The Panic Attack Questionnaire

1. Abstract

The Panic Attack Questionnaire (PAQ) is a seminal self-report psychometric instrument designed to comprehensively assess the phenomenology, frequency, symptom severity, situational contexts, and cognitive appraisals associated with panic attacks. Developed originally by G. Ron Norton and colleagues in 1985 at the University of Winnipeg, the PAQ was constructed to bridge clinical nosology—specifically the diagnostic criteria introduced in the Diagnostic and Statistical Manual of Mental Disorders, Third Edition (DSM-III)—and empirical psychological research across both clinical and nonclinical populations. The questionnaire opens with a standardized descriptive definition of a panic attack to establish conceptual equivalence among respondents before querying lifetime occurrence, recent attack frequency, and specific attack typologies (e.g., spontaneous vs. cued or situational). The central diagnostic section systematically evaluates the presence and perceived intensity of the classic physical and cognitive symptoms of panic, such as palpitations, dyspnea, trembling, dizziness, paresthesias, derealization, and catastrophic cognitions (fear of dying, going crazy, or losing control), typically measured on continuous or multi-point Likert scales (ranging from 1 [not at all] to 5 [extremely severe]). Additional qualitative and quantitative subcomponents assess anticipatory anxiety, agoraphobic avoidance, physiological triggers, medical consultation patterns, and concurrent substance or medication use. Psychometric evaluations of the PAQ and its revised variants (e.g., PAQ-R, PAQ-IV) indicate robust internal consistency across symptom severity scales (Cronbach’s α typically exceeding .85), stable test-retest reliability across brief non-interventional intervals, and strong convergent validity with gold-standard anxiety measures such as the Beck Anxiety Inventory (BAI) and the Anxiety Sensitivity Index (ASI). By capturing the dimensional nuances of panic experiences beyond binary diagnostic thresholds, the PAQ remains a cornerstone instrument in clinical psychology, psychiatric epidemiology, and cognitive-behavioral research.

2. Keywords

Panic Attack Questionnaire, PAQ, panic disorder, agoraphobia, panic phenomenology, DSM-III, cognitive appraisal, anxiety sensitivity, self-report assessment, psychometrics, nonclinical panickers, symptom severity

3. Authors

The Panic Attack Questionnaire was formulated and introduced to the empirical literature by a research team led by G. Ron Norton, Ph.D., along with colleagues Bruce Harrison, M.A., Jerry Hauch, B.A., and Linda Rhodes, B.A., within the Department of Psychology at the University of Winnipeg in Manitoba, Canada.

Dr. G. Ron Norton was a prominent clinical psychologist whose research program significantly advanced the scientific understanding of anxiety disorders, particularly the epidemiology and cognitive-behavioral mechanisms of panic attacks occurring in nonclinical college student and community samples. Norton and his collaborators were pioneers in demonstrating that discrete panic attacks are not exclusively restricted to patients presenting with psychiatric diagnoses in clinical settings, but rather represent a prevalent dimensional phenomenon in the general population. In subsequent years, additional psychometric validation and structural refinements of the instrument were conducted in collaboration with notable anxiety researchers, including Brian J. Cox, Ph.D., and Richard P. Swinson, M.D., across Canadian academic medical centers.

4. Purpose

The primary purpose of the Panic Attack Questionnaire is to provide an empirical, standardized, and theoretically grounded methodology for identifying, cataloging, and quantifying panic attacks. In psychiatric nosology prior to the publication of the DSM-III in 1980, acute episodic anxiety was frequently subsumed under broad classifications such as anxiety neurosis. When the DSM-III formally isolated panic disorder as a distinct diagnostic entity characterized by sudden, unexpected surges of intense apprehension or terror accompanied by discrete autonomic and cognitive symptoms, researchers lacked comprehensive self-report tools to assess this newly defined phenomenology across diverse populations.

