Appendix VIII. Serum Transferrin and Bioelectrical Impedance Analysis

Two indicators of protein-energy status (serum transferrin and bioelectrical impedance analysis) were not deemed valid measures of nutritional status in MD patients by the a priori definition (median panel rating 7 or above), but were considered by the Work Group to be worthy of brief discussion. Both were limited by a lack of specificity as nutritional indicators.

Serum Transferrin

Serum transferrin has been used extensively as a marker of nutritional status, and particularly the visceral protein pools, in individuals with or without CRF.17 It has been suggested that serum transferrin may be more sensitive than albumin as an indicator of nutritional status, possibly because transferrin has a shorter half-life than albumin (~ 8 versus ~ 20 days, respectively).17 Transferrin is a negative acute-phase reactant and is limited by many of the same conditions that limit albumin and prealbumin as indicators of nutritional status. Moreover, the serum transferrin concentration is affected by iron status (ie, serum transferrin increases in iron deficiency and declines following iron loading). Thus, increased iron requirements induced by chronic blood loss from sequestration of blood in the hemodialyzer, blood drawing, or gastrointestinal bleeding and by erythropoietin therapy and the frequent intravenous administration of iron may complicate interpretation of serum transferrin levels.

There is insufficient evidence that serum transferrin is a more sensitive index of PEM than serum albumin in MD patients. Furthermore, its lesser degree of specificity renders it less clinically useful than other serum proteins in this population. Serum transferrin may be more useful in nondialyzed patients with advanced CRF who are less likely to have increased blood loss and who are not receiving erythropoietin or iron therapy.85

Bioelectrical Impedance Analysis (BIA)

BIA is an attractive tool for nutritional assessment of individuals undergoing MD because it is relatively inexpensive to perform, noninvasive and painless, requires minimal operator training, and provides input data that has been correlated with several aspects of body composition.261 Numerous population-based studies have shown a strong direct correlation (r > 0.9) between BIA (height-adjusted resistance) and total body water (TBW). The estimation of other, more complex body compartments (eg, edema-free lean body mass and body cell mass) has proved more difficult, in part because of the relative unavailability of gold standards for estimating compartment sizes. Population-specific regression equations for edema-free lean body mass and body cell mass have not been developed in ESRD. Therefore, systematic bias might magnify the error obtained using regression models derived from nonrenal populations. Errors may compound if multiple compartments are estimated (eg, body cell mass = lean body mass - extracellular water). Therefore, using regression-adjusted BIA parameters (resistance and reactance) to estimate body composition is not sufficiently reliable or valid to recommend its use in MD patients, in contrast to DXA (Guideline 11).

A more compelling argument for the use of BIA is the evidence linking phase angle* with survival in hemodialysis patients.200,264 Although phase angle has been shown to correlate with some nutritional variables (eg, SGA, anthropometric measures, nPNA, and serum albumin, prealbumin, and creatinine), the physiologic basis for the correlation between phase angle and protein-energy nutritional status is not clearly established.200 As with other nutritional indicators (eg, serum albumin; Guideline 3, Rationale), it is not clear that the relation between phase angle and survival is related to nutritional status.

Exploring the link between reactance, resistance, and derivations thereof (eg, phase angle), survival, and nutritional status represents an exciting area of inquiry. If BIA is to be used in the clinical setting, it is recommended that focus be placed on these direct impedance parameters, rather than on regression estimates of edema-free lean body mass or other body compartments.

*Phase angle reflects the relative contributions of fluid (resistance, or R) and cell membranes (reactance, or Xc) to the observed impedance in a biological system. Mathematically, phase angle equals the arc tangent of Xc/R.264