Validation of Zung's Self-rating Depression Scale among the Colombian general population
Main Article Content
In this study we aimed to validate Zung’s Self-rating Depression Scale (SDS; 1965) among Colombian people living in Bucaramanga, Colombia. Although used frequently in Colombian investigations to identify depressive disorders, the SDS had not been validated formally among the general Colombian population. Participants were a random sample of people dwelling in an urban area, with a mean age of 37.4 years (SD = 12.7). Participants filled out the SDS, and were then interviewed by psychiatrists using the Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I; First, Spitzer, Gibbon, & Williams, 1999) to diagnosis a major depressive episode (MDE) during the last month. Forty was taken as a cut-off point. SDS scores ranged from 21 to 62 (M = 36.5, SD = 9.1). Using the SDS, 95 (35.7%) persons reported clinically meaningful depressive symptoms. The SCID-I interview identified 44 (16.5%) persons with MDE. Cronbach’s alpha was 0.832. The sensitivity was 88.6% (95% CI 74.6-95.7), the specificity 74.8% (95% CI 68.4-80.2), the positive predictive value 41.1% (95% CI 31.2-51.6), the negative predict value 97.1% (95% CI 92.9-98.9), half Cohen’s kappa coefficient 0.433 (95% CI 0.327-0.539), and area under ROC curve 0.901 (95% CI 0.857-0.945). The SDS was found to be a useful tool for screening MDE among the general community.
The major depressive disorder is the most common mental disorder among the Colombian general population (WHO World Mental Survey Consortium, 2004); however, many patients do not request medical assistance (Posada-Villa, Aguilar-Gaxiola, Magaña, & Gómez, 2004), or they fail to be diagnosed properly when they attend for medical evaluation (Kessler, Lloyd, Lewis, & Pereira, 1999).
The Zung Self-rating Depression Scale (SDS) was designed to establish the severity of a major depressive episode in clinical settings (Zung, 1965). In spite of this, the SDS has been shown to be a useful tool for screening clinically meaningful depressive symptoms among people dwelling in the community. Making use of the depression rating scale as a tool for identifying persons with a current major depression disorder is a practical, rapid and economic strategy in different contexts (Pignone et al., 2002); although in Colombia the SDS is the most popular screening depression scale applied in several studies (Amézquita, González, & Zuluaga, 2003; Arboleda, Gutiérrez, & Miranda, 2001; Campo-Cabal & Gutiérrez, 2001; González, Campo, & Haydar, 1999; Miranda, Gutiérrez, Bernal, & Escobar, 2000; Pareja-Ángel & Campo-Arias, 2004; Posada & Torres, 1995), just one study has involved validation among university students (Díaz, Campo, Rueda, & Barros, 2005). In addition, it is necessary to keep in mind that sociocultural and linguistic factors may play an important role in accurate answers (Rubio-Stipec, Hicks, & Tsuang, 2002), and that the cut-off point must be adjusted according to the investigated group; the cut-off point has to be different in order to achieve good sensitivity and specificity (Blacker & Endicott, 2002). Validation among Colombian university students suggested that a good level of sensibility and specificity is obtained taking a cut-off point 13 points higher than the traditional 40 (Díaz, Campo, Rueda, & Barros, 2005). This point has been reported and suggested in other studies in different settings (Aragonés, Masdeu, Cando, & Coll, 2001; Dugan, McDonald, Passik, Rosenfeld, Theobald, & Edgerton, 1998; Zung, MacDonald, & Zung, 1988).
It is very important to be able to rely on free use of an easy and quick scale permitting the identification of possible depression cases in people living in the community. This method could be useful for monitoring the incidence and prevalence of depressive disorders, and for improving the incidence of seeking professional help. Some empirical observations have suggested that these screenings have increased self-recognition of symptoms, and frequency of medical attendance (Greenfield et al., 1997). Many persons do not receive treatment because of several myths about depressive symptoms or sometimes because physicians do not make an accurate diagnosis when patients report these symptoms (Kessler, Llyod, Lewis, & Pereira, 1999).
Since the SDS has not been validated formally among any Colombian population other than female university students (Campo-Arias, Diaz-Martinez, Rueda-Jaimes, & Barros-Bemúdez, 2005), the objective in this investigation was validating the SDS in a general population from Colombia.
Method
This is a criterion validity study of a screening test, conducted among the general population of Bucaramanga, Colombia, during 2004. The Review Board from Universidad Autónoma de Bucaramanga approved this study and participants signed an informed consent establishing the objectives and that there was minimum risk for the study subjects (Ministerio de Salud de Colombia, 1993).
