A Meta-Analysis of the Association between Diabetes Mellitus and Traditional Chinese Medicine Constitution

Objective To explore the distribution of constitution types of diabetes mellitus (DM) in traditional Chinese medicine (TCM) and to provide evidence-based medicine basis for the prevention and treatment of diabetes. Methods PubMed, Embase, Web of Science, and three Chinese databases were searched to include research literature on the relationship between diabetes and TCM constitution. The single rate study of cross-sectional literature was conducted with RStudio software, and the control meta-analysis of the diabetic and nondiabetic population was performed with Review Manager 5.3 software. Two independent reviewers assessed the methodological quality of the studies' data. The main outcomes included the distribution of constitutional types in the diabetic population and the odds ratio (OR) between the two. Effect sizes are expressed as proportions or ORs with 95% confidence intervals (CI). Results A total of 28,781 diabetic cases were included in 87 articles. Yin-deficiency, phlegm-dampness, and qi-deficiency accounted for 18% (95% CI (15%, 20%), P < 0.01), 17% (95% CI (15%, 19%), P < 0.01), and 13% (95% CI (11%, 15%), P < 0.01) of the total diabetic cases. The risk of diabetes in people with yin-deficiency and phlegm-dampness was 3.06 (95% CI (1.38–6.78), P=0.006) and 1.89 (95%CI (1.05–3.42), P=0.03) times higher than that in those with other constitutions, respectively. The distribution of TCM constitution of DM patients varied significantly in different regions and ages. Conclusion Yin-deficiency and phlegm-dampness are the common constitution types of diabetic people, and they may also be the risk factors of diabetes. Balanced constitution may be a protective factor of diabetes. More high-quality cohort and case-control studies need to be designed to provide more valuable evidence-based basis for assessing the correlation between DM and TCM constitution.


Introduction
Diabetes mellitus is a common chronic metabolic disease. With the development of society and the change of people's lifestyles, the incidence of DM has increased year by year, and it has become the most harmful noncommunicable disease on human health after tumor and cardiovascular disease [1]. According to the data released by the International T2DM Federation in 2013, there were 387 million diabetes patients in the world in 2013. China has the largest number of T2DM patients in the world, with 98 million T2DM patients, which is expected to increase to 143 million by 2035 [2]. e high incidence of chronic diseases and the rapid increase in medical costs highlight the importance of public health. In the face of the current high prevalence of diabetes, it is particularly important to carry out effective prevention and treatment measures. Traditional Chinese medicine (TCM) has been used for thousands of years of history as a form of health management, and this health care system has been widely used in many countries of the world [3]. TCM regards that everyone is different from others, so there is no recognized diagnosis and treatment model in the TCM treatment of diabetes, but more emphasis on individual treatment. Different constitution types make individuals suffer from different diseases. Individualized prevention and treatment measures for different types of constitution personalized treatment, so as to achieve a better effect than conventional treatment. e constitution of traditional Chinese medicine is the grasp of "preventive treatment of the disease" [4]. In April 2009, the Constitution in Chinese Medicine Questionnaire (CCMQ) was published and recommended by the China Association of Chinese Medicine as a standard for constitution measurement because of its good reliability and validity [5]. According to the CCMQ, the body constitutions of Chinese population are divided into nine basic types, namely balanced, qi-deficiency, yang-deficiency, yin-deficiency, phlegm-dampness, dampness-heat, blood stasis, qi stagnation, and inherited special. Among them, the balanced constitution is normal, while the other eight are biased and prone to relevant diseases [6]. TCM constitution identification has been widely used in nondisease treatment centers and community health service centers all over the country because of its simple operation, easy to understand, and easy to accept. In recent years, a large number of clinical studies on the relationship between diabetes and TCM constitution have been carried out in various regions, which provides data support for analyzing the distribution law of the TCM constitution of diabetic patients. is study aims to obtain a larger sample of data on the distribution of TCM constitution types of diabetic patients, to determine the types of constitution prone to diabetes, and to provide evidence-based medical evidence for the effective prevention and treatment of diabetes for "preventive treatment of disease." is was achieved by a meta-analysis of the current clinical research literature on the TCM constitution of diabetes.

Retrieval Strategy.
Clinical studies on the correlation between body constitution and diabetes were searched in PubMed, Embase, Web of Science, China National Knowledge Infrastructure (CNKI), VIP Database (VIP), and Wanfang Database from April 2009 (the CCMQ's publication time) to March 31, 2021. e search terms included "diabetes mellitus," "tangniaobing" (Chinese pinyin of diabetes), "tizhi" (Chinese pinyin of constitution), and "constitution." No language, nationality, or publication restrictions were applied. As an example of one specific strategy, the search terms for PubMed were as follows: (diabetes mellitus [Title/Abstract]) AND constitution [Title/ Abstract]).

