Development and Validation of the Yin Deficiency Scale

Abstract

Objectives:

Yin Deficiency (YD) is a pathophysiologic pattern that manifests with emaciation, dry mouth, tidal fever, Five Center Heat, night sweats, and malar flushing. The purpose of the present study was to develop and validate a new YD metric, the Yin Deficiency Scale (YDS), to define an optimum cutoff point for the YDS, and to examine the relationships between YD and Cold–Heat patterns.

Methods:

A total of 169 outpatients and 83 college students were asked to complete the YDS. The college students were also asked to complete the Yin Deficiency Questionnaire (YDQ) and Cold–Heat Questionnaire (CHQ). Twelve clinicians determined whether or not the 169 outpatients had YD. YDS and YDQ data were used to estimate the internal consistency, construct validity, and concurrent validity of the YDS. CHQ data were used to examine the relationships between YDS and CHQ subscale scores. Total YDS scores and the clinicians' diagnoses of the outpatients were used to define an optimum cutoff score for the YDS.

Results:

The 27-item YDS had satisfactory internal consistency (α=0.885) and concurrent validity, with correlations between the YDS and YDQ subscale scores. A total of eight factors were extracted in the test of construct validity, using principal component analysis, and the overall variance explained by all factors was 63.1%. Among the eight factors, cough, fever, and skin-related factors were associated only with Heat pattern, whereas pain, urine, and fatigue-related factors were associated with both Cold and Heat patterns. In the receiver operating characteristic curve analysis, the sensitivity, specificity, and area under the curve of the YDS were 73.7%, 84.8%, and 0.875, respectively. The optimum cutoff score was defined as 10 points.

Conclusions:

The results of this study suggest that the YDS is a reliable and valid instrument for evaluating YD.

Introduction

I n East Asian medicine, a “pattern” is a diagnostic conclusion based on pathologic changes at a certain stage of a disease.¹ Identifying patterns is important in clinical cases, because patterns not only give practitioners information about the origins and lesions of pathogens, but also guide these practitioners toward treatment principles, such as those involving acupoints or herbal formulae.²

Among the methodologies available for identifying diverse patterns, the Eight Guiding Principles (EGP) classify diseases or syndromes into eight subpatterns grouped in opposing pairs: Yin–Yang; Exterior–Interior; Cold–Heat; and Deficiency–Excess.¹ The EGP are brief, well-combined with each subpattern, and emphasize the balance between healthy qi and each pathogen, which is important for holistic treatment of pathologic patterns or diseases. For example, the combination of Cold and Deficiency patterns refers to weak Healthy-Qi, and the combination of Heat and Exterior patterns refers to a strong pathogen. For these reasons, the EGP are frequently used in clinical cases to determine the fundamental characteristics of diseases or syndromes.³ Yin Deficiency (YD), a subpattern of the EGP, indicates fever caused by insufficient Yin and the consequent failure of Yin to control Yang.⁴ Clinical manifestations of YD are mainly associated with Heat Deficiency, such as emaciation, dry mouth or throat, tidal fever, palpitations, insomnia, Five Center (the palms, soles, and chest) Heat, night sweats, malar flushing, a red tongue with little coating, and a thready pulse.⁴ However, it is generally accepted that the clinical spectra of YD are broader than those related to Heat Deficiency.^3,5 YD originally results from the Kidney. However, Two-Viscera groups—such as the Lungs and Kidney, Liver and Kidney, Spleen and Kidney, and Heart and Kidney—are often involved simultaneously and can induce YD.³ Therefore, in addition to the typical symptoms related to Heat Deficiency, cough, fatigue, and tinnitus resulting from viscera other than the Kidney can occur in YD. Complex YD can also affect other regions of the body—such as the joints, bone, and skin—which are considered to be linked with the Five Viscera through meridians and the Triple Energizer. For example, Lungs have the function of disseminating Fluid and activating Defending Energy to the skin so that the surface of the skin can be kept moist and bright.⁶ When this function does not work properly because of Lung and Kidney YD, the skin becomes itchy and rough. Many studies have shown that Liuwei Dihuang Wan (a bolus of six drugs, including rehmannia), one of most frequently used prescriptions for treating YD, is effective for treating broad clinical manifestations such as respiratory, gastrointestinal, dermatologic, urinary, arthritic, spinal, musculoskeletal, hormonal, and gynecologic diseases.⁷ Therefore, to identify YD in clinical cases, Liuwei Dihuang Wan may be beneficial not only for treating symptoms and signs associated with Heat Deficiency but also for determining if a patient's symptoms and signs are associated with other viscera or regions of the body.

