Updated Wednesday, May 22, 2019
JCIM & JIM
Open Access
THIS ARTICLE
-  Abstract
-  Full text
-  Download PDF file
-  Send to a friend
-  Related articles in JCIM & JIM
-  Cited in JCIM & JIM
-  Reader's comments
-  Send a comment
 Language Polishing
Research Article
Journal of Integrative Medicine: Volume 13, 2015   Issue 4
Tongue diagnosis: relationship between sublingual tongue morphology in three tongue protrusion angles and menstrual clinical symptoms
Tim Hideaki Tanaka (The Pacific Wellness Institute, Toronto, Ontario, Canada M5S 2V1 )

ABSTRACT

OBJECTIVE: The morphological and color characteristics of the tongue sublingual veins (SLVs) can manifest differently within the subjects, depending on the way their tongue is curled upward. This study was conducted in order to investigate the clinical relevancy of tongue SLV diagnosis in relation to menstrual clinical symptoms (pain, clots, heavy, and scanty), using three different inspection procedures (IP1, IP2, and IP3).

METHODS: Three-hundred and seventy-seven female patients were asked to stick out their tongues in three specific ways which were intended to create different tongue protrusion angles. The SLV parameters for thickness (TK), length (LE), color (CL), shape (SP), and nodules (ND) were then evaluated.

RESULTS: According to the results of the Wald χ2 test, IP1 provides the best model for pain (R2 = 0.155), IP3 for clots (R2 = 0.437), IP2 for heavy (R2 = 0.268), and scanty (R2 = 0.192). Abnormal SLV diagnostic parameters were most strongly associated with the clinical symptom of clots (R2 = 0.492).

CONCLUSION: While the study showed the relations between tongue SLV features and menstrual clinical symptoms, as well it showed that IP2 was the best overall predictor for the symptomatic indexes used in this study, and using one particular SLV inspection procedure may not be sufficient. The application of a particular inspection method alone may cause under- or over-estimation of SLV abnormalities.

Citation: Tanaka TH. Tongue diagnosis: relationship between sublingual tongue morphology in three tongue protrusion angles and menstrual clinical symptoms. J Integr Med. 2015; 13(4): 248–256.

  

1 Introduction

Tongue diagnosis has been an integral part of Chinese medical practice since the latter’s inception around the third century BCE[1]. A description of general tongue characteristics in relation to diseases appears in the classic Chinese medicine texts Huangdi Neijing (Yellow Emperor’s Inner Classic) and Shanghan Lun (Discussion of Cold-Induced Disorders). A description of tongue diagnosis specific to sublingual veins (SLV), however, does not appear until much later, in the thirteenth century CE[2]. Abnormality in SLV features is generally associated with compromised smooth flow of blood, traditionally known as blood stasis. Other tongue signs associated with blood stasis include a purple or dark red tongue body and the existence of brown or purple spots (stasis speckles) on the tongue body surface[3].
Many physical signs and symptoms are associated with blood stasis. A sharp and fixed pain anywhere in the body could be caused by blood stasis. Blood stasis is commonly seen among women of childbearing age who are experiencing menstrual symptoms, including dysmenorrhoea and clotting menstrual blood[3].
It is important to note that although much clinically useful information can be obtained through observation of the tongue body, coating, and SLV, various precautionary steps are required to avoid false or distorted findings. When it comes to the inspection of SLVs in particular, most traditional Chinese medicine texts offer ambiguous procedural information. There is no consensus as to precise SLV inspection methodology, and SLV images taken with non-uniform inspection methods have been appearing in some case reports and textbooks[4–8].

 
  

