Translation and Psychometric Evaluation of the Persian Version of Lupus Erythematosus Quality of Life Questionnaire (LEQoL)

DOI: https://doi.org/10.21203/rs.3.rs-436424/v1

Abstract

Background: Recently a comprehensive and multi-diversionary scale for assessing Quality of Life (QoL) in Patients with Lupus has been developed. This study aimed to evaluation psychometric properties of the Persian version of Lupus Erythematosus Quality of Life Questionnaire (LEQoL) in Iranian patients with Lupus Erythematosus.

Method: We used the forward-backward translation, and cognitive interview for linguistic translation. A cross-sectional design was utilized. We recruited a convenience sample of 165 lupus patients aged 19 years or over from the Iranian social media and Tehran city rheumatology clinics. Patients completed the 36-Item Short Form Survey (SF-36), The Symptom Checklist-90-R (SCL-90-R), and the translated LEQoL. The psychometric properties of the LEQoL were examined to establish test-retest reliability, internal consistency with Cronbach's alpha coefficient (COA), divergent-convergent validity, and construct validity. Also, we used Confirmatory factor analysis (CFA) and exploratory factor analysis (EFA) for assessing factor structures.

Results: The total COA was α = 0.86. All subscales internal consistency ranging from 0.85–0.94. The findings of test-retest reliability for the overall scale was 0.93 and the subscales ranging between 0.82–0.92. The results from a CFA indicated that goodness‐of‐fit are satisfactory (χ2/df = 1.28, RMSEA = 0.042, CFI = 0.975). Also EFA showed that the Persian version of LEQoL with five-factor can explain 73.7% of the variances. For assessing validity, every factor of LEQoL has a correlation with some SF.36 and SCL-90-R subscales. This matrix indicates that the LEQoL subscales are somewhat independent, and this matter is one strength issue. Only depression (SCL-90 subscale) has a significant correlation with all LEQoL subscales.

Conclusion: The translated Persian version of the LEQoL is a suitable scale for assessing QoL in Iranian patients with lupus.

Introduction

One of the chronic, heterogeneous, inflammatory, and autoimmune disease with broad dimensions of symptoms is Lupus Erythematosus (LE)(1). LE characterized by immune dysregulation, production of autoantibodies, damages in various organs (e.g. Skin and the brain). LE divided in two main subsets: lupus erythematosus (SLE) and cutaneous lupus erythematosus (CLE) (2). The lupus erythematosus and cutaneous lupus erythematosus can occur co-morbid and separately(3). The prevalence of LE is at least five million worldwide, and the result showed that the prevalence of SLE in Iran is estimated at 40 per 100,000 Iranian(3). About gender differences, in every 10 LE patients, nine of them are female(4).

In chronic diseases context, one of the therapeutic aim is enhancing quality of life. As interventions for treating LE are in first steps, so one of the primary and important therapeutic aims for these patients is increasing their quality of life (57).The scientific literature showed that many psychiatric and neuropsychiatric symptoms can occur in LE patients. These symptoms include mood changes, chorea, seizures, anxiety, suicide, sleep disturbance, cognitive dysfunction, and even psychotic (810). So, this disease has bio-psycho-social nature and patient’s quality of life is multidimensional (e.g. emotional and physical). For assessing LE quality of life, there are some tools such as Lupus Quality of Life (Lupus QoL)(11), Systemic Lupus Erythematosus Quality of Life Questionnaire (L-QoL) (12), and Skindex to measure dermatology(13). These scales showed promising results in English patients and used widely. However, this scale has various and important limitation. First, these scales did not assess LE in general. In fact, they evaluate skin pathology, systemic lupus erythematosus, or other subsets instead of all domains. Second, as mentioned above QoL has various dimensions (emotional, physical, Cognition, Interpersonal, and Appearance), But, theses scales only evaluating one or two QoL-related dimensions(13). Finally, it is also suggested that a holistic point of view can describe patients in a comprehensive way and expand the clinician’s attention regarding making treatments. (12, 13).

Recently, a Lupus Erythematosus Quality of Life Questionnaire (LEQoL) has been developed by Castellano-Rioja et.al (2020). This scale measures Lupus Erythematosus patient’s Quality of Life in five subscale: Physical Factor, Appearance factor, Emotional factors, Cognition factor, and Relationship factor(14).

