Abstract
BACKGROUND: Idiopathic normal pressure hydrocephalus (iNPH) overlaps clinically and pathologically with Alzheimer’s disease (AD), making differential diagnosis and prognostication challenging. The identification of reliable and minimally invasive plasma biomarkers would have major clinical implications for distinguishing iNPH from AD and for monitoring disease course and surgical outcomes. METHODS: We conducted three studies: (Study 1) cross‑sectional comparisons of plasma phosphorylated tau 217 (p‑Tau217) and amyloid-β oligomers (AβO) among healthy controls (HCs; n = 22), iNPH (n = 76), and AD (n = 23) groups, accompanied by receiver operating characteristic (ROC) analysis; (Study 2) correlation analyses within preoperative iNPH group to assess associations between biomarkers and clinical indices, as well as two‑group comparisons between iNPH group with and without AD pathology stratified by CSF p‑Tau181 > 30 pg/mL; and (Study 3) longitudinal pre- and postoperative assessments following shunt surgery. RESULTS: In the three‑group analyses, plasma p‑Tau217 effectively discriminated iNPH from AD (AUC = 0.79, 95% = CI 0.700–0.882), whereas discrimination between HCs and iNPH was poor (AUC = 0.48). Plasma AβO moderately discriminated HCs from iNPH (AUC = 0.71, 95% CI = 0.567–0.848). In the preoperative iNPH group, CSF and plasma p‑Tau217 showed a strong correlation (ρ = 0.736, p < 0.001), whereas CSF and plasma AβO showed no significant correlation (ρ = −0.047, p = 0.689). Among iNPH group with AD pathology (CSF p-Tau181 > 30 pg/mL), plasma p-Tau217 was higher, plasma AβO was lower, and Frontal Assessment Battery (FAB) scores were reduced compared with those without AD pathology. For detecting coexistent AD pathology in iNPH group, AUCs were modest for plasma p‑Tau217 (0.64), AβO (0.63), and p‑Tau217/AβO ratio (0.66). Postoperatively, iNPH grading scale (GS), Mini-Mental State Examination (MMSE), and FAB scores improved significantly; CSF and plasma AβO levels decreased, CSF p‑Tau217 increased, whereas plasma p‑Tau217 levels remained stable. CONCLUSIONS: Plasma p-Tau217, reflecting AD-related pathology, and plasm AβO, reflecting disease stage, provide complementary information for distinguishing iNPH from AD, assessing comorbidity of AD pathology in iNPH and evaluating postoperative changes. Their combined measurement may support differential diagnosis, aid the preoperative identification of concomitant AD pathology in iNPH, and serve as a minimally invasive screening and triage tool as well as a means of postoperative monitoring, while acknowledging that discrimination between HCs and iNPH at individual level remains limited. TRIAL REGISTRATION: Not applicable. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12987-026-00784-8.