Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.315974
Source of Support: None, Conflict of Interest: None
Keywords: Aura, migraine, stroke, vascular
Migraine is a common disabling neurologic illness, and affects approximately 20% of the world's population, with the highest prevalence in women of reproductive age group. As per the “Global Burden of Diseases Study” migraine is amongst the leading causes of disability worldwide in individuals less than 50 years of age. A growing body of evidence in literature identifies migraine as a risk factor for stroke and CVD. The association is strongest between migraine with aura (MA) (a third of migraineurs may have an aura associated with headache) and ischemic stroke. Data about hemorrhagic stroke, myocardial infarction, and cardiovascular mortality is less clear. There is a general agreement in the literature that women with migraine are at a higher risk of vascular disease as compared to men. This may be because migraine is more common in women or maybe due to changes in endogenous and exogenous hormone levels. A comparative lack of data for men is an important confounding factor. In this review article, we give an update on the association between some common vascular diseases in persons suffering from migraine and discuss the clinical implications.
Migraine and stroke
Migraine and ischemic stroke
The association between migraine and ischemic stroke has both explicit and fuzzy aspects. The association is particularly stronger for MA, female gender, young age (<45 years old), and use of oral contraceptives and smoking. Many studies including five meta-analyses of observational studies support this contention.[4–8] In these meta-analyses, the relative risk of ischemic stroke in migraine has been reported between 1.73 (1.31, 2.29) to 2.16 (1.89, 2.48) for overall migraine, 1.56 (1.30, 1.87) to 2.41 (1.81, 3.43) for MA, and 1.02 (0.68, 1.51) to 1.83 (1.06, 3.15) for migraine without aura. In general, the risk of stroke in MA is nearly twice as that in people without migraine. The risk is 10 fold in young women who have MA and are current cigarette smokers and oral contraceptive users. The risk is also higher in people with active migraine and frequent attacks., The 'Stroke Prevention in Young Women Study' found that the stroke risk was higher in women with more than 12 attacks per year. The severity of attacks and the longer duration of disease does not seem to play a significant role. The risk in men and in the elderly is more uncertain, except for elderly who smoke, or elderly with the onset of MA in later life (after 50 years of age).[12–14] The functional outcomes are reported to be good., MA and comorbid cardiovascular risk factors are also associated with increased risk of perioperative ischemic stroke. As far as ischemic stroke subtypes are concerned, the 'Atherosclerosis Risk in Communities study' (involving 1622 migraineurs from a community-based cohort of 12,758 participants) found an association between migraine with visual aura in late middle age and increased risk of cardioembolic stroke.
Migraine and transient ischemic attack (TIA)
MA is linked to an increased risk of TIA, although the association is confounded by a possibility of misdiagnosis as migraine aura and TIA are many a time difficult to differentiate. Particularly challenging are case scenarios wherein the aura is atypical, occurs late in life (when the risk of TIA also increases), occurs in a patient with vascular risk factors, and is not associated with headache. Also, a TIA (or any cerebrovascular event) is known to trigger migraine-like headaches, leading to misinterpretation of stroke as 'complicated migraine' and undermining the use of 'associated headache' in the differentiation. Clinically the most relevant points for the distinction between the two entities include onset, progression, and phenomenology. Sudden onset, lack of march/progression (or if at all it progresses the nadir is reached in seconds to a minute), and negative symptomatology, would favor a TIA over migraine aura. Nevertheless, if the aura is different from the previous attacks, atypical, or involves the basilar artery, it would be prudent to do a tissue imaging and a detailed clinical workup for TIA. Arterial spin labeling is a new technique that can be used to differentiate aura from TIA. The hypoperfusion in migraine aura overlaps with adjacent vascular areas and does not follow the strict distribution of cerebral vascular territories. These perfusion changes are reversible.
