CLARK, R. R., C. BARTOK, J. C. SULLIVAN, and D. A. SCHOELLER. Minimum Weight Prediction Methods Cross-Validated by the Four-Component Model. Med. Sci. Sports Exerc., Vol. 36, No. 4, pp. 639–647, 2004. The National Collegiate Athletic Association (NCAA) requires prediction of minimum weight (MW) for collegiate wrestlers. The rule was implemented to minimize unhealthy weight loss practices and requires assessment of body composition before the competitive season.
Purpose: This study cross-validated the body composition methods of dual energy x-ray absorptiometry (DXA), leg-to-leg bioelectrical impedance analysis (BIA), hydrostatic weighing (HW), and skinfolds (SF) for predicting MW using a four-component criterion (4C).
Methods: Criterion MW was calculated by the 4C model using independent measurement of body density (BD), bone mineral content (BMC), and total body water (TBW). Subjects were 53 Division I athletes from the University of Wisconsin (mean ± SD; age = 19.7 ± 1.3 yr, height = 176.2 ± 7.4 cm, weight = 75.6 ± 8.9 kg). Accuracy, precision, and systematic bias were examined in the predictions.
Results: There were no significant differences in mean MW from HW (70.5 ± 7.3 kg, P = 0.57), SF (70.5 ± 7.2 kg, P = 0.29) BIA (70.6 ± 7.6 kg, P = 0.39), DXA (70.3 ± 7.5, P = 0.97), and the 4C criterion (70.3 ± 7.4 kg). The regression for the relationships between 4C and HW (y = 0.994 × HW + 0.077 kg), 4C and SF (y = 1.003 × SF–0.437 kg), 4C and DXA (y = 0.942 × DXA + 4.034 kg), and 4C and BIA (y = 0.896 × BIA + 6.987 kg) did not significantly deviate from the line of identity. Pure error (PE) values ranged from 1.34 kg for HW to 3.08 kg for BIA.
Conclusion: Comparable means, high correlations, regression lines that did not significantly deviate from the line of identity, and no systematic bias were found. However, the methods differed widely in precision. The best precision, based on SEE and PE values, were seen in the HW and SF methods. In conclusion, this rigorous four-component cross-validation study supports the NCAA methods as the most accurate and precise MW prediction methods in this sample.