A blood-based algorithm for the detection of Alzheimer's disease

Area of Science:

• Neurology
• Biomarker Discovery
• Diagnostic Accuracy

Background:

• Previous serum-based algorithm demonstrated high diagnostic accuracy for Alzheimer's disease (AD).
• Current study aimed to refine the AD diagnostic algorithm by reducing serum protein markers and incorporating clinical laboratory data.
• Investigated the association between the refined biomarker risk score and neuropsychological performance.

Purpose of the Study:

• To simplify and enhance a previously developed serum-based Alzheimer's disease diagnostic algorithm.
• To assess the diagnostic performance of a reduced panel of serum proteins combined with clinical data.
• To explore the relationship between the biomarker risk score and cognitive function in Alzheimer's disease.

Main Methods:

• Serum protein multiplex biomarker data analyzed from 197 Alzheimer's disease patients and 203 cognitively normal controls.
• A 30-protein panel, identified as most significant, was used to develop the refined algorithm.
• Clinical laboratory data and demographic information were integrated with the biomarker data.

Main Results:

• The 30-protein risk score achieved a sensitivity of 0.88, specificity of 0.82, and AUC of 0.91.
• Incorporating demographic data and clinical labs improved performance: sensitivity 0.89, specificity 0.85, AUC 0.94.
• The biomarker risk score showed the strongest correlation with neuropsychological tests assessing language and memory.

Conclusions:

• The Alzheimer's disease diagnostic algorithm can be effectively reduced to 30 serum proteins while maintaining high diagnostic accuracy.
• The refined biomarker risk score demonstrates significant associations with cognitive performance, particularly in language and memory domains.
• This optimized algorithm offers a more streamlined approach to Alzheimer's disease diagnosis and assessment.