Introduction: Haemodialysis dosing is traditionally based on urea clearance (Kt/V). Aiming for the same Kt/V target, some racial groups have better survival. We investigated whether body composition differs with ethnicity and may lead to differences in Kt/V delivered. Methods: We compared total body water (TBW) measured by multifrequency bioelectrical impedance analysis (MF-BIA) that calculated from standard anthropometric equations. Results: Three hundred and seventy-one adult patients, with a mean age of 58.2 ± 16.6 years, 60.6% of whom were male, 29.1% diabetic, 38.5% Caucasoid, 29.4% African/Afro-Caribbean, 24.8% South Asian and 5.4% East Asian, were studied. TBW measured by MF-BIA differed significantly from that predicted by anthropometric equations. Body fat of women and diabetics was greater, and muscle mass in South Asians was reduced. The difference between the TBW MF-BIA measurement and that of the equation by Watson et al. [11] was associated with % muscle mass (β -10.8, p < 0.001), age (β 0.23, p < 0.001), serum albumin (β -0.24, p < 0.001), body mass index (β 0.91, p = 0.001) and racial origin (β -9.86, p = 0.04). Conclusions: Variation in body composition between ethnic groups potentially leads to over-estimation of delivered dose for some ethnic groups and underestimation for others when using anthropometric equations. MF-BIA assessments of body water should be evaluated as a method for dosing dialysis patients.

Background/Aims: Intradialytic hypotension is the most common complication of modern day haemodialysis (HD). Convective modalities, including haemodiafiltration (HDF) are reported to result in greater cardiovascular stability compared to standard HD, which has been suggested to be due to improved solute transport between compartments. We therefore investigated the effect of treatment on body water by bioimpedance. Methods: We measured the change in extracellular water (ECW) and intracellular water (ICW) in 263 outpatients attending for HD using cooled dialysate and 134 patients for HDF. Results: Patient cohorts were matched for demographics, dialysate composition, ultrafiltration rate, and session duration. The fall in systolic blood pressure following HD was -11.8 mm Hg (-25.3 to 2.3) and not different from that following HDF -12 mm Hg (-27 to 6). Similarly there were no differences in pretreatment serum sodium and dialysate sodium gradient [HD 1 mmol/l (-1 to 3) vs. HDF 2 mmol/l (1 to 4)], or change in serum sodium posttreatment [HD 0 mmol/l (-2 to 2) vs. HDF 1 mmol/l (-1 to 3)]. There were no differences in ICW or ECW pretreatment, and following treatment the reduction in ICW and ECW did not differ [ICW HD -3.5% (-5.7 to -1.8) vs. -4.1% (-6.0 to -1.7), ECW HD -7.1% (-9.4 to -4.7) vs. HDF -7.1% (-9.7 to -4.9)]. Conclusion: We were unable to demonstrate any advantage for HDF over HD using cooled dialysate in terms of changes in blood pressure during a treatment session, or differences in the relative changes in ICW or ECW volumes.

Falls in blood pressure are common during dialysis, and greater for those starting dialysis with the highest systolic pressures, greater dialysate to serum sodium concentration gradient, and also those with the least ECW in the arm. As such, segmental bioimpedance may be useful in highlighting patients at greatest risk for a fall in blood pressure with dialysis.