Stress Research Institute

The Symptom Checklist-core depression (SCL-CD) scale: validation of a brief six item scale for the assessment of major depression in population surveys

BackgroundIt is well known that major depressive disorders are common and have sub-stantial negative impact on individuals and society. Still, there is no generally accepted, convenient and homoge-nous scale for the measurement of major depression in larger population surveys.

MethodsThis study examines the psychometric qualities of a brief dimensional scale based on (Hopkins) Symptom Check-list 90 (SCL-90) covering 6 core symp-toms of depressive disorder, referred to as the Symptom Checklist-core depres-

CONTACT Linda Magnusson Hanson, Stress Research Institute, Stockholm University

E-mail linda.hanson@stressforskning.su.se Phone +46 8 5537 8916

Linda L Magnusson Hanson1, Hugo Westerlund1,2, Reiner Rugulies3-5, Walter Osika1, Constanze Leineweber1, Tores Theorell1, Per Bech6 (1) Stress Research Institute, Stockholm University, Stockholm, Sweden (2) Department of Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden (3) National Research Centre for the Working Environment, Copenhagen, Denmark (4) Department of Public Health, University of Copenhagen, Denmark (5) Department of Psychology, University of Copenhagen, Denmark (6) Psychiatric Research Unit, Mental Health Centre North Zealand, Hillerød, Denmark

Stress Research Institute is a knowledge centre in the area of stress and health. The Institute is part of the Faculty of Social Science, Stockholm University, Sweden and conducts basic and applied research on multidisciplinary and interdisciplinary methodological approaches. E-mail info@stressforskning.su.se Website www.stressresearch.se

ResultsThe scale showed a high internal consistency and homogeneity, as well as high correlation with Center for Epidemiological Depression Scale (CES-D) and MDI. Results for the SCL-CD were generally com-parable or slightly better than for CES-D. In ROC analysis, AUC was 0.97, indicating excellent accuracy, and 17 (on a scale from 0-24) was identified as the best cutpoint for major depression with a sensiti-vity of 0.68 and specificity of 0.98. The scale predicted subsequent purchase of antidepressants and hospitalization with a depressive episode.

ConclusionOn the basis of self-reports and agreement with the MDI as the cri-terion instrument, the brief SCL-CD appears to be a valid, and unidi-mensional depression scale. A cutpoint suitable for epidemiological research was determined which should be validated in other set-tings and against psychiatric diagnosis.

Results & Conclusions

The scale showed a high internal consistency and homogeneity, high correlation with other depression scales, and appears to be valid and unidimensional scale.

sion (SCL-CD) scale. We assessed unidi-mensionality by exploratory factor analyses concurrent/criterion and con-struct validity by means of correlations with related scales and performed re-ceiver operating characteristics (ROC) analysis in a random sample of 1476 residents in Stockholm County aged 18-64 years. Weighted kappa statistics was used to derive a cutpoint, indicati-ve of major depression for epidemiolo-gical research, using Major Depression Inventory (MDI) severity score (>26) as criterion measure. Predictive validity was further assessed, based on a sam-ple of 5985 participants in the Swedish Longitudinal Occupational Survey of Health (SLOSH).

Table 1. Predictive validity

Use of antidepressants according to the Prescribed drug register

Risk of hospitalization with a depressive episode

OR 95 % CI OR 95 % CI

Major depression according to the SCL-CD >17

4.84 3.66-6.40 12.2 3.40-43.4