Correlations among depression rating scales and a self-rating anxiety scale in depressive outpatients
Toru Uehara, M.D., Ph.D.,
Tetsuya Sato, M.D., Ph.D.,
Sakado, M.D., Ph.D.
We investigated relationships among the Hamilton Rating Scale for Depression (HRSD), the Beck Depression Inventory (BDI), the Inventory to Diagnose Depression (IDD), and the Self-rating Anxiety Scale (SAS) for 44 outpatients with major depression. The IDD showed significant correlation with the HRSD (r=.58), the BDI (r=.42), and the SAS (r=.39). The BDI showed significant correlation with the HRSD (r=.39) and the SAS (r=.40). Results from multiple regression analysis showed that the BDI and age were selected as the significant variables to correlate with the SAS (P<.05). Although self-rating depression scales, such as the BDI and IDD, are typically more influenced by anxiety than the objective rating(HRSD), the IDD may prove to be as useful and reliable as the HRSD for evaluating, relatively independent of anxiety, the severity of depressive symptoms.
key words: self-report scale, symptom, diagnose, depression, anxiety
Many rating scales for depression are currently in use, including the Hamilton Rating Scale for Depression (HRSD: Hamilton, 1960), the Beck Depression Inventory (BDI: Beck, 1979), the Zung Self-rating Depression Scale (Zung, 1965), and the Center for Epidemiological Studies Depression Scale (Radloff, 1977). The HRSD and the BDI are the most popular scales in Japan, just as they are in Western countries. Some of the scales for depression were initially developed only to evaluate the severity of symptoms, but have been misused for diagnosing depression. The Inventory to Diagnose Depression (IDD: Zimmerman et al., 1987) is a self-rating scale for use both with diagnosing major depression (MD) and evaluating the severity of symptoms, and the reliability and validity of the Japanese version of the IDD have already been confirmed (Uehara et al., 1995). Haaga et al. (1993) investigated the concurrent validity of the scale with a student sample, and they reported that the IDD showed a high positive correlation with the BDI, as did the Beck Anxiety Inventory (BAI: Beck et al., 1988). But they did not investigate the relationship between the IDD and other objective clinical ratings such as the HRSD. One purpose of the present study is to examine the inter-correlation among depression ratings (the HRSD as an objective clinical rating, and the BDI as a self-rating) including the IDD in a clinical sample. It has been suggested that anxiety symptomatically influences the diagnosis of depression (Beck & Clark, 1988), and in fact, symptom co-occurence in anxiety and depression frequently exists in non-clinical samples (Gottlib & Cane, 1989). In addition, most depression-rating scales include some items which are congruent to anxiety-rating scales. The IDD, which is based on diagnostic criteria of DSM-III-R, may be independent of anxiety. To clarify whether these rating scales measured depressive symptoms independent of anxiety symptoms, we also investigated the correlation of three depression scales and their demographic features with the Self-rating Anxiety Scale (SAS: Zung, 1971) using multiple regression analysis.
Subjects were selected from new patients who consecutively sought treatment at our psychiatric clinics at the Niigata City General Hospital from April 1996 to November 1996. Forty-four depressive outpatients were included because they were diagnosed as having a major depressive episode based on the DSM-IV at baseline. Our reliability to diagnose depression were high (kappa=.80, n=15) by our preliminary analysis. Subjects did not have sever physical complications, and they also agreed to cooperate in our study. The mean age of subjects was 42.3 (SD=10.7) years old (range from 18 to 61). There were 18 male (40.9%) and 26 female patients. Average education was 12.9 (SD=2.7) years. There were 19 mild and 25 moderate cases according to the DSM-IV (APA: 1994) criteria of major depression. Informed consent was obtained from all patients. To investigate the concurrent validity of popular depression-rating scales, we administered three depression ratings: the IDD (IDD: Zimmerman et al., 1987), the BDI (BDI: Beck, 1979), the 17 item HDRS (HRSD: Hamilton, 1960) at baseline. The IDD contains 22 items that include the entire range of symptoms used by DSM-III to diagnose major depression, which are as follows: low mood, decreased energy, psychomotor agitation, psychomotor retardation, decreased interest, decreased pleasure, decreased libido, guilt, worthlessness, suicidal thoughts, decreased concentration, indecisiveness, decreased appetite, weight loss, increased appetite, weight gain, insomnia, hypersomnia, anxiety, hopelessness, irritability and somatic complaints. One anxiety rating (the SAS: Zung, 1971) and the BDI were first given to all patients before the examination. After completing those scales, clinical evaluations and interview were conducted by authors (T.U., and K.S.) including one social functioning assessment (Global Assessment for Functioning; the GAF: APA, 1994) and the HRSD. If patients accepted to participate in our study, the IDD was carried out. We calculated Pearson's correlation coefficients among the total scores of these ratings. And to analyze the connection of anxiety to depression ratings while controlling for age, sex, and education, we used a multiple regression analysis. We used the "STATISTICA" program (Statsoft, Oklahoma) for statistical analyses.
Mean and SD of individual ratings were 38.0 and 12.1 on the IDD, 19.7 and 4.0 on the HRSD, 27.8 and 7.6 on the BDI, 45.8 and 7.6 on the SAS, and 55.9 and 9.3 on the GAF respectively. Table 1 shows the results of the simple correlation among the total scores of these ratings. The total score of the IDD showed strong correlation with those of the HRSD (r=.57, P=.001) and the BDI (r=.41, P=.006), and weak correlation with that of the SAS (r=.38, P=.010). The HRSD was correlated significantly with the BDI (r=.39, P=.009). The BDI showed significant correlation with the SAS (r=.40, P=.008). The GAF score was not correlated with any rating scales.
