Migraine and ischaemic stroke

Migraine and ischaemic stroke are both common conditions. It is hardly surprising, therefore, that the two can coexist in the same patient—but does a causal relation exist? The meta-analysis by Etminan et al in this issue provides further evidence that it does.¹ Previous work showed no increase in the risk of haemorrhagic stroke in people with migraine.²

The increased relative risks reported by Etminan et al were 1.8 in migraine without aura, 2.3 in migraine with aura, and 8.7 in women with migraine who are taking the oral contraceptive pill. These figures may not be accurate because their meta-analysis has several problems: similar populations were not included; case-control studies are vulnerable to recall bias; control for confounding risk factors such as family history, smoking, diabetes, hypertension, and common treatments may not have been uniform; high risk patients with migrainous symptoms due to other conditions were not excluded; uncertainty existed in some studies about the diagnosis of migraine and even of stroke (transient ischaemic attacks were sometimes included); no distinction is made between users of oral contraceptive pills containing high doses of oestrogen and those containing low doses or only progesterone. This may be crucial because risks with the latter are lower or even non-existent.^3-5 Most importantly they do not give information on the influence of age.

Age is the most important risk factor for stroke. In young people the absolute risk of stroke is very low, and so the increased relative risks reported here, even in women taking oral contraceptive pills, need not be alarming. What does the increased risk of stroke mean for older people with migraine? Confirmation of another suspected risk factor could perhaps be taken as good news because it offers the hope of prevention. The risk of stroke in older people is due to the influence of several risk factors and may exceed the sum of their individual relative risks. Older people with migraine should therefore perhaps be assessed more carefully.

Migraine is not one condition. Worrying subtypes include migraine with aura, hemiplegic migraine, and migraine recurring in elderly people after years of freedom or occurring for the first time. Some medical conditions associated with an increase in the risk of stroke can also cause migraine—for example, cerebral autosomal dominant arteriopathy with subcortical infarcts and leucoaraiosis (CADASIL) and the antiphospholipid syndrome. Others such as acute dissection of the carotid or vertebral artery, subarachnoid haemorrhage, cranial arteritis, and occasionally cerebral tumours may produce migrainous symptoms. However, these are in the minority, and the increased risks reported by Etminan et al are unlikely to be due solely to them.

Similarly, ischaemic stroke is not a single disease. Small or large vessels may be affected, with local thrombosis or thromboembolism from artery to artery or cardiac embolism. In migraine, cerebral blood flow has been shown to be reduced in certain regions and platelet activity is thought to be increased⁶,⁷—two factors that might contribute to the risk of thrombosis. The possibility that treatments used in acute migraine might affect the risk of stroke needs consideration. Using vasoconstrictors such as ergotamine and the triptans in the presence of focal ischaemia is a theoretical concern. However, patients feel better with them, and we have no evidence that such treatment increases the risk of infarction.⁸ Furthermore, one would expect that some agents used in the prophylaxis of migraine—for example, β blockers and aspirin—should reduce the risk of stroke.

Most primary care doctors will not see a neurological deficit persisting after migraine. Even in neurological practice this is rare and then usually causes a visual field defect.⁹ In some instances, the ischaemic event is not related to an obvious acute attack of migraine with cephalgia.² In such a situation, another condition may possibly cause both—the ischaemic event and migraine.

In the absence of robust evidence of what might help, what advice may be appropriate? In an otherwise healthy young person, no cause for concern exists because of the very low absolute risk of stroke. Advice to stop smoking seems prudent, as does prescribing oral contraceptive pills containing low dose oestrogen or only progesterone to young women with migraine. Anecdotal evidence has prompted most neurologists to advise stopping oral contraceptive pills if migraine becomes more frequent or changes in character, with a more prominent or prolonged aura, although some evidence indicates that these worries are not necessarily real.² If a middle aged woman with migraine continues to smoke or if other risk factors exist, coming off the oral contraceptive pills should be considered. Although discontinuation seems prudent, we have no evidence that this will reduce the risk. And as migraine is usually under-reported,¹⁰ doctors, when they are prescribing oral contraceptive pills, should ask these women if they have migraine and see if any change in its frequency or character occurs.