The PAQ was specifically engineered to fulfill three principal research and clinical mandates:

• Epidemiological and Subclinical Detection: Standard diagnostic interviews were resource-intensive and predominantly deployed within psychiatric clinics. The PAQ provided a reliable screening mechanism capable of differentiating nonclinical panickers (individuals who experience occasional, full-blown, or limited-symptom panic attacks without meeting diagnostic criteria for panic disorder or agoraphobia) from non-panickers and from clinically diagnosed patients.
• Phenomenological Profiling: Beyond establishing whether an individual had experienced a panic attack, the PAQ was designed to explore the granular parameters of the attack. This includes evaluating attack onset characteristics (e.g., uncued/spontaneous attacks versus situationally bound or cued attacks), peak latency, duration of the acute episode, and the precise constellation of somatic, neurovegetative, and catastrophic cognitive symptoms.
• Functional and Treatment-Related Assessment: The questionnaire captures the secondary clinical consequences of panic, such as anticipatory dread, fear of somatic sensations, behavioral avoidance patterns resembling agoraphobia, medical service utilization (e.g., emergency department visits for perceived cardiac emergencies), and self-medication strategies involving prescription pharmaceuticals, alcohol, or illicit substances.

In clinical environments, the PAQ assists clinicians in mapping a patient’s idiosyncratic panic profile prior to initiating evidence-based treatments such as Cognitive Behavioral Therapy (CBT) or pharmacotherapy. In experimental psychopathology, it serves as a baseline assessment tool for biological challenge paradigms (such as carbon dioxide inhalation, sodium lactate infusion, or hyperventilation provocations) and longitudinal studies evaluating the natural history of anxiety pathology.

5. Psychological Construct

The Panic Attack Questionnaire measures the psychological, physiological, and behavioral dimensions of the panic attack construct. In clinical psychopathology, a panic attack is defined as an abrupt surge of intense fear or intense discomfort that reaches a peak within minutes, during which time a cluster of specific physical and cognitive symptoms occurs. The PAQ conceptualizes panic not as an all-or-nothing binary event, but as a complex multi-dimensional phenomenon comprising distinct symptom clusters and associated functional sequelae.

1. Autonomic and Cardiorespiratory Somatic Sensations

This dimension encompasses the acute physiological manifestations mediated by sudden sympathetic nervous system hyperarousal and hyperventilation. The PAQ systematically evaluates the presence and subjective severity of individual autonomic symptoms, including:

• Palpitations, pounding heart, or accelerated heart rate (tachycardia)
• Shortness of breath, dyspnea, or smothering sensations
• Choking sensations or difficulty swallowing (globus hystericus)
• Chest pain, pressure, or precordial discomfort
• Profuse sweating or diaphoresis
• Trembling, shaking, or localized tremors
• Nausea, abdominal distress, or gastrointestinal spasms

2. Vestibular, Neurological, and Temperature Dysregulation

A second somatic dimension captures sensations arising from cerebral vasoconstriction secondary to hypocapnia (induced by rapid breathing) and altered peripheral vascular tone:

• Dizziness, lightheadedness, unsteady feelings, or faintness
• Paresthesias (numbness or tingling sensations, particularly in the perioral region and distal extremities)
• Hot flushes, chills, or sudden alterations in perceived body temperature

3. Catastrophic Cognitive Appraisals and Perceptual Distortions

Central to modern cognitive conceptualizations of panic, this construct assesses the interpretive catastrophic misappraisals of somatic sensations:

• Fear of Impending Death: Misinterpreting cardiovascular or respiratory symptoms as immediate signs of myocardial infarction, cardiac arrest, or respiratory failure.
• Fear of Losing Control or Going Crazy: Misinterpreting intense autonomic arousal, derealization, or racing thoughts as indicators of psychotic decompensation, acute mania, or irreversible loss of mental equilibrium.
• Derealization and Depersonalization: Subjective feelings of unreality, detachment from one’s environment, or detachment from one’s physical body or cognitive processes.

4. Temporal and Contextual Attack Dynamics

The construct measured by the PAQ extends into the contextual parameters of the episode:

• Temporal Parameters: Estimated duration from onset to peak intensity, and total time elapsed until symptoms resolve to baseline.
• Contextual Cues: Categorization into spontaneous (unexpected) attacks occurring in the absence of an identifiable environmental or somatic trigger versus cued (expected) attacks occurring upon exposure to, or in anticipation of, specific situational phobic triggers (e.g., enclosed spaces, crowds, public speaking).
• Nocturnal Manifestations: Assessment of panic attacks waking the individual from non-REM sleep without an obvious external dream-related stimulus.

