Hospital Anxiety and Depression Scale (HADS)

1. Abstract

The Hospital Anxiety and Depression Scale (HADS) is an internationally validated, self-report screening instrument developed by Anthony S. Zigmond and R. Philip Snaith in 1983 to detect states of anxiety and depression among patients in non-psychiatric medical outpatient and inpatient settings. Comprising 14 items divided equally into two distinct 7-item subscales—Anxiety (HADS-A) and Depression (HADS-D)—the instrument was deliberately designed to eliminate physical, somatic, and autonomic symptoms of distress (e.g., headaches, fatigue, insomnia, weight fluctuations, and gastrointestinal disturbances) that frequently confound psychiatric evaluation in medically ill populations. Items are scored on a four-point Likert-type scale ranging from 0 to 3, yielding dimensional subscale scores from 0 to 21, with established clinical cut-offs identifying non-cases (0–7), mild/borderline cases (8–10), moderate distress (11–14), and severe psychiatric morbidity (15–21). Psychometric investigations across hundreds of clinical populations—ranging from oncology, cardiology, and palliative medicine to rheumatology and primary care—have firmly substantiated its robust internal consistency (Cronbach’s alpha typically spanning 0.80 to 0.93 for HADS-A and 0.76 to 0.90 for HADS-D) and test-retest reliability ($r > 0.80$). Extensively translated into more than 115 languages, the instrument demonstrates high diagnostic sensitivity and specificity (both approximately 0.80) against gold-standard psychiatric interviews based on DSM and ICD criteria. Despite ongoing debates regarding its underlying latent dimensionality—alternating between classical two-factor, three-factor tripartite, and bifactor structures—the HADS remains a preeminent dimensional psychometric benchmark for the clinical appraisal of psychological distress within somatic health ecosystems.

2. Keywords

Hospital Anxiety and Depression Scale, HADS, psychometrics, anxiety screening, depression screening, medical illness, anhedonia, somatic confounding, factor structure, construct validity

3. Authors

The Hospital Anxiety and Depression Scale was conceived, calibrated, and standardized by two British consultation-liaison psychiatrists:

• Anthony S. Zigmond, M.B., Ch.B., FRCPsych: Formerly Consultant Psychiatrist at St. James’s University Hospital, Leeds, United Kingdom. Dr. Zigmond’s clinical research centered on the intersection of liaison psychiatry, general hospital management, and the objective assessment of emotional disturbance in physically compromised patients.
• R. Philip Snaith, M.D., FRCPsych: Formerly Senior Lecturer in Psychiatry at the University of Leeds and Honorary Consultant Psychiatrist at St. James’s University Hospital, Leeds, United Kingdom. Dr. Snaith was an international authority on clinical psychometrics, affective disorders, and the phenomenology of anhedonia, who also co-developed the Snaith-Hamilton Pleasure Scale (SHAPS).

4. Purpose

The primary clinical and epidemiological objective of the Hospital Anxiety and Depression Scale is to offer a rapid, reliable, and valid screening mechanism to identify clinically significant affective morbidity—specifically generalized states of anxiety and depressive anhedonia—in individuals receiving care in general medical and surgical facilities. When medically compromised patients present with serious physical disorders (such as coronary artery disease, stroke, chronic obstructive pulmonary disease, end-stage renal failure, or neoplastic conditions), their medical illness or pharmacotherapy frequently precipitates somatic symptoms that overlap almost entirely with conventional diagnostic criteria for mood disorders.

Traditional diagnostic tools and classic rating inventories, such as the Beck Depression Inventory (BDI) or the Hamilton Rating Scale for Depression (HAM-D), incorporate heavy weightings on neurovegetative markers including fatigue, lethargy, psychomotor retardation, sleep disturbance, anorexia, dizziness, and somatic tension. When administered to medically ill patients, these traditional instruments produce elevated false-positive rates, erroneously classifying physiological signs of systemic illness or medication side-effects as psychiatric disorders. Zigmond and Snaith (1983) specifically set out to construct a questionnaire devoid of these confounding neurovegetative indices, thereby capturing purely cognitive, affective, and emotional markers of distress.

