Printer Friendly

The use of infrared spectrophotometry for measuring body water spaces.

Total body water has been used to assess human body composition since the demonstration that a relatively fixed fraction of lean body tissue is occupied by water (1). The deuterated water ([sup.2][H.sub.2]O) isotope dilution method for measuring total body water has now replaced the earlier radioactive method ([sup.3]H) (2). However, the use of gas isotope ratio mass spectrometry to measure [sup.2][H.sub.2]O is expensive, requires specialist knowledge for operation and maintenance, and is somewhat tedious because the water is first transformed to hydrogen (3, 4) or equilibrated with hydrogen (5, 6).

Other methods for the determination of deuterium ([sup.2]H) in water have been proposed (7-10), including infrared Spectrophotometry (11-16). This latter technique has received particular attention because of its relatively high potential sensitivity, which suggests that only a small quantity of deuterium need be administered. Unfortunately, the absorption band centered at ~2500 [cm.sup.-1], attributable to excitation of the D-O bond, is present as a shoulder on the band at ~2130 [cm.sup.-1], which is attributable to the H-O bond. If a single wavelength is monitored (as is usually done with a dispersive instrument), this leads to uncertainty in the baseline reading and consequently limits the accuracy of the technique at low enrichments. In recent years, there have been considerable advances in infrared instrumentation through the application of Fourier transform infrared (FTIR) techniques. This allows rapid monitoring of the whole of the absorption region (17-19). However, these do not appear to have been used to measure enrichment in human plasma or other body fluids. Therefore, the aim of this study was to assess the precision of the new methodology and to compare results obtained from several physiological fluids with those obtained with mass spectrometry.

Subjects and Methods

Five subjects took part in this study (mean weight, 75.75 [+ or -] 4.0 kg; mean body mass index, 24.2 [+ or -] 1.70 kg/[m.sup.2]). Each subject arrived at the laboratory after an overnight fast (12 [+ or -] 2 h) and received an oral dose of 1.5 mol (30 g) of [D.sub.2]O, 99.9% (Sigma Chemicals) followed by ~200 g of tap water [to produce an enrichment of 450-750 [micro]mol/mol (ppm)]. The straw and dosing bottle used for the administration of the [D.sub.2]O were retained and weighed to obtain the exact amount of dosing solution consumed. Samples of blood (5- to 10-mL samples collected in heparin-containing stoppered tubes) and saliva (4-mL samples collected into clean, dry tubes) were obtained immediately before and at 4, 5, and 6 h after dosing. Spot urine samples were also collected before and at 4, 5, and 6 h after dosing. The subjects were asked to void their bladders 30 min before the collection time and to refrain from drinking during the 30 min prior to each saliva collection. Otherwise, they were allowed to drink small amounts of water ad lib (<200 mL during the study).


The saliva and plasma samples were centrifuged immediately after collection, and the supernatants were retained. All samples were stored frozen at -20 [degrees]C. Prior to analysis, the samples were thawed, vortexed, and centrifuged. Preliminary work showed that the predose urine samples were not suitable for use as background samples for estimation of deuterium by infrared spectroscopy because of the large variations in composition of sequential urine samples. To overcome these variations, we freeze-dried weighed aliquots of the postdose urine samples and then reconstituted them gravimetrically with local tap water containing naturally occurring deuterium to provide samples of the same constituents, but without deuterium enrichment, to act as background samples for the postdose urine samples. Local tap water was also analyzed [mass spectrometry value, 147.7 [micro]mol/mol (ppm; this value was essentially identical to values obtained from background samples of physiological fluids, e.g., saliva and plasma)].


The CVs for the new method of analysis were assessed using 8, 16, 32, and 64 scans for solutions containing 200, 400, 800, and 1000 [micro]mol/mol deuterium. Five readings were averaged to produce a single analytical result and repeat analyses for each sample. The same procedure was used to establish the CVs in plasma, saliva, and urine, using 8 or 64 scans per sample. The temperature of the sample compartment of the instrument was found to be stable (32 [+ or -] 1 [degrees]C) without thermostatic control. When not in use, the sample holders were kept in the instrument with the instrument switched on to maintain them at operating temperature.

