Brief report
The Calgary Depression Rating Scale for schizophrenia in a healthy control group: Psychometric properties and reference values

https://doi.org/10.1016/j.jad.2005.04.005Get rights and content

Abstract

Background

Assessment of depression in schizophrenia is of great importance as depressive signs and symptoms and suicidality are highly prevalent in patients with schizophrenic disorders. The Calgary Depression Rating Scale (CDSS) is the standard assessment instrument for that purpose due to its proven reliability and validity. However, so far no reference values derived from an adequate healthy sample have been published.

Methods

The present study analyzed CDSS item scores and summary scores in 154 healthy subjects (49% female, mean age 32.8 ± 11.7 years) selected from studies as controls for schizophrenic patients.

Results

The total CDSS score was 2.6 ± 2.7 (range 0–12), about 1/3 of healthy subjects had scores = 0. Gender differences were observed with respect to “early awakening” and “suicidal ideation” with higher scores in females. Total CDSS scores (r =  0.17, P = 0.03) and single items were slightly negatively correlated with age. The psychometric properties of the CDSS scale were satisfying (corrected item-score correlations rtc = 0.33–0.61, internal consistency Cronbach's α = 0.76, one-dimensional factor structure). CDSS total scores showed a skewed distribution fitting the upper part of a normal distribution. Based on the empirical distribution of CDSS values, reference values were listed for the total group of healthy subjects, and for females and males separately.

Conclusions

The results corroborate the favourable psychometric properties of the CDSS found previously in clinical samples in a healthy control group. The reference values reported here can assist the clinical use and interpretation of CDSS values and emphasize the usefulness of the CDSS for a thorough assessment of depression and suicidality in schizophrenia.

Introduction

The Calgary Depression Rating Scale (CDSS) (Addington et al., 1990) is recommended to estimate the severity of depression in patients with schizophrenia (Collaborative Working Group on Clinical Trial Evaluations, 1998) and shows superiority compared to the Hamilton Depression Rating Scale (HAMD), the Montgomery-Asberg Depression Rating Scale (MADRS), and other depression scales (Addington et al., 1996, Collins et al., 1996, Reine et al., 1998, El Yazaji et al., 2002).

The superiority of the CDSS originates from its psychometric properties, particularly internal consistency, inter-rater reliability, sensitivity, and specificity (Lancon et al., 1999, Müller et al., 1999, Reine et al., 2000, Sarro et al., 2004). Discriminant and convergent validity have been proven in a large number of studies (e.g. Bernard et al., 1998, Sarro et al., 2004). CDSS total scores show in most settings only low to modest overlap with positive and negative symptoms, and no substantial correlation with extrapyramidal symptoms (Addington et al., 1994, Lancon et al., 1999, Kontaxakis et al., 2000a, Kontaxakis et al., 2000b, Lancon et al., 2000, Müller, 2002, Müller et al., 2002) during acute schizophrenia; in chronic illness states the overlap between positive and depressive symptoms seems somewhat higher (Lancon et al., 2001).

Converging results have shown that cut-off scores of 4–5 points for minor depression and 6–7 points for major depression show diagnostic congruence of the CDSS with conventional classification manuals (DSM-IV) (Bressan et al., 1998, Sarro et al., 2004). As substantial depression seems to be prevalent in 50% of schizophrenic patients (Siris, 1995), for an adequate interpretation of CDSS scores in schizophrenia the score distribution in clinically non-depressed and non-psychotic subjects should be considered. Somehow different to positive and—to a lesser degree-negative symptoms, single depressive symptoms are rather frequent in the general population and depression severity scores seem to be continuously distributed according to a quantitative gene-environment interaction effect (Lesch, 2004).

Section snippets

Design

154 healthy subjects were recruited as controls from four yet unpublished studies on schizophrenia and the validation of the CDSS. All subjects gave informed consent prior to the study. All studies were approved by the local ethics committee.

Participants and assessment

An extensive psychiatric interview and a medical examination took place. The psychiatric assessment comprised a diagnostic interview according to DSM-IV and ICD-10 (DIPS) and standardized rating scales. Subjects with a lifetime diagnosis of affective or

Results

Mean age (32.8 ± 11.7 years, range 18–73 years), and education (mean 11.0 ± 1.7 years, range 8–13 years) were comparable in males and females (48.7% female, n = 75). Table 1 and Fig. 1 present results of descriptive analyses, gender differences, and age-dependence of CDSS items and total scores. Total CDSS scores and single item scores were negatively correlated with age, i.e. higher depression scores were found in younger subjects. Correlation coefficients reached only values of r < 0.30 indicating

Discussion

The study provides psychometric properties and reference values of the CDSS in healthy subjects with low mean depression severity. The nine items of the CDSS comprise a highly sensitive, one-dimensional scale with high internal consistency in healthy subjects corroborating findings in schizophrenic samples (Addington et al., 1992, Addington et al., 1993, Addington et al., 1996, Lancon et al., 1999, Müller et al., 1999). Gender differences were only marginal, and no substantial effect of age and

References (27)

Cited by (0)

View full text