Lifestyle Advice for Pediatric Migraine: Blaming the Patient, or Evidence Based?

Amy A. Gelfand, MD, MAS; Samantha L. Irwin, MBBS


Semin Neurol. 2020;40(3):277-285. 

In This Article

Section 3: "Avoidance of Skipping Meals"

What is the Evidence?

In the survey study of adolescents in Spain, eating breakfast before school was less common among those with headaches (n = 494) versus those without (n = 1,125): 70.2 versus 77.9%, p = 0.022. However, no other differences in eating habits were seen.[6] In a univariate analysis, those who skipped breakfast had higher odds of headaches (OR: 1.5; 95% CI: 1.2–1.9); however, this was no longer a significant predictor in the multivariate analysis. Of the 494 participants with recurrent headaches, 184 (11.4% of the total sample) fulfilled criteria for probable migraine. There were no observed differences in dietary habits between those with migraine and those with other headaches.

In an internet-based survey study of 256 U.S. high-school students with migraine, approximately 90% of teens reported skipping breakfast at least once a week. The average number of missed breakfasts was 3 out of a 5-day school week.[30] Those who skipped breakfast at least once a week did not have a higher (self-reported) migraine frequency than those who did not. In fact, the observed effect was in the opposite direction, although the finding was not statistically significant (estimated effect of breakfast skipping on headache days per month: −1.5 days; 95% CI: −4.0 to 1.1).[30]

In the "Adequate Hydration" section where adolescents were asked to self-report how much water they drank, adolescents "on average reported skipping one meal per week despite recommendations of consistently eating three meals per day."[8] Hence, it seems clear that many adolescents with migraine sometimes skip meals. Whether this influences their migraine frequency remains somewhat unclear.

In adults, headaches are more likely to occur on days of religious fasts.[31–34] These fasts are of 25-hour duration (Yom Kippur) or from dawn to dusk for a month in the case of Ramadan. These data suggest that day-long fasts may activate migraine or other headache disorders. However, there may also be other changes in activities or behaviors on these days that may be contributing to headache, including perhaps decreased hydration status and/or caffeine withdrawal. Of note, in the Spanish adolescent survey study, caffeine intake was associated with increased odds of having headaches (OR: 1.5, 95% CI: 1.1–2.2). Tobacco use was also independently associated with increased odds of headaches in that study (OR: 2.3, 95% CI: 1.5–3.4).[6]

Interestingly, pretreating with a long half-life nonsteroidal anti-inflammatory drug (NSAID) right before beginning a religious fast can decrease the occurrence of headache in adults. In a study of 195 adults who typically experience headache during the Yom Kippur fast, participants were randomized to pretreatment with etoricoxib (n = 99) or placebo (n = 96). In the treatment group, 36.4% experienced headache, whereas 67.7% of the placebo group experienced headache (p < 0.0001).[32] Similarly, on the first day of Ramadan, adults randomized to take etoricoxib (n = 96) in the predawn before fasting were less likely to have headache that day than those who took placebo (n = 92): 21 vs. 46%, p < 0.001. Interestingly, headaches associated with Ramadan are most likely to occur on the first day of Ramadan, suggesting that to an extent it may be a change in feeding pattern—rather than fasting per se—that is associated with headache. In the aforementioned study, the proportion who experienced headache in the placebo group was 46% on day 1, 20% on day 6, and 15% on day 14 of Ramadan.[33]

In a related example, adults fasting preoperatively were more likely to develop a preoperative headache if their anesthesia induction was done after noon, or if they usually consumed >400 mg of caffeine per day.[35] The onset of their fast was presumed to have been at midnight. Hence, duration of fasting may be an important determinant of likelihood of headache.

Some individuals with migraine may fast involuntarily. In a large, population-based survey study of 48,645 Canadians aged ≥12 years, respondents were queried regarding migraine status and food insecurity in the home. The prevalence of migraine was 10.2% and food insecurity was present in 7.6%, based on responses to the 18-question Canadian Food Security Status Module. Food insecurity is defined by the U.S. Department of Agriculture as "household-level economic and social conditions of limited or uncertain access to adequate food."[36] Those with migraine had a more than twofold increase in odds of experiencing food insecurity (OR: 2.4, 95% CI: 2.0–2.8). After adjusting for age, sex, household income, mood disorders, and other health and socioeconomic variables, migraine was still independently associated with odds of food insecurity (OR: 1.5, 95% CI: 1.2–1.8). Diabetes and bowel disorders were not associated with higher odds of food insecurity, but migraine, arthritis, and asthma were. The authors hypothesized the impaired mobility may play a role in being able to access food on a regular basis, or that disability from migraine may prevent individuals from shopping and/or cooking.[37]

