Alzheimers disease and psychiatric polygenic risk are disease-specific and independent of plasma biomarkers in a real-world young-onset dementia cohort.

Background: Young onset dementia is frequently misdiagnosed, particularly where neurodegenerative and primary psychiatric disorders (PPD) overlap. Plasma biomarkers measure current pathology, whereas polygenic risk scores (PRS) capture lifelong genetic risk. Whether PRS derived from late-onset and population samples transfer to young-onset presentations, and how they relate to plasma biomarkers, is unknown. Methods: We studied 687 participants of European ancestry from the Markers in Neuropsychiatric Disorders (MiND) Study, a prospective young-onset neuropsychiatry cohort: 195 healthy controls (HC), 283 PPD, 55 Alzheimers disease (AD) and 154 non-AD neurodegenerative disease (oNDG). PRS for AD, a cross-disorder psychiatric factor (PSY), frontotemporal dementia, Lewy body dementia and Parkinsons disease were computed with PRS CS, excluding the APOE locus. Plasma NfL, GFAP and p-tau217 were available for 332 participants. Models were adjusted for age, sex and APOE e4 dose. Results: Relative to HC, the AD PRS was associated with AD (OR 1.69 per SD, 95% CI 1.17-2.45) but not PPD (0.95, 0.77-1.16). The PSY PRS showed the reciprocal pattern (PPD 1.29, 1.05-1.58; AD 1.00, 0.72-1.40). No PRS was associated with any plasma protein. In distinguishing PPD from oNDG, NfL was the strongest discriminator (OR 2.15, 1.62-2.83). The AD PRS was also associated with oNDG (1.38, 1.11-1.71), without improving discrimination (AUC 0.764 for NfL alone versus 0.773 combined). Conclusion: In young onset presentations, AD and psychiatric polygenic risk are disease-specific and independent of plasma biomarkers of neurodegeneration. Their combined diagnostic value requires evaluation in larger, longitudinal pre-clinical cohorts.