Lipidomic markers for the prediction of progression from mild cognitive impairment to Alzheimer's disease

Abstract Dementia is a well‐known syndrome and Alzheimer's disease (AD) is the main cause of dementia. Lipids play a key role in the pathogenesis of AD, however, the prediction value of serum lipidomics on AD remains unclear. This study aims to construct a lipid score system to predict the risk of progression from mild cognitive impairment (MCI) to AD. First, we used the least absolute shrinkage and selection operator (LASSO) Cox regression model to select the lipids that can signify the progression from MCI to AD based on 310 older adults with MCI. Then we constructed a lipid score based on 14 single lipids using Cox regression and estimated the association between the lipid score and progression from MCI to AD. The prevalence of AD in the low‐, intermediate‐ and high‐score groups was 42.3%, 59.8%, and 79.8%, respectively. The participants in the intermediate‐ and high‐score group had a 1.65‐fold (95% CI 1.10 to 2.47) and 3.55‐fold (95% CI 2.40 to 5.26) higher risk of AD, respectively, as compared to those with low lipid scores. The lipid score showed moderate prediction efficacy (c‐statistics > 0.72). These results suggested that the score system based on serum lipidomics is useful for the prediction of progression from MCI to AD.