Abstract
Introduction
The higher-than-expected incidence of vascular diseases reported in migraineurs suggests that migraine may, in some cases, be a dangerous condition rather than just a distressing but harmless disorder. We provide a systematic review of data linking migraine to vascular diseases.
Migraine and vascular diseases
Available data indicate an increased risk of ischemic stroke in subjects suffering from migraine with aura. In addition, evidence suggests an association between migraine with aura and cardiac disease, intracerebral hemorrhage, retinal vasculopathy and mortality that needs to be further corroborated; consequently, for those conditions, migraine with aura can be only considered among the less-well-documented risk factors. As the absolute risk of ischemic stroke in the overall migraineur population is low, subjects suffering from migraine with aura should be made aware of the possible link but not unduly alarmed. It is a common notion that the vascular risk of migraineurs may be further increased by the presence of easily treatable vascular risk factors such as arterial hypertension, cigarette smoking and oral contraceptive use.
Conclusions
Forthcoming guidelines should appropriately recommend supporting migraineurs not only with measures aimed at decreasing headache frequency, thus improving quality of life, but also with general measures and preventive strategies aimed to reduce the overall vascular risk. In fact, headache specialists should take care not only of relieving pain but also of assessing and treating concurrent vascular risk factors, while gynecologists, in particular, should routinely consider the presence and type of migraine before prescribing oral contraceptives.
Keywords
Introduction
Migraine is one of the most common primary headaches, having a one-year prevalence period of about 12% (1). The prevalence is three times higher in women than in men and peaks in the age group 30–39 years (1). Migraine with aura (MwA) accounts for one-third of all migraines (2). A higher-than-expected incidence of vascular diseases (VDs) in migraineurs (3–7) suggests that migraine may be, in some cases, a dangerous condition rather than just a distressing but harmless disorder.
MwA can rarely cause an ischemic stroke (IS) as a direct consequence of an acute attack (as in migrainous infarction) or predispose to an IS occurring remotely in time from a typical migraine attack (3–5). More recently, MwA has been associated with other VDs including intracerebral hemorrhage (ICH) (8), cardiac disease (6,7), retinal vasculopathy (9) and vascular mortality (10,11), while available studies do not allow any conclusion about a possible association between migraine and peripheral artery disease (PAD) (12,13). Moreover, incidental brain lesions, including infarct-like lesions, have also been reported in migraineurs (14–17). In this paper we review the available evidence linking migraine to VDs and propose some strategies to manage the overall risk in migraineurs.
Methods
Data for this review were obtained through search in BIOSIS, the Cochrane database, Embase, Google Scholar, ISI Web of Science, and Medline from their first availability up to December 2011. The search terms included “migraine” OR “headache” AND “stroke”, “transient ischemic attack” (TIA), “cerebral ischemia”, “intracerebral hemorrhage”, “subarachnoid hemorrhage” (SAH), “death”, “mortality”, “myocardial infarction”, “angina”, “coronary heart disease”, “coronary artery disease”, “retinal disease”, “retinal vasculopathy”, “peripheral artery disease”, “vascular disease”, “brain lesion”, “infarct-like lesion”, “silent infarct” and “white matter hyperintensity”. We also searched the reference lists of identified articles as well as papers quoting the identified articles. We reviewed only studies published in English. Titles and abstracts of all the retrieved studies were examined by two of us (SS and SR) for possible inclusion. Disagreements were resolved by discussion with a third party (AC) to reach consensus. We selected original research articles or meta-analyses describing migrainous infarction or the association between migraine and IS or TIA, migraine and hemorrhagic stroke, migraine and cardiac events, migraine and retinal vasculopathy, migraine and PAD, migraine and vascular and non-vascular mortality, and the presence of brain lesions in migraineurs. We reviewed all retrieved studies but, due to the plethora of data, we report, when available, only results from meta-analyses, avoiding the description of results from original research articles that were already included in the meta-analyses. When new original research articles were published after the release of the meta-analyses, we added the results from these new articles. The section of our article dealing with management of vascular risk is based on the personal experience of the authors.
