Covid Headache

An observational study was conducted to estimate headache frequency in patients with confirmed COVID-19. To characterize the headache phenotype attributed to COVID-19, and compare patients depending on their need for hospitalization and sex, Between March 8th (first case), and April 11th, 2020, we screened all eligible patients in a reference population that included 261,431 people. A survey was completed by a physician to assess demographic and clinical information and the phenotype for the headache. 

The study found that 2194 of the patients at high risk for COVID-19 were diagnosed during the period. 514/2194 patients reported experiencing headaches (23.4%; 95% CI 221.7-25.3%), which included 383/1614 (23.7%) and 131/580 (22.6%) outpatients. The headache phenotypes were studied in detail in 458 patients (mean, 51 years old; 72% women; previous history of headaches, 49%). 

The most common symptom of COVID-19 was headache. The median time it took to get headaches started was 24 hours. In 13% of cases, the headaches lasted for at least 1 month. Bilateral pain was prevalent (80%), predominantly frontal (71%), and of extreme intensity (75%). 98% of patients had systemic symptoms. Patients with and without hospitalization had similar headache frequency and phenotype. Female patients were more severe than male patients.

Trial registration: This study was funded by the Institute of Health Carlos III, code 07.04.467804.74011, and Regional Health Administration Gerencia Regional de Salud Castilla y Leon, (GRS 2289/A/2020).


Headache is a common non-respiratory symptom of coronavirus infection 2019 (COVID-19).1,2. The reported prevalence is quite different, with 14-70% of COVID-19 victims describing it depending on the study1,2,3,4,5. It is the most common neurological symptom and it affects 14.8% of hospitalized patients. Patients between 18 and 49 years of age experience it. It is not clear what the true incidence and phenotype are of headaches with COVID-19, since most studies involve hospitalized patients1,2,3,4,5.6,6.

In most cases, headaches manifested within 72 hours of diagnosis in COVID-19 patients7,8,9. This phenotype is characterized by a bilateral headache, with frontal predominance, a oppressive quality, and moderate-to severe intensity8. The majority of patients meet the International Classification of Headache Disorders criteria for “Acute headache attributed systemic viral infection”.

However, 54% of patients also met the phenotypic ICHD criteria to treat tension-type headaches (TTH), and 25% for migraine7.10. Although the phenotype might be a chameleon for a primary headache (TTH), all patients in hospital-based series had at least one red flag. In COVID-19 patients, it is not known if headaches can be mistakenly diagnosed as primary headache disorders. The presence of headaches in hospitalized patients predicted lower mortality2 and lower admission to intensive care unit11,12. 

Patients with headaches also reported a high level of disability and the need for frequent acute treatment7. This study aims to determine the prevalence of headaches in COVID-19 patients and to identify the clinical phenotype. We compared the frequency and clinical phenotype of patients who were admitted to hospital with those who were managed on an outpatient basis. We also compared both genders.


Study design and setting

This study is observational analytic with a cross-sectional design. The Strengthening In the Reporting of Observational studies in Epidemiology statement13 was followed. Patients with COVID-19 confirmed and headaches at any stage of the COVID-19 course were included in the study population. The study was conducted in Valladolid East Health Area which includes a tertiary academic medical center (Hospital Clinico Universitario de Valladolid), and 22 primary care centers. The total reference population was 261,431 people. The primary care center included in the study had an average of 1896 to 20,930 patients. This includes both urban and rural areas.

Study the population

The eligibility criteria was based on prior studies2,7,9; patients were included if: (1) they had a confirmed diagnosis of COVID-19, either by real-time reverse-transcription polymerase chain reaction (RT-PCR) assay from a respiratory tract sample and/or by the presence of anti-severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) antibodies; (2) new-onset headache presented during the course of COVID-19 that was not better accounted for by another secondary cause, according to the International Classification of Headache Disorders, 3rd edition (ICHD-3)10 criteria; and (3) they were older than 18 years. Patients who were unable to take part due to severe or unstable medical conditions, (2) a history of cognitive impairment, dementia, or speech or language problems, were exempted from the study. (4) They died during COVID-19 or declined to participate. We also excluded patients that we couldn’t reach after at most three attempts with all contact options.

Screening patients

Patients with COVID-19 symptoms were instructed to contact a 24-hour hotline for COVID-19 pandemic. Patients with symptoms that suggested COVID-19 were contacted by phone or in person every day. The COVID-19 symptoms checklist included headaches. Additional Fig. 1). Supplementary materials provide information on the criteria used to admit patients to hospitals according both to local and national protocols14. 

