M Steijaert1, P Deurenberg, L Van Gaal, I De Leeuw. 1. Department of Endocrinology, Metabolism and Clinical Nutrition, Faculty of Medicine, University Hospital Antwerp, Belgium.
Abstract
OBJECTIVE: To validate the assessment of total body water (TBW) and extracellular water (ECW) by multi-frequency bioelectrical impedance. SUBJECTS: Twenty-five overweight but otherwise healthy subjects and 20 lean subjects. DESIGN: Cross-sectional. MEASUREMENTS: TBW and ECW were determined by dilution techniques. Prediction equations from the literature were used to calculate TBW and ECW from measured impedance at 100 and 50 kHz or 1 and 5 kHz, respectively. In 18 of the obese subjects, impedance was also measured with the electrodes placed at proximal sites. RESULTS: In lean and obese subjects, significant correlations were observed between the impedance index (H2/Z) at high frequencies with TBW (r = 0.90, P < 0.001 in lean and r = 0.80, P < 0.001 in obese subjects) and at low frequencies with ECW (r = 0.87, P < 0.001 and r = 0.77, P < 0.001 respectively). Proximal placement of electrodes slightly improved the correlation between the impedance index and TBW (from r = 0.83 to r = 0.90 at 50 kHz and from r = 0.85 to r = 0.90 at 100 kHz) and ECW (from r = 0.77 to r = 0.83 at 1 kHz and from r = 0.79 to r = 0.85 at 5 kHz). The association of ECW and TBW with H2/Z was different for obese and lean subjects: in obese subjects a given amount of TBW or ECW corresponded with a lower index. An equation consisting only of the impedance index could predict TBW and ECW with small mean errors in lean (1.3 and 0.8 kg respectively) and obese subjects (0.1 and 0.0 kg respectively). Applying a more specific equation, including other subject characteristics, resulted in larger prediction errors in obese subjects, illustrating the population specificity of prediction equations. Furthermore an association was observed of the prediction bias of TBW and ECW with TBW (r = 0.48, P < 0.01) and ECW (r = 0.70, P < 0.001) respectively, and with body water distribution (r = -0.38 and r = 0.33 respectively, P < 0.05). TBW and ECW were also associated with weight (r = 0.76 and r = 0.71 respectively, P < 0.001) and body mass index (BMI) (r = 0.54 and r = 0.53 respectively, P < 0.001). CONCLUSION: It appeared from this study that the accuracy of TBW and ECW estimation with the impedance technique is dependent on the absolute amount of TBW and ECW. A higher amount of TBW and ECW in obese subjects may contribute to a difference in prediction error between lean and obese individuals.
OBJECTIVE: To validate the assessment of total body water (TBW) and extracellular water (ECW) by multi-frequency bioelectrical impedance. SUBJECTS: Twenty-five overweight but otherwise healthy subjects and 20 lean subjects. DESIGN: Cross-sectional. MEASUREMENTS: TBW and ECW were determined by dilution techniques. Prediction equations from the literature were used to calculate TBW and ECW from measured impedance at 100 and 50 kHz or 1 and 5 kHz, respectively. In 18 of the obese subjects, impedance was also measured with the electrodes placed at proximal sites. RESULTS: In lean and obese subjects, significant correlations were observed between the impedance index (H2/Z) at high frequencies with TBW (r = 0.90, P < 0.001 in lean and r = 0.80, P < 0.001 in obese subjects) and at low frequencies with ECW (r = 0.87, P < 0.001 and r = 0.77, P < 0.001 respectively). Proximal placement of electrodes slightly improved the correlation between the impedance index and TBW (from r = 0.83 to r = 0.90 at 50 kHz and from r = 0.85 to r = 0.90 at 100 kHz) and ECW (from r = 0.77 to r = 0.83 at 1 kHz and from r = 0.79 to r = 0.85 at 5 kHz). The association of ECW and TBW with H2/Z was different for obese and lean subjects: in obese subjects a given amount of TBW or ECW corresponded with a lower index. An equation consisting only of the impedance index could predict TBW and ECW with small mean errors in lean (1.3 and 0.8 kg respectively) and obese subjects (0.1 and 0.0 kg respectively). Applying a more specific equation, including other subject characteristics, resulted in larger prediction errors in obese subjects, illustrating the population specificity of prediction equations. Furthermore an association was observed of the prediction bias of TBW and ECW with TBW (r = 0.48, P < 0.01) and ECW (r = 0.70, P < 0.001) respectively, and with body water distribution (r = -0.38 and r = 0.33 respectively, P < 0.05). TBW and ECW were also associated with weight (r = 0.76 and r = 0.71 respectively, P < 0.001) and body mass index (BMI) (r = 0.54 and r = 0.53 respectively, P < 0.001). CONCLUSION: It appeared from this study that the accuracy of TBW and ECW estimation with the impedance technique is dependent on the absolute amount of TBW and ECW. A higher amount of TBW and ECW in obese subjects may contribute to a difference in prediction error between lean and obese individuals.
Authors: Jennifer R Mager; Shalamar D Sibley; Tiffany R Beckman; Todd A Kellogg; Carrie P Earthman Journal: Clin Nutr Date: 2008-08-03 Impact factor: 7.324
Authors: Jeffrey L Segar; Kirthikaa Balapattabi; John J Reho; Connie C Grobe; Colin M L Burnett; Justin L Grobe Journal: Am J Physiol Regul Integr Comp Physiol Date: 2020-10-21 Impact factor: 3.619