Share this article on:

The Effect of Dehydration on Wrestling Minimum Weight Assessment


Medicine & Science in Sports & Exercise: January 2004 - Volume 36 - Issue 1 - p 160-167
doi: 10.1249/01.MSS.0000106855.47276.CD
APPLIED SCIENCES: Physical Fitness and Performance

BARTOK, C., D. A. SCHOELLER, R. R. CLARK, J. C. SULLIVAN, and G. L. LANDRY. The Effect of Dehydration on Wrestling Minimum Weight Assessment. Med. Sci. Sports Exerc., Vol. 36, No. 1, pp. 160–167, 2004. Given that some wrestlers arrive for minimum weight (MW) testing in a dehydrated condition, it is important to understand the effects of dehydration on MW assessment methods.

Purpose To determine the effect of dehydration on the assessment of MW by three-site skinfolds with the Lohman formula (SF), leg-to-leg bioelectrical impedance analysis (BIA), and multifrequency bioelectrical impedance spectroscopy (BIS) compared with a four-component (4C) criterion.

Methods Twenty-two male collegiate wrestlers (mean ± SD, age: 19.9 ± 1.4 yr, height: 174.0 ± 6.8 cm, body mass: 77.4 ± 9.1 kg) had their body composition assessed by the 4C criterion, hydrostatic weighing (HW), SF, BIA, and BIS in euhydration (EUH) and dehydration (DEH). Subjects dehydrated 2–5% of body weight through fluid restriction and exercise in a hot environment.

Results In EUH, the total error (TE) for HW (1.75 kg) and SF (2.15 kg) were not significantly different, but the TE for HW and SF methods were significantly lower than the TE for both BIS (3.68 kg) and BIA (3.77 kg). In DEH, SF, BIA, and BIS methods had a TE approaching or exceeding 4 kg (8.8 lb). Dehydration increased the TE for SF and BIA through an artificial lowering of body weight and for BIS by an increased error in intracellular water prediction.

Conclusion Acute thermal dehydration violates assumptions necessary for the accurate and precise prediction of MW by SF, leg-to-leg BIA, and multifrequency BIS.

1Department of Nutritional Sciences, University of Wisconsin, Madison, WI; and

2U. W. Health Sports Medicine Center, Madison, WI

Address for correspondence: Cynthia Bartok, 2 Hidden Ledge Dr., Conway, MA 01341; E-mail:

Submitted for publication November 2002.

Accepted for publication September 2003.

Over 250,000 high school and college athletes participate in the sport of wrestling each year (23,26). To maintain fairness and prevent injuries, wrestlers are paired with opponents of similar weight through the use of weight classes. In order to “make weight” for competitions, some wrestlers dehydrate and restrict food intake, placing themselves at increased risk for health problems (15). To control these practices, several states have implemented minimum weight (MW) programs to regulate high school wrestling (8), and the National Federation of State High School Associations has recommended that all states have MW programs in place by 2004 (25). In addition, the National Collegiate Athletic Association implemented a MW program regulating collegiate wrestling in 1998 (24). These MW programs assign wrestlers a minimum competition weight based on body composition testing (8).

The NCAA Competitive Safeguards Committee has approved hydrostatic weighing (HW), skinfolds (SF) using the Lohman equation (18), and air displacement plethysmography (36) for collegiate MW assessment (24). At the high school level, MW assessment is made using HW, SF, or in some states, bioelectrical impedance analysis (BIA) (25). Several studies have established the validity of SF testing in high school and collegiate wrestlers (6,9,10,35) and have demonstrated potential for the use of whole-body (7,29) and leg-to-leg (37) BIA techniques. However, concerns have arisen about the effects of dehydration (DEH) on MW assessment by these techniques (7,9). DEH reduces body weight and total body water, alters the distribution and compartmentalization of body water, lowers the hydration of fat-free mass, and alters the density of fat-free mass (11,28).

This is a significant area of concern because recent research has demonstrated that some wrestlers arrive for MW testing in a dehydrated state to secure a lower predicted MW. Unannounced reweighing of wrestlers within days of MW testing showed that almost 25% had gained more than 1.4 kg (3 lb), with a maximum weight gain of over 8 kg (18 lb) (8). In response to this problem, the NCAA and some high school programs have added hydration testing to the MW testing protocol. Wrestlers must produce a urine sample with a specific gravity ≤ 1.020 before assessment of MW to demonstrate adequate hydration (24,25). However, concerns remain about the sensitivity and specificity of urine specific-gravity testing as well as the ability to mask signs of DEH with acute ingestion of hypotonic fluids (28).

