Serum prealbumin: Is it a marker of nutritional status or of risk of malnutrition?

Hospitalized patients who are undernourished are more likely to develop clinical complications and have relatively poor outcomes, with increased length of stay (LOS)1 and higher mortality compared with well-nourished patients. Provision of adequate nutritional support reduces the complication rate and improves outcome (1 ). Considerable efforts have therefore been made to identify patients at risk of malnutrition, with a view to early provision of nutritional support. A full nutritional assessment is a complex process, involving detailed assessment of nutritional intake, changes in body composition, signs or symptoms of nutritional deficiency or excess, and laboratory tests, and it should include not only protein-energy status but also vitamins and essential trace elements. Because of this complexity, rapid screening tests have been sought to identify patients who may already be malnourished or are at risk of malnutrition, who can then undergo a more detailed nutritional assessment. The screening tools with the most validation for proteinenergy malnutrition include body mass index (weight/ height) in conjunction with recent changes in weight and a simple assessment of illness severity (2 ). In many patients, however, obtaining an accurate measurement of current and previous weight to allow calculation of rate of weight loss may not be possible, so clinicians have sought a rapid, reliable laboratory method, usually involving plasma proteins, to obtain comparable information. Serum albumin is of virtually no value in assessment or monitoring of nutritional status (3 ) but is mentioned here because, surprisingly, there still remain some clinicians who use it as part of their nutritional assessment. The main factor affecting plasma albumin concentration in patients is the rate of transcapillary escape into the interstitial fluid. This transcapillary escape of albumin is markedly increased in disease [as part of the systemic inflammatory response syndrome (SIRS)], leading to decreased plasma albumin concentrations (4 ). It is inevitable that postoperative patients and patients with severe infection will have low plasma albumin concentrations. The more severe the disease, the lower the albumin, and therefore the lower the albumin, the worse the prognosis. Prealbumin, also known as transthyretin, has a half-life in plasma of 2 days, much shorter than that of albumin. Prealbumin is therefore more sensitive to changes in protein-energy status than albumin, and its concentration closely reflects recent dietary intake rather than overall nutritional status (5 ). Because of this short half-life, however, the concentration of prealbumin falls rapidly as a result of the fall in its synthetic rate when there is a reprioritization of synthesis toward acute-phase proteins such as C-reactive protein (CRP), fibrinogen, or 1-acid glycoprotein. Moreover, prealbumin concentration in plasma, like that of albumin, is affected by changes in transcapillary escape. Hence, interpretation of plasma prealbumin is difficult in patients with infections, inflammation, or recent trauma (4 ). Despite this difficulty, interest in prealbumin as a potential marker of nutritional status in certain groups of patients led to the First International Congress on Transthyretin in Health and Disease in 2002 (6 ). Some studies have screened patients on the basis of their prealbumin on admission, with values 100 mg/L being regarded as indicating severe risk of protein-energy malnutrition, 100–170 mg/L moderate risk, and 170 mg/L no risk. This type of classification, however, may often reflect severity of illness and the magnitude of the SIRS rather than nutritional status. When screening protocols that use prealbumin have been compared with a 2-stage process involving a screening questionnaire followed by an assessment by a professional dietitian, the prealbumin protocols identified many more patients considered to be malnourished (7, 8). The authors have tended to interpret this finding as showing the increased sensitivity of prealbumin in detecting malnutrition, rather than the lack of specificity of this test. Nonetheless, these results do suggest a place for prealbumin measurement soon after admission. In this issue of Clinical Chemistry, Devoto et al. (9 ) report their investigation of the concordance of prealbumin measurement, made on day 3 after admission, with a Detailed Nutritional Assessment (DNA) as a reference method to detect protein-energy malnutrition. Intriguingly, they found excellent correlation of prealbumin with the DNA, in patients with and without increased CRP ( 5 mg/L). Devoto et al. (9 ) interpret this correlation as indicating that prealbumin is a good screening tool for malnutrition, in both the presence and absence of SIRS. Closer examination raises some concerns, however. First, the DNA score is not affected only by nutritional status—it contains variables affected both by nutritional status and by inflammation. Low albumin and low cholesterol, both of which are influenced by SIRS, may account for up to 50% of DNA scores classified as “malnourished”. Thus, in the group with increased CRP, it is not surprising that there is good concordance between prealbumin and DNA. Similarly, because the DNA does contain true nutritional indicators such as low nutritional intake or weight loss that lead to low prealbumin, it can be expected that in patients without increased CRP, there would also be good concordance between DNA and prealbumin concentration. Moreover, nearly half the patients in their study either had undergone trauma or had an infection, so their CRP was probably stabilizing or decreasing on treatment during the 3 days before prealbumin was measured. As noted below, an intake of as little as 66% of the nutritional 1 Nonstandard abbreviations: LOS, length of stay; SIRS, systemic inflammatory response syndrome; CRP, C-reactive protein; DNA, Detailed Nutritional Assessment; ICU, intensive care unit. Editorial