This scale tried to eliminate the limitations that exist in other similar scales. In fact, it is a multifactorial tool that examines the quality of life in a wide range of patients' problems. It also has a holistic view of the physical injuries and physical activities of patients. Finally, this tool is not only for SLE and is used for all LE patients. To be placed. Therefore, the aim of the present study was to investigate the psychometric properties of the Persian version of LEQoL in the Iranian population with SLE.

Materials And Methods

Study design

This Cross-sectional translation and psychometric evaluation of the Persian version of Lupus Erythematosus Quality of Life Questionnaire (LEQoL) was conducted in 2021 at the Department of Clinical psychology, Kermanshah University of Medical Sciences (KUMS). Permission was obtained from Giménez-Espert (corresponding author of the original version of scale).

Study population

A cross-sectional study was done with a convenience sample. Participants included 172 patients with LE from Iran. From these 172 patients, 165 patients met the inclusion criteria for the study: diagnosed as systemic or cutaneous/discoid lupus erythematosus, and motivated to take participating in research” were included. All 165 patients were over 18 years old and signed an informed consent form. The data collection process occurred in March to April 2021.

Linguistic validation

The linguistic validation of the LEQoL from Spanish to Persian consisted of three steps: a forward translation, a backward translation, and cognitive interviews. Two independent, bilingual Persian-native speakers with medical backgrounds (psychologists) translated the original Spanish version into Persian. A consensus version was developed by discussion and revision of the translated versions by the authors. Backward translation of the consensus version was performed independently by one Spanish native speaker without a psychological background, who was familiar with the culture of the Spanish and translated language. Adequacy of the translated version was proven by comparison of the original with the backward-translated versions. Next, the patient examination was conducted. The personal interviews were conducted during which the interviewer asked whether the participant had any trouble in understanding the LEQoL and checked the participant’s interpretation of all items. Finally, proofreading was performed(15, 16).

Procedures

Initially a battery of tests including the Persian version of LEQoL, Symptom Checklist-90-R (SCL-90), and 36-Item Short Form Survey (SF-36). These tools were used to check validity.Then a Google Form containing all the mentioned questionnaires was created. Then a public announcement was spread to the Iranian population through the social networks Instagram and Telegram. Due to the low prevalence of LE, the notice asked individuals to call or send a message or send information or notifications if they have or know a person with lupus. The corresponding author of the article also referred to rheumatology clinics in Tehran (the capital of Iran, the most populous city in Iran) and asked clients to participate in this study. Finally, people with lupus were interviewed for 20 minutes. Regarding inclusion criteria, 165 people were included in the study. Then, the tools were provided to these people online and they were asked to send them to the researchers as soon as they completed them. The information will be collected confidentially and anonymously and the results will not be shared with others.

Instrument

Quality of Life of Patients with Lupus Erythematosus Instrument” (LEQoL): This scale consist of 21 items in five subscales: Physical Factor, Appearance factor, Emotional factors, Cognition factor, and Relationship factor. Results showed that original version of LEQoL has suitable psychometric properties(14). The higher scores indicate higher impairments in patients with Lupus. In this research we tried to assess psychometric properties of the Persian version of LEQoL.

Psychological symptoms: To evaluate the psychological symptoms of the patients, the SCL-90-R (Symptom Checklist-90-R) was utilized. The SCL-90-R evaluates the current level of the symptoms which occur during the last seven days. The aim of this scale is to carry out a brief evaluation of the type and severity of the patient's symptoms by self-assessment and has eight subscales, besides a total score (Sleeping Problems, Sensitivity, Anxiety, Depression, Obsessive-Compulsive, Somatization, Agoraphobia, and Hostility). We used the Persian manual for examining the SCL-90 data. Item scores range from zero (none) to four (severe). Cronbach's alpha's for the Iranian SCL-90 was 0.90(17).

General Quality of Life: For assessing the general quality of life in participants we used 36 Items Short Form Survey (SF-36). This scale consists of 36 items classified into eight subscales: role impairment due to physical health/role physical (4 items), mental health (5 items), energy and fatigue/vitality (4 questions), physical functioning (10 questions), social functioning (2 questions), general health (5 questions), body pain (2 items), and role impairment due to emotional health/role emotional (3 items). Two other general subscales are obtained by combining the subscales known as Mental Component Summary (MCS), and Physical Component Summary (PCS). In this scale, high scores described as higher levels of QoL, and vice versa. Psychometric researches showed that the Persian version of SF-36 has suitable validity and reliability in the Iranian population(18).

