About Migraine Triggers
What types of studies have doctors and scientists done to figure out what triggers migraines?
Four main categories of studies have been reported in the scientific literature: (1) Challenge studies in which migraine sufferers are deliberately exposed to a potential trigger (and usually, on another occasion, to a placebo) to see whether a headache occurs; (2) Treatment studies in which migraine patients avoid exposure to a set of potential triggers for some period of time, to see whether this reduces the rate of headaches; (3) Diary studies in which patients keep a record of their headaches and what potential triggers they have been exposed to, and their temporal relationship is examined; and (4) Survey studies in which patients are asked for their opinions about what causes their headaches.
What has been learned from challenge studies?
Challenge studies potentially provide the strongest evidence for cause-and-effect linkages. Over the years, many convincing positive results have been reported. For example, writing in The Lancet, Littlewood et al. (1988) reported a study of 11 migraine patients who felt certain that red wine triggered their headaches. Nine of the 11 experienced migraines after being challenged with red wine, whereas zero out of 8 had headaches after drinking vodka with a comparable alcohol content. (The beverages were chilled and drank through a straw to conceal their identity.) Other challenge studies showed that certain individuals' headaches were brought on by fasting (Blau and Cumings, 1966), fatigue, chocolate (Gibb et al., 1991) and the chocolate ingredient phenylethylamine (Sandler et al., 1974), although in some cases, only about half of the subjects reporting sensitivity to these things experienced headaches when challenged. In a placebo-controlled study, Henderson and Raskin (1972) demonstrated that sodium nitrite (found in hotdogs and other smoked meat) reliably produced headaches in one migraine sufferer who was tested on many occasions (some involving placebo).
Have all challenge studies produced positive findings?
No, some challenge studies have failed to find very strong effects. However, most of these studies did not focus specifically on individuals reporting a special susceptibility to the trigger under investigation. Thus, these weaker results may simply reflect the fact that people differ in their triggers, rather than invalidating the positive results discussed above (see Radnitz, 1990, for a thoughtful and detailed review of this literature).
What have treatment studies shown?
Treatment studies are a mixed bag, with some studies reporting beneficial effects of diets that eliminate large numbers of foods (Egger et al, 1983; Mansfield et al., 1985) and others reporting no benefits. The studies have tended to have high drop-out rates, weakening the conclusions (Radnitz, 1990). Similarly, compliance has been a problem with therapeutic use of elimination diets, which is not surprisingly since patients are being asked to give up a large number of different foods. Most neurologists do not view elimination diets as a generally useful therapeutic approach.
What have diary studies shown?
The largest diary study that we are aware of (Woeber et al., 2007) collected a comprehensive diary from 327 migraine sufferers over a 3 month period and used statistical methods somewhat similar to those used on MyMigraineJournal.com to determine what environmental triggers were causing headaches the following day. The authors concluded that menstruation in women had a strong triggering effect, with real but smaller effects of muscle tension in the neck, psychic tension, tiredness, noise and odors. No food and drink items were identified as triggers, although there was some evidence for a completely unexpected protective effect of drinking beer the night before (i.e., fewer headaches after drinking beer), which the authors guessed might be a statistical "false alarm". Although the analysis was statistically sophisticated, it assumed that triggering operated uniformly across migraineurs, and thus the negative group-level results for food and beverage items might simply reflect triggering effects varying from one individual to the next.
An important study by Prince et al. (2004) had 77 migraine patients record their headaches, and then examined the connection between occurrence of headaches and the actual weather occurring at that time in the individual’s location (this was obtained from US National Weather Service). The authors concluded that weather did have an effect for about half of the people studied (the most common linkage being triggering of headaches by hot/humid weather).
What about survey studies?
The majority of studies of migraine triggering have simply asked migraine sufferers to record their opinions about what causes their headaches. These studies make it clear that most migraine sufferers do believe that their headaches can be precipitated by various triggers. Silberstein (1995) surveyed 500 migraine sufferers and found the following percentages believing their headaches were prompted by various categories of triggers:
| Stress | 79.0% |
| Changes in Weather | 44.0% |
| Before Menstruation | 37.2% |
| Changes in Light | 33.8% |
| Eating certain Foods | 30.0% |
| During Menstruation | 26.8% |
| Fatigue/insomnia/other | 6.6% |
| Other Events (combined) | 13.2% |
Another fairly large and comprehensive study was conducted by Kelman (2007), who surveyed 1207 migraine patients treated at a headache clinic. 75% of the patients reported believing that their headaches were sometimes or always triggered, with the average patient specifying 6.7 triggers. The list of frequently reported triggers was similar to that shown in the table above, with a substantial number of migraine patients also reported triggering by lack of food or skipped meals (57.3%), odors (43.7%), and sleeping late (32.0%). Similar results have been found in "prospective" studies, in which patients were followed for some period of time and asked to speculate about the causes of individual headaches (e.g., Robbins, 1994).
The list of potential trigger questions that you will be offered if you sign up for MyMigraineJournal.com includes most of the factors mentioned in the survey studies (with the most-mentioned and most-controllable items being pre-checked by default).
What do we know about whether people’s opinions of what causes their headaches are accurate or not?