Migraine and hemorrhagic stroke
The association between migraine and hemorrhagic stroke remains uncertain. Sacco et al. performed a systematic review and meta-analysis (eight studies comprising of four case-control and four cohort studies, involving 320539 individuals with 1600 hemorrhagic strokes) on the association between migraine and hemorrhagic stroke. They reported an overall pooled effect size of 1.48 (95% CI, 1.16, 1.88) for hemorrhagic stroke in participants with any migraine. However, unlike in ischemic stroke, subgroup analysis did not identify increased predisposition in the sub-groups tested, except for higher effect size in women and younger patients (<45 years). Few studies have shown a trend towards an association between hemorrhagic stroke and MA but the conclusions are not firm. Migraine is also associated with hemorrhagic stroke during pregnancy. As per a large population-based study on inpatients in the United States, migraine in the peripartum period was associated with hemorrhagic stroke with an odds ratio of 9.1 (95% CI, 3.0, 27.8). In contrast to the outcome of ischemic stroke in migraine, hemorrhagic strokes are associated with poor outcomes.
Mechanisms of migraine and stroke
Several plausible hypotheses have been contemplated linking stroke with migraine [Table 1], and are discussed briefly.
Cortical spreading depression: Cortical spreading depression is a spontaneous wave of depolarization that spreads across the cortex at a rate of 3–5 mm/min. The depolarization is associated with increased glucose expenditure and activation of metalloproteinases, thus increasing the blood-brain barrier permeability and local hyperemia. This hyperemia is followed by vascular unresponsiveness and prolonged oligemia. It is proposed as a basis of migraine aura and also predisposes the vulnerable brain tissue to ischemic insult.
Endothelial dysfunction: Endothelial dysfunction is thought to affect cerebrovascular reactivity, although studies in this area in migraine depict conflicting results. Migraine particularly with aura is thought to be associated with an inflammatory milieu. Cytokines released during migraine aura contribute to the prothrombotic state. People with migraine and MA harbor biomarkers of inflammation and endothelial injury like C reactive protein, calcitonin gene-related peptide, vascular endothelial growth factor, von Willebrand factor, stable nitric oxide metabolites and, have fewer endothelial progenitor cell levels.,
Traditional vascular risk factors: Regarding the role of traditional vascular risk factors, studies have shown an increased association of hypertension, obesity, smoking, and dyslipidemia with migraine, but the results are highly variable.[28–30] Obesity is also thought to affect migraine frequency, severity, duration, and chronification, while smoking increases the risk of stroke in women with MA. Many studies have assessed the burden of traditional risk factors in people with migraine. But the results are extremely variable, probably because of different study designs, selection and sampling criteria, method of capturing the risk factors (questionnaires, medical records, in-person visits), changing definitions, and residual confounding factors. This is particularly true for the association between migraine and hypertension where the data are quite heterogeneous refuting conclusive statements. For migraine and diabetes, there seems to be a negative association, while for factors like obesity, smoking, and family history of CVDs, the literature supports a positive relationship. Migraine also has an association with an unfavorable lipid profile, but the clinical importance of this relationship is not established. Some studies report reduced alcohol consumption in migraine which may be a disadvantage due to the cardioprotective effect of moderate alcohol consumption. There are reports of an increased carotid intima-media thickness in people with migraine, but the association is not conclusive.,
How do these traditional risk factors play between the association of migraine and CVDs? Most studies have shown that the connection between migraine and CVDs, particularly stroke, is independent of the usual vascular risk factors. Most of the strokes in migraineurs are not atherothrombotic. In fact, the stroke risk in migraineurs is increased the most in those who have a favorable cardiovascular risk profile except for cigarette smoking which is known to increase stroke risk in these patients. In the WHS, the risk of stroke in MA was proven only in women with the lowest Framingham risk score. These findings suggest that the migraine aura-ischemic stroke connection is probably non-atherosclerotic. On the contrary myocardial infarction (as mentioned above), was seen in the highest Framingham risk score group.
