Evaluation of the BOD POD for Estimating Percent Body Fat... : The Journal of Strength & Conditioning Research (original) (raw)
Introduction
Many athletes are aware of the importance of body composition and percent body fat (%BF) as it relates to optimal performance (8). In sports that involve power, speed, and endurance, excessive body fat may impede the athletes' ability to jump, run fast, or improve aerobic capacity. Conversely, low %BF is associated with conditions such as disordered eating, iron deficiency anemia, amenorrhea, premature osteoporosis, and sports injuries (3,10). Appearance is known to be a major factor influencing body composition in female athletes, often placing them at a risk for developing health problems caused by inadequate nutrition (17). Women in general and female athletes in particular are very susceptible to developing body image problems because of high demands from society or the sport in which they compete (9,17). Monitoring body composition during the training and competition season with an accurate and reliable method is a useful strategy to ensure proper nutrition for health and athletic competition.
Hydrostatic weighing (HW) is a densitometric method of body composition analysis based on Archimedes' principle. It has been considered the “gold standard” by which other methods of body composition assessment have been validated (8). However, it is time consuming, unpleasant for some subjects, and requires a highly trained technician (18,19). A few studies (2,10,16) have used dual-energy X-ray absorptiometry (DXA) instead of HW as a criterion method for body composition analysis of athletes, suggesting that it might be a new gold standard. Unlike HW, DXA accounts for differences in lean mass (bone mineral density and soft tissue), and therefore its accuracy in measuring %BF might be better for the athletic population. Still, the use of DXA has higher financial costs.
Another commonly used body composition analysis method is the skinfold (SF) measurement. The SF technique is quick and reliable, if done by an experienced technician. Depending on the equation used, the prediction error of percent body fat by SF can vary between 2.2 and 3.5% (8). The skinfold technique is a good alternative for athletes and coaches who do not have access to a laboratory to perform HW for body composition assessment.
Similar to HW, air-displacement plethysmography (ADP) is a densitometric method. The BOD POD Body Composition System (Life Measurement Instrument, Concord, California) is a commercial device using ADP that was developed to provide an alternative method for estimating %BF other than the traditional HW (5). The BOD POD measures body volume by air displacement to calculate body density (Db). Previous studies comparing ADP with HW in men and women have reported close agreement (7,13,18), whereas other investigations found the results to be inconsistent (1,4,19,21).
It is beneficial to compare the BOD POD with HW because both are densitometric methods and, theoretically, should produce similar results. In addition, it is valuable to investigate and compare %BF estimation by the BOD POD with the DXA measurement to establish their relationship in body composition assessment of the athletic population. The primary purpose of this study was to examine the accuracy of %BF estimations obtained by the BOD POD (BFBP) in a group of Division I collegiate track and field female athletes (N = 30) using HW as the gold standard. A secondary purpose was to compare and determine the relationship between BFBP and the estimation of percent body fat by SF (BFSF) and DXA (BFDXA) in this group of female athletes.
Methods
Experimental Approach to the Problem
A female athlete's health and athletic performance are greatly affected by her %BF. The availability of a reliable, accurate, and easy-to-administer body composition analysis tool allows many coaches and athletes to monitor body composition for health and performance. Hydrostatic weighing requires specialized facilities, a trained technician, and ample time to complete. Additionally, it may be unpleasant for some subjects. The BOD POD was developed to create an administratively advantageous method for estimating %BF other than HW.
This investigation examines the accuracy of the BOD POD estimation of %BF in a group of track and field female athletes (N = 30) while comparing it with the results from three other body composition analysis tools (HW, DXA, and SF) commonly used on this population. A short health questionnaire was completed by each subject before data collection. Data were collected at two laboratories at a Southwestern university in the United States. Excluding the residual volume (Rv), which was measured earlier, all measurements were taken on the same day for each subject. Because the BOD POD requires the subject's body to be completely dry, the testing order was DXA and BOD POD first, followed by SF and HW.
Subjects
Thirty Division I collegiate female athletes (18-24 years) volunteered from the university track and field team during their competition season. Physical characteristics of the female athletes included in this study are presented in Table 1. The ethnic composition of the sample was 60% Caucasian, 20% Hispanic, 10% black, 3.3% Asian, and 6.7% other (mixed ethnicities). Before the study, the primary investigator explained the purposes, goals, and procedures of the study to each volunteer. All subjects signed a consent form approved by the human research review committee of the university. Subjects were asked not to eat or exercise for 4 hours before testing. Because the DXA assessment involves exposure to radiation, a pregnancy test was conducted on every woman not taking birth control peels before testing to ensure she was not pregnant.