2 Methods

2.1 Subjects
Utilizing the filtering functions of our patient database, a list was extracted consisting of females between 20 and 45 years old who initially visited our clinic between April 2010 and March 2014 (n=458). Patients had provided their consent for their tongues to be photographed and the images included in the study were further extracted (n=422). Patients with amenorrhea, who were pregnant, or who were wearing intrauterine devices were excluded from the study (n=17). Some patients who were unable to move their tongue according to the instructions were also excluded (n=26). In addition, missing or inadequate images resulting from technical issues were also discarded (n=2). Finally, 377 patients were selected for this study.
2.2 SLV image capture procedures
Patients were asked not to consume hot drinks, caffeine, alcohol, and colored foods for at least 1 h prior to the tongue assessments. At each assessment, patients were asked to stick out their tongues in three specific ways: for inspection procedure 1 (IP1), subjects were asked to curl the tongue upward and gently rest the tip against the upper lip; for IP2, subjects were asked to curl the tongue upward and gently rest the tip between the root of the upper central incisor and the transverse palatine folds; for IP3, subjects were asked to curl the tongue upward and gently rest the tip on or near the top of the palate. The variations were intended to create three different tongue protrusion angles (approximately 30, 45, and 80 degrees for IP1, IP2, and IP3, respectively). The subjects held each tongue position for approximately 1 s in a relaxed manner while a photo was taken. An interval of about 5 s occurred between each of the three tongue positions. If patients required more than 3 s to set the proper tongue position in a relaxed manner, the image was also retaken after 30 s rest, in order to avoid distension or engorgement of SLVs due to prolonged curling attempts of their tongue. In the rare event, patients who repeatedly failed to perform the movements as instructed were disqualified from the study. Utilizing a random number table generated by Microsoft Excel, the order of observation was randomly assigned among patients, with 6 different orders occurring.
Subjects’ tongue images were captured using a digital single lens reflex camera with a macro lens, attached with a twin flash system (Olympus Imaging Corp, Tokyo). Focal length (70 mm), aperture (F19), color temperature (5 500 K), and other settings remained consistent throughout the examinations. The distance between the lens and the subject’s tongue was approximately 15 cm. The output power ratio of left and right flashes was set to 1:1. The acquired images (3 648 × 2 736 pixels) were transferred to a personal computer and carefully examined on a color-calibrated in-plane switching monitor. The morphological and color characteristics of the SLVs were then evaluated.
 2.3 SLV image evaluation
The following SLV appearances were considered to be normal SLV features[6,10]: thickness (TK): less than 2.7 mm in diameter; length (LE): no more than three-fifths of the entire length of the underside of the tongue; color (CL): color of SLVs should be light and not of a dark-blue or purple tone; shape (SP): one straight vein on each side, without tortuousness or varicosis; nodules (ND): no nodules along the veins. The SLV features were assessed by the author. For TK and LE, the digital and analog rulers were utilized as needed. For each parameter, the particular SLV appearance within the norm was coded as (0) and those outside the norm were coded as (1), for later statistical analysis.

2.4 Menstrual signs and symptoms

Patients were asked to self-evaluate the following clinical symptoms related to their menstruation: pain and cramps during menstruation (pain); observation of clotting in their menstrual blood (clots); heavy menstrual flow (heavy); scanty menstrual flow (scanty). The above indexes were inputted into an Excel sheet. No symptom was coded as (0) and the existence of symptoms was coded as (1), for later statistical analysis. The data input and analysis were performed by personnel who were unfamiliar with the details of the study.

2.5 Statistical analysis

For each of the four dependent variables (clinical symptoms), multiple logistic regression for binary responses was used to investigate the effects of the independent variables (IP) on the probability that a case is a member of one of the categories of the dependent variable. In multiple logistic regression, the effect of a specific independent variable on the dependent variable was adjusted for the effects of all the other independent variables in the model. The Type 3 analysis of effects based on the Wald χ2 test was used to determine whether an effect was statistically significant. A P value of less than 0.05 resulted in the rejection of the null hypothesis.

 In multiple logistic regression, collinearity occurs when two or more independent variables in the model are approximately determined by a linear combination of other independent variables in the model. A tolerance of less than 20 indicates a potential collinearity problem[11]. Therefore, the tolerance statistic[11,12] was calculated in order to check the collinearity among the independent variables,

 

  