Patent foramen ovale: Patent foramen ovale (PFO) is thought to be mediate cryptogenic stroke and migraine aura through 'paradoxical embolization.' Numerous but not all studies have reported a high prevalence of PFO in people with migraine, especially MA., Studies have also reported a PFO in approximately 65% of patients with migrainous infarction and a lower mean age of stroke in these patients. Although observational studies have reported a beneficial effect of PFO closure on migraine occurrence, the evidence and the risk-benefit ratio are not enough to recommend routine screening and closure of PFO in migraine. PFO should be screened for if there is a cryptogenic stroke in the young or unusual and frequent aura with migraine attacks. Factors like the size of the PFO, right to left shunt at rest, and the presence of an atrial septal aneurysm, should be considered for the decision of closure.
Genetics: Certain genetic disorders like cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL), familial hemiplegic migraine (FHM), mitochondrial encephalopathy with lactic acidosis and stroke (MELAS), hereditary endotheliopathy with retinopathy, nephropathy and stroke (HERNS), and cerebroretinal vasculopathy connect migraine and stroke or stroke-like episodes. CADASIL predisposes to stroke through a non-arteriosclerotic, non-amyloid, fibrotic vasculopathy (NOTCH 3 gene mutation on chromosome 19p13 codes for smooth muscle cell receptors on blood vessels). FHM increases susceptibility to a cortical spreading depression. Genome-wide analysis studies also report a shared genetic susceptibility to migraine and stroke., The association is stronger with migraine without aura than MA, and large artery and cardioembolic stroke than small vessel disease.
Oral contraceptive: The use of oral contraceptive pills has been implicated as a risk factor for vascular events in migraine. Estrogens can impact the association between migraine and vascular disease. Endogenous estrogens are known to protect against vascular damage via anti-inflammatory and immunomodulatory activity in women. Exogenous estrogens differ from endogenous estrogens in that they increase C reactive protein in women of reproductive age, increase the production of pro-coagulation factors, reduce the levels of anticoagulant factors, and affect the hepatic synthesis of lipoproteins and triglycerides., Estrogens are also thought to mediate an increase in platelet reactivity and adhesion ability. Studies in the past, including a meta-analysis, have shown a higher risk of ischemic stroke (approximately seven times higher risk) in women with migraine using oral contraceptives. The risk directly correlates with the dose of ethinylestradiol contained in the formulation. 50 micrograms or more of ethinylestradiol has a higher risk of ischemic stroke as compared to low-dose formulations. As per current evidence, low-dose oral contraceptives containing 25 micrograms of ethinylestradiol or less do not increase the vascular risk in women with migraine. A study comparing 20 micrograms EE to three different formulations (30 micrograms EE + drospirenone, vaginal ring, and patch) reported an increased risk of ischemic stroke and myocardial infarction with 30 micrograms oral preparation only, more so in women aged 35 years or older. Few studies have assessed the risk differences of hormonal contraceptive use between migraine with and without aura. In this context, the 'Stroke Prevention in Young Women study' reported an increased risk of ischemic stroke in those who had MA and used hormonal contraceptives along with smoking, but no increase in risk with the use of hormonal contraceptives alone. Another recent study found a six-fold higher risk of stroke in women who had MA and used oral contraceptives, in comparison to women without these risk factors. In women, without aura, the stroke risk was high but unchanged with the use of oral contraceptives. In a systematic review on stroke risk and use of estrogen-containing contraceptives in women with a migraine the odds of stroke ranged from 2.08 to 16.9. The authors recognized that the quality of evidence is low due to small studies and wide confidence intervals. As per existing guidelines, hormonal contraceptives are contraindicated in women with MA.[47–49] WHO also restricts the use to women younger than 35 years of age.
