Evidence-Based Science

We compare how many migraines our app users get when they first start the app to how many they get in the third month (precisely, weeks 8-12). Based on tracking data in the app, 85% of users reduce or eliminate migraines.

We believe this is an enormous effect and should be touted at every opportunity! That’s a big deal. Read on for more info.

Knowing that 85% of users reduce or eliminate migraines…  we got curious on a second question: How much of this user improvement is correlated to using the app? 

Let’s consider newly diagnosed users of our app. Users that start the app when newly diagnosed are likely starting a medication at the same time. We would expect a high percentage of users would show a reduction in migraine frequency. When we look at users that are newly diagnosed at the time of starting the app, over 90% show a reduction in frequency. 

There is no way to know from our tracking data what part of their improvement is new treatments versus the app making an impact.

But! Let’s consider the opposite group of users. Here’s where the statistics get really interesting. For people who have had migraines more than 3 years before starting and who did not start a new medication when starting the app –  72% show improvement in the 8-12 week timeframe. For this group, Migraine Insight is their new treatment, essentially. And, it’s showing up in the app’s data as more effective than most medications on the market. 

In summary, when we look at all users – those newly diagnosed or not, 85% of users that use the app for 12 weeks show a reduction or elimination in migraines. To be incredibly clear again, the data we use to evaluate effectiveness is not clinical trial data, but data from the app itself. This is “task-contributed data” or TCD. TCD is not a substitute for clinical trial data. They are not the same.

We believe TCD is very valuable in proper context. Please read further to better understand why we are excited about TCD data.

Clinical Data: We use clinical trial data to inform the design of our system. Clinical studies, their results, and their data help us make good decisions on how best to create a robust, useful migraine app experience. We strive to make an experience that helps users better understand their own life situation and work better with their care team. The design of the app is strongly informed by insights from publicly available clinical trial data.

Task Contributed Data (TCD): The data in our app – collected by users as they track their own lives – is “task-contributed data” or TCD. (Examples: migraines, symptoms, medications, trigger events). This data is primarily entered as personal-use data that people generate when seeking solutions to their own migraine. This data has an important secondary use as pooled data for population insights. TCD is exciting data that’s likely to be highly accurate since it’s contributed in real time. It’s collected in a way that highly motivates people to enter data accurately as they try and better understand their own condition. This accuracy is hard to quantify, but would logically exceed the accuracy of self-reported time-delayed data collected in clinical trials, when used in that setting.

For example, in clinical trials with nutrition surveys – people just don’t remember how many servings of broccoli they had two months ago with a lot of accuracy. But, tracking that in an app to help with a medical condition, where they are highly motivated to get that right and the app makes it easy for them to enter the data in real time and refer to it later? That’s going to be more accurate. 

To be clear, TCD is not clinical research data, nor is it medical data.
 Although TCD is not clinical, research or medical data, it could be incredibly valuable. For example, we use it in the app as a way to find patterns across different sets of users. Also, TCD data is used to evaluate app effectiveness. Perhaps someday it could be used in a setting for people who are looking to design better clinical trials. An  anonymous pool of TCD data with rich user behavior that was tracked in the past – this could help shape better clinical trials of the future.

No. To be clear … unlike our competitors and most apps that track personal information, we do not collect data for the express purpose of selling to a third party, and we do not sell data to third parties. Ever. 

Despite not selling data, we use GDPR best practices when collecting, storing and using all data, including TCD (
“task-contributed data”).

We use TCD collected from users to help other users in several ways – all within the app, and all de-identified/anonymous. 
This is done by the use of anonymous pooled data, and only if the the user has opted-in. Opting-in means the user has allowed that their data can be de-identified and put into an anonymous pool of data. This pooled data helps Migraine Insight help other people through features in our app like population insights and generalized learnings.

When evaluating migraine outcomes, we address frequency of attacks, duration, and severity.

Identifying triggers is recommended in standard medical practice. For example, the AAN’s Headache Quality Measurement Standing Work Group recommends the practice (e.g. p14).