Migrainous infarction
Migraine may be directly associated with an IS (migrainous infarction; International Classification of Headache Disorders, second edition [ICHD-II], code 1.5.4) (4,5,18,19). This condition is very rare, with an estimated incidence of 0.8/100,000 cases per year (20). In the past, before the development of the ICHD diagnostic criteria, it was vastly overestimated. Migrainous infarction can be diagnosed only in patients with a definite history of MwA. To meet the diagnostic criteria, the stroke must occur during a MwA attack that is typical of previous attacks, except for the persistence, for more than 60 minutes, of one or more aura symptoms; brain neuroimaging must demonstrate an ischemic infarction in a relevant cerebral area and the clinical condition must not be attributed to another disorder, even though stroke risk factors may be present (19). As most strokes in migraineurs occur outside migraine attacks, only a minority of IS in migraineurs meet these criteria. Migrainous infarctions tend to occur in young subjects and the lesions are located mainly in the posterior circulation territory (21). The prognosis, in terms of survival and functional outcome, is usually good (9). The pathogenesis is probably related to severe hypoperfusion during the aura phase, although the precise causative mechanism remains to be established (3,22).
Risk of vascular disease
Ischemic stroke and transient ischemic attacks
A large body of published data indicates an increased risk of IS in migraineurs (6,7,23–42). Whereas this increased risk has been statistically proven in MwA, only a non-significant trend toward the association has been observed in migraine without aura (MwoA). Two recent meta-analyses reappraised data on this issue (43,44). In the most recent one, referring to IS occurrence, the overall pooled adjusted-effect estimate was 2.04 (95% confidence interval [CI] 1.72–2.43) for subjects with any migraine versus no migraine (Figure 1), 2.25 (1.53–3.33) for subjects with MwA versus no migraine, and 1.24 (0.86–1.79) for subjects with MwoA versus no migraine (Figure 2), indicating that the presence of any migraine and of MwA is associated with a two-fold increase in the risk of IS (44). The other meta-analysis yielded similar results, finding the pooled relative risk (RR) of IS to be 1.73 (95% CI 1.31–2.29) in subjects with any migraine versus no migraine (Figure 1), 2.16 (1.53–3.03) in subjects with MwA versus no migraine, and 1.23 (0.90–1.69) in subjects with MwoA versus no migraine (Figure 2) (43). Compared with the findings of a previous meta-analysis, published in 2005 (45), those two meta-analyses (43,44) showed a similar risk of IS in individuals with any migraine and with MwA, but did not confirm the significant association between MwoA and IS (Figure 2). This discrepancy is due to the fact that several recent studies (6,35,40) did not find an association between MwoA and IS. The American Migraine Prevalence and Prevention (AMPP) study, a population-based investigation involving 120,000 US households (24), which became available only after the publication of the most recent meta-analyses (43,44),confirmed the above reported evidence. In detail, the AMPP study found an association between any migraine and IS (odds ratio [OR] 1.61; 95% CI 1.19–2.18) and between MwA and IS (3.14; 2.25–4.38), but no association between MwoA and IS (0.89; 0.61–1.30) (24).
Effect estimates for vascular events and mortality in subjects with any migraine versus no migraine in available meta-analyses. CI: confidence interval; CVD: cardiovascular disease. Effect estimates for ischemic stroke in subjects with migraine with aura versus no migraine and in subjects with migraine without aura versus no migraine in available meta-analyses. CI: confidence interval.
The risk of suffering a TIA seems to be increased in migraineurs, although this issue has not been extensively investigated and the possibility of a bias linked to misdiagnosis cannot be excluded, given that migraine aura may mimic TIA and vice versa. In an analysis of data from the Women’s Health Study (WHS), women with MwA versus non-migrainous women had an increased risk of TIA on multivariable-adjusted analysis (RR, 1.55; 95% CI 1.03–2.34), whereas women who had MwoA or a past history of migraine did not show any increased or decreased risk of TIA (39).
