Cluster Headache: What Could Be Causing It

What Is Cluster Headache and Why Does It Happen?

Cluster headache belongs to a family of disorders called trigeminal autonomic cephalalgias (TACs). The pain is strictly one-sided, centered around the eye or temple, and accompanied by autonomic features like tearing, nasal congestion, or eyelid drooping on the same side. Attacks group into "clusters" lasting weeks to months, separated by remission periods that can stretch for a year or longer.

The Hypothalamic Clock Theory

The leading explanation centers on the hypothalamus. Functional MRI studies first published by May et al. In The Lancet demonstrated ipsilateral posterior hypothalamic activation during cluster headache attacks [1]. The hypothalamus serves as the body's master circadian pacemaker, which explains why attacks tend to strike at predictable times, often 1 to 2 hours after falling asleep or during the early morning hours.

This circadian signature distinguishes cluster headache from other severe headache types. Positron emission tomography (PET) studies have confirmed that the hypothalamic gray matter is structurally different in cluster headache patients compared to controls, with increased neuronal density in the region responsible for sleep-wake regulation [2].

The Trigeminovascular Pathway

Once the hypothalamus initiates the cascade, the trigeminal nerve and its connections to the superior salivatory nucleus drive the pain and autonomic symptoms. Calcitonin gene-related peptide (CGRP) and vasoactive intestinal peptide (VIP) levels rise sharply in the jugular venous blood during an attack [3]. CGRP causes vasodilation of meningeal blood vessels and transmits nociceptive signals. This biochemical fingerprint has guided the development of newer targeted therapies, including galcanezumab, an anti-CGRP monoclonal antibody now approved for episodic cluster headache prevention.

Who Gets Cluster Headache? Risk Factors and Epidemiology

Cluster headache has a lifetime prevalence of roughly 124 per 100,000 people according to a meta-analysis published in Cephalalgia (N=302,725 across 15 population-based studies) [4]. Men are affected three to four times more often than women, though this gap has narrowed in recent decades, possibly reflecting improved diagnosis in women rather than a true shift in incidence.

Genetics

A first-degree relative of someone with cluster headache carries a 14- to 39-fold increased risk of developing the condition compared to the general population [5]. The HCRTR2 gene, which encodes the hypocretin (orexin) receptor 2, has been implicated in multiple genome-wide association studies. Hypocretin neurons reside in the posterior hypothalamus and regulate arousal, reinforcing the hypothalamic origin theory.

Age of Onset and Smoking History

Peak onset falls between ages 20 and 40. Historically, 65% to 80% of cluster headache patients were current or former smokers, although smoking does not appear to trigger individual attacks and smoking cessation does not reliably end cluster periods [6]. The association may reflect shared genetic vulnerability or lifestyle patterns rather than a direct causal link.

Comorbid Conditions

Sleep apnea appears more frequently in cluster headache patients than in the general population. A study in the Journal of Headache and Pain found obstructive sleep apnea in 41 of 100 consecutive cluster headache patients evaluated by polysomnography [7]. Treating the sleep apnea reduced cluster headache burden in a subset of those individuals, suggesting that nocturnal hypoxemia may serve as a modifiable trigger.

Common Triggers During an Active Cluster Period

Outside of a cluster bout, most triggers have no effect. This is a defining paradox of the disorder: alcohol will provoke an attack within minutes during an active cluster period but has zero headache-inducing effect during remission.

Alcohol and Vasodilators

Alcohol is the single most consistent trigger reported across patient surveys, provoking attacks in 50% to 80% of patients during active bouts [8]. Nitroglycerin and other nitric oxide donors can reproduce a full cluster attack in susceptible individuals during a bout, a finding that helped researchers confirm the vascular component of the pain pathway. Histamine infusion produces the same effect.

Altitude, Heat, and Sleep Disruption

Changes in barometric pressure, exposure to strong solvents, and disrupted sleep schedules are commonly reported triggers. Shift workers with cluster headache often experience longer and more severe bouts. The connection to sleep architecture is not incidental: cluster attacks preferentially emerge during REM sleep, and melatonin secretion is blunted in patients during active cluster periods [9].

What Does Not Cause Cluster Headache

Stress, dietary triggers like aged cheese or chocolate, and hormonal fluctuations (common migraine triggers) do not reliably provoke cluster headache attacks. This distinction matters clinically. Patients who arrive with a misdiagnosis of migraine may have spent years avoiding irrelevant triggers while the actual pattern, a circadian and seasonal rhythm with autonomic features, went unrecognized.

How Cluster Headache Is Diagnosed

There is no blood test or imaging study that confirms cluster headache. Diagnosis is clinical, based on the International Classification of Headache Disorders, 3rd edition (ICHD-3) criteria [10].

ICHD-3 Diagnostic Criteria

The ICHD-3 requires at least five attacks of severe unilateral orbital, supraorbital, or temporal pain lasting 15 to 180 minutes if untreated. Each attack must include at least one ipsilateral autonomic sign (lacrimation, conjunctival injection, nasal congestion, rhinorrhea, forehead and facial sweating, miosis, ptosis, or eyelid edema) or a sense of restlessness and agitation. Attacks occur with a frequency between one every other day and eight per day during a bout.

The Role of Neuroimaging

Brain MRI with contrast is recommended at initial presentation to rule out secondary causes. The European Headache Federation guidelines state: "All patients presenting with a TAC phenotype for the first time should undergo brain MRI, preferably with attention to the pituitary region and cavernous sinus" [11]. Pituitary adenomas, cavernous sinus meningiomas, and vertebral artery dissections have all been reported to mimic cluster headache.

