Page 166 - Migraine, the heart and the brain

Chapter 9
different. Only one other population based study also reported longitudinal data like our study. The Atherosclerosis Risk in Communities cohort (ARIC) study cross- sectionally found an association between migraine and white matter hyperintensities which was in line with our nding, but they did not nd differences in progression over time.5 Furthermore they did not nd a gender effect in this association. Possibly these differences can be explained by the fact that the ARIC study, like most studies in the previous mentioned meta-analysis, did not distinguish between deep and peri-ventricular white matter lesions. Also the headache assessment was only done at baseline, so some controls might have become migraine patients during follow- up and were not analyzed as such. Recently a cross-sectional study with only female twins with and without migraine also did not nd an association between deep white matter hyperintensities and migraine. Subjects with cardiovascular history were excluded in this study, and speci cly migraineurs with aura were excluded on this history. Also responder rates in migraine were higher than in controls. Both might have biased ndings in this study.6
Our study was unique as we had a long follow up of both MRI as well as of clinical and migraine speci c variables. In this way we were able to show that progression of white matter hyperintensities was not dependent on persistence migraine activity, but also occurred in migraineurs who stopped having attacks. Likewise, progression of lesions was not dependent on attack frequency or total amount of attacks during life time
We have shown that a migraine attack related mechanism does not explain the association between migraine and deep white matter lesions. Thus which attack unrelated mechanism can explain this association? An underlying factor influencing, and causing, both migraine and lesions may be possible, such as a shared genetic factor. An example of this is the co-occurrence of migraine with aura in one third of patients with CADASIL who develop also white matter lesions and stroke and is caused by mutations in the Notch3 gene. Other possible explanations may be non- genetic factors that are associated with migraine as well as white matter lesions, such as endothelial dysfunction. Furthermore, a reduced resting cerebral flow in the brain in the white matter, as reported to be lower in female migraineurs compared to controls, could also explain the higher prevalence of white matter lesions. 7
For patients and treating doctors is seems reassuring that the migraine attacks itself does not seem to be causative for the lesions. Therefore, preventative migraine medication is not advocated to prevent white matter hyperintensities in women with migraine.
164