Research Review By Dr. Joshua Plener©


Download MP3

Date Posted:

October 2020

Study Title:

Sentinel headache as a warning symptom of ischemic stroke


Lebedeva E, Ushenin A, Gurary N et al.

Author's Affiliations:

Department of Emergency Neurology, the Ural State University, Russia; International Headache Centre “Europe-Asia”, Russia.

Publication Information:

The Journal of Headache and Pain 2020; 21: 70.

Background Information:

Sentinel headache is characterized by a sudden, intense, persistent headache, with features different from any usual, previous headache. Sentinel headache occurs in approximately 15-60% of spontaneous subarachnoid hemorrhage patients, preceding the vascular event by days or even weeks (1). Put simply, a sentinel headache warns against an impending disease.

Currently, there is a paucity of literature examining sentinel headache in ischemic stroke patients. The aim of this study was to compare patients with and without sentinel headache in order to evaluate triggers, determine potential risk factors and provide clinical recommendations in order to assist in identifying patients with sentinel headache. The focus of the study is on headaches occurring within 7 days of an ischemic stroke. These are the patients most relevant to chiropractors, as it is commonly thought that many patients who suffer a stroke in close temporal relation to a neck manipulation, may have presented for care with the stroke already in progress…one of the signs of this scenario could be a sentinel headache!

Pertinent Results:

Out of 2,995 ischemic stroke patients who were prospectively examined, 550 patients were included in the study (306 men and 244 women, average age 63). 192 control patients were included and consisted of patients admitted to the ER for a range of diagnoses such as lumbar spine osteochondrosis, pancreatitis and gastrointestinal ulcer.

The results indicate that factors associated with stroke in general included male sex, smoking, intake of strong alcoholic beverages, arterial hypertension, diabetes mellitus, hyperglycemia, atrial fibrillation, low physical activity, hypercholesterolemia, angina pectoris, myocardial infarction and a family history of stroke.

Headaches within the last year, excluding the week before the ischemic stroke:

Out of the 550 ischemic stroke patients, 13.3% compared to 8.9% in the control group suffered from migraine headaches with aura within the last year, excluding the week preceding the stroke (discussed below). Furthermore, 65.2% of ischemic stroke patients compared to 70.3% in the control group suffered from tension-type headaches within the last year, excluding the week preceding their stroke (remember, tension headaches are very common). The only significant finding in this study for headaches within the last year was females with migraines in the last year had an odds ratio of 2.0 for the development of a stroke.

Headache within the week before the ischemic stroke:

94 patients out of 550 (17.1%) experienced a headache during the 7 days preceding their stroke. In the control group, only 12 patients (6.2%) had a headache 7 days before their ER admission. Therefore, a headache within 1 week before an ischemic stroke resulted in an odds ratio of 3.9. Out of the 94 ischemic stroke patients with a headache, 13 (~14%) had headaches that were not different than any previous headaches, 54 (57%) had headaches that were altered in their characteristics, such as being more severe, longer lasting and/or more frequent, and 27 (29%) experienced a new type of headache.

Sentinel Headache:

A sentinel headache was represented as a new type of headache or a previous headache with altered characteristics. 81 stroke patients (14.7%) fit this category, which consisted of 52 females and 29 males with an average age of 62.4 years. Out of the 81 sentinel headaches, 46 (~57%) had a tension type-like headache, 33 (~41%) had a migraine-like headache and two (~2%) had a thunderclap headache. 28.4% of headaches arose within 60 min to 10 hours before their stroke, 32.1% arose 1-2 days before their stroke and the remaining (~40%!) arose during the 3-7 days before their stroke. The majority of patients had their headache disappear within 24 hours of stroke onset and no cases were observed where the headache ceased before the onset of their stroke.

Possible Triggers and Risk Factors:

The only trigger significantly associated with stroke patients compared to the control group was an attack of arrhythmia, which had an odds ratio of 2.3 in this dataset.

