Hospital Anxiety and Depression Scale

Abstract

The Hospital Anxiety and Depression Scale (HADS) is an internationally recognized, 14-item self-report screening instrument originally designed by A. S. Zigmond and R. P. Snaith in 1983 to assess clinically significant states of anxiety and depression in non-psychiatric hospital outpatient clinics. Composed of two distinct seven-item subscales—HADS-Anxiety (HADS-A) and HADS-Depression (HADS-D)—the instrument intentionally excludes somatic symptoms such as fatigue, insomnia, weight fluctuations, and bodily pain. This design choice was implemented to prevent confounding psychological distress with concurrent physical pathophysiology, making the scale exceptionally valuable in general medical, oncology, cardiology, palliative care, and primary care environments. Each item is rated on a 4-point response scale ranging from 0 to 3, yielding potential subscale scores from 0 to 21, with conventional clinical thresholds identifying normal (0–7), borderline abnormal (8–10), and clinically abnormal or caseness (11–21) states. Psychometrically, the HADS demonstrates robust internal consistency across diverse populations, with Cronbach’s alpha typically spanning 0.80 to 0.93 for HADS-A and 0.81 to 0.90 for HADS-D. Extensive exploratory and confirmatory factor analyses generally corroborate its two-dimensional structure, although alternative tripartite (e.g., Clark and Watson’s tripartite model of affect) and bifactor configurations have been substantiated in specialized somatic cohorts. Concurrent validity analyses against gold-standard psychiatric interviews and instruments—such as the Beck Depression Inventory (BDI) and the State-Trait Anxiety Inventory (STAI)—confirm optimal diagnostic sensitivity and specificity (both generally approximating 0.80) at the standard cut-off score of 8 or above. Consequently, the HADS remains an indispensable, brief, and clinically resilient psychometric measure for evaluating emotional distress in medically compromised individuals.

Keywords

Hospital Anxiety and Depression Scale, HADS, psychometrics, screening instrument, anhedonia, somatic confounding, internal consistency, factor structure, affective disorders, clinical caseness

Authors

The Hospital Anxiety and Depression Scale was developed by Anthony S. Zigmond and R. Philip Snaith in 1983 while based in the Department of Psychiatry at the University of Leeds and St James’s University Hospital, Leeds, United Kingdom. Dr. Anthony S. Zigmond served as a Consultant Psychiatrist with research interests in liaison psychiatry, psychometrics, and the psychiatric manifestations of chronic systemic disease. Dr. R. Philip Snaith (1933–2003) was an internationally distinguished clinical psychiatrist, researcher, and psychometrician whose foundational work focused on affective disorders, anhedonia, and the clinical assessment of depressive illness. Throughout his tenure at the University of Leeds, Snaith contributed extensively to establishing clear boundaries between cognitive-emotional symptoms and biological-somatic complaints in psychiatric nosology.

Subsequent cross-cultural adaptations and language validations have involved various prominent clinical researchers. The widely utilized Dutch validation was conducted by Frans Pouwer, Frank J. Snoek, and Henk M. van der Ploeg in 1997 at the Vrije Universiteit Medical Center (VUmc) in Amsterdam, Netherlands. Dr. Pouwer and Dr. Snoek are internationally recognized leaders in medical psychology and psychodiabetology, examining the complex intersections of metabolic disorders and psychiatric comorbidity.

Purpose

The primary clinical and theoretical impetus behind the development of the HADS was the profound diagnostic challenge of identifying affective disturbance among patients admitted to general medical wards or attending outpatient medical clinics. Prior to the introduction of the HADS, standard psychological and psychiatric assessment tools—such as the Hamilton Rating Scale for Depression (HAM-D) and the Beck Depression Inventory (BDI)—relied heavily on neurovegetative and somatic symptoms. These included lethargy, psychomotor retardation, sleep disturbances, gastrointestinal upset, diffuse somatic pain, appetite loss, and weight fluctuation.

