Plasma Protein Panels for Mild Cognitive Impairment Among Elderly Chinese Individuals with Different Educational...

To explore plasma protein panels as potential biomarkers to screen for mild cognitive impairment (MCI) among elderly Chinese individuals with different educational backgrounds. Forty-four illiterate, 36 lower education (1-6 years), and 55 higher education (7 years or more) elderly individuals were included in the present study. Among all subjects, 67 were healthy individuals and 68 were diagnosed with MCI. Fifty plasma proteins in blood samples collected from these subjects were analyzed via the Luminex assay. Binary logistic regression was utilized to explore diagnostic models for MCI among the three educational subgroups. Then, receiver operating characteristic (ROC) curves were conducted for the clinical validity of the MCI models. Among the analyzed proteins, clusterin was used in the model of MCI among the total sample with a sensitivity (se) of 67.6%, a specificity (sp) of 59.7%, and a classification rate of 63.68%. The MCI model for the illiterate group included cystatin C, plasminogen activator inhibitor-1, and apolipoprotein A-I (se: 71.4%, sp.: 82.6%, accuracy: 77.25%). The sensitivity, specificity, and classification rate of the diagnostic model of MCI in elderly adults with lower education (human serum albumin) were each 75.0%. Additionally, the sensitivity, specificity, and accuracy rate of the diagnostic model for MCI elderly individuals with higher education (alpha-acid glycoprotein + soluble intercellular adhesion molecule-1 + pancreatic polypeptide) were 77.8%, 89.3%, and 83.60%, respectively. The performance of diagnostic models for MCI based on different educational levels is superior to that of diagnostic models for MCI without grouping by educational level.