Validation of bioimpedance analysis as a measure of change in body cell mass as estimated by whole-body counting of potassium in adults.

Abstract

The body cell mass (BCM) is an important measure of macronutrient status, but measurements are difficult to obtain outside of sophisticated research laboratories. Bioimpedance analysis (BIA) is a simple technique that holds promise as a means of estimating body composition. The purpose of this study was to evaluate the ability of BIA to estimate changes in BCM as measured by whole body counting of 40K (TBK). Paired studies of BCM, including both TBK and BIA, were compared in 87 human immunodeficiency virus-positive subjects and in 62 healthy, weight-stable control adults. Potential errors in the predictions were examined. BCM change by TBK and BIA correlated closely (r = .755). After accounting for errors related to repeat measures of TBK, the correlation coefficient was .784, with a standard error of the estimate of 1.24 kg. The differences between predicted and measured BCM change were consistent with a normal distribution. However, there was a systematic error in prediction, with BIA underpredicting the magnitudes of both gains and losses in BCM by TBK. BIA is a useful surrogate for measuring changes in BCM in clinical circumstances. Because TBK assesses only intracellular potassium, whereas BIA reflects all intracellular cations, the underprediction of BCM change by BIA compared with TBK could be related to changes in intracellular potassium concentration as a result of malnutrition or its treatment.