Possible risk factors that were determined in the study included the female sex (odds ratio of 2.4), atrial fibrillation (odds ratio of 1.9), angina pectoris (odds ratio of 2.0), and a previous history of headache (odds ratio of 4.0). When separated, a previous history of migraine had an odds ratio of 2.5 and a previous history of tension type headache had an odds ratio of 2.3.

Clinical Application & Conclusions:

The main findings of this study include:
  1. The prevalence of any headache during the seven days before a stroke was significantly higher in stroke patients compared to controls.
  2. For patients who developed a new type of headache, a migraine-like headache was more common than a tension-like headache or thunderclap headache.
  3. The only potential trigger factor identified for the development of a sentinel headache prior to an ischemic stroke was an arrhythmia attack.
  4. Significant factors associated with the development of a sentinel headache included the female sex, atrial fibrillation, angina pectoris, previous history of migraine and tension type headache.
Possible Mechanism of a Sentinel Headache:

Overall, the mechanism behind a sentinel headache still requires further research. In this study, the findings of an increased prevalence of arrhythmia and atrial fibrillation seen in sentinel headache patients compared to patients who did not develop a sentinel headache provide support that an embolus may be the underlying mechanism. The proposed mechanism is that the emboli liberates cytokines, which dilate the cerebral artery in addition to being pro-inflammatory. This potentially affects perivascular nerve endings leading to nociception. No literature currently supports this possible mechanism, but other authors have postulated a similar theory (2).

In a clinical environment, patients with an unusual and severe headache require a thorough investigation. This starts with an in-depth history in order to understand their headache characteristics and how this headache may be different from previous headaches experienced (be sure to ask!). Examples of important characteristics to understand are pain intensity, duration, character, aggravating factors, accompanying symptoms and the effectiveness of any drugs they have taken. Furthermore, in older adults, such as over 50 years of age, these findings are considered red flags and depending on the headache characteristics, imaging should be performed.

Reviewers note: This study provides an interesting analysis and was an important step in determining possible trigger and risk factors for sentinel headaches. However, only 94 out of 550 stroke patients developed a headache within 7 days prior to their ischemic stroke. Therefore, clinicians need to be attentive (as always) to other signs and symptoms of stroke, as only relying on one finding, such as the presence or absence of a sentinel headache, can also have serious negative consequences for the patient.

Study Methods:

This was a prospective study conducted from September 2012 to October 2015. The interviews were performed at “New Hospital” (Yekaterinburg) in Russia.

Inclusion Criteria:
  • Ischemic stroke with the presence of a new infarct on MRI with diffusion-weighted imaging or on CT
  • No history of previous stroke or transient ischemic attack
  • No serious pathology of the nervous system (i.e. brain tumor, traumatic brain injury, multiple sclerosis etc.)
  • No serious somatic pathology
  • No impaired consciousness
  • No memory or speech impediments that would affect the collection of study information
  • Participants could provide a clear description of their headache and agreed to a five year follow up period to study post stroke persistent headaches and cardiovascular events
Exclusion criteria:
  • Patients who were comatose/stuperose/intubated after admission
  • Patients not consenting to the examination
  • Patients with dementia
  • Cognitive impairment as measured by a score less than 26 on the Mini-Mental State Examination
  • Patients with aphasia
  • Patients not able to identify headache onset and stroke symptoms
The control group consisted of individuals who were aged match to study participants. These patients were admitted to the ER without acute neurological deficits or serious neurological or somatic disorders.

Headache Definitions:

Headaches were defined according to the International Classification of Headache Disorders (ICHD-3). Headaches were recorded for the experimental and control group for the last year and within the week (7 days) preceding their stroke. Headaches were categorized into three groups:
  1. headache without a change of characteristics compared to previous headaches,
  2. headache with altered characteristics compared to previous headaches, or
  3. and a new type of headache that arose within the last week before a stroke.
Sentinel headache was defined as a new type of headache or previous kind of headache with altered characteristics (i.e. change in intensity, frequency, drug effect) which arose within seven days before their stroke.

