People with chronic migraine have a high frequency of attacks (> 15 days/month), yet little is known about their day-to-day experience of pain. We looked at the relationship between 12 different clinical factors, including mood, sleep, and stress, and day-to-day pain severity scores in adults with chronic migraine. We found that people's experience of pain severity and its relationship to different clinical factors was very individualized, with a lot of variability between people. Description

Alcohol has long been suspected as a common migraine trigger but evidence for this is inconclusive. Accordingly we assessed in an observational, prospective, cohort study whether alcohol intake is associated with the onset of migraine attacks up to 2 days after consumption, in individuals with episodic migraine (EM). N1-Headache™ was used to track headache symptoms and alcohol intake for at least 90 days and the association of alcohol intake (Yes/No and number of units) in the 2 days preceding a migraine attack was determined.

A total of 487 individuals, reporting 5913 migraine attacks and a total of 40,165 diary days, were included in the analysis. The presence of migraine on day-2 and its interaction with alcohol intake on day-2 were not significant and removed from the model. At the population level, alcohol intake 2 days before attacks was associated with a lower probability of migraine attack (OR [95% CI] = 0.75 [0.68, 0.82]; event rate 1006/4679, 21.5%). The effect of alcohol intake the day before attacks was not significant (OR [95% CI] = 1.01 [0.91, 1.11]; event rate 1163/4679, 24.9%) after adjusting for sex, age, and average weekly alcohol intake. Similar results were obtained when analysing for the number of beverages consumed.

We conclude that in this cohort of individuals with EM, who mostly identified themselves as low-dose alcohol consumers, there was no significant effect on the probability of a migraine attack in the 24h following consumption, and a slightly lower likelihood of a migraine attack from 24 to 48h following consumption. Description

When headaches occur on half of days of the month or more, with features consistent with migraine for at least 8 days, for more than 3 months, the current ICHD definition of chronic migraine (CM) is met. Individuals with CM have a low frequency of migraine attacks because the attacks last long periods, which makes analysis on migraine attack onset not useful and impractical. For that reason, we aimed to describe the patterns of peak attack severity from day-to- day within individuals with CM and analyze those patterns in relation to perceived stress. Data on 136 participants contributing 8216 migraine days into the analysis showed that 70.6% of participants reported the same peak migraine severity on more than 50% of their headache days: 13.2% reported mild peak severity, 47.1% reported moderate peak severity and 10.3% reported severe peak severity.

The relationship between peak severity and perceived stress was also modeled and adjusted for sex, age, continuous headache, presence of menstrual bleeding, day of the week, and disability. Model results showed that for every one unit increase in perceived stress, the odds of reporting a higher peak severity were 10% higher (adjusted OR [95% CI] = 1.10 [1.07–1.14]). Moreover, the inclusion of individual effects on the model allowed the investigators to assess that there is a substantial amount of variation between individuals in this relationship. Description

In this paper, our statistical methodology group and collaborators investigated the application of Compositional Data Analysis to generate new visualizations and insights from association rule mining. Association rule mining is a powerful data analytic technique used for extracting information from big transactional databases to identify unique patterns and rules which define relationships and interactions between different items.

The aim of this work was to investigate the association of negative mood emotions to various types of headache/migraine events. Detecting mood associations with headache type is important because it might help develop early interventions that might improve patient condition. In this study, the items were eight ‘negative’ mood factors (stress, anxiety, irritability, lack of happiness, sadness, angriness, boredom, lack of relaxedness) each categorized into low/high at an individual level and the type of headache events (non-headache, headache-only, possible migraine, definite migraine or migraine with aura). Data on 462 N1-Headache™ users (and 65,929 daily questionnaire entries) showed that there were no differences on the association of negative mood factors across the different types of headaches, however non-headache days were more likely found associated with low levels of negative mood emotions. Description

