Objective: Bioelectrical impedance analysis (BIA) is widely used as bedside assessment of body composition. Body cell mass (BCM) and intracellular water (ICW) are clinically important body compartments. Estimates of ICW obtained from BIA by different calculation approaches were compared to a reference method in male HIV-infected patients.
Patients: Representative subsample of clinically stable HIV-infected outpatients, consisting of 42 men with a body mass index of 22.4 +/- 3.8 kg/m(2)(range, 13-l31 kg/m(2)).
Methods: Total body potassium was assessed in a whole body counter, and compared to 50 kHz monofrequency BIA and multifrequency bioelectrical impedance spectroscopy. Six different prediction equations for ICW from BIA data were applied. Methods were compared by the Bland-Altman method.
Results: BIA-derived ICW estimates explained 58% to 73% of the observed variance in ICW (TBK), but limits of confidence were wide (-16.6 to +18.2% for the best method). BIA overestimated low ICW (TBK) and underestimated high ICW (TBK) when normalized for weight or height. Mono- and multifrequency BIA were not different in precision but population-specific equations tended to narrower confidence limits.
Conclusion: BIA is an unreliable method to estimate ICW in this population, in contrast to the better established estimation of total body water and extracellular water. Potassium depletion in severe malnutrition may contribute to this finding but a major part of the residual between methods remains unexplained.
Copyright 2000 Harcourt Publishers Ltd.