Hydration abnormalities in Nigerian patients on chronic hemodialysis

The state of hydration affects the outcomes of chronic dialysis. Bioelectrical impedance analysis (BIA) provides estimates of body water (V), extracellular volume (ECFV), and fat-free mass (FFM) that allow characterization of hydration. We compared single-frequency BIA measurements before and after 14 hemodialysis sessions in 10 Nigerian patients (6 men, 4 women; 44+/-7 years old) with clinical evaluation (weight removed during dialysis, presence of edema) and with estimates of body water obtained by the Watson, Chertow, and Chumlea anthropometric formulas. Predialysis and postdialysis values of body water did not differ between BIA and anthropometric estimates. However, only the BIA estimate of the change in body water during dialysis (-0.8+/-2.9 L) did not differ from the corresponding change in body weight (-1.3+/-3.0 kg), while anthropometric estimates of the change in body water were significantly lower, approximately one-third of the change in weight. Bioelectrical impedance analysis correctly detected the intradialytic change in body water content (the ratio V/Weight) in 79% of the cases, while anthropometric formula estimates of the same change were erroneous in each case. Compared with patients with clinical postdialysis euvolemia (n=7), those with postdialysis edema (n=5) had higher values of postdialysis BIA ratios V/FFM (0.77+/-0.01 vs. 0.72+/-0.03, p<0.01) and ECFV/V (0.53+/-0.02 vs. 0.47+/-0.06, p<0.05), respectively. Bioelectrical impedance analysis appeared to underestimate body water and extracellular volume in a patient with massive ascites and bilateral pleural effusions. Anthropometric formulas are not appropriate for evaluating the state of hydration in patients on chronic hemodialysis. In contrast, BIA provides estimates of hydration agreeing with clinical estimates in the same patients, although it tends to underestimate body water and extracellular volume in patients with large collections of fluid in central body cavities.