In routine clinical practice, the HADS performs several functions:

• Triage and Case-Finding: Differentiating patients with normative, transient situational distress from those exhibiting clinically significant affective states requiring formal psychiatric consultation, psychotherapy, or pharmacotherapy.
• Dimensional Severity Monitoring: Tracking fluctuations in anhedonia and psychic tension over time, particularly across longitudinal trajectories of surgical interventions, chemotherapy cycles, organ transplantation, or palliative management.
• Clinical Trial Endpoint: Serving as a standardized patient-reported outcome measure (PROM) in pharmacotherapeutic trials and psychological intervention studies within psycho-oncology, psychosomatic medicine, and behavioral cardiology.
• Epidemiological Surveillance: Quantifying the prevalence and severity of affective distress across diverse chronic disease populations without inflating estimates through somatic cross-contamination.

5. Psychological Construct

The HADS is designed around two distinct but interconnected psychological dimensions: Psychic Anxiety and Depressive Anhedonia. Each dimension is operationalized through seven items intentionally calibrated to assess affective states experienced over the past week.

Hospital Anxiety (HADS-A)

The anxiety subscale measures generalized subjective tension, apprehension, dread, and psychic restlessness, focusing heavily on mental and affective components rather than autonomic overactivity. Unlike general anxiety assessments that evaluate panic-induced autonomic discharge (such as tachycardia, diaphoresis, hyperventilation, or peripheral tremors), HADS-A emphasizes the cognitive and emotional experiences of worry, nervous tension, and sudden panic:

• Cognitive Dread and Apprehension: Captured by Item 3 (“I get a sort of frightened feeling as if something awful is about to happen”) and Item 5 (“Worrying thoughts go through my mind”). These items evaluate intrusive, catastrophic cognitive processing and anticipatory hypervigilance.
• Psychic Tension and Inability to Relax: Captured by Item 1 (“I feel tense or ‘wound up’”) and Item 7 (“I can sit at ease and feel relaxed”). This taps into internal psychic agitation and the physiological inability to maintain subjective quiescence.
• Motor Restlessness and Internal Agitation: Captured by Item 11 (“I feel restless as if I have to be on the move”). This targets the psychomotor expression of internal anxiety without referencing somatic exhaustion.
• Visceral Apprehension and Panic Attacks: Items 9 (“I get a sort of frightened feeling like ‘butterflies’ in the stomach”) and 13 (“I get sudden feelings of panic”) index the sudden onset of acute panic-like anxiety and classic affective butterflies, remaining distinct from chronic medical gastrointestinal pathology.

Hospital Depression (HADS-D)

The depression subscale is anchored in Snaith’s conceptualization of anhedonia—the inability to experience pleasure or anticipated pleasure from typically rewarding experiences. Snaith posited that anhedonia represents the primary psychopathological marker of true endogenomorphic or melancholic depression, distinguishing true affective pathology from the somatic sequelae of medical disease. The construct avoids queries concerning weight loss, insomnia, or physical fatigue, measuring hedonic tone and energy across cognitive-affective domains:

• Consummatory and Anticipatory Anhedonia: Operationalized by Item 2 (“I still enjoy the things I used to enjoy”) and Item 12 (“I look forward with enjoyment to things”). These items evaluate both current hedonic capacity and prospective hedonic anticipation.
• Humor and Affective Reactivity: Evaluated by Item 4 (“I can laugh and see the funny side of things”) and Item 6 (“I feel cheerful”). These questions assess positive affectivity and the responsiveness of mood to cheerful environmental stimuli.
• Cognitive Slowing and Avolition: Addressed by Item 8 (“I feel as if I am slowed down”). Although slowing can have biological origins, within the HADS context it gauges the subjective sense of psychic lethargy and reduced psychomotor drive.
• Loss of Self-Care and Engagement: Addressed by Item 10 (“I have lost interest in my appearance”) and Item 14 (“I can enjoy a good book or radio or TV programme”). These items focus on motivational deficits, apathy, and the loss of passive hedonic engagement with common cultural and self-preservation behaviors.

6. Theoretical Framework

The theoretical framework of the HADS rests on two fundamental pillars: the Phenomenological Model of Hedonic Deficit developed by Snaith and colleagues, and the Tripartite Model of Anxiety and Depression formulated by Clark and Watson (1991).

