Migraine with aura and increased risk of ischaemic stroke

In the linked systematic review (doi:10.1136/bmj.b3914), Schürks and colleagues assess the association between migraine and cardiovascular disease, including stroke, myocardial infarction, and death as a result of cardiovascular disease.1

Migraine is a highly prevalent chronic condition characterised by a hyper-responsive nervous system that predisposes to recurrent episodes of severe headache and autonomic disturbance.2 Roughly a quarter of people who have migraine experience temporary neurological symptoms, known as aura, before some or all of their headaches. Aura is distinguished from other causes of brief recurrent neurological disturbance, such as transient ischaemic attacks, by its gradual onset and disappearance, the presence of both positive and negative features, and its duration of an hour or less.

Visual disturbances are the most common sort of aura, although sensory or motor auras can also occur. Patients with aura can be identified using the visual aura rating scale (table⇓), which assigns a point value to common aura features.3 Its developers make the clinically relevant observation that to identify patients with aura it is necessary only to detect visual aura, because 99% of people with non-visual auras also have visual aura occasionally. Sensitivity to light and visual blurring or fatigue are common accompaniments to migraine but should not be confused with aura.4

Doctors have long suspected a connection between migraine and vascular events such as stroke. A previous meta-analysis of the association between migraine and ischaemic stroke provided support for this link, but it could not clearly distinguish risks for migraine subgroups and did not examine the risk of other vascular events.5 The linked study by Schürks and colleagues incorporates results of several subsequent large, population based studies.1 Importantly, the results clearly show that the increased risk of ischaemic stroke is largely confined to migraineurs with aura, who have roughly double the risk of ischaemic stroke compared with people who do not have migraine. Migraine was also associated with an increased risk of transient ischaemic attacks and angina, but not haemorrhagic stroke.

Schürks and colleagues did not find a significant association between migraine and myocardial infarction or death from cardiovascular disease. Their findings should be interpreted cautiously, however, because the number of studies was too small to examine risks in subgroup. It is possible that when additional studies are available they will show that the size of the association between migraine and cardiac outcomes is different in people with and without aura.

The clinical implications of these findings are that patients who have migraine with aura should be followed closely and treated aggressively for modifiable cardiovascular risk factors. Counselling patients about their increased risk of stroke may increase adherence to recommended treatment, but the information should be put in context—the absolute risk of stroke for most patients with migraine is low, so a doubling of risk is not cause for panic. At a population level, however, this risk deserves attention because the prevalence of migraine is so high.

Schürks and colleagues found that the risk of ischaemic stroke in patients with migraine was magnified by the combination of smoking and oral contraceptive use, and that it was most apparent in women under 45. Previous work indicates that the association between migraine and ischaemic stroke in women directly correlates with the frequency but not severity of individual attacks.6 7

All of this suggests that clinicians need to identify young women with migraine, particularly those who are seeking oestrogen containing hormonal contraception. Those who have migraine without aura and are otherwise appropriate candidates for hormonal contraception should not be denied its benefits, because we have no convincing evidence of a clinically meaningful increased risk of stroke. For women who have aura, however, decisions about the use of hormonal contraception are more difficult.

Many experts recommend against the use of oestrogen containing contraception in women who have migraine with aura because they judge the risk of stroke to be unacceptably high.8 Certainly this seems true for women who have aura and additional stroke risk factors, such as smoking and hypertension. Blanket proscription is hard to justify, however, because combination oestrogen-progestin contraceptives are the most effective form of birth control. They have other benefits as well. These must be weighed against other harms, which include the relatively small increased risk of stroke in women with no risk factors beyond aura. Although the risk of stroke is probably modified by the frequency of aura, we do not have good information on how other aspects of aura affect risk. Women who have longer auras, more complex auras, or whose auras appear or worsen in the context of hormone use might be at higher risk for stroke than those with only occasional aura.

In the absence of clear evidence, clinical judgment and tailored decision making are important. In most cases, women with aura should consider their reasons for using oestrogen containing contraception. Persuasive reasons for use despite an increased risk of stroke might include the treatment of conditions such as endometriosis. Another compelling reason for use would be contraception for a woman who is not willing or able to use alternative methods and would not consider termination of pregnancy if contraception failed. The lowest tolerated oestrogen dose should be used, and treatment should be stopped if auras become more frequent or complex. The risk of venous thromboembolism varies with the type of progesterone. Risk is probably lowest for contraceptives with levonorgestrel and norethisterone.9