Lee et al. developed and validated the Yin Deficiency Questionnaire (YDQ)⁸ to assess YD in clinical settings. In their study, two factors were extracted, using a principal component analysis (PCA): Heat Deficiency and accompanying symptoms. However, these factors did not include broad clinical manifestations. Moreover, the optimum cutoff point for the YDQ was not defined. Thus, the purpose of the current study was to develop and validate a new Yin Deficiency Scale (YDS; see Appendix) that includes broad clinical manifestations and to define an optimum cutoff point for the YDS. The newly developed YDS may be beneficial for identifying YD when treating broad clinical manifestations associated with YD.

The relationships between YD and Cold–Heat patterns were also examined in this study. Although YD mainly results from Heat Deficiency, other subpatterns, such as dual Deficiency of Yin and Yang, and pure YD without Heat Deficiency also exist.⁹ In Lee's study, symptoms or signs other than Heat-Deficiency factor were grouped together rather than being classified separately.⁸ Kim et al. reported the development of the Cold–Heat Questionnaire (CHQ),¹⁰ and Ryu et al. developed and validated another version of the CHQ.¹¹ However, few studies have addressed the relationships between YD and Cold–Heat patterns. Given that the Cold score of the CHQ refers to a Yang Deficiency pattern and the Heat score of the CHQ refers to Yang Excess caused by YD,¹⁰ examining these relationships may be beneficial to clarify the following features of YD: Heat Deficiency; dual Deficiency of Yin and Yang; and pure YD without Heat Deficiency.

The current authors' previous study presented a pilot version of the YDS consisting of 30 items using the Delphi method.⁵ That study examined the reliability and validity of the pilot version of the YDS and defined optimum cutoff criteria using receiver operator characteristic (ROC) curve analysis. Finally, the relationships between YD and Cold–Heat patterns were examined.

Methods

Subjects and data collection

One hundred and sixty-nine outpatients (169; group A) who were able to understand and answer the YDS questionnaire items were asked to complete the YDS from May 2009 to June 2009. A separate group of 83 college students (group B) was asked to complete the YDQ and CHQ in addition to the YDS during November 2009. Table 1 lists the age and gender distributions of groups A and B. Twelve clinicians—who were all members of the Society of Korean Medical Etiology and had >8 years of clinical experience—determined the presence or absence of YD for each patient in group A in the clinicians' East Asian medicine clinics. Among the 169 outpatients, 57 were diagnosed with YD and 112 were defined as non-YD. Informed consent was obtained from all subjects prior to enrollment in the study.

Table 1.

Age and Gender Distributions of the Subjects

Variable	Group A (n=169) Men/women	Group B (n=83) Men/women
Number	39/130	58/25
Age (years)	42.1±14.7 / 43.5±14.9	21.0±2.7 / 20.1±1.9