At the population level, studies have shown that the odds of migraine increase during the perimenstrual window (first day of bleeding 2 days). These studies have been neither long nor large enough to assess associations of migraine with menstruation from an individual perspective. In this study we investigated whether there are differences in how menstruation and migraine are associated, not only between women, but also within a woman across her menstrual cycles. Participants were 203 actively menstruating adult women (mean age 35.58 8.70 years) who used N1-Headache™ to track their migraine symptoms and risk factors daily. A total of 53,302 days, 18,520 of which were migraine days, and a total of 2,126 menstrual cycles were analyzed using generalized multilevel models. Results confirmed that, for an average woman’s typical attack, the perimenstrual window was associated with a 34% increase in odds of a migraine day (OR=1.34, 95% CI: 1.23-1.45, p < .0001). Interestingly, the relationship between menses and migraine varied more within-person across cycles than among women. These data highlight that, when treating women for short-term prevention of menstrual migraine, practitioners should consider both the strength of the “average” association between menses and migraine as well as within-person consistency. Description

In this study we aimed to describe individual and episode-level patterns of perceived stress across stages of the migraine cycle. Perceived stress was assessed daily on a 0-10 scale with a single question. Days were categorized into phases of the migraine cycle: Ppre = pre-migraine headache (the 2 days prior to the first day with migraine headache), P0 = migraine headache days, Ppost = post-migraine headache (the 2 days following the last migraine day with migraine headache), and Pi = interictal days (all other days).

Data from 351 participants and 2115 episodes were included in this analysis. Cluster analysis revealed 3 common patterns of perceived stress variation across the migraine cycle. For cluster 1 (16.7% of episodes), the “let down” pattern, perceived stress in the interictal phase (Pi) falls in the pre-headache phase (Ppre) and then decreases more in the migraine phase (P0) relative to Pi. For cluster 2 (59.2%), the “flat” pattern, perceived stress is relatively unchanging throughout the migraine cycle. For cluster 3 (24%), the “stress as a trigger/symptom” pattern, perceived stress in Ppre increases relative to Pi, and increases further in P0 relative to Pi. Twelve participants (13.4%) had more than 50% of their episodes fall into the “let-down” pattern, 216 participants (61.5%) had more than 50% of their episodes fall into the ”flat” pattern, and 25 participants (7.1%) had more than 50% of their episodes fall into the “stress as a trigger/symptom” pattern. In conclusion, on an aggregate level, perceived stress peaks during the pain phase of the migraine cycle. However, on an individual and episode level, there are 3 dominant patterns of perceived stress variation across the migraine cycle. Elucidating how patterns of perceived stress vary across the migraine cycle may contribute insights into disease biology, triggers and protective factors, and provide a framework for targeting individualized treatment plans. Description

Curelator Inc. continues investing in Research and Development to incorporate evidence-based, validated methods in the N1‑Headache™ platform, and support the migraine community. In this paper, our statistical methodology group investigated whether a recently developed method (Compositional Data analysis) could be successfully applied to individual data (N-of-1) for the self-prediction of migraine attacks. In general, prediction of migraine attacks is thought to be difficult because most people with migraine present great (between and within-subject) variability in non-headache symptoms preceding the attack onset, frequency, duration and periodicity of attacks. Using data from 1,159 individuals, with a total of 28,441 migraine days and 136,780 non-migraine days, we concluded that Compositional Data analysis is a promising, valid methodology to aid individuals in self-predicting the onset of their migraine attacks. Description

Self-monitoring is commonly used to monitor health behaviors and disease activity. Curelator Headache^® (now called N1‑Headache™) used a smart-phone application to evaluate the factors associated with adherence to mobile headache diaries. Participants were asked to complete baseline questionnaires and daily recordings of headache symptoms and other factors for at least 90 days. Response options were coded on an 11-point Likert scale ranging from not at all) to 10 (a lot). Out of 4207 who registered, 40.3% (1561) met the criteria for analysis and were included in the final sample. 90% reported migraine as their most common headache type and the majority (79.8%) reported a physician diagnosis of migraine. Of the 1561 participants 505 (32.4%) completed 90 days of self-monitoring. A higher level of anxiety was significantly associated with lower odds of completion of 90 days of data entry. It was also found that age, account type, and gender were associated with adherence rate. Methods designed specifically to improve adherence, such as increasing incentives or reducing the burden of diary entries, could target individuals with high anxiety. Description

Certain chronic diseases such as migraine result in episodic, debilitating attacks for which neither cause nor timing is well understood. Curelator Headache^® (now called N1‑Headache™) applied n=1 statistical analysis to a 326-migraine-patient database from a study in which patients used paper-based diaries for 90 days to track 33 potential factors associated with their migraine attacks. The average number of factors associated with attacks was four per patient. Factor profiles were highly individual and were unique in 85% of patients with at least one identified association. Correct identification of migraine triggers in individuals may allow more effective self-management of the condition. Description

