Precise Detection of Cataracts with Specific High-Risk Factors by Layered Binary Co-Ionizers Assisted Aqueous Humor Metabolic Analysis

Abstract

Diabetes and high myopia as well-known high-risk factors can aggravate cataracts, yet clinical coping strategy remains a bottleneck. Metabolic analysis tends to be powerful for precisely detection and mechanism exploration since most of diseases including cataracts are accompanied by metabolic disorder. Herein, a layered binary co-ionizers assisted aqueous humor metabolic analysis tool is proposed for potentially etiological typing and detection of cataracts, including age-related cataracts (ARC), cataracts with diabetes mellitus (CDM), and cataracts with high myopia (CHM). Startlingly, taking advantage of the optimal machine learning algorithm and all metabolic fingerprints, 100% of accuracy, precision, and recall rates are achieved for arbitrary comparison between groups. Moreover, 11, 9, and 7 key metabolites with explicit identities are confirmed as markers of discriminating CDM from ARC, CHM from ARC, and CDM from CHM, and the corresponding area under the curve values of validation cohorts are 0.985, 1.000, and 1.000. Finally, the critical impact of diabetes/high myopia on cataracts is revealed by excavating the change levels and metabolic pathways of key metabolites. This work updates the insights of prevention and treatment about cataracts at metabolic level and throws out huge surprises and progresses metabolic diagnosis toward a reality.