聚类分析对肾虚症状的探索性研究

Study on kidney-deficiency symptoms with cluster analysis

目的用聚类分析方法寻找肾虚症状的自然类群,探讨肾虚证候结构与症状之间的关联关系以及聚类分析应用的局限性。方法用流行病学的思想和原则收集大样本的肾虚症状资料,对这些资料进行科学的整理和分析,利用SPSS统计软件进行相关分析和聚类分析。结果在调查样本中,肾虚症状经过不同角度的分析,其症状群落的结构和关系与中医理论的描述基本一致,为中医诊断学中有关肾虚的症状结构提供了科学的解释。结论中医学在长期的临床积累中经验性地形成了证候与一组症状约定性的关系,但这种关系以及他们的密切程度尚缺乏流行病学基础上的系统研究。聚类方法可以将一组或一群症状自然地聚为一族或一群,从而揭示一组症状与证候之间的密切关系;然而由于聚类分析方法的局限性,对证候与症状的这种分析尚有诸多缺陷。

Objective To look for the natural people group with kidney-deficiency symptoms by cluster(analysis),and discuss the relationship between syndrome structure and symptoms of kidney-deficiency and the limitation of cluster analysis.Method To collect the large-sampled data of kidney-deficiency(according) to the principle of epidemiology,and sort out and analyse the data.The SPSS software was used in related and cluster analysis.Result The structure and relationship of symptom groups of(kidney-)deficiency were accordance with the theory of Chinese medicine after the analysis on kidney-deficiency symptoms from different aspects in the sample,which gave a scientific explanation to the structure of(kidney-)deficiency symptoms in the diagnostics of Chinese medicine.Conclusion A kind of relationship has been formed between syndrome manifestations and a group of symptoms in long time clinical practice of Chinese medicine,while this relationship and the closed degree have not been analysed on the(epidemologic) base.Cluster analysis can gather some symptoms to a natural group and explain the close relationship between the syndrome manifestations and a group of symptoms.This analysis,however,has some shortcomings because of the limitation of the cluster analysis.