Migraine headaches can seemingly occur spontaneously but are often reportedly precipitated by events or factors, known as migraine triggers. Singularly, the triggers are generally necessary but not sufficient enough to bring on headache by themselves and, compounding of those triggers is usually required. How triggers affect a given individual can vary based on intensity and exposure of the trigger, the susceptibility of the individual to the trigger, and the time interval between exposure and the ensuing migraine attack. Causative trigger identification and subsequent behavioral changes have the potential to benefit people with migraine. However, it is only achievable if reliable daily information is collected, if the analysis is performed using appropriate statistical modelling, and if trigger-modification testing is executed to prove causality. Description

The paradigm of Mr. Average has had enormous clinical (as well as social and behavioral) consequences for humans who are outliers. Migraine, one of the leading causes of disability worldwide, is a model condition if we want to study variation between individuals and the therapeutic implications of these differences. Patient susceptibility and response to a wide range of potential factors represent an important spectrum of real-world markers for studying individual variation in genetic, physiological, psychological, and ultimately biochemical domains. In addition, many migraine population studies have already been done, generating aggregate data about the average migraine patient, which we can use to benchmark against variation in the individual patient. Therefore it seems that an important next step would be to acknowledge the need to understand both the degree of, and the basis for, individual variation in chronic disease. And if the average approach is limiting, or possibly even causing harm in some individuals, then optimizing individual therapeutic pathways and outcomes may be the most effective way forward for patients with chronic diseases. Description

Chocolate is rarely statistically associated with migraine attack occurrence, and surprisingly just as likely to act as a migraine protector as a migraine trigger. N1‑Headache™ analyzed headache intensity in people with chronic migraine to assess chocolate as a potential trigger associated with more severe pain days and a potential protector associated with less severe pain days. Of 109 individuals, chocolate association with migraine attack intensity was identified statistically in very few individuals, with no difference in associations with “high” or “low“ pain days. Description

Stress is one of the most common self-reported triggers of migraine. N1‑Headache™ analyzed stress throughout the phases of the migraine cycle on an individual level in 730 people with migraine. Aggregate-level analysis of how stress varied across the migraine cycle suggests there is a homogeneous pattern whereby stress peaks during the pain phase compared to the premonitory, prodromal and interictal phases. However, there was a wide variation among these individual patterns, with distinct dominant patterns of stress variation in the migraine cycle that differ significantly from individual to individual. Description

The relationship between stress and migraine attack occurrence is complex, with evidence that changes in stress level, either an increase or a decrease (‘let down’) can be a trigger. N1‑Headache™ analyzed the relationship between daily, self-reported stress level and the intensity of headache pain. Of 120 individuals with chronic migraine, stress was associated with increased risk of a “high” pain day in 34; there was no association found between stress and headache severity in the other 86. Nearly all individuals who responded to the question suspected stress as a migraine trigger, yet there was no clear association between degree of suspicion and the proportions of individuals in whom an association was statistically identified. Description

Curelator studied headache intensity from individual attacks to assess potential triggers associated with more severe “high” pain days and potential protectors associated with less severe “low” pain days. Of 141 individuals with chronic migraine, the most common risk factors associated with “high” pain days were symptoms typically associated with migraine attacks which could be either potential triggers or prodromal symptoms. This compares to those potential protectors associated with “low” pain days, where the top factors were related to good sleep and positive mood. The aggregated data of the most common individual risk factors was consistent with current beliefs about the pathophysiology of migraine. However, there was considerable variability observed between individuals, underscoring the need for analysis at the individual level. Description

N1‑Headache™ studied the association of factors that modify risk of occurrence and influence severity of migraine attacks in the same person. Individual n-of-1 analyses in people with migraine demonstrated limited within patient overlap of occurrence factors and severity modifiers, suggesting that factors associated with migraine frequency and severity may act differently in the same patient. Description

Cutaneous allodynia is a common feature of migraine attacks. In those with migraine, allodynia is associated with inadequate response to acute migraine treatments and is a risk factor for migraine chronification. N1‑Headache™ explored the correlation within individuals between allodynia and migraine attacks. For 774 individuals, cutaneous allodynia was reported by two-thirds. Self-reported allodynia was associated with greater migraine attack frequency and greater migraine days/month compared to those who never reported allodynia. Description

