点击显示 收起
1 From the Research and Development department, Fresenius Medical Care, Bad Homburg, Germany (PWC, PW, and UMM); the Institut für Humanernährung und Lebensmittelkunde, Christian Albrechts-Universität, Kiel, Germany (MJM, AB-W, and OK); and the MRC Childhood Nutrition Research Centre, Institute of Child Health, London, United Kingdom (NJF)
2 Data from the Institut für Humanernährung und Lebensmittelkunde were obtained through funding from Fresenius Medical Care. 3 Reprints not available. Address correspondence to PW Chamney, Research and Development, Fresenius Medical Care, Bad Homburg, Germany.
ABSTRACT
Background: Excess fluid (ExF) accumulates in the body in many conditions. Currently, there is no consensus regarding methods that adequately distinguish ExF from fat-free mass.
Objective: The aim was to develop a model to determine fixed hydration constants of primary body tissues enabling ExF to be calculated from whole-body measurements of weight, intracellular water (ICWWB), and extracellular water (ECWWB).
Design: Total body water (TBW) and ECWWB were determined in 104 healthy subjects by using deuterium and NaBr dilution techniques, respectively. Body fat was estimated by using a reference 4-component model, dual-energy X-ray absorptiometry, and air-displacement plethysmography. The model considered 3 compartments: normally hydrated lean tissue (NH_LT), normally hydrated adipose tissue (NH_AT), and ExF. Hydration fractions (HF) of NH_LT and NH_AT were obtained assuming zero ExF within the diverse healthy population studied.
Results: The HF of NH_LT mass was 0.703 ± 0.009 with an ECW component of 0.266 ± 0.007. The HF of NH_AT mass was 0.197 ± 0.042 with an ECW component of 0.127 ± 0.015. The ratio of ECW to ICW in NH_LT was 0.63 compared with 1.88 in NH_AT. ExF can be estimated with a precision of 0.5 kg.
Conclusions: To calculate ExF over a wide range of body compositions, it is important that the model takes into account the different ratios of ECW to ICW in NH_LT and NH_AT. This eliminates the need for adult age and sex inputs into the model presented. Quantification of ExF will be beneficial in the guidance of treatment strategies to control ExF in the clinical setting.
Key Words: Excess fluid • tissue hydration • normal hydration • body composition • adipose tissue • ECW:ICW ratio
INTRODUCTION
Diseases such as cardiac impairment and kidney failure often lead to an accumulation of excess fluid (ExF), increasing the body's state of hydration. ExF may be regarded as an expansion of the extracellular or total body fluid compartments of the body but is not required by the body to maintain homeostasis. In patients with kidney failure, it is essential to remove the ExF to avoid long-term cardiovascular mortality (1-4). Despite the occurrence of ExF, little progress has been made in the body-composition field to develop a method for its identification.
Well-established techniques, such as hydrometry (5, 6), dual-energy X-ray absorptiometry (DXA) (7-9), and underwater weighing (10, 11), are available to obtain estimates of fat-free mass (FFM). The major drawback with these methods, however, is that ExF cannot be distinguished from FFM. Although ExF may be reflected by a rise in ECW (12) or TBW, quantification of ExF is only possible once a hydration reference has been established. The hydration reference represents normal values of ECW and TBW found in healthy control subjects, although the proportions of ECW and TBW vary according to body composition. Where body composition is assumed constant in a given subject group, a hydration reference may be established allowing ExF to be calculated (13). The ratio of ECW to TBW also offers the basis of a hydration reference, and this approach has been applied in patients with corrections for age (14). The hydration of FFM (HFFM), although regarded constant at 0.73 (15), may be influenced by several health factors (16) thus limiting its use for quantification of ExF. Additionally, in more recent work, the hydration of lean soft tissue and its relation with the ratio of ECW to ICW was investigated (17).
One factor influencing the ratio of ECW to ICW (and hence its use as a reference) is the diversity of major body tissues such as the distinct adipose tissue and lean (nonadipose) tissue (18-20). To take into account the dissimilar hydration of relevant tissues and satisfy the need for a method to quantify ExF, we developed a new body-composition model.
SUBJECTS AND METHODS
Subjects and measurements
Data were used from 89 healthy adults recruited in Kiel, Germany. These data were supplemented with those from 15 healthy adults gathered in a previous study conducted in Cambridge United Kingdom (Morgan M, Madden A, Jennings G, Elia M, Fuller N, unpublished observations). Ethical approval was obtained from the ethical board of the Christians Albrechts Universität. The subjects were specifically chosen such that the full extent of the body composition range (in terms of percentage fat) could be investigated. The subject characteristics derived by combining the data (n = 104) from the 2 centers are reported in Table 1.
