Epidemiology, Assessment, and Diagnosis of Migraine

Please subscribe to use our print features or to download PDF files.

DR. BUSE: It has been empirically proven that asking a patient about the impact of headaches on a patient’s life leads to improvement in the accuracy of diagnosis, better understanding of the true scope and burden of illness, development of more comprehensive and effective treatment plans, and better communication and rapport between the patient and HCP.2,3


Validated instruments include the Migraine Disability Assessment Scale (MIDAS) for headache-related disability,4 which is available for use free of charge; the Headache Impact Test 6 (HIT-6) for headache impact5; and several disease-specific and general instruments that measure health-related quality of life.6

Research has shown that simply asking open-ended questions such as “How do your headaches affect your life?” can provide a wealth of useful information; improve HCP-patient rapport; and, surprisingly, at the same time, slightly decrease the total time of a visit.5,7 The American Migraine Communications Study8 videotaped 60 patient-provider interactions during office visits. They found that an average migraine office visit lasted 12 min, and that HCPs asked an average of 13 closed-ended questions, of which more than 90% were closed-ended or answered with short responses.

Questions focused primarily on the frequency of attacks and severity of headache symptoms, triggers, and other features, but HCPs rarely inquired about headache-related disability or quality of life. There was a lack of agreement between patients and HCPs about the frequency and severity of headache when assessed during separate follow-up interviews. They also missed opportunities for creating effective treatment plans, including using preventive medications.

In response to the communication issues observed in the American Migraine Communication Study-I (AMCS-I), researchers developed and tested an educational intervention for improving communication during office visits. As part of the ACMS-II, they taught HCPs 2 communication strategies: use of open-ended questions and the “ask-tell-ask” strategy.9 They advised HCPs to use the “ask-tell-ask” strategy to clarify the number of headache days and distinguish the number of headache days from headache attacks by asking the patient “How many headache attacks have you had in the last month?” and “How many days did each attack last?” The patient would reply, at which point the HCP would assimilate and rephrase what was said back to the patient, for example “So that means you had X number of headache days in the past month,” He or she would then provide the patient with the opportunity to agree or clarify the statement. Then, the HCP would follow-up with one more opportunity to ask whether the patient had any more questions.

The “ask-tell-ask” is a simple and precise strategy for assessing the number of headache days in a month, which is needed to provide an accurate diagnosis of several conditions, including chronic migraine. However, it can be adapted for use in almost any type of medical communication.

DR. LIPTON: You also mentioned the importance of assessing psychiatric comorbidities. Why is this important and is there a specific approach you might recommend for that?

DR. BUSE: Assessing psychiatric comorbidities is important for several reasons. Once a diagnosis of migraine is established, an HCP should also have a heightened index of suspicion for psychiatric comorbidities,10 especially in the case of chronic migraine.11 Psychiatric comorbidities with migraine have been associated with poor treatment outcomes, treatment refractoriness, and problems with adherence. For example, patients with depressive symptoms demonstrate a poorer response to both pharmacologic and behavioral interventions.12,13 Psychiatric comorbidities are also related to poor quality of life14 and the potential for serious consequences including suicide.15 There is also evidence that several psychiatric conditions are risk factors for new-onset chronic migraine or transformation from episodic to chronic migraine.16,17

The existence of comorbidities has implications for treatment, both creating opportunities and imposing limitations. HCPs may want to consider opportunities for “therapeutic two-fers” or using treatments that will have benefit for both conditions when possible, although medications must be administered at therapeutic doses for both conditions. On the other hand, HCPs should avoid medications that may have adverse events that exacerbate or complicate the comorbidity.

There are several scientifically sound and easy-to-use instruments that can be incorporated into clinical practice to aid in the assessment and monitoring of psychiatric comorbidities.

The Patient Health Questionnaire (PHQ)18 is a paper-and-pencil screening tool for a range of psychiatric disorders and related common psychological issues. The PHQ-919 is a 9-item subset that screens for depression. The 7-Item Generalized Anxiety Disorder (GAD-7) assessment20 is a 7-item scale that screens for anxiety. The PHQ-421 is a very brief screening test for depression and anxiety. HCPs in a busy practice may want to use the PHQ-4 as a screening tool at the time of the initial or follow-up visit and then follow-up with one of the full-length instruments if the patient screens positive.

All of these instruments have strong psychometric properties, well-validated scoring systems, use Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) criteria22 for assigning diagnoses, and are in the public domain and available for use at no cost.

DR. LIPTON: Are there other comments on the assessment section, Drs. Silberstein or Kurth?

DR. SILBERSTEIN: I think general health issues such as smoking and hypertension and hormonal contraception in this population should also be assessed. Many of the patients we see are women of childbearing potential. Many are often taking hormonal contraception, which can be a risk factor for complications of migraine, but also a benefit. I think we need to ascertain the use of hormonal contraception for that reason and especially if we’re going to put a woman on regular medications so that we don’t prescribe drugs that may have the potential to cause abnormalities in the fetus.