Distinguishing Cluster Headache From Migraine

Misdiagnosis remains a significant problem. A U.S. Survey of 1,134 cluster headache patients found a mean diagnostic delay of 5.3 years, with 77% receiving at least one incorrect diagnosis before the correct one [12]. Migraine was the most common misdiagnosis.

The key clinical differences: cluster attacks are shorter (15-180 minutes vs. 4-72 hours for migraine), always unilateral without side-shifting, and accompanied by ipsilateral autonomic signs. Cluster patients pace, rock, or bang their heads. Migraine patients seek dark, quiet stillness.

Dr. Peter Goadsby, a neurologist at King's College London and a leading researcher in TAC biology, has written: "The most useful diagnostic question in headache medicine is, 'What do you do during an attack?' A patient who lies still likely has migraine. A patient who cannot sit still almost certainly has a trigeminal autonomic cephalalgia" [13].

Differential Diagnosis: Other Conditions That Mimic Cluster Headache

Several conditions produce severe unilateral head pain with autonomic features. Ruling them out shapes the treatment plan.

Paroxysmal Hemicrania

Paroxysmal hemicrania shares the same pain location and autonomic signs but produces shorter attacks (2-30 minutes), higher daily frequency (more than five per day), and responds completely to indomethacin. An indomethacin trial is mandatory when the clinical picture is ambiguous. The European Headache Federation considers the "indomethacin response" a diagnostic criterion for paroxysmal hemicrania [11].

SUNCT and SUNA

Short-lasting unilateral neuralgiform headache attacks with conjunctival injection and tearing (SUNCT) and short-lasting unilateral neuralgiform headache attacks with cranial autonomic symptoms (SUNA) last only 1 to 600 seconds. Their brevity separates them from cluster headache, though overlap cases exist.

Secondary Causes

Carotid or vertebral artery dissection can present with severe periorbital pain and a partial Horner syndrome, closely mimicking a cluster attack. Pituitary apoplexy, Tolosa-Hunt syndrome, and cavernous sinus thrombosis belong on the differential as well. Any new-onset cluster-like headache in a patient over 50 warrants urgent imaging and vascular studies [14].

Acute Treatment: Stopping an Attack

Two therapies have the strongest evidence base for aborting a cluster headache attack.

High-Flow Oxygen

Inhalation of 100% oxygen at 12 to 15 liters per minute through a non-rebreather mask for 15 minutes provides relief in roughly 78% of attacks. The landmark randomized controlled trial by Cohen et al. In JAMA (N=109) showed that oxygen was superior to medical air (78% vs. 20% response at 15 minutes, P<0.001) [15]. Oxygen has no contraindications, no drug interactions, and no systemic side effects. Its main limitation is portability.

Subcutaneous Sumatriptan

Subcutaneous sumatriptan 6 mg relieves headache within 15 minutes in approximately 75% of attacks [16]. Oral triptans are too slow for a condition where attacks peak within 5 to 10 minutes. Sumatriptan nasal spray (20 mg) provides an intermediate option for patients who cannot tolerate injections, though response rates are lower.

The American Headache Society guideline notes: "Subcutaneous sumatriptan and high-flow oxygen are the two treatments with Level A evidence for acute cluster headache" [17].

Treatments to Avoid

Opioids are ineffective and carry addiction risk. Over-the-counter analgesics like ibuprofen and acetaminophen act too slowly. Ergotamine suppositories have historical use but limited modern evidence and more side effects than triptans.

Preventive Therapy: Reducing Attack Frequency

Preventive treatment begins as soon as a cluster bout is recognized and continues until the bout's expected end.

Verapamil

Verapamil is the most widely used preventive agent worldwide. Effective doses range from 240 mg to 960 mg per day, divided into two or three doses. ECG monitoring is required because verapamil can prolong the PR interval and cause heart block. A double-blind trial published in Neurology demonstrated that verapamil 360 mg/day significantly reduced attack frequency compared to placebo in the second week of treatment [18].

Galcanezumab

The phase 3 CGAL trial (N=106) evaluated galcanezumab 300 mg subcutaneous injection monthly for episodic cluster headache. The galcanezumab group experienced a mean reduction of 8.7 attacks per week at weeks 1 through 3, compared to 5.2 in the placebo group (P=0.04) [19]. The FDA approved galcanezumab (Emgality) for episodic cluster headache in 2019, making it the first drug to receive this specific indication. It did not show efficacy in chronic cluster headache in a separate trial.

Transitional (Bridge) Therapies

Short courses of oral corticosteroids (prednisone 60-80 mg tapered over 2-3 weeks) or a greater occipital nerve block with corticosteroid and local anesthetic can suppress attacks while waiting for verapamil to reach therapeutic levels. A randomized trial found that suboccipital steroid injection reduced attack frequency by 57.7% at one week compared to placebo [20].

Lithium and Other Options

Lithium carbonate (600-1200 mg/day) has decades of observational support for chronic cluster headache but requires serum level and thyroid monitoring. Topiramate, melatonin (10 mg at bedtime), and civamide intranasal capsaicin have weaker but supportive evidence for selected patients.

Living With Cluster Headache: Practical Steps

Patients who keep an attack diary that records timing, duration, and potential triggers provide their clinicians with the pattern data needed to optimize treatment. Maintaining a consistent sleep-wake schedule, avoiding alcohol during cluster periods, and having an oxygen setup accessible at home and work reduces the impact of breakthrough attacks.

Cluster headache carries a measurable psychological burden. A European survey of 294 patients found that 55% reported depression and 64% reported anxiety during cluster bouts [21]. Screening for mood disorders and referral for psychological support should be part of routine management.

The Cluster Headache Support Group (OUCH UK and Clusterbusters in the U.S.) provides peer networks. Patients report that connecting with others who understand the severity of the pain, sometimes called "suicide headache" in older literature, reduces isolation and improves coping during active bouts.