In theory, DEH violates assumptions critical to the estimation of MW by HW, SF, and impedance methods. However, to date, there are no published studies specifically aimed at understanding the effects of DEH on MW assessment. The goals of this study were to assess the accuracy and precision of MW assessment techniques under strictly controlled conditions of adequate hydration (euhydration, EUH) and DEH as well as to determine the effects of DEH on these MW assessment techniques. The test techniques included SF, leg-to-leg BIA, and proximal tetrapolar multifrequency bioelectrical impedance spectroscopy analysis (BIS). We hypothesized that all techniques would have unacceptable total errors (>4 kg) in DEH due to violations of body composition assumptions critical to these techniques.

Back to Top | Article Outline



Twenty-five healthy male volunteers were recruited for the protocol. Subjects were young (18–23 yr), current collegiate wrestlers, and represented all major weight classes of wrestling. The protocol was approved by the Human Subjects Committee of the University of Wisconsin and was in conformance with the human experimentation policy statement of the American College of Sports Medicine. Before participation, all subjects provided written informed consent.

Subjects reported to the General Clinical Research Center of the University of Wisconsin Hospital and Clinics at 1700 h of day 1. Subjects were in season and had participated in practice the afternoon of the day of admission. Subjects received a defined diet (evening meal) and 30 mL·kg−1 bottled water to ensure adequate hydration. Subjects remained under observation at the hospital overnight to minimize the potential for DEH.

During the morning of day 2, subjects underwent a battery of body composition tests in EUH. At 0800 h, baseline blood samples were collected for deuterium and bromide analysis. Subjects voided (void discarded) and then a baseline weight was measured in minimal dry clothing. At 0820 h, subjects were dosed with 10 mg·kg−1 NaBr and 40 g of 10% deuterium solution for the determination of ECW and TBW. From this point on and until the conclusion of the study, all urine output was collected and the volume measured for analysis of lost bromide and deuterium. Then subjects were transported to the University of Wisconsin Sports Medicine Exercise Science Laboratory. Between 0900 and 1030 h, subjects underwent hydrostatic weighing (HW), lung residual volume estimation, bone mineral assessment by dual x-ray absorptiometry (DXA), and skinfold (SF) measurement. These measurements were made using standard laboratory procedures described previously (10). Subjects were then transported back to the Research Center. At 1030 h, subjects voided to empty their bladders so that a spot urine sample could be collected 1 h later. From 1130 to 1200 h, EUH phase measurements were made. Blood was drawn for deuterium/bromide analysis as well as hydration testing. Subjects emptied their bladders to provide a spot urine sample for hydration testing. BIA and BIS measurements were made and weight was measured in the same minimal dry clothing as in earlier measurements. At noon, subjects received two to four Boost Bars (Mead Johnson Nutritionals, Evansville, IN) and 120 mL of water. Then, they were moved to a hot room (90°F, 25% relative humidity) and exercised until they reached a level of DEH between 2 and 5% of body weight. Staff nurses monitored weight loss (in minimal dry clothing), vital signs, and exercise time during the DEH intervention. After subjects completed this intervention, they were returned to a room of normal temperature (20°C) to shower, rest, and cool down. No beverages or food were allowed. Exactly 1 h after being moved from the hot room, blood was drawn for deuterium/bromide analysis and hydration testing as well as a spot urine sample collected for hydration testing. Then SF, BIA, and BIS measurements were made. Final, dehydrated weight was measured in the same minimal dry clothing as all previous measurements. At most, three subjects were scheduled on the same day, and two investigators collected data. One investigator (CB) performed all DXA, BIS, and BIA scans, whereas the other (RC or JS) performed all residual volume, underwater weighing, and SF measurements. The same investigator performed the pre- and postdehydration SF measurements in a given individual. The two SF investigators co-train on SF measurements regularly and have excellent reliability (r = 0.98–1.0) for SF measurements at various sites.

Back to Top | Article Outline

Body water assessment.

TBW was quantified using the deuterium dilution, isotope ratio mass spectrometry technique, and equations of Schoeller (32). Corrections were made for lost dosage of deuterium in urine, sweat, and breath vapor, as well as nonaqueous hydrogen exchange (21,33,39). ECW was quantified using the anion exchange high-pressure chromatography technique and equations of Miller et al. (19,20), except that we added a correction of 0.987 for the concentration of water in serum ultrafiltrate (2). Corrections were made for lost dosage of bromide in urine and sweat. Sweat samples were collected on five subjects using Tegaderm patch (3M Health Care, St. Paul, MN) covered gauze on three sites (arm, back, and chest). Undiluted sweat samples were analyzed for bromide using the same anion exchange high-pressure chromatography technique as for blood. Urinary bromide was isolated using ion exchange chromatography methods (2). ICW was calculated as the difference between TBW and ECW.