5. Secondary Functional Sequelae and Coping Profiles

Finally, the instrument captures the secondary impact of the panic experience, including anticipatory anxiety regarding future episodes, development of situational or interoceptive avoidance, help-seeking behavior (such as calling emergency medical services), and behavioral coping strategies (e.g., distraction, deep breathing, fleeing the situation, or chemical anxiolysis).

6. Theoretical Framework

The conceptual foundation of the Panic Attack Questionnaire is anchored in the convergence of modern psychiatric descriptive taxonomy, the biological model of panic, and cognitive-behavioral models of anxiety pathology.

The DSM-III Paradigm Shift

Historically, psychoanalytic paradigms viewed panic and acute anxiety as symptomatic manifestations of unresolved unconscious conflict or signal anxiety. The introduction of the DSM-III in 1980 represented an ontological revolution, adopting an empirical, neo-Kraepelinian framework that emphasized reliable descriptive criteria. Panic disorder was demarcated from generalized anxiety disorder based on the qualitative distinctiveness of acute, circumscribed panic attacks versus continuous, free-floating anxiety. Norton and colleagues developed the PAQ precisely to operationalize these standardized descriptive criteria for psychometric testing.

Clark’s Cognitive Model of Panic

The theoretical framework of the PAQ strongly aligns with David M. Clark’s (1986) cognitive model of panic attacks. Clark posited that panic attacks occur when individuals possess a relatively enduring tendency to interpret normal or benign bodily sensations (such as palpitations or dizziness triggered by exercise, caffeine, fatigue, or mild emotional arousal) as signs of an impending catastrophic physical or mental disaster. This catastrophic misinterpretation creates a vicious feedback loop: perception of threat enhances sympathetic autonomic arousal, which intensifies bodily sensations, thereby confirming the initial catastrophic appraisal and culminating in a full-blown panic attack. The PAQ’s explicit measurement of individual somatic sensations alongside specific catastrophic thoughts directly mirrors this cognitive formulation.

Barlow’s Triple Vulnerability and Interoceptive Conditioning Models

The PAQ also draws upon David H. Barlow’s biological and psychological vulnerability framework. Barlow suggested that panic involves a biological false alarm—an inappropriate activation of the fight-or-flight defense system in the absence of objective environmental danger. Subsequent learned alarms occur through interoceptive conditioning, wherein internal conditioned stimuli (e.g., elevated heart rate, slight dyspnea) become paired with intense fear. By assessing contextual variations and interoceptive triggers, the PAQ accounts for how conditioned interoceptive stimuli contribute to the chronic maintenance of panic symptomatology.

The Continuity Hypothesis

Underpinning the construction of the PAQ was Norton and colleagues’ “continuity hypothesis” of panic disorder. Contrary to the view that panic disorder represents a qualitatively distinct biomedical disease entity separated from normality by a discrete categorical boundary, Norton proposed that panic attacks exist along a continuum of frequency and severity. By utilizing the PAQ in large university and community populations, Norton demonstrated that a substantial proportion of individuals who do not meet clinical criteria for panic disorder nonetheless experience occasional panic attacks phenomenologically indistinguishable from those observed in clinical psychiatric cohorts. The PAQ was designed to provide the dimensional measurement necessary to study this continuum.

7. Validity

The validity of the Panic Attack Questionnaire has been supported across numerous empirical investigations involving nonclinical community samples, undergraduate cohorts, and clinical patient samples presenting to specialized anxiety disorder clinics.

Construct and Criterion Validity

Criterion validity was established in foundational studies by Norton et al. (1985, 1986), wherein the PAQ reliably discriminated between individuals independently diagnosed with panic disorder, those exhibiting infrequent nonclinical panic attacks, and non-panicking controls. Patients with diagnosed panic disorder scored significantly higher on symptom intensity ratings, reported a higher frequency of spontaneous attacks, and exhibited significantly greater levels of anticipatory anxiety and behavioral disruption than nonclinical panickers.