Hemorrhagic stroke
The few studies that assessed the association between migraine and hemorrhagic stroke (including ICH and SAH) provided conflicting results (8,25,28–30,46), hindering any definite conclusion about the possible association. With those limitations, the available evidence suggests that the risk, if any, is low, limited to subjects with MwA, and mostly related to ICH rather than to SAH.
Early data on the possible migraine-hemorrhagic stroke association were provided by a case-control study (28). Case subjects were women aged 20–44 years, hospitalized for IS, hemorrhagic stroke or unclassified stroke, and control subjects were age-matched women hospitalized for other reasons. No association emerged between hemorrhagic stroke and any migraine (adjusted OR 1.10; 95% CI 0.63–1.94), MwoA (1.84; 0.77–4.39) or MwA (0.86; 0.44–1.67). Among migraineurs, the ORs for hemorrhagic stroke were significantly increased only in women who had a history of high blood pressure and who were current smokers. A further study evaluating subjects with and without migraine failed to find any association between hemorrhagic stroke and migraine (adjusted hazard ratio [HR] 1.34; 95% CI 0.90–1.99) (30). Another case-control study that specifically investigated the possible link between SAH and headache, including migraine, found no association between a past history of headache or migraine and the presence of SAH (46). Conversely, peripartum migraine has been associated with an increased incidence of ICH (OR 9.1; 95% CI 3.0–27.8) but not of SAH (3.4; 0.5–24.1), although conclusions should be drawn with care due to the wide CIs (25). More recent data from the WHS reported an association between migraine and hemorrhagic stroke (8). In detail, the study found an increased risk of hemorrhagic stroke in women with MwA (age-adjusted HR 2.31; 95% CI 1.15–4.64) but not in women with MwoA (0.52; 0.16–1.66). The increased risk was more marked in the later years of the follow-up and the association was evident for ICH (2.78; 1.09–7.07) but not for SAH (1.73; 0.53–5.65); in addition, there emerged a stronger risk of fatal (3.56; 1.23–10.31) than of non-fatal hemorrhagic stroke (1.96; 0.78–4.93). However, the available data cannot be taken as definitive proof of the association, as the number of reported hemorrhagic strokes was low.
Cardiac events
The risk of cardiac events in migraineurs varies greatly among studies, ranging from a lower-than-average to a moderately increased risk (6,7,24,47–49). A major limitation is represented by the availability of data disaggregated for migraine type from only two studies (6,24). With those limitations, evidence suggests an increased risk of myocardial infarction in subjects suffering from MwA while no conclusion can be drawn with reference to any migraine or MwoA.
A recent meta-analysis did not show an increased risk of myocardial infarction in subjects with any migraine versus no migraine (pooled RR 1.12, 95% CI 0.95–1.32) (Figure 1) (43). In fact, only one (6) of the studies included in the meta-analysis reported disaggregated data for migraine type; in that study, which was part of the WHS, at the end of a mean 10-year follow-up, MwoA was not associated with an increased risk of any vascular event, whereas MwA was associated with an increased risk of major vascular events (multivariable-adjusted HR 2.15; 95% CI 1.51–2.78), including IS (1.91; 1.17–3.10), myocardial infarction (2.08; 1.30–3.31), coronary revascularization (1.74; 1.23–2.46), angina (1.71; 1.16–2.53) and death due to VD (2.33; 1.21–4.51). Only after the publication of the aforementioned meta-analysis (43), the AMPP study found an association between migraine (any migraine, MwA and MwoA) and myocardial infarction (24).
Other vascular comorbidities
There are insufficient data to draw any definite conclusion about the possible association between migraine and retinal vasculopathy. Several case reports have linked migraine with retinal infarction (50,51) and retinal vein occlusion (52,53). A population-based study in white Australians reported smaller retinal arteriolar caliber in subjects with a history of MwoA versus individuals without any history of migraine (54). Results from the Atherosclerosis Risk in Communities (ARIC) study indicated that middle-aged persons with migraine and other headaches, in the absence of arterial hypertension and diabetes, were more likely to have signs of retinopathy, including retinal hemorrhages, micro-aneurysms, soft or hard exudates, macular edema, intraretinal microvascular abnormalities, venous beading, swelling or laser photocoagulation scars (9). The association was stronger in patients with MwA.
