Abstract:Cerebrospinal fluid (CSF) biomarkers are essential for the accurate diagnosis of Alzheimer’s disease (AD), yet their measurement levels vary widely across centers and regions, leaving no uniform cutoff values to date. Diagnostic cutoff values of CSF biomarkers for AD are lacking for the Chinese population. As a member of the Alzheimer’s Association Quality Control program for CSF biomarkers, we aimed to establish diagnostic models based on CSF biomarkers and risk factors for AD in a Chinese cohort. A total of 64 AD dementia patients and 105 age- and sex-matched cognitively normal (CN) controls from the Chongqing Ageing & Dementia Study cohort were included. CSF Aβ42, P-tau181, and T-tau levels were measured by ELISA. Combined biomarker models and integrative models with demographic characteristics were established by logistic regression. The cutoff values to distinguish AD from CN were 933 pg/mL for Aβ42, 48.7 pg/mL for P-tau181 and 313 pg/mL for T-tau. The AN model, including Aβ42 and T-tau, had a higher diagnostic accuracy of 89.9%. Integrating age and APOE ε4 status to AN model (the ANA’E model) increased the diagnostic accuracy to 90.5% and improved the model performance. This study established cutoff values of CSF biomarkers and optimal combined models for AD diagnosis in a Chinese cohort.

Establishment of combined diagnostic models of Alzheimer’s disease in a Chinese cohort: the Chongqing Ageing & Dementia Study (CADS)