Furthermore, the PAQ demonstrated strong construct validity in laboratory biological challenge studies. When participants classified as panickers or non-panickers via the PAQ were exposed to hyperventilation or carbon dioxide inhalations, those identified as panickers exhibited significantly greater subjective anxiety, more intense somatic symptoms matching their PAQ profiles, and a higher frequency of laboratory-induced panic episodes, corroborating the behavioral validity of the scale.

Convergent Validity

Convergent validity has been repeatedly demonstrated through significant positive correlations with established measures of panic, anxiety sensitivity, and generalized distress. In an extensive psychometric study by Cox, Cohen, Direnfeld, and Swinson (1996), the PAQ symptom severity scales exhibited moderate-to-strong correlations with the Beck Anxiety Inventory (BAI), with correlation coefficients typically ranging between r = .50 and r = .68 (p < .001). This confirms that the PAQ accurately captures the physiological and neurovegetative symptom clusters assessed by gold-standard anxiety indices.

Additionally, the cognitive symptom subscales of the PAQ (e.g., fear of dying, fear of losing control) correlate significantly with the Anxiety Sensitivity Index (ASI; Reiss et al., 1986), yielding coefficients between r = .45 and r = .62. Significant correlations have also been observed between the PAQ agoraphobic avoidance items and the Mobility Inventory for Agoraphobia (MIA; Chambless et al., 1985), supporting the instrument’s capacity to reflect secondary avoidance patterns.

Discriminant Validity

The PAQ demonstrates satisfactory discriminant validity when compared to instruments measuring non-somatic or unrelated constructs. Although panic attacks often co-occur with depressive symptoms, correlations between the PAQ somatic symptom subscales and measures of depression such as the Beck Depression Inventory (BDI) are consistently lower (typically r = .25 to .38) than those observed between the PAQ and anxiety-specific scales. Furthermore, when controlling for general negative affectivity, the somatic and cognitive symptom items of the PAQ retain unique predictive variance for panic-specific behavioral outcomes, indicating that the instrument does not simply assess generalized neuroticism or demoralization.

8. Reliability

The psychometric evaluation of the Panic Attack Questionnaire indicates that it possesses sound reliability, characterized by high internal consistency for its symptom rating scales and adequate-to-good test-retest stability across various observation windows.

Internal Consistency

When evaluated as a continuous metric of panic symptom severity, the symptom scale of the PAQ demonstrates high internal consistency across diverse cohorts:

• Overall Symptom Scale: In both clinical and nonclinical samples, Cronbach’s alpha coefficients for the full 13-to-14 symptom intensity inventory consistently range between α = .84 and α = .91 (Cox et al., 1996; Norton et al., 1985).
• Somatic Subscale: The cluster of items measuring physiological and autonomic activation (cardiovascular, respiratory, neurological) yields internal consistency estimates typically between α = .82 and α = .88.
• Cognitive Subscale: The items assessing catastrophic cognitions (fear of dying, going crazy, losing control, derealization) yield reliability estimates ranging from α = .74 to α = .82, which is acceptable given the smaller number of items in this subscale.

Test-Retest Reliability

Evaluating the test-retest reliability of the PAQ requires separating stable trait-like constructs (such as characteristic symptom severity during an attack) from episodic temporal variables (such as attack frequency over a brief interval):

• Symptom Severity Ratings: Across a two- to three-week retest interval in untreated nonclinical cohorts, the stability coefficient for typical panic attack symptom intensity ratings ranges from r = .72 to r = .83, indicating robust temporal consistency in how individuals report their typical panic phenomenology.
• Classification Stability: When categorizing individuals as panickers versus non-panickers based on standardized classification criteria, the concordance rate (Cohen’s κ) over a 2-week to 1-month interval generally ranges from κ = .65 to κ = .78. Shifts in classification typically occur among infrequent panickers whose attack frequency hovers near zero, whereas individuals with frequent attacks demonstrate high stability.
• Long-term Stability: Across extended follow-up periods (e.g., 6 to 12 months) without therapeutic intervention, panic frequency naturally fluctuates in nonclinical samples, reflecting the episodic, waxing-and-waning course of subclinical panic attacks in community populations.

9. Factor Analysis

Numerous exploratory factor analyses (EFA) and confirmatory factor analyses (CFA) have been conducted on the symptom severity sections of the Panic Attack Questionnaire to clarify the latent structure of panic attacks as defined by the DSM criteria.

