Altered body water distribution in subjects with juvenile rheumatoid arthritis and its effects on the measurement of water compartments from bioelectric impedance.

OBJECTIVE To assess the reliability of bioelectric impedance analysis (BIA) for predicting total body water (TBW) and extracellular water (ECW) in children affected by juvenile rheumatoid arthritis (JRA). SUBJECTS Thirty-nine children affected by JRA and 23 healthy children of similar age (11.0 +/- 3.6, range 3.0-19.0 y) were recruited for the study. METHODS TBW and ECW were measured by deuterium oxide and bromide dilution, respectively. Bioelectric impedance (Z) was measured at frequencies of 5, 50 and 100 kHz. The prediction of TBW and ECW from BIA was based on the impedance index (ZI = height2/Z, cm2/omega). RESULTS TBW standardized per kg of body weight and ECW standardized per litre of TBW were significantly higher in JRA as compared to control patients (59.7 +/- 2.4 vs 57.7 +/- 2.7% and 44.5 +/- 4.6 vs 38.1 +/- 7.9%, with P < 0.005 and P < 0.0001, respectively). Moreover, intracellular water standardized per litre of TBW was significantly lower in JRA than in control subjects (55.5 +/- 4.6 vs 62.5 +/- 8.1, with P < 0.0001). In both controls and patients, the use of ZI at 5kHz offered the more accurate prediction of ECW. However, the use of ZI at 100 kHz did not offer a better prediction of TBW as compared to its value of 50 kHz. Control-generated formulae for predicting water compartments from BIA [TBW = 0.716 x ZI at 100 kHz-1.504, r = 0.934, s.e.e. = 2.2 l;:ECW = 0.430 x ZI5-3.652, r = 0.869(7) s.e.e. = 1.7 l] underestimated TBW and ECW in JRA patients. However, population-specific formulae [TBW (1) = 0.766 x ZI at 100 kHz-0.053, r = 0.939, s.e.e. = 2.8 l; ECW (l) = 0.399 x ZI at 5 kHz-0.283, r = 0.886, s.e.e. = 1.7 l] allowed an accurate prediction of TBW and ECW in JRA patients, taking into account their altered body water distribution. CONCLUSIONS Altered water distribution impedes the use of formulae developed on healthy children to predict TBW and ECW from BIA and JRA patients. It is hypothesized that chronic inflammation and subclinical malnutrition may be responsible for the altered body water distribution of JRA patients. Traditional body composition models may require adjustments for use in JRA children due to their altered body hydration and water distribution.