3 Results

3.1 Analysis results of the four multiple logistic regressions (one for each of the dependent variables (pain, clots, heavy, and scanty) with independent variables (IP))
The tolerance results were all > 0.2, indicating that there was no multicollinearity problem in the fitted model. Tables 1–3 summarize the results of the four multiple logistic regressions which show the two-way frequency table of the dependent variables and the independent variables.
3.1.1 Relationship between IP1 and clinical symptoms
The Wald χ2 test results when clinical symptom = pain, and the independent variables = IP1-TK, IP1-LE, IP1-CL, IP1-SP, and IP1-ND: R2 = 0.155; Wald χ2 DF = 1. There was a statistically significant association between pain and IP1-TK (χ2 (1, N = 377) = 8.059, P = 0.005), between pain and IP1-LE (χ2 (1, N = 377) = 8.656, P = 0.003), or between pain and IP1-SP (χ2 (1, N = 377) = 3.943, P = 0.047). The Wald χ2 test results when clinical symptom = clots, and the independent variables = IP1-TK, IP1-LE, IP1-CL, IP1-SP, and IP1-ND: R2 = 0.222; Wald χ2 DF = 1. There was a statistically significant association between clots and IP1-SP (χ2 (1, N = 377) = 21.567, P <0.001), and between clots and IP1- ND (χ2 (1, N = 377) = 11.387, P = 0.007). The Wald χ2 test results when clinical symptom = heavy, and the independent variables = IP1-TK, IP1-LE, IP1-CL, IP1-SP, and IP1-ND: R2 = 0.110; Wald χ2 DF = 1. There was a statistically significant association between heavy and IP1-SP (χ2 (1, N = 377) = 14.476, P = 0.001). The Wald χ2 test results when clinical symptom = scanty, and the independent variables = IP1-TK, IP1-LE, IP1-CL, IP1-SP, and IP1-ND: R2 = 0.149; Wald χ2 DF = 1. There was a statistically significant association between heavy and IP1-SP (χ2 (1, N = 377) = 20.916, P < 0.001).
3.1.2 Relationship between IP2 and clinical symptoms
The Wald χ2 test results when clinical symptom = pain, and the independent variables = IP2-TK, IP2-LE, IP2-CL, IP2-SP, and IP2-ND: R2 = 0.148; Wald χ2 DF = 1. There was a statistically significant association between pain and IP2-LE (χ2 (1, N = 377) = 3.929, P = 0.048), and between pain and IP2-ND (χ2 (1, N = 377) = 9.919, P = 0.002). The Wald χ2 test results when clinical symptom = clots, and the independent variables = IP2-TK, IP2-LE, IP2-CL, IP2-SP and IP2-ND: R2 = 0.406; Wald χ2 DF = 1. There was a statistically significant association between clots and IP2-CL (χ2 (1, N = 377) = 6.129, P = 0.013), between clots and IP2-SP (χ2 (1, N = 377) = 62.739, P < 0.001), and between clots and IP2-ND (χ2 (1, N = 377) = 8.177, P = 0.004). The Wald χ2 test results when clinical symptom = heavy, and the independent variables = IP2-TK, IP2-LE, IP2-CL, IP2-SP, and IP2-ND: R2 = 0.267; Wald χ2 DF = 1. There was a statistically significant association between heavy and IP2-TK (χ2 (1, N = 377) = 5.760, P = 0.016), between heavy and IP2-SP (χ2 (1, N = 377) = 11.536, P = 0.007), and between heavy and IP2-ND (χ2 (1, N = 377) = 17.353, P < 0.001). The Wald χ2 test results when clinical symptom = scanty, and the independent variables = IP2-TK, IP2-LE, IP2-CL, IP2-SP, and IP2-ND: R2 = 0.192; Wald χ2 DF = 1. There was a statistically significant association between scanty and IP2-TK (χ2 (1, N = 377) = 8.874 6, P = 0.003), between scanty and IP2-LE (χ2 (1, N = 377) = 6.339, P = 0.012), and between scanty and IP2-SP (χ2 (1, N = 377) = 10.293, P = 0.001).
3.1.3 Relationship between IP3 and clinical symptoms
