Impact of COVID-19 Pandemic on Migraine in Migraine Buddy Users

Background: The nature of COVID-19 pandemic measures has altered the clinical management of migraine, and has also created barriers to evaluate the impact of such measures of migraine patients. Using the Migraine Buddy smartphone application, we assessed the impact of the COVID-19 pandemic on migraine in users residing in the United States. Methods: Migraine Buddy is a smartphone application by individuals to record their migraine headache episodes, characteristics, and coping mechanisms. For this study, anonymized self-reported data from 163,176 adult Migraine Buddy users in the United States between January 2020 and May 2020, were analyzed for migraines associated with stress. A stress-related migraine is dened as one in which stress or anxiety was reported as a trigger or symptom. A questionnaire on the impact of COVID-19 on migraine and its management was also completed by 1,322 users in the app between April 2020 and May 2020. Results: 88% of the Migraine Buddy database extract and 83% of the respondents are female, with a mean age of 36.2 years. The proportion of stress-related migraine attacks peaked at 50% on March 17, although the number of migraine attacks decreased. This followed the declaration of the COVID-19 national emergency on March 13 and a spike in the number of COVID-19 cases in the United States. Questionnaire respondents felt that the following added more stress: social isolation (22.6%), information overdose (21.2%), access to essentials (food, medication, etc.) (18.7%), and nancial concerns (17.8%). To help migraine respondents suggested stress and diet coaching programs and resources (medical articles, etc.) (34.0%), having the option for home delivery of medication (30.6%) and tele-consulting (25.5%). Conclusion: Here, we report the change in the proportion of self-reported stress-related migraine in relation to evolution of the COVID-19 pandemic, as well as its impact of migraine management. Our data will help increase the understanding of patients' needs and help with planning and execution of mitigating strategies.


Introduction
The World Health Organization declared COVID-19 a global pandemic on March 11, 2020. At the time of writing (December 30), there have been 19,432,125 con rmed cases and 337,419 deaths in the United States [1]. New York City went into lockdown on March 22, requiring residents to stay home. As of May 30, New York has been the virus' epicenter in the United States with 23,780 deaths and 368,234 cases [2].
Under such conditions, it is di cult to assess the impact of COVID-19 for migraine patients without faceto-face consultations.
With the recent developments in information technology, real-world big data studies have attracted increasing attention in the eld of medicine. Obtaining clinical insights is an important part of the wider scope of real-world medical data, and its standardized use is critical for the generation of high-quality real-world evidence to improve the core functioning and competitiveness of clinical care data as well as provide high-quality medical services for patients.
Migraine Buddy is the most widely used smartphone application by individuals worldwide to record their patterns of migraine headaches, characteristics (triggers, pain intensity, symptoms) and coping mechanisms (medication intake, relief methods) [3]. By tracking the migraine headaches over time, users and their headache specialists are able to better understand and manage the condition. Through the app, Migraine Buddy users can also provide real-time insights for research purposes by responding to surveys and working with the Healint data analytics platform to extract key aggregated ndings. Using the Migraine Buddy app, this study aimed to assess impact of the COVID-19 pandemic on migraine in Migraine Buddy users residing in the United States.

Methods
The study is a non-interventional, retrospective analysis of data that is self-reported via the Migraine Buddy smartphone application. Participants provided authorization for their data to be used for research, and deidenti ed data were used and analyzed for this study. Participants have the option to refuse to this condition and are able to use the app without any restriction if they choose to do so. As such, study subjects were not placed at risk by being included in the study sample, and the study was approved by the Institutional Ethics Committee at the Saitama Neuropsychiatric Institute (approval number: SNI 20 − 006).
A retrospective analysis was conducted using data captured through the free smartphone application Migraine Buddy. When self-reporting a migraine headache episodes, users are prompted to provide the following: attack duration, attack type, pain intensity, location of pain onset, medication taken and its effectiveness, relief methods, symptoms, aura and prodrome, affected activities, location when the attack occurred, along with weather and prior sleep patterns collected by the app.. For this study, anonymized self-reported data from 163,176 adult Migraine Buddy users in the United States over the 5-month period from January 1, 2020 through May 31, 2020 were extracted from the Migraine Buddy database. The variables examined included the following: demographic characteristics (age, gender), migraine frequency and triggers. Each migraine record represents a single migraine attack.
A stress-related migraine is de ned as one in which stress or anxiety were self-reported as a trigger or symptom. The stress-related migraine group includes individuals who reported at least 1 stress-related migraine in each calendar month of the study period. The change of Dow Jones Industrial Average (DJIA) and the number of new COVID-19 cases in the United States were used as reference indexes [1].
In addition, a questionnaire survey was sent to Migraine Buddy users and completed in the app between April 2020 and May 2020. The survey was designed by authors to obtain responses of the impact of COVID-19 on migraine and migraine management during this period.