Patients with migraine will often ask whether there are certain foods they should avoid. This is a complex area in need of more research; however, based on available evidence to date, there does not seem to be an evidence-based reason to counsel patients to avoid certain types of foods. In a study of 327 adults with migraine, participants completed a headache diary each night for 90 days, as well as a diary indicating whether they had or had not experienced 52 different physical, environmental, or dietary factors thought to be related to migraine. Food types queried included cheese, chocolate, coffee/caffeine-containing beverages, dairy products, fast food, fish, vegetables, nuts, citrus fruits, other fruits, aspartame, glutamate, and sodium nitrite. Data from 28,325 patient-days were analyzed, with headache occurring on 8,648. Hazards of headache occurrence were higher on menstruation and premenstrual days, and on days where the person had experienced neck pain or tiredness the day before. None of the above-mentioned dietary factors were associated with increased risk of headache. The authors conclude, "Our study provides evidence for the limited importance of nutrition in the precipitation of migraine."[38]

In subsequent analyses of the same dataset, researchers sought to determine whether dietary factors might be associated with migraine in some individuals, even though they were not for the group overall. For these N-of-1 analyses, the 90-day diary data for each individual were analyzed to see if that individual was more likely to have a headache after eating a certain food versus when they had not eaten that food. A few of the participants had one or more dietary associations identified; however, the proportions of participants who did was small: caffeine, 7.7%; nuts, 7.4%; MSG, 6.1%; citrus fruits, 6.1%; sparkling wine, 4.6%; artificial sweetener, 4.0%; nitrates, 3.4%; white wine, 2.5%; chocolate, 2.5%; red wine, 2.1%; spirits, 2.1%; beer, 1.1%; and cheese, 0.9%. By comparison, missed meals were associated with migraine attacks in 20.6% of the participants, tiredness in 33.1%, and neck pain in 43.6%.[39]

Nonetheless, many patients do perceive certain foods to be the trigger.[40] Part of the explanation for this might be the difficultly in sorting out premonitory phase symptoms from migraine triggers.[39] Before the headache phase of migraine, many patients experience what is called a premonitory phase during which there are already changes occurring in the brain as the migraine attack is beginning.[41] Premonitory symptoms can include increased yawning, neck pain, irritability, fatigue, increased micturition, and food cravings,[42] and premonitory phase symptoms commonly occur in pediatric migraine as well.[43,44] If the brain has entered the premonitory phase of migraine, the person may crave chocolate and eat it. When the headache phase follows, the person may conclude that chocolate eating caused migraine; however, what actually happened is the opposite: migraine (specifically the premonitory phase of migraine) caused chocolate eating. Indeed, empirically chocolate was no more likely to trigger migraine than the control substance in a randomized controlled trial in adults, regardless of whether the person thought chocolate was a trigger for them.[45]

Another Possible Example. Does caffeine intake cause migraine, or does premonitory phase fatigue lead people to drink more caffeine than they usually would and a headache then follows? In a recent study of 98 adults with migraine, participants kept prospective diaries for at least 6 weeks, yielding 4,467 days of data—of which 825 were migraine days. Drinking one or two caffeinated beverages in a day was not associated with increased risk of headache; however, drinking three or more was associated with a trend in increased odds of headache.[46] Is this a dose effect? That is, does taking in high amounts of caffeine activate migraine, whereas lower amounts do not? Or, were people more likely to drink excess caffeine when they were feeling tired due to premonitory fatigue? Given the significant complexity in figuring out what activates migraine, it is perhaps no surprise that perceiving migraine triggers accurately can be challenging.[47]

Recommendations for the Clinician

Counseling. "Adolescents who skip breakfast may be more likely to have headaches. I recommend that you eat something each day before noon. You don't have to avoid any particular types of foods, though I recommend that you follow a generally healthy diet." And, if relevant, "If you plan to take part in a religious fast, or if you are going to need to fast for a procedure, we can discuss whether you would like to pre-treat with a medication that might make headache less likely to occur during your fast, and/or whether you need to decrease caffeine intake beforehand."

Facilitating Implementation. 504 letters can include a recommendation to allow students to eat snacks during the day. Providing a prescription for a long-acting NSAID to those who participate in religious fasts, or who may need to fast preoperatively for procedures, may make it easier for them to avoid headaches on those days, as may tapering down caffeine use prior to fasting events. Screening for food insecurity in clinic may help identify families who need public assistance to ensure adequate food availability.

Recommendations for Research

Studies are needed to determine whether meal skipping and/or fasting duration influences migraine frequency in children and adolescents. It will be important to separate out effects of not eating from that of caffeine withdrawal or inadequate hydration. The influence of diet quality and content on migraine in youth is also worthy of investigation.