SEARCH

SEARCH BY CITATION

Keywords:

  • body density;
  • body composition;
  • air displacement plethysmography;
  • body hair;
  • body volume

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Objective: The objective of this study was to determine the effect of body hair (scalp and facial) on air displacement plethysmography (BOD POD) estimates of percentage of body fat.

Research Methods and Procedures: A total of 25 men (31.4 ± 8.0 years, 83.4 ± 12.2 kg, 181.8 ± 6.9 cm) agreed to grow a beard for 3 weeks to participate in the study. Total body density (g/cm3) and percentage of body fat were evaluated by BOD POD. To observe the effect of trapped isothermal air in body hair, BOD POD measures were performed in four conditions: criterion method (the beard was shaven and a swimcap was worn), facial hair and swimcap, facial hair and no swimcap, and no facial hair and no swimcap.

Results: The presence of only a beard (facial hair and swimcap) resulted in a significant underestimation of percentage of body fat (16.2%, 1.0618 g/cm3) vs. the criterion method (17.1%, 1.0597 g/cm3, p < 0.001). The effect of scalp hair (no swim cap worn) resulted in a significant underestimation in percentage of body fat relative to the criterion method, either with facial hair (facial hair and no swimcap; 14.8%, 1.0649 g/cm3) or without facial hair (no facial hair and no swimcap; 14.8%, 1.0650 g/cm3, p < 0.001 for both).

Discussion: A significant underestimation of percentage of body fat was observed with the presence of facial hair (∼1%) and scalp hair (∼2.3%). This underestimation in percentage of body fat may be caused by the effect of trapped isothermal air in body hair on body-volume estimates. Thus, excess facial hair should be kept to a minimum and a swimcap should be worn at all times to ensure accurate estimates of body fat when using the BOD POD.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Air displacement plethysmography (BOD POD) has gained widespread use in the determination of percentage of body fat partially because it is a fast and noninvasive alternative to hydrostatic weighing. The BOD POD applies basic gas laws for the determination of body volume using the pressure-volume relationship. Boyle's Law describes the pressure-volume relationship when temperature remains constant (isothermal conditions): PV = k; where P is pressure, V is volume, and k is the constant (1). However, it is difficult to maintain constant temperature throughout the testing procedures (2). Therefore, Poisson's Law, PVγ = k, more accurately describes the pressure-volume relationship when temperature is changing (adiabatic conditions) (1), where γ is the ratio of the specific heat of the gas at constant pressure to the specific heat of the gas at constant volume (3). The distinction between the two air conditions is important because isothermal air behaves differently from adiabatic air. Isothermal air is 40% more compressible than adiabatic air, and produces a lower pressure output signal for a given body volume; as a result, body volume would be underestimated in the presence of isothermal air if instrument software did not make appropriate corrections (4).

Potential complications caused by the different air conditions (adiabatic vs. isothermal) in the BOD POD are minimized by the use of an air circulation mechanism and an oscillating diaphragm that causes sinusoidal perturbation (4). However potential sources of isothermal air are present in the lungs, skin surface area, clothing, and body hair. The BOD POD software (4) corrects for the effect of isothermal air from the lungs and skin surface area. Recently, it has been shown that different clothing schemes can result in an underestimation of percentage of body fat by BOD POD (∼5%) compared with hydrostatic weighing (5). However until now no study has investigated the effect of body hair on BOD POD estimates of percentage of body fat. Therefore, the purpose of this study was to determine the effect of facial (beard) and scalp hair on BOD POD estimates of total body density and percentage of body fat.

Research Methods and Procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Subjects

A total of 25 men (31.4 ± 8.0 years, 83.4 ± 12.2 kg, 181.8 ± 6.9 cm) were recruited from the Birmingham, Alabama metropolitan area. Individuals agreed to engage in a 3-week period of beard growth where no shaving was permitted, before testing. All testing was performed in the Physiology and Metabolism Laboratory in the Department of Nutrition Sciences, University of Alabama at Birmingham. Approval was obtained from the University of Alabama at Birmingham Institutional Review Board for human use, and written informed consent was obtained before testing.

Experimental Design

All subjects were tested in a “Speedo” (Speedo, Los Angeles, CA) swimsuit supplied by the laboratory. To observe the effect of trapped isothermal air present in facial (beard) and scalp hair, individuals were tested in four conditions. Condition 1: criterion method (the beard had been shaven and a swim cap was worn). This condition ameliorated the effect of trapped isothermal air present in beard and scalp hair. Condition 2: facial hair and swimcap (the beard was present and a swim cap was worn). This condition studied the effect of trapped isothermal air present in only the beard. Condition 3: facial hair and no swimcap (the beard was present and no swim cap was worn). This condition studied the effect of trapped isothermal air present in both beard and scalp hair. Condition 4: no facial hair and no swimcap (the beard was shaven and no swim cap was worn). This condition had the effect of trapped isothermal air from the beard removed, but trapped isothermal air in the scalp was still present. After completion of Condition 2 and Condition 3, facial hair was removed with a standard beard and moustache trimmer (CONAIR, East Windsor, NJ). Facial hair was collected and weighed to the nearest 0.001 g on an T120 electronic balance scale (Ohaus, Florham Park, NJ).