Back to Top | Article Outline

Body composition assessment.

Upon admission, height was measured to the nearest mm using a stadiometer. At each weight measurement point, weight was measured to the nearest 0.05 kg using a digital platform scale. Subjects were measured without shoes and in the same minimal, dry clothing at each point. Minimal dry clothing was either dry gym shorts with underwear or dry swim trunks. Percent DEH by weight (% DEH WT) was calculated as:EQUATION

where WTEUHend = weight (kg) at the end of the EUH phase, WT DEHend = weight (kg) at end of DEH phase, solidsIN = intake (kg) in solids from nutritional bars, and solidsOUT = solids lost via respiration estimated as 1 g·min−1 exercise (21). For all body composition methods, MW was calculated as body weight at 5% body fat using the equation MW = fat-free mass (FFM)/0.95 (24). For four-component (4C), HW, and SF techniques, fat mass (FM) was calculated as (% body fat/100)*weight, and FFM was calculated as weight minus FM.

The criterion 4C body composition assessment took place at the UW Sports Medicine Exercise Science Laboratory. A Norland XR-36 dual energy x-ray absorptiometry (DXA) machine (Norland, Fort Atkinson, WI) with software pack 3.7.4/2.1.0 was used to measure total body bone mineral content. Body density was determined by HW at residual volume, as described by Behnke and Wilmore (4) using a custom built tank (Louis Hoffman Company, Milwaukee, WI) and a Chatillon cadaver scale (John Chatillon and Sons, Kew Gardens, NY). Ten underwater weights were measured, and the average of the three heaviest was used to calculate body density. Residual volume was measured by the oxygen dilution method described by Wilmore (38) using a modified Collins 13.5-L respirometer (Braintree, MA) and a Med Science model 505 N2 analyzer (St. Louis, MO). The subject’s residual volume was measured outside the tank in a seated position simulating that used during HW. The mean of two trials within 75 mL was used for calculations. An additional correction of 100 mL, to account for gastrointestinal gas, was used in the HW calculation (4). 4C percent body fat was calculated using the modified Selinger formula (3):EQUATION

where Db is equal to the body density, W is the TBW expressed as a fraction of body weight, and M is body mineral expressed as a fraction of body weight. M was calculated from total body bone mineral by DXA assuming a constant relationship between bone mineral and nonosseous mineral in the body (body mineral = bone mineral/0.824) (5). Hydrostatic weighing (HW) percent fat was calculated from body density using Brozek’s formula (5).

Estimation of body density by SF was made using three skinfold sites and the Lohman equation (18). Skinfold thickness at triceps, subscapular, and abdominal sites was measured using a Lange ski-fold caliper. Each subject’s measurements were taken in duplicate, by the same investigator, on the right side of the body. If the two measures differed by more than 0.5 mm, a third measurement was made and the results averaged. The Brozek equation was used to estimate percent body fat from the predicted body density (5).

A Tanita TBF-300GS leg-to-leg bioelectrical impedance analyzer (Tanita Corporation of America, Arlington Heights, IL) was used for all BIA measurements. Subjects stood motionless without socks on the measurement platform while a low amperage alternating current (500 μA, 50 kHz) was passed through the lower half of the body. Weight, FM, FFM, TBW, and percent body fat were given as output using the “athlete” mode.

A Xitron Hydra 4200 BIS analyzer with Cole-Cole modeling software (Xitron Technologies, San Diego, CA) was used for all BIS measurements. Four electrodes were placed in the proximal (elbow/knee) position for the measurement of resistance and reactance (31). The output variables of the Cole-Cole analysis included resistance of extracellular fluid (Re) and resistance of the intracellular fluid (Ri) in ohm. TBW, ECW, and ICW were calculated using previously published proximal technique equations for the general adult population (13): ICW (kg) = 0.191(Ht2/Ri) + 12.8 TBW = ECW + ICW, EQUATION

where Ht = height in cm. FFM was calculated using the equation FFM = TBW/0.73 (32).

Back to Top | Article Outline

Other measures.

