Inflammatory Biomarkers and Post-Intensive Care Syndrome: A Prospective Cohort Study

Background and Objective: Post-Intensive Care Syndrome is a multidimensional sequela of critical illness that includes physical, cognitive, and psychological impairments after intensive care unit discharge. Systemic inflammation has been proposed as one potential mechanism contributing to selected PICS domains, but the direction, timing, and clinical relevance of this relationship remain uncertain. This study assessed whether serial concentrations of C-reactive protein, procalcitonin, and interleukin-6 were associated with PICSQ severity in ICU survivors. Materials and Methods: This prospective single-center cohort study included 267 adult ICU patients in the primary complete case analysis. CRP, PCT, and IL-6 were measured at predefined time points during hospitalization. PICS severity was assessed using the Post-Intensive Care Syndrome Questionnaire at ICU discharge and at 2 and 3 months after discharge. The primary endpoint was total PICSQ severity at 3 months. Correlation analyses, threshold-based group comparisons, and logistic and multivariable linear regression models adjusted for age, sex, reason for ICU admission, and length of hospitalization were performed. Secondary and domain-specific analyses were considered exploratory. Results: Correlation analyses did not show a consistent association between inflammatory biomarkers and total PICSQ severity at ICU discharge, 2 months, or 3 months. Weak inverse associations were observed between selected CRP, PCT, and IL-6 measurements and cognitive domain scores, indicating lower cognitive impairment scores among patients with higher biomarker values in some analyses. These findings were not consistent across time points and should be interpreted cautiously. In unadjusted threshold-based analyses at 3 months, selected associations were observed for CRP after one week, PCT on day 4, and IL-6 on days 2 and 4, mainly in the psychological domain. In fully adjusted linear regression models, none of the selected biomarker thresholds remained statistically significant at p < 0.05. In adjusted linear regression, PCT ≥ 2 ng/mL on day 4 showed a borderline association with the total PICSQ score at 3 months, with β = 0.44, 95% CI -0.07 to 0.94, p = 0.089, N = 267. IL-6 > 10 pg/mL on day 2 showed a borderline association with psychological domain severity at 3 months, with β = 0.46, 95% CI from -0.005 to 0.92, p = 0.052. In adjusted logistic regression, IL-6 > 10 pg/mL on day 2 was associated with higher odds of psychological domain score ≥ 4 at 3 months, with adjusted OR 3.61, 95% CI 1.26 to 10.37, p = 0.017, N = 266. No consistent association was observed for the physical domain. Conclusions: In this exploratory cohort, routinely available inflammatory biomarkers were not consistently associated with global PICSQ severity across all assessment time points. Selected time-dependent and domain-specific associations, particularly involving PCT and IL 6 at 3 months, may indicate a possible link between inflammatory activation and later psychological PICS burden. These findings do not establish causality or predictive performance and require external validation before CRP, PCT, or IL 6 can be used for PICS risk stratification.