Statistical analysis

Data have been analyzed by SPSS V.25 and AMOS 26 software’s. In order to analyze the data using descriptive statistics (mean, standard deviation, frequency and percentage) as well as Cronbach's alpha method to assess internal consistency, the correlation of LEQoL with SF-36 and SCL-90 were used to assess convergent validity. Also, test-retest reliability (from 25 participants) with two week was assessed by intraclass correlation coefficient (ICC). Confirmatory factor analysis (CFA) was used to examine the adequacy of the resulting factor model. To assess model fit, this research utilized a range of incremental and absolute model fit indices, including the ratio of chi-square to degrees of freedom (X2/df), root mean square error of approximation (RMSEA),comparative fit index (CFI), The goodness of fit index (GFI), and S-Bχ2 (df). Exploratory factor analysis (EFA) was applied to examine instrument dimensionality. Kaiser-Meyer Olkin measure of sampling adequacy (KMO) and Bartlett’s test of sphericity were analyzed first to determine the suitability of the data to undergo factor analysis.

Results

Sample characteristics

The Iranian patients with lupus diagnosis (n = 165) had a mean age of 40.49 ± 13 (range 19 to 65) years. Near eleven percent (n = 17) were male, and 71.5% (n = 118) had suffered systemic lupus. Table 1 showed a comprehensive picture about Participants’ sociodemographic and clinical characteristics.

Table 1

Participants’ sociodemographic and clinical characteristics (n = 165)

Variable

Mean ± SD / N (%)

Gender

Male

17 (10.3%)

female

148(89.7%)

Education level

Under diploma

73(44.2%)

Diploma

62(37.6%)

University graduated

30(18.2%)

Age

 

40.49 ± 13

Lupus erythematosus type

Cutaneous

47(28.5%)

Systemic

118(71.5%)

Time to diagnosis

Less than six months

25(15.2%)

Six to twelve months

34(20.6%)

More than one year

106(64.2%)

Marital status

Single

23(13.9%)

Married

104(63%)

In a relationship

9(5.5%)

Other status/ prefer not to say

29(17.5%)

 

Internal Consistency

The Cronbach’s α value for the Persian version of the LEQoL was measured at 0.862, which implies satisfactory internal consistency. Table 2 showed that all items if removed (except 18th item), lead to decrease the value of alpha, which means that items are appropriate. The various dimensions showed alpha values of between 0.85 and 0.94 (Physical Factor α = 0.88, Appearance factor α = 0.85, Emotional factors α = 0.93, Cognition factor α = 0.85, Relationship factor α = 0.94).

Table 2

Item-Total Statistics

Items

Scale Mean if Item Deleted

Scale Variance if Item Deleted

Corrected Item-Total Correlation

Cronbach's Alpha if Item Deleted

LEQoL1

91.5939

457.011

.586

.851

LEQoL2

91.6121

466.849

.507

.854

LEQoL3

91.9030

471.295

.428

.857

LEQoL4

91.6909

469.020

.451

.856

LEQoL5

91.8303

466.312

.440

.856

LEQoL6

91.6364

461.794

.538

.853

LEQoL7

91.3636

459.452

.438

.857

LEQoL8

91.3818

452.579

.539

.853

LEQoL9

91.4909

465.386

.453

.856

LEQoL10

91.5333

458.214

.504

.854

LEQoL11

91.7879

463.863

.432

.857

LEQoL12

91.3152

466.778

.416

.857

LEQoL13

91.2000

462.478

.474

.855

LEQoL14

91.3697

460.966

.529

.853

LEQoL15

91.2606

461.182

.502

.854

LEQoL16

91.6485

483.266

.256

.863

LEQoL17

91.6545

475.520

.374

.859

LEQoL18

91.7212

485.910

.257

.863

LEQoL19

91.8545

474.015

.404

.858

LEQoL20

91.8061

473.974

.410

.858

LEQoL21

91.7394

475.877

.395

.858

 

Test Retest Reliability

To test-retest reliability evaluation, the 30 patients were chosen by the computer version of the random number table. They completed the scale a second time two weeks after the baseline assessment without being presented the scale they had previously filled out. The test-retest reliability coefficient was found to be 0.964 for the whole scale (Lower Bound: 0.925, Upper Bound: 0.983) and this measure was significant in 0.01 significance level (P < 0. 01).This measure indicate that this scale has strong test re-test reliability and is suitable for interventional studies.