An 8 × 12 × 2 cm plasma conductivity cell was constructed of clear plastic and four aluminum alloy transecting conducting plates. Plasma samples were placed in a 0.7-cm-diameter well in the center of the conductivity cell. Current injection and detection clips were attached to the conducting plates. Resistance was measured at 5, 50, 500, and 1000 kHz. An infrared thermometer (Minitemp FS, Raytek, Santa Cruz, CA) was used to monitor skin temperature at the injection and detection electrode sites used in BIS measurements. EUH and DEH measurements were made immediately before electrodes were placed and BIS measurements were made. Hydration tests were completed by the hospital laboratory shortly after collection of blood and urine samples. An IRIS 900 UDEX automated urinalysis system (International Remote Imaging Systems, Chatsworth, CA) was used to test for urine specific gravity by the harmonic oscillation densitometry method. An Advanced Micro Osmometer (Advanced Instruments Inc., Norwood, MA) freezing point osmometer was used to measure the osmolality of serum samples.

Back to Top | Article Outline


Power analysis suggested 21 subjects were needed to detect differences in MW between impedance methods and the criterion. Assuming 10% subject attrition and/or sample loss, a sample size of 25 was set. Data are presented as mean ± standard deviation unless otherwise noted. BIA, BIS, HW, and SF were evaluated against the 4C criterion method using the methods of Lohman (18), Guo and Chumlea (14), and Bland and Altman (1). Statistical evaluation of MW estimates included: 1) calculation of the mean difference, standard error of estimate, total (pure) error, simple bivariate correlation; 2) comparison of means using a 1 × 8 repeated measures ANOVA design with eight factors being 4C, HWeuh, SFeuh, BIAeuh, BISeuh, SFdeh, BIAdeh, and BISdeh; and 3) assessment of potential bias using Bland-Altman analysis. Post hoc analysis consisted of paired two-tailed t-tests because we had a limited number of desired contrasts (all 7 factors compared with 4C only). Significant differences in total error values were tested for using the F-max (variance comparison) test. For this test the ratio of highest TE2 to lowest TE2 was calculated and the ratio’s significance was determined in reference to the F-distribution. For all analyses, a P value less than 0.05 was considered significant.

Back to Top | Article Outline


All 25 subjects completed the protocol activities without difficulty. Two wrestlers who competed in the heavyweight category were removed from the sample because their skinfold measurements were statistical outliers. One additional subject was removed from the sample due to missing EUH BIS data (instrument malfunction). This produced a pool size of 22. The physical characteristics of this subject pool, including % body fat and MW, are presented in Table 1. Due to scheduling constraints, SF data in DEH are available for 20 subjects. Comparisons between EUH and DEH SF are made on this subset of 20 subjects.



During the DEH phase of the study, subjects acutely lost between 1.9 and 5.2% of body weight (average = 3.2%, 2.39 kg) using thermal DEH, exercise, and fluid restriction. According to dilution techniques, DEH resulted in a average loss of 2.18 ± 0.81 kg TBW, 1.21 ± 1.1 kg ECW, and 0.97 ± 1.2 kg ICW (all P < 0.001). This corresponds to an average 4.5% decrease in TBW, 7.1% decrease in ECW, and 3.1% decrease in ICW. Plasma osmolality rose from 288 ± 3.9 mOsm·L−1 in EUH to 298 ± 5.2 mOsm·L−1 in DEH (P < 0.001). Likewise, urine specific gravity rose from 1.015 ± 0.0042 in EUH to 1.028 ± 0.0044 in DEH (P < 0.001).

MW data for subjects in EUH and DEH are presented in Table 2 and Table 3. The 1 × 8 repeated measures ANOVA revealed significant differences between mean MW values for the 20 subjects with complete data (F = 6.49, P = 0.002). Paired t-tests were used to determine whether significant differences existed between 4C and HWeuh, SFeuh, BIAeuh, BISeuh, BIAdeh, and BISdeh using the complete data set (N = 22, shown in Table 2). Paired t-tests for the subset of subjects (N = 19) having SF measures in EUH and DEH are shown separately in Table 3. As seen in Table 2, mean MW predicted by HW (P = 0.17), SF (P = 0.40), and BIA (P = 0.11) were not significantly different from the 4C criterion in EUH. The BIS-predicted MW was significantly (P = 0.006) greater than the 4C criterion. In EUH, the TE for SF (2.15 kg) was not significantly different from the TE for HW (1.75 kg, P = 0.17). However, the TE for BIS (3.68 kg) and BIA (3.77 kg) were significantly greater than the TE for both HW (P < 0.001 for both) and SF (P < 0.01 for both) methods. Bland-Altman tests showed no significant bias for any method versus 4C (R2 = 0–0.34).