In patients with systemic medical illnesses (e.g., end-stage renal disease, cardiovascular disease, systemic lupus erythematosus, and neoplastic conditions undergoing chemotherapy), these physical manifestations are frequently direct biological sequelae of the organic illness or its pharmacological treatment rather than indicators of an underlying mood disorder. Utilizing somatic-heavy instruments in these populations predictably leads to substantial false-positive rates, diagnostic inflation, and inappropriate psychiatric attribution. Zigmond and Snaith recognized this fundamental psychometric flaw and designed the HADS to isolate psychological and cognitive-emotional indicators, eliminating all physical and somatic symptomatology.

Clinically, the HADS is utilized for:

• Rapid initial screening for clinically significant anxiety and depressive states in general medical, surgical, and specialist outpatient clinics (e.g., oncology, cardiology, neurology, gastroenterology, endocrinology).
• Differentiating psychological distress from organic somatic disease symptoms.
• Monitoring longitudinal symptom progression and therapeutic response following medical, surgical, or psychotherapeutic interventions.
• Quantifying dimensional distress in epidemiological studies, clinical trials, and health-related quality of life (HRQoL) research programs.

The instrument is widely applied across inpatient settings, outpatient secondary care, palliative care environments, and community-based general practice, where its brief completion time (consistently under five minutes) prevents cognitive and physical burden in fragile or severely debilitated patients.

Psychological Construct

The HADS is designed around two discrete yet clinically interrelated psychological constructs: generalized anxiety and anhedonic depression. The conceptual architecture of both subscales intentionally prioritizes mental and phenomenological distress over physical manifestations.

HADS-Anxiety (HADS-A)

The HADS-A subscale assesses cognitive, psychic, and subjective manifestations of anxiety. It reflects sustained autonomic apprehension, psychomotor restlessness, and catastrophic anticipation without querying biological somatic symptoms (such as tachycardia, diaphoresis, or tachypnea). The core construct centers on the clinical entity of tension and psychic hyperarousal. Specifically, the items explore:

• Psychic tension and feeling ‘wound up’: Inability to relax mentally and a subjective sense of inner turmoil.
• Anticipatory apprehension: Subjective dread and fear that impending misfortune or catastrophe is about to materialize.
• Intrusive worry: Rumination, mental preoccupation, and repetitive anxious thoughts that prove difficult to control or dispel.
• Visceral apprehension: Experiencing fear-induced visceral sensations, such as “butterflies in the stomach,” which Snaith framed as psychic distress rather than a sign of gastrointestinal pathology.
• Motor agitation: Restlessness and the compelling urge to remain constantly active or physically mobile due to psychological tension.
• Panic attacks: Sudden, unprovoked waves of panic and terror.

HADS-Depression (HADS-D)

The HADS-D subscale assesses depressive affect by evaluating anhedonia—the loss of interest in activities and the diminished capacity to experience pleasure. Snaith argued that anhedonia represents the central diagnostic core of melancholic and clinically responsive depressive illness. By focusing predominantly on hedonic tone rather than dysphoric mood or neurovegetative signs (such as sleep disruption, fatigue, or appetite reduction), the subscale successfully prevents medical illness symptoms from falsely inflating depression scores. The construct evaluates:

• Consummatory anhedonia: The inability to derive pleasure in real time from previously rewarding experiences (e.g., reading a good book, listening to radio, or watching television).
• Anticipatory anhedonia: The absence of forward-looking pleasure or excitement when anticipating positive future events.
• Affective blunting and loss of humor: The inability to experience amusement, laugh, or perceive the humorous dimension of everyday life events.
• Loss of subjective energy and psychomotor slowing: A perceived internal sluggishness distinct from biological muscle weakness.
• Loss of motivation: Abandonment of self-care, grooming, and personal appearance.

Theoretical Framework

The conceptual formulation of the HADS is grounded in the classical descriptive psychopathology of European clinical psychiatry and anticipates modern structural models of emotion, notably the Tripartite Model of Anxiety and Depression articulated by Lee Anna Clark and David Watson (1991). The Tripartite Model posits that anxiety and depression share a broad dimension of non-specific negative affectivity (general distress, subjective suffering, and emotional instability), but are structurally differentiated by specific unique factors: autonomic hyperarousal (specific to anxiety) and low positive affectivity/anhedonia (specific to depression).