Measures

YDS. For the current authors' previous study, the medical literature was searched, including Yi Xue Ru Me ¹², Dong Eui Bo Gam ¹³, and Bing Yin Lun ¹⁴ for YD references, and 75 questionnaire items were identified.⁵ The Delphi method was utilized to refine the questionnaire by asking 50 clinicians—with an average of 8 years of clinical experience each—to rate the importance of each item for YD. Through two iterations of the Delphi method, using e-mail or fax, all of the clinicians agreed upon the final 30 items included in the pilot version of the YDS (Table 2). Each item was rated on a 7-point Likert scale: 1=disagree very strongly; 2=disagree strongly; 3=disagree; 4=neither agree nor disagree; 5=agree; 6=agree strongly; and 7=agree very strongly. The scores of each item were used to examine reliability, construct validity, and concurrent validity. Additional scoring of the YDS was carried out using a dichotomous response system similar to the General Health Questionnaire (GHQ)¹⁵ and Phlegm Pattern Questionnaire (PPQ).¹⁶ On the basis of the dichotomous response system, the current authors transformed Likert scores of 1, 2, 3, and 4 to 0 point and Likert scores of 5, 6, and 7 to 1 point. The dichotomous scoring method has the advantage of eliminating errors caused by to end-users and middle-users, because this method will yield the same score irrespective of whether end-users and middle-users tend to prefer endpoints (such as 1, 2, 6, and 7) or middle points such as (3, 4, and 5).¹⁶ The totals of the 169 outpatient dichotomous YDS scores and YD determinations from the clinicians were used to examine the sensitivity and specificity, and to define an optimum cutoff score for the YDS.

Table 2.

Pilot Version of the 30-Item Yin Deficiency Scale (YDS)

Pathologic etiology	Condition	Symptom
Urine-related	I urinate frequently	Frequent urination
	My urine is dark yellow	Dark yellow urine
	I feel residual urine	Residual urine
	I can't contain my urine	Difficulty containing urine
	I wake to urinate in the night	Wake due to night urination
Dryness pattern-related	My stool is hard	Hard stool
	My hair falls out	Falling-out hair
	I have a rough skin	Rough skin
	I have an itch in the night	Night itch
	I have a dry mouth	Dry mouth
	My heel is dry and cracked	Dry and cracked heel
	I do not gain weight despite eating fully	Weight maintenance
Fire flare-up related	I have a fever in the night	Night fever
	I have a fever in the afternoon	Afternoon fever
	I have a flush in the afternoon	Afternoon flush
	My soles are hot in the night	Night hot soles
	I feel hot deep in the body, e.g., in the bone	Bone steaming
	I prefer cold beverages to warm beverages	Cold beverage preference
	I am susceptible to heat and cold	Susceptibility to heat and cold
	I sweat during sleep	Sweating during sleep
	My ear rings	Ringing ear
Cough-related	I have a cough in the afternoon	Afternoon cough
	I have a cough in the night	Night cough
	My cough lasts for a long time	Persistent cough
Fatigue-related	I feel tired or languid	Fatigue
	I feel tired in the morning	Morning fatigue
Pain-related	I feel low back pain	Low back pain
	I feel dull pain in my ankle or knee	Dull pain of the ankle or knee
	I feel heavy or weak in my lower limbs	Weakness of the lower limbs
	I feel dull pain in my heel	Dull pain of the heel

YDQ⁸ and CHQ.¹⁰ The YDQ consists of ten items rated on a 100-mm visual analogue scale. In the current study, the total scores of the two factors (Deficiency Heat and accompanying symptoms) suggested by Lee et al.⁸ were calculated to examine the relationships between YDS and YDQ subscale scores in group B. The CHQ consists of seven Heat pattern–related and eight Cold pattern–related items rated on a 7-point Likert scale: 1=disagree very strongly; 2=disagree strongly; 3=disagree; 4=neither agree nor disagree; 5=agree; 6=agree strongly; and 7=agree very strongly. The total scores of the Cold and Heat subscales were used to examine the relationships between the YDS and CHQ.