Three cases of individuals with migraine who were receiving Botox while using the N1‑Headache™ digital platform were selected to illustrate how the platform can monitor individual tracked days, total headache days, migraine days and disablity due to migraine over time. After 90 days of data collection, a disability score was calculated using the same method as the MIDAS score, but from data captured prospectively on headache days. It was determined that using a digital platform and n=1 statistical algorithms is an effective way to monitor migraine outcomes remotely. Description

N1‑Headache™ compared chocolate intake and migraine attack occurrence in individuals to see how many suspected chocolate as a trigger and for how many an association between chocolate ingestion and migraine attack occurrence can be identified statistically. Of 774 individuals, chocolate was suspected as a trigger by half at baseline. However, an association between self-reported intake and risk of migraine attacks was identified statistically in only a few individuals, where chocolate could act as either a trigger or protector of migraine attacks. Description

N1‑Headache™ explored the belief in neck pain/tension as a migraine trigger and the statistical correlation between neck pain/tension and occurrence of migraine attacks. Each day the question was asked: ‘Did you notice neck pain or tension today?’ In 774 individuals, a high proportion suspected that neck pain is a trigger of their migraine attacks. However a statistical association between neck pain/tension and increased risk of attacks was identified statistically in 32% of individuals. There was an association between the strength of neck pain/tension being suspected as a trigger and the frequency of statistical confirmation of an association with their attacks. The temporal association between neck pain/tension and migraine attacks suggests that neck pain is part of migraine symptomology, rather than a migraine trigger. Description

N1‑Headache™ explored whether daily variation in nitrate intake in food is associated with risk of having migraine attacks. For 774 individuals, nitrates in food was suspected as a trigger by nearly half at baseline but in only a small proportion of these individuals was an association identified statistically between daily self-reported intake and increased risk of migraine attacks. Description

Repetitive yawning is a common premonitory migraine symptom, and has a high predictive value for a migraine attack. N1‑Headache™ explored how many individuals recorded excessive yawning and for how many an association between excessive yawning and migraine attacks could be identified statistically. Out of 483 individuals, excessive yawning was a discernible symptom and sensitive predictor of migraine in 27%. Description

It has been reported that the food additive, monosodium glutamate (MSG) causes headaches in susceptible individuals and may trigger migraine attacks. N1‑Headache™ was used to determine how many individuals suspected MSG as a migraine trigger and how many had a statistical association between MSG intake and migraine attacks. Of 774 individuals, half suspected MSG as a trigger at baseline, but in only a very small proportion of individuals was an association between self-reported intake of MSG and increased risk of migraine attacks identified statistically. Description

People with migraine often report an increase in the frequency and/or severity of attacks on non-working days, often referred to as “weekend headache.” Curelator explored the association between type of day and risk of migraine attack. Through individual analysis of self-reported patient data in 707 individuals with migraine, N1‑Headache™ was able to show that there is minimal association between the type of day and risk of migraine attack. Description

In this study, we hypothesize that modifying potential migraine risk factors may also impact the severity of an attack (severity modifiers). N1‑Headache™ identified factors associated with headache pain severity during migraine attacks in individuals, and compared the prevalence and the effect size of these severity modifers in chronic vs. episodic migraineurs. Pain severity was collected on days with headache through the question “How bad was your headache pain at its worst today?” with three possible answers Mild/Moderate/Severe. After studying the data of 775 participants, it was concluded that severity modifier profles were highly individualized. For some factors, the prevalence of severity modifers was higher in chronic compared to episodic migraine and varied with migraine frequency. Description

N1‑Headache™ compared suspected triggers with those identified statistically from prospective data in 488 individuals with episodic or chronic migraine. An average of 23 migraine triggers were suspected per individual, but a mean of less than 3 factors were associated with increased risk of migraine attack. Statistical associations between suspected triggers and occurrence of migraine attacks were confirmed in less than 15% of individuals for most factors, in both the episodic migraine and chronic migraine groups. Description

N1‑Headache™ analyzed which risk factors were identified as possible “triggers” of migraine initiation as opposed to being associated with severity of a migraine. For 479 participants, the profiles of risk factors that were associated with incidence and severity of migraines varied considerably between individuals. Overall, more risk factors were associated with migraine attack occurrence compared to migraine attack severity, both in terms of prevalence and types. Description