Factor Solutions in the Literature

Early psychometric evaluations conducted on both clinical cohorts and nonclinical panickers revealed that the somatic and cognitive symptoms of panic do not form a strictly unidimensional construct. Instead, exploratory analyses consistently yield a multi-factor structural model, most frequently resolving into three or four correlated latent factors:

1. Cardiorespiratory and Autonomic Arousal Factor

This primary factor typically accounts for the largest proportion of common variance (often between 25% and 35%). It is defined by high factor loadings (> .60) from physical items including:

• Palpitations / pounding heart
• Shortness of breath / smothering sensations
• Chest pain or discomfort
• Choking sensations
• Sweating and trembling

2. Catastrophic Cognitive and Dissociative Factor

A second distinct factor captures the catastrophic psychological appraisal and altered sensory perception components of the attack:

• Fear of dying (loadings typically .65 – .80)
• Fear of going crazy or losing control (loadings .70 – .82)
• Feelings of unreality (derealization) or detachment (depersonalization)

3. Vestibular and Neurosensory Factor

A third factor, frequently extracted across both clinical and student samples, isolates neurosensory and vasomotor symptoms:

• Dizziness, unsteadiness, or feeling faint
• Paresthesias (numbness or tingling sensations)
• Hot flashes or chills
• Nausea or gastrointestinal distress

Confirmatory Factor Analysis and Model Fit

Subsequent confirmatory factor analytic studies (e.g., Cox et al., 1996; Norton et al., 1992) evaluating competing structural configurations (unidimensional vs. hierarchical vs. oblique multi-factor models) have demonstrated that a multidimensional, oblique three- or four-factor model provides superior fit to the empirical data compared to a single-factor model. In CFA evaluations, acceptable fit indices have been reported for multi-factor configurations:

• Comparative Fit Index (CFI): Values typically exceed .92 – .95.
• Tucker-Lewis Index (TLI): Consistently above .90.
• Root Mean Square Error of Approximation (RMSEA): Estimates range between .048 and .065, indicating good absolute fit.
• Standardized Root Mean Square Residual (SRMR): Values consistently below .06.

These findings substantiate the clinical conceptualization that while panic is an integrated acute syndrome, individuals exhibit distinct symptomatic profiles—such as predominantly respiratory, predominantly cardiac, or predominantly cognitive phenotypes.

10. Instrument / Measurement Tool

The Panic Attack Questionnaire is a comprehensive, multi-section self-report instrument. It combines diagnostic instructional definitions, quantitative Likert-type scales, categorical forced-choice items, and open-ended qualitative prompts to construct a comprehensive profile of panic phenomenology.

Structural Organization and Format

• Administration Type: Paper-and-pencil or digital self-report questionnaire; can be self-administered or utilized as a semi-structured interview schedule.
• Respondent Population: Adults and adolescents aged 16 years and older; validated in both nonclinical community/student populations and clinical anxiety disorder cohorts.
• Completion Time: Approximately 15 to 25 minutes, depending on the individual’s panic history (respondents with no history of panic complete only initial screening sections).
• Major Sections:
  • Diagnostic Definition and Screening: Presents respondents with an explicit, formal description of a panic attack based on DSM criteria, asking whether they have ever experienced such an episode.
  • Frequency and Temporal Parameters: Assesses lifetime attack occurrence, past-year frequency, past-month frequency, typical onset speed (seconds to minutes), and average attack duration.
  • Symptom Profile and Severity: A structured checklist of core somatic, cognitive, and sensory panic symptoms rated on a quantitative intensity scale.
  • Situational Triggers and Context: Explores the context of attacks (e.g., uncued/spontaneous, situationally predisposed, public vs. private settings, nocturnal panic).
  • Anticipatory Anxiety and Avoidance: Evaluates persistent worry about future attacks, changes in daily behavior, and avoidance of specific environments.
  • Medical History and Treatment: Qualitative and categorical items assessing medical evaluations (e.g., ER visits, cardiology workups), current and past psychotropic medication, psychotherapy, and somatic substance use (caffeine, alcohol, stimulants).
• Response Scales:
  • Categorical: Yes / No responses for initial screening, presence of individual symptoms, and medical consultation.
  • Numerical Frequency: Direct numeric estimation for attack counts (e.g., number of attacks in the past 3 weeks / 12 months).
  • Severity Likert Scales: Symptom intensity is typically rated on a 5-point Likert scale: 1 = None / Not at all present, 2 = Mild, 3 = Moderate, 4 = Severe, 5 = Very severe (some research variants utilize a 0 to 4 scale).
• Scoring and Interpretation Procedures:
  • Total Symptom Severity Score: Calculated by summing the intensity ratings across all evaluated somatic and cognitive symptom items (theoretical range varies by version, typically 13 to 65 for a 13-item 5-point scale). Higher scores indicate greater acute panic severity.
  • Diagnostic Thresholding: To satisfy criteria for a “full-blown” panic attack according to classic DSM criteria, the episode must feature the sudden onset of intense fear or discomfort accompanied by at least 4 of the core somatic/cognitive symptoms rated at or above a predefined severity threshold (e.g., rated ≥ 2 or ≥ 3). Episodes featuring fewer than 4 symptoms are classified as “limited-symptom attacks.”
  • Subscale Scores: Separate somatic (cardiorespiratory, neurosensory) and cognitive subscores can be computed by summing items within their respective factor dimensions.