A random general population sample was evaluated. The estimated prevalence of current major depressive disorders was 15%. First, blocks were selected, then a house in each block, and finally one person per home. All selections were done employing a random number list taken from Epi Info (Dean et al., 1997). Two hundred and ninety-nine people were asked to participate in the research; however, 33 people refused (11.0%). Finally, we interviewed 266 people aged from 18 to 65 years old, the average age was 37.4 years (SD = 12.7); and the formal education ranged from 0 to 25 years, the mean was 9.8 years (SD = 4.7). Other demographic characteristics of the study sample appear in Table 1. This sample represents the adult Colombian population.
Table 1. Demographic Characteristics of 266 Persons Interviewed for a Major Depressive Episode
The SDS is a 20-item Likert-type scale which is filled out writing clearly and precisely on the symptom frequency during the last two weeks. Each item gets 1 to 4 points, and the total score ranges from 20 to 80. The original version used 40 points as a cut-off point for clinically meaningful depressive symptoms (Zung, 1965). To compare results on the SDS, the section for a current major depressive disorder from the Spanish Structured Clinical Interview for DSM-IV Axis I disorders, Clinical Version, was applied by experienced psychiatrists who had been trained to conduct this interview (First, Spitzer, Gibbon, & Williams, 1999).
All participants received instructions to complete the SDS given by research assistants and, after that, they were interviewed by a psychiatrist blinded to SDS results; this interview was taken as a gold standard. Persons who met the criteria for a major depressive disorder or other mental disorders were referred to their own medical services.
Internal consistency was established using Cronbach’s alpha test (Cronbach, 1951). Sensitivity, specificity, positive predictive value, negative predictive value, and likelihood ratio for a positive or negative test result were calculated. Also, concordance was determined using half Cohen’s kappa coefficient (Cohen, 1960). Finally, the best cut-off point was found making a receiver operating characteristics (ROC) curve (Rey, Morris-Yates, & Stanislaw, 1992). All statistical tests were done in STATA 8.0 (Stata Corporation, 2003).
Results
SDS scores ranged from 21 to 62, mean 36.5 (SD = 9.1); 95 (35.7%) of the people reported clinically meaningful depressive symptoms. On the other hand, the psychiatric structured interview identified 44 (16.5%) persons who met the criteria for a current major depressive episode.
The Cronbach’s alpha value was 0.832. Using the traditional cut-off point of 40, sensitivity was 88.6% (95% CI 74.6-95.7); specificity, 74.8% (95% CI 68.4-80.2); positive predictive value, 41.1% (95% CI 31.2-51.6); negative predictive value, 97.1% (95% CI 92.9-98.9); likelihood ratio for a positive test result, 3.52; likelihood ratio for a negative test result, 0.15; and half Cohen’s kappa coefficient, 0.433 (95% CI 0.327-0.539). Under the ROC curve area was 0.901 (95% CI 0.857-0.945); this curve is presented in Figure 1. According to the ROC curve, the cut-off point that best classified persons who had a major depressive episode was 49; with sensitivity, 50.0% (95% CI 34.8-65.2); specificity, 94.6% (95% CI 90.5-97.0); positive predict value, 64.7% (95% CI 46.5-79.7); negative predict value, 90.5% (95% CI 85.8-93.8); likelihood ratio for a positive test result, 9.26; likelihood ratio for a negative test result, 0.53; and half Cohen’s kappa coefficient, 0.491 (95% CI 0.372-0.609).
Figure 1. ROC Curve for SDS among 266 adults from the general population.
Discussion
This investigation shows that SDS screening is a useful tool for identifying meaningful depressive symptoms. The SDS presents a high sensibility, good specificity, and good concordance with the psychiatric interview.
The SDS is a well-known scale for detecting depressive symptoms among primary-care patients and medical inpatients (Aragonés, Masdeu, Cando, & Coll, 2001; Dugan et al., 1998). However, there was no previously published validation with a Colombian general population. Data suggest that the SDS could be used for identifying clinically meaningful depressive symptoms among the general community relying on its psychometric properties in clinical and primary-care samples (Greenfield et al., 1997; Posada & Torres, 1993).
