Self-Prediction of Headaches, Stress Letdown Are Targets for Early Migraine Intervention

For many migraine sufferers, current strategies to treat headaches, or to prevent their occurrence, are insufficient to ward off the near-daily occurrence of head pain. The conventional targets for migraine treatment were challenged by Albert Einstein Medical Center’s Richard Lipton, MD, director of the Montefiore Headache Center in New York City, during a headache-focused scientific session at the American Academy of Neurology’s 2014 annual conference in Philadelphia, PA. During his presentation, Lipton reviewed targets and opportunities for pre-emptive migraine treatment, along with the possibility of preventing progression in the natural history of migraine, among other topics.

Conventional migraine treatment may be acute, designed to ward off or stop a single headache attack; also, treatment may be prophylactic, where medication is taken daily to reduce the overall incidence of headache for a sufferer. Finally, non-pharmacologic treatment such as behavioral therapy or lifestyle modification may be employed.

Opportunities exist to pre-empt single attacks of headache by intervening earlier in the headache cycle. Further, finding a way for a migraine sufferer to experience fewer overall headaches may help prevent progression of episodic migraines to chronic migraines. Currently, there is some treatment initiated during known periods of increased migraine risk, as when a woman is prescribed triptans for short-term use just before and during menses to prevent menstrual migraine. However, Lipton said there is still an unmet need for accurate prediction of migraine attacks to further personalize headache treatment and ensure that treatment is delivered in the brief, critical window before a headache occurs.

Discussing ways to achieve more accurate prediction of headache attacks, Lipton enumerated known predictors, which may be grouped broadly into: triggers, or exogenous factors, such as red wine consumption; premonitory features, such as excessive yawning; and self-prediction, where the sufferer somehow senses an impending attack. There are no known biomarkers for impending attacks, though this would also be a logical target for prediction.

For the migraine sufferer, there are four possible phases of the disease state: first, the interictal period, when there is no headache and no prodrome; next, the premonitory state (aura will occur during this time period, but it exists even for those who do not experience aura); third, the ictal state, during which headache occurs; and finally, the postical or postdromal state, when effects of the headache linger. The goal is to minimize time spend in the ictal phase, decreasing overall headache burden and maximizing the interictal period.

Premonitory features are important to capture, and pushing intervention earlier in the headache circle will require researchers to tease out the earliest possible signaling of an impending headache for individual sufferers. One study used electronic diaries to collect information from headache sufferers about premonitory symptoms. Features predictive of headache within 72 hours included difficulty speaking, difficulty reading or writing, emotional lability, yawning, and blurred vision, all of which at least doubled the likelihood of headache in the next three days. Stress was not premonitory of headaches in this study.

Lipton also referred to his work, published in the journal Neurology this year, which examined about 1,000 headache days in 110 headache attacks suffered by 17 patients, who averaged three to 10 attacks per month. In this group, he further examined the relationship between stress and headache, and tightened the length of the premonitory period to a more clinically useful 24 hours. Stress on the preceding evening, he found, is not predictive of headache in the next 24 hours.

However, he also examined headache occurrence when stress declines‑‑to test the “letdown” theory of migraine occurrence. Here, he found that a reduction in stress resulted in an odds ratio of about 4 for headache within six hours, meaning that recent reduction in stress was highly predictive of a headache attack. Theories for the causality of letdown headaches range from psychological factors to post-stress glucocorticoid withdrawal. In any case, they seem to be a real phenomenon and a target for well-timed intervention to stave off progression to headache, in addition to exploring ongoing stress management to avoid highs and lows.

In this same group of subjects, nearly half were also very good at headache self-prediction without any clearly identifiable premonitory factors. This group’s knowledge of an impending headache, sometimes hours before any clear symptoms emerge, also presents a clinically useful target for intervention, Lipton noted.

In addition to continuing research using electronic diaries to capture information for these individuals, he recommended imaging during the premonitory phase to identify the very earliest brain changes that may precede headache onset. The ongoing goal is to catch headaches before the earliest symptoms occur, both to alleviate suffering and to modify the natural history of this disease.