By Dara Jamieson, MD
Associate Professor of Clinical Neurology, Weill Cornell Medical College
Dr. Jamieson reports she is a consultant for Roche.
SYNOPSIS: A longitudinal cohort study of twins found no increased stroke risk related to migraine overall, but there was a modestly increased risk for stroke related to migraine with aura. Familial factors and vascular biomarkers associated with migraine with aura may explain its correlation with vascular disease.
SOURCES: Lantz M, et al. Migraine and risk of stroke: A national population-based twin study. Brain 2017;140:2653-2662.
Tietjen GE, et al. Migraine and vascular disease biomarkers: A population-based case-control study. Cephalalgia 2017; Jan 1:333102417698936. doi: 10.1177/0333102417698936 [Epub ahead of print].
A population-based cohort study used data from the Swedish Twin Registry to follow twins without cerebrovascular disease to evaluate migraine as a risk factor for eventual stroke. Twins who were born in the years 1935 to 1958 answered a headache questionnaire during the 1998 to 2002 time period; twins born between 1959 and 1985 answered a headache questionnaire during the 2005 to 2006 time period. Migraine with and without aura and probable migraine were defined according to the International Classification of Headache Disorders criteria. Cerebral ischemia and intracerebral hemorrhage diagnoses were obtained from national patient and cause of death registries. Twins were followed longitudinally, by linkage of national registers, from the date of interview until the date of first stroke, death, or until the study ended late in 2014. In total, 8,635 twins had any migraine/probable migraine headache (3,553 with migraine with aura and 5,082 with non-aura migraine/probable migraine headache) and 44,769 twins had no migraine headache history. Hypertension, peripheral arterial disease, and obesity were more common in non-migraineurs, with atrial fibrillation being less common. During the mean follow-up time of 12 years, there were 1,297 incident (1,073 ischemic and 276 hemorrhagic) strokes, with a mean age at the end of follow-up of 57 years. The data were analyzed using a Cox proportional hazards model, with results finding that any migraine/probable migraine headache and non-aura migraine/probable migraine headache were not associated with an increased risk for stroke. Migraine with aura was associated with a barely significant 27% increased risk for stroke in the initial analysis; but, there were no significant associations between migraine and specific stroke type. The age- and gender-adjusted hazard ratio (HR) for stroke related to migraine with aura was 1.27 (95% confidence interval [CI], 1.00-1.62; P = 0.05) and 1.07 (95% CI, 0.91-1.26; P = 0.39) related to any migraine/probable migraine headache. The estimated HR for stroke was non-significantly higher in twins younger than 50 years of age and in females. The 2,142 twin pairs discordant for migraine with aura showed an HR for stroke of 1.09 (95% CI, 0.81-1.46), which attenuated the association compared to the primary analysis. Genetic markers for stroke may contribute to this migraine with aura and stroke association.
The association of migraine and vascular disease biomarkers was evaluated in 300 women and 117 men (mean age 48 years) in the CAMERA 1 (Cerebral Abnormalities in Migraine, an Epidemiologic Risk Analysis) substudy of the Dutch general population-based Genetic Epidemiology of Migraine. There were 155 migraineurs with aura, 128 migraineurs without aura, and 134 controls with no severe headaches. The vascular disease biomarkers compared between groups were: fibrinogen, Factor II, D-dimer, high-sensitivity C-reactive protein (hs-CRP), and von Willebrand factor antigen. Stroke-associated clinical phenotypes, including migraine with aura, female sex, and long duration and high attack frequency of aura and of headache, also were evaluated. Fibrinogen and hs-CRP were elevated in migraineurs compared to headache-free controls. Migraine was associated with an increased likelihood of elevated Factor II and hs-CRP. Fibrinogen and Factor II were associated with migraine with aura in women, but not in men. The vascular biomarker hs-CRP was correlated with both increased years and numbers of migraine aura attacks. An increased number of attacks was a significant predictor of elevated von Willebrand factor antigen, D-dimer, and fibrinogen.
Epidemiological studies have shown that migraine with aura is associated with an increased risk of cerebral and cardiac ischemia. However, an increased cerebrovascular risk has not been shown for individuals with migraine without aura, the predominant migraine subtype. The reason for this consistently demonstrated increase in vascular risk associated with migraine with aura is not clear, with theories including cortical spreading depression, endothelial dysfunction, hypercoagulability, arterial dissection, and embolization through a patent foramen ovale. The Swedish twin study lends credence to a genetic theory linking migraine with aura and ischemic stroke; however, the relatively weak twin correlation indicates a multifactorial association. Given that migraine in general, as well as the specific migraine type, has a very strong familial correlation, a twin association could be expected, even if the migraine type was discordant between twin pairs. Elevated vascular biomarkers were associated with migraine, particularly migraine with aura, as well as with increased years of aura and number of aura attacks. Other studies, including the CAMERA study, have linked increased frequency and duration of migraine auras with neuroimaging findings that have the appearance of white matter ischemic disease.
Elevated vascular biomarkers should be investigated to determine which markers, if any, can be used to stratify risk of vascular events in patients with migraine with aura. All patients at risk of vascular disease should be counseled on the management of the more well-established vascular risk factors. However, migraine with aura in combination with some estrogen-containing contraceptives appears to increase the risk of cerebral ischemia. The ability to counsel women with migraine with aura about contraceptive choices potentially could be enhanced by using biomarker screening to determine the degree of vascular risk.