In DEH, BIS-predicted MW was significantly (P = 0.03) higher than 4C (P = 0.03), and BIA-predicted MW was significantly lower than 4C (P = 0.04). The total error for BIS and BIA increased to 4.89 and 4.64 kg, respectively. Table 3 shows the effect of DEH on MW assessment by SF for a subset of 20 subjects. DEH reduced the predicted MW by approximately 2 kg over the EUH prediction (P < 0.001), and the total error grew from 2.40 kg to 3.81 kg.

The effects of DEH on MW assessment are described in detail in Tables 4–6. As Table 4 shows, DEH results in a decrease in the sum of three SF by approximately 1 mm (P < 0.05), which significantly raised the predicted body density and lowered percent body fat. In this sample, an increase in SF of 1 mm would increase predicted MW by 0.19–0.38 kg (0.43–0.84 lb). However, when the significant decrease in body weight with DEH was factored in, FFM and MW were significantly (P < 0.001) reduced as compared with EUH.







Table 5 shows the effects of DEH on MW assessment by leg-to-leg BIA. DEH did not result in a significant change in impedance; however, it resulted in a significant decrease in the measured body weight (P < 0.001). Although the predicted TBW, percent fat, and FM significantly decreased, FFM and MW were not statistically different between EUH and DEH.

The effects of DEH on BIS-predicted MW are shown in Table 6. The resistance of the ECW increased with DEH (P < 0.001), which resulted in a decrease in predicted ECW (P < 0.01). In contrast, the resistance of the ICW decreased (P < 0.001) in DEH, which lead to an increase in predicted ICW (P < 0.01). Thus, there was no significant net change in TBW with DEH (P = 0.76) and the true TBW was overestimated. Therefore, predicted FFM and MW were close to that estimated during EUH.

Figure 1 shows the average ECW, ICW, and TBW predicted by BIS as well as the average TBW predicted by BIA in EUH and DEH. At the EUH time point, BIS-predicted ECW and ICW were not significantly different from the dilution-measured compartments. However, TBW was significantly overestimated by BIS. In DEH, all body water compartments were significantly overestimated by BIS as compared with dilution. The TBW predicted by BIA was not significantly different from dilution at EUH and approached significance (P = 0.054) at DEH.



Figure 2 depicts the average within-subject change in water volumes from EUH to DEH by the impedance methods. The average predicted change in ECW was not significantly different between BIS and dilution methods. However, because BIS was insensitive to the acute loss in ICW with DEH, the average BIS-predicted change in TBW was significantly less than the true changes as measured by dilution or weight change. Similarly, BIA significantly underestimated the loss in TBW as compared with dilution and weight change data.



Plasma conductivity was monitored at 5, 50, 500, and 1000 kHz in a subset of subjects (N = 20). The average resistance of the plasma at these frequencies was unchanged from EUH (607 ± 16 Ω) to DEH (604 ± 22 Ω). To test for differences in skin temperature in EUH and DEH, the temperatures at the four electrode sites was monitored in a subset of subjects (N = 10). The average skin temperature increased 2.2 ± 3.5°C from EUH to DEH (P < 0.05).

Back to Top | Article Outline


This study determined how acute thermal dehydration (DEH) affects body composition methods by analyzing differences in measured and predicted MW variables using subjects during euhydration (EUH) as their own controls for DEH. Test methods included hydrostatic weighing (HW), skinfolds (SF), leg-to-leg bioelectrical impedance analysis (BIA) calibrated for athletes, and multifrequency bioelectrical impedance spectroscopy (BIS). These test methods were compared to a four-component (4C) criterion.

Back to Top | Article Outline

MW estimation in euhydration.

In EUH, MW was most accurately and precisely predicted by HW. It is not a surprise that HW had the highest correlation, lowest mean difference, lowest standard error of estimate, and lowest TE or of any technique. The 4C method utilizes the body density value from HW, but then includes an actual measure of TBW and body mineral rather than assuming reference values for these components (3). Thus, the HW and 4C methods are highly related because they share a certain portion of methodology.

As seen in other studies (6,9,10,35), SF also performed well in relation to the criterion method in EUH. SF had a TE = 2.15 kg, which is comparable with previous studies (6,7). This TE was significantly lower than the TE of both BIA (TE = 3.77 kg) and BIS (TE = 3.68 kg) in EUH. It is important to note that the prediction of MW by BIS would likely be improved with the development and use of wrestling-specific equations (34).

Back to Top | Article Outline

MW estimation in dehydration.