The Wald χ2 test results when clinical symptom = pain, and the independent variables = IP3-TK, IP3-LE, IP3-CL, IP3-SP, and IP3-ND: R2 = 0.133; Wald χ2 DF = 1. There was a statistically significant association between pain and IP3-LE (χ2 (1, N = 377) = 6.881, P = 0.009), and between pain and IP3-SP (χ2 (1, N = 377) = 13.614, P < 0.001). The Wald χ2 test results when clinical symptom = clots, and the independent variables = IP3-TK, IP3-LE, IP3-CL, IP3-SP, and IP3-ND: R2 = 0.437; Wald χ2 DF = 1. There was a statistically significant association between clots and IP3-SP (χ2 (1, N = 377) = 72.481, P < 0.001), and between clots and IP3-ND (χ2 (1, N = 377) = 14.379, P < 0.001). The Wald χ2 test results when clinical symptom = heavy, and the independent variables = IP3-TK, IP3-LE, IP3-CL, IP3-SP, and IP3-ND: R2 = 0.207; Wald χ2 DF = 1. There was a statistically significant association between heavy and IP3-SP (χ2 (1, N = 377) = 16.316, P < 0.001) and between heavy and IP3-ND (χ2 (1, N = 377) = 16.767, P < 0.001). The Wald χ2 test results when clinical symptom = scanty, and the independent variables = IP3-TK, IP3-LE, IP3-CL, IP3-SP, and IP3-ND: R2 = 0.054; Wald χ2 DF = 1. There was a statistically significant association between scanty and IP3-TK (χ2 (1, N = 377) = 6.150, P = 0.013).
Table 4 summarizes the R2 for each dependent variable (clinical symptom) and procedure (IP1, IP2, and IP3). The R2 could be used to compare models for the same dependent variables. According to the results of R2, when clinical symptom = pain, IP1 provides the best model, in terms of power of prediction; when clinical symptom = clots, IP3 provides the best model; when clinical symptom = heavy, IP2 provides the best model; when clinical symptom = scanty, IP2 provides the best model. If considering all 4 clinical symptoms, IP2 provides an overall better power of prediction than IP1 and IP3.
3.2 Relation between clinical symptoms and SLV parameters
Tables 5–8 show the frequency counts of SLV diagnostic parameters (total TK, total LE, total CL, total SP, and total ND) by each clinical symptom. Note that total TK score = Sum of IP1-TK (0 or 1), IP2-TK (0 or 1), and IP3-TK (0 or 1). Total LE score = sum of IP1-LE (0 or 1), IP2-LE (0 or 1), and IP3-LE (0 or 1). Total CL score = sum of IP1-CL (0 or 1), IP2-CL (0 or 1), and IP3-CL (0 or 1). Total SP score = sum of IP1-SP (0 or 1), IP2-SP (0 or 1), and IP3-SP (0 or 1). Total ND score = sum of IP1-ND (0 or 1), IP2-ND (0 or 1), and IP3-ND (0 or 1). The results of tolerance were all > 0.2, indicating that no multicollinearity problem existed in the fitted model. The analysis results of the logistic regression are summarized below.
3.2.1 Pain and total SLV parameters
The Wald χ2 test results when dependent variable = pain, and the independent variables = total TK, total LE, total CL, total SP, and total ND: R2 = 0.192; Wald χ2 DF = 1. There was a statistically significant association between pain and the total LE score (χ2 (1, N = 377) = 15.925, P < 0.001).
3.2.2 Clots and total SLV parameters
The Wald χ2 test results when dependent variable = clots, and the independent variables = total TK, total LE, total CL, total SP, and total ND: R2 = 0.492; Wald χ2 DF = 1. There was a statistically significant association between clots and the total CL score (χ2 (1, N = 377) = 6.123, P = 0.013), between clots and the total SP score (χ2 (1, N = 377) = 57.604, P < 0.001), and between clots and the total ND score (χ2 (1, N = 377) = 12.853, P = 0.003).
3.2.3 Heavy and total SLV parameters
The Wald χ2 test results when dependent variable = heavy, and the independent variables = total TK, total LE, total CL, total SP, and total ND: R2 = 0.261; Wald χ2 DF = 1. There was a statistically significant association between heavy and the total SP score (χ2 (1, N = 377) = 12.153, P < 0.001), and between heavy and the total ND score (χ2 (1, N = 377) = 9.838, P = 0.002).
3.2.4 Scanty and total SLV parameters
The Wald χ2 test results when dependent variable = scanty, and the independent variables = total TK, total LE, total CL, total SP, and total ND: R2 = 0.120; Wald χ2 DF = 1. There was a statistically significant association between scanty and the total TK score (χ2 (1, N = 377) = 4.000, P = 0.046), and between scanty and the total SP score (χ2 (1, N = 377) = 12.608, P < 0.001).