Inclusion/Exclusion Criteria.
All clinical literature on the relationship between diabetes and TCM constitution types (including cross-sectional studies, case-control studies, and cohort studies) were included, without limiting the type of research and the form of publication. e details are as follows: (1) Research object: the research object is the diabetic patients with a definite diagnosis; (2) Research tool: the measurement tool used for the constitution identification of the research object is the standard of Classification and Judgment of TCM Constitution issued by the China Institute of TCM in 2009; and (3) the sample size of the study is clear, and the data of the constitution composition are complete.
Exclusion criteria: (1) literature lacking in basic information reports or without statistics of constitution composition; (2) included subjects combined with other systematic serious diseases that may affect their TCM constitution types; (3) the study is limited to a certain kind of constitution population, such as simple phlegm-dampness constitution research; and (4) repeatedly published literature.

Data Extraction and Quality Evaluation on Methodology.
Make a data extraction form, including the literature title, the type of research design, the area of the study, the sex of the object of study, the average age, the number of people of each type of constitution, the total sample size, and so on.
e Newcastle-Woodward scale (NOS) [7] was used to assess the methodological quality of the cohort study and case-control study. e scale was compared with 8 items from three aspects: the selection of the study population, the comparability between groups, and the measurement of exposure factors, with a total score of 9 points. e crosssectional study adopted the standard (referred to as AHRQ standard) recommended by the Agency for Healthcare Research and Quality (AHRQ) [8]. It is divided into 11 items, including data sources, inclusion criteria, observation time, continuity of research objects, subjective factors of evaluators, quality control, and so on, with a total score of 11 points.

Data Analysis.
e meta-analysis of the individual rate of each constitution type was carried out by using RStudio-1.1.463 software [9]. Analysis models were chosen by the results of heterogeneity tests. e percentage of 8 different biased constitution types of diabetic people and their 95% confidence interval (CI) were calculated. When the heterogeneity was too large, subgroup analysis was carried out according to regional and age factors, and the inverted funnel chart was used to analyze the degree of publication bias assessed by funnel plot analysis. e Review Manager 5.3 software provided by the Cochrane collaboration network was used to analyze the data synthesis between the diabetic group and nondiabetic group, and the odds ratio (OR) and its 95% CI were used to describe the effect value of a single study. e effect model was selected according to heterogeneity: the fixed effect model was used when heterogeneity was less than 50%, and the random effect model was used when heterogeneity was greater (I 2 > 50%).

Retrieval Process and Results
. Duplicate references were excluded. All titles and abstracts were then screened to eliminate duplicates and obviously irrelevant citations. Following screening of 1240 citations found in the literature search, 93 papers potentially met the inclusion criteria and were examined in full text. 87 studies were included in the final review. e literature screening process and results are shown in Figure 1.

Basic Characteristics and Quality Evaluation of the Inclusion Study.
Among all the studies we included, 1 was a Chinese-Malaysian constitution fitness study, and 86 were domestic clinical studies, including 22 provinces, autonomous regions, and municipalities directly under the central government. e study was published in 2009 at the earliest, and the number of studies is increasing year by year. A total of 28,581 patients with diabetes were included in each study, with an average sample size of 329. According to the NOS (9 items) and AHRQ cross-sectional study evaluation criteria (11 items), 7 case-control studies and 80 cross-sectional studies were evaluated. e scores of cross-sectional studies were all 3-7, which were medium-and low-quality literatures, which did not have high research quality and had a high risk of bias. See Table 1 for details.

Meta-Analysis of Biased Constitution Distribution in TCM.
According to the number of diabetic patients reported in each study, meta-analysis was carried out. A total of 87 studies reported these data, with a total sample size of 28,581 cases. e results showed that the distribution of 8 biased constitution types in patients with diabetes was yindeficiency, phlegm-dampness, and qi-deficiency in the highest proportion, and the results of meta-analysis were shown in a forest map. e proportion of the other five types of constitution is less than 10%. e results of the metaanalysis are described in a table.

Yin-Deficiency
Constitution. 86 studies involving 28,342 cases reported the proportion of yin-deficiency constitution in DM patient population. e random effects model was adopted due to the great heterogeneity of included studies. e results showed that the proportion of yin-deficiency constitution in DM patients was 18% (95% CI (15%, 20%), P < 0.01) ( Figure 2).

Phlegm-Dampness
Constitution. 86 studies involving 28,342 cases reported the proportion of phlegm-dampness constitution in DM patient population. e random effects model was adopted due to the great heterogeneity of included studies. e results showed that the proportion of phlegm-dampness constitution in DM patients was 17% (95% CI (15%, 19%), P < 0.01) (Figure 3).

Qi-Deficiency
Constitution. 84 studies involving 27,608 cases reported the proportion of qi-deficiency constitution in DM patient population. e random effects model was adopted due to the great heterogeneity of included studies. e results showed that the proportion of qideficiency constitution in DM patients was 13% (95% CI (11%, 15%), P < 0.01) (Figure 4).

Other TCM Constitutions.
e distribution proportion of the other five biased constitution in DM patients is less than 10%, and the order from high to low is yangdeficiency, dampness-heat, blood stasis, qi stagnation, and inherited special constitution.
e statistical results show that the I 2 values of all types of constitution are more than 50%, suggesting that there is a large heterogeneity, so the random effect model should be used for meta-analysis ( Table 2). e results are shown in Table 3.