This study clearly shows that DEH has a profound impact on the total error in MW estimation. The TE of methods increased 23% for leg-to-leg BIA, 33% for multifrequency BIS, and 59% for SF. TE for SF approached 4 kg (3.8 kg), whereas BIA and BIS methods had a TE exceeding 4 kg. A TE of 4 kg means that predicted MW is within 4.0 kg (8.8 lb) 68% of the time and within 8.0 kg (17.6 lb) 95% of the time. These error levels would be unacceptable for MW estimation because they span several weight classes. DEH affected the estimation of MW through various violations of assumptions in body composition assessment. For SF assessment, DEH had opposing effects on the measured variables. The small but significant decrease in skinfold thickness results in a higher body density and lower percent body fat. This would ultimately result in a higher calculated MW if body weight does not change. However, when the effects of DEH-induced weight loss are factored in, the result is a 2.2 kg lower MW. It is interesting to note that the weight loss of 3.2 kg produces a net effect of a 2.2 kg lower MW. Thus, on average, 69% of the weight lost in DEH is propagated into a reduction in MW. For this method to accurately measure MW in DEH, a euhydrated weight must be used in the prediction equations.

DEH also had a significant effect on body composition determination by the leg-to-leg bioimpedance system. It is interesting to note that despite a statistically significant loss of 0.8 kg of BIA-predicted TBW (2% body water), the measured bioimpedance did not change (0.5% change, NS). The only measured variable that changed significantly was body weight (3-kg loss). In addition, the loss in body weight in DEH was interpreted as an average 2% loss in body fatness, 1.8-kg loss in FM and a 0.7-kg loss in FFM. Several previous studies also have documented decreases in percent fat with DEH in the presence or absence of changes in resistance (16,30). This raises questions about how these instruments predict body composition variables such as % fat, FFM, and TBW from impedance. Our data and the data from another study (30) suggest that body weight is likely a large factor in the prediction of body composition and body composition changes by some impedance instruments. It is likely that these instruments incorporate body weight into the regression analysis equations that predict body composition from resistance values.

Thermal DEH greatly affected the accuracy and precision of BIS-predicted MW. The resistance of the ECW significantly increased over EUH, which resulted in a predicted loss of 0.9 kg ECW. This change was similar to the dilution-predicted loss in ECW. However, the resistance of the intracellular pathway significantly decreased, which significantly raised ICW by over 1 kg, despite a dilution-measured loss in ICW. The net effect was no change in BIS-predicted TBW levels with DEH. These findings are in contrast to the results of O’Brien et al. (27), who observed that BIS-predicted TBW changes were more accurate under conditions of hypertonic DEH than isotonic DEH. They hypothesized that if the impedance equations were changed to reflect the change in body fluid resistivity in hypertonic DEH, the prediction of TBW could be further improved in DEH. To investigate this possibility, we measured the conductivity of subjects’ plasma in both EUH and DEH. No significant difference was observed between the resistivity of the blood in EUH and DEH, despite changes in serum osmolality and natremia.

Our data suggest BIS lacks accuracy and precision in ICW estimation with hypertonic DEH and that this is the primary problem with BIS’s prediction of TBW changes in DEH. The increase in predicted ICW with DEH was noted in 15 of the 22 subjects. Of these 15 subjects, 10 had predicted ICW increases under 2 kg and five had increases greater than 2 kg, with a maximum increase of 5.1 kg. Inspection of the raw data showed that the Ri values were highly variable in DEH and that the elevations in predicted ICW were not due to simple outliers in the data. It may be that the resistivity of the intracellular water changed and that this disruption in body chemistry was not propagated into the ECW compartment. Without the ability to measure ICW resistivity directly, we cannot confirm this hypothesis. In an attempt to minimize potential artifacts produced by changes in skin temperature with exercise, we placed subjects in a cool room, had them take a cool shower, and rest for 1 h before making measurements in the DEH state. This limited the increase in skin temperature to 2.2°C. This effect of this temperature increase would be minimal (12) and would not explain the lack of accuracy and precision in Ri values. It also is possible that physical exercise produced a redistribution of body fluids to the exercising limbs, which could increase the hydration of the limbs and conceal the DEH in the trunk (22). This hypothesis has been used to explain a lack of accuracy in postexercise impedance measurements even when skin temperature effects was controlled (17).

Back to Top | Article Outline

Conclusions and implications.

This study evaluated MW predictions by leg-to-leg bioelectrical impedance analysis, multifrequency bioelectrical impedance spectroscopy, and skinfold measurement techniques as compared with a four-component criterion method in both a euhydrated and dehydrated state in a group of 25 collegiate wrestlers. In the euhydrated state, the total error for hydrostatic weighing and skinfold measurements were not significantly different, whereas BIA and BIS had total errors that were significantly greater than both hydrostatic weighing and skinfold measurements. In the dehydrated state, all techniques had a total error approaching or exceeding 4 kg, which could span several weight classes. Detailed analysis of MW predictions showed that dehydration affected the skinfold technique through a reduction in body weight, BIA through a reduction in weight, and BIS through a disruption in intracellular water predictions. It is clear that dehydration poses a significant limitation to body composition assessment for all the methods tested in this study.