Physical characteristics of subjects (N = 30).
Procedures
Body Height and Weight
Height was measured to the nearest 0.1 cm using a wall-mounted stadiometer (Holtain Ltd., Crymych, Wales). Subjects stood erect, without shoes, with their hands at their sides. Height was recorded at the end of inspiration. After emptying their bladders, body weight was measured to the nearest 0.01 kg using a calibrated digital platform scale (Tanita Corporation, BWB 627A, Japan). Subjects wore only swimsuits.
Air-Displacement Plethysmography
Body volume was measured by the BOD POD using standardized published procedures (13). The BOD POD is a dual-chamber ADP machine, previously described by Dempster and Aitkens (5). The BOD POD was calibrated according to the manufacturer's guidelines using a 50-L cylinder. Subjects wore clothing according to the manufacturer's recommendation (a swimsuit and a swim cap) to rule out air trapped in clothes and hair. As previously noted, each subject was weighed on a calibrated digital scale and then entered the BOD POD chamber. Body volume was measured twice by the machine to ensure measurement reliability. If the first two readings for body volume differed by more than 150 ml, a third measurement was taken. If additional readings were needed, the BOD POD was recalibrated and the measurements were repeated for that subject. Each subject's BFBP was calculated from the Db obtained by the BOD POD (DbBP) using the Siri equation (15).
Thoracic Gas Volume
Thoracic gas volume (Tgv) was measured at the time of the actual BOD POD test following the manufacturer's recommendations explained in detail elsewhere (13). This value was integrated into the calculation of body volume.
Dual-Energy X-Ray Absorptiometry
Each subject's BFDXA was obtained by performing a whole-body scan (slow mode) with a DXA scanner (Lunar DPX, Lunar Radiation Corp., Madison, Wis., using NT software version 3.50.176). All jewelry and any clothing that contained metal were removed before the scan. The machine was calibrated before each day of testing with the manufacturer's “standard block” (a bone-simulating substance of known composition and attenuating capacity), according to the manufacturer's recommendations.
Hydrostatic Weighing
Underwater weight was measured using a four-load cell system (Interface, N. Scottsdale, Ariz.) attached to a platform, and a digital scale (Precision Biomedical, State College, Pa.) integrated to an analog signal-acquisition system (Biopac, MP100, Santa Barbara, Calif.). Subjects entered the tank, removed air bubbles from their swimsuits and hair to minimize possible measurement error, and then positioned themselves with both hands and knees on the load cell platform. When ready, subjects were instructed to go under the water and maximally exhale air from their lungs while fully submerged. The measurement was taken when air bubbles were no longer seen. Subjects performed as many trials as needed until three measurements within 100 g were obtained. The closest three measurements were then averaged and used to calculate body volume and Db. Each subject's BFHW was calculated from Db obtained from HW (DbHW) using the Siri (15) equation.
Residual Lung Volume
Residual lung volume was measured before the HW using a closed-circuit helium dilution technique (Collins, GS modular PFT, Braintree, Mass.). Each subject sat on a chair outside the hydrostatic weighing tank, breathing through a closed circuit. Subjects were given as many trials as needed to obtain two values within 100 ml. The two closest trials were then averaged and used in the calculations of Db and %BF.
Skinfolds
The same trained technician performed the SF measurements for all subjects. Measurements were taken in triplicate in a rotational pattern on the right side of the body, using calibrated Lange SF calipers (Cambridge Scientific Instrument, Cambridge, Md.). Four different sites were measured: triceps, suprailiac, abdominal, and thigh. The average of three measurements for each site was used to calculate body density according to the Jackson and Pollock (11) equation for female athletes ages 18-29. Body density from the SF (DbSF) equation was converted to %BF using the Siri equation.
Statistical Analyses
All analyses were produced using SPSS (SPSS for Windows, version 13, Chicago, Ill.). A repeated-measures analysis of variance (ANOVA) followed by Tukey post hoc comparisons were performed to detect differences between the four body composition analysis methods used in this study. Linear regression and Pearson correlation analyses were used to determine the relationship between the different body composition analysis methods. Criterion alpha level for significance was set at p ≤ 0.05 for all analyses.
