Summarized by Masters of Longevity from Frontiers - Health.
Presents a multicenter cross-sectional machine learning model that combines Traditional Chinese Medicine constitution and clinical indicators to estimate concurrent probable mild cognitive impairment risk in older adults with possible sarcopenia.
Key Takeaways
- A seven-feature model retained education, sex, smartphone use, sleep quality, depressive symptoms, qi-deficiency, and damp-heat constitutions.
- The multilayer perceptron achieved internal AUC 0.853 and external AUC 0.799 for classifying screening-defined probable MCI.
- External validation showed modest calibration and limited transportability, indicating the need for prospective multicenter validation and recalibration.
