IntroductionNeurodegenerative dementia (ND) is characterized by progressive cognitive decline. The role of autoantibodies in ND remains controversial, as they may contribute to neurological damage or, under certain pathophysiological conditions, exert protective effects. In this study, we aimed to analyze the clinical characteristics of patients with autoantibody-positive ND and investigate potential molecular mechanisms using proteomics.MethodsThe study included 13 patients with autoantibody-positive ND (ND +), 13 with autoantibody-negative ND (ND -), and 13 cognitively normal controls. Differentially expressed proteins (DEPs) were identified through proteomic analysis. Bioinformatics approaches were employed to explore potential pathways and mechanisms. Multiple linear regression analysis was performed to examine the effects of DEPs on cognitive function and their interaction with autoantibody status.ResultsThe ND + and ND - groups displayed similar overall cognitive status, levels of depression or anxiety symptoms, and mental and behavioral abnormalities. The ND + group demonstrated better daily living abilities and superior frontal lobe cognitive function. Proteomic analysis identified 162 DEPs associated with autoantibodies enriched in pathways related to endocytosis, endoplasmic reticulum protein processing, tight junctions, and cellular adhesion. In the ND + group, beta-arrestin-1 (ARRB1, beta = 3.31, 95% CI: 1.56-5.05) and calpain-2 (CAPN2, beta = 0.32, 95% CI: 0.09-0.55) were positively associated with Frontal Assessment Battery (FAB) scores, whereas no such associations were observed in the ND - group. Conversely, in the ND- group, UDP-glucose:glycoprotein glucosyltransferase 1 (UGGT1, beta = - 2.38, 95% CI: -4.12 to-0.64) was negatively associated with FAB, with no corresponding effect in the ND + group.ConclusionThis exploratory study provides preliminary evidence of distinct clinical and molecular signatures in autoantibody-positive neurodegeneration. ARRB1 and CAPN2 may contribute to cognitive resilience in this group, while UGGT1 appears associated with cognitive decline in autoantibody-negative cases. Owing to the limited sample size, these findings should be considered hypothesis-generating and warrant validation in larger, independent cohorts before consideration as biomarkers or therapeutic targets.