Longitudinal course of vasomotor symptoms in perimenopausal migraineurs

Abstract

Objective: To examine the longitudinal course of vasomotor symptoms (VMS) in women with a history of migraine in comparison to women without a history of migraine disease.

Methods: The study sample consisted of 467 women with a self-reported prior migraine diagnosis and 2,466 women without prior migraine diagnosis who were assessed longitudinally during menopausal transition as part of the Study of Women's Health Across the Nation. Linear mixed regression models with backward elimination were used to evaluate longitudinal associations between VMS and migraine while adjusting for baseline and time-varying demographic, socioeconomic, psychological, and reproductive factors. Additional analyses were performed to further assess the specificity of the association between migraine and VMS that included evaluating the association between migraine and vaginal dryness and between back pain and VMS.

Results: A history of migraine predicted an increased frequency of VMS but not vaginal dryness during menopausal transition. Significant interaction between history of migraine and menopausal status for the prediction of VMS was also identified. Burden of VMS was found to be higher during late-stage perimenopause in women with migraine. In contrast, the history of back pain did not predict the frequency of VMS.

Interpretation: This is the first study to delineate that a history of migraine predicts an increased frequency of VMS in women during menopausal transition. Hypothalamic abnormalities and thermoregulatory dysfunction against a milieu of decreasing estradiol concentrations during menopausal transition may explain the increased frequency of VMS in migraineurs during menopausal transition. ANN NEUROL 2019;85:865-874.

FIGURE 1:

Adjusted mean vasomotor symptoms predicted by the model. Displayed are the mean values predicted by the fixed effects of the model for the migraine groups (control [CTRL] vs migraine [MIG]) at different menstrual status stages for frequency of hot flashes and night sweats. These are covariate adjusted means, specifically means estimated at the grand means of other covariates in the model. Asterisks indicate p < 0.05 (p values are adjusted for multiple pairwise post hoc tests using the Tukey method). The bar heights represent the average estimated values, and the error bars represent the standard error associated with each estimate. Peri = perimenopause.

FIGURE 2:

A model for shared pathway between migraine and menopausal transition for vasomotor symptoms (VMS). We propose a model for the potential shared pathway between migraine and VMS during menopausal transition that may explain the mechanism of increased frequency of VMS in migraineurs. We suggest that hypothalamic dysfunction in migraineurs in the presence of declining levels of estrogen during menopause may lead to increased sympathetic hypersensitivity and greater response to norepinephrine (NE) facilitating more frequent VMS. We suggest that genetic vascular susceptibility and trait faster estrogen decline rate further contribute to increased susceptibility to frequent VMS in migraineurs. This process parallels the hypothalamically mediated homeostatic/thermoregulatory changