DASS-21 Measurement Properties: A Systematic Review

The Depression Anxiety Stress Scales–21 (DASS-21) is a concise, widely used, and freely available self-report instrument designed to measure emotional symptoms of depression, anxiety, and stress. Originating as a shorter version of the DASS-42, it was initially developed to assess negative emotional symptoms related to depression and anxiety, with a third construct—stress, encompassing irritability, tension, and agitation—empirically emerging during its development. Each of its three subscales (Depression, Anxiety, and Stress) contains seven items. The DASS-21’s broad utility has led to its application across diverse clinical and non-clinical populations in numerous countries, with translations available in 44 languages. However, its psychometric properties, particularly structural validity, have generated varied findings across studies, prompting the need for a comprehensive review.

A systematic review, conducted by Lee, Lee, and Moon (2019), aimed to thoroughly evaluate the measurement properties of the DASS-21 by applying the updated COnsensus-based Standards for selection of health Measurement INstruments (COSMIN) methodology. The study meticulously searched MEDLINE, Embase, and CINAHL databases, ultimately including 48 studies from 41 articles that met specific eligibility criteria, such as focusing on DASS-21 measurement properties in populations aged 14 years and older. The methodological quality of the included studies was assessed using the updated COSMIN Risk of Bias checklist, and the quality of evidence for measurement properties was rated using a modified GRADE approach, categorizing evidence as high, moderate, low, or very low.

Key Findings on DASS-21 Measurement Properties:

• Structural Validity: The review highlighted that the bifactor structure is the optimal model for the DASS-21, demonstrating sufficient high-quality evidence. This model posits that DASS-21 items load onto a general factor representing a “Negative Emotional state” (accounting for common variance) and also onto orthogonal group factors for Depression, Anxiety, and Stress subscales (explaining unique item covariance). While a three-factor structure also showed sufficient moderate-quality evidence, the bifactor model better captures the instrument’s underlying structure given the substantial interfactor correlations observed.
• Internal Consistency: Consistent with the bifactor structure, the DASS-21 exhibited sufficient high-quality evidence for internal consistency. For the bifactor structure, Cronbach’s alpha values ranged from 0.90–0.95 for the total scale, 0.82–0.92 for Depression, 0.74–0.88 for Anxiety, and 0.76–0.90 for Stress.
• Criterion Validity: The Depression subscale demonstrated sufficient high-quality evidence for criterion validity, with an Area Under the Curve (AUC) ranging from 0.77–0.91 when compared to the psychiatrist-administered Structured Clinical Interview for DSM-IV Axis 1 Diagnoses (SCID) for depression. However, the Anxiety subscale showed insufficient moderate-quality evidence.
• Hypothesis Testing for Construct Validity: The DASS-21 exhibited sufficient high-quality evidence for hypothesis testing of construct validity.
  • Convergent Validity was supported by high pooled correlation coefficients with established comparator instruments:
    • DASS-21 Depression correlated strongly with BDI (r=0.73), HADS Depression (r=0.69), and PANAS Negative Affect (r=0.56).
    • DASS-21 Anxiety correlated strongly with BAI (r=0.75), HADS Anxiety (r=0.66), and PANAS Negative Affect (r=0.55).
    • DASS-21 Stress correlated strongly with PANAS Negative Affect (r=0.66).
  • Known-groups validity was also well supported, with 100% of hypotheses supported for Depression and Anxiety, and 80% for Stress.
• Content Validity: Overall, there was sufficient moderate-quality evidence for the content validity of the DASS-21, particularly sufficient high-quality evidence for comprehensiveness and sufficient moderate-quality evidence for comprehensibility. However, the relevance of items for the construct, target population, and context of use showed only sufficient very-low-quality evidence, as this aspect was not adequately evaluated by experts or patients in the studies.
• Reliability: There was insufficient low-quality evidence for the reliability of each DASS-21 subscale, primarily due to inconsistent results from studies utilizing Pearson’s correlation coefficients and methodological concerns regarding the use of these correlations for instruments measuring states that fluctuate over time.
• Cross-cultural validity/measurement invariance: Measurement invariance across gender demonstrated inconsistent moderate-quality evidence for the three-factor structure. However, evidence for bifactor structure measurement invariance could not be adequately assessed due to limited studies.
• Responsiveness: The Depression and Stress subscales showed sufficient low-quality evidence for responsiveness, indicating significant changes in scores following interventions. However, the Anxiety subscale exhibited insufficient very-low-quality evidence, and the methodology (paired t-tests) used in many studies for assessing responsiveness was noted as inappropriate.

Practical Implications and Future Directions:

The DASS-21’s psychometric robustness and wide applicability make it a valuable tool for understanding negative emotional states in diverse populations. The review emphasizes that both the total score and the individual subscale scores should be calculated and interpreted separately to provide a comprehensive understanding of general negative emotional status as well as specific symptoms of depression, anxiety, and stress.

Despite its strengths, the review identified several areas for further research:

• Establishing cut-off points for the DASS-21 is crucial to enhance its practicality as a screening tool, as these are currently lacking unlike other established measures.
• Further validation studies regarding responsiveness are needed, particularly if the DASS-21 is to be widely used as an outcome measure in intervention studies.
• More research is required to evaluate measurement invariance reflecting a bifactor structure, especially across different genders or languages.
• Additional studies are also recommended to rigorously assess the reliability and measurement error of the DASS-21.
• Furthermore, evaluating the relevance component of content validity by experts and patients is strongly recommended to ensure the instrument’s items are truly pertinent to the construct and target population.
• Finally, applying the DASS-21 to populations younger than 14 years requires further dedicated validation studies.

In conclusion, the DASS-21 is a psychometrically sound instrument with robust evidence for its bifactor structural validity, internal consistency, construct validity, and criterion validity (particularly for the depression subscale), making it comparable to other well-known emotional symptom measures. Addressing the identified gaps will further strengthen its utility and ensure its continued effective application in research and clinical practice.

The Depression Anxiety Stress Scales–21 (DASS-21) is a concise, widely used, and freely available self-report instrument designed to measure emotional symptoms of depression, anxiety, and stress. Originating as a shorter version of the DASS-42, it was initially developed to assess negative emotional symptoms related to depression and anxiety, with a third construct—stress, encompassing irritability, tension, and agitation—empirically emerging during its development. Each of its three subscales (Depression, Anxiety, and Stress) contains seven items. The DASS-21’s broad utility has led to its application across diverse clinical and non-clinical populations in numerous countries, with translations available in 44 languages. However, its psychometric properties, particularly structural validity, have generated varied findings across studies, prompting the need for a comprehensive review.A systematic review, conducted by Lee, Lee, and Moon (2019), aimed to thoroughly evaluate the measurement properties of the DASS-21 by applying the updated COnsensus-based Standards for selection of health Measurement INstruments (COSMIN) methodology. The study meticulously searched MEDLINE, Embase, and CINAHL databases, ultimately including 48 studies from 41 articles that met specific eligibility criteria, such as focusing on DASS-21 measurement properties in populations aged 14 years and older. The methodological quality of the included studies was assessed using the updated COSMIN Risk of Bias checklist, and the quality of evidence for measurement properties was rated using a modified GRADE approach, categorizing evidence as high, moderate, low, or very low.

Reference:

Lee, J., Lee, E. H., & Moon, S. H. (2019). Systematic review of the measurement properties of the Depression Anxiety Stress Scales–21 by applying updated COSMIN methodology. Quality of Life Research, 28(7), 1735–1753. https://doi.org/10.1007/s11136-019-02177-x