Employing Machine Learning Algorithms to Quantify the Enhancement of Sarcopenic Skeletal Muscle Preservation through a Hybrid Exercise Program (Preprint)

Abstract

Sarcopenia is characterized by the loss of skeletal muscle mass and muscle function with increasing age. The skeletal muscle mass of the elderly who suffer from sarcopenia may be improved via the practice of strength training and Tai Chi. However, it remains unclear if the hybridization of strength exercise training (SET) and traditional Chinese exercise (TCE) will have a better effect.

Therefore, we designed a strength training and Tai Chi exercise hybrid program to improve sarcopenia in the elderly. Moreover, Explainable Artificial Intelligence was utilized to predict post-intervention sarcopenic status and quantify the feature contribution.

To assess the influence of sarcopenia in the elder group, ninety-three participated as experimental subjects in 24-week randomized controlled trial and were randomized into three intervention groups, namely the Tai Chi exercise and strength training hybrid group (TCSG, n = 33), the strength training group (STG, n = 30), and the control group (CG, n = 30). Abdominal computed tomography (CT) was employed to evaluate the skeletal muscle mass at the third lumbar (L3) vertebra. Analysis of demographic characteristics of participants at baseline used one-way ANOVA and chi-square test and repeated measures ANOVA to analyze experimental data. In addition, ten machine-learning classification models were resorted to calculate if these participants could reverse the degree of sarcopenia after the intervention.

A significant interaction effect was found in skeletal muscle density at L3 (L3 SMD), skeletal muscle area at L3 (L3 SMA), grip strength, muscle fat infiltration (MFI), and relative skeletal muscle mass index (RSMI). Grip strength, RMSI, and L3 SMA were significantly improved after the intervention for subjects in the TCSG and STG. After post hoc tests, we found that participants in the TCSG had a better effect on L3 SMA than those in the STG and participants in the CG. The LightGBM classification model had the greatest performance in accuracy (88.4%), recall score (74.0%) and F1 score (76.1%).

Skeletal muscle area of sarcopenic older adults may be improved by a hybrid exercise program composing of strength training and Tai Chi. In addition, we identified that the LightGBM classification model had the best performance to predict the reversion of sarcopenia.