Snaith’s Hedonic Deficit Paradigm

In developing the HADS, Philip Snaith drew upon his earlier psychopathological studies on the core components of depressive illness. He argued that general medical illness inevitably induces biological exhaustion, disrupted sleep architecture, altered appetite, and pain. If a psychometric inventory treats these manifestations as symptoms of psychiatric distress, it creates an artificial correlation between physical disease and mental illness. Snaith posited that loss of pleasure (anhedonia) is the single most specific and invariant manifestation of depressive illness that remains unaffected by medical illness. By designing the depression subscale to measure hedonic tone, the instrument assesses genuine emotional disturbance rather than secondary somatic fatigue.

The Tripartite Model Integration

Although published eight years before Clark and Watson’s formal introduction of the Tripartite Model, the HADS anticipates its three-part architecture:

• Negative Affectivity (General Distress): A shared dimension of non-specific distress, emotional lability, and subjective demoralization common to both anxiety and depressive states.
• Positive Affectivity (Low Positive Affect / Anhedonia): Specific to depression, characterized by fatigue of interest, apathy, joylessness, and loss of enthusiasm. HADS-D aligns with low positive affectivity.
• Physiological Hyperarousal: Specific to anxiety, characterized by somatic tension and autonomic hyperactivity. HADS-A deliberately circumvents peripheral autonomic discharge, capturing instead subjective psychic hyperarousal and tension.

Because HADS-A measures psychic tension and HADS-D measures anhedonia, the two subscales maintain functional separation. This conceptual distinction minimizes the shared variance driven by general negative affectivity, allowing clinicians to independently evaluate co-occurring anxiety and depression.

7. Validity

The validity of the HADS has been examined across diverse inpatient, outpatient, and general population cohorts, establishing solid psychometric properties across diagnostic paradigms.

Construct and Discriminant Validity

The discriminant validity of the HADS lies in its capacity to detect affective disturbance without inflation from somatic disease. In a meta-analytic review of 747 published papers, Bjelland et al. (2002) confirmed that the correlation between the HADS-A and HADS-D subscales spans $r = 0.40$ to $0.74$ (mean $r = 0.56$), showing moderate overlap while confirming that the two scales index distinct clinical constructs.

Criterion and Concurrent Validity

Concurrent validity has been established by correlating HADS subscale scores against established psychiatric rating inventories and structured psychiatric diagnostic interviews (e.g., SCID, SCAN, CIDI):

• HADS-A Concurrent Validity: Exhibits correlations ranging between $r = 0.60$ and $0.75$ with the State-Trait Anxiety Inventory (STAI), the Beck Anxiety Inventory (BAI), and the Hamilton Anxiety Rating Scale (HAM-A).
• HADS-D Concurrent Validity: Yields correlation coefficients between $r = 0.62$ and $0.79$ with the Beck Depression Inventory (BDI), the Montgomery-Åsberg Depression Rating Scale (MADRS), and the Hamilton Depression Rating Scale (HAM-D).

Diagnostic Sensitivity and Specificity

Receiver Operating Characteristic (ROC) analyses across diverse medical settings (including oncology, palliative care, primary care, stroke units, and cardiac rehabilitation) show area under the curve (AUC) values between 0.84 and 0.93. When utilizing the standard cut-off score of $ge 8$ points on each subscale:

• HADS-A: Achieves a mean sensitivity of 0.89 and a mean specificity of 0.75 against DSM-defined generalized anxiety disorder and adjustment disorders.
• HADS-D: Demonstrates a mean sensitivity of 0.83 and a mean specificity of 0.79 against DSM-defined major depressive disorder.

At the higher diagnostic threshold of $ge 11$, specificity rises to between 0.90 and 0.95, optimizing the tool for definitive case identification and minimizing false positives in specialty medical clinics.

8. Reliability

The internal consistency, test-retest stability, and measurement precision of the HADS have been supported by empirical evaluations spanning four decades.