The ultimate outcome when treating headache and migraine is to reduce the frequency of headache. Working with the patient to identify potential migraine triggers and counseling them on lifestyle factors they can change can help reduce the severity and number of migraines. Reduction of headaches is associated with improved health-related quality of life.”

By identifying triggers, people can choose, with the oversight of their medical team, to avoid those triggers.

Many studies show that reducing exposure to triggers improves outcomes for people living with migraines. Examples:

Study: Perceived migraine triggers: do dietary factors play a role? [1]
Study: Migraine and triggers: post hoc ergo propter hoc? [2]

[1] Camboim Rockett, F., Castro, K., Rossoni de Oliveira, V., da Silveira Perla, A., Fagundes Chaves, M. L., & Schweigert Perry, I. D. (2012). Perceived migraine triggers: do dietary factors play a role?. Nutricion hospitalaria, 27(2), 483–489. https://doi.org/10.1590/S0212-16112012000200020 – “In a cross-sectional study at an outpatient headache clinic, 98% of patients identified dietary triggers, but these included fasting, in addition to foods and alcohol.” (link)

[2] Hoffmann, J., & Recober, A. (2013). Migraine and triggers: post hoc ergo propter hoc?. Current pain and headache reports, 17(10), 370. https://doi.org/10.1007/s11916-013-0370-7 (link)

De-identification is a process by which personal identifiers are removed from digital information. This allows the use of data in capacities where no individual can be identified. De-identified data is information that does not identify an individual because it lacks name, email address, phone numbers, and any other personally identifying information. De-identified data is also held to a standard in which there is no reasonable basis to believe that the information included in the dataset can be used to identify an individual. For example, if a particular type of dataset revealed personal information, and had a result where there was only one person in a specific zip code, and had the capability to look at user’s personal information by zip code – this example scenario would not qualify as ‘de-identified’ information in that system. 

We take data privacy seriously. 

This app does not contain doctor-patient communications, and therefore does not fall under HIPPA (USA). Despite not having a legal requirement to do so, we look to HIPPA for guidance when going beyond the legal minimum requirements. One example is the HHS’s “Guidance Regarding Methods for De-identification of Protected Health Information in Accordance with the Health Insurance Portability and Accountability Act (HIPAA) Privacy Rule.” 

More information: https://www.hhs.gov/hipaa/for-professionals/privacy/special-topics/de-identification/index.html 

In a study of food allergies associated with migraine using 56 migraineurs and their matched controls, the study found that 100% of patients but only 26% of healthy controls had elevated titers of IgG antibodies against different foods. [3]


[3] Arroyave Hernández, C. M., Echavarría Pinto, M., & Hernández Montiel, H. L. (2007). Food allergy mediated by IgG antibodies associated with migraine in adults. Revista alergia Mexico (Tecamachalco, Puebla, Mexico : 1993), 54(5), 162–168.

Migraine remains second among the world’s causes of disability, and first among young women. [4]

Migraines are more than just a headache. People who suffer from migraines have higher suicide rates. For chronic migraine with aura patients, the attempted suicide rate is alarmingly high at 13.9% (vs. 2.8% in migraine-free controls). Suicide ideation in that group is 47.2%, according to one study. [5]


[4] Steiner, T.J., Stovner, L.J., Jensen, R. et al. Migraine remains second among the world’s causes of disability, and first among young women: findings from GBD2019. J Headache Pain 21, 137 (2020). https://doi.org/10.1186/s10194-020-01208-0 (link)

[5] Lin, Y. K., Liang, C. S., Lee, J. T., Lee, M. S., Chu, H. T., Tsai, C. L., Lin, G. Y., Ho, T. H., & Yang, F. C. (2019). Association of Suicide Risk With Headache Frequency Among Migraine Patients With and Without Aura. Frontiers in neurology, 10, 228. https://doi.org/10.3389/fneur.2019.00228 (link)

Yes. We do offer reduced or free subscriptions to those in need. There is no minimum financial requirement. 

If you would like a premium subscription but can’t afford one due to financial reasons, please make contact at hello@migraineinsight.com.

We are a small, dedicated team motivated by helping others. We want you to better understand your migraines so you and your care team can make the best decisions possible. A financial burden should be in the way of you getting results in our app.