Abstract
In our study ‘Migraine, headache, and the risk of depression: Prospective cohort study’ (1), we examined the associations between migraine, including aura status, as well as nonmigraine headache and incident depression. We observed an approximately similar increase in the risk of depression for women with nonmigraine headache (relative risk 1.44, 95% CI 1.32–1.53), migraine with aura (relative risk 1.53, 95% CI 1.35–1.74), migraine without aura (relative risk 1.40, 95% CI 1.25–1.56) and past history of migraine (relative risk 1.56, 95% CI 1.37–1.77) compared to those with no history of headache. To the best of our knowledge, there is only one other population-based study that has evaluated the association between migraine aura status and depression. This study assessed the lifetime prevalence of major depression in 536 participants with migraine, 162 participants with severe headache and 586 controls. The authors did observe a statistically significant difference in prevalence of depression by migraine aura status (Wald chi-square = 6.198, 1 df, p = 0.013) (2). In his letter, Tfelt-Hansen argues that in our study we did not observe a statistically significant difference because our migraine aura classification suffers from misclassification (3).
The evaluation of migraine aura in population-based studies is challenging as aura specifics may be difficult to assess via self-administered questionnaires. We note that in our prospective cohort study random misclassification is possible and this also includes migraine aura status. If we assume that we are completely unable to correctly classify migraine aura, we would expect to find no differences between migraine aura groups even if there truly is a migraine aura effect. However, for biologically plausible distinctions between migraine with and migraine without aura, such as the association with stroke, data from the WHS indicate a significant association only for migraine with aura (4). Thus, although a certain degree of misclassification of migraine aura may be present, we believe that a large proportion of patients who experience migraine with aura have been correctly identified in the Women's Health Study.
Both our study and the previous study observed that those who experience migraine are at increased risk of depression. However, differences in population characteristics and the ascertainment of migraine and depression may have resulted in different magnitudes of effect and in different findings with regard to migraine aura. Based on current evidence, it remains unclear whether the association between migraine and depression is modified by migraine aura status.