Several Colombian studies have used the SDS without adjusting a cut-off point and reporting an unusually high prevalence of depression instead of prevalence of clinically meaningful depressive symptoms. A general population-based survey reported a “prevalence of depression” of 26.7% (Posada & Torres, 1995); there were similar findings (27.2%) among psychology students (González, Campo, & Haydar, 1999); 33.3% to 36.4% was reported in medical students (Arboleda, Gutiérrez, & Miranda, 2001; Miranda, Gutiérrez, Bernal, & Escobar, 2000); 38.1% to 49.8% was documented in university students (Amézquita, González, & Zuluaga, 2003; Campo-Cabal & Gutiérrez, 2001); and 52.7% among primary-care patients suffering from migraine (Pareja-Ángel & Campo-Arias, 2004). These “prevalences of depression” are two- to five-fold higher than prevalence found when a structured interview is used. The prevalence of a major depressive episode was 2.0% among the general population, according to the Composite International Diagnostic Interview, which is applied by lay interviewers (WHO World Mental Survey Consortium, 2004); while 18.8% was reported among university students using the clinical version of the Structured Clinical Interview for DSM-IV Axis I disorders applied by psychiatrists (Díaz, Campo, Rueda, & Barros, 2005). These discrepancies confirm that validating is always necessary to vouch for using a screening scale in each population (Blacker & Endicott, 2002). Moreover, this investigation suggests that researchers must modify and recognize the best cut-off point depending on the population setting (Blacker & Endicott, 2002; Rubio-Stipec, Hicks, & Tsuang, 2002).
In addition, the validation of a scale like the SDS is very important because it is accepted that the report of depressive symptoms changes in the same way as does the sociocultural context. For instance, one study found that non-Hispanic whites were more likely than were Mexican-American (Latin Americans) people to report depressed mood, appetite disturbances, sleep problems, fatigue, death ideation, decreased sexual interest, feeling of worthlessness, and difficulty concentrating (Golding, Karno, & Rutter, 1990). Many people in developing countries do not consider the major depressive disorder as a mental disease but they believe it is an indicator of psychological weakness (Manson, 1995).
Validating a scale among the general population is an expensive and hard task, which explains the delayed validation of the SDS. Nevertheless, it is very important to be sure about what a scale like the SDS is measuring in people dwelling in the community. At present, we know that the SDS is very useful to identify clinically meaningful depressive symptoms among the Colombian general population. In addition the SDS is an easy and cheap way to screen depressive symptoms in a large population, which is thought to improve the number of people who seek professional treatment for clinically meaningful depressive symptoms (Greenfield et al., 1997). Furthermore, a recent investigation reported that self-report depression scales are also useful for following up patients and for evaluating treatment response (Zimmerman, Posternak, & Chelminski, 2004).
We conclude that the SDS is useful in identifying clinically meaningful depressive symptoms in the Colombian general population using the traditional cut-off point. The SDS needs to be validated with other groups such as school students of both sexes, and male university students, also inpatients and outpatients with a medical or psychiatric condition.
Amézquita, M., E., González, R. E., & Zuluaga, D. (2003). Prevalence of depression, anxiety, and suicidal behaviors in pregrade students of the Universidad de Caldas, year 2000 [In Spanish]. Revista Colombiana de Psiquiatría, 32, 341-356.
Aragonés, E., Masdeu, R. M., Cando, G., & Coll, G. (2001). Diagnosis validity of the Self-rating Depression Scale of Zung primary care patients [In Spanish]. Actas Españolas de Psiquiatría, 29, 310-316.
Arboleda, A., Gutiérrez, J. C., & Miranda, C. A. (2001). Prevalence of depressive symptoms in medical students and residents of the Universidad del Valle, 1999 [In Spanish]. Médicas UIS, 15, 4-7.
Blacker, D., & Endicott, J. (2002). Psychometric properties: Concepts of reliability and validity. In A. J. Rush, H. A. Pincus, & M. B. First (Eds.), Handbook of psychiatric measures. Washington, DC: American Psychiatric Association.
Campo-Arias, A., Diaz-Martinez, L. A., Rueda-Jaimes, G. E., & Barros-Bermúdez, J. A. (2005). Validation of Zung’s Self-rating Depression Scale among female university students from Bucaramanga, Colombia [In Spanish]. Revista Colombiana Psiquiatria, 34, 54-62.
Campo-Cabal, G., & Gutiérrez, J. C. (2001). Psychopathology among university students of the Facultad de Salud, Univalle [In Spanish]. Revista Colombiana de Psiquiatría, 30, 351-358.
Cohen, J. (1960). A coefficient of agreement for nominal scales. Educational & Psychological Measurement, 20, 27-46.
Cronbach, L. J. (1951). Coefficient alpha and the internal structure of test. Psychometrika, 16, 297-334. Dean, A. G., Dean, J. A., Coulombier, D., Burton, A. H., Brendel, K. A., Smith, D. C., et al. (1997).
Epi Info 6.04c. A word processing, database, and statistic program for public health. Center for Disease Control and Prevention (Atlanta, USA), and World Health Organization (Geneva, Switzerland).
Díaz, L. A., Campo, A., Rueda, G., & Barros, J. A. (2005). Proposing a short version of Zung’s Self- rating Depression Scale [In Spanish]. Colombia Médica, 36, 168-172.