Instrumentation

Whole body density was evaluated with the BOD POD version 1.69 (Body Composition System; Life Measurement Instruments, Concord, CA). BOD POD procedures are described elsewhere (5). Percentage of body fat was determined by the Siri equation (6). Two repeat measures of percentage of body fat on the same day with the criterion method (beard had been shaven and a swim cap worn) in all 25 subjects had an intraclass correlation of r = 0.98 and a SE of the estimate of 0.7%.

Statistics

Regression analyses were performed comparing each hair state (facial hair and swimcap, facial hair and no swimcap, and no facial hair and no swimcap) with the criterion method. If the regression slope did not significantly deviate from one, and the intercept from zero, the percentage of body fat estimates under the two conditions were not considered different. Additionally, paired t tests were used to compare group means between the various hair states. To account for multiple paired t tests and an increase in Type I error, the Bonferroni statistical procedure was used. Furthermore, potential bias in percentage of body fat estimates by the BOD POD was examined using the Bland-Altman procedure (7). All data were analyzed using SPSS software (version 8.0; SPSS Inc., Chicago, IL). Results are given for percentage of body fat; the total body density (g/cm3) value used for calculation of percentage of body fat is also shown for reference.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

The regression of percentage of body fat by the criterion method against the percentage of body fat with facial hair and swimcap was used to assess the effect of facial hair on BOD POD estimates of percentage of body fat. This relationship significantly deviated from the line of identity (Figure 1, top). However, no significant bias between the two measures of percentage of body fat was observed (Figure 1, bottom). A paired t test revealed that the presence of facial hair only (facial hair and swimcap) yielded an estimate of percentage of body fat (16.2%, 1.0618 g/cm3) that was significantly lower than the criterion method (17.1%, 1.0597 g/cm3, p < 0.001). Additionally, the magnitude of the difference between percentage of body fat determined by the criterion method and that determined with facial hair and swimcap had no relationship with beard weight (Figure 2).

image

Figure 1. Top: Regression of percentage of fat by the criterion method against percentage of fat determined with facial hair and swimcap. The dotted line is the line of identity (regression slope = 1 and regression intercept = 0). The regression line significantly deviated from the line of identity. Bottom: Bland–Altman plot. The middle line represents the mean difference between percentage of fat determined with facial hair and swimcap and percentage of fat determined by the criterion method; the upper and lower dashed lines represent ±2SD from the mean. No bias was observed as indicated by the nonsignificant p value.

Download figure to PowerPoint

image

Figure 2. The regression of the difference between percentage of fat determined with facial hair and swimcap and the percentage of fat determined by the criterion method on the weight of facial hair (beard) in grams. No statistically significant association was observed (p = 0.31).

Download figure to PowerPoint

The next two analyses were performed to determine to effect of scalp hair on BOD POD estimates of percentage of body fat. The regression of percentage of body fat by the criterion method against percentage of body fat determined with facial hair and no swimcap significantly deviated from the line of identity (Figure 3, top); however, no bias was found across the range of fatness (Figure 3, bottom). Paired t test results showed a significant underestimation of percentage of body fat determined with facial hair and no swimcap (14.8%, 1.0649 g/cm3) vs. the criterion method (17.1%, 1.0597 g/cm3, p < 0.001). A similar observation was made between the criterion method vs. that with no facial hair and no swimcap. The regression of percentage of body fat by the criterion method and percentage of body fat determined with no facial hair and no swimcap significantly deviated from the line of identity (Figure 4, top). No significant bias across the range of fatness was observed (Figure 4, bottom). A paired t test showed a significant underestimation of percentage of body fat when determined with no facial hair and no swimcap (14.8%, 1.0650 g/cm3) vs. the criterion method (17.1%, 1.0597 g/cm3, p < 0.001).

image

Figure 3. Top: Regression of percentage of fat determined by the criterion method against percentage of fat determined with facial hair and no swimcap. The dotted line is the line of identity (regression slope = 1 and regression intercept = 0). The regression line significantly deviated from the line of identity. Bottom: Bland–Altman plot. The middle line represents the mean difference between percentage of fat determined with facial hair and no swimcap and percentage of fat determined by the criterion method; the upper and lower dashed lines represent ±2SD from the mean. No bias was observed as indicated by the nonsignificant p value.

