基于残差拆分思想的中医“证”与临床指标关联关系再分析的数学模型的建立及应用

A mathematical model for re-analysis of the relationship between essence of syndromes in traditional Chinese medicine and clinical biochemical indicators based on the residual-split method and its application

中医证实质研究是中医药基础理论研究的热点和难点,其中探讨中医证实质与生化指标关联关系的临床研究是重要方面之一。然而,由于各临床研究选取的生化指标与特定疾病关系密切,而传统的数据处理方法无法有效评价疾病因素的影响,故所得结论会随着研究对象疾病类型的不同而变化,出现"同证异理"现象,并成为中医证实质研究的主要困惑之一。本研究基于信息学理论,从指标变化量的残差拆分入手,通过定量化计算中医的"证"概念于病理过程中提取的不同于"疾病"概念的新信息在指标变化上的表达,并结合其是否存在的显著性检验,建立一个能够客观分析影响临床指标变化的独立的证信息成分的大小及统计学意义的数学模型。将其应用于阳虚证环核苷酸研究文献数据分析的结果显示:第一,消除疾病因素后,对于环腺苷酸(cyclic adenosine monophosphate,cAMP),阳虚证信息成分量基本为负值,而对于环鸟苷酸(cyclic guanosine monophosphate,cGMP),阳虚证信息成分量基本为正值,在所有纳入文献中表现一致,提示阳虚证患者组出现较为明显的cAMP降低趋势和cGMP升高趋势,与国内相关研究结论相符;第二,对于cAMP和cGMP,所有纳入文献的阳虚证信息成分量与0值差异的统计学检验结果不尽一致,可能与原文献数据质量有关;第三,在某些情形下,疾病因素而非中医"证"因素可能是造成特定疾病阳虚证组与正常组相比有显著性差异的主要原因,在临床研究中应加以注意。本数学模型旨在为病证结合的中西医临床研究中疾病因素和"证"因素的处理提供数学工具,为中医证实质与临床指标关联关系研究提供新方法。

Studying the essence of syndromes（Zheng）in traditional Chinese medicine（TCM）is a fundamental challenge in basic theoretical research of TCM.The relationship between any given syndrome and biochemical indicators is one of the important aspects of the study.As the indexes selected in each study are specific to a particular Western medical disease diagnosis,and the disease factor is inevitably introduced into the study,the effect of disease factor on the index changes cannot be assessed effectively by traditional data processing methods.This is known as ＂the same syndrome with different reasons＂,which has resulted in confusion in TCM research.This study aimed at providing a mathematical tool to address this issue.Based on information theory and the residual-split method,the syndrome information,which was covered in the index variation,was quantitatively calculated in this paper as an independent part of the disease factor.A mathematical model capable of objectively assessing and statistically testing the effect of the syndrome factor on the index changes was established.Applying this model to literature data of studies on the relationship between cyclic nucleotides and yang-deficiency syndrome showed following results.First,the values of yang-deficiency syndrome information were negative for cyclic adenosine monophosphate（cAMP）while positive for cyclic guanosine monophosphate（cGMP）in all included literature.This indicated that the group of yang-deficiency syndrome was correlated with an obvious trend of reduced cAMP levels and increased cGMP levels.Second,the statistical test results of yang-deficiency syndrome information of the two indexes were different among the literature included.The quality of original data was considered as a possible reason.Third,the significant differences between the yang-deficiency group of a specific disease and the normal group may,in some cases,be caused by a disease factor rather than a syndrome factor.The mathematical model provided a reasonable mathematical tool for the analysis of disease factor and syndrome factor in clinical research of TCM,suggesting that the mathematical model may give rise to innovative ideas and methods in the study of syndromes.