Plasma lipidomic biomarker analysis reveals distinct lipid changes in vascular dementia

• DOI: 10.1016/j.csbj.2020.06.001
• Publication Date: 2020-06-09T06:47:46Z
• Authors: Yue Liu; Daniel K.Y. Chan; Anbupalam Thalamuthu; Wei Wen; Jiyang Jiang; Matthew Paradise; Teresa Lee; John Crawford; Matthew Wai Kin Wong; Ying Hua Xu; Anne Poljak; Russell Pickford; Perminder S. Sachdev; Nady Braidy
• Keywords: Neurocognitive; Vascular dementia; Lipid Profile; Univariate analysis

Vascular dementia (VaD) is a complex neurocognitive disorder secondary to variety of cerebrovascular lesions. Numerous studies have shown that lipid metabolism involved in the pathobiology disease. We examined plasma profiles VaD, with expectation identifying reliable biomarkers for VaD. 49 VaD patients and 48 healthy controls were recruited from Bankstown-Lidcombe Hospital Sydney, Australia. Lipids extracted by single phase 1-butanol/methanol, untargeted analysis was performed liquid chromatography coupled-mass spectrometry (LC-MS/MS). Univariate variance used examine differences classes individual lipids between control groups. In an independent sample 161 subjects Older Australian Twins Study (OATS), elastic net penalization generalized linear model (Glmnet) Random Forest applied levels subcategorise into vascular cognitive impairment controls. Most belonging ceramides (Cer), cholesterol esters (ChE) phospholipids significantly lower plasma, while glycerides elevated compared Levels ChE, Cer two together achieved best accuracy discriminating controls, more than 80% accuracy. The probable group OATS predicted showed greater most domains, especially attention processing speed executive function but did not differ white matter hyperintensities DTI measures. As conclusion, levels, particular are abnormal may help discriminate them Understanding basis these provide insights