Objectives: To determine the relationship between prealbumin, nutrient intake, and indicators of inflammation for recuperative and rehabilitative care patients.
Design: Prospective cohort.
Setting: Recuperative Care Unit within a Veterans Administration Nursing Home Care Unit.
Participants: One hundred eleven men (100 white; mean age 80, range 64-93).
Measurements: Prealbumin and seven markers of inflammation (C-reactive protein (CRP), tumor necrosis factor, and interleukin-6 (IL-6) and their soluble receptors) were measured at admission and discharge (median length of stay 23 days, interquartile range 15-40 days). Detailed calorie counts were performed daily, and intake was expressed as a percentage of estimated requirements for protein (1.5 g/kg body weight per day) and energy (Harris-Benedict equation). The study objective was examined using least-squares regression analysis.
Results: Discharge prealbumin and the change in prealbumin were positively correlated with protein and energy intake and inversely correlated with markers of inflammation, particularly CRP and IL-6. When all covariates were included in a multivariable regression analysis, the markers of inflammation predominantly accounted for the variance in prealbumin change (56%), whereas discharge protein intake accounted for 6%.
Conclusion: For older recuperative care patients, prealbumin and its change during hospitalization are positively associated with protein intake, but inflammation or changes in inflammation appear to exert a much more-powerful influence on prealbumin concentration. Given the potential confounding effects of inflammation, monitoring the change in prealbumin is not an adequate substitute for a more-detailed nutritional assessment in this population.