Even available data on the possible association between migraine and PAD do not allow any definite conclusion, as the only two studies assessing the issue provided contradictory results (12,13). In both studies the presence of PAD was assessed by measuring the ankle-brachial-index (ABI), an easy-to-obtain, reliable indicator of PAD, as well as a marker of cardiovascular risk (55,56). The first of those two studies found a lower mean ABI value in migraineurs than in controls and showed that an ABI < 0.9 was independently associated with the presence of migraine (12). At variance, the second study did not report any difference in mean ABI values between migraineurs and controls (13). However, it is worth mentioning that the difference in the mean ABI value between migraineurs and controls found in the first of the two studies, despite significance, was low, and that in migraineurs the mean ABI value was in the normal range, indicating the absence of PAD. Moreover, in both studies the number of subjects suffering from MwA was low, hindering any conclusion as to the possible association between PAD and migraine type.
Vascular and non-vascular mortality
Although data on mortality in subjects with migraine are scarce, the available evidence points to high mortality rates in subjects with MwA with respect to non-migraineurs while subjects with any migraine or MwoA are not at increased risk of death (6,7,10,11,30,47,57,58). The 2009 meta-analysis, already mentioned, which evaluated data from five studies, did not find an overall association between any migraine and mortality from VD (pooled OR 1.03; 95% CI 0.79–1.34) (Figure 1) (43). Only one of the included studies, which was part of the WHS, provided data on mortality according to migraine type, reporting an increase in subjects with MwA (HR 2.33; 95% CI 1.21–4.51) and no increase in subjects with MwoA (1.06; 0.46–2.45) (6). The relationship between migraine and mortality was assessed more recently in a further meta-analysis of 10 cohort studies, in four of which MwA and MwoA were evaluated separately (Figure 1) (11). Results showed that the presence of any migraine did not alter the risk of all-cause (pooled RR 0.90; 95% CI 0.71–1.16), VD (1.09; 0.89–1.32) or coronary artery disease mortality (0.95; 0.57–1.60). Moreover, only MwA increased the risk of all-cause mortality (data from the Reykjavik study (10)), VD mortality (1.57; 0.89–2.78) and coronary artery disease mortality (1.29; 1.12–1.50).
Epidemiological factors influencing the relationship between migraine and vascular disease
Other important issues are whether migraine may represent a risk factor for VD in both sexes and across all age groups, and whether there is an association between increased migraine frequency and VD. So far, the available data are insufficient to draw any conclusive answer. As the early studies focused on women (27,41,42), there exists a huge body of evidence linking IS with migraine in the female gender; data on migraine in men are scarce and lacking in details, mostly referring to migraine type. Even with these limitations, the available evidence suggests a definite increase in the risk of vascular events in the cerebral and cardiac vascular districts in women suffering from MwA. Since no direct estimates of the risk in men versus women are available, it is impossible to establish whether it is higher in one gender. The 2009 meta-analysis showed an increased risk of IS in women with any migraine versus women with no migraine (pooled RR 2.08, 95% CI 1.13–3.84), but not in men with any migraine versus men with no migraine (1.37, 0.89–2.11) (43). The presence of any migraine in the WHS and in the Physicians’ Health Study (PHS) was associated with a greater risk of major vascular events, mostly due to an increase in myocardial infarction but not to an increase in IS (6,7). Women, but not men, were also found to be at increased risk of coronary revascularization, angina and death from VD. In the AMPP study, stroke was more likely to occur in subjects with MwA than in those with MwoA in both sexes. The same study reported an increased risk of vascular events in men and women affected by any migraine, by MwA and by MwoA (24). Although, as already mentioned, direct comparisons between men and women are lacking, the ORs were higher in men than in women, while CIs were mostly overlapping. Data on mortality are conflicting; in the WHS, women affected by MwA showed an increased risk of death from VD, while in the PHS, men suffering from any migraine did not present an increased risk of death from VD (6,7). Conversely, in the Reykjavik study, mortality from VD was marginally greater in men than in women affected by any migraine and by MwA (10).