Download figure to PowerPoint

image

Figure 4. Top: Regression of percentage of fat determined by the criterion method against percentage of fat determined with no facial hair and no swimcap. The dotted line is the line of identity (regression slope = 1 and regression intercept = 0). The regression line significantly deviated from the line of identity. Bottom: Bland–Altman plot. The middle line represents the mean difference between percentage of fat determined with no facial hair and no swimcap and percentage of fat determined by the criterion method; the upper and lower dashed lines represent ±2SD from the mean. No bias was observed as indicated by the nonsignificant p value.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

To our knowledge, this is the first study to demonstrate the effect of scalp and facial hair on BOD POD estimates of percentage of body fat. Underestimation of percentage of body fat due to the presence of facial hair was small (∼1%) but significant; a greater underestimation was observed with the presence of scalp hair (∼2.3%).

Underestimation of percentage of body fat due to scalp and facial hair may result from an underestimation in body volume because of the presence of isothermal air trapped in the hair fibers. Isothermal air is 40% more easily compressed than adiabatic air; therefore, if isothermal air is trapped in hair fibers a lower output signal for a given body volume is produced (4). As a result, body density is overestimated and percentage of fat is underestimated. These results suggest that facial hair may be a source of trapped isothermal air not accounted for in BOD POD program software.

A greater underestimation of percentage of fat was observed with scalp hair (∼2.3%) than with facial hair (∼1%), indicating a greater effect caused by the trapped isothermal air in scalp hair. The effect of scalp hair is ameliorated by a swim cap (as recommended by the manufacturer); however, the influence of facial hair is not addressed. No relationship between beard weight and the degree of underestimation of percentage of body fat was found. This suggests that hair quantity may not influence body volume measurements; instead, the amount of isothermal air trapped within the hair may be the source of error. This may be attributable to the different air-trapping capabilities of varying beard morphologies.

Taylor et al. (8), describing the early development of an air displacement plethysmography technique, reported poor agreement between air displacement plethysmography measures of volume compared with hydrostatic weighing volumes in rats. However, after fur was removed from the animals, agreement was considerably improved. It seems based on our findings, that a similar effect of scalp and facial hair on estimation of body volume occurs in humans. Moreover, preliminary results from our laboratory suggest that other sources of body hair (e.g., chest, legs, and arms) result in an additional underestimation of percentage of body fat up to ∼3% (D.A. Fields, P.B. Higgins, and G.R. Hunter, unpublished observation). Further research on the effect of total body hair on BOD POD measures of percentage of body fat is warranted.

In conclusion, body hair can cause an underestimation of BOD POD percentage of body fat. BOD POD program software currently recommends the use of a swim cap; this study further supports the use of a swim cap for all testing and indicates that the presence of facial hair should be considered a potential source of error in the determination of body composition by the BOD POD. Our results are of importance in longitudinal studies using the BOD POD for body composition measurement. It is recommended that men remove all facial hair for BOD POD testing or maintain body hair status when tested in longitudinal studies. If it is deemed necessary to correct for facial hair, the following correction equations are recommended: percentage of body fat = 1.44 + 0.97 × (BOD POD percentage of body fat) and total body density = 0.038 + 0.96 × (total body density from the BOD POD).

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

This work was supported by the Division of Physiology and Metabolism, Department of Nutrition Sciences, UAB, and the Clinical Nutrition Research Center (NIADK P30 DK56336).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  • 1
    Zumdahl, S. S. (1989) Chemistry. Gases. 2nd ed. D.C. Heath and Co: Lexington, MA. 176219.
  • 2
    Gundlach, B. L., Nijkrake, H. G. M., Hautvast, JGAJ (1980) A rapid and simplified plethysmometric method for measuring body volume. Hum Biol. 52: 2333.
  • 3
    Sly, P. D., Lanteri, C., Bates, J. H. T. (1990) Effect of the thermodynamics of an infant plethysmograph on the measurement of thoracic gas volume. Pediatr Pulmonol. 8: 203208.
  • 4
    Dempster, P., Aitkens, S. (1995) A new air displacement method for the determination of human body composition. Med Sci Sports Exerc. 27: 16921697.
  • 5
    Fields, D. A., Hunter, G. R., Goran, M. I. (2000) Validation of the BOD POD with hydrostatic weighing: influence of body clothing. Int J Obes Relat Metab Disord. 24: 200205.
  • 6
    Siri, W. E. (1961) Body composition from fluid spaces and density: analysis of methods. In Brozek, J. and Hencshel, A. (eds). Techniques for Measuring Body Composition. National Academy of Sciences/National Research Council: Washington, DC. 223224.
  • 7
    Bland, J. M., Altman, D. G. (1986) Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 8: 307310.
  • 8
    Taylor, A., Scopes, J. W., du Mont, G., Taylor, B. A. (1985) Development of an air displacement method for whole body volume measurement of infants. J Biomed Eng. 7: 917.