Knowing more by fewer measurements: about the (In)ability of bioelectric impedance to enhance obesity research in children.

Coordination Center for Clinical Trials Leipzig, Leipzig, Germany. goetz.gelbrich@kksl.uni-leipzig.de

The prevalence of obesity is increasing worldwide. The implications for human health can already be observed in children. Consequently, it is desirable to provide good quantitative descriptions of the relationship of body fat and health risks, such as hypertension. Bioelectric impedance analysis has been frequently praised to be useful for assessing body fat. Devices to analyse body composition based on this technique seem to be selling well, while the real gain in information they provide is unclear. Here we show in a cohort of 2,218 schoolchildren that the body mass index (which is more easily and less costly to determine) is a better predictor of hypertension than the data delivered by impedance analysis. Moreover, we demonstrate that the output of a random number generator is competitive with impedance measurement for this purpose. It is explained by simple arguments from physics why the formulas for the computation of body fat from bioelectric impedance obliterate rather than clarify the relationship of obesity and hypertension. As a consequence, we suggest questioning the opinion propagated by others that bioelectric impedance analysis is a useful tool in field studies on body fat in children. Measurements requiring more effort (compared to simpler methods) should be proved to add worthwhile information, otherwise they should be avoided.

PMID: 15813605 [PubMed - indexed for MEDLINE]