When considering age, migraine tended to be regarded as a risk factor for stroke in the young, because the earlier studies were carried out in young patients aged less than 35 or 45 years; the same studies found that the risk of IS in those patients was influenced by age, being greater in younger (less than 35 years) than in older patients (between 36 and 45 years) or in the overall group (27,41,42). Thereafter, data from the WHS became available (6), in which women aged 45 years or older were included, that corroborated the association between IS and MwA in older patients as well, even though the association was more evident in women between 45 and 50 years (6). The migraine-IS association was also found to be influenced by age in men, as shown by data from the PHS, which included men aged 40–84 years at study entry (7). The age-adjusted HR of IS was 1.84 (95% CI 1.10–3.08) in migraineurs aged less than 55 years compared with men who did not report migraine, in the absence of any significant association in the older age groups. In the 2005 meta-analysis, the risk of IS was increased among men and women of less than 45 years of age (RR 2.36; 95% CI 1.92–2.90) (45). This same risk in the 2009 meta-analysis was greater in migraineurs of less than 45 years of age than in the overall group (2.65, 1.41–4.97) (43).
Few studies reported data on the impact of migraine frequency on the occurrence of stroke and other vascular events (7,28,32,35,59,60). This was mainly due to the low number of outcome events in available studies, causing uncertainty in the interpretation of subgroup analyses-based estimates. Overall data suggested that an increased frequency of attacks was mainly related to a higher incidence of IS in MwA. In patients suffering from MwA, an initial migraine frequency of 13 or fewer attacks per year was associated with a ten-fold increase in the risk of IS, while a lower frequency of the attacks was not associated with an increased risk (59). In agreement with those results, the Stroke Prevention in Young Women (SPYW) study reported an increased risk of IS in women with MwA experiencing more than 12 migraine attacks per year (35). In the same study, recent onset of migraine (within the previous year) was related to an eight-fold increased risk of IS (35). At variance, in another study, restricted to women using oral contraceptives, no association was found between higher migraine frequency and stroke (28). The WHS is the only study which included cardiac events when evaluating the association between migraine frequency and incidence of VD. In an early report no association was found between migraine frequency and IS (32). Subsequent analyses showed that attack frequency did not change the risk of VD in patients suffering from MwoA (60). Different results were found in women with MwA, in whom a frequency of less than one attack per month was linked to an approximately two-fold increase in the risk of major vascular events, including myocardial infarction, IS, coronary revascularization and angina (60). In the same group, a monthly attack frequency was not related to any change in the risk, while experiencing at least one attack per week was associated with a four-fold increase in the risk of IS (60). In the PHS, men who had four or more migraine attacks did not show any further increased risk of VDs than did men with a lower attack frequency (7).
Migraine and brain lesions
Since brain magnetic resonance has been available, incidental brain lesions have been reported in migraineurs. Migraine has been associated with white-matter hyperintensities (WMH) at brain MRI (14,15) and infarct-like lesions (16,17). Even in this case, data point toward a clear association with MwA whereas data referring to any migraine or MwoA are controversial.