Study population
The study sample included 1,116,605 migraine records from 163,176 users. 83% of users were female, with a mean age of 36.2 years. A summary of demographics and migraine characteristics of the study population is shown in Table 1, 2. The stress-related migraine group had a higher proportion of women than the migraine group that did not reported stress-related migraine (p < 0.0001). Although there were no difference in pain intensity and monthly headache days between the two groups, the usage rate of acute and preventive treatment was higher in stress-related migraine group than non-stress-related migraine group (p < 0.0001).  Recorded migraine attacks Questionnaire survey A total of 1,322 respondents completed the questionnaire survey. Of the respondents, 83% were female; 12% were aged 35 to 44 years, 10% were aged 25 to 34 years, and 10% were aged 45 to 54 years ( Fig. 2 −  1).
57.5% of respondents reported an increase in migraine days, with 26.0% describing a major increase and 31.5% as slight increase (Fig. <link rid=" g2">2</link>-2). The migraine severity was reported to worsen compared to before the pandemic by 41.8% of respondents, with 12.8% describing this as signi cant and 29.0% as claiming a slight worsening (Fig. 2-3). Sources of stress were reported as follows: social isolation (22.6%), information overdose (21.2%), access to essentials (food, medication etc.) (18.7%) and nancial concerns (17.8%) (Fig. 2-4). In terms of migraine management, more than half of responders (58.1%) experienced an interruption of medication, health/emergency services, and doctor consultations, with a quarter of responders (26.3%) experiencing di culty in maintaining migraine-friendly diet (Fig. 2-5). According to the respondents, the most helpful way in managing migraine was having a coaching program (stress management, diet etc.) and resources (medical articles etc.) (34.0%), followed by medication home delivery (30.6%) and tele-consulting with a health specialist (25.5%) (Fig. 2-6).