This study applied an n=1 analytic framework to individual daily diary data to explore both intra and inter individual differences in the perimenstrual occurence of migraine. Results showed that menstruation was associated with increased odds of migraine in 66% women studied, with substantial variation in the association between migraine and menstruation based on the ICHD-3 beta time window (days -2 to +3). In the future, it may be possible to incorporate specific logorithms into the pathophysiology and management of menstrual migraine. Description

Alcohol has long been suspected as a migraine trigger, but numerous studies have produced inconclusive results. N1‑Headache™ explored this question by statistically comparing the intake of alcohol and occurrence of migraine. Two different time relations, lagged (previous day) and co-occurring (same day) effect were analyzed. Overall, alcohol was suspected as a trigger at baseline by more than half of 272 individuals, but an association between daily self-reported intake and increased risk of migraine attacks was identified statistically in only a small proportion of individuals. There was no clear association between the strength of alcohol being suspected as a trigger and frequency of statistical confirmation of alcohol being associated with their migraines. An important limitation of this study is that the current study determines simple associations and not causality. Description

To clarify whether migraines are distinct compared to other headaches, N1‑Headache™ investigated associated risk factors at the N=1 level in patients who experience both migraine and non-migraine headaches based on ICHD-3beta criteria. A sample of 479 participants reported daily on headache occurrence, symptoms of migraine attack, degree of pain associated with the headache, and exposure to 70 potential migraine risk factors. Individual-level analyses revealed that nearly all risk factors (triggers and protectors) were usually associated with occurrence and severity of either migraine or non-migraine headaches, but not both. Description

Daily diary studies of women with migraine reveal that at the population level, the probability of migraine is increased shortly before and shortly after the onset of menstrual bleeding. The ICHD-3 beta criteria for migraine defines the time of increased risk due to menstruation as the time from 2 days before until 3 days after bleeding begins.This population average data does not consider that migraine risk may vary from woman to woman in relation to bleeding. Here, we consider that headache risk may vary from woman to woman on particular days or particular segments of the menstrual cycle. The analysis sample consisted of data collected from 50 menstruating females reporting on a median of 200 days using the N1‑Headache™ app. Using a categorical approach, analysis of pooled data from all women indicated that migraine risk was highest during actively bleeding days, elevated to a lesser degree on pre-bleeding days and more comparable on post bleeding days. A continuous time method of analysis showed that each woman’s migraine risk often varied not only across menstrual stages but also within specific stages. Understanding these individual patterns may prove useful in clarifying the mechanisms of menstrual migraine and in developing individualized approaches to short-term preventive treatment. Description

The ICHD-III classifies Menstrually Related Migraine (MRM) as a subtype of migraine without aura if attacks occur on 2 out of 3 menstrual windows (defined as the five days centered on the first day of bleeding) in women who also have non menstrual attacks. Because of concerns that MRM diagnoses set by this criterion may lead to unacceptable diagnostic error rates, a novel statistical method to diagnose MRM was compared against the ICHD-III 2/3 criterion. Analysis of pooled data from 106 women using the N1‑Headache™ digital platform was compared to 123 women attending the City of London Migraine Clinic whose data have previously been published. All women had logged migraine attacks during at least 3 consecutive natural menstrual cycles and MRM was diagnosed by the standard ICHD-III criterion. Based on the results of the statistical analysis, the ICHD-III criteria did not select women with prolonged menstrual attacks. It was concluded that the proposed standard of 3-cycle observations prior to setting an MRM diagnosis should be extended with at least one perimenstrual window to obtain sufficient information for statistical processing. Description

As a critical component of self-management of their condition, this study examines the ability of episodic migraineurs to self-predict attacks on an individual level. Individual participants used N1‑Headache™ daily for at least 90 days, entering details about headaches and also migraine expectation for the next 24 hours. In this study of 146 individuals with episodic migraine, 9% were good migraine predictors (defined as having an accuracy of ≥75%) and 27% were bad predictors (defined as those with ≤25% accuracy). The difference between each of these groups and the rest of the study population was statisticaly significant (p<0.001). Individuals were much better at predicting migraine-free days, with 71% good predictors and <1%% bad predictors. In the 5% of people that can accurately predict migraine and migraine-free days within 24 hours with high accuracy (and the 9% of good migraine predictors), their associated disability and headache frequency remains higher than in the total sample. Description