Statistical analyses

Reliability is concerned with the repeatability or reproducibility of measurement.¹⁷ Among the reliability tests, internal consistency reliability is defined as the extent to which tests or procedures assess the same characteristic, skill, or quality.¹⁸ Item–total correlations are concerned with the relationship between each item and the total scores of all items.¹⁹ Higher item–total correlations may suggest higher discriminant ability for the issues of interest, compared with other items, and this parameter is known to be associated with reliability.¹⁹ In this study, internal consistency reliability was examined using Cronbach's coefficient-α, and the corrected item–total correlations were calculated together with Cronbach's coefficient-α. A low α-value (α<0.600) or item–total correlation (r<0.300) suggested that some items either had very high variability or were not all measuring the same thing.²⁰

Validity is concerned with the accuracy of data; that is, ensuring that responses are a true reflection of the issues of interest.¹⁷ Among the validity-related parameters, construct validity is used to examine how many latent variables or factors all of the measurement variables have.¹⁹ Concurrent validity is used to demonstrate the accuracy of a measure by comparing it with another measure considered to be valid.¹⁹ In this study, construct validity was examined using a PCA with varimax rotation. The Kaiser criterion was used, and only factors with eigenvalues >1.0 were retained. Concurrent validity was examined by Pearson's correlations between the YDS and YDQ subscale scores. Pearson's correlations were also used to examine the relationships between the YDS and CHQ subscale scores.

Finally, the accuracy of a diagnostic test is characterized by its sensitivity and specificity.²¹ The ROC curve is widely accepted as a method for selecting an optimal cutoff point for a test. In this study, the cutoff point for the YDS was determined at the maximum Youden index (J=sensitivity+specificity −1) level that corresponds to the maximum accuracy level.²¹ Together with the Youden index, the area under the curve (AUC) was examined; the AUC expresses discriminative ability to determine YD. It is generally accepted that a test with an AUC>0.9 has high accuracy, 0.7–0.9 indicates moderate accuracy, and 0.5–0.6 indicates low accuracy.²¹ All statistical analyses were performed with SPSS 15 for Windows. Values were presented as mean±standard deviation, and p<0.05 indicated statistical significance.

Results

Cronbach's coefficient-α was 0.880 for the 30 items for group A. However, two items, “cold beverage preference” and “weight maintenance,” not only led to an increased Cronbach's coefficient-α if deleted but also produced an item–total correlation below 0.3. One item, “hard stool,” yielded an item–total correlation below 0.3. These results indicated that these three items did not measure the same things as the other items and, therefore, the three items were removed from the validity tests. Cronbach's coefficient-α was 0.885 for these items.

To examine construct validity, 27 items (excluding “cold beverage preference,” “weight maintenance,” and “hard stool”) were subjected to a PCA, using the data obtained from group A. As a result, eight factors were extracted, and the total percentage of variance was 63.1% (Table 3). The eight factors extracted can be described as follows: factor 1 consisted of cough-related items (cough factor); factor 2 was made up of fever-related items occurring in the afternoon and night (fever factor); factor 3 consisted of pain- and weakness–related items in the lower limbs (pain-weakness factor); factor 4 consisted of four urine-related items (urine factor); factor 5 consisted of fatigue-related items (fatigue factor); factor 6 consisted of pain- and Heat–related items mainly in the feet and bones (feet-bone steaming factor); factor 7 consisted of Kidney- and Liver–related items (Kidney-Liver Deficiency factor); and factor 8 was made up of dermatologic items (skin–hair factor).

Table 3.

Factor Loadings for the 27 Items of the Yin Deficiency Scale (YDS) and the Variance Explained by the Eight Factors

