Evaluation of between-methods agreement of extracellular water measurements in adults and children

Abstract

Background: Extracellular water (ECW), a relevant molecular level component for clinical assessment, is commonly obtained by 2 methods that rely on assumptions that may not be possible to test at the time the measurements are made.

Objective: The aim of the current study was to evaluate the degree of agreement between ECW assessment by the sodium bromide dilution (ECW(NaBr)) and total body potassium (TBK; whole-body (40)K counting) to total body water (TBW; isotope dilution) methods (ECW(TBK-TBW)) in an ethnically mixed group of children and adults.

Design: ECW was measured with the ECW(NaBr) and ECW(TBK-TBW) methods in 526 white and African American males and females (86 nonobese children, 193 nonobese adults, and 247 obese adults). Fat mass was assessed with dual-energy X-ray absorptiometry. Multiple regression analysis was used to examine the variables related to between-ECW method differences.

Results: Significant but generally small group mean (+/-SD) differences in ECW were found in the obese adults (1.28 +/- 2.54 kg) and children (-0.71 +/- 1.78 kg). The magnitude of the differences was related to mean ECW in obese adults, children, and nonobese adults, and the relations between these variables were modified by sex for nonobese adults. ECW differences were also dependent on age, weight, sex, and race or on interactions between these variables.

Conclusions: Overall, although good between-method agreement was found across the 3 groups, the degree of agreement varied according to subject characteristics, particularly at the extremes of ECW and body weight. We advance a possible mechanism that may link subject characteristics with the degree of agreement between ECW measurement methods and their underlying assumptions.

FIGURE 1

Mean (±SE) differences between methods for determining extracellular water (ECW) in the sample of nonobese adult white males (n = 45), white females (n = 70), African American males (n = 18), and African American females (n = 60) adjusted to the observed sample mean values of age, weight, and mean ECW using the following model: ECW_TBK-TBW minus ECW_NaBr = −6.509 + [white: 4.948 (P = 0.022); African American: 0] + 0.479 × mean ECW + mean ECW × [male: −0.092 (P = 0.001); female, 0] + weight × [white: −0.081 (P = 0.001); African American: −0.028 (P = 0.256)] + age × [white: −0.006 (P = 0.501); African American: 0.033 (P = 0.004)].

FIGURE 2

Relation between adjusted differences between methods for determining extracellular water (ECW) and the mean ECW for nonobese adults: males (r = 0.290, P = 0.021) and females (r = 0.263, P = 0.002), after adjustment for the effects of race, race by age, and race by weight.

FIGURE 3

Relation between adjusted differences between methods for determining extracellular water (ECW) and mean ECW for obese females, after adjustment for the effects of mean ECW and for the effects of body weight.

FIGURE 4

Relation between adjusted differences between methods for determining extracellular water (ECW) and mean ECW for children, after adjustment for the effects of mean ECW and for the effects of body weight.

FIGURE 5

Agreement between methods using Bland-Altman analysis. The solid line represents the mean differences between the methods. The dashed lines represent the 95% limits of agreement (±1.96 SD). The trend line represents the association between the differences in the methods and the mean of both methods. A: Nonobese adults (r = 0.078, P = 0.278); B: Children (r = 0.135, P = 0.214); C: Obese females (r = 0.250, P < 0.001).