11. Permissions & Fee and Test Year

• Year of Initial Publication: 1985 (with subsequent modifications in 1986, 1992, and the PAQ-R / PAQ-IV variants in the 1990s).
• Original Developers: G. Ron Norton, Bruce Harrison, Jerry Hauch, and Linda Rhodes (University of Winnipeg).
• Copyright and Availability: The Panic Attack Questionnaire was developed under academic research auspices and published within peer-reviewed academic literature (Journal of Abnormal Psychology). The instrument is not commercially distributed by a proprietary commercial psychometric testing corporation.
• Usage Permissions: The PAQ is generally accessible for non-commercial academic, scientific, and clinical research purposes under fair-use conventions. Researchers wishing to reproduce, modify, or translate the questionnaire into other languages are expected to properly cite the foundational publications and obtain permission from the copyright holders (the original authors or the American Psychological Association, depending on the specific published version).
• Fee: There is no standard commercial purchase fee or per-use royalty for non-commercial educational and research administration.

12. References

• Barlow, D. H. (2002). Anxiety and its disorders: The nature and treatment of anxiety and panic (2nd ed.). Guilford Press.
• Chambless, D. L., Caputo, G. C., Jasin, S. E., Gracely, E. J., & Williams, C. (1985). The Mobility Inventory for Agoraphobia. Behaviour Research and Therapy, 23(1), 35–44. https://doi.org/10.1016/0005-7967(85)90140-8
• Clark, D. M. (1986). A cognitive approach to panic. Behaviour Research and Therapy, 24(4), 461–470. https://doi.org/10.1016/0005-7967(86)90011-2
• Cox, B. J., Cohen, D. C., Direnfeld, D. M., & Swinson, R. P. (1996). Does the Beck Anxiety Inventory measure anything beyond panic attack symptoms? Behaviour Research and Therapy, 34(11–12), 949–954. https://doi.org/10.1016/S0005-7967(96)00067-1
• Norton, G. R., Cox, B. J., & Malan, J. (1992). Nonclinical panickers: A critical review. Clinical Psychology Review, 12(2), 121–139. https://doi.org/10.1016/0272-7358(92)90111-A
• Norton, G. R., Dorward, J., & Cox, B. J. (1986). Factors associated with panic attacks in nonclinical subjects. Behavior Therapy, 17(3), 239–252. https://doi.org/10.1016/S0005-7894(86)80054-0
• Norton, G. R., Harrison, B., Hauch, J., & Rhodes, L. (1985). Characteristics of people with infrequent panic attacks. Journal of Abnormal Psychology, 94(2), 216–221. https://doi.org/10.1037/0021-843X.94.2.216
• Reiss, S., Peterson, R. A., Gursky, D. M., & McNally, R. J. (1986). Anxiety sensitivity, anxiety frequency and the prediction of fearfulness. Behaviour Research and Therapy, 24(1), 1–8. https://doi.org/10.1016/0005-7967(86)90143-9

13. Items of the Scale