Item	Factor 1	Factor 2	Factor 3	Factor 4	Factor 5	Factor 6	Factor 7	Factor 8
Night cough	0.840	0.123	0.027	0.140	0.020	0.136	0.095	0.048
Persistent cough	0.822	0.171	0.069	0.109	0.176	0.093	0.037	−0.098
Afternoon cough	0.785	0.232	0.085	0.107	−0.018	0.110	0.074	0.109
Afternoon fever	0.228	0.760	0.230	0.021	0.010	0.072	0.184	0.173
Afternoon flush	0.289	0.738	0.052	0.084	0.091	0.165	−0.035	0.249
Dry mouth	0.115	0.703	0.110	0.300	0.172	0.061	0.047	−0.098
Night fever	0.123	0.604	0.129	−0.008	−0.048	0.225	0.540	0.132
Weakness of the lower limbs	0.205	0.167	0.812	0.033	0.116	0.108	0.055	0.175
Dull pain of the ankle or knee	0.009	0.161	0.751	−0.080	0.312	0.170	0.071	−0.094
Low back pain	−0.045	0.155	0.641	0.231	0.188	0.090	0.199	0.088
Wake due to night urination	0.008	0.171	−0.050	0.676	0.118	0.191	0.088	−0.079
Frequent urination	0.145	0.019	0.031	0.621	0.096	−0.216	0.231	0.217
Residual urine	0.308	0.261	0.084	0.569	0.072	0.141	−0.035	0.242
Difficulty containing urine	0.150	−0.061	0.266	0.553	−0.002	0.192	−0.169	0.084
Morning fatigue	0.078	−0.007	0.265	0.039	0.724	−0.015	0.250	0.118
Susceptibility to heat and cold	0.104	0.035	0.027	0.036	0.703	0.284	0.012	0.190
Fatigue	0.023	0.205	0.222	0.245	0.682	−0.113	0.043	−0.050
Dry and cracked heel	0.022	0.227	−0.012	0.146	0.138	0.684	0.052	−0.126
Night itch	0.320	0.082	0.252	−0.049	0.090	0.555	0.143	0.225
Dull pain of the heel	0.141	0.054	0.342	0.156	0.031	0.483	0.075	0.240
Bone steaming	0.188	−0.018	0.417	0.240	−0.210	0.483	0.027	0.138
Night hot soles	0.153	0.364	0.152	−0.085	0.030	0.469	0.450	0.169
Sweating during sleep	0.183	0.102	0.017	−0.033	0.188	0.189	0.690	0.025
Ringing ear	−0.071	−0.008	0.314	0.274	0.066	−0.068	0.522	0.087
Falling-out hair	−0.058	0.103	0.057	0.157	0.044	0.132	0.106	0.725
Rough skin	0.033	0.404	0.227	−0.069	0.321	0.145	−0.151	0.559
Dark yellow urine	0.332	0.084	0.060	0.183	0.129	−0.115	0.308	0.496
Variance explained	10.0 (%)	9.9 (%)	9.1 (%)	7.4 (%)	7.3 (%)	7.3 (%)	6.1 (%)	6.0 (%)

Bolding indicates factor loading >0.450.

Table 4 lists the Pearson's correlations between the eight YDS, two YDQ, and two CHQ subscale scores. All scores for the YDS factors were positively correlated with those of Heat Deficiency and accompanying symptoms of the YDQ. This indicated that the YDS was satisfactory in terms of concurrent validity. The Heat score of the CHQ was positively correlated with all the scores for the YDS factors except for factor 7 (Kidney-Liver Deficiency). However, scores for factor 3 (pain-weakness), factor 4 (urine), and factor 5 (fatigue) were positively correlated with both the Cold and Heat scores of the CHQ. This indicated that factors 3, 4, and 5 could be grouped into the dual Deficiency of Yin and Yang category.

Table 4.

Correlation Coefficients (r) Between the YDS, YDQ, and CHQ Subscale Scores

	YDQ		CHQ
YDS	Heat Deficiency	Accompanying symptoms	Heat score	Cold score
Factor 1^a	0.411^**	0.455^**	0.322^**	0.023
Factor 2^b	0.643^**	0.570^**	0.391^**	0.073
Factor 3^c	0.500^**	0.610^**	0.439^**	0.270^*
Factor 4^d	0.434^**	0.482^**	0.450^**	0.234^*
Factor 5^e	0.218^*	0.447^**	0.387^**	0.548^**
Factor 6^f	0.404^**	0.458^**	0.226^*	0.107
Factor 7^g	0.334^**	0.480^**	0.185	0.157
Factor 8^h	0.294^**	0.263^*	0.278^*	0.043

p<0.05; ^** p<0.01.

