Richard Lipton, M.D., is the Edwin S. Lowe Professor and Vice Chair of Neurology, Professor of Epidemiology and Population Health and Professor of Psychiatry and Behavioral Sciences at the Albert Einstein College of Medicine, Bronx, US. Among his research interests is headache, where he focuses on the epidemiology of migraine and on clinical trials. His epidemiologic studies have evaluated trigger factors for headache attacks and risk factors for headache progression. Lipton recently spoke by telephone with Shana Burrowes, a graduate student and epidemiologist at the University of Maryland, Baltimore, US, to discuss what epidemiologists have learned about migraine, including the role of triggers of the condition. Below is an edited transcript of their conversation.
From the perspective of epidemiology—the study of the distribution and determinants of health and disease—what have been the biggest changes in the understanding of migraine since you became involved in this area?
One of the biggest changes is an increasing recognition of the global burden of migraine, which has come through the efforts of the World Health Organization, which estimates disability due to diseases on a worldwide basis. The metric used is called disability-adjusted life years, and papers published in the journal The Lancet over the last few months show that migraine is the world’s second most disabling disorder.
The reason that’s important is that the goal of treatment is to relieve pain and reduce disability. The therapeutic opportunities for migraine are really tremendous, particularly during the peak productive years when migraine is most common.
Another insight that emerges from epidemiology is that headache frequency within individuals fluctuates very markedly over time. In the field we distinguish between episodic and chronic migraine, and when you analyze clinical trial data, you look at group means [i.e., averages]. The group mean data looks very smooth, but if you plot individual trajectories, you observe that people have good weeks and bad weeks, good months and bad months—this is something that patients know. That variation is an important part of the disease burden.
What is the usefulness of categorizing patients as having episodic or chronic migraine? Do we need to include additional factors to better understand patients?
Episodic migraine, if it’s mild and responds well to triptans [a class of drugs used for migraine and some other conditions], or to non-steroidal anti-inflammatories [a different class of medications that includes aspirin, ibuprofen and naproxen], may be a medical inconvenience. But chronic migraine, when it’s frequent and disabling, can completely overwhelm the person’s life. Distinguishing the mild spectrum of migraine and the severe spectrum of migraine is really worthwhile.
We distinguish episodic and chronic migraine based on the number of headache days per month. People with chronic migraine have headaches on 15 or more days per month, while people with episodic migraine have less than 15 headache days per month. Patients don’t respect the bright lines we draw in the sand and move in and out of the categories we’ve defined. Some treatments, like onabotulinum toxin A (Botox), work for chronic migraine, but not for episodic migraine. In this sense the distinction matters in terms of treatment. Factors other than headache days per month may also predict treatment response.
People with episodic migraine sometimes develop chronic migraine, a phenomenon known as headache progression. Epidemiologic studies have identified a number of factors that predict progression. First of all, the more headaches you get, the more likely you are to have additional headaches. Further, allodynia, the experience of ordinarily non-painful stimuli as painful, is a risk factor for progression of migraine. Other risk factors include obesity, depression, anxiety, and medication overuse. There are certain respiratory disorders like asthma and rhinitis, and other pain disorders, that are also risk factors for progression.
We know a lot about prognostic factors in migraine and about the natural course of the disease, but it’s a challenge to figure out how to define subgroups of migraine patients that will give patients and the doctors who care for them optimal information.
What is the relationship between migraine and some of the other disorders you mentioned that migraine patients often suffer from?
While for some disorders we know the relationship is bidirectional, for others we don’t really know. For depression and anxiety, we know the relationship is bidirectional: If you follow people with depression over time, they develop migraine at increased rates, and if you follow people with migraine over time, they develop depression at increased rates. The same is true for anxiety, and for epilepsy.
We know that obesity is associated with migraine progression, but we don’t know if having migraine leads to obesity. I suspect that whether there is a bidirectional relationship or not depends on the particular disorder that migraine patients also suffer from.
Taking obesity as an example, we know that inflammation leads to pain, and we know that obesity is a pro-inflammatory state, so it seems plausible that obesity or asthma might lead to worsening of migraine.
For pain disorders, we know that the more pain you have, the more likely you are to have worse headaches in the short and long term. One possibility is that the pain systems in various regions of the body overlap, so it may be that pain begets pain through the mechanism of allodynia and central sensitization [an increased sensitivity to pain because of changes in the brain and spinal cord].
Many migraine research studies now utilize magnetic resonance imaging (MRI) and other brain imaging techniques. How is this helping to advance the understanding of migraine?