Results
Body Fat
Body fat percentages obtained from the BOD POD, HW, DXA, and SF for all female athletes tested are presented in Table 2 along with the results for each track and field specialty group. The repeated-measures ANOVA results revealed a significant difference (p < 0.05) between BFBP, BFHW, BFDXA, and BFSF. Post hoc comparisons showed that BFBP estimations were significantly (p < 0.05) higher than BFHW and significantly (p < 0.05) lower than BFDXA. However, the post hoc comparisons showed no significant (p < 0.05) difference between BFBP and BFSF. Figure 1 illustrates the mean percentages of body fat and their SD of the HW, SF, BOD POD, and DXA.
Comparison of percent body fat (%BF) and body density (Db) (N =30 ).
Mean percent body fat and SD for each method (N = 30).
A Pearson correlation test revealed that BFBP and BFHW were significantly correlated (r = 0.88, SEE = 2.30). This correlation is presented in Figure 2. The correlation of BFSF and BFBP was good (r = 0.85, SEE = 2.05) and is presented in Figure 3. A linear regression analysis produced an _R_2 = 0.71. On the other hand, as shown in Figure 4, BFBP and BFDXA had a poor correlation (r = 0.25, SEE = 5.73). A linear regression analysis produced an _R_2 = 0.064. Figure 5 illustrates the correlation between BFBP, BFSF and BFHW, showing similar trends for each subject.
Scatter plot of the relationship between percent body fat values obtained from hydrostatic weighing (BFHW) and percent body fat values obtained from the BOD POD (BFBP).
Scatter plot of the relationship between percent body fat values obtained from skinfolds (BFSF) and percent body fat values obtained from the BOD POD (BFBP).
Scatter plot of the relationship between percent body fat values obtained from dual-energy X-ray absorptiometry (BFDXA) and percent body fat values obtained from the BOD POD (BFBP).
Subject trends in percent body fat values obtained from the BOD POD (BFBP), percent body fat values obtained from skinfolds (BFSF), and percent body fat values obtained from hydrostatic weighing (BFHW).
Body Density
The repeated-measures ANOVA results revealed a significant difference (p < 0.05) between DbBP, DbHW, and DbSF. Post hoc comparisons show that DbHW was significantly greater than both DbBP and DbSF. In addition, DbSF was significantly (p < 0.05) greater than DbBP.
A Pearson correlation test revealed that DbBP and DbHW as well as DbBP and DbSF were significantly correlated (r = 0.88, SEE = 0.005, and r = 0.85, SEE = 0.005, respectively). A linear regression analysis produced a coefficient of determination _R_2 = 0.77 for DbBP and DbHW and is presented in Figure 6. For the comparison of DbBP and DbSF, a linear regression analysis produced an _R_2 = 0.71 and is presented in Figure 7.
Scatter plot of the relationship between body density values obtained from hydrostatic weighing (DbHW) and body density values obtained from the BOD POD (DbBP).
Scatter plot of the relationship between body density values obtained from skinfolds (DbSF) and body density values obtained from the BOD POD (DbBP).
Discussion
Body composition can be estimated from a two-component model based on the measurement of Db. For densitometric methods, HW is regularly used as the gold standard method for estimation of %BF (8). Because the BOD POD is also based on densitometry, HW was used as the reference method in the present study. In addition, the BOD POD was compared with two other (SF and DXA) commonly used methods of body composition analysis.
McCrory et al. (13) were the first to evaluate the BOD POD in comparison with HW on human subjects. Their data, which were expressed in terms of %BF, reported a nonsignificant mean difference of 0.3% (p < 0.05) (BOD POD was higher) and concluded that the BOD POD was a valid instrument for determining %BF in adult men and women. Since then, other studies (4,6,7,14,18-21) have shown diverse results when examining the validity of the BOD POD against HW. Vescovi et al. (20) have reported a nonsignificant (p < 0.05) difference in the estimation of %BF between the two methods in a sample of heterogeneous adults. In contrast, Millard-Stafford et al. (14) tested a heterogeneous group and found the BOD POD to significantly (p < 0.05) underestimate %BF by 2.8%. On the other hand, Wagner et al. (21) have reported a significant (p < 0.05) overestimation of nearly 2% for black men with the BOD POD.