Snaith and Zigmond’s intuitive operationalization in 1983 remarkably preceded Clark and Watson’s formal empirical model. Snaith posited that while general demoralization is common among physically ill individuals, true depressive illness is defined by pervasive anhedonia. By anchoring HADS-D items almost entirely in low positive affect (e.g., lack of cheerfulness, blunted humor, diminished anticipation of enjoyable events), the subscale isolates the distinct component of depression. Conversely, the HADS-A subscale targets psychic apprehension, hypervigilance, and subjective tension, which reflect heightened emotional hyperarousal without relying on somatic indicators that might stem from an underlying medical condition (such as chronic obstructive pulmonary disease, hyperthyroidism, or autonomic neuropathy).

Cognitive theory (Aaron T. Beck) also underpins the scale’s emphasis on cognitive mediation. Rather than assessing bodily manifestations, the instrument measures cognitive schemas—evaluations of impending catastrophe (anxiety) and the cognitive appraisal of future hopelessness and loss of reward responsiveness (depression). Consequently, the HADS functions as a clean operational measure of affective psychopathology in medical populations.

Validity

The psychometric validity of the Hospital Anxiety and Depression Scale has been established across hundreds of clinical cohorts spanning oncology, neurology, cardiology, chronic musculoskeletal pain, respiratory disorders, and primary care.

Construct and Structural Validity

Construct validity is evidenced by the scale’s internal dimensional cohesion and its theoretical alignment with diagnostic classification systems such as the DSM-5-TR and the ICD-11. Meta-analytic evaluations, such as the seminal systematic review by Bjelland, Dahl, Haug, and Neckelmann (2002), reviewing 747 identified papers, confirmed that the correlation between the HADS-A and HADS-D subscales is moderate to strong (ranging from r = 0.40 to 0.74; mean r = 0.56). This supports both their concurrent overlap under general affective distress and their meaningful construct divergence as separate dimensions.

Concurrent and Criterion Validity

Criterion-related validity has been evaluated against structured psychiatric clinical interviews, including the Structured Clinical Interview for DSM (SCID), the Present State Examination (PSE), and the Composite International Diagnostic Interview (CIDI). In the benchmark meta-analysis by Bjelland et al. (2002), the sensitivity and specificity of both the HADS-A and HADS-D across multiple clinical populations were found to be optimal at an identical threshold:

• HADS-A Cut-off ≥ 8: Sensitivity ranges from 0.78 to 0.93; Specificity ranges from 0.70 to 0.90 (AUC typically ≥ 0.84 to 0.92).
• HADS-D Cut-off ≥ 8: Sensitivity ranges from 0.72 to 0.90; Specificity ranges from 0.75 to 0.89 (AUC typically ≥ 0.82 to 0.91).

When evaluated against other self-report inventories, concurrent validity is robust. HADS-A correlates strongly with the State-Trait Anxiety Inventory (STAI; r = 0.64 to 0.81) and the Beck Anxiety Inventory (BAI; r = 0.60 to 0.78). HADS-D correlates strongly with the Beck Depression Inventory (BDI; r = 0.62 to 0.73) and the Center for Epidemiologic Studies Depression Scale (CES-D; r = 0.58 to 0.72).

Discriminant Validity

The scale effectively discriminates between somatic symptoms caused by physical disease and affective disturbance. In cancer patients undergoing radiation or chemotherapy, HADS scores demonstrate minimal correlation with biological markers of tumor burden or non-specific physical fatigue scales (such as the Functional Assessment of Chronic Illness Therapy-Fatigue; r < 0.30), whereas somatic-laden scales (e.g., BDI-I) correlate artificially higher due to physical item contamination.

Reliability

The reliability of the HADS has been examined across diverse demographic cohorts, medical subspecialties, and cross-cultural adaptations.

Internal Consistency

Internal consistency estimates via Cronbach’s alpha consistently satisfy and exceed established psychometric thresholds for screening and clinical monitoring:

• HADS-Anxiety (HADS-A): Cronbach’s α values systematically range from 0.80 to 0.93 across clinical cohorts (mean α ≈ 0.83). In general population studies, α remains stable between 0.82 and 0.86.
• HADS-Depression (HADS-D): Cronbach’s α values systematically range from 0.81 to 0.90 in medical samples (mean α ≈ 0.82), occasionally falling to 0.76–0.79 in healthy, non-clinical populations where anhedonic variance is more restricted.
• Total Score (HADS-T): When operationalized as a global distress index, total score Cronbach’s α coefficients consistently range between 0.88 and 0.93.