Headache Evaluation:

The headache evaluation followed the “gold standard” for headache diagnosis (according to the authors), which is a semi-structured face-to-face interview by a neurologist. Two neurologists performed the interviews after patients received the necessary clinical examination and the MRI or CT were completed. The following information was collected:
  • Past history of headache such as migraine, tension type headache, cluster headache etc.
  • Description of the headache onset and exact day the headache developed.
  • Headache characteristics including location, side, frequency, duration of the headache, character of the headache (i.e. pressing, pulsating), severity of the headache (i.e. mild, moderate or severe), aggravating factors (i.e. physical activity), presence or absence of aura and the effect of drugs for pain relief. If headaches began the same day as the stroke, patients were questioned about when the headache first developed compared to when the patients first noticed signs of a stroke.
Results of specialist consultations, imaging and any other investigations related to the stroke, as well as previous and current treatment were also recorded.

Risk factors for a sentinel headache were identified from past literature and evaluated in the current study:
  • Acute alcohol abuse defined as alcohol intake of more than 40g within 24 hours preceding a stroke or greater than 150g in the week leading up to the stroke
  • Heavy physical exertion such as lifting heavy weights, unusual physical exercise
  • Psychological stress
  • Overwork
  • Lack of sleep
  • Hypertensive crisis defined as greater than 180 mm Hg systolic and 120 mmHg diastolic
  • Clinical infection
  • Arrhythmia attack where the patient felt and described severe unusual heart palpitations
  • Overheating
  • Female sex
  • Smoking (current smoker)
  • Consumption of light alcoholic beverages and strong alcoholic beverages
  • Increased body mass index (BMI greater than 25)
  • Low physical activity defined as less the 30 minutes of physical activity one time per week
  • Family history of stroke
  • Cardiovascular conditions (such as arterial hypertension, diabetes mellitus and atrial fibrillation)
Statistical analysis: Percentages and odds ratios were used to calculate crude prevalence of headache disorders, means and categorical variables were used to summarize continuous variables and chi-squared tests were used to compare distributions of categorical variables between groups. The difference between the prevalence of triggers and probable risk factors between patients with and without a sentinel headache were statistically examined by an unpaired t-test and chi-square test. A p-value of < 0.05 was considered statistically significant. No sample size was originally calculated, as limited research exists regarding sentinel headache in ischemic stroke.

Study Strengths / Weaknesses:

  • This study enrolled a large number of participants.
  • This is the first study to convincingly demonstrate the existence of sentinel headache in ischemic stroke patients.
  • This study provides important clinical recommendations that should be implemented into practice.
  • A matched control group was utilized to help reduce bias.
  • A 7-day window was utilized, however if a longer pre-stroke window was used, more sentinel headaches may have been recorded, as a sentinel headache can arise prior to 7 days before an ischemic stroke. However, the 7 day cut-off was chosen in an attempt to limit recall bias, so it is understandable why they did this.
  • The study was controlled but not blinded.
  • A significant number of patients were excluded due to reasons such as language impairment, cognitive impairment etc. This limits the generalizability of the results, as these patients may have differed in meaningful ways from the patients ultimately enrolled in the study.

Additional References:

  1. Pereira JL, de Albuquerque LA, Dellaretti M et al. Importance of recognizing sentinel headache. Surg Neurol Int 2012; 3: 162.
  2. Gorelick PB, Hier DB, Caplan LR, et al. Headache in acute cerebrovascular disease. Neurology 1986; 36(11): 1445–1450

Contact Tech Support  Contact Dr. Shawn Thistle
RRS Education on Facebook Dr. Shawn Thistle on Twitter Dr. Shawn Thistle on LinkedIn Find RRS Education on Instagram RRS Education (Research Review Service)