Validity

In order to investigate the validity, the correlation between the LEQoL, SCL-90-R and SF-36were calculated. As shown in Table 3, some convergent-divergent correlations are significant at the 0.01 level. Therefore, LEQoL and its subscales have appropriate Validity levels.

Table 3. Pearson’s coefficient correlations among variables

SCL: SCL-90-R, OCD: obsessive-compulsive disorder, interpersonal: interpersonal sensitivity, phobic phobic anxiety, physical: Physical functioning, Role: Role limitations due to physical health, EP: Role limitations due to emotional problems, Emotional: Emotional well-being, Social: Social functioning, General: General health

Bolded correlations indicate the correlation is significant at the 0.01 level (2-tailed).

Underlined correlations indicate the correlation is significant at the 0.05 level (2-tailed).

Exploratory Factor Analysis (EFA)

The responses of patients (165) were utilized for the evaluation of the EFA of CSS. KMO test showed strong sampling adequacy (0.828). Bartlett's Test was significant (chi-square value = 2175.647, P-value < 0.001) that explained the items are correlated and factor analysis can be fitted. Latent factors were achieved by the principal components analysis. The extracted factors were rotated with the varimax way. This criterion suggested the five-factor, which explained 73.7% of the variance (Table 4 & Fig. 1). The pattern matrix of extracted factors is shown in Table 5.

Table 4

Extracted principal components with Eigenvalues, Cumulative percentage of explained variance

Component

Initial solution

Rotated solution

Cumulative % of variance explained

Eigenvalue

Cumulative % of variance explained

1

27.468

27.468

18.469

2

14.095

41.562

34.707

3

11.749

53.312

49.782

4

10.549

63.861

62.657

5

9.927

73.788

73.788

Table 4

Pattern matrix of scale

 

Component

1

2

3

4

5

LEQoL1

.807

       

LEQoL2

.786

       

LEQoL3

.758

       

LEQoL4

.768

       

LEQoL5

.816

       

LEQoL6

.766

       

LEQoL7

   

.810

   

LEQoL8

   

.781

   

LEQoL9

   

.745

   

LEQoL10

   

.744

   

LEQoL11

   

.783

   

LEQoL12

 

.913

     

LEQoL13

 

.921

     

LEQoL14

 

.899

     

LEQoL15

 

.879

     

LEQoL16

       

.889

LEQoL17

       

.876

LEQoL18

       

.842

LEQoL19

     

.927

 

LEQoL20

     

.922

 

LEQoL21

     

.941

 

Extraction Method: Principal Component Analysis.

Rotation Method: Varimax with Kaiser Normalization.

a. Rotation converged in 5 iterations.

 

Confirmatory factor analysis

To confirm the latent structure of CSS, the CFA was evaluated. The analysis was carried out on the whole sample (n = 160). Factor solution with two to five factors was modelled to check if the 5-factor is the best. Indicators of models were shown in Table 5. RMSEA and parsimony goodness of fit index (PGFI) for the 5-factor model are smaller than other models and CFI is larger. Also, the model has the best indexes that show the five-factor extracted model is the best representation of the structure of the data. Figure 2 showed the best model fit for five-factor solution.

Table 5

indicators to five-factor model about model fitness

No. of factors

χ2/df

CFI

Robust RMSEA (90% CI)

PGFI

5 factors

1.28

0.975

0.042 (0.024 0.057)

0.692

Discussion

The present paper aimed to translate and evaluate the psychometric properties of the Persian version of the LEQoL. As individuals with lupus are reportedly suffering from psychosocial and physical limitations due to their symptoms, it is important to adequately measure QoL to assess the effects of interventional researches. The present study found to support that a 21-item Persian version of LEQoL could be used to assess QoL in individuals who have lupus.

The procedure of forward-back translation was executed fluently in this investigation. Regarding the reliability of the scale, the total Cronbach's alpha coefficient had very well qualifies (α = 0.86), which was similar to the values in the original Spanish version (α = 0.92). All give subscales also had good and excellent internal consistency (ranging from 0.85–0.94, which was consistent with the original Spanish version (ranging from 0.82–0.92). Moreover, the findings of test-retest reliability for the overall scale (ICC = 0.93) and the five subscales (ranging between 0.82–0.92) indicated that the LEQoL possesses good stability over time, which was consistent with the original Spanish version.

