Abstract
Frozen shoulder (FS) and osteoporosis (OP) are common age-related degenerative diseases, occurring more frequently in females, which suggests potential molecular links between them. This study aimed to identify shared genetic features and pathways of OP and FS using bioinformatics and machine learning approaches. Gene expression data for OP and FS were obtained from the Gene Expression Omnibus database. Common differentially expressed genes (DEGs) were identified. Functional enrichment analysis, protein-protein interaction (PPI) networks construction, and machine learning algorithms were applied to screen key genes. Diagnostic value was evaluated by receiver operating characteristic (ROC) curve analysis. Immune infiltration and regulatory networks involving transcription factors and miRNAs were explored. Potential therapeutic compounds were also predicted. A total of 111 common DEGs were identified, enriched in pathways related to neurological development, cellular signaling, and immune regulation. PPI analysis revealed 14 hub genes, with SDC1 and ELN identified as key diagnostic markers by machine learning. ROC curves confirmed their diagnostic efficacy for both OP and FS. Immune infiltration analysis revealed distinct immune cell patterns in OP, correlating with the expression of key gene. Regulatory network analysis demonstrated complex transcriptional regulation of SDC1 and ELN. Drug prediction identified five candidate small molecules targeting these genes. This study uncovered shared genetic features of FS and OP through comprehensive bioinformatics analysis, enhancing understanding of their co-morbidity mechanisms. These findings provide a theoretical basis for identifying novel diagnostic biomarkers and therapeutic targets, facilitating the development of precise diagnostic strategies for OP with FS.