  

4 Discussion

In this study, the clinical relevancy of tongue SLV features for menstrual symptoms was investigated.
First, the relations between menstrual clinical symptoms (pain, clots, heavy, and scanty) and SLV diagnostic parameters (TK, LE, CL, SP, and ND) were analyzed using three different inspection methods (IP1, IP2, and IP3). According to the results of the Wald χ2 test, when clinical symptom = pain, IP1 appears to be the best model, in terms of power of prediction based on R2 value. In particular, IP1-LE showed the most significant association with pain. When clinical symptom = clots, IP3 provides the best model. In particular, IP3-SP showed the most significant association with clots. When clinical symptom = heavy, IP2 provides the best model. In particular, IP2-ND showed the most significant association with heavy menstrual bleeding. When clinical symptom = scanty, IP2 provides the best model. In particular, IP2-SP showed the most significant association with scanty menstrual bleeding. Considering all four clinical symptoms, the results indicate that IP2 provides a better overall power of prediction than IP1 and IP3.
The second part of the analysis was intended to investigate the relationship between clinical symptoms and total SLV diagnostic parameter scores. Pain showed a significant association with the total LE score. Clots, heavy and scanty showed the most significant association with the total SP score. Overall, abnormal SLV diagnostic parameters were most strongly associated with the clinical symptoms of clots. This finding is in agreement with traditional Chinese medicine teachings. As indicated earlier, abnormalities in SLV features are frequently associated with blood stasis. Blood stasis is a frequent cause of menstrual symptoms such as pain, heavy or scanty menstrual blood flow, and clotting menstrual blood. However, clotting blood is generally considered to be most specifically involved in blood stasis, while the other clinical symptoms can be involved with other multiple underlying causes[3,16]. While previous literature indicated a relationship between abnormal SLV features and blood stasis and a relationship between blood stasis and menstrual symptoms, there had been no scientific study that investigated the direct link between abnormal SLV features and menstrual symptoms. This study demonstrated the link between abnormal SLV diagnostic parameters and menstrual-related symptomatic parameters.
Besides SLVs, other tongue signs, such as a dark purple tongue body and stasis speckles, are also indicators of blood stasis[3]. However, the morphological and color changes of SLV are considered more time-sensitive parameters that reflect early stages of blood stasis more precisely than do other tongue stasis signs[17]. Despite its claimed clinical usefulness, the diagnostic outcomes of the aforementioned SLV parameters could change substantially due to procedural bias when the inspection procedure is not uniform (Figure 1). Preliminary analysis of 189 cases found that the appearance of SLVs varied substantially depending on slight variations in tongue elevation angles[9]. It should be noted that the sublingual veins can become more superficial and/or engorged with any tongue angles, when the patients use too much force to stretch or contract the tongue. Therefore, various precautionary steps, as described in the Method section, were followed to avoid false or distorted findings. Still, notable alterations in every SLV parameter were observed in multiple cases depending on the tongue protrusion angle.
Based on the results from Analysis I, IP2 appears to be the best overall predictor for the clinical symptoms tested in this study. However, IP1 and IP3 also showed significant associations for certain clinical symptoms such as pain and clots. The results suggest that using one particular SLV inspection procedure may not be sufficient for diagnosis. The application of a particular inspection method alone may cause under- or over-estimation of SLV abnormalities and may yield ineffective or erroneous clinical decisions.
The major limitation of this study is the lack of an independent assessor conducting the SLV examinations. Therefore, potential assessor bias is a concern, and inter-practitioner assessment reliability cannot be assumed. It should be noted, however, that tongue assessment in general is a subjective form of diagnostic technique. A study by Kim et al[18] reported a low level of inter- and intrapractitioner agreement regarding tongue surface parameters. The reliability of SLV assessment has not been evaluated.
The symptomatic indexes used in this study, such as “heavy flow” and “scanty flow,” are also subjective parameters; each patient could have different interpretations of these metrics. In future studies, it is important to incorporate objective clinical parameters.
It should be also noted that the SLV observation and evaluation methods used in this study are different from the procedures traditionally performed in acupuncture and Chinese medicine practices. In many clinical practices, tongue inspection is performed under insufficient or mixed lighting conditions, which can substantially alter tongue appearance — in particular, the color of SLVs. As well, a practitioner typically has only a few seconds to observe and memorize patients’ tongue features while a patient is holding a single tongue position. In this study, on the other hand, three tongue images with different protrusion angles were digitally captured under standardized lighting conditions. The captured images were then carefully observed on a large high-resolution screen without any time restrictions. The outcomes of SLV assessments from this study might, as a result, differ from outcomes derived from common clinical methods.
Lastly, while this study showed that SLV features are a meaningful predictor for menstrual symptoms, their applicability to other clinical ailments is unknown. Abnormal SLV appearance may also be suggestive of conditions including hepatitis, coronary heart disease, and hypertension[5,6]. It would be interesting to conduct studies investigating the diagnostic usefulness of SLV features on other demographics or disease populations.
In conclusion, the study showed the relations between tongue SLV features and menstrual clinical symptoms. IP2 was the best predictor for the symptomatic indexes used in this study, while the clinical symptom of clots was most significantly associated with SLV abnormality.