Using imaging is helpful and MRI shows, among other things, that pain has correlates in brain structure and function. In particular, there is research showing that when people have allodynia, they also have increased activation in the thalamus, which is the sensory relay station in the brain. That makes anatomic sense and may help explain the way pain disorders interact.
Why is it so hard to understand the importance of diet, sleep, and other so-called migraine triggers?
By definition, exposure to a migraine trigger factor increases the chances of getting a migraine over a brief period of time, often hours or days. Red wine may trigger one person’s headaches while chocolate triggers another person’s headaches. We don’t usually assess a trigger factor and then the onset of the headache within an individual. Most studies of triggers survey people with migraine and ask them about what triggers their headache. Using alcohol as an example, the triggers that people report are often written up as, “60% of people with migraine have alcohol as a trigger.” The problem is that these surveys of triggers assess people’s beliefs about triggers, and not whether the trigger is actually associated with an increased risk of migraine.
One way to study triggers is to do a randomized clinical trial where you expose individuals to a trigger, such as an artificial sweetener like NutraSweet, and compare them to people taking a placebo, and then see if headaches are more likely in the group exposed to the trigger. But there have been very few clinical trials like this.
The other approach, which is also not common, is to repeatedly measure both the trigger and the occurrence of headache in a large sample of people. Then you would ask if the trigger is associated with an increased risk of headache, over some relatively brief period of time.
The famous example of this is the association between the phenomenon in Canada called Chinook winds and migraine. There was a paper published in the journal Neurology where patients were asked to keep daily diaries. Researchers found that Chinook winds did trigger headache in some people, some people had Chinook winds as a trigger and didn’t know it, and other people had the belief that Chinook winds triggered their headache even though they didn’t. In Canada, the notion that Chinook winds trigger migraine is a very popular notion, so people’s beliefs about triggers are an important factor that complicates the understanding of triggers.
It’s a difficult area. I think triggers are important in some people, but it’s important not to mistake beliefs about triggers for demonstration that a trigger matters in a particular person. Of course, clinically it matters because if red wine triggers an individual’s headache, that person shouldn’t drink red wine, but if red wine doesn’t trigger your headaches, then there’s no reason not to drink red wine.
The use of headache diaries is very important in clinical and in research settings. Given that headache frequency fluctuates, and triggers can vary, what is an appropriate length of time to observe a migraine patient?
There are a couple of answers to that. First, studying triggers is not really driven by how long you observe, but by the number of headaches a person has and how frequently that person is exposed to the trigger. If you want to do an analysis of very common trigger factors, you probably need at least 10 or 15 headaches in each individual. If the trigger you’re interested in is the menstrual cycle, then you might need a year of data to do a careful analysis of the association of menstrual cycle with migraine within an individual. So how long an observation period you need depends on headache frequency, frequency of exposure to the trigger factor, and whether you’re asking the question for population mean data or for an individual.
Second, in terms of getting a baseline measurement for a study, if the trial is a chronic migraine study, participants need to have 15 or more headache days a month during the placebo run-in phase [the period before a clinical trial starts] to be eligible. Therefore, if you have 14 headaches in a particular month, you’re out of the study, which means that people who get into the study on average are having bad months. This means that their performance regresses to the mean [those with high headache frequencies tend to move towards average frequencies over time] , so if you’re enrolling people who are above a cut-off point, that actually enhances placebo rates in randomized trials, [which makes it difficult to disentangle the true contribution of the trigger to migraine].
Typically, when we are doing a preventive treatment trial, we do a four-week diary run-in phase and then a 12-week treatment phase. This may not be a long enough run-in period as the estimate of headache frequency during the treatment phase is more stable than the estimate of headache frequency during the run-in phase, and that could pose a problem.
Do migraine patients receive adequate care? Are there barriers to treatment?
Of people with migraine and unmet treatment needs, about 60% with episodic migraine have gone to the doctor for headache in the past year. Of those about 75% actually get a migraine diagnosis. Then, some fraction of people with episodic migraine get what I call minimally effective acute treatment, which I define as an NSAID or a triptan. But when you multiply out, did you go to the doctor, did you get a diagnosis, and did you get treatment, only 25% of people with episodic migraine and headache-related disability traverse all three of those barriers. One of the problems is that things get worse because you have to traverse multiple barriers to get minimally effective acute treatment.
If you look at chronic migraine, these people are less likely to have visited the doctor in the last year compared to those with episodic migraine. If you look at the chances of receiving a diagnosis like chronic migraine, only 40% of chronic migraine patients go to the doctor and a small fraction of those get a diagnosis. Among those who get a diagnosis of chronic or transformed migraine [which progressed from episodic migraine] or chronic daily headache, a small fraction gets preventive treatment. Again, there are barriers at multiple levels: seeking care, getting an appropriate diagnosis, and receiving appropriate treatment.