Hormone replacement therapy: The association of vascular risk with hormone replacement therapy (HRT) in patients with migraine lacks clarity due to conflicting data. Both high dose and low-dose formulations have been linked to increased risk in different studies. There are no formal trials comparing different formulations but local HRT, low-dose estrogens, and natural estrogens may be allowed keeping in mind some increased risk of thromboembolism. However, if in a patient on HRT, migraine recurs or worsens HRT should be discontinued. And if the patient develops new-onset headache and aura, a thorough investigation on lines of transient ischemic attack should be done. Also, in postmenopausal women who already have subclinical vascular disease, estrogens are thought to be detrimental to vascular endothelium. If given for a prolonged period in late menopause, even low-dose estrogens associate with a pro-inflammatory state. Thus the age, duration of treatment, and time point of starting HRT in relation to the menopause onset are all important considerations while deciphering the vascular risk with HRT in migraine. Starting HRT late in menopause and prolonged use in an older woman with vascular risk factors would pose a greater risk of vascular disease than a prompt initiation of treatment in the perimenopausal period or young women with atherosclerosis and a relatively shorter administration.
Drugs: Ergot alkaloids and triptans can cause vasoconstriction and increase the risk of ischemic vascular events in a patient with migraine who has comorbid cardio/cerebrovascular disease or uncontrolled vascular risk factors. Prior history of migraine is frequently reported in the reversible vasoconstriction syndrome (RCVS) which is characterized by acute severe headache with or without seizures and transient/persistent focal neurological deficits, multifocal constriction of cerebral arteries, spontaneous resolution within three months). Triptans and ergots are known to precipitate RCVS. Active migraine during pregnancy is another potential predisposing factor for stroke. Women with migraine and stroke risk factors should be considered for high-risk obstetric monitoring.
The absolute risk of stroke with acute migraine therapies (triptans) is estimated to be small (1:100,000). A population-based cohort study (25-80 years of age) from Denmark (2003–2011) reported a borderline higher relative risk for stroke among migraineurs (identified by triptan utilization) in respect to all strokes in the general population (RR 1.07, 95% CI: 1.01–1.14). However, a higher risk of hemorrhagic stroke was observed in women, particularly those aged 25-45 (RR ~ 1.7). As triptan utilization was used as a proxy measure to identify migraine, the risk differences between triptan users and other migraineurs, between migraine with and without aura, are not available from this study. Ergotamine overuse increases the risk of CVD and cerebral ischemia, particularly in patients with established CVD. A retrospective nested case-control study reported an odds ratio of 2.55 (95% CI 1.22, 5.36) for ischemic complications with ergotamine overuse. Overuse of triptans with or without the use of cardiovascular drugs did not increase this risk. The cardiovascular safety of triptans in established CVD is not an issue as per evidence, but as per the label they are contraindicated in patients with stroke, TIA, unstable angina, acute coronary syndrome, and uncontrolled vascular risk factors. Also, as already mentioned triptans and ergotamines can induce vasoconstriction and potentially precipitate RCVS. People with RCVS frequently have a history of migraine. The thunderclap headache of RCVS may be erroneously thought to be a 'bad migraine attack' if the history of time taken for the attack to peak is not properly elicited. Triptans and ergots given in such a setting may aggravate the vasoconstriction. Among the migraine preventives, the recently developed calcitonin gene-related peptide (CGRP) antagonists have been postulated to carry a theoretical risk of impaired compensatory vasodilation during ischemic conditions (CGRP mediate vasodilation). A study compiled data on vascular adverse events from four double-blind, placebo-controlled studies of Erenumab in migraine (chronic or episodic). The incidence was similar across the placebo and Erenumab treatment groups over a period of 12 weeks.
'Migrainous infarction' versus 'migraine with stroke'
From a clinical perspective, ischemic stroke in a patient with migraine can be categorized into one of the following three categories: cerebral infarction of other cause coexisting with migraine, cerebral infarction of other cause presenting with symptoms similar to MA (ischemic cerebral events can trigger migraine-like attacks) or 'migrainous infarction'. Of these, migrainous infarction is defined [by International Classification of Headache Disorders (ICHD)-3] as an attack of MA, where one or more aura symptoms last more than 60 minutes (aura is prolonged, but symptoms are typical of previous attacks for that particular patient), neuroimaging reveals an ischemic brain lesion in the relevant territory and the stroke cannot be attributed to another disorder. It is thus a diagnosis of exclusion. It is uncommon, seen in 0.2%–0.5% of all ischemic strokes and most often involves the posterior circulation. It is postulated that the blood flow during the prolonged aura phase touches ischemic levels.