It’s important to ascertain the smoking status as well, which can be a risk factor for aggravation of headache and a contraindication of hormonal contraception. We also look for hypertension and other general medical problems that may be a contraindication to the drugs we use to treat migraine, both from an acute and a preventive point of view.

DR. BUSE: In addition, healthy lifestyle habits are important in achieving the best outcomes. I recommend asking about sleep habits and problems, exercise and activity, nutrition, smoking habit, and alcohol or other substance use as Dr. Silberstein recommended. Problems in any of these areas may warrant referral to a behavioral or mental health provider, physical or occupational therapist, nutritionist, or other allied healthcare provider.

DR. LIPTON: Thank you, Dr. Silberstein and Dr. Buse. Let’s move on to the epidemiology. Dr. Kurth, how common is migraine in its various forms?

DR. KURTH: Well, migraine is by far the most common neurological disorder in the general population. It’s very clear that women are more affected than men—nearly 3 to 4 times more. We know that migraine prevalence is age-dependent, and we have a peak 1-year prevalence between the ages of 25 to approximately 55 years, with a decreased prevalence thereafter but not complete cure in the elderly.23

Overall, approximately 10% to 20% of the general population has active migraine.24 Dr. Lipton has extensively researched this, reporting that approximately 18% of women and 6% of men have a history of migraine.23 The question of how common the various subforms are, specifically migraine with aura and migraine without aura, is quite difficult to answer in population-based studies because aura is challenging to ascertain even in a clinical setting. When we go into population studies and try to determine specific neurological features, such as visual disturbances, it’s very difficult to get the right figures.

If you look at the population level, migraine with aura ranges from approximately 10% up to 30% and sometimes even higher.25 It’s likely that the ascertainment tools are different. Maybe the right number is between these 2 figures, approximately 20%.

With regard to chronic migraine in the population, most studies now indicate that chronic migraine has a prevalence of approximately 1% to 2%,26 which is a very high value because chronic migraine is certainly a condition that has the highest burden for the patient and society.

DR. LIPTON: There’s a distinction between chronic daily headache and chronic migraine. My understanding of the literature is that chronic daily headache, primary headache for 15 or more days per month, has a prevalence of 3% to 5%. A review of studies estimating chronic migraine using several definitions in countries worldwide reported most prevalence rates in the range of 1% to 3%.26

DR. KURTH: That’s correct. A recent examination of the US population reported an overall rate of just below 1%, with higher rates in women in midlife.27 Another issue is migraine in the elderly. Although we believe that migraines are less common in the elderly, an increasing number of studies are showing that approximately 3% to 5% of the elderly (>65 years) in the population still report migraine,28 which is still a high number.

DR. LIPTON: Dr. Buse, how do you understand the relationship between episodic migraine and chronic migraine?

DR. BUSE: The relationship between episodic migraine and chronic migraine is fluid and bidirectional, and individuals may move in either direction during the course of their lifetime. Moving from episodic to chronic migraine has been referred to as “transformation,” “chronification,” or “progression.” An analysis of data from the American Migraine Prevalence and Prevention (AMPP) study showed that among individuals with episodic migraine in the general US population, an average of 2.5% developed chronic migraine in a subsequent year.29 This rate is probably higher in clinic-based populations. We know there are several risk factors for new onset of chronic migraine, which can be divided into nonmodifiable and potentially modifiable risk factors. Modifiable risk factors and potentially modifiable risk factors are important to identify because they may provide targets for intervention and treatment.30

Nonmodifiable risk factors include the female sex, a lower education level, low socioeconomic status, history of traumatic brain injury, and genetic factors. Potentially modifiable factors include headache attack frequency, certain classes of medication overuse, caffeine use, obesity, snoring, stressful life events, depression, anxiety, and adverse childhood experiences.

DR. LIPTON: Dr. Kurth, what is the relationship between migraine and stroke?

DR. KURTH: Well, the relationship between migraine and stroke was first described long ago and for approximately 40 years, the relationship has been confirmed through numerous population-based and clinic-based studies. We have to understand, however, that there are certain subgroups that are at a higher risk than other groups. Overall, there’s an approximately twofold increase in the risk of stroke from migraine, which is only observable in patients with migraine with aura in most studies.31 There is no increased risk of stroke with migraine without aura.31

In subgroups, the increased risk seems to be particularly evident in younger women. Smoking and oral contraceptives seem to play an important role. If a young woman with migraine with aura smokes and uses an oral contraceptive, the risk of stroke increases up to approximately 10-fold or even more in some studies.31

We have to understand, however, that although we’re talking about doubling the relative risk, the absolute risk for an individual is still low. It was estimated that approximately 4 additional ischemic stroke events would occur annually per 10,000 women with migraine with aura,32 which is still a very small number. Despite the publicity about this topic, we have to be realistic about the fact that we cannot predict which patient with migraine with aura will eventually have a stroke.

Some evidence suggests that migraine is also a risk factor for hemorrhagic stroke, but the evidence is scanty as compared to that for ischemic stroke because ischemic stroke is a much more common event.