Am J Geriatr Psychiatry. 2025 Apr 22:S1064-7481(25)00329-X. doi: 10.1016/j.jagp.2025.04.210. Online ahead of print.
ABSTRACT
OBJECTIVES: To identify the distinct delusional subtypes in drug-naïve, amyloid PET-positive Alzheimer’s disease (AD) and analyze their interrelationships via network analysis.
DESIGN: Cross-sectional observational study.
SETTING: A dementia clinic at Soonchunhyang University Cheonan Hospital in South Korea, which maintains a comprehensive dementia registry.
PARTICIPANTS: One hundred two patients with mild-to-moderate, amyloid PET-positive AD who exhibited delusions on the Korean Neuropsychiatric Inventory (K-NPI).
MEASUREMENTS: Delusional subtypes were defined using the K-NPI; global cognitive function was measured by the Korean Mini-Mental State Examination (K-MMSE). Network analysis examined central (hub) and bridging symptoms.
RESULTS: Theft delusion was the most frequent subtype (89.2%), followed by reduplicative paramnesia (46.1%). Network analysis identified reduplicative paramnesia as the most central delusion, strongly linked to others, while theft delusion also emerged as a central node. Infidelity delusion was peripheral and negatively correlated with theft delusion, suggesting distinct etiological pathways. No significant associations were found between any subtype and K-MMSE scores. Stability metrics supported the robustness of these interconnections.
CONCLUSIONS: In amyloid PET-positive, drug-naïve AD, paranoid and spatial misidentification themes-particularly reduplicative paramnesia-appear highly interconnected and may share underlying pathophysiological processes. Targeting core misinterpretation and paranoia (e.g., via consistent environmental cues or focused cognitive strategies) could potentially alleviate multiple delusional themes. In contrast, jealousy-driven beliefs may require more interpersonal or relational interventions. Further longitudinal research is needed to clarify how these networks evolve with advancing disease and whether core nodes shift as global cognition declines.
PMID:40340196 | DOI:10.1016/j.jagp.2025.04.210
AI-Assisted Evidence Search
Share Evidence Blueprint
Search Google Scholar