An electronic diary, N1‑Headache™ was used to compare medication use and possible overuse patterns in individuals with migraine in the United States (US) and the United Kingdom (UK). Of 467 individuals, medication overuse was identified in 29% of individuals in a predominantly physician referred group in the US and in 19% in a predominantly population recruited group in the UK. The number of medications or medication classes per user was significantly higher for chronic migraine (CM) than episodic migraine (EM) in both countries. Patients with CM in the US took more medication per headache than UK CM patients, but UK EM patients used more medication per headache than US EM patients. Surprisingly, the average number of medications and classes of medications used on a per headache basis was significantly higher in EM than in CM group in both countries. In both the US and UK, triptans are most frequently used by EM patients and NSAIDs are most frequently used by CM patients. Opioids were used more frequently by CM patients in the US than the UK. Both CM and EM patients showed different rates of medication use per individual, as well as different medication overuse and medication use frequencies. Description

N1‑Headache™, a platform that demands daily tracking of both risk factors and headache symptoms, was used to record profile demographic data, as well as real time daily collection of headache data including frequency, possible triggers, treatment and disability score (MIDAS) among other variables. An analysis was performed to compare characteristics of those who completed 90 days of data collection compared to non-completers. Of 2678 individuals with migraine registered with the platform, at 90 days only 493 individuals (18.4%) reliably completed data entry. Completers were older, less likely to be employed, had slightly less severe pain although they visited the ER less frequently, utilized less caffeine and smoked less cigarettes when compared to all non-completers (drop-outs). Description

Since Dr. Edda Hanington's study in 1967, which suggested a connection between tyramine-containing foods and migraine attacks, doubts have been raised about the prevalence of this sensitivity in the migraine population. To explore this question, N1‑Headache™, a digital platform that enables daily data capture and novel n-of-1 analytics, was used to statistically compare self-reported intake of tyramine-containing foods and occurrence of migraine attacks. Of 488 individuals using the platform in this study, almost half suspected tyramine was a migraine trigger at baseline. However, data from 49% was not analyzable due to low or unchanged intake of tyramine. In those with analyzable data, an association between self-reported intake of tyramine and risk of migraine attacks was identified statistically in a small portion of individuals. Tyramine was associated with decreased attack risk (potential protector) in a similar proportion of individuals. Description

Migraine attacks may be triggered by combinations of internal and external risk factors which differ markedly between individuals. In the migraine literature many factors have been proposed as triggers, however these are usually based on retrospective recall which is subject to misinterpretation and recall bias. This study describes beliefs about personal triggers and compares these beliefs in those with episodic migraine (EM) and chronic migraine (CM). Using the N1‑Headache™ digital platform, an initial questionnaire presented a list of potential migraine risk factors and asked individuals to rate their importance as personal migraine triggers on a 0 to 10 scale (0=none; 10=maximal). An average of 23 triggers were suspected by 488 individuals at baseline. When comparing those with EM (n=391) and CM (n=97), there are no clear differences in terms of the range and type of suspected triggers other than the CM group suspecting slightly more triggers on average than the EM group. Description

N1‑Headache™ is a digital platform that collects daily data about an individual’s migraine attack risk factors (i.e., up to 80 factors including mood, stress, diet, activity, weather etc.) and then statistically identifies associations between these factors and the occurrence of migraine attacks. The resulting associations include potential triggers (both 'true' triggers associated with increased risk of attack and premonitory symptoms) as well as potential protectors (factors associated with reduced risk of migraine attack occurence). After 557 Individuals used the N1‑Headache™ app daily for 90 days entering details about headaches and tracking daily factors that may affect migraine attack occurrence, potential protectors were found in 471 (84.4%) individuals. On average individuals had 3 such associations (range 0 - 16). Identifying ‘protectors’ may be as clinically useful as identifying potential triggers in people with migraine. Description

Historically, factors believed to trigger migraine attacks were identified through analysis of aggregated data from populations of patients, but may not apply to a given individual. This study applied n=1 statistical methods to analyze a 326-migraine patient database from the PAMINA study to determine for each individual which triggers are associated with their attacks. In 87% (285) of migraine patients analyzed, “potential trigger” profiles were generated, with an average of four factors significantly associated with occurrence of migraine attacks per patient. These “potential trigger” profiles were highly individual and unique in 85% of the patients who had at least one identified significant factor. Identification of associations between factors and attacks is not proof of causality as these factors may be triggers and/or may be warning signs of their attacks. Description

