Unveiling CT-Scan Radiodensity Leg Asymmetry Impact on Common Aging Comorbidities

This study utilizes advanced CT scan radiodensity analysis of soft tissues, specifically, the novel nonlinear trimodal regression analysis (NTRA), to evaluate soft tissue health in the lower extremities and its association with type 2 diabetes (DM) and hypertension (HTN) in the elderly. We analyzed 3157 participants from the AGES-Reykjavik dataset using various machine learning models to assess the predictive power of NTRA-derived asymmetry indicators. The results show that Random Forest models achieved the highest accuracy, with 0.89 for DM and 0.88 for HTN. This research highlights the significance of muscle, fat, and connective tissue in aging health and suggests future studies should explore more machine learning techniques and consider lifestyle factors. Despite some limitations, such as the necessary use of oversampling techniques, our study advances the understanding of how soft tissue composition impacts elderly health, in particular finding that the fat tissue density differences between legs significantly impact the common aging comorbidities of DM and HTN.