The precision of the standard procedure for measuring areas under the curve, which is incorporated into the software of FTIR spectrophotometer, and a modified procedure (20) was determined using the techniques described below. The infrared spectra of aqueous samples were measured in the range 1800-2800 [cm.sup.-1] on a Mattson Genesis series FTIR spectrophotometer (Mattson Instruments), equipped with an automated sample shuttle and a pair of matched calcium fluoride sample cells with a 0.1-mm pathlength. The specifications of the instrument were as follows: infrared ceramic emitter; detector, LaTa[O.sub.3] (uncooled); beam sputter, KBr; spectral range, 5500-225 [cm.sup.-1]; and resolution, 1 [cm.sup.-1]. When the matched cells were switched, the error was less than [+ or -]1 [micro]mol/mol (SD). A sample volume of 0.2 mL is adequate for the total procedure.

Urine samples were measured using a single-sample system and a cell near the end of its useful life because ammonia and phosphate are both known to react with calcium fluoride, causing the crystal to become opaque. Postdose (or calibrator) samples were read at the same time as predose (background) samples under the same ambient conditions. Before the spectrum of each sample was recorded, that of a background sample with natural deuterium abundance was measured. An automatic sample shuttle made it possible, without disturbing the spectrometer, to mimic a double-beam system. After correction for natural abundance, the digitized spectra were exported to a spreadsheet for subsequent processing. Calibration was made after measurements of eight samples. The calibration procedure involved preparation of [sup.2][H.sub.2]O calibrators by gravimetric dilution of [sup.2][H.sub.2]O (99.9 APE; Sigma) with local tap water. The enrichment of these calibrations was confirmed by mass spectrometry (see Fig. 1 for a typical calibration curve).


The infrared spectrum of [D.sub.2]O in water obtained from the FTIR instrument is shown in Fig. 2. The signal from the deuterium-oxygen bond appears as a weak shoulder on the intense nondeuterated water band. Baseline subtraction can be performed with unenriched water as a reference, using the sample shuttle, but it is still difficult to determine the zero enrichment background with any great precision. This is demonstrated in Fig. 3, in which the zero background enrichment is subtracted from the postdose spectrum. The line at 2500 [cm.sup.-1] shows the single frequency at which most previous workers have made measurements (13, 16).



To overcome this uncertainty in baseline positioning, a similar approach to that developed in this laboratory for gas chromatography/mass spectrometry isotope mass ratio determination was used (20). The spectra were digitized in the region 2665-2415 [cm.sup..sup.-1]. The sample spectrum was then compared with the sum of the reference spectrum and its first two derivatives, generated numerically. The inclusion of the derivative functions allowed correction for slight spectral energy shift. Baseline subtraction was achieved by the inclusion of a linear function of light energy in the fitting procedure. This method compared the spectrophotometer outputs for sample and a reference across the whole of the relevant peak, separating the responses into that which is common to the absorption peak itself and that which is solely a property of the linear baselines beneath the peaks. Admixture of the first, second, and subsequent derivatives of the observed lineshape for the reference material allowed for nonsuperimposibility of the spectra because of instrumental drift. The equation used is:

S([bar.[nu]]) = [rho][delta]/2 ([delta] + 1)R([bar.[nu]]+ 1) + [rho](1 - [[delta].sup.2])R([bar.[nu]]) + [rho][delta]/2 ([delta] - 1) R([bar.[nu]] - 1) + A[bar.[nu]] + B

where [nu] is the wavenumber, S([bar.[nu]]) is the sample spectrum, R([bar.[nu]]) is the reference spectrum, [rho] is the ratio between the peak intensities, b is the instrumental shift along the wavenumber axis between the sample and reference spectra, and A and B are constants related to the slopes and intercepts of the backgrounds of the spectra. In practice, the n points of the digitized spectra were exported from the spectrophotometer software directly into a spreadsheet and the following matrix expression was solved:


where the (n - 2) row vector [] is made up of the digitized data for the sample, omitting the first and last points. ?? is a matrix of five columns with (n - 2) rows. The first column contains the spectrum for the reference, omitting the last two readings. The second row contains the same data, but now starts at the second reading, omitting the last point; the third column contains the same data again, but omitting the first two points. The fourth column is the wavenumber, and each element of the fifth column is equal to unity. [] is a vector of the best estimates of the parameters of the fit. Most importantly, the ratio of spectral intensities, [rho], can be obtained by summing the first three elements of []. The relative amounts of deuterium in the sample can then be estimated by assuming the Beer-Lambert law to hold. In calculating the enrichment of deuterium in salivary and plasma water, we assumed (21) (and confirmed experimentally) that the hydration fraction of saliva was 99.5% [+ or -] 0.2% (in kg/L), and that of plasma, 94% [+ or -] 0.4%. The hydration fraction of the urine samples was calculated from the freeze-dried samples.



