2007•The Journal of Strength and Conditioning ResearchOpen access
Reliability and Detecting Change Following Short-Term Creatine Supplementation: Comparison of Two-Component Body Composition Methods
Liam P. Kilduff, Sarah J. Lewis, Mike I.C. Kingsley, Nick Owen, Rebecca E. Dietzig
Open full text 28 citations
Abstract
The purpose of the present study was twofold: firstly, to assess the reliability of various body composition methods, and secondly, to determine the ability of the methods to estimate changes in fat-free mass (FFM) following creatine (Cr) supplementation. Fifty-five healthy male athletes (weight 78.3 +/- 10.3 kg, age 21 +/- 1 years) gave informed consent to participate in this study. Subjects' FFM was estimated by hydrostatic weighing (HW), air-displacement plethysmography (ADP), bioelectrical impedance analysis (BIA), near-infrared spectroscopy (NIR), and anthropometric measurements (ANTHRO). Measurements were taken on 2 occasions separated by 7 days to assess the reliability of the methods. Following this, 30 subjects returned to the laboratory for an additional test day following 7 days of Cr supplementation (20 g.d(-1) Cr + 140 g.d(-1) dextrose) to assess each method's ability to detect acute changes in FFM. In terms of reliability, we found excellent test-retest correlations for all 5 methods, ranging from 0.983 to 0.998 (p < 0.001). The mean biases for the 5 methods were close to 0 (range -0.1 to 0.3 kg) and their 95% limits of agreement (LOAs) were within acceptable limits (HW = -1.1 to 1.7 kg; ADP = -1.1 to 1.2 kg; BIA = -1.0 to 1.0 kg; NIR = -1.4 to 1.4 kg); however, the 95% LOAs were slightly wider for ANTHRO (-2.4 to 2.6 kg). Following Cr supplementation there was a significant increase in body mass (from 77.9 +/- 10.1 kg to 78.9 +/- 10.3 kg, p = 0.000). In addition, all 5 body composition techniques detected the change in FFM to a similar degree (mean change: HW = 0.9 +/- 0.6 kg; ADP = 0.9 +/- 0.6 kg; BIA = 0.9 +/- 0.6 kg; NIR = 0.8 +/- 0.5 kg; ANTHRO = 1.0 +/- 0.7 kg; intraclass correlation coefficient = 0.962). We conclude that between-day differences in FFM estimation were within acceptable limits, with the possible exception of ANTHRO. In addition, all 5 methods provided similar measures of FFM change during acute Cr supplementation.