Is leg-to-leg BIA valid for predicting minimum weight in wrestlers?

Background: The National Collegiate Athletic Association (NCAA) and several state associations require prediction of minimum weight (MW) for collegiate and high school wrestlers. The rule requires assessment of body composition before the competitive season to minimize unhealthy weight-loss practices. Leg-to-leg bioelectrical impedance analysis (BIA) has been suggested for use with wrestlers. Purpose: To evaluate leg-to-leg BIA against a four-component (4C) criterion to determine whether leg-to-leg BIA predicted MW within acceptable limits for the sport of wrestling. Methods: Criterion MW was calculated by the 4C equation of Lohman (19) using independent measurement of body density (BD) by hydrostatic weighing, bone mineral content (BMC) by dual x-ray absorptiometry (DXA), and total body water (TBW) by deuterium dilution. Subjects were 57 wrestlers (mean SD; age = 19.7 +/- 1.3 yr, height = 176.6 +/- 7.3 cm, weight = 77.7 +/- 12.4 kg). Hydration was confirmed by the NCAA guidelines. Accuracy, precision, and systematic bias were examined. Results: Comparable mean values (72.2 +/- 9.7 vs 72.2 +/- 10.3 kg), a high correlation (r = 0.94), and a regression line similar to the line of identity were found between BIA and 4C. However, large individual differences and systematic bias were seen across the range of MW. BIA predicted MW within 3.5 kg 68% of the time and within 7.0 kg 95% of the time. MW residuals ranged from - 10.4 kg to +6.9 kg. When using 2.0 kg as an acceptable cutoff for error, only 40% of the BIA values were within 2.0 kg of the criterion. Conclusion: Large individual variation was seen, and, by definition, the precision was poor when estimating MW for individuals. In practical terms, the prediction error may span multiple weight classes, thus making leg-to-leg BIA unacceptable for prediction of MW in this sample under the conditions of the study.