Excess sample was isotopically equilibrated with hydrogen gas for 3 days, using platinum on alumina powder as a catalyst (6). The deuterium content of the resultant gas was compared with that of known calibrators, using a Sira 10 (VG Instruments) equipped with a dual inlet, and the results were converted to the SMOW/SLAP scale.


Samples enriched in deuterium by known amounts were prepared by the gravimetric addition of [D.sub.2]O (99.9%; Sigma-Aldrich) to an aliquot of natural abundance water (Cambridge tap water) and then adjustment to a known volume with tap water. A portion of this enriched sample was further diluted in Cambridge tap water to give samples enriched above natural abundance by 1000, 800, 600, 400, and 200 [micro]mol/mol.


Results were expressed as means and SD. Comparisons of results obtained with the two methods and between different physiological fluids collected at the same time were made using the Bland and Altman (22) method.


The study was approved by the local ethics committee, and informed written consent was obtained from all the participants.


At all levels of enrichment, the method of Bluck and Coward (20) for measuring spectral intensity was found to be more precise than the traditional integration method used by the instrument (Table 1). There was little or no significant difference between results obtained using 8, 16, 32, and 64 scans per reading (Table 2). The time taken to obtain 64 scans was 2.5 min. The precision for measurement of deuterium enrichment plasma and saliva was very similar to that of calibrator solutions (Table 3).

The results in Table 4 show that there is close agreement between both methods for mass spectrometry and infrared deuterium analysis when measurements were made in plasma, saliva, and urine (none of the comparisons were significantly different from each other; paired t-test). There also was close agreement in the enrichment between the different types of fluid analyzed (Table 5) and no significant increase in the difference between methods as deuterium enrichment increased. It was possible to analyze samples within 30 min of receiving them (including time for centrifugation).


This study demonstrates that infrared spectroscopy, which incorporates an improved analytical algorithm for measuring spectral intensity, can be used to measure the enrichment of [D.sub.2]O in plasma and saliva rapidly and precisely (CV, 0.1-0.9% for samples with deuterium enrichment of 400 [micro]mol/mol or more). The technique has a precision (CV, 0.1-0.9%)that is several fold better than the traditional method for measuring [D.sub.2]O enrichment with FTIR (CV, 3-6% over the same range of enrichment; see text).

This study also demonstrates good agreement between the infrared and mass spectrometry measurements of deuterium enrichment in saliva, plasma, and urine. Furthermore, the differences between plasma and salivary measurements obtained with the infrared technique were similar to those obtained with mass spectrometry. However, the dose of deuterium required to achieve a CV <1% for repeated measurements of the same sample is several fold greater with the infrared method than with mass spectrometry. On the other hand, changes in background enrichment, which may occur after the administration of intravenous fluids or the drinking of distilled water, will influence the estimation of total body water by the infrared method to a lesser extent than mass spectrometry, which usually involves measurement of smaller increments in deuterium enrichment. Furthermore, the operation of mass spectrometers requires more expertise than the operation of the infrared spectrophotometer, which can yield results shortly after the samples are obtained (plasma and saliva). A particular problem with urine analysis is the presence of variable amounts of interfering urinary substances, which makes predose urine samples less suitable for background measurements than the corresponding plasma and saliva samples. Although appropriate predose urine samples can be prepared by removing the enriched water from the urine (freeze-drying) and replacing it with tap water, the process is time-consuming and unsuitable for bedside analysis. Another problem with urine analysis is that the standard calcium fluoride cells are not suitable for prolonged and repeated use, although our preliminary studies with zinc selenide cells suggest that they are a suitable alternative because they are not "attacked" by urine. Sapphire cells may also be appropriate.

Finally, there have been reports that the use of a single-cell system to measure enrichment of deuterium in urine (instead of the matched-cell system for plasma and saliva measurements) could explain the higher CV for the urinary measurements (Table 3) and the greater discrepancy between urinary and alternative fluid measurements when the infrared technique is used, compared with the mass spectrometry method (Table 5). There have been reports of temperature dependence of the infrared spectrum of both [H.sub.2]O and [D.sub.2]O in the shorter wavelength 5000-16 000 [cm.sup.-1] region (23); therefore, some workers have used thermostatically controlled cells (13,16). However, because the operating temperature was stable and the primary purpose of this work was to assess the technique as a simple, rapid, and inexpensive method for determination of total body water, no attempt was made to thermoregulate the sample.