View this table:
TABLE 1. Characteristics of healthy control subjects1
Each subject fasted overnight for 10 h. At the end of this period, 5-mL venous blood samples were taken for assay calibration and to establish baseline concentrations of deuterium and bromide. A mixture of deuterium oxide (0.4g/kg body weight) and NaBr (50 mg/kg body weight) was administered to each subject. According to past studies, a homogenous distribution of deuterium is achieved within 4 h (5). During this equilibration period any ingestion of food or water was prohibited. At the end of the equilibration period, a second set of samples was taken. Plasma water was obtained from the resulting serum samples by ultrafiltration to remove plasma solids. In Kiel, the volume of TBW was determined from the concentration of deuterium in plasma water by using Fourier transform infrared spectroscopy (FTS 2000; Varian, Deutschland GmbH, Darmstadt, Germany). Deuterium space was corrected for nonaqueous exchange factors by multiplying by 0.945 (21). The same technique was used for analysis of deuterium in the Cambridge data (ATI Mattson Genesis, Cambridge, United Kingdom). Plasma water was assayed additionally for bromide concentration by evaluation of the peak area resulting from HPLC anion exchange chromatography (Waters GmbH, Eschborn, Germany). ECW was determined from bromide space as described by Miller et al (22) from the following equation:
RESULTS
After normalization of fat, TBW, ECWWB, and ICWWB, a clear reduction in body water compartments was observed with increasing body fat content as shown in Figure 2. By linear regression of normalized TBW and ECWWB, as defined by Equations C4 and C8 against DXA fat mass, the following parameters were obtained: aTBW = –0.672, bTBW = 0.703, aECW = –0.185, and bECW = 0.266. By using these regression parameters, the tissue hydration parameters were calculated with Equations C5, C9, C11, and C12. The process was repeated by using fat mass calculated with the Siri equation and the 4-C model. However, these particular methods are not entirely independent of either body weight or TBW and so a small degree of mathematical coupling occurs, which slightly influences bias. The hydration parameters obtained with all 3 methods for fat determination are shown in Table 3 along with those derived by using DXA for comparison. Literature values obtained with the use of in vitro techniques are shown in Table 4. By substitution of the DXA hydration parameters given in Table 3 into Equation B5, the expression for the mass of ExF may be reduced to the following:
DISCUSSION
Central to the new model presented is the concept of normally hydrated lean tissue (NH_LT) and normally hydrated adipose tissue (NH_AT). The sum of these 2 tissues provides a hydration reference against which excess fluid (ExF) can be identified. In the current study, we used deuterium oxide and sodium bromide as the dilution references for TBW and ECW. An advantage of the model is that the hydration parameters in a group of healthy control subjects may be established by using any alternative dilution reference of choice such as total body potassium (TBK). Nevertheless, the hydration parameters obtained in our study for NH_LT and NH_AT were in the range reported in other studies (19, 20, 30, 33). NH_AT was found to have a much lower water content but a higher ratio of ECW to ICW than did NH_LT, which is consistent with the findings of others (19, 20). By definition, HTW_NH_LT cannot exceed HFFM because adipose water is included in FFM but not in NH_LT. Because the range of HFFM in healthy control subjects is 0.69–0.77 (16), the value of 0.8 for HTW_NH_LT obtained by Wang et al (19) may be slightly overestimated.
Although Morse and Soeldner (20) observed no difference in the hydration of NH_AT between obese and nonobese subjects, as seen in Table 4, a more recent study by Martin et al (33) indicated the contrary. This suggests that more detailed investigation of the hydration properties of adipose tissue may be necessary in differing degrees of obesity, not only in terms of total water content but also in terms of intra- and extracellular phases. The model developed in the present study assumed fixed hydration parameters for adipose tissue, and Figure 2 indicates that this serves as a good approximation to the measured data. Furthermore, an improvement in the reproducibility of the measurement methods is necessary before a more complicated model of adipose tissue can be justified.
It is evident from Figure 2 that whereas the ECWWB is clearly lower than the ICWWB when NH_LT dominates body weight, the converse is observed as NH_AT becomes the principle body weight component. This would appear to be the basis of the relative expansion of ECWWB in obese subjects observed in other studies (27). As the ratio of ECW to ICW in NH_AT is a least twice that of NH_LT, then the whole-body proportions of NH_LT and NH_AT would explain the significant differences in the ratios of ECWWB to ICWWB and ECWWB to TBW between the lean and obese subjects, as shown by categorical divisions of body fat (Figure 3).
