al.
2003). This parameter showed a good correlation with BF% in athletes (r = 0.65).
Comparing the results of body fat percent in athletes and non athletes, it was shown that there is
a good correlation between values derived from SFT and BIA in the present study (r = 0.82 for
runners, r = 0.87 for handball players, r = 0.88 for all athletes).
In conclusion, for estimation of body fat percent in athletes, we recommend skinfold
thickness method, rather than bioimpedance, because of inter individual and inter sports
variations in arms length and regional muscle mass. Body mass index is not a good predictor of
body fat percent. It does not provide specific information about body fatness, but rather body
Baumgarthner, R. N., Chumlea, W. C., & Roche, A. F. (1989). Estimation of body composition
from bioelectrical impedance of body segments.
The American Journal of Clinical
Bottaro, M. F., Heyward, V. H., Bezerra, R. F. A., & Wagner, D. R. (2002). Skinfold method vs
dual-energy X-ray absorptiometry to assess body composition in normal and obese
Journal of Exercise Physiology
Deurenberg, P., Andreoli, A., Borg, P. et al.
(2001). The validity of predicted body fat
percentage from body mass index and from impedance in samples of five European
European Journal of Clinical Nutrition
Erselcan, T., Candad, F., Saruhan, S., & Ayca, T. (2000). Comparison fo body composition
analysis methods in clinical routine.
Annals of Nutrition and Metabolism
J. Popadi„ Ga„eöa et al.
Eveleth, P. B., & Tanner, J. M. (1976).
World-Wide Variation in Human Growth
Cambridge University Press.
Fuller, N. J., & Elia, M. (1989). Potential use of bioelectrical impedance of the whole body and
of body segments for the assessment of body composition: comparison with densitometry
The American Journal of Clinical Nutrition, 43
Goodpaster, B. H. (2002). Measuring body fat distribution and content in humans.
Opinion in Clinical Nutrition & Metabolic Care, 5,
Heyward, H. V., & Stolarczyk, M. L.
Applied Body Composition Assessment
Champaign, IL: Human Kinetics.
Loy, S. F., Likes, E. A., Andrews, P. M. et al.
(1998). Easy grip on body composition
ACSM's Health & Fitness Journal, 2
(1921). The testing of physical efficiency.
American Journal of Physical
Prentice, A. M., & Jebb, S. A. (2001). Beyond body mass index.
The International Association
for the Study of Obesity, Obesity reviews,
Snijder, M. B., Kuyf, B. E. M., & Deurenberg, P. (1999). Effect of body build on the validity of
predicted body fat from body mass index and bioelectrical impedance.
Nutrition and Metabolism,
Srdic, B., Stokic, E., & Polzovic, A. (2003). Parameters defining size and distribution of body fat
tissue: relationship analysis.
Obesity-Preventing and Managing the Global Epidemic
. Geneve: Report of
WHO consultations on obesity.
Submitted October 9, 2011
Accepted December 15, 2011