3.4.2.
Age. According to the average age of participants reported in each study, all the studies were divided into 3 subgroups by average age (≤45, 46-60, and >60). e random effects model was adopted due to the great heterogeneity of included studies. Meta-analysis showed that among the population of DM patients, the distribution of yin-deficiency, phlegm-dampness, and qi-deficiency constitution over 45 years old was significantly higher than that of adults under 45 years old (Table 4).

Meta-Analysis of Distribution of TCM Constitution in DM and General Population.
e three most common types of constitution including yin-deficiency, phlegm-dampness, and qi-deficiency were identified in the meta-analysis of the proportion of constitution in DM patients. en, we further compared it with general population. Seven studies [  Evidence-Based Complementary and Alternative Medicine reported the relevant data, so they were included in the meta-analysis.

Publication Bias.
Taking the proportion of yin-deficiency constitution of participants with DM as an index, the inverted funnel diagram was analyzed in the included literature. Considering that TCM constitution research is mostly carried out in different regions and people, and the differences in climate, environment, and lifestyle lead to obvious heterogeneity of each research, so it presents a relatively scattered state, which is related to the particularity of TCM constitution research ( Figure 6).

Analysis of the Correlation between DM and the Constitution of TCM.
Prevention is the one of the most important essences of TCM, indicating the preventive treatment of diseases. e body constitution in TCM could indicate the patients' overall condition and could form the basis of prevention theory; thus, it has attracted much attention from researchers and clinicians [97]. e TCM constitution refers to the comprehensive, relatively stable, and inherent characteristics of the morphological structure, physiological function, and psychological state formed on the basis of innate endowment and acquired disposition in the course of human life [98]. It can reflect an individual's current health status and future health trends in constitution differences, life processes, psychological condition, and adaptability to natural and social environments. e specificity of individual constitution often causes body's susceptibility to a specific pathogenic factor. For instance, people with phlegmdampness constitution easily suffer from disease such as diabetes mellitus, metabolic syndrome, and other diseases, with weak adaptability to rainy season and moist environment [99]. e TCM constitution scale provided an objective, generally recognized technical tool and professional standard. rough the constitution classification of TCM, we can determine the individual constitution type of the population and guide the individualized treatment and disease prevention. Correlative studies on constitution and clinical disease have promoted the application of TCM in public health. Relevant research shows that yin-deficiency constitution of hypertension, diabetes, constipation, female menopause, and osteoporosis showed high distribution [100]. is is also consistent with our research results.

Limitations of the Study.
From the perspective of the methodology of research design, the literature reports on the basic information related to the research is incomplete, such as the time of inclusion of the research object, basic data, and the original data of each constitution. In the aspect of data statistical analysis, some studies only describe the number and proportion of various constitution types and there is a lack of research on the correlation between constitution types and family heredity, external environment, illness and other factors. e causes of diabetes are complex, but due to the lack of reported data in the original literature, the analysis angle has some limitations, so it is difficult to deeply analyze the various factors that affect the constitution of people with diabetes. is study is different from other types of metaanalysis. e articles included are mainly observational studies, which can only hint but not prove the causal relationship between disease and constitution. e results of this study need to be confirmed by more high-quality prospective cohort studies and case-control studies. erefore, there may be a certain risk of deviation in the result. e included studies are heterogeneous. Due to the large amount of literature, there are significant differences in research time, measurement methods, sample size, age, gender, region, and other aspects. Different from intervention research, individual differences have a great impact on TCM constitution, which is also the characteristic of TCM constitution. erefore, this study conducted   phlegm-dampness constitution is closely related to various metabolic diseases, which is of great significance to guide the accurate prevention and treatment of metabolic diseases. In the future clinical practice, we can determine the target of early intervention of diabetes by screening the high-risk groups of yin-deficiency and phlegm-dampness and carry out targeted drug and nondrug intervention according to the constitution and syndrome of TCM, in order to adjust biased constitution and reduce the risk of disease.

Conclusion
e current systematic review of 87 included studies found that yin-deficiency, phlegm-dampness, and qi-deficiency are the main types of constitution in diabetes patients, of which yin-deficiency and phlegm-dampness are associated with the occurrence of diabetes. ey are also risk factors for the incidence of diabetes, while balanced constitution is a protective factor for diabetes. It is suggested that in the future, we should focus on people over 45 years old with yindeficiency in North China and phlegm-dampness people over 45 years old in South China and increase TCM constitution intervention measures as soon as possible to improve biased constitution and reduce the incidence of diseases. More high-quality observational studies should be designed in the future to provide a more scientific evidencebased basis for the use of personalized diagnosis and treatment of TCM in the prevention and treatment of diabetes.

Data Availability
e relevant data used to support the results of this study are included in this article.

Conflicts of Interest
e authors declare that there are no conflicts of interest regarding the publication of this paper. Evidence-Based Complementary and Alternative Medicine 11