Although the fat content of female subjects tends to be higher than that of male subjects, as seen in Table 1, these differences are reflected in the relative proportions of NH_LT and NH_AT. Therefore, there was no need to differentiate by sex in our model, because body composition (in terms of fat content) is taken into account from the input measurements of ECWWB, ICWWB, and body weight. It is also unlikely that age contributes fundamentally to the ratio of ECWWB to ICWWB, but is simply due to the proportions of NH_LT and NH_AT, as suggested by Wang et al (19). It can be argued that age-associated increases in fatness may lead to a higher proportion of NH_AT.
In our study, the mean value of HFFM was found to be 0.723, which is consistent with the findings of others (24, 34, 35). In circumstances where ExF accumulates, a rise in HFFM can be expected (16), but the lack of sensitivity of HFFM to variations in ExF has not been emphasized in past studies. This can be illustrated by considering, for example, a subject with 55 kg FFM and 40 L TBW, which leads to a HFFM of 0.727. If this subject now gains 5 kg ExF, HFFM rises to 0.750, a change of just 3%. It is clear that because ExF appears in both the FFM and the TBW, the effect of ExF largely cancels out in HFFM. This renders HFFM a poor choice for providing a reliable hydration reference against which any ExF can be detected.
A significant difference in the ECWWB-to-ICWWB and ECWWB-to-TBW hydration ratios was observed between the lean and obese subgroups, as seen in Figure 3. A similar rise in ratio of ECWWB to ICWWB has been observed in other studies with increasing BMI (27) and with increasing age (36). Therefore, any method for estimation of ExF that assumes fixed ECWWB-to-ICWWB or ECWWB-to-TBW values for the entire population will result in bias errors at extremes of relative fat, as shown in Table 5. In the method described by Lopot et al (14), such effects of body composition have been taken into account by introduction of corrections for age and sex in the ratio of ECWWB to TBW.
Considerable quantities of ExF were found to be present on application of the new model in a previous study of malnourished patients (32), as seen in Table 6. These results are reasonable given the high ratios of ECW to ICW resulting in these subjects coupled with the observation that a significant decrease in the ratio of plasma volume to ECW occurred in the severely malnourished subgroup (32). Methods to obtain the mass of LT or FFM via DXA, for example, do not differentiate ExF from these tissues (37). If a large proportion of the LT or FFM is occupied by ExF, the estimated protein content of these tissues will be reduced. This could lead to ambiguous conclusions regarding nutritional status. In the new model, by contrast, the mass of NH_LT applies regardless of the degree of ExF presented; that is, removal or accumulation of large volumes of fluid do not change the mass of NH_LT. It is proposed, therefore, that estimation of NH_LT mass may offer a more reliable alternative to measures such as LBM and FFM in disease.
APPENDIX A
General model definitions
Whole-body mass (MWB) is given by the sum of the 3 compartments, namely normally hydrated adipose tissue mass (MNH_AT), normally hydrated lean tissue mass (MNH_LT), and excess fluid (ExF) mass (MExF), as follows:
APPENDIX B
Calculation of excess fluid
The sum of the intracellular components of MNH_LT and MNH_AT yields ICWWB, as given by Equation B1:
APPENDIX C
Calculation of principal body-composition parameters
The purpose of the following derivation was to extract the principal body-composition parameters defined by Equations A2-A7. To proceed, it was assumed that the ExF in healthy control subjects has a mean value of zero, ie, MExF = 0. By introducing Equation A1, the mass of total body water, TBW, may be expressed as follows:
ACKNOWLEDGMENTS
The authors thank M Morgan, M Elia, A Madden, and G Jennings for the use of their Cambridge data.
PWC and PW developed the concepts for excess fluid calculation and prepared the manuscript. UMM contributed to technical aspects of the data analysis. MJM organized the study in Kiel and made a number of recommendations to simplify the manuscript. AB-W and OK undertook the studies in Kiel, performed all of the measurement assays and provided valuable scientific support. NJF contributed to refinement of the concepts, and input for alterations to the draft manuscript. PWC, PW, and UMM are employed at the Research and Development department at Fresenius Medical Care. None of the other authors has a conflict of interest to disclose.
REFERENCES