In conclusion, this report describes an improved infrared analytical method for measuring [D.sub.2]O enrichment in physiological fluids. It is simple, precise, rapid, and of potential value to the clinician who monitors and adjusts fluid therapy on a daily basis.

Received February 18, 1999; accepted April 15, 1999.


(1.) Pace N, Rathburn E. Studies on body composition. III. The body water and chemically combined nitrogen content in relation to fat content. J Biol Chem 1945;158:685-91.

(2.) Panaretto B. Estimation of body composition by the dilution of hydrogen isotopes. Body Composition in animals and man. Washington DC: National Academy of Sciences-National Research Council, 1968:200-17.

(3.) Coleman ML, Shepherd TJ, Durham JJ, Rouse JE, Moore GA. Reduction of water with zinc for hydrogen isotope analysis. Anal Chem 1982;54:993-5.

(4.) Barrie A, Coward WA. A rapid analytical technique for the determination of energy expenditure by the doubly labeled water method. Biomed Mass Spectrom 1985;12:535-41.

(5.) Coplen T, Wildman JD, Chen J. Improvements in the gaseous-hydrogen water equilibration technique for hydrogen isotope ratio analysis. Anal Chem 1991;63:910-2.

(6.) Scrimgeour CM, Rollo MM, Mudambo SMKT, Handley LL, Prosser SJ. A simplified method for deuterium hydrogen isotope ratio measurements on water samples of biological origin. Biol Mass Spectrom 1993;22:383-7.

(7.) Schloerb R, Frus-Hansen J, Edelman IS, Sheldon DB, Moore SD. The measurement of deuterium oxide in body fluids by the falling drop method. J Lab Clin Med 1951;37:652-62.

(8.) Reaser P. Determination of deuterium oxide in water by measurement of freezing point. Science 1958;128:415-6.

(9.) Mendez J, Prokop E, Piton-Reategui E, Akers R, Buskirk ER. Total body water by [D.sub.2]O dilution using saliva samples and gas chromatography. J Appl Physiol 1970;28:354-7.

(10.) Nielsen WC, Krzywicki HJ, Johnson HL, Consolazio CF. Use and evaluation of gas chromatography for determination of deuterium in body fluids. J Appl Physiol 1971;31:957-61.

(11.) Thornton V, Condon FE. Infrared spectrometric determination of deuterium oxide in water. Anal Chem 1950;22:690-1.

(12.) Turner MD, Neely WA, Hardy JD. Rapid determination of deuterium oxide in biological fluids. J Appl Physiol 1960;15:309-10.

(13.) Byers F. Extraction and measurement of deuterium oxide at tracer levels in biological fluids. Anal Biochem 1979;98:208-31.

(14.) Haschke F. Body composition of adolescent males. Acta Paediatr Stand 1983;(Suppl 307):13-23.

(15.) Wang J, Pierson R, Kelly WG. A rapid method for the determination of deuterium oxide in urine: application to the measurement of total body water. J Lab Clin Med 1973;82:170-8.

(16.) Lukaski H, Johnson P. A simple, inexpensive method of determining total body water using a tracer dose of [D.sub.2]O and infrared absorption of body fluids. Am J Clin Nutr 1985;41:363-70.

(17.) Pang KS, Xu N, Goresky CA. [D.sub.2]O as a substitute for [sup.3][H.sub.2]O, as a reference indicator in liver multiple-indicator dilution studies. Am J Physiol 1991;261:6929-36.

(18.) Fusch C, Spririg N, Moeller H. Fourier transform infrared spectroscopy measures [sup.1]H/[.sup.2]H ratios of native water with a precision comparable to that of isotope ratio mass spectrometry. Eur J Clin Chem Clin Biochem 1993;31:639-44.

(19.) Veryaminor SY, Prendergast FG. Water ([H.sub.2]O and [D.sub.2]O) molar absorptivity in the 1000-4000 [cm.sup.-1] range and quantitative infrared spectroscopy of aqueous solutions. Anal Biochem 1997; 248:234-45.