Test-Retest Reliability and Temporal Stability

Test-retest reliability reflects the scale’s stability while remaining sensitive to genuine clinical changes. In clinically stable outpatient cohorts over retest intervals ranging from 1 to 14 days, intraclass correlation coefficients (ICC) and Pearson product-moment coefficients range between r = 0.80 and 0.92 for HADS-A, and between r = 0.84 and 0.91 for HADS-D. Over longer observation periods (e.g., 6 weeks in non-interventional post-myocardial infarction or stroke recovery studies), stability coefficients remain high (r ≈ 0.70 to 0.75), while systematically dropping in response to pharmacotherapy or cognitive behavioral therapy, confirming the scale’s sensitivity to treatment response.

Factor Analysis

The underlying latent dimensional structure of the HADS has been one of the most widely investigated topics in medical psychometrics. While Zigmond and Snaith originally proposed an orthogonal two-factor structure (Anxiety and Depression), subsequent exploratory factor analyses (EFA) and confirmatory factor analyses (CFA) have sparked scholarly discussion.

Two-Factor Models

The classical two-factor model—comprising an independent Anxiety factor (Items 1, 3, 5, 7, 9, 11, 13) and Depression factor (Items 2, 4, 6, 8, 10, 12, 14)—has been supported in hundreds of studies. Standard CFA evaluations using maximum likelihood estimation demonstrate acceptable fit indices:

• Comparative Fit Index (CFI) ≥ 0.90 to 0.94
• Tucker-Lewis Index (TLI) ≥ 0.89 to 0.93
• Root Mean Square Error of Approximation (RMSEA) ≈ 0.05 to 0.07
• Standardized Root Mean Square Residual (SRMR) ≤ 0.06

Factor loadings for the respective latent dimensions generally exceed 0.50. However, Items 7 (“I can sit at ease and feel relaxed”) and 8 (“I feel as if I am slowed down”) frequently exhibit minor cross-loadings or lower primary loadings (0.35–0.48) in elderly or chronically ill medical cohorts, as psychomotor slowing can reflect physical illness.

Three-Factor Models

Several psychometricians (notably Clark and Watson, 1991; Dunbar et al., 2000; and Martin et al., 2003) have argued that a three-factor structural model better captures the latent variance of the HADS. The Dunbar three-factor model divides the scale into:

1. Anhedonia/Depression: Items 2, 4, 6, 8, 10, 12, 14.
2. Autonomic/Psychic Anxiety: Items 3, 9, 13.
3. Negative Affectivity/Tension: Items 1, 5, 7, 11.

In various chronic illness cohorts (such as end-stage cardiovascular disease, cancer, and chronic pain), CFA fit indices for this tripartite model often display incremental improvement over the classical two-factor structure (CFI ≥ 0.95, RMSEA ≤ 0.048).

Bifactor and Unidimensional Models

Recent psychometric research utilizing bifactor modeling suggests that HADS items reflect a dominant general factor of “Psychological Distress” alongside two distinct subscale group factors. When calculated, omega hierarchical coefficients (ω_h) for the general distress factor often exceed 0.80, leading many researchers to advocate for calculating both subscale scores and the composite total HADS score (HADS-T) in clinical oncology and general hospital studies.

Instrument / Measurement Tool

The Hospital Anxiety and Depression Scale is configured as follows:

