Early detection of mild cognitive impairment due to Alzheimer’s disease using combinations of an online cognitive assessment and blood biomarkers

Abstract Background Accessible measurements to a large screening and early differentiation of mild cognitive impairment due to Alzheimer’s disease (MCI‐AD) from healthy elderly is increasing urgent with the prevalence of Alzheimer’s disease (AD) expanded. Method an online cognitive assessment (Memtrax) and MoCA were performed, on 99 clinically diagnosed cognitively normal (CON) and 101 MCI‐AD participants. The plasma levels of phosphorylated tau (p‐tau)181 and neurofilament light (NfL) and amyloid beta (Aβ)42/40 were measured using Simoa assay. The relevance between Memtrax and blood biomarkers were conducted by Spearman correlation analysis. Discrimination models were built using assorted machine learning approaches combining Memtrax with blood biomarkers, and model performance was measured using Area Under the Receiver Operating Characteristic Curve (AUC). Result The AUC of Memtrax and MoCA in differentiating MCI‐AD from CON was similar. Memtrax has a strong correlation with phosphorylated tau (p‐tau)181 and neurofilament light (NfL), while less related to amyloid beta (Aβ)42/40. Combining Memtrax and blood biomarkers, the AUC of best model reach to 0.975 (95% CI: 0.950‐0.999). Conclusion The combination of Memtrax and blood biomarkers is promising to the large screening and early detection in MCI‐AD.