Internal Consistency

In the landmark systematic psychometric review conducted by Bjelland et al. (2002), the internal consistency of both subscales was examined across dozens of independent medical cohorts:

• HADS-A Cronbach’s Alpha: Ranged from 0.68 to 0.93, with a mean estimate across clinical trials of $\alpha = 0.83$. Values above 0.80 confirm strong homogeneity among the seven anxiety items without excessive item redundancy.
• HADS-D Cronbach’s Alpha: Ranged from 0.67 to 0.90, with a mean estimate of $\alpha = 0.82$. This confirms that the hedonic items form a cohesive operationalization of depressive affectivity.

Test-Retest Reliability

Evaluation of temporal stability across intervals ranging from one day to six weeks has yielded test-retest correlation coefficients consistently above $r = 0.80$ (commonly ranging between 0.82 and 0.92) in medically stable outpatients. In longitudinal psychiatric interventions, the HADS tracks symptom reduction following pharmacotherapy or cognitive behavioral therapy, demonstrating sensitivity to clinical change while maintaining stability in untreated control groups.

Item-Total Correlations and Inter-Item Homogeneity

Corrected item-total correlation coefficients for the individual items range between $r = 0.45$ and $r = 0.68$, meeting psychometric criteria ($r > 0.30$) for scale integrity. Item 10 (“I have lost interest in my appearance”) occasionally demonstrates lower corrected item-total correlations ($r \approx 0.35 – 0.42$) in elderly male populations or severe ICU settings, reflecting differing cultural and medical baseline attitudes toward grooming during acute illness.

9. Factor Analysis

The structural dimensionality of the HADS has been the subject of ongoing empirical investigation within clinical psychometrics. While Zigmond and Snaith originally posited a strict two-factor orthogonal model, structural equation modeling (SEM) and confirmatory factor analysis (CFA) have documented nuanced latent configurations.

Exploratory Factor Analysis (EFA)

Early exploratory factor analyses with varimax or oblimin rotation typically confirmed the clean emergence of two dominant factors corresponding to HADS-A and HADS-D, with items loading between 0.50 and 0.82 on their designated factor and cross-loadings remaining low (< 0.25). The seven anxiety items loaded primarily on Factor I, while the seven depression items defined Factor II.

Confirmatory Factor Analysis (CFA) and Competing Models

Subsequent modern CFA studies evaluating goodness-of-fit indices (including RMSEA, CFI, TLI, and SRMR) have contrasted multiple competing latent models:

• The Bidimensional Model (Zigmond & Snaith, 1983): A two-factor oblique model that provides acceptable fit across general somatic cohorts ($\chi^2/df < 3.0$, $\text{CFI} > 0.92$, $\text{RMSEA} < 0.06$).
• The Tripartite Three-Factor Model (Clark & Watson / Dunbar et al., 2000): Separates the scale into (1) Anhedonic Depression, (2) Autonomic Anxiety/Agitation, and (3) Generalized Negative Affectivity. This model often yields improved fit indices ($\text{CFI} > 0.95$, $\text{RMSEA} < 0.05$) in psychiatric and oncology populations.
• The Four-Factor Model (Dunbar et al., 2000; Caci et al., 2003): Splits anxiety into psychic anxiety and motor restlessness, while retaining anhedonia and psychomotor retardation as separate depressive dimensions.
• The Bifactor Model (Norton et al., 2013): Hypothesizes a single general factor of overall “Psychological Distress” explaining the majority of common item variance, alongside two specific orthogonal group factors corresponding to residual anxiety and residual anhedonia. Many contemporary psychometricians suggest that the total HADS score serves as an effective index of general emotional distress, while the subscales provide granular insight into specific symptom profiles.

10. Instrument / Measurement Tool

The HADS is a self-administered, paper-and-pencil or digital screening questionnaire designed for rapid administration without clinical supervision.