Few studies have examined the accuracy of the BOD POD on athletes using HW as the reference method (4,6,18,19). Results from previous studies on athletes are presented in Table 3. There is no apparent pattern relating the accuracy of %BF estimation by the BOD POD to that of HW. Investigating the accuracy of the BOD POD on a group of collegiate football players, Collins et al. (4) found its estimation of %BF to be 1.9% lower (significant underestimation) than that of HW. Vescovi et al. (19) showed an inverse trend (significant overestimation) when they tested it on a group of college female athletes. In contrast, both Utter et al. (18) and Dixon at al. (6), who tested the BOD POD against HW on two different groups of collegiate wrestlers, found the difference to be nonsignificant (−0.33 ± 2.34 and 0.7 ± 1.78%, respectively).
A comparison of past research testing the accuracy of the BOD POD.
Similar to Vescovi et al. (19), in the present study the BOD POD was found to significantly (p < 0.05) overestimate %BF of the female athletes. Additionally, a previous investigation by Vescovi et al. (20), who divided their heterogeneous subjects into three groups (lean, average, and overweight), showed no significant difference in %BF between the two methods for the average and overweight groups. They did report a significant difference for the lean group, with the BOD POD overestimating by 2.3 ± 3.5%. A _t_-test between BFHW of the subjects in the present study and the lean group of the Vescovi et al. (20) study shows that the two groups are alike (p = 0.22). Similar to their study, the average %BF for the track and field female athletes in the present study was significantly overestimated (3.9 ± 2.3%, p < 0.05) by the BOD POD. It is important to note that in the Vescovi et al. (20) investigation, the lean group consisted mainly of women. Supported by the findings of Vescovi et al. (20), two possible reasons why the BOD POD differs significantly from HW in the present study are (a) the overestimation of the lean group might suggest a possible limitation in measuring lean subjects, and (b) the overestimation of a group consisting mainly of women might suggest a possible gender bias by the BOD POD.
The strong correlation between %BF estimated by the BOD POD and HW suggests that it has the potential to be useful for body composition measurement of female athletes. However, additional research with this instrument on this population is needed to establish agreement between the two methods before it can be used as an alternative to HW for this population. Until then, practitioners should reconsider the use of the BOD POD for the purpose of estimation of %BF of lean female athletes and, perhaps, use SF for that purpose. Because SF is a field method that is quick and reliable when done by a trained technician, one can save the effort involved with going to a laboratory to use the BOD POD.
The relationship between %BF estimations by the BOD POD and that measured with DXA has been examined a few times in the past. Similar to the present study, Collins et al. (4) found that the %BF estimated by the BOD POD was significantly lower than that measured by DXA, reporting a smaller difference (2%) than the one in the present study (3.7 ± 6.3%). On the other hand, Ballard et al. (2) have recently reported a valid measurement of body composition by the BOD POD when compared with DXA. They found the two methods to be highly correlated and report no significant difference between the two instruments (mean BFBP = 22.5%, BFDXA = 22.0%). Possible explanations for the disagreement in results of this study and the one done by Ballard et al. include the following: (a) their female athlete population consisted of only non-Hispanic, Euro-American women, which might suggest ethnic bias measurement, (b) the DXA scan was done using a different DXA machine (Holigic QDR 4500A software version 12.01, Waltham, Mass.), and (c) a whole-body scan in the present study took 10-15 minutes, and they reported theirs to be approximately 5 minutes. Different than both Collins et al. (4) and Ballard et al. (2), in the present study the correlation between the two methods was poor, indicating a weak relationship between the BOD POD and DXA. As previously mentioned by Lohman (12), differences in calibration procedures, software version, and the instrument's company and model might lead to differences in the results of validation studies. The poor correlation presented here between %BF estimated by the BOD POD and those measured by DXA, as well as the significant difference (p < 0.05) between DXA and HW, should lead researchers to reconsider the use of DXA as a gold standard with female athletes in research until specifications for standardized instrumentation and software are established.
Practical Applications
Although the BOD POD has been available for several years for estimating %BF, its accuracy, especially in athlete populations, is not well established. The use of the skinfold method on female athletes is a good alternative that saves the time and effort of going to a laboratory setting. The good correlation between HW and the BOD POD might eventually secure the use of this instrument as a good body composition analysis tool, but its accuracy will need to be improved in the female athlete population before it can replace HW.