What other barriers are there?
Within the healthcare delivery system, there are many barriers. Everyone with chronic migraine should get a preventive treatment but only about 13% of people do. Preventive treatments are underutilized in part because doctors may not offer them and in part because many people with migraine deny that what they have is really a chronic disabling illness. Also, some physicians deliver the message that patients have nothing to worry about—that it’s just a migraine so if they take some ibuprofen or naproxen they’ll be just fine. So there are barriers at the level of the individual with denial of illness, chronicity, and the burden of migraine, and then there are barriers at the level of recognition by doctors that migraine disability is incredibly treatable and that they can do a tremendous amount to make the lives of their patients better, if they try.
Another major barrier to effective care is recognizing the patients who need non-oral acute treatment and giving them some options that will work while their digestive system is paralyzed. The digestive system is severely affected by migraine. If you take a pill and then vomit, it’s obvious the medicine wasn’t absorbed and it can’t help you, but more subtle is a phenomenon called gastric paresis, where the digestive system is paralyzed by migraine. A patient takes a pill, but it sits in the stomach until the attack is over. It’s absorbed when the digestive system starts working again, but it’s too late to be of any benefit. For people with prominent nausea and vomiting, or for people with gastric paresis, sometimes taking acute treatments, but not by mouth, works out much better.
What about non-drug interventions for migraine, such as cognitive behavioral therapy?
With non-pharmacological therapies, learning to identify and avoid triggers is important, particularly if stress is a trigger or if comorbid depression, or comorbid anxiety, is present. Cognitive behavioral therapy, biofeedback and progressive muscle relaxation are all evidence-based therapies for migraine, where it’s clear that they add value.
I direct the Montefiore Headache Center where we have two health psychologists on staff and we make extensive use of non-pharmacological treatments. The key thing is if you say to a patient, ‘I want to send you to a health psychologist,’ they may hear that as, ‘I think your problem is not a real medical problem, it’s all in your head—it’s a psychological problem.’
Of course, nothing could be further from the truth because biofeedback, cognitive behavioral therapy, progressive muscle relaxation, and trigger identification and management are treatments for, and not a denial of, the medical condition. The referral is evidence the medical condition is taken seriously.
What’s clear is that migraine treatment is not a one size fits all. Treatment needs to be individualized based on the frequency and severity of the patient’s attacks, their trigger factors, comorbidities and their preferences.
Recent research has shown that migraine patients may be at an increased risk of cardiovascular disease, and so triptans and NSAIDs should not be used, since these types of drugs have effects on the cardiovascular system. What do you think about this issue?
The U.S. Food and Drug Administration (FDA) recommends that triptans not be used in people who have active cardiovascular disease. I avoid triptans in those people, and then you have to find other medicines to use. NSAIDs also can exacerbate cardiovascular disease, but that’s much more with chronic use. I would certainly use NSAIDs, drugs that block the neurotransmitter dopamine, and muscle relaxants in people who have migraine and cardiovascular disease, though I would do so cautiously. There are also devices that can be used, such as those that provide transcranial magnetic stimulation [which uses magnetism to stimulate the brain], which are safe in the setting of cardiovascular disease.
What is the most exciting thing happening in migraine research right now?
There are a number of exciting things. One is that there are oral drugs in development for acute treatment that don’t have the cardiovascular risk of triptans.
One family of those treatments is oral drugs that block a molecule called calcitonin gene-related peptide [CGRP] receptor. These drugs offer real promise of migraine relief in people who either don’t respond to triptans or can’t take them.
Another exciting class of drugs in development is antibodies that target the CGRP system [see RELIEF related content]. There are four of those: one targets the CGRP receptor [the molecule to which CGRP binds], and the other three target CGRP itself by binding to it directly. All four of these drugs have been shown to be effective in the preventive treatment of both episodic and chronic migraine, and they all have very favorable side effect profiles so far.
There’s another class of acute treatments in development that is free of cardiovascular risk, and these target the serotonin system. There are also a number of treatments in development for people who have migraine and cardiovascular disease who can’t take triptans, and for people with migraine who don’t respond to triptans.
Prevention is getting better and safer. Acute treatment is also getting better and safer. This is a good time to be a doctor interested in migraine and a good time to be a patient who suffers from migraine because there are many new options on the horizon.
Homepage image credit: lightwise/123RF Stock Photo.