(20.) Bluck UC, Coward WA. Peak measurement in gas chromatographic mass spectrometric isotope studies. J Mass Spectrom 1997;32:1212-8.

(21.) Snyder WS, Cook MJ, Nasset ES, Karhausen LR, Howells GP, Tipton IH. Report of the Task Group of Committee 2 of the International Commission on Radiological Protection. Oxford, UK: Pergamon Press, 1976:335-467.

(22.) Bland JM, Altman DG. Statistical methods of assessing agreement between two methods of clinical measurement. Lancet 1986;1:307-10.

(23.) Waggener W. Absorbance of liquid water and deuterium oxide between 0.6 and 1.8 microns. Anal Chem 1958;30:1569-70.


[1] MRC Dunn Clinical Nutrition Centre, Hills Road, Cambridge CB2 2DH, United Kingdom.

[2] MRC Human Nutrition Research, Downhams Lane, Milton Road, Cambridge CB4 1XJ, United Kingdom.

* Author for correspondence. Fax 44 1223 426617.
Table 1. Comparison of the CV (n = 5 separate samples)
obtained using two different methods for measuring the
area under the [D.sub.2]O peak. (a)

 CV, %
[D.sub.2]O calibrator integrated
[micro]mol/mol method New method (b)

1000 3.7 0.35
800 6.2 0.39
400 3.0 0.87
200 5.0 1.4

(a) Eight scans per reading, and five measurements per sample.
Improved algorithm used for measuring area under the curve
(see text).

Table 2. CV (%) for the infrared method using 8, 16, 32,
and 64 scans per reading. (a)

 CV, %
[micro]mol/mol 8 scans 16 scans 32 scans 64 scans
1000 0.35 0.49 0.11 0.20
800 0.39 0.23 0.37 0.25
400 0.87 0.56 0.75 0.30
200 1.40 1.30 1.40 1.4

(a) Five measurements per scan of calibrator solutions of [D.sub.2]O.
Analysis used an improved algorithm for measuring area under the curve
(see text) (17).

Table 3. CV (%) for infrared measurements of enrichment of
deuterium in saliva, plasma, and urine. (a)

Enrichment of CV, %
[micro]mol/mol Saliva Plasma Urine

1000 0.30 0.20 0.50
800 0.22 0.24 0.54
400 0.41 0.31 0.76
200 1.7 1.4 2.0

(a) Sixty-four scans per reading and five measurements per sample on
15 separate samples of each fluid. Single-cell analysis for urine and
dual matched-cell analysis for saliva and plasma, using improved
algorithm for measuring area under the curve (see text).

Table 4. Postdose comparison of deuterium enrichment
measurements ([micro]mol/mol) in plasma, saliva, and urine
obtained with mass spectrometry and the infrared method. (a)

 Plasma, Saliva, Urine,
 [micro]mol/mol [micro]mol/mol [micro]mol/mol

Mean of methods 610 [+ or -] 71 614 [+ or -] 69 617 [+ or -] 67
and infrared)

Difference 6 [+ or -] 11 4 [+ or -] 1 -2 [+ or -] 12
methods (mass
spectrometry -

(a) Separate samples (n = 15) were used for each fluid, with 64 scans
perreading. The improved algorithm was used for measuring the area

under the curve.

Table 5. Postdose enrichments and differences in
enrichments (bias [+ or -] SD) between physiological fluids,
according to method used. (a)


 Mass spectrometry Infrared
 Plasma 613 [+ or -] 68 607 [+ or -] 74
 Saliva 616 [+ or -] 70 612 [+ or -] 68
 Urine 616 [+ or -] 69 619 [+ or -] 65

Differences in enrichment

 Plasma-saliva -3 [+ or -] 15 -5 [+ or -] 13
 Plasma-urine -3 [+ or -] 15 -11 [+ or -] 13
 Saliva-urine 1 [+ or -] 3 -5 [+ or -] 10

(a) Separate samples (n = 15) of each fluid were used, with 64
scans per reading.
COPYRIGHT 1999 American Association for Clinical Chemistry, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1999 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Nutrition
Author:Jennings, Graham; Bluck, Leslie; Wright, Antony; Elia, Marinos
Publication:Clinical Chemistry
Date:Jul 1, 1999
Previous Article:Plasma porphyrins in the porphyrias.
Next Article:Automated albumin method underestimates pharmaceutical-grade albumin in vivo.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |