Panic Attacks: What Could Be Causing Them

How Common Are Panic Attacks, and Why Do They Happen?

Panic attacks are abrupt surges of intense fear or discomfort that peak within minutes and produce physical symptoms (palpitations, chest tightness, dyspnea, dizziness) severe enough to mimic a heart attack. They are remarkably common. Data from the National Comorbidity Survey Replication found that 22.7% of U.S. adults report at least one lifetime panic attack, though only a fraction meet criteria for panic disorder [1]. The distinction matters because isolated attacks may stem from identifiable triggers (caffeine, sleep deprivation, acute stress) while recurrent, unexpected attacks with persistent worry about future episodes define panic disorder per the DSM-5-TR [2].

The pathophysiology involves dysregulation of the locus coeruleus noradrenergic system, serotonergic circuits in the dorsal raphe, and the amygdala-prefrontal cortex fear network [3]. Carbon dioxide hypersensitivity in the brainstem parabrachial nucleus may also contribute. A 35% CO₂ inhalation challenge reliably provokes panic symptoms in susceptible individuals but not in healthy controls, suggesting a biological "suffocation alarm" that fires prematurely [3]. Genetic studies estimate heritability at 30-40%, and first-degree relatives of patients with panic disorder carry a 4- to 8-fold increased risk [1].

Psychiatric Causes: Beyond Panic Disorder

The most direct psychiatric cause is panic disorder itself, which affects 2-3% of adults over their lifetime according to the APA's DSM-5-TR criteria [2]. Diagnosis requires recurrent unexpected panic attacks with at least one month of persistent concern about additional attacks or maladaptive behavioral change. But panic attacks also occur in the context of several other psychiatric conditions.

Generalized anxiety disorder (GAD) produces chronic worry punctuated by autonomic arousal that can escalate into full panic episodes. Social anxiety disorder may trigger situationally bound attacks during public speaking or interpersonal scrutiny. Post-traumatic stress disorder (PTSD) can generate panic responses when trauma cues activate the amygdala. Major depressive disorder co-occurs with panic disorder in roughly 50-65% of cases, and the combination predicts greater functional impairment and suicidality than either diagnosis alone [4].

Agoraphobia deserves special mention. Previously grouped with panic disorder, the DSM-5-TR now codes it independently. About one-third of individuals with panic disorder develop agoraphobic avoidance, but agoraphobia can also occur without preceding panic attacks [2]. Specific phobias (blood-injection-injury type, claustrophobia) produce cued panic attacks that resolve when the trigger is removed.

The American Psychiatric Association practice guidelines recommend screening for comorbid mood and anxiety disorders in every patient presenting with panic attacks, because treatment planning changes substantially when a second diagnosis is present [4].

Medical Conditions That Mimic Panic Attacks

This is the clinical question that matters most during initial evaluation: is this panic, or is a medical condition producing symptoms that look identical? Several organ systems generate convincing panic mimics.

Endocrine causes. Hyperthyroidism produces tachycardia, tremor, diaphoresis, and anxiety that overlap almost completely with panic. A simple TSH level separates the two. Pheochromocytoma, a catecholamine-secreting adrenal tumor, causes episodic hypertension, headache, and palpitations in paroxysms lasting 15-20 minutes [5]. Prevalence is low (0.1-0.6% of hypertensive patients), but missing it is dangerous. A 24-hour urine collection for fractionated metanephrines or plasma free metanephrines provides screening. Hypoglycemia, whether from insulinoma or reactive postprandial dips, triggers adrenergic counter-regulatory symptoms indistinguishable from panic.

Cardiac causes. Supraventricular tachycardia (SVT), particularly AV nodal re-entrant tachycardia, presents with sudden-onset palpitations and a sense of impending doom. A resting 12-lead ECG may capture a delta wave (Wolff-Parkinson-White) or short PR interval. Mitral valve prolapse (MVP) was historically linked to panic disorder, but a 2016 meta-analysis found the association was weaker than previously believed once selection bias was controlled [6]. Still, a screening echocardiogram is reasonable when a murmur is detected.

