A non-invasive, on-line deuterium dilution technique for the measurement of total body water in haemodialysis patients

Abstract

Background: Despite its importance, total body water (TBW) is usually estimated rather than measured due to the complexity of isotope dilution methods. The aim of this study was to demonstrate the applicability in haemodialysis (HD) patients of a recently developed on-line breath test, previously validated in healthy subjects, that uses the gold standard deuterium dilution method to measure TBW. In particular we wished to show that a pre-dialysis estimation was as good as a post-dialysis equilibrated measurement in order to avoid patients needing to remain behind after dialysis treatment.

Methods: The dispersal kinetics of breath HDO, measured using a flowing afterglow mass spectrometer (FA-MS) following ingestion of D(2)O immediately post-dialysis, were determined in 12 haemodialysis patients and used to calculate the absolute TBW(PostHD) after full equilibration. TBW(PreHD) was then determined from breath samples taken immediately prior to the next dialysis. This measurement was adjusted for the interdialytic weight change and urine output (TBW(PreHD-adjusted)) and compared to the TBW(PostHD). The accuracy and precision of FA-MS was also assessed using known concentrations of deuterium-enriched water samples.

Results: Mean TBW(PostHD) was 50.0 +/- 9.3 L and TBW(PreHD-adjusted) was 50.7 +/- 9.0 L. They were highly correlated (R = 0.99, P < 0.001) with a CV of 2.6%. The mean difference was +0.74 L (SEM 0.35, 95% CI -0.03 to 1.51 L, P = 0.059), compatible with a daily insensible loss of 0.37 L. Accuracy and precision of FA-MS were comparable to the previous validation work.

Conclusions: This non-invasive adaptation of the D isotope dilution method for determining TBW can be applied to haemodialysis patients who show deuterium equilibration kinetics identical to normal subjects; a pre-dialysis estimation may be used to determine TBW, and so avoiding the necessity to remain behind after dialysis making this suitable for application in the clinical setting.

Figure 1

The subject exhales gently into a disposable tube; breath is sampled from a calibrated capillary into the flow tube where it mixes with protonated water (H₃O⁺) created by a microwave discharge. Whilst flowing down the tube, driven by a constant stream of the inert gas (e.g. Helium), breath water vapour forms clusters with the H₃O⁺, and those with a m/z ratio of 74 will contain Deuterium. A relative increase in these water clusters when compared to m/z 75 detected by the mass spectrometer enables calculation of deuterium enrichment.

Figure 2

Flow diagram of the experimental protocol

Figure 3

Bland and Alman plot of observed versus expected measurements of D abundance in the vapour phase above known dilutions of HDO in tap water. Each point represents the mean (±95% CI) difference at 8 different dilutions each measured at 9 independent experimental sessions. The average difference across the range of dilutions was −0.25 ppm; the outer limits for all the 95%CI are ±20 ppm, 3% of a typical measurement.

Figure 4. Deuterium Kinetics

(a) Average equilibration curve of breath D following oral ingestion; the initial large peak is due to D still present in the mouth. Subsequently there is gastrointestinal absorption and then equilibration by 2 hours. (b) Over the next few days there is exponential clearance of D from the breath, (mean values, those marked PRE are pre-dialysis) predominantly during dialysis treatments (shown as vertical arrows) such that it is cleared by 10 days.

Figure 4. Deuterium Kinetics

(a) Average equilibration curve of breath D following oral ingestion; the initial large peak is due to D still present in the mouth. Subsequently there is gastrointestinal absorption and then equilibration by 2 hours. (b) Over the next few days there is exponential clearance of D from the breath, (mean values, those marked PRE are pre-dialysis) predominantly during dialysis treatments (shown as vertical arrows) such that it is cleared by 10 days.

Figure 5

Bland and Altman plot comparing the measurement of TBW following post-dialysis equilibration and immediately prior to the next dialysis session, having adjusted for weight changes and urine output. The small systematic difference can be explained by insensible losses; there is no bias across the wide range of TBW volumes studied.