Subscales are as follows: ^aCough, ^bFever, ^cPain-weakness, ^dUrine, ^eFatigue, ^fFeet-bone steaming, ^gKidney-Liver Deficiency, ^hSkin-Hair.

Yin Deficiency Scale; Yin Deficiency Questionnaire; Cold–Heat Questionnaire.

In the ROC curve analysis, the calculated AUC value was 0.875. This indicated that the YDS had moderate discriminative ability to determine YD.¹⁸ Table 5 lists the sensitivity, specificity, and corresponding Youden index cutoff points. Based on the maximum Youden index, the optimum cutoff point of the YDS score was determined to be 10 points, and the sensitivity and specificity at that level were 73.7% and 84.8%, respectively.

Table 5.

Cutoff Score, Sensitivity, and Specificity of the Total YDS Score for Determining Yin Deficiency Pattern Using a Receiver Operating Characteristic (ROC) Curve

Cutoff score	Sensitivity (%)	Specificity (%)	Youden index (J)
1	1.000	0.036	0.036
2	1.000	0.107	0.107
3	1.000	0.196	0.196
4	1.000	0.313	0.313
5	0.982	0.429	0.411
6	0.982	0.473	0.456
7	0.930	0.580	0.510
8	0.842	0.661	0.503
9	0.772	0.768	0.540
10	0.737	0.848	0.585^a
11	0.667	0.893	0.560
12	0.509	0.946	0.455
13	0.456	0.964	0.420
14	0.263	0.991	0.254
15	0.228	1.000	0.228
16	0.211	1.000	0.211
17	0.193	1.000	0.193

Optimum cutoff point corresponds to the maximum Youden index (J=sensitivity+specificity −1).

Discussion

In this study, the 27-item YDS yielded satisfactory reliability (Cronbach's coefficient-α=0.885 and Pearson's correlations were >0.300). The total variance explained using PCA was 63.1%, and the eight factors of the YDS were significantly related to the two factors of the YDQ. Through ROC analysis, the optimum cutoff point was defined as 10 points, and the sensitivity and specificity at that level were 73.7% and 84.8%, respectively. The AUC of the ROC was 0.875; therefore, the discriminative ability of the YDS was considered to be moderate. In summary, these results suggest that the YDS is a reliable and valid questionnaire to determine YD.

“Cold beverage preference,” “weight maintenance,” and “hard stool” were removed from the reliability analysis because they did not appear to measure YD. The removal of “cold beverage preference” may have occurred because it is more closely related to individual taste than to Heat Deficiency. In YD, Fluid, Blood, and Essence are burned by Heat Deficiency, causing emaciation.³ In a previous study using the Delphi method, most clinicians considered emaciation itself to be rare in clinical cases and, therefore, inappropriate to use as a questionnaire item. Instead, the clinicians proposed using “I do not gain weight despite eating fully.”⁵ However, this item was excluded from the reliability test. Therefore, it appears that “emaciation” is not correlated with meal size or weight. “Hard stool” is considered one of the main symptoms of YD.^3
–5 However, “hard stool” actually manifests within diverse patterns: Qi Deficiency; Qi Stagnation; Stomach Heat; Dry Blood; Condensed Cold; and YD.²¹ Although “hard stool” is considered an important clue for YD, these diverse etiologies may decrease item–total correlation.