Some interesting data on this point comes from the Prince et al. (2004) weather study mentioned above. As noted, the authors found strong objective evidence that weather really did trigger some headaches. However, they also looked at migraine patients’ reports about whether they believed that weather affected their headaches. A large fraction did believe that weather played a role, but there was no sign that individuals who believed weather triggered their headaches had any greater susceptibility to weather-induced headaches than did other subjects in the study. This suggests that many migraine sufferers have drawn erroneous conclusions about what triggers their headaches. This is also consistent with the results of the challenge studies described above: on the one hand, these studies verified some individuals' beliefs about what triggers could produce headaches in them, but they also found that a sizable fraction of individuals did not have the expected reaction to the challenge (e.g., Gibb et al., 1991)
The conclusion that people are often tricked about what causes a headache or other symptom should probably not be surprising. Cognitive psychologists have performed many laboratory experiments to assess people's ability to detect causal linkages in events that they experience sequentially, and have found that performance is often rather poor, and becomes worse when the time between events is lengthened, even over periods much shorter than one day (Buehner and May, 2002; Shanks, Pearson, and Dickinson, 1989). Moreover, the task facing a migraine sufferer who wants to figure out what triggers his or her headaches is harder than the task facing subjects in the psychological studies just mentioned, because there is a long list of potential headache triggers to be considered (the studies involved events that varied on only a few dimensions).
In our opinion, a reasonable (but tentative) conclusion from the literature on migraine triggering would be that while some people are able to correctly identify some migraine triggers, impressions formed about what causes events like headaches are often wrong, and thus many migraine patients are probably avoiding things that do not cause their headaches--as well as failing to avoid things that do.
Does the approach being followed at MyMigraineJournal.com make sense in light of the scientific literature discussed above?
As described above, the migraine literature suggests that migraine triggering does exist, and may well occur for many or most migraine sufferers; however the triggers probably differ from one person to another, and migraine sufferers are often frustrated in their efforts to figure out what triggers cause their headaches. Therefore, making a systematic effort to objectively determine one’s own migraine triggers would seem rational for those suffering from frequent headaches--especially if this can be done conveniently and accurately. MyMigraineJournal is intended to facilitate that. As discussed elsewhere on this site, users should plan to provide information to the website for a number of months before clear findings are likely to emerge. We urge users to make a habit of using the system daily. Note that the actual time expenditure needed to enter daily data for 6 months adds up to only about 90 minutes, based on estimated daily usage time of 30 seconds/day. If you develop new hypotheses about possible headache triggers that you did not include in your initial sign-up, you can always add them on MySettings (this does not disrupt analysis of already collected data; why?). Also, keep in mind that if you miss some days or even weeks, this does not eliminate the value of the existing data; if that happens, just return to daily usage as soon as you can.
References
Blau, J. N., & Cumings, J. N. (1966). Method of precipitating and preventing some migraine attacks. British Medical Journal, 2, 1242-1243.
Buehner, M. J., & May, J. (2002). Knowledge mediates the timeframe of covariation assessment in human causal induction. Thinking and Reasoning, 8, 269-295.
Chabriat, H., Danchot, J., Michel, P., Joire, J. E., & Henry, P. (1999). Precipitating factors of headache. A prospective study in a national control- matched survey in migraineurs and nonmigraineurs. Headache, 39, 335-338.
Egger, J., Carter, C. M., Wilson, J., Turner, M. W., & Soothill, J. F. (1983). Is migraine food allergy? A double-blind controlled trial of oligoantigenic diet treatment. Lancet, 2(8355), 865-869.
Gibb, C. M., Davies, P. T., Glover, V., Steiner, T. J., Clifford Rose, C., & Sandler, M. Chocolate is a migraine-provoking agent. Cephalgia, 11, 93-95.
Henderson, W. R., & Raskin, N. (1972). Hot dog headache: Individual susceptibility to nitrite. Lancet, 2, 1162-1163.
Kelman, L. (2007). The triggers or precipitants of the acute migraine attack. Cephalgia, 27, 394–402.
Littlewood, J. T., Glover, V., Davies, P. T. G., Gibb, C., Sandler, M., and Rose, F. C. (1988). Red wine as a cause of migraine. The Lancet, I(8585), 558-559.
Mansfield, L. E., Vaughan, T. R., Waller, S. F., Haverly, R. W., & Ting, S. (1985). Food allergy and adult migraine: double-blind and mediator confirmation of an allergic etiology. Annals of Allergy, 55, 126-129.
Marcus, D. A., Scharff, L., Turk, D., & Gourley, L. M. (1997). A double-blind provocative study of chocolate as a trigger of headaches. Cephalgia, 11, 93-95.
Prince, P.B., Rapoport, A. M., Sheftell, F.D., Tepper, S.J., and Bigal, M.E. (2004). The effect of weather on headache. Headache, 44, 596-602.
Radnitz, C. L. (1990). Food-triggered migraine: a critical review. Annals of Behavioral Medicine, 12, 51–65.
Robbins, L. (1994). Precipitating factors in migraine: a retrospective review of 494 patients. Headache, 34, 214-216.
Sandler, M., Youdim, M., & Hanington, E. (1974). A phenylethylamine oxidising defect in migraine. Nature, 350, 335-337.
Shanks, D. R., Pearson, S. M., & Dickinson, A. (1989). Temporal contiguity and the judgment of causality by human subjects. Quarterly Journal of Experimental Psychology B: Comparative and Physiological Psychology. 41, 139-159
Shaklee, H. (1983). Human covariation judgment: Accuracy and strategy. Learning and Motivation, 14, 433-448.
Silberstein S. D. (1995). Migraine symptoms: results of a survey of self-reported migraineurs. Headache, 35, 387–396.
Woeber, C., Brannath, W., Schmidt, K., Kapitan, M., Rudel, E., Wessely, P., Woeber-Bingoel, C., & the PAMINA Study Group (2007). Prospective analysis of factors related to migraine attacks: the PAMINA study. Cephalgia, 27, 304-314.