Most Patients With Frequent Cluster Headaches Can Predict a Coming Bout

woman with a headache
The researchers aimed to describe the prevalence of pre-cluster (premonitory) symptoms and examine the predictability of an upcoming cluster headache bout.

Patients with episodic cluster headaches (eCH) commonly have pre-cluster symptoms that predict upcoming cluster headache bouts. These findings from a retrospective cross-sectional study were published in Cephalalgia.

Patients (N=100) with frequent eCH were recruited at the Danish Headache Center from January 2019 to April 2020 to participate in this study. During a semi-structured interview, researchers assessed patients for 9 specific local symptoms and 15 general signs or symptoms.

Patients were a mean age of 42.2 years (interquartile range [IQR], 34-51), the male-to-female ratio was 3.8 to 1, patients had had CH for an average of 15 years (IQR, 6-20), and 10% had comorbid migraine.

Most patients (86%) had 1 or more pre-cluster symptoms that occurred an average of 6.8 days (IQR, 3-14) prior to their CH. The most frequently reported pre-cluster symptoms were dull or aching sensations (62%; mean, 5.8 days; IQR, 3-13), shadow attacks (60%; mean, 6.5 days; IQR, 4-14), and neck pain (41%; mean, 4.8 days; IQR, 2-7). Shadow attacks are those that do not require abortive treatment.

Nearly a third of patients (29%) had at least 1 cranial autonomic symptom, such as lacrimation, ptosis, nasal congestion, or eye redness, at an average of 4.5 days (IQR, 2-10) before the CH bout. They frequently reported symptoms of fatigue (37%; mean, 5.3 days; IQR, 2-10), irritability (33%; mean, 5.2; IQR, 3-7 days), and sleep difficulties (19%; mean, 5.3 days; IQR, 3-7).

Most patients (57%) reported they could predict a bout of CH an average of 4.6 days (IQR, 2-7) prior to onset. Among patients who could predict their CH bout, 32% said they were more than 95% certain of the coming bout.

Patients who could predict their bout onset had experienced more overall bouts (P =.046), more symptoms (6.7 vs 3.8 symptoms), and experienced shadow attacks (odds ratio [OR], 3.06; 1.19-7.88; P =.020).

During remission periods, 58% of patients still experienced symptoms, and 19% of those patients could distinguish between bout-related pre-cluster symptoms and remission symptoms. The patients who could distinguish them indicated remission symptoms were milder, shorter, or fewer than pre-cluster symptoms, or they were associated with the wrong time-of-year.

This study likely included some amount of recall bias because patients were asked to self-report previous symptoms, in some cases spanning more than 1 year.

These data indicated most patients with eCH could predict coming CH bouts days or even weeks prior to their onset. Additional studies are needed to determine whether these predictions may allow for early or preventive interventions to be effectively applied.

Disclosure: Several authors declared affiliations with the pharmaceutical industry. Please refer to the original article for a full list of disclosures.


Pedersen AS, Snoer A, Barloese M, Petersen A, Jensen RH. Prevalence of pre-cluster symptoms in episodic cluster headache: is it possible to predict an upcoming bout? Cephalalgia. Published online January 20, 2021. doi:10.1177/0333102421989255