In this research, the results from a CFA on data from 165 patients with lupus indicated that the psychometric properties of the LEQoL questionnaire are satisfactory. Specifically, most of the values used to evaluate the goodness-of‐fit are satisfactory (χ2/df = 1.28, RMSEA = 0.042, CFI = 0.975). Also, for EFA results showed that for Persian version of LEQoL with five-factor can explain 73.7% of the variances which is almost equal with the original scale EFA with 75.3%.

Moreover, we used divergent-convergent correlations for assessing validity. We found that every factor of LEQoL has correlation with some SF.36 and SCL-90-R subscales. This matrix indicate that the LEQoL subscales are somewhat independent, and this matter is one strength issue. Only depression (SCL-90 subscale) has significant correlation with LEQoL subscales which means the common problem for every LEQoL factor is depression and clinicians must pay attention to this issue for all patients with lupus. This result is in accordance with pervious researches. The research literature showed that every reduction in QoL has strong correlation with depression incidence. In fact, depression and QoL can affect each other(1921).

Besides existing other scales for assessing QoL in patients with lupus, this scale is multi-diversionary, and brief. Brief scales with adequate subscales and self-referred nature are more suitable rather than comprehensive, difficult, and unclear scales(14, 22). In Iran, because of poor insurance procedures and lack of an adequate number of clinicians all around the country, brief and clear scales can help patients to use self-assessment. Also, these scales can help clinicians to obtain a comprehensive view of patients' quality of life in a short time.

The current research besides important strengths has a number of limitations. First, the study is limited by the prevalent limitations in psychological research, including using voluntary participation and self-report scales. Second, notwithstanding the sufficiency of the sample size, it may not nationally representative (Iran consists of various cultures and lifestyles). Third, for the validity assessment of the LEQoL, we used convergent and construct (only confirmatory factor analysis) validity. Totally, future researches can evaluate other types of construct validity such as predictive, and discriminant. Also, future research can develop family-reported and clinician-reported of this scale.

Conclusion

The translated Persian version of the LEQoL scale used in this research is a reliable and consistent tool that showed suitable internal consistency and validity. Clinicians can assess Qol in patients with lupus by this scale, and following patients' progress during treatment. Also, researchers can assess their intervention efficacy by using this scale. Finally, the patients and their families can use LEQoL as a self-reported measure for a better understanding of their disease.

Abbreviations

EFA: Exploratory factor analysis, CFA: Confirmatory factor analysis, GFI: Goodness-of-fit index, AGFI: Adjusted goodness of fit index, CFI: Comparative fit index RMSEA: Root mean square error of approximation SRMR: Standardized root mean square residual, QoL: Quality of Life, LEQoL: Lupus Erythematosus Quality of Life Questionnaire.

Declarations

Acknowledgements:

The authors wish to thanks all participants for their assistance and participation in this study.

Funding:

None.

Competing interests:

The authors does not have any competing interests to declare.

Ethics approval and consent to participate:

All procedures used in collecting survey data on which this article relies on, are in accordance with the ethical standards of the Helsinki Declaration of 1964 and subsequent amendments or ethical standards. All data were collected anonymously, and no association could be established between the questionnaires and the responders.

Availability of data and materials:

Data are available upon reasonable request from the researchers.

Contributions:

M.D and S.M.A (Majid) have made substantial contributions to conception and design. S.R and Z.R collecting date. S.M.A (Mojtaba) analyzing data and drafting result subsection. All author take participate in Drafting the manuscript. Then, First author and corresponding author revised and submitted manuscript.