To investigate the direct relationship with anxiety controlling for the demographic variables, we conducted a multiple regression analysis. Our dependent variable was the SAS score. The IDD, the HRSD, the BDI, sex, age, and education were entered in as independent variables at first. Using stepwise method, we selected the BDI and age as independent variables in the final regression model (F=6.20, R=.58, df=3,40, P<.01). The beta coefficients of the BDI (.32 ) and age (-.29) were significant (P<.05) respectively.
Table 1: Pearson's coefficients among the rating scales
|*** = P<0.001, ** = P<0.01, * = P<0.05 df=42|
IDD; Inventory to diagnose depression
HRSD; Hamilton rating scale for depression
BDI; Beck depression inventory
SAS; Self-rating anxiety scale
GAF; Global assessment of functioning
We here consider that the total score of the IDD correlated significantly with the depression scales HRSD (r=.57) and the BDI (r=.41). The IDD, which was developed for diagnosing MD, seems to be as useful as a subjective, or even an objective rating for evaluating the severity of depressive symptoms. The HRSD was correlated with the BDI significantly (r=.39), but this was weaker than what we typically expected. On the other hand, the GAF did not correlate with any of the rating scales. This result would seem to indicated that the GAF score not only reflects symptoms, but also social function and interpersonal relationships. Many depression-rating scales have items which also evaluate anxiety symptoms, or cooperate with the anxiety-rating scale, and depression often combines with anxiety (Gottlib, 1984; Luteijin & Bouman, 1988). Actually, the IDD was correlated with the SAS (P=.01). And the BDI, which is a self-report of depression like the IDD, also correlated with the SAS (P<.01); however, the HRSD did not show significant correlation. This means that the self-rating depression scales may be more influenced by anxiety than the objective ratings. However, controlling for demographics, only the BDI was correlated with the SAS significantly by multiple regression analysis. Haaga et al. (1993) showed that the IDD highly correlated with the BDI (r=.88) and the BAI (BAI: Beck et al., 1988) (r=.67) for 115 students. And these authors also reported that the BDI correlated strongly with the BAI (r=.70) same as the IDD. These findings and the results of our study lead us to think that self-rating depression scales are influenced by anxiety in clinical and non-clinical subjects, but the IDD may be able to evaluate depressive symptoms more independently from anxiety than the BDI especially in a clinical sample. Our study is unique in its report of usefulness and relatively independence from anxiety of the IDD. However, our data should be viewed with caution, since our sample is small, and we used only a self-rating to measure anxiety symptoms. In addition, our results might be affected by the reliability with which major depression was diagnosed and the HRSD was administered. We will present a detailed reliability study in a separate report.
American Psychiatric Association (1994) Diagnostic and Statistical Manual for Mental Disorders Fourth Edition . Washington, D.C : APA.
Beck AT, Epstein N, Brown G, Steer RA (1988) An inventory for measuring clinical anxiety: Psychometric properties. Journal of Consultation and Clinical Psychology 56,893-897.
Beck AT, Rush AJ, Shaw BF, Emery G (1979) Cognitive therapy for depression. New York, Guilford.
Beck AT, Clark DA (1988) Anxiety and depression: An information processing perspective. Anxiety Research: An International Journal 1,23-36.
Gottlib IH & Cane DB (1989) Self-report assessment of depression and anxiety. In Anxiety and Depression: Distinctive and overlapping features (eds. Kendall PC & Watson D). pp131-169, San Diego, Academic.
Gottlib IH (1984) Depression and general psychopathology in university students. Journal of Abnormal Psychology 93,19-30.
Haaga AFD, McDermut W, Ahrens AH (1993) Discriminant validity of the inventory to diagnose depression. Journal of Personality Assessment 60,285-289.
Hamilton M (1960) A rating scale for depression. Journal of Neurology Neurosurgery and Psychiatry 23,56-62.
Luteijin F & Bouman TK (1988) The concepts of depression, anxiety and neuroticism in questionnaires. European Journal of Personality 2,113-120.
Radloff LS (1977) The CES-D scale: a self-report depression scale for research in the general population. Applied Psychological Measurement 1,385-401.
Uehara T, Sakado K, Sato T, Sato S, Kawashima Y, Machizawa S
(1995) The reliability and validity of the Japanese version of
the inventory to diagnose depression (IDD). Japanese Journal of
Zimmerman M, Coryell M, Corenthal C, Wilson S (1986) A self-report scale to diagnose major depressive disorder. Archives of General Psychiatry 43,1076-1086.
Zung WWK (1971) A rating instrument for anxiety disorders. Psychosomatics 12,371-379.
Zung WWK (1965) A self-rating depression scale. Archives of General Psychiatry 12, 63-70.
From the Department of Psychiatry, Niigata
University School of Medicine
(T.U., K.S.), the Department of Psychiatry, Niigata City General Hospital
(T.U., K.S.), and the Department of Psychiatry, Fujita Health University
School of Medicine (T.S.).
Address for correspondence and reprint requests :
T. Uehara, Department of
Psychiatry, Niigata University School of Medicine, Asahimachi 1-757,
Niigata 951, Japan.
Tel: 81-25-223-6161(Ext. 2632), Fax: 81-25-223-3882