We used the primary care COVID-19 team database to identify patients. All patients who were examined between March 8, 2020 (first COVID-19 confirmed case in the healthcare sector) and April 11, 2020 were screened. Electronic health records were used to determine if a patient had headache. A member of the study team was comprised of 16 primary care physicians and two neurologists. All covid-19 patients were also interviewed about their headaches by the physician. Patients could be seen in person or were assessed over the telephone. To ensure late-onset headaches were detected, all patients were examined at least one month after onset of symptoms. The questionnaire asked patients about their headaches and they were then invited to fill out the rest of the questionnaire.

Study goals

These two main goals were to estimate the incidence of headaches over the course COVID-19 in general population, and characterize the clinical phenotypes of headaches in COVID-19 patients.

Secondary objectives were to evaluate the frequency of red flags in headache patients,  compare the clinical manifestations of headaches between patients who required hospitalization and those who didn’t;( evaluate the clinical phenotype based on the gender of the patients.


Before the study began, all researchers received training. The study questionnaire was standardised and based upon previous studies7,15. The protocol was followed and data accuracy and completeness were checked by a coordinator.

Demographic data included the following: age, sex and country of origin. The research team was accompanied by a physician who reviewed the family history and medical history for migraine. Prior medical history was evaluated for the presence of hypertension, diabetes, a smoking habit, cardiac disorders, pulmonary disorders, cancer, immunosuppression, and chronic neurologic disorders (full definition available in supplementary material 1). 

The physician also reviewed the prior history of headaches, including the exact diagnosis, source, frequency, and degree of similarity between the headaches experienced during COVID-19, and their usual headaches, on a scale of 0-100% (0: totally different, 100%: the same). A physician asked participants about their past headaches in relation to any prior infections.

The clinical presentation of COVID-19 was based on radiological abnormalities. We determined that oxygen therapy was required. We analyzed whether COVID-19 was diagnosed using RT-PCR or serological tests. We evaluated the first symptom as well as the occurrences of other systemic symptoms. (full list in the supplementary materials). The time between the onset of the headache and the onset of the symptoms was evaluated. 

The variables that affected headache phenotype were: laterality (bilateral, unilateral, or midline headache), topography (quality of pain, absence of associated symptoms like nausea, vomiting, phonophobia and osmophobia), cranial autonomic symptoms such as conjunctival injections, lacrimation and nasal congestion, rhinorhea, migraines, headache-related disability (rated on an 0-100% numeric rating system, 0: no disability, 10: worst case), need for acute medication and factors that made the headaches neighting, including bending, coughing, nose, moving around the eyes, or sneight, or sneight, or sneight, or sneight, or sneight, walking, snea headache, or sneensneea headache, sneacusnearithma scale, the a numeric numeric rating, We also screened for headache-related red flags using a standardized screening method that was previously used in COVID-19 patients suffering from headache9 (supplementary appendix).

Statistic analysis

As percentages and frequencies, we present ordinal and qualitative variables. Variables that have missing data are represented as the percentage of patients who exhibit the variable compared to the total number of patients with valid information. Continuous variables are defined as means and standard deviations, medians and interquartile intervals (IQR), depending on their distribution. 

The Kolmogorov–Smirnov test was used to determine the normality of the distribution. The incidence of headache was determined as the number of cases per 100 patients (%) using the 95% confidence interval. (CI). Data were analysed per intention to treat (ITT), by dividing the number COVID-19 patients presenting with headaches by the total COVID-19 case count (i.e. The entire area was screened. 

Fisher’s exact tests were used to compare qualitative categorical variables. If the distribution was uniform and homogeneous, Student’s T-tests were used to assess quantitative variables. In all other cases, the Mann-Whitney U-test was used. The sample size was not known in advance, and the analysis was done using all available data. Complete case analysis was used to manage missing data. In the event of multiple comparisons, the significance level was set at P = 0.05 and corrected using the Benjamini–Hochberg procedure17. SPSS statistical software package (26.0) was used for statistical analysis on Mac (IBM Corp. Armonk NY).

Ethical aspects

The International Conference on Harmonization Guidelines for Good Clinical Practice was used to guide the study. The Valladolid East ethics board approved both the hospital and primary care series (PI20-1738) All participants signed informed consent.

Participation requires consent and approval from the ethics committee

Both the hospital series (PI20-1738) as well as the primary care series, PI20-1881 were approved by the local ethics review board. All participants gave informed consent verbally or in writing.

By Cary Grant

Leave a Reply

Your email address will not be published.

You May Also Like