Reply to K de Meer

Body Composition Laboratory, USDA/ARS Children's Nutrition Research Center, Baylor College of Medicine, Houston, TX, E-mail: kellis{at}bcm.tmc.edu

Dear Sir:

We are pleased to see that our recent article (1) in the Journal received such a detailed examination by de Meer. The following is in response to his questions and requests.

We reexamined the data sets used to generate the mean values presented in Table 2. We found that 3 total body water (TBW) values were incorrect and we thank de Meer for bringing them to our attention. The correct values are given in the corresponding erratum. These errors were traced to a transposed entry and an incorrect entry for the 2 male groups and to an incorrect age classification for one girl (13 y rather than 3 y). No other changes were found in the reanalysis of the remaining values presented in Table 2. When we calculated the mean extracellular water (ECW) values as the difference between the mean TBW and mean intracellular water (ICW) values, we found acceptable agreement (after the above-noted corrections) when an uncertainty of only ±0.1 L was allowed for the mean TBW and ICW values, which is much less than the corresponding SEM for these values.

We did not provide resistance values in this article or in our previous article (2), partially because bioelectrical impedance spectroscopy (BIS), unlike single-frequency bioelectrical impedance analysis (BIA), generates an impedance locus plot that must be modeled. Furthermore, others have shown that the resistance values expressed in units can be instrument dependent; hence, we did not want to add to the confusion. However, if one simply rearranges the terms in Eq 1, substitutes the values for weight and height from Table 1, V_ECF from Table 3, and uses the k_ECF constant provided with the Xitron instrument, a mean value for R_e can be easily calculated. Likewise, if the appropriate values (including that derived for R_e) are substituted in Eq 2, one can obtain a mean value for R_L.

It is not too surprising to us that the results for the "constants" used with the BIS calculation are population specific or that they can be dependent on the criterion or reference method chosen for calibration. We already summarized in the Discussion (1) the various values derived for k_ECF and k_p as reported in the literature by other investigators. In a previous study that included adults (2), we compared the BIS estimates with those based on the dilution techniques. Although the dilution techniques, in principle, are conceptually straightforward, the actual measurements and assays are difficult to perform. A series of constants are needed, for example, to convert the deuterium and bromine spaces into estimates for TBW and ECW, respectively. Each step in the assay can introduce errors that are cumulative and we reported that the CV for k_ECF, based on dilution, was >19%, whereas that for k_p was 25–28%. The mean value for k_ECF by dilution was 11% higher than that obtained in the present study. We found that the major difficulties with the dilution reference model were mainly with the bromine technique, a concern that has been expressed by others (3, 4). The CVs for dual-energy X-ray absorptiometry (DXA) and total body potassium (TBK), on the other hand, are of the order of 1–2% (5, 6), which is much better than for the dilution methods. Use of the DXA + TBK model as the reference resulted in significant improvements in the CVs for the BIS constants. The CV for k_ECF, for example, was 8%. We believe that the estimates derived for the BIS constants in the present study are much more reliable than are those based on the dilution techniques. We also believe that the prediction precision when BIS is used can be improved further if age-adjusted constants are used.

As noted in the first National Institutes of Health review of the BIA technology (7), and recently updated by an experienced expert panel (8), there are many technical advantages associated with the bioelectrical impedance methods. We believe that continued investigation of such techniques for use in various clinical applications will better show their strengths and weaknesses.

REFERENCES

1. Ellis KJ, Shypailo RJ, Wong WW. Measurement of body water by multifrequency bioelectrical impedance spectroscopy in a multiethnic pediatric population. Am J Clin Nutr 1999;70:847–53.
2. Ellis KJ, Wong WW. Human hydrometry: comparison of body water by multifrequency bioelectrical impedance with ²H₂O and bromine dilution. J Appl Physiol 1998;85:1056–62.
3. Van Marken Lichtenbelt WD, Kester A, Baarends EM, Westerterp KR. Bromine dilution in adults: optimal equilibration time after oral administration. J Appl Physiol 1996;81:653–6.
4. Kim J, Wang Z, Gallagher D, Kotler DP, Ma K, Heymsfield SB. Extracellular water: sodium bromide dilution estimates compared with other markers in patients with acquired immunodeficiency syndrome. JPEN J Parenter Enteral Nutr 1999;23:61–6.
5. Ellis KJ, Shypailo RJ, Pratt JA, Pond WG. Accuracy of dual-energy X-ray absorptiometry for body composition studies in children. Am J Clin Nutr 1994;60:660–5.
6. Ellis KJ, Shypailo RJ. Whole-body potassium measurements independent of body size. Basic Life Sciences 1993;60:371–5.
7. Bioelectrical impedance analysis in body composition measurement: National Institutes of Health Technology Assessment Conference Statement. Am J Clin Nutr 1996;64(suppl):524S–32S.
8. Ellis KJ, Bell SJ, Chertow GM, et al. Bioelectrical impedance methods in clinical research: a follow-up to the NIH technology assessment conference. Nutrition 1999;15:874–80.