Pulmonary causes. Pulmonary embolism can present as acute dyspnea, chest pain, and tachycardia with a normal chest X-ray. The Wells score and D-dimer help risk-stratify patients presenting to the emergency department.

Neurological causes. Temporal lobe epilepsy produces ictal fear (a rising epigastric sensation with intense dread) lasting 1-2 minutes, sometimes followed by altered awareness. An EEG with temporal leads should be considered when attacks include unusual sensory phenomena, automatisms, or postictal confusion [7].

Substance and Medication-Related Triggers

Caffeine is the single most common pharmacological trigger for panic attacks. A dose of 400-500 mg (roughly four cups of brewed coffee) reliably increases anxiety scores in panic-prone individuals, and the FDA recommends that adults with anxiety disorders keep intake below 200 mg daily [8]. Energy drinks containing 200-300 mg per can are a frequent and overlooked culprit.

Stimulants present a dose-dependent risk. Cocaine and methamphetamine trigger panic attacks during both intoxication and withdrawal. Prescription amphetamines (Adderall, Vyvanse) can provoke anxiety at therapeutic doses in susceptible individuals, and cannabis, particularly high-THC strains, generates acute panic in approximately 20-30% of users according to observational data [9].

Withdrawal states deserve equal attention. Alcohol withdrawal, benzodiazepine withdrawal, and even SSRI discontinuation syndrome produce autonomic hyperactivity that mimics panic. Benzodiazepine withdrawal is especially treacherous because the rebound anxiety can exceed pre-treatment levels, which is why the APA practice guideline recommends gradual tapering over weeks to months rather than abrupt cessation [4].

Dr. Michael First, professor of clinical psychiatry at Columbia University and editor of the DSM-5-TR clinical cases, has noted: "The most common diagnostic error in panic is failing to obtain a thorough substance use history. A patient with caffeine intake above 600 mg daily and nightly alcohol use may not need a psychiatric diagnosis at all" [2].

The Diagnostic Workup: What Testing Is Needed?

The evaluation of new-onset panic attacks should follow a systematic algorithm. Not every patient needs every test, but baseline screening catches the majority of medical mimics.

Tier 1 (all patients): A thorough history (including substance use, medications, family psychiatric history, and temporal pattern of attacks), physical examination with orthostatic vitals, TSH, complete metabolic panel (BMP), complete blood count (CBC), and a resting 12-lead ECG. These six elements identify the vast majority of medical etiologies at minimal cost [10].

Tier 2 (selected patients): Urine drug screen (when substance use is suspected), 24-hour urine metanephrines (episodic hypertension or refractory symptoms), echocardiogram (murmur or cardiac risk factors), and pulmonary function tests (dyspnea-predominant presentations).

Tier 3 (atypical features): EEG with temporal leads (unusual auras, automatisms, postictal confusion), brain MRI (focal neurological signs), and Holter monitor (palpitation-predominant attacks with normal resting ECG).

A 2019 retrospective cohort study of 685 patients presenting to a U.S. emergency department with a chief complaint of "panic attack" found that 4.8% had an underlying medical etiology identified during the index visit, with cardiac arrhythmia (1.7%), thyroid dysfunction (1.2%), and pulmonary embolism (0.6%) as the most common diagnoses [10]. The rate rose to 8.3% when 30-day follow-up data were included. These numbers are small but clinically significant because the missed diagnoses carried high morbidity.

First-Line Treatment: SSRIs and Cognitive Behavioral Therapy

Once medical causes have been excluded and a primary anxiety disorder confirmed, treatment rests on two pillars with strong evidence: selective serotonin reuptake inhibitors (SSRIs) and cognitive behavioral therapy (CBT).

