Latent structure model evaluation of syndrome elements and features in 826 hypertriglyceridemia patients

Objective To evaluate main Chinese medical syndrome elements and features of hypertriglyceridemia patients. Methods Using latent structure model (LSM) method, the latent structure diagram of 826 hypertriglyceridemia patients were established. Hypertriglyceridemia syndrome elements and features were interpreted by using latent probability,conditional probability,mutual information, and cumulative information coverage to quantify symptoms/syndromes data,as well as using manual interpretation methods. Results The accumulative information coverage rate reaching 95% was taken as the judgment standard for major syndrome elements. In the 826 hypertriglyceridemia patients, moderate and severe symptoms/syndromes (with the latent probability being 35% and 60% respectively) were dominant. The syndrome elements mainly included qi deficiency,qi stagnation,fire heat,stasis blood,yin deficiency,and yang deficiency. The main targets were dominated in Xin, Gan, and Shen. Conclusion LSM based syndrome element evaluation method could quantify the association degree of each variable (syndrome element; Chinese medical symptoms) and the occurrence probability.