The PAMINA study analyzed collectively factors increasing and decreasing the risk of attacks in patients with migraine and showed that a reduction in the risk of attacks was associated with a number of factors: including days of holiday or days off, Tuesdays, consumption of beer and (less robustly in the analyses) progresterone-only contraception and being divorced. Here, we report a new analysis of pseudonymized data from 326 participants in the PAMINA study to identify potential 'protectors' in individual patients. This study used n=1 analytical methods to determine statistically significant associations between a variety of potential migraine risk factors and occurrence of migraine attacks. In the PAMINA study patients, we found 0-6 'protectors' per patient: 35% of patients had at least one. Identifying protective factors may be a useful addition to the management of migraine, potentially of equal importance to identifying triggers. However, association is not evidence of causality: it is important to guide each individual to think about their behaviors and self-analyse why and in what context such factors - whether potential protectors or triggers - may affect their condition. Description

Migraine attacks may be triggered by combinations of internal (e.g. hormonal) and external (diet, environment, lifestyle) factors which differ markedly between individuals. Most migraineurs suspect certain factors as triggers: this is usually based on retrospective recall and is subject to recall bias and misinterpreted associations. This study compared suspected triggers from 390 individuals with migraine against statistically identified triggers after 90 days use of Curelator Headache^® (now called N1‑Headache™). Overall, of 7317 suspected triggers, only 1186 (16%) were shown to be statistically associated with migraine attack occurrence. An average of 4.6 factors were statistically associated with increased risk of attacks for each individual. Of these, an average of 3 triggers were previously suspected by an individual and 1.6 were not previously suspected. Increasing an individual’s knowledge about their potential migraine triggers may enable them to improve their condition by avoiding or coping with their triggers differently. Description

Curelator Headache^® (now called N1‑Headache™) is a digital platform which collects daily data about potential migraine attack risk factors and then statistically identifies individual associations between factors and migraine attack occurrence. This analysis also identifies factors statistically associated with reduced risk of attack occurrence called potential ‘protectors.’ In 284 individuals, 87% had at least one ‘potential protector‘. Most common were waking refreshed (46%), sleep quality (23 %), sleep duration (11%), feeling happy (33%), feeling relaxed (23%), physical activity (20%) and caffeine (14%). Surprisingly, a number of factors generally thought of as triggers were identified as ‘protectors’ in some individuals, such as travel and alcohol. Therefore, identifying factors that are associated with reduced incidence of migraine attacks may be a useful addition to the management of migraine, potentially of equal importance to identifying triggers. Description

Curelator Headache^® (now called N1‑Headache™) is a novel digital platform that allows individuals with migraine to discover their personal triggers and protective factors and eliminate factors that are not associated with their attacks. To maximize both patient compliance and accuracy of data entry, a freely licensable semiotic system called Visual Migraine Language^® (VML) was developed and tested on people with migraine. VML codifies migraineurs’ environments, possible interventions and clinically relevant aspects of their condition into a series of intuitive symbols. Use of the VML simplifies the user interface, enhances user experience, and encourages daily data entry compliance and adherence compared to text-based interfaces. Description

The episodic, largely unpredictable, nature of migraine attacks implies they occur as a result of a complex interaction between an individual’s genetics and their personal environment. Individuals appear to be sensitive to different sets of triggers. In addition, ‘protective factors’ may at least partially counteract the effect of triggers. An approach is needed that objectively identifies important triggers and protectors for individuals with migraine. Curelator Headache^® (now called N1‑Headache™) is a novel digital platform that allows individuals with migraine to discover their personal triggers and protective factors and eliminate factors that are not associated with their migraine attacks. To test the data entry interface, 511 people with migraine were recruited on-line and registered to use the pilot versions of Curelator Headache^® for 90 days. Of those who registered, 375 used the platform for>2 days and 218 (58%) entered adequate data to allow generation of personal ‘Trigger Map^®s’. It is expected that improved knowledge of potential migraine triggers and protective factors will help individuals, and their clinicians, manage their migraine more effectively. Description