  

5 Acknowledgements

The author would like to thank Yuhua Su, Ph.D. for her statistical assistance.

  

6 Conflict of interests

The author declares that there is no conflict of interests regarding the publication of this paper.

  

Figures and Tables in this article:



Figure 1 Examples of intra-subject SLV variations

The images of the same individual’s tongue SLV with three different tongue protrusion angles. The subjects held each tongue position for approximately 1 s in a relaxed manner while a photo was taken. An interval of about 5 s occurred between each of the three tongue positions. A: 29-year-old female with IP1; B: 29-year-old female with IP2; C: 29-year-old female with IP3; D: 41-year-old male with IP1; E: 41-year-old male with IP2; F: 41-year-old male with IP3. SLV: sublingual veins.
 

Table 1 Clinical symptoms (pain, clots, heavy, and scanty) vs. IP1-TK, IP1-LE, IP1-CL, IP1-SP, and IP1-ND

(n (%))

IP1: inspection procedure 1; TK: thickness; LE: length; CL: color; SP: shape; ND: nodules.
 

Table 2 Clinical symptoms (pain, clots, heavy, and scanty) vs. IP2-TK, IP2-LE, IP2-CL, IP2-SP, and IP2-ND

(n (%))

IP2: inspection procedure 2; TK: thickness; LE: length; CL: color; SP: shape; ND: nodules.
 

Table 3 Clinical symptoms (pain, clots, heavy, and scanty) vs. IP3-TK, IP3-LE, IP3-CL, IP3-SP, and IP3-ND

(n (%))

IP3: inspection procedure 3; TK: thickness; LE: length; CL: color; SP: shape; ND: nodules.


Table 4 The R2 for each dependent variable (clinical symptom) and procedure (IP1, IP2, and IP3)

IP: inspection procedure.


Table 5 Frequency counts of total TK, total LE, total CL, total SP, and total ND, by clinical symptom = pain

Freq (frequency of occurrences): “0” indicates SLV abnormality was not observed with any of the SLV inspection procedures; “1” indicates SLV abnormality was observed with one of the SLV inspection procedures (IP1, IP2, and IP3); “2” indicates SLV abnormality was observed with two of the SLV inspection procedures (IP1, IP2, and IP3); “3” indicates SLV abnormality was observed with all of the SLV inspection procedures (IP1, IP2, and IP3). SLV: sublingual vein; TK: thickness; LE: length; CL: color; SP: shape; ND: nodules.


Table 6 Frequency counts of total TK, total LE, total CL, total SP, and total ND, by clinical symptom = clots

Freq (frequency of occurrences): “0” indicates SLV abnormality was not observed with any of the SLV inspection procedures; “1” indicates SLV abnormality was observed with one of the SLV inspection procedures (IP1, IP2, and IP3); “2” indicates SLV abnormality was observed with two of the SLV inspection procedures (IP1, IP2, and IP3); “3” indicates SLV abnormality was observed with all of the SLV inspection procedures (IP1, IP2, and IP3). SLV: sublingual vein; TK: thickness; LE: length; CL: color; SP: shape; ND: nodules.
 