Eight factors were extracted from the YDS in the construct validity test using PCA. As expected, not only symptoms and signs related to Heat Deficiency, but also cough, pain, fatigue, and skin-related symptoms and signs were found to be independent factors. These results suggest that, in addition to Heat Deficiency symptoms or signs caused by YD, other symptoms or signs related to complex YDs—such as symptoms related to Lung-Kidney, Spleen-Kidney, and Liver-Kidney—should be considered when determining YD in clinical cases. It is interesting that “dark yellow urine” was grouped into the skin–hair factor and not into the urine factor. For Lung-Kidney YD, the body Fluid decreases, the skin becomes rough, and the urine is condensed. This may explain why “dark yellow urine” was grouped into the skin-hair factor.

In examining Pearson's correlations between the YDS and CHQ subscale scores, “cough,” “fever,” and “skin–hair’ factors were related only to Heat score, not to Cold score. This may have been because these three factors are pure factors related to Heat Deficiency. However, it is interesting that “pain-weakness,” “urine,” and “fatigue” factors were related to both Heat and Cold scores. Because the Cold score of the CHQ refers to a Yang Deficiency pattern, and the Heat score of the CHQ refers to Yang Excess such as Heat Deficiency caused by YD,¹⁰ these three factors may be associated with a dual Deficiency of Ying and Yang and not with only YD. The “Kidney-Liver Deficiency” factor was not related to the Cold or Heat score. This suggests a pattern of pure YD that is not associated with Heat Deficiency caused by YD or with a dual Deficiency of Ying and Yang. In summary, the results suggest that Cold and Heat scores may be beneficial for clarifying the characteristics of YD in clinical cases.

The ROC results indicate that the optimum cutoff of the YDS is 10 points. This cutoff point (10/27) is much lower than that of the Chalder Fatigue Scale (3/4).²³ For this reason, it appears that the YDS, like the PPQ,¹⁶ consists of many factors and that one or two factors may be in agreement with clinician decisions. It is generally accepted that YD becomes prevalent in patients with febrile or obstinate diseases, such as cancer.^3,24 Therefore, the definition of a cutoff point for the YDS may be beneficial for identifying appropriate Yin-nourishing treatment for these clinical cases.

This study refined and verified the reliability of the YDS; however, the study had some limitations. First, this study was conducted using a sample collected entirely in Seoul, Korea; therefore, the study has limited population validity. Second, data to examine correlations among the YDS, YDQ, and CHQ were not obtained from the patient group; therefore, concurrent validity and relationships between the YDS and CHQ for the outpatient group could not be assessed. Third, each outpatient was not examined by all of the clinicians; therefore interobserver reliability could not be evaluated. Fourth, the YDS was not analyzed by the test–retest reliability method; therefore, repeatability of the questionnaire items could not be evaluated. Further studies are needed to overcome the limitations of population validity and to examine concurrent validity, interobserver reliability, and repeatability for the patient group.

Conclusions

In this study, the current authors developed and validated the YDS for 169 outpatients and 83 college students. The 27-item YDS yielded satisfactory reliability (Cronbach's coefficient-α=0.885 and Pearson's correlations were >0.300). The total variance explained using PCA was 63.1%, and the eight factors of the YDS were significantly related to the two subscale scores of the YDQ. Results for the relationships between the YDS and CHQ suggested the following categories of YD: Deficiency Heat; dual Deficiency of Ying and Yang; and pure YD without Deficiency Heat. Through ROC analyses, the optimum cutoff point was defined as 10 points, and sensitivity, specificity, and the AUC at that level were 73.7%, 84.8%, and 0.875, respectively. Further studies are needed to overcome the limitations of population validity and to examine concurrent validity, interobserver reliability, and repeatability for the patient group.

Footnotes

Disclosure Statement

No competing financial interests exist.

Appendix. Yin Deficiency Scale

We would like to know more about any problems you have experienced recently. Please answer ALL of the questions by ticking the answer that most closely applies to you.

1: disagree very strongly, 2: disagree strongly, 3: disagree, 4: neither agree nor disagree, 5: agree, 6: agree strongly, 7: agree very strongly.

References

World Health Organisation. WHO International Standard Terminologies on Traditional Medicine in the Western Pacific Region. Manila: WHO Western Pacific Regional Office, 2007.