• Instrument Name: Hospital Anxiety and Depression Scale (HADS)
• Authors: Anthony S. Zigmond and R. Philip Snaith (1983)
• Test Format: Self-report rating inventory; paper-and-pencil or computer-assisted administration
• Target Population: Non-psychiatric hospital inpatients, outpatients, primary care patients, and the general adult population (aged 16–65+ years)
• Number of Items: 14 items total (7 items for HADS-A; 7 items for HADS-D)
• Subscale Allocation:
  • Anxiety Subscale (HADS-A): Items 1, 3, 5, 7, 9, 11, 13
  • Depression Subscale (HADS-D): Items 2, 4, 6, 8, 10, 12, 14
• Recall Period: Past week (“over the past week”)
• Administration Time: 2 to 5 minutes
• Response Format: 4-point Likert-type options (ranging from 0 to 3), with item-specific textual response anchors reflecting frequency, intensity, or change in hedonic capacity.
• Scoring Architecture:
  • Each item is scored from 0 to 3.
  • Reverse-worded items: To guard against acquiescence response bias, several items are phrased positively (e.g., Items 2, 4, 6, 7, 12, 14) and scored reversely so that higher numerical totals consistently denote more severe psychopathology.
  • Subscale Scores: Summation of items for HADS-A yields a range of 0–21; summation of items for HADS-D yields a range of 0–21.
  • Total Score (Optional General Distress Index): 0–42 points.
• Clinical Cut-Off Categories:
  • 0–7: Normal (non-case; absence of clinically significant anxiety or depression)
  • 8–10: Mild / Borderline case (indicates possible presence of an affective disorder; warrants ongoing monitoring or brief intervention)
  • 11–14: Moderate disturbance (indicates probable presence of an affective disorder; indicates need for detailed diagnostic evaluation)
  • 15–21: Severe psychiatric disturbance (indicates significant affective pathology requiring urgent clinical or psychiatric attention)

11. Permissions & Fee and Test Year

The Hospital Anxiety and Depression Scale was originally formulated and published in 1983 in the peer-reviewed journal Acta Psychiatrica Scandinavica. The legal and licensing history of the instrument has evolved over time:

• Historical Copyright: The scale was originally published by Munksgaard International Publishers (subsequently absorbed by John Wiley & Sons). For several decades, worldwide commercial licensing, translations, and paper distribution rights were managed exclusively by GL Assessment (formerly NFER-Nelson) in London, United Kingdom.
• Current Status: Following the expiration of certain commercial contractual agreements, user permissions vary depending on the context. Commercial entities, pharmaceutical clinical trials, and proprietary healthcare organizations are typically required to obtain formal licenses or per-administration permissions through the copyright holders or GL Assessment.
• Academic and Non-Commercial Research: Zigmond and Snaith’s 1983 publication remains widely referenced, and researchers routinely use the instrument in unfunded, investigator-initiated research and student theses under fair-use provisions, provided full bibliographic attribution is maintained and the items remain unaltered. Clinicians and researchers should consult the current intellectual property holder (GL Assessment / Wiley) to confirm licensing requirements prior to deployment in commercial or subsidized clinical environments.

12. References

Below are primary peer-reviewed literature citations and foundational psychometric texts supporting the Hospital Anxiety and Depression Scale, formatted according to APA 7th edition standards:

• 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
• Caci, H., Baylé, F. J., Mattei, V., Boyer, P., & Dos Reis, H. (2003). How many dimensions in the Hospital Anxiety and Depression Scale? An alternative approach. Psychiatry Research, 118(2), 177–187. https://doi.org/10.1016/S0165-1781(03)00078-4
• 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
• Cosco, T. D., Doyle, F., Ward, M., & McGee, H. (2012). Latent structure of the Hospital Anxiety and Depression Scale: A 10-year systematic review. Journal of Psychosomatic Research, 72(3), 180–184. https://doi.org/10.1016/j.jpsychores.2011.11.008
• Dunbar, M., Ford, G., Hunt, K., & Der, G. (2000). A confirmatory factor analysis of the Hospital Anxiety and Depression scale: Comparing alternative models. British Journal of Health Psychology, 5(4), 365–387. https://doi.org/10.1348/135910700168991
• 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
• McDowell, I. (2006). Measuring health: A guide to rating scales and questionnaires (3rd ed.). Oxford University Press. https://doi.org/10.1093/acprof:oso/9780195165678.001.0001
• Norton, S., Cosco, T., Doyle, F., Done, J., & Sacker, A. (2013). The Hospital Anxiety and Depression Scale: A meta-analytic approach to establish the internal bound of dimensionality. British Journal of Clinical Psychology, 52(3), 304–316. https://doi.org/10.1111/bjc.12015
• Snaith, R. P. (2003). The Hospital Anxiety and Depression Scale. Health and Quality of Life Outcomes, 1(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