The Cerebral Abnormalities in Migraine, an Epidemiological Risk Analysis (CAMERA) study showed in a population-based sample of men and women aged 20–60 years, that women, but not men, with any migraine, were at increased risk of deep WMH, and that the risk also increased with increasing attack frequency (14). Incidentally, a meta-analysis of some case-control studies showed that migraine was associated with a four-fold increased risk (OR 3.9; 95% CI 2.26–6.72) of WMH (61). More recently, data from the Epidemiology of Vascular Ageing Study, a large, cross-sectional, population-based sample of older participants, showed that the association with WMH was not specific to migraine headache but was present also for non-migraine headaches mostly represented by tension-type headaches (16). In the same study, any lifetime history of severe headaches was associated with an increased risk of higher volumes of WMH (62). The pattern of the association was similar for both migraine (multivariable-adjusted OR of being in the highest volume third for total WMH 1.8; 95% CI 1.0–2.9, p for trend 0.03) and non-migraine headache (2.7; 1.2–5.9, p for trend 0.01). When considering the association between WMH and migraine type, the study showed a lack of any significant association, with the only exception of a positive correlation between MwA and lesions located in the deep white matter (12.4; 1.6–99.4 for the highest third, p for trend 0.005).
The CAMERA study showed that subjects with migraine had a seven-fold increased risk for infarct-like lesions in the cerebellum compared with controls, an association that was stronger for subjects suffering from MwA and for those with a high attack frequency (14). Those findings are in agreement with the Reykjavik study, which found that women with MwA in midlife had an increased risk of cerebellar infarct-like lesions in later life (17). In that study, the risk was independent of the presence of cardiovascular risk factors; there was no increased risk for cortical or subcortical lesions in the overall group of women or in those with MwoA or non-migraine headache. In the Epidemiology of Vascular Ageing study, there was no evidence that a history of severe headache in general was associated with brain infarcts, as only participants with MwA had a greater than three-fold increased risk of brain infarcts (OR 3.4, 95% CI 1.2–9.3). Furthermore, there was a suggestion that participants with MwA were at increased risk of multiple infarcts (3.7, 0.8–17.3) (16). Most of the available evidence indicated that infarct-like lesions were more commonly located in the posterior cerebral circulation (14,16), at variance with the results of the Epidemiology of Vascular Ageing study, in which most of the infarcts were located outside of the cerebellum or the brain stem (16).
Management
To date, no peculiar features have been identified that may indicate which subjects, across the overall MwA population, are at the highest risk of vascular event. Even if a high frequency of migraine attacks may represent a possible indicator of an increased stroke risk, only future research will allow the eventual identification of other markers. Identification of those features is relevant to locating the subgroup of migraineurs at the highest vascular risk. In addition, the pathophysiological mechanisms linking MwA to IS, and to VD in general, have not been identified, although several possibilities have been hypothesized and reviewed (3–5,63,64). This is a further point of relevance for the proper management of vascular risk in migraineurs.
To date, MwA has been categorized among the less-well-documented or potentially modifiable risk factors for IS and its population-attributable risk is estimated at 3.5% (65); however, it is not reported among factors usually considered when addressing the cardiovascular risk in the general population (66). This represents an underestimation of the role of MwA, as available data on the association between MwA and IS do allow for considering MwA to be a well-documented risk factor for IS even if it is associated with a small absolute risk (lower than the risk associated with other recognized stroke risk factors). By contrast, most evidence indicates that MwoA cannot be regarded as a risk factor for IS. In addition, we must consider that evidence linking MwA and MwoA to vascular events in districts other than the cerebral one are incomplete and indicate that MwA should be considered among the inadequately documented risk factors for cardiac diseases and vascular mortality and that MwoA should not be considered in any way as a risk factor for these conditions.
The risk of IS in migraineurs is magnified in the presence of some acknowledged vascular risk factors (28,29,35,41,45,67). In the presence of cigarette smoking, the risk of IS in migraineurs is increased three- to nine-fold and four- to eight-fold in the presence of oral contraceptive use (28,41–45,67). The combination in the same subject of smoking and oral contraceptive use is associated with a ten-fold increase in the risk with respect to the presence of migraine only (28,42). Unfortunately, there are no adequately powered studies to separately determine IS risk for patients with MwA and MwoA taking oral contraceptives and to draw definite conclusions referring to migraine type. We should consider as well that some studies also indicate that arterial hypertension, methylenetetrahydrofolate reductase gene polymorphisms, and procoagulant states may affect the risk of IS in migraineurs (68–70). At variance, there are no data clearly identifying factors that may affect the risk of vascular events other than stroke.
