Early change in serum prealbumin adds predictive value for Day-7 enteral nutrition response in critically ill adults: a retrospective cohort study

Background and aims Single serum prealbumin (PAB) measurements are difficult to interpret during critical illness because they are strongly influenced by inflammation and disease severity. We investigated whether early changes in PAB after enteral nutrition (EN) initiation were associated with an operationally defined Day 7 EN response and added predictive value to that of routinely available clinical variables. Methods This single-center retrospective cohort study screened 341 critically ill adults. The primary cohort comprised 263 patients with PAB measured at baseline, Day 3, and Day 7; an expanded sensitivity cohort comprised 271 patients with baseline and Day 3 PAB and ascertainable Day 7 EN outcomes, irrespective of Day 7 PAB availability. Day 7 EN response required energy target achievement ≥70%, protein delivery ≥1.0 g/kg/day, no gastrointestinal (GI)-related EN interruption >24 h, and no supplemental parenteral nutrition (PN) before assessment. We used multivariable logistic regression analyses, component-specific analyses, receiver operating characteristic (ROC) curve analysis, calibration, decision-curve analysis, restricted cubic splines, and 2,000-resample bootstrap validation. Results In the primary cohort, 154 of the 263 patients (58.6%) were responders. Each 10 mg/L increase in Day-3 ΔPAB was independently associated with higher odds of response (adjusted odds ratio 2.67, 95% confidence interval [CI] 1.89–3.76; p < 0.001). Adding Day 3 ΔPAB to the clinical model increased the area under the curve (AUC) from 0.804 to 0.869 (ΔAUC 0.065, bootstrap 95% confidence interval 0.031–0.107; DeLong p < 0.001). In the expanded cohort, the corresponding adjusted odds ratio was 2.34 (1.72–3.19), and the AUC increased from 0.783 to 0.849 (ΔAUC 0.066; DeLong p = 0.001). Component analyses showed associations with energy and protein target attainment, whereas estimates for prolonged gastrointestinal interruption and supplemental parenteral nutrition were imprecise because these events were uncommon. Conclusion An early increase in PAB was independently associated with Day 7 EN response and improved prediction beyond routine clinical, inflammatory, and nutritional variables. Day 3 ΔPAB should be interpreted as an adjunctive dynamic signal rather than a stand-alone marker of nutritional status, and external validation is required.