Cluster Headache Clinical Phenotypes: Tobacco Nonexposed (Never Smoker and No Parental Secondary Smoke Exposure as a Child) versus Tobacco-Exposed: Results from the United States Cluster Headache Survey

Headache. 2018 May;58(5):688-699. doi: 10.1111/head.13295. Epub 2018 Mar 14.


Objective: To present results from the United States Cluster Headache Survey comparing the clinical presentation of tobacco nonexposed and tobacco-exposed cluster headache patients.

Background: Cluster headache is uniquely tied to a personal history of tobacco usage/cigarette smoking and, if the individual cluster headache sufferer did not smoke, it has been shown that their parent(s) typically did and that individual had significant secondary smoke exposure as a child. The true nontobacco exposed (no personal or secondary exposure) cluster headache sufferer has never been fully studied.

Methods: The United States Cluster Headache Survey consisted of 187 multiple choice questions related to cluster headache including: patient demographics, clinical headache characteristics, family history, triggers, smoking history (personal and secondary), and headache-related disability. The survey was placed on a website from October through December 2008.

Results: One thousand one hundred thirty-four individuals completed the survey. One hundred thirty-three subjects or 12% of the surveyed population had no personal smoking/tobacco use history and no secondary smoke exposure as an infant/child, thus a nontobacco exposed population. In the nonexposed population, there were 87 males and 46 females with a gender ratio of 1.9:1. Episodic cluster headache occurred in 80% of nonexposed subjects. One thousand and one survey responders or 88% were tobacco-exposed (729 males and 272 females) with a gender ratio of 2.7:1. Eighty-three percent had a personal smoking history, while only 17% just had parents who smoked with secondary smoke exposure. Eighty-five percent of smokers had double exposure with a personal smoking history and secondary exposure as a child.

Significant highlights from the survey: Nonexposed cluster headache subjects are significantly more likely to develop cluster headache at ages 40 years and younger, while the exposed sufferers are significantly more likely to develop cluster headache at 40 years of age and older. Nonexposed patients have a statistically significant higher frequency of a migraine family history. The exposed population is statistically significantly more likely to have a history of head trauma 19% vs the nonexposed population 10% (P = .02). Tobacco exposed are significantly more likely to transition from episodic to chronic cluster headache (23% vs 14%, P = .02). Cranial autonomic symptoms as well as agitation are more common in tobacco exposed. Nonexposed are less likely to have specific cluster headache triggers. Exposed are significantly more likely to be triggered by alcohol. Tobacco exposed are significantly heavier caffeine users than nonexposed. Nonexposed are significantly more likely to have cluster headache cycles that vary throughout the year than exposed (52% vs 40%, P = .02). Exposed are much more likely to develop cluster headache from 12 am to 6 am than non exposed. Exposed experience significantly more frequent attacks per day and longer duration cycles than nonexposed. A significantly larger percent of the exposed population (57%) has suicidal ideations with their syndrome than nonexposed (43%) (P = .003). In regard to disability, both subtypes are disabled by their headaches, but exposed have more work related disability and lost home-days from headache. Both subgroups have a poor overall response to preventive and abortive medication outside of inhaled oxygen and injectable sumatriptan.

Conclusion: Cluster headache sufferers who were never exposed to tobacco (personal or secondary as a child) appear to present uniquely compared to the tobacco exposed subgroup. The tobacco exposed clinical phenotype appears to have a more severe syndrome based on attack frequency, cycle duration, and headache related disability. Tobacco exposure is associated with cluster headache chronification. The nonexposed subtype appears to have an earlier age of onset, higher rate of familial migraine, and less circadian periodicity and daytime entrainment, suggesting a possible different underlying pathology than in the tobacco exposed sub-form.

Keywords: cigarette; cluster headache; secondary exposure; smoking; tobacco.

MeSH terms

  • Adult
  • Age of Onset
  • Chronic Disease / epidemiology
  • Cluster Headache / chemically induced
  • Cluster Headache / epidemiology*
  • Cluster Headache / etiology
  • Cluster Headache / physiopathology*
  • Comorbidity
  • Craniocerebral Trauma / epidemiology*
  • Disease Susceptibility
  • Female
  • Health Surveys / statistics & numerical data
  • Humans
  • Male
  • Middle Aged
  • Migraine Disorders / epidemiology*
  • Smoking / adverse effects
  • Smoking / epidemiology*
  • Suicidal Ideation*
  • Tobacco Smoke Pollution / adverse effects
  • Tobacco Smoke Pollution / statistics & numerical data*
  • United States / epidemiology


  • Tobacco Smoke Pollution