Possible cross‑talk between sarcopenia and obstructive sleep apnea revealed by WGCNA analysis and machine learning.

Abstract

Background: Emerging evidence suggests a potential link between obstructive sleep apnea (OSA) and sarcopenia. OSA-induced hypoxia and sleep disturbances may promote the loss of muscle mass, thereby increasing the risk of sarcopenia. This study employed bioinformatics approaches to identify key diagnostic candidate genes underlying the interaction between OSA and sarcopenia.

Methods: In this study, gene expression datasets for OSA and sarcopenia were retrieved from the GEO. WGCNA was employed to identify shared common genes. Subsequently, two machine learning algorithms, SVM-RFE and RF, were utilized to determine candidate hub genes for diagnosing OSA combined with sarcopenia. The diagnostic performance of candidate hub genes was evaluated in both the discovery and validation datasets using a nomogram and ROC curve analysis.

Results: Our analysis identified 20 shared common genes associated with the comorbidity of sarcopenia and OSA. From this gene set, LGR6 and ARRDC2 were identified as candidate hub genes. The developed diagnostic model demonstrated robust predictive performance in both the discovery and validation datasets.

Conclusion: Our findings identify LGR6 and ARRDC2 as potential biomarkers for diagnosing comorbid OSA and sarcopenia. Moreover, a diagnostic model based on these two hub genes can predict the risk of sarcopenia in patients with OSA.