Table 7 Frequency counts of total TK, total LE, total CL, total SP, and total ND, by clinical symptom = heavy

Freq (frequency of occurrences): “0” indicates SLV abnormality was not observed with any of the SLV inspection procedures; “1” indicates SLV abnormality was observed with one of the SLV inspection procedures (IP1, IP2, and IP3); “2” indicates SLV abnormality was observed with two of the SLV inspection procedures (IP1, IP2, and IP3); “3” indicates SLV abnormality was observed with all of the SLV inspection procedures (IP1, IP2, and IP3). SLV: sublingual vein; TK: thickness; LE: length; CL: color; SP: shape; ND: nodules.
 

Table 8 Frequency counts of total TK, total LE, total CL, total SP, and total ND, by clinical symptom = scanty

Freq (frequency of occurrences): “0” indicates SLV abnormality was not observed with any of the SLV inspection procedures; “1” indicates SLV abnormality was observed with one of the SLV inspection procedures (IP1, IP2, and IP3); “2” indicates SLV abnormality was observed with two of the SLV inspection procedures (IP1, IP2, and IP3); “3” indicates SLV abnormality was observed with all of the SLV inspection procedures (IP1, IP2, and IP3). SLV: sublingual vein; TK: thickness; LE: length; CL: color; SP: shape; ND: nodules.

  
References
1. Chen ZL. Brief history of tongue inspection[J]. Chin Med J (Engl), 1987, 100(1) : 38–44.
  
2. Chen ZL, Xie JW, Chen JM, Xu DS, Wang PZ, Shen JZ, Dai HL. An analysis on tongue features of 5403 healthy subjects[J]. J Tradit Chin Med, 1982, 2(2) : 151–156.
  
3. Wiseman N, Ellis A. Fundamentals of Chinese medicine[M]. Brookline, Massachusetts: Paradigm Publications, 1996. 23–149.
4. Schnorrenberger C, Schnorrenberger B. Pocket atlas of tongue diagnosis[M]. Stuttgart: Thieme, 2005. 230–231.
5. Kirschbaum B. Observation of sublingual veins[J]. J Chin Med, 2003, 71: 30–31.
6. Beppu S, Serada K. Tongue diagnosis for clinicians[M]. Tokyo: Ishiyaku Publishers, 2005. 12–97.
7. Kirschbaum B. Atlas of Chinese tongue diagnosis. Vol 2[M]. Seattle: Eastland Press, 2003. 187–237.
8. Ashida M, Ikejiri K, Ito R, Okada M, Shigemi M. The tongue and the pluse in diagnosis of traditional Chinese medicine[M]. Tokyo: Ishiyaku Publishers, 1993. 27.
9. Tanaka T. Tongue examination of sublingual veins: intrasubject variability in four diagnostic parameters. ICMART 2011 World Congress on Medical Acupuncture. 2011.
10. Chen ZL, Hu QF. Recent development in research on tongue inspection[J]. Chin Med J (Engl), 1986, 99(6) : 444–456.
  
11. Menard S. Applied logistic regression analysis. Thousand Oaks, CA: Sage University. 1995.
12. Allison P. Logistic regression using SAS: Theory and application. Cary, NC: SAS Institute Inc. 2012.
13. Menard S. Coefficients of determination for multiple logistic regression analysis[J]. Am Stat, 2000, 54(1) : 17–24.
[Tandfonline]  
14. Nagelkerke N. A note on a general definition of the coefficient of determination[J]. Biometrika, 1991, 78: 691–692.
[Jstor]  
15. SAS/STAT 9.3 user’s guide. Cary, NC: SAS Institute Inc. 2012.
16. Maciocia G. The foundations of Chinese medicine[M]. London: Elsevier, 2005. 385–449.
17. Maciocia G. Tongue diagnosis in Chinese medicine[M]. Seattle: Eatland Press, 1987. 62–63.
18. Kim M, Cobbin D, Zaslawski C. Traditional Chinese medicine tongue inspection: an examination of the inter- and intrapractitioner reliability for specific tongue characteristics[J]. J Altern Complement Med, 2008, 14(5) : 527–536.
  