The authors would like to thank University of Wisconsin wrestling coach Barry Davis, athletic trainer Jason Williams, and the volunteer subjects who participated in the study.

This work was supported by grant M01 RR03186 from the General Clinical Research Centers Program of the National Center for Research Resources, National Institutes of Health; grants from the Gatorade Sports Science Institute, the National Collegiate Athletic Association, and the Sports Medicine Fund of the University of Wisconsin Sports Medicine Center; and by a training grant from the National Institutes of Health.

Back to Top | Article Outline


1. Altman, D. G., and J. M. Bland. Measurement in medicine: the analysis of method comparison studies. Statistician 32: 307–317, 1983.
2. Bartok, C., R. L. Atkinson, and D. A. Schoeller. Measurement of nutritional status in simulated microgravity by bioelectrical impedance spectroscopy. J. Appl. Physiol. 95: 225–232, 2003.
3. Bartok-Olson, C. J., D. A. Schoeller, J. C. Sullivan, and R. R. Clark. The “B” in the Selinger four-compartment body composition formula should be body mineral instead of bone mineral. Ann. N. Y. Acad. Sci. 904: 342–344, 2000.
4. Behnke, A. R., and J. H. Wilmore. Evaluation and Regulation of Body Build and Composition. Englewood Cliffs, NJ: Prentice-Hall, Inc., 1974, pp. 39–50.
5. Brozek, J., R. Grande, J. T. Anderson, and A. Keys. Densitometric analysis of body composition: revision of some quantitative assumptions. Ann. N. Y. Acad. Sci. 110: 113–140, 1963.
6. Clark, R. R., J. M. Kuta, and R. A. Oppliger. The Wisconsin Wrestling Minimal Weight Project: cross-validation of prediction equations. Pediatr. Exerc. Sci. 4: 117–127, 1992.
7. Clark, R. R., J. M. Kuta, J. C. Sullivan, W. M. Bedford, J. D. Penner, and E. A. Studesville. A comparison of methods to predict minimal weight in high school wrestlers. Med. Sci. Sports Exerc. 25: 151–158, 1993.
8. Clark, R. R., and R. A. Oppliger. Minimal weight standards in high school wrestling: the Wisconsin model. Orthop. Phys. Ther. Clin. North Am. 7: 23–45, 1998.
9. Clark, R. R., R. A. Oppliger, and J. C. Sullivan. Cross-validation of the NCAA method to predict body fat for minimum weight in collegiate wrestlers. Clin. J. Sport Med. 12: 285–290, 2002.
10. Clark, R. R., J. C. Sullivan, C. Bartok, and D. A. Schoeller. Multi-component cross-validation of minimum weight predictions for college wrestlers. Med. Sci. Sports Exerc. 35: 342–347, 2003.
11. Going, S. B. Densitometry. In. Human Body Composition, A. F. Roche, S. B. Heymsfield, and T. G. Lohman (Eds.). Champaign, IL: Human Kinetics, 1996, pp. 3–24.
12. Gudivaka, R., D. A. Schoeller, and R. F. Kushner. Effect of skin temperature on multifrequency bioelectrical impedance analysis. J. Appl. Physiol. 81: 838–845, 1996.
13. Gudivaka, R., D. A. Schoeller, R. F. Kushner, and M. J. G. Bolt. Single- and multifrequency models for bioelectrical impedance analysis of body water compartments. J. Appl. Physiol. 87: 1087–1096, 1999.
14. Guo, S. S., and W. C. Chumlea. Statistical methods for the development and testing of prediction equations. In: Human body composition, A. F. Roche, S. B. Heymsfield, and T. G. Lohman (Eds.). Champaign, IL: Human Kinetics, 1996, pp. 191–202.
15. Harms, R. Wrestling the scale not the opponent. W. M. J.September: 37–38, 1998.
16. Khaled, M. A., M. J. Mccutcheon, S. Reddy, P. L. Pearman, G. R. Hunter, and R. L. Weinsier. Electrical impedance in assessing human body composition: the BIA method. Am. J. Clin. Nutr. 47: 789–792, 1988.
17. Koulmann, N., C. Jimenez, D. Regal, et al. Use of bioelectrical impedance analysis to estimate body fluid compartments after acute variations of the body hydration level. Med. Sci. Sports Exerc. 32: 857–864, 2000.
18. Lohman, T. G. Skinfolds and body density and their relation to body fatness: a review. Hum. Biol. 53: 181–225, 1981.
