Analysis of Factors Influencing Irisin Levels in T2DM Patients with Sarcopenia Based on Machine Learning

Background: Type 2 diabetes mellitus and sarcopenia are significant public health concerns. Irisin, a myokine involved in energy metabolism, is implicated in the pathophysiology of both conditions. Objective: This study aimed to explore the relationship between Irisin levels and type 2 diabetes mellitus with sarcopenia and to develop predictive models for early risk identification using machine learning. Methods: A total of 200 subjects were enrolled and divided into four groups: healthy controls, sarcopenia only, type 2 diabetes mellitus only, and type 2 diabetes mellitus with sarcopenia. Five machine learning algorithms were applied to evaluate the influence of Irisin levels. Model performance was assessed using accuracy, Kappa coefficient, mean absolute error, and relative absolute error. Results: The optimized decision tree model demonstrated superior performance (accuracy=99.0%). Knowledge rules revealed that individuals with Irisin levels >94ng/ml exhibited normal muscle function, while those with levels <67ng/ml had a higher risk of sarcopenia and metabolic disturbances. Analysis of variance confirmed significant differences in Irisin levels across the groups. Conclusions: Irisin is a promising biomarker for assessing the risk of type 2 diabetes mellitus with sarcopenia. The decision tree model facilitates the early identification of high-risk individuals, offering valuable insights for clinical practice. Further large-scale validation is warranted.