Michael Eller; Peter J Goadsby


Expert Rev Neurother. 2013;13(3):263-273. 

In This Article

Population-based MRI Brain Studies of Patients With Migraine

A separate set of incidental findings in migraineurs have been highlighted within the past decade. A Dutch-based population study found no statistical difference in clinically silent 'infarct-like' changes between control and migraine patients (mean age: 48 years) who underwent 1.5-T MRI brain scan (8.1 and 5.0%, respectively).[28] This prevalence is similar to that seen in patients in the general population in a larger contemporaneous study in nearby Rotterdam.[24] However, the diagnosis of migraine predisposed patients to subclinical imaging changes in the posterior circulation (p = 0.02). The odds ratio (OR) was higher in patients with aura (13.7), and those with higher frequency (more than one per month) migraine attacks (OR: 9.3) or both (OR: 15.8). In women with migraine, more deep white matter changes were seen (OR: 2.1).

These findings were elaborated in a follow-up study, where infarct-like changes on MRI in the posterior circulation were found to be disproportionately localized to the border zones of the cerebellum in migraineurs compared to controls. Of the 435 total patients who underwent a 1.5-T MRI brain, 60 patients demonstrated discrete posterior circulation changes on MRI, 39 of which were present in the cerebellum. Of the total, these latter changes on MRI were found in 0.7% of controls, 2.2% of MO patients and 7.5% of MA patients.[29] Independently, a large population-based study based in Reykevik was in agreement with these findings,[30] in contrast to a further study weakened by small numbers of MA patients.[4] The cardiovascular risk profile did not alter the risk of these changes. The authors suggest that these changes on MRI represent true infarcts, and hypothesize the mechanism relates to hypoperfusion and/or embolism, rather than atherosclerosis and small-vessel disease. Clinicopathological series are awaited in order to understand the nature of these MRI changes.

The 9-year follow-up of the Dutch study described above demonstrated no progression of changes characterized above in migraineurs, aside from a small (0.05 ml; 95% CI: 0.01–0.27 ml) but statistically significant increase in the volume of deep white matter hyperintensities in women.[31] These changes were not related to migraine frequency, duration, semiology or hypertension. There were no documented changes in cognition between groups over time, consistent with a separate population-based study.[4] This study following more than 200 patients over almost a decade provides tangible evidence that MRI changes seen in migraineurs are unlikely to be of clinical import.