Vitals Knowledge Vault

Cluster Headache

6 min read

Cluster Headache

TL;DR

  • Cluster headache is a severe strictly unilateral orbital / supraorbital / temporal pain, lasting 15–180 minutes, with ipsilateral autonomic symptoms and/or marked restlessness/agitation (ICHD-3).
  • Pain location alone is not enough to label this. The full phenotype must fit.
  • The biggest safety risk is false reassurance — dangerous mimics (glaucoma, carotid dissection, infection) can present similarly.
  • Acute treatment (high-flow oxygen, subcutaneous sumatriptan) and preventive treatment (verapamil, prednisone bridge, galcanezumab) are separate lanes.
  • Vitals should limit itself to attack-time logging, sleep-timing review, and red-flag escalation — not diagnosis.

Why it matters for Vitals

  • A user reporting severe one-eye pain needs safety triage first, not a cluster-headache label.
  • Vitals can support structured logging: attack time, side, duration, autonomic features, restlessness, nocturnal wake time, alcohol exposure, and bout patterns.
  • Vitals can support sleep regularity review and possible sleep-apnea flagging, which has a known association with cluster headache.
  • Vitals must not attempt to diagnose cluster headache from wearable data alone, and must not substitute for urgent medical evaluation when red flags are present.
  • The acute vs preventive split matters for coaching: users who confuse them will be frustrated by delayed results.

Key facts

  • Phenotype anchor (ICHD-3): Severe strictly unilateral orbital / supraorbital / temporal pain; 15–180 minutes; 1 attack every other day up to 8/day; ipsilateral cranial autonomic symptoms and/or marked restlessness/agitation.
  • Episodic vs chronic: Most cases are episodic (bouts separated by remissions). Chronic cluster headache persists >1 year without remission.
  • Circadian pattern: Nocturnal attacks are common; many patients describe attacks at consistent times of day. This is real but not universal.
  • Diagnostic delay: Registry data show cluster headache is frequently mislabeled as migraine or sinus disease for years, increasing disability burden.
  • Behavioral differentiator from migraine: Cluster patients tend to pace, rock, or appear unable to keep still. Migraine patients more often seek dark, quiet, and immobility.
  • TAC framing: Cluster headache belongs to the trigeminal autonomic cephalalgias (TACs), not the migraine family. Key differential within TACs: paroxysmal hemicrania (shorter attacks, higher frequency, absolute indomethacin response).

Mechanism summary

  • Working model: trigeminal-autonomic reflex disorder — trigeminal pain signaling + ipsilateral parasympathetic activation → unilateral pain, tearing, conjunctival injection, nasal symptoms, ptosis, or miosis.
  • Strong hypothalamic and circadian involvement is well-supported; posterior hypothalamic activation is a consistent imaging finding.
  • CGRP (calcitonin gene-related peptide) plays a role in trigeminovascular activation.
  • Formal autonomic testing supports dysautonomia as part of the physiology.
  • The REM-sleep theory is contested — attacks are not cleanly locked to REM and the association is weaker than older literature suggested.

What the current evidence suggests

Confirmed

  • ICHD-3 diagnostic criteria are stable and guideline-accepted.
  • High-flow oxygen (100% O₂ at 12 L/min via non-rebreather mask) is one of two strongest acute abortive therapies.
  • Subcutaneous sumatriptan is the fastest pharmacologic acute option with strong RCT support.

Supported

  • Oral triptans are too slow for most classic cluster attacks (route + onset speed issue).
  • Verapamil is the core preventive (not acute) therapy with controlled-trial support; requires ECG-aware follow-up.
  • Prednisone bridge therapy has randomized evidence for short-term attack reduction while prevention is being established.
  • Galcanezumab has randomized-trial support for episodic cluster headache prevention.
  • Actigraphy-level sleep timing and attack logging are defensible wearable-adjacent data uses.
  • Sleep apnea has a recognized association with cluster headache.

Contested / uncertain

  • REM-sleep as a specific attack trigger is not reliably supported.
  • Response to oxygen or triptans is therapeutically useful but not diagnostic proof — these therapies also help other headache disorders.
  • The precise mechanistic pathway from hypothalamus to attack is still incomplete.

Gap

  • No validated consumer wearable can diagnose, detect, or predict cluster headache attacks in real time.
  • No blood test, imaging biomarker, or wearable signal is established for confirmation.

Red flags — when to interrupt cluster framing

The following features should push toward urgent medical evaluation, not routine cluster self-management:

  • Vision loss or major visual change
  • Fixed, abnormal, or nonreactive pupil
  • Corneal haze or red-eye pattern that looks ocular rather than autonomic
  • Proptosis or ophthalmoplegia
  • Fever or systemic illness
  • Focal neurologic deficits
  • Thunderclap onset
  • Neck pain with new Horner-pattern symptoms
  • New onset after age 50

These are not cluster-defining. They widen the differential toward ocular, vascular, infectious, inflammatory, or structural causes.

Acute vs preventive treatment

LaneTherapiesOnset
Acute (abortive)High-flow oxygen, SC sumatriptanMinutes
Bridge (short-term)PrednisoneDays
PreventiveVerapamil, GalcanezumabDays to weeks

Key practical note: These are separate lanes. Predicting that verapamil will abort an active attack is a category error. Coaching users on this distinction reduces frustration and unsafe experimentation.

Vitals wearable boundaries

Defensible:

  • Attack-time logging (timestamp, duration, side, autonomic features, restlessness, nocturnal wake)
  • Sleep timing and regularity review
  • Sleep-disruption correlation with attack patterns
  • Sleep-apnea screening flags when clinically relevant

Not defensible:

  • Real-time attack detection from HRV, HR, motion, or sleep stage
  • HRV-only diagnosis
  • REM-stage prediction for attack timing
  • Oxygen-response-as-proof-of-diagnosis logic

Risks and uncertainty

  • Diagnosis is clinical, not biomarker-confirmed. The phenotype can be misidentified by patients and clinicians.
  • Dangerous secondary causes can mimic the phenotype. Red-flag screening is not optional.
  • The strongest evidence base covers episodic adult male patients; chronic and atypical presentations are harder to manage.
  • The sleep literature is more complicated than simple “cluster patients sleep less” — actigraphy data do not support a simple story.
  • Treatment response does not equal diagnosis. Both oxygen and triptans can help other headache types.

Best stack context

  • Cluster headache is primarily a safety and triage topic for Vitals, not a stack-optimization topic.
  • If a user is on preventive therapy (verapamil, galcanezumab), Vitals coaching can support adherence tracking, timing review, and attack-frequency logging.
  • No peptide or supplement is established as a standard cluster-headache intervention in the evidence base reviewed.

What stays inside this hub

The following are not split into separate notes because they are not independently retrieval-worthy across the vault:

  • Paroxysmal hemicrania differential (context within this hub)
  • Indomethacin response (tied to paroxysmal hemicrania)
  • CGRP mechanism details (cluster-specific, not broadly reusable)
  • Hypothalamic chronobiology specifics (cluster-specific, not broadly reusable)
  • SUNCT/SUNA differential (mentioned here; broader TAC note not yet warranted)
  • Specific acute backup options (octreotide, intranasal zolmitriptan) — these are in the source monograph, not needed as separate vault notes

Evidence grade: Supported (monograph source has Confirmed evidence anchors for phenotype, oxygen, SC sumatriptan; Supported for prevention; Gap for wearable detection). Review: 2026-04-20.

Graph View

Backlinks