PURPOSE: Determine if: 1) dual energy x-ray absorptiometry (DEXA), bioelectrical impedance spectroscopy (BIS) and bioelectrical impedance analysis scale (S) can be used interchangeably as accurate surrogate measurements of % body fat (%BF) and % total body water (%TBW) before and after a 100 mile footrace; 2) If % body mass (%BM) change is an accurate indicator of %TBW change during such exercise. METHODS: BM was measured using S without shoes in running clothes. %TBW and %BF were obtained before and after the race using the S, DEXA and BIS sequentially. RESULTS: Pre-race %BF and %TBW values among 18 runners were not different for the three measurement techniques and correlation coefficients (Pearson's r) showed positive associations among all three measurement techniques. Post-race comparison measures in 10 runners (6 finishers; 4 non-finishers covering 30-62 miles) revealed significant differences in mean (±SD) %BF with BIS (12.0 ± 5.4) vs. DEXA (18.7 ± 3.6; p<0.05) and vs. S (18.2 ± 4.8; p<0.05) and in %TBW with BIS (64.1 ± 4.0) vs. DEXA (56.7 ± 4.1; p<0.05) and vs. S (59.6 ± 3.5; p<0.05). Significant correlations in post-race measurements were noted only between BIS and DEXA in %BF (r=0.81; p<0.05) and in %TBW (r=0.80; p<0.05). However, Bland-Altman plots revealed DEXA values measuring consistently higher than BIS values in %BF with a mean difference between the two averages of 3.6 and limits of agreement (LOA) between +5.4 and -12.6. Similarly, Bland-Altman plots for %TBW revealed a mean difference of 2.4 with BIS measuring consistently higher than DEXA and LOA between +9.4 and -4.6. Mean %BM loss during the race of 1.6 ± 2.3 was not significant. Mean percent decrease (D) in %BF with DEXA (-0.5 ± 0.6), BIS (-4.1 ± 4.6) and S (-1.0 ± 1.0), and mean percent increase (D) in %TBW with DEXA (0.4 ± 1.2), BIS (2.7 ± 3.8) and S (0.7 ± 0.7) were not significantly different between techniques with high ±SD's. CONCLUSIONS: DEXA, BIS and S can be used interchangeably to measure %BF and %TBW in steady-state (pre-race) conditions but are not accurate surrogate measures in non-steady-state (post-race) scenarios. %BW loss during a 100 mile footrace cannot be solely attributed to changes in %TBW and therefore other changes in body composition (%BF) should be factored into consideration. Sponsored by Western States Endurance Run Foundation and ImpediMed™
Read full abstract