Migraine, white matter lesions, and cognition
Migraine is also associated with multiple T2 -FLAIR periventricular and deep white matter hyperintensities (4%-59% of migraine population). A meta-analysis of six population-based and 13 clinic-based studies revealed that white matter abnormalities, silent infarct-like lesions and volumetric changes in grey, and white matter regions were more common in participants with migraine than in control groups. The association was strongest for MA (OR 1.68, 95% CI: 1.07, 2.65). The odds of these white matter lesions have been reported to increase with the increasing frequency of migraine (OR 2.6, 95% CI: 1.2, 6.0). Traditional vascular risk factors and migraine-specific drugs do not mediate this association. The clinical significance of these lesions is not clear. A population-based study on 1170 participants (938 – no severe headache, 65 – nonmigraine headache, 167 - migraine out of which 24 had MA), with a mean age of 69 years, did not find a faster decline in cognition over four to five years of follow-up. Longitudinal studies are required to assess the structural progression and the corresponding clinical changes. Systematic reviews on the association between migraine and cognitive decline, in general, have presented heterogeneous results. Studies both support and refute an association between migraine and vascular dementia. There is no role of stroke prophylaxis in patients with white matter hyperintensities or silent infarcts.
Migraine and cardiac disorders
Evidence also suggests that migraine is associated with common cardiovascular risk factors like hypertension, Raynaud's syndrome, ischemic heart disease, and myocardial infarction. As per a large prospective population-based study (Nurse's Health Study II: 115541 women, aged 25–42 years at baseline, free of angina and CVD (CVD) initially, 20 years follow-up) the hazard of myocardial infarction, angina/coronary revascularization, and cardiovascular mortality amounts to a hazard ratio of 1.39, 1.73, and 1.37, respectively. The Women's Health Study (WHS), another large prospective cohort study (27519 women, aged greater than equal to 45 years at study entry, free from CVD at baseline, 11.9 years follow-up) reported increased incident myocardial infarction in patients who had MA and an adverse vascular risk profile (based on Framingham risk score). The Danish nationwide population-based cohort study (51032 patients with migraine, 510320 people from the general population, 19 years follow-up) reported a positive association between migraine and myocardial infarction (adjusted HR 1.49, 95% CI 1.36 to 1.64) and atrial fibrillation/atrial flutter (adjusted HR 1.25, 95% CI 1.16 to 1.36). Autonomic dysfunction may play a role in the pathophysiology of both migraine and AF, and this may predispose older migraineurs to cardioembolic stroke. MA is also associated with an increased risk of venous thromboembolism in younger patients (age <55 years, adjusted HR 3.32, 95% CI, 1.51, 7.31). The association is stronger in people with aura. At present, there is no evidence to say that migraine prevention reduces cardiovascular risk or, to institute primary prevention with antiplatelet or antithrombotics in patients with migraine.
Fallacies behind the association
Despite the recognized associations between vascular disease and migraine discussed above [summarized in [Table 2]], there are caveats that one needs to consider and contemplate upon.
The prevalence of the vascular disease is low in young individuals. Indeed, most of the studies investigating the association between migraine and vascular events are retrospective. Few prospective studies are available and some do not differentiate between migraine subtypes. The methodologies used to capture the migraine diagnosis and type (with or without aura) are neither robust nor consistent across different studies. For example, in the Women's Health Study, patients having active migraine were classified into active MA and active migraine without aura by asking if they had an “aura or any indication a migraine is coming”. The possibility that premonitory symptoms are classified as aura by the patient may lead to wrong subtyping. Also, most prospective cohort studies derive the associations using a 'baseline' history of migraine and a 'later' occurrence of stroke/vascular event. Immediate temporal correlation is not available. 'Headache with stroke' may be a manifestation of stroke (for example, hemorrhagic stroke), and stroke may trigger migraine further confounding the associations.