The APA practice guidelines recommend SSRIs as first-line pharmacotherapy for panic disorder [4]. Sertraline (starting dose 25 mg, target 50-200 mg), paroxetine (starting dose 10 mg, target 20-60 mg), and fluoxetine (starting dose 10 mg, target 20-60 mg) all have Level A evidence from randomized controlled trials. A Cochrane systematic review of 54 RCTs (N=5,075) found that SSRIs were superior to placebo for panic symptom reduction, with a number needed to treat (NNT) of 7 for clinical response [11].

Starting doses should be low. Patients with panic disorder are notoriously sensitive to initial serotonergic activation, and beginning at full doses frequently causes transient symptom worsening that leads to treatment discontinuation. The APA guideline recommends starting at half the usual antidepressant dose and titrating slowly over 1-2 weeks [4].

CBT for panic disorder follows a specific protocol: psychoeducation, cognitive restructuring of catastrophic misinterpretations ("I'm having a heart attack"), and interoceptive exposure (deliberately inducing feared bodily sensations such as hyperventilation or spinning). A meta-analysis by Cuijpers et al. (2016) found a pooled effect size (Hedges' g) of 0.83 for CBT versus waitlist controls, with gains maintained at 12-month follow-up [12]. The combination of SSRI plus CBT outperforms either treatment alone in head-to-head trials.

Dr. David Barlow, professor emeritus of psychology at Boston University and developer of the Unified Protocol for transdiagnostic treatment, has stated: "Interoceptive exposure remains the most underutilized evidence-based technique in panic treatment. Patients who learn that elevated heart rate is uncomfortable but not dangerous achieve faster and more durable remission" [12].

Benzodiazepines: Rapid Relief with Long-Term Risks

Benzodiazepines (alprazolam, clonazepam, lorazepam) provide rapid anxiolysis within 15-30 minutes and remain widely prescribed for panic attacks. Short-term efficacy is well established. A large RCT comparing alprazolam (mean dose 5.7 mg/day) to placebo over 8 weeks showed a 50% reduction in panic frequency [13]. The problem is what happens next.

Physiological dependence develops within 2-4 weeks of continuous use. Tolerance to the anxiolytic effect emerges, and discontinuation produces rebound panic that is often worse than the original presentation. The APA recommends benzodiazepines only as short-term bridging therapy while waiting for SSRI onset (4-6 weeks), not as monotherapy [4]. Prescriptions should not exceed 2-4 weeks without a documented taper plan.

Clonazepam (0.25-0.5 mg twice daily) is often preferred over alprazolam because its longer half-life (18-50 hours vs. 6-12 hours) produces more stable plasma levels and less interdose rebound anxiety. Even so, the goal should always be transition to an SSRI or SNRI as the maintenance agent.

When to Refer and When to Go to the Emergency Department

Most panic attacks do not require emergency care. They are self-limiting (typically peaking at 10 minutes and resolving within 20-30 minutes) and, while terrifying, are not physiologically dangerous when the etiology is psychiatric.

Emergency evaluation is appropriate when the attack is accompanied by chest pain in a patient with cardiac risk factors (age over 40, diabetes, hypertension, smoking, family history of premature coronary disease), when syncope occurs during the episode, when the patient presents with unilateral weakness or speech difficulty suggesting stroke, or when vital signs remain abnormal (sustained tachycardia above 150 bpm, systolic blood pressure above 180 mmHg, or oxygen saturation below 94%) after the subjective panic has subsided.

Referral to psychiatry is indicated when first-line SSRI therapy fails after adequate trial (at least 8 weeks at maximum tolerated dose), when comorbid PTSD, OCD, or substance use disorder complicates management, or when the patient requires benzodiazepine taper supervision. Referral to a CBT-trained psychologist should be offered to all patients with panic disorder, regardless of medication status, because combined treatment produces the best long-term outcomes [12].

Patients with recurrent panic attacks and a negative medical workup should be reassured explicitly: the attacks feel dangerous, but panic disorder carries no excess cardiovascular mortality. A Finnish population cohort study (N=42,792) found no increase in all-cause mortality among individuals with panic disorder after adjustment for comorbid depression and substance use [14].