Dugan, W., McDonald, M. V., Passik, S. D., Rosenfeld, B. D., Theobald, D., & Edgerton, S. (1998). Use of the Zung Self-rating Depression Scale in cancer patients: Feasibility as a screening tool. Psychooncology, 7, 483-493.
First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. W. (1999). Structured clinical interview for DSM-IV axis I diagnosis - clinical version [In Spanish]. Barcelona; Masson.
Golding, J. M., Karno, M., & Rutter, C. M. (1990). Symptoms of major depression among Mexican-Americans and Non-Hispanic Whites. American Journal of Psychiatry, 147, 861-866.
González, A., Campo, A., & Haydar, R. (1999). Depressive symptoms among psychology students. [In Spanish]. Acta Psiquiátrica y Psicológica de América Latina, 45, 234-239.
Greenfield, S. F., Reizes, J.M., Magruder, K.M., Muenz, L. R., Kopans, B., & Jacobs, D. G. (1997). Effectiveness of community-based screening for depression. American Journal of Psychiatry, 154, 1391-1397.
Kessler, D., Lloyd, K., Lewis, G., & Pereira, D. (1999). Cross sectional study of symptom attribution and recognition of depression and anxiety in primary care. British Medical Journal, 318, 436-440.
Manson, S. (1995). Culture and major depression. Psychiatric Clinic of North America, 18, 487-501.
Ministerio de Salud, Colombia. (1993). Resolution 008430 de 1993, for establishing scientific, technical, and administrative norms for health investigation [In Spanish]. Santafé de Bogotá: Ministerio de Salud, Colombia.
Miranda, C. A., Gutiérrez, J. C., Bernal, F., & Escobar, C. A. (2000). Prevalence of depression in medicine students of the Universidad del Valle [In Spanish]. Revista Colombiana de Psiquiatría, 29, 251-260.
Pareja-Ángel, J., & Campo-Arias, A. (2004). The prevalence of symptoms of anxiety and depression in female migraine sufferers [In Spanish]. Revista de Neurología, 39, 711-714.
Pignone, M. P., Gaynes, B. N., Rushton, J. L., Burchell, C. M., Orleans, C. T., Mulrow, C. D., et al. (2002). Screening for depression in adults: A summary of the evidence for the U. S. Preventive Services Task Force. Annals of Internal Medicine, 136, 765-776.
Posada-Villa, J. A., Aguilar-Gaxiola, S. A., Magaña, C. G., & Gómez, L. C. (2004). Prevalence of mental disorders and use of service: Preliminary results from a National Study of Mental Health, Colombia, 2003 [In Spanish]. Revista Colombiana de Psiquiatría, 33, 241-262.
Posada, J. A., & Torres, Y. (1995). National study of mental health and substance use Colombia 1993 [In Spanish]. Bogotá: Ministerio de Salud.
Rey, J. M., Morris-Yates, A., & Stanislaw, H. (1992). Measuring the accuracy of a diagnostic test using receiver operating characteristics (ROC) analysis. International Journal of Methods in Psychiatric Research, 2, 39-50.
Rubio-Stipec, M., Hicks, M. H. R., & Tsuang, M. T. (2002). Cultural factors influencing the selection, use, and interpretation of psychiatric measures. In: A. J. Rush, H. A. Pincus, & M. B. First (Eds.), Handbook of psychiatric measures. Washington, DC: American Psychiatric Association.
State Corporation (2002). STATA for Windows 8.0. College Station, Texas: Stata Corporation.
WHO World Mental Health Survey Consortium (2004). Prevalence, severity, and unmet need for treatment of mental disorders in the World Health Organization World Mental Health Surveys. Journal of the American Medical Association, 291, 2581-2590.
Zimmerman, M., Posternak, M. A., & Chelminski, I. (2004). Using a self-report depression scale to identify remission in depressed outpatients. American Journal of Psychiatry, 161, 1911-1913.
Zung, W. W. K. (1965). Self-report depression scale. Archive of General Psychiatry, 12, 63-70.
Zung, W. W. K., MacDonald, J., & Zung, E. M. (1988). Prevalence of clinically significant depressive symptoms in black and white patients in a family practice setting. American Journal of Psychiatry, 145, 882-883.
Table 1. Demographic Characteristics of 266 Persons Interviewed for a Major Depressive Episode
Figure 1. ROC Curve for SDS among 266 adults from the general population.
The authors thank Liliana Fabiola Ruiz and all collaborating medical students for help with this research. The study was supported by a grant from the Centro de Investigaciones Biomé
dicas of the Universidad Autó
noma de Bucaramanga (code 2105).
Appreciation is due to reviewers including
Erika Szadoczky
National Institute of Psychiatry and Neurology
Huvosvolgyi ut 116
Budapest
Hungary 1021. Email