Metabolomics and lipidomics study on serum metabolite signatures in Alzheimer's disease and mild cognitive impairment.

Alzheimer's disease (AD) and mild cognitive impairment (MCI) are major causes of dementia in the elderly, with metabolic alterations not fully understood. In this study, we quantitatively profiled serum metabolites of participants in 2 independent cohorts. Based on the data from cohort 1 (22 AD patients, 19 MCI patients and 19 cognitively normal participants (CN)), we identified 32 differential metabolites in AD and 49 in MCI serum. Notably, differential metabolites related to amino acid (AA), organic acid, fatty acid (FA), phosphatidylcholine (PC), sphingomyelin (SM) metabolism in AD and free fatty acid (FFA), acylcarnitine, PC, SM in MCI were strongly associated with cognitive level, memory, attention and execution function as evaluated by cognitive scales. Pathway analysis based on the differential metabolites revealed perturbation in pathways related to phospholipid metabolism, sphingolipid metabolism, AAs metabolism, beta oxidation of FAs, and carnitine metabolism. Using random forest (RF), support vector machine (SVM) and Boruta analysis for classification and validated by gradient boosting (GB), logistic regression (LR) and RF diagnostic model, we identified panels of 10 metabolites in AD and 13 metabolites in MCI that effectively discriminate AD and MCI individuals from CN with high accuracy, sensitivity and specificity. The diagnostic accuracy of the models was further validated in an independent cohort 2, consisting of 20 AD, 20 MCI and 20 CN individuals, with consistent results.