Discussion
In the present study, we reported a clear increase in the proportion of stress-related migraine as the COVID-19 crisis deteriorated. We chose this metric to illustrate the changing patterns in migraine, as the number of recorded migraine attacks is affected by the number of active users in Migraine Buddy. In our questionnaire survey, 57.5% of respondents reported an increase in migraine frequency and 41.8% indicated worse severity during pandemic period. Hence, we argue that the COVID-19 pandemic has not only increased the proportion of stress-related migraine but also the migraine frequency and severity. Similar to our ndings, an online survey conducted in Kuwait demonstrated that the majority of respondents had reported increase in migraine frequency and severity during pandemic period [4].
Stress is known as the most common trigger of migraine attacks [5]. COVID-19 pandemic, as a global health crisis, is perceived as a major stressful event. In a recent study from Spain, 41% of participants reported feeling moderate to severe stress and depressive symptoms during the pandemic, and 25% experiecned mild to severe levels of anxiety [6]. Young women, who consisted the majority of respondents in our study, felt the strongest negative impact [6]. COVID Stress Scales [7], which are scales measuring COVID-related stress and anxiety symptoms, may offer promise as tools for better understanding patients with COVID stress-related migraine.
Next, information overdose was reported as the second most common source of stress during pandemic in our questionnaire survey, with another study demonstrating that attention paid to COVID-19 media coverage as a risk factor for psychological distress in migraineurs [8]. This suggests that migraineurs could avoid paying too much attention on the media coverage of the pandemic.
Most respondents experienced signi cant life changes during a short period of time. In the light of questionnaire survey, respondents had di culty maintaining migraine-friendly diets, scheduling doctor consultations, and accessing health/emergency services. A recent study reported that disturbance of eating habits was one of the factors affecting increase migraine frequency and severity [4]. A recent survey of 155 countries conducted by the World Health Organization found that nearly half of the patients with noncommunicable diseases failed to receive their regular medical care and medications since COVID-19 pandemic began [9]. With the various lockdown measures and shifts in healthcare systems towards COVID-19, it is plausible that this has resulted in a shift in patient behaviour towards self-management and may have exacerbated migraine.
From our questionnaire, 13.5% of respondents reported in migraine frequency and 11.2% of respondents reported improvement of migraine severity during pandemic period. This could be explained by the reduction in work-related pressure as workplaces shift towards a working from home arrangement, and echoed by our observation that migraine records tend to decrease over the weekend. This is similar to a recent study from Italy that reported fewer migraine attacks and lesser pain [10]. The authors attributed the ndings to a reduction of trigger as a result of less work. Another recent study of Italian working women revealed that most of the additional housework and childcare associated with the COVID-19 pandemic were managed on not their partners but women [11]. Therefore, less o ce work attributed to COVID-19 pandemic may also be another trigger of migraine for women.
Interestingly, the ratio of stress-related migraine showed inverse correlation with the DJIA index. Rising fears and global economic shutdown due to the COVID-19 pandemic are believed to be main causes of the stock market crash [12]. During the study observation period, we observed that the bottom of the DJIA index overlaps with the peak of the ration stress-related migraine, suggestive that the stock market performance has an impact on stress-related migraine.
The ongoing threat of COVD-19, or other future pandemics, will require strategies for patients to better manage their migraine. The solutions identi ed by the survey respondents include coaching programs, medication home delivery and telemedicine consultation. Thus, it is a positive sign that organisations such as the American Migraine Foundation remain dedicated to providing support, education and comfort to those with migraine [13]. These resources will be bene ts for patients managing migraine during these di cult times. Headache specialists should also consider taking steps to rapidly develop remote teleconsultation services. There is an urgent need to identify clinicians who are trained in telemedicine, expand existing remote consultation technologies with community hospitals and emergency departments and offer medication home delivery.
There are known limitations inherent to these real-world studies that rely on self-reported data from a smartphone app. First, there is potential bias as the study analyzed data reported by users, and no physician diagnosis is available to con rm a migraine diagnosis. However, the majority of respondents have reported the use of common migraine medication within the Migraine Buddy app.
Secondly, selection bias may arise owing to the requirement to have access to a mobile smartphone and comfort level in operating digital apps and, therefore, the population may not be entirely representative of the migraine population. Our respondents were unbalanced for sex and age, as the population to be young females. However, migraine is known to be more common in females, with the highest prevalence in the 20-50 age groups. Also, younger population usually constitutes the majority of smartphone app users.
Thirdly, being an observational study conducted via a smartphone app, the establishment of a direct causal relationship between migraine and COVID-19 pandemic can be di cult. Our goal is to present ndings that can be representative of the real-world impact of COVID-19 on patients with migraine, given the rather large sample size. This will help increase the understanding of patients' needs, and help with the planning and executive of mitigating strategies.
During COVID-19 pandemic, the clinical management of migraine as well as mental health should not be neglected. The stress-related psychosocial impact of the pandemic is evident and complex. Migraine attacks can be associated with a decline in perceived stress [14]. It is plausible to experience an increase in migraine cases once the COVID-19 pandemic situation has stabilized. Long-term follow-up data are necessary and smartphone apps like the Migraine Buddy can play a big role in monitoring changes in migraine managemnet. In future, the analysis of information about the body, collected via wearables, together with self-reported data entered into the migraine diary, can be performed by arti cial intelligence to provide migraine patients with feedback to optimize their migraine management.

Conclusion
Here, we leveraged the Migraine Buddy smartphone app to query and report the change in the proportion of self-reported stress-related migraine as the COVID-19 pandemic evolves, as well as its impact of migraine management. From these responses, we propose that headache specialists continue to develop appropriate strategies to help patients manage migraine and mental health. Our data will help increase the understanding of patients' needs and help in the planning and execution of such mitigating strategies.
Abbreviations DJIA, Dow Jones Industrial Average Declarations Authors' contribution WP had the original idea for the study, and planned the overall design. YK wrote the initial draft. WP and ZH were responsible for data acquisition. YK, WP and TS were responsible for analysis of the result. WP, ZH, FC, TS and YM critically reviewed the initial manuscript. The authors have read, revised and approved the nal manuscript.
The graph shows trend about number of recorded migraine attacks, the ratio of stress-related migraine, Dow Jones Industrial Average and the number of new COVID-19 cases over time. The ratio of stressrelated migraine reached the peak value of 50% on March 17, though the number of recorded migraine attacks decreased. During this time, Dow Jones Industrial Average index decreased sharply.

Figure 2
Results of questionnaire survey