• Test Type: Self-administered psychometric screening questionnaire (can also be administered by a clinician or trained researcher).
• Format: Paper-and-pencil or secure computer/tablet-based digital administration.
• Item Count: 14 items total (7 items in the Anxiety subscale; 7 items in the Depression subscale).
• Assessment Window: Instructs respondents to reflect on feelings and emotional states experienced over the past week (a 4-week recall window has also been validated in certain clinical and European adaptations).
• Response Scale: 4-point scale scored from 0 to 3 (utilizing item-specific response anchors corresponding to symptom intensity and frequency).
• Scoring and Directionality:
  • Directly scored items (higher raw value = higher distress): 1, 3, 5, 8, 9, 10, 11, 13.
  • Reverse-scored items (0 becomes 3, 1 becomes 2, 2 becomes 1, 3 becomes 0; higher score reflects greater distress): 2, 4, 6, 7, 12, 14.
• Subscale Score Derivation:
  • HADS-Anxiety (HADS-A): Sum of items 1 + 3 + 5 + 7 + 9 + 11 + 13 (Range: 0 to 21).
  • HADS-Depression (HADS-D): Sum of items 2 + 4 + 6 + 8 + 10 + 12 + 14 (Range: 0 to 21).
  • Total Score (HADS-T): Sum of all 14 items (Range: 0 to 42), used to reflect generalized emotional distress.
• Standard Clinical Cut-off Scores (per subscale):
  • 0 to 7: Normal (non-case).
  • 8 to 10: Borderline abnormal (borderline caseness; warrants monitoring or low-intensity clinical attention).
  • 11 to 21: Abnormal (clinical case; indicates probable psychiatric anxiety disorder or clinical depression, warranting comprehensive diagnostic clinical interview).
• Administration Time: Typically completed in 2 to 5 minutes.

Permissions & Fee and Test Year

The Hospital Anxiety and Depression Scale was originally published in 1983 by Anthony S. Zigmond and R. Philip Snaith in Acta Psychiatrica Scandinavica. For several decades following its publication, the scale was widely treated as an open-access clinical screening tool and published freely in clinical manuals and research papers.

Subsequently, the commercial copyright for the standardized English edition was acquired by GL Assessment (formerly NFER-Nelson) based in the United Kingdom. While the original 1983 publication remains widely referenced, commercial applications, pharmaceutical clinical trials, and standardized operational use may require commercial licensing and user fees through GL Assessment or its national distributors.

Academic and non-commercial health researchers typically administer the instrument under academic fair-use guidelines or via licensed regional validation translations (e.g., the Dutch version validated by Pouwer, Snoek, and van der Ploeg in 1997). Researchers should verify institutional and commercial licensing requirements with GL Assessment prior to launching commercial clinical trials or integrating the tool into electronic health record (EHR) platforms.

References

• Bjelland, I., Dahl, A. A., Haug, T. T., & Neckelmann, D. (2002). The validity of the Hospital Anxiety and Depression Scale: An updated literature review. Journal of Psychosomatic Research, 52(2), 69–77. https://doi.org/10.1016/S0022-3999(01)00296-3
• Clark, L. A., & Watson, D. (1991). Tripartite model of anxiety and depression: Psychometric evidence and taxonomic implications. Journal of Abnormal Psychology, 100(3), 316–336. https://doi.org/10.1037/0021-843X.100.3.316
• Dunbar, M., Ford, G., Hunt, K., & Der, G. (2000). A confirmatory factor analysis of the Hospital Anxiety and Depression scale: Comparing empirical and theoretical models. British Journal of Clinical Psychology, 39(1), 79–94. https://doi.org/10.1348/014466500163121
• Herrmann, C. (1997). International experiences with the Hospital Anxiety and Depression Scale—A review of validation data and clinical results. Journal of Psychosomatic Research, 42(1), 17–41. https://doi.org/10.1016/S0022-3999(96)00216-4
• Martin, C. R., & Newell, R. J. (2004). Factor structure of the Hospital Anxiety and Depression Scale in individuals with facial disfigurement. Journal of Psychosomatic Research, 57(5), 415–420. https://doi.org/10.1016/S0022-3999(03)00624-X
• Pouwer, F., Snoek, F. J., & van der Ploeg, H. M. (1997). De Hospital Anxiety and Depression Scale (HADS): Een betrouwbare en valide screeningsvragenlijst voor depressieve gevoelens en angstklachten bij somatische patiënten. Nederlands Tijdschrift voor Geneeskunde, 141(48), 2326–2330.
• Snaith, R. P. (2003). The Hospital Anxiety and Depression Scale. Health and Quality of Life Outcomes, 1, Article 29. https://doi.org/10.1186/1477-7525-1-29
• Zigmond, A. S., & Snaith, R. P. (1983). The Hospital Anxiety and Depression Scale. Acta Psychiatrica Scandinavica, 67(6), 361–370. https://doi.org/10.1111/j.1600-0447.1983.tb09716.x

13. Items of the Scale (Questionnaire)