Serum Levels of Albumin and Prealbumin Do Not Correlate With Nutrient Delivery in Surgical Intensive Care Unit Patients

Background: Serum albumin and prealbumin levels, may be more strongly associated with inflammation than with nutrient delivery. Their predictive value has not been extensively described in surgical intensive care unit (ICU) patients. Methods: We analyzed a registry of adult surgical ICU patients receiving enteral nutrition. Subjects with at least 1 serum albumin, prealbumin, or C-reactive protein (CRP) level were included. Demographic, nutrition, and clinical outcome data were collected. Results: A total 252 subjects were included. A subset had serial measurements: albumin (n = 194), prealbumin (n = 13), CRP (n = 9), white blood cell (WBC) (n = 131), and neutrophil–lymphocyte ratio (NLR) (n = 86). Serum albumin level was inversely correlated with all 3 inflammatory biomarkers (CRP, ρ = −0.24, P <0.02; WBC, ρ = −0.15, P <0.001; and NLR, ρ = −0.26, P < 0.001). Change in serum albumin level was inversely correlated with change in NLR (ρ = −0.22, P = 0.044) but not with CRP or WBC. Admission serum albumin level was significantly higher in nourished vs. moderately and/or severely malnourished patients (3.2 [2.7–3.7] vs. 2.7 [2.3–3.0], P = 0.004). Admission serum prealbumin level was significantly higher in nourished vs. moderately and/or severely malnourished patients (9 [7–12] vs. 4 [3–5], P = 0.001). Serum albumin level was inversely correlated with Charlson Comorbidity Index (ρ = 0.20, P = 0.001). Calorie and/or protein delivery in the ICU was not correlated with changes in serum albumin or prealbumin levels. Conclusions: In the ICU, initial serum albumin levels and serial trends are inversely correlated with inflammation. Although initial serum albumin levels are reflective of baseline nutrition status, neither serum albumin level nor serum prealbumin level trends correlate with calorie or protein deficits and should not be used to assess adequacy of nutrition delivery.