Summary and clinical implications
To summarize, evidence links migraine to an increased risk of CVDs (CVD) and stroke, but the overall absolute risk is small. Are there reliable indicators as to which individuals across the migraineurs are most susceptible? As of now, this information is limited. In general, most of the migraine patients do not have an aura, and migraine without aura has a smaller risk, which is reassuring. MA particularly portends a higher risk of ischemic stroke especially in younger people (<45 years age), female gender, and with the use of oral contraceptives and smoking. Late-onset of migraine aura (after 50 years of age), active migraine, and frequent attacks also increase the risk of ischemic stroke. It is important to differentiate between a migraine aura and TIA (onset, duration, progression, and phenomenology hold the key). If a patient with aura presents with atypical symptoms, symptoms different than his usual aura, symptoms, and signs suggestive of posterior circulation involvement, or increased frequency of aura in later life, magnetic resonance imaging of the brain and detailed vascular workup should be done. ASL perfusion imaging, a contrast-free method may be suitable to assess the reversible perfusion changes during MA and differentiation from TIA. Migrainous infarction should always be considered as a diagnosis of exclusion after a thorough workup for other causes. The association between migraine and hemorrhagic stroke is less clear, younger women (< 45 years age) and the peripartum period is more vulnerable. The association with aura is not clearly established.
The increased risk of stroke in these individuals does not seem to associate with traditional vascular risk factors other than smoking, although a plethora of studies reports an increased prevalence of an adverse vascular risk factor profile in migraine (smoking, hypertension, higher body mass index, dyslipidemia, and family history of CVD). Migraine-specific pathophysiology might play a role. On the contrary, myocardial infarction is more likely to occur in MA and a comorbid bad vascular risk status. It is imperative therefore in daily clinical practice to treat the vascular comorbidities if any and to encourage smoking cessation. At present, there are no recommendations to start drugs like aspirin and anti-thrombotic for the prevention of vascular events in these patients, unless clear nonmigraine indications are present. The role of migraine preventives in reducing future CVD/strokes is not yet established. Drugs like angiotensin-converting enzyme inhibitors and angiotensin receptor-blockers have shown an initial promise as migraine prevention therapies and are known to have a cardiovascular benefit independent of blood pressure reduction. Whether they can modify the cardiovascular risk in migraine, if used as a preventive drug, is yet to be seen.
Ergot alkaloids and triptans can be safely used in recommended dosages in patients who do not have a cardio or cerebrovascular disease or uncontrolled vascular risk factors. Oral hormonal contraceptives should be avoided in women who have MA, are above 35 years of age, smoke, or have an adverse vascular risk profile or pre-existing thrombophilia. Low estrogen-based oral contraceptives may be used in younger women who have migraine without aura. In the peri-menopausal period, hormone replacement therapy if required should be instituted early. Local formulations, low-dose estrogens and natural estrogens have a lower risk of thromboembolism. Long-duration HRT started in the late post-menopause period can be detrimental to vascular health in older women with migraine. The presence of white matter hyperintensities on brain imaging should not prompt starting antiplatelets or anti-thrombotics, in the absence of other indications for these drugs. The clinical relevance of these lesions needs further evaluation in prospective long-duration cohorts.
To conclude, migraine particularly MA is associated with increased risk of stroke and CVD, the strongest association being with ischemic stroke. The specific pathophysiological mechanism is unclear. Aura-associated circulating coagulation markers, systemic endothelial dysfunction, comorbid conditions (like adverse vascular risk factors, PFO, cervical artery dissection, RCVS), exogenous estrogen acting on a diseased endothelium, and acute migraine drugs may play a role. We need to elucidate the specific upstream/downstream pathophysiological pathways that can explain the 'migraine-vascular risk' connection. Future studies should also explore clinical, imaging, and blood biomarkers that can predict an adverse vascular outcome in migraine. Targeted intervention trials on such enriched cohorts may provide useful future strategies to prevent stroke and CVD in people with migraine.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Table 1], [Table 2]