19. Miller, M. E., and C. J. Cappon. Anion-exchange chromatographic determination of bromide in serum. Clin. Chem. 30: 781–783, 1984.
20. Miller, M. E., J. A. Cosgriff, and G. B. Gorbes. Bromide space determination using anion-exchange chromatography for measurement of bromide. Am. J. Clin. Nutr. 50: 168–171, 1989.
21. Mitchell, J. W., E. R. Nadel, and J. A. Stolwijk. Respiratory weight losses during exercise. J. Appl. Physiol. 32: 474–476, 1972.
22. Monnier, J. F., E. Raynaud, J. F. Brun, and A. Orsetti. Influence of meal consumption and exercise on the assessment of body composition by bioelectrical impedance. Sci. Sports 12: 256–258, 1997.
23. National Collegiate Athletic Association. 1982–2001 NCAA sports sponsorship and participation report. Indianapolis, IN: National Collegiate Athletic Association, 2002, pp. 145.
24. National Collegiate Athletic Association. NCAA Wrestling Rules and Interpretations. Indianapolis, IN: National Collegiate Athletic Association, 2003, pp. WR23–WR34.
25. National Federation of State High School Associations. Wrestling weight management program. Indianapolis, IN: National Federation of State High School Associations, 2001, pp. 25–34.
26. National Federation of State High School Associations. 2002 High school participation survey. Indianapolis, IN: National Federation of State High School Associations, 2002, pp. 191.
27. O’Brien, C., C. J. Baker-Fulco, A. J. Young, and M. J. Sawka. Bioimpedance assessment of hypohydration. Med. Sci. Sports Exerc. 31: 1466–1471, 1999.
28. Oppliger, R. A., and C. Bartok. Current opinion: hydration testing. Sports Med. 32: 959–971, 2002.
29. Oppliger, R. A., D. H. Nielsen, and C. G. Vance. Wrestlers’ minimal weight: anthropometry, bioimpedance, and hydrostatic weighing compared. Med. Sci. Sports Exerc. 23: 247–253, 1991.
30. Saunders, M. J., J. E. Blevins, and C. E. Broeder. Effects of hydration changes on bioelectrical impedance in endurance trained individuals. Med. Sci. Sports Exerc. 30: 885–892, 1998.
31. Scheltinga, M. R., D. O. Jacobs, T. D. Kimbrough, and D. W. Wilmore. Alterations in body fluid content can be detected by bioelectrical impedance analysis. J. Surg. Res. 50: 461–468, 1991.
32. Schoeller, D. A. Hydrometry. In: Human Body Composition, A. F. Roche, S. B. Heymsfield, and T. G. Lohman (Eds.). Champaign, IL: Human Kinetics, 1996, pp. 25–44.
33. Schoeller, D. A., R. F. Kushner, P. Taylor, W. H. Dietz, and L. Bandini. Measurement of total body water: isotope dilution techniques. In: Ross Conference on Medical Research, 1985. Columbus, OH: Ross Laboratories.
34. Segal, K. R. Use of bioelectrical impedance analysis measurements as an evaluation for participating in sports. Am. J. Clin. Nutr. 64( Suppl.): 469s–471s, 1996.
35. Thorland, W. G., G. O. Johnson, C. J. Cisar, and T. J. Housh. Estimation of minimal wrestling weight using measures of body build and body composition. Int. J. Sports Med. 8: 365–370, 1987.
36. Utter, A. C., F. L. Goss, P. D. Swan, G. S. Harris, R. J. Robertson, and G. A. Trone. Evaluation of air displacement for assessing body composition of collegiate wrestlers. Med. Sci. Sports Exerc. 35: 500–505, 2003.
37. Utter, A. C., J. R. Scott, R. A. Oppliger, et al. A comparison of leg-to-leg bioelectrical impedance and skinfolds in assessing body fat in collegiate wrestlers. J. Strength Cond. Res. 15: 157–160, 2001.
38. Wilmore, J. H. A simplified method for determination of residual lung volumes. J. Appl. Physiol. 27: 98–100, 1969.
39. Wong, W. W., W. J. Cochran, W. J. Klish, E. O. Smith, L. S. Lee, and P. D. Klein. In vivo isotope-fractionation factors and the measurement of deuterium- and oxygen-18-dilution spaces from plasma, urine, saliva, respiratory water vapor, and carbon dioxide. Am. J. Clin. Nutr. 47: 1–6, 1988.


©2004The American College of Sports Medicine