References

  1. Pamuk ON, Hasni S. Correspondence on 'Blood-brain barrier leakage in systemic lupus erythematosus is associated with gray matter loss and cognitive impairment'. Annals of the rheumatic diseases. 2021.
  2. Charras A, Smith E, Hedrich CM. Systemic Lupus Erythematosus in Children and Young People. Current rheumatology reports. 2021;23(3):20.
  3. Akbarian M, Faezi ST, Gharibdoost F, Shahram F, Nadji A, Jamshidi AR, et al. The epidemiology of systemic lupus erythematosus in Iran: a survey on 2143 cases. Tehran University Medical Journal. 2010;68(5).
  4. Singh JA, Shah NP, Mudano AS. Belimumab for systemic lupus erythematosus. The Cochrane database of systematic reviews. 2021;2:Cd010668.
  5. Seguier J, Jouve E, Bobot M, Whalen E, Dussol B, Gentile S, et al. Paradoxical association between blood modular interferon signatures and quality of life in patients with systemic lupus erythematosus. Rheumatology (Oxford, England). 2020;59(8):1975-83.
  6. Kusnanto K, Sari N, Harmayetty H, Efendi F, Gunawan J. Self-care model application to improve self-care agency, self-care activities, and quality of life in patients with systemic lupus erythematosus. Journal of Taibah University Medical Sciences. 2018;13(5):472-8.
  7. Applyrs DL, Williams EM, Faith TD, Kamen DL, Vazques E, Jurkowski JM. Cultural and quality-of-life considerations when administering corticosteroids as a therapeutic strategy for African American women living with systemic lupus erythematosus. Patient preference and adherence. 2018;12:1007-14.
  8. Huang HC, Liu CM, Liu CC. Psychiatric manifestations in systemic lupus erythematosus mimic psychotic prodrome: a case report. General hospital psychiatry. 2010;32(2):e3-4.
  9. Eldeiry D, Zandy M, Tayer-Shifman OE, Kwan A, Marzouk S, Su J, et al. Association between depression and anxiety with skin and musculoskeletal clinical phenotypes in systemic lupus erythematosus. Rheumatology (Oxford, England). 2020;59(11):3211-20.
  10. Iacobas D, Wen J, Iacobas S, Schwartz N, Putterman C. Remodeling of Neurotransmission, Chemokine, and PI3K-AKT Signaling Genomic Fabrics in Neuropsychiatric Systemic Lupus Erythematosus. Genes. 2021;12(2).
  11. McElhone K, Abbott J, Teh LS. A review of health related quality of life in systemic lupus erythematosus. Lupus. 2006;15(10):633-43.
  12. Doward LC, McKenna SP, Whalley D, Tennant A, Griffiths B, Emery P, et al. The development of the L-QoL: a quality-of-life instrument specific to systemic lupus erythematosus. Annals of the rheumatic diseases. 2009;68(2):196-200.
  13. Chren M-M, Lasek RJ, Flocke SA, Zyzanski SJ. Improved Discriminative and Evaluative Capability of a Refined Version of Skindex, a Quality-of-Life Instrument for Patients With Skin Diseases. Archives of Dermatology. 1997;133(11):1433-40.
  14. Castellano-Rioja E, Giménez-Espert MdC, Soto-Rubio A. Lupus Erythematosus Quality of Life Questionnaire (LEQoL): Development and Psychometric Properties. International Journal of Environmental Research and Public Health. 2020;17(22):8642.
  15. Cooper HE, Camic PM, Long DL, Panter A, Rindskopf DE, Sher KJ. APA handbook of research methods in psychology, Vol 1: Foundations, planning, measures, and psychometrics: American Psychological Association; 2012.
  16. Miles J, Gilbert P. A handbook of research methods for clinical and health psychology: Oxford University Press on Demand; 2005.
  17. Ardakani A, Seghatoleslam T, Habil H, Jameei F, Rashid R, Zahirodin A, et al. Construct validity of symptom checklist-90-revised (SCL-90-R) and general health questionnaire-28 (GHQ-28) in patients with drug addiction and diabetes, and normal population. Iranian journal of public health. 2016;45(4):451.
  18. Montazeri A, Goshtasebi A, Vahdaninia M, Gandek B. The Short Form Health Survey (SF-36): translation and validation study of the Iranian version. Quality of life research. 2005;14(3):875-82.
  19. Hunt SM, McKenna SP. The QLDS: a scale for the measurement of quality of life in depression. Health policy. 1992;22(3):307-19.
  20. Katschnig H, Jaidhauser K, Subasi B, Zoghlami A, Serim M. Quality of life in depression. European Psychiatry. 1997;12(S2):116s-s.
  21. Ay-Woan P, Sarah CP, LyInn C, Tsyr-Jang C, Ping-Chuan H. Quality of life in depression: predictive models. Quality of life research. 2006;15(1):39-48.
  22. Paniagua FA, Yamada A-M. Handbook of multicultural mental health: Assessment and treatment of diverse populations: Academic Press; 2013.