References
1. Ball, SD and Altena, TS. Comparison of the BOD POD and dual energy x-ray absorptiometry in men. Physiol Meas 25: 671-678, 2004.
2. Ballard, TP, Fafara, L, and Vukovich, MD. Comparison of BOD POD and DXA in female collegiate athletes. Med Sci Sports Exerc 36: 731-735, 2004.
3. Cameron-Donaldson, ML. The female athlete triad-a growing health concern. Orthop Nurs. 22: 322-324, 2003.
4. Collins, MA, Millard-Stafford, ML, Sparling, PB, Snow, TK, Rosskopf, LB, Webb, SA, and Omer, J. Evaluation of the BOD POD for assessing body fat in collegiate football players. Med Sci Sports Exerc 3: 1350-1356, 1999.
5. Dempster, P and Aitkens, S. A new air displacement method for the determination of human body composition. Med Sci Sports Exerc 27: 1692-1997, 1995.
6. Dixon, CB, Deitrick, RW, Pierce, JR, Cutrufello, PT, and Drapeau, LL. Evaluation of the BOD POD and leg-to-leg bioelectrical impedance analysis for estimating percent body fat in national collegiate athletic association Division III collegiate wrestlers. J Strength Cond Res 19: 85-91, 1995.
7. 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 13: 1232-1237, 2005.
8. Heyward, VH and Wagner, DR. Applied Body Composition Assessment (2nd ed.). Champaign: Human Kinetics, 2004.
9. Hopkinson, RA and Lock, J. Athletics, perfectionism, and disordered eating. Eat Weight Disord 9: 99-106, 2004.
10. Houtkooper, LB, Mullins, VA, Going, SB, Brown, CH, and Lohman, TG. Body composition profiles of elite American heptathletes. Int J Sport Nutr Exerc Metab 11: 162-173, 2001.
11. Jackson, AS, Pollock, ML, and Ward, A. Generalized equations for predicting the body density of women. Med Sci Sports Exerc 12: 175-181, 1980.
12. Lohman, TG. Dual energy X-ray absorptiometry. In: Human Body Composition. A.F. Roche, S.B. Heymsfield, and T.G. Lohman, eds. Champaign: Human Kinetics, 1996. pp. 63-78.
13. McCrory, MA, Gomez, TD, Bernauer, EM, and Mole, PA. Evaluation of air displacement plethysmography for measuring human body composition. Med Sci Sports Exerc 27: 1686-1691, 1995.
14. Millard-Stafford, ML, Collins, MA, Evans, EM, Snow, TK, Cureton, KJ, and Rosskopf, LB. Use of air displacement plethysmography for estimating body fat in a four-component model. Med Sci Sports Exerc 33: 1311-1317, 2001.
15. Siri, WE. Body composition from fluid spaces and density: analysis of methods. In: Techniques for Measuring Body Composition. J. Brozek, A. Henschel, eds. Washington, DC: National Academy of Sciences, National Research Council, 1961. pp. 223-224.
16. Stewart, AD and Hannan, WJ. Prediction of fat and fat-free mass in male athletes using dual energy x-ray absorptiomentry as the reference method. J Sports Sci 18: 263-274, 2000.
17. Sundgot-Borgen, J. Risk trigger factors for the development of eating disorders in female elite athletes. Med Sci Sports Exerc 26: 414-419, 1994.
18. Utter, AC, Goss, FL, Swan, PD, Harris, AS, Robertson, RJ, and Trone, GA. Evaluation of air displacement for assessing body composition of collegiate wrestlers. Med Sci Sports Exerc 35: 500-505, 2003.
19. Vescovi, JD, Hilderbrandt, L, Miller, W, Hammer, R, and Spiller, A. Evaluation of the BOD POD for estimating percent fat in female college athletes. Int Strength Cond Res 16: 599-605, 2002.
20. Vescovi, JD, Zimmerman, SL, Miller, WC, Hilderbrandt, L, Hammer RL, and Fernhall, B. Evaluation of the BodPod for estimating percentage body fat in a heterogeneous group of adult humans. Eur J Appl Phys 85: 326-332, 2001.
21. Wagner, DR, Heyward, VH, and Gibson, AL. Validation of air displacement plethusmography for assessing body composition. Med Sci Sports Exerc 32: 1339-1344, 2000.
Keywords:
air-displacement plethysmography; body composition;
© 2008 National Strength and Conditioning Association