Serum neurofilament light chain in hydrocephalus and surgical controls: baseline comparison and 24-hour perioperative dynamics under general anesthesia

Background/Objectives neurofilament light chain (NfL) is a blood-accessible marker of neuroaxonal injury. We assessed the clinical utility of serum NfL in adults with idiopathic normal-pressure hydrocephalus (iNPH). Methods in a single-center prospective cohort (registered at ClinicalTrials.gov; partial overlap with a prior registry detailed in Supplementary Table S1), we enrolled 41 adults aged ≥ 50 years. The hydrocephalus arm included only anticipated ventriculoperitoneal (VP) shunt responders (Group A); non-hydrocephalus surgical patients served as controls (Group B). The primary analysis compared log-transformed serum NfL between groups using Welch’s t-test. A prespecified paired subset was sampled before premedication (baseline) and 24 h after surgery start. As a sensitivity analysis, we fitted age-aware linear models for log(NfL) with group as the exposure and age and sex as covariates. Results serum NfL was right-skewed; analyses used log-transformed values. The primary between-group contrast showed no difference (geometric mean ratio [GMR] 1.03; 95% CI 0.51–2.10; p = 0.93). In age-adjusted models (log[NfL] ~ group + age + sex), the between-group difference remained non-significant with an effect magnitude comparable to the unadjusted GMR. In paired analyses, serum NfL did not change from baseline to 24 h overall (24 h vs. baseline GMR 1.13; 95% CI 0.89–1.45; p = 0.305) and remained non-significant within both Group A (n = 10) and Group B (n = 11). Conclusions in adults ≥ 50 years, serum NfL neither distinguished anticipated VP-shunt responders with iNPH from surgical controls nor exhibited an acute change at 24 h after surgery start under general anesthesia. These findings suggest limited standalone diagnostic/prognostic utility of serum NfL in this setting. Interpretation should be age-aware and integrated with complementary clinical data and biomarkers; future studies should use standardized sampling and multimodal panels to refine risk stratification.