A total of 20 CI patients and 20 healthy controls (HC) were enrolled. Fecal microbiota was profiled using 16S rRNA gene high-throughput sequencing. Serum metabolomics and proteomics were analyzed using ultra-high-performance liquid chromatography–tandem mass spectrometry (UPLC–MS/MS) and data-independent acquisition (DIA) proteomics, respectively. Spearman correlation and multi-omics integration were applied to explore the associations among microbiota, metabolites, proteins, and clinical indicators.