  
Reader's comments
1. After completing their trip and a series of challenges on the way, the top winner will be given an additional $5,000 worth of future travel from CheapOair. coach factory outlet http://coachfactoryoutletstore.backpacks2015.com/ (2015/11/24 0:00:00)
2. Artificial limbs, foot, leg; arm can be attached to the body to replace the damaged one. The exterior or interior parts that cannot be manufactured synthetically or artificially, an identical part from another human being or even of an animal can be transplanted. A kidney, an eye, the liver, a lung, and even the heart can be and has also been transplanted. Science can give new hearts in place of old ones. nike roshe run http://www.agoraevangelism.net/47/ (2015/11/24 0:00:00)
3. Now before you quit your job and pursue your passion, realize that it doesn??t always work out that way. cheap christian louboutin shoes http://www.hickorysociety.com/71/ (2015/11/24 0:00:00)
4. What I really like as a cyclist is that the lenses are tall and, when my head is down, I can still look up through the glasses. nike zooms http://www.nikeruntheone.com/ (2015/11/24 0:00:00)
5. We know that eyes are very important for us and wed better protect it from any danger. Nike Free Mens http://www.golfleecreek.net/07/ (2015/11/24 0:00:00)
6. Thank you a bunch for sharing this with all of us you really recognise what you're speaking approximately! Bookmarked. Please also discuss with my web site =). We will have a hyperlink trade agreement among us duvetica paris http://www.okatent.com/it/?it-duvetica-paris-18792.html (2015/12/8 0:00:00)
7. Heya! I'm at work browsing your blog from my new apple iphone! Just wanted to say I love reading through your blog and look forward to all your posts! Carry on the great work! ray ban wayfarer dobravel http://www.merchantservicesbedfordshire.co.uk/?uk-ray-ban-wayfarer-dobravel-24653.html (2015/12/13 0:00:00)
8. If you desire to take a great deal from this post then you have to apply such methods to your won web site. canada goose coat for cheap http://www.sanctus1.co.uk/?uk-canada-goose-coat-for-cheap-27746.html (2015/12/27 0:00:00)
9. Wow that was unusual. I just wrote an really long comment but after I clicked submit my comment didn't show up. Grrrr... well I'm not writing all that over again. Anyway, just wanted to say excellent blog! oakley radar path pas cher xy http://www.cc-des-sucs.fr/?fr-oakley-radar-path-pas-cher-xy-21821.html (2016/1/8 0:00:00)
10. I think he reminded many of Jordan, as they race like two peas in a pod. nike free advantage training shoes womens http://www.ciplore.com/m317/ (2016/2/6 0:00:00)
11. ???????????????????? (2016/3/26 0:00:00)
12. ???????????????????? (2016/3/27 0:00:00)
13. ??????????????????????????????????????????? ??????????? http://okdgg.com (2016/3/27 0:00:00)
14. B0MHT6 Say, you got a nice post.Much thanks again. Will read on (2017/9/20 0:00:00)
15. B0MHT6 Say, you got a nice post.Much thanks again. Will read on (2017/9/20 0:00:00)
16. abCAa3 raspberry ketone lean advanced weight loss (2017/11/8 0:00:00)
17. abCAa3 raspberry ketone lean advanced weight loss (2017/11/8 0:00:00)
18. PSCBZS There is noticeably a bundle to know about this. I assume you made sure nice points in options also. (2018/8/13 0:00:00)
19. PSCBZS There is noticeably a bundle to know about this. I assume you made sure nice points in options also. (2018/8/13 0:00:00)
20. jvolBe I truly appreciate this article post.Much thanks again. (2018/8/13 0:00:00)
21. jvolBe I truly appreciate this article post.Much thanks again. (2018/8/13 0:00:00)
22. 1hVcwS wow, superb blog post.Really pumped up about read more. Really want more. (2018/8/13 0:00:00)
23. HZ8UCG Pretty! This has been an incredibly wonderful post. Thanks for supplying these details. (2019/4/19 0:00:00)
24. HZ8UCG Pretty! This has been an incredibly wonderful post. Thanks for supplying these details. (2019/4/19 0:00:00)

Send a comment

 Home | Current Issue | Past Issues | SearchRSS
Copyright © 2013-2018 by JIM Editorial Office. All rights reserved. ISSN 2095-4964; CN 31-2083/R. 沪ICP备110264号