Patients suffering from migraine, when presenting at a headache center, are mostly treated to relieve pain. They are not usually screened for the presence of vascular risk factors and even in the presence of vascular risk factors they do not receive hints for their management because headache specialists do not consider themselves primarily involved in cardiovascular prevention. Although subjects suffering from MwA should not be alarmed, they should be made aware that their basic vascular risk may increase in some circumstances. Cigarette smoking and arterial hypertension represent well-documented risk factors for stroke and vascular disease in the general population, and therefore, basic recommendations about their management are similar for migraineurs and non-migraineurs, whereas prescription of combined oral contraceptives deserves special caution even in the absence of evidence-based studies on the prescription of oral contraceptives in migraineurs. Because MwoA is not a definite risk factor for stroke, however, no specific restrictions are warranted in women suffering from this condition, especially in the absence of comorbidities. Oral contraceptive use should be discouraged in women suffering from MwA, as they may contribute to an unacceptable increased vascular risk. Their prescription is also contraindicated in women with MwA and other comorbid vascular risk factors or congenital or acquired thrombophilia. An awareness-raising recommendation on this issue is needed for gynecologists, who, with few exceptions, when prescribing oral contraceptives are often unaware of the association between migraine and VD and of the implications of oral contraceptive use in migraineurs.
No drugs are currently recommended for the vascular prevention in migraineurs. Indeed, as the risk of developing a stroke in this population is small, no studies on vascular risk reduction in migraineurs have been conducted. It has not yet been understood whether treatments used to prevent migraine can also reduce the risk of IS and of other vascular diseases (71). Some authors consider migraine as a treatable stroke risk factor and consequently a target for stroke prevention strategies (44). The American Heart Association/American Stroke Association (AHA/ASA) 2011 guidelines for the primary prevention of IS recommend that “because there is an association between higher migraine frequency and stroke risk, treatments to reduce migraine frequency might be reasonable, although there are no data showing that this approach would reduce the risk of a first stroke (Class IIb; Level of Evidence C)” (65). In any case, the rationale that migraine prevention strategies may affect the risk of stroke is weak (71). If stroke would occur as a consequence of repeated migraine attacks, it could be hypothesized that any treatment able to reduce migraine frequency is able to reduce the risk of IS. However, with the only exception of the rare cases of migrainous infarction, this hypothesis is unlikely and does not explain the occurrence of vascular events in extra-cerebral arterial districts. A possible alternative is represented by a common underlying alteration that may favor migraine, stroke and even other vascular events (72). This possibility is supported by the existence of some genetic disorders presenting primarily with both migraine and vascular diseases (72) and by the presence of systemic alterations in arterial and endothelial function in migraineurs (73,74). Consequently, drugs that reduce the frequency of migraine attacks would hardly reduce the risk of vascular events. The use of drugs concurrently effective in treating migraine and preventing stroke should, at least in theory, be more reasonable. To date, however, the available migraine prevention treatments do not show vascular-protective effects. Beta-blockers are still widely used to treat arterial hypertension, but their benefits are related only to the reduction of blood pressure values in subjects suffering from arterial hypertension (75). Angiotensin-converting enzyme inhibitors and angiotensin-receptor blockers, which, in preliminary studies, seemed promising in preventing migraine attacks (76–82), were also associated with a reduction in the risk of vascular events in normotensive subjects in the presence of benefits beyond those that could be expected due only to the reduction of blood pressure values, suggesting the presence of a peculiar effect on the arterial or endothelial function (83). However, the effectiveness of these drugs has not been demonstrated for migraine prevention and we are far from having any clear evidence indicating that they may contribute to favorably modulating the vascular function of migraineurs.
Conclusions
Recommendations for improving care of patients suffering from migraine by targeting their vascular risk.
Footnotes
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Disclosures
AC has received honoraria and travel support for educational workshops from Sanofi-Aventis, Bristol-Myers-Squibb, Bayer and Solvay Pharma. SS and SR declare no conflict of interest.