Berle

, Cobbin

, Smith

et al. A novel approach to evaluate Traditional Chinese Medicine treatment outcomes using pattern identification. J Altern Complement Med, 2010; 16:357–367.

Zhu

, Gao

, Ji

. Textbook of Traditional Chinese Medical Diagnosis [in Chinese] Beijing: Ren Min Wei Sheng Chu Ban She, 2002; 508–535.

Zou

, Tian

, Cong

. Diagnostics of Traditional Chinese Medicine: Differentiation of Syndromes. Shanghi: Shanghi University of Traditional Chinese Medicine, 1999; 292–299.

Cho

, Park

, Kim

et al. Development of questionnaire for pathogenesis analysis of Yukmijiwhang-tang symptom. J Korean Inst Orient Med Diagn, 2009; 12:61–73.

Zhang

, Shi

, Zhang

. Basic Theory of Traditional Chinese Medicine (II): The Theory of Viscera and Bowels. Shanghi: Shanghi University of Traditional Chinese Medicine, 1999; 80–88.

Liu

, Xiao

, Dai

. Contemporary Studies and Applications of Liuwei Dihuang Wan. Beijing: Ren Min Wei Sheng Chu Ban She, 2002.

Lee

, Park

, Lee

et al. Development and validation of Yin-Deficiency Questionnaire. Am J Chin Med, 2007; 35:11–20.

Leng

, Wang

. Criteria for Chinese Medical Pattern Identification. Beijing: Ren Min Wei Sheng Chu Ban She, 1996.

10.

Kim

, Park

. Development of questionnaire for Cold–Heat patternization [sic] [in Korean] J Korean Inst Orient Med Diagn, 2003; 7:64–75.

11.

Ryu

, Lee

, Kim

et al. Reliability and validity of a Cold–Heat pattern questionnaire for Traditional Chinese Medicine. J Altern Complement Med, 2010; 16:663–667.

12.

Chan

. An Introduction to Medicine. Seoul: Dae Sung Press, 1982.

13.

Heo

Joon

. An Oriental Medical Treasure. Seoul: Nam San Dang Press, 1998.

14.

Kim

. Oriental Medical Pathology. Seoul: Sun Press, 2003.

15.

Goldberg

, Williams

. A User's Fuide to the General Health Questionnaire. Windsor, Canada: NFER-Nelson, 1988.

16.

Park

, Park

, Kim

et al. Development of a valid and reliable Phlegm Pattern Questionnaire. J Altern Complement Med, 2011; 17:851–858.

17.

Smith

. Survey Research: Instruments, Validity and Reliability. Smith

. Research Methods in Pharmacy Practice. London: Pharmaceutical Press, 2002; 43–84.

18.

Writing@CSU. Reliability and Validity. Colorado State University. http://writing.colostate.edu/guides/research/relval/com2a4.cfm. 2011 Nov 21.

19.

Tak

. Psychological testing:understanding of development and evaluation of the psychological testing. Tak

. Item Analysis. Seoul: Hak-Ji Press, 2001; 135–150.

20.

Yun

, Wang

, Lee

et al. Validation study of the Korean version of the Brief Fatigue Inventory. J Pain Symptom Manage, 2005; 29:165–172.

21.

Akobeng

. Understanding diagnostic tests 3: Receiver operating characteristic curves. Acta Paediatrica, 2007; 96:644–647.

22.

Lee

, Park

, Kim

. Textbook of Oriental Medicine: Differential Diagnosis. Seoul: Sung Bo Sa, 1988; 148–149.

23.

Chalder

, Berelowitz

, Pawlikowska

et al. Development of a fatigue scale. J Psychosom Res, 1993; 37:147–153.

24.

Lin

, Huang

, Liu

et al. Severity of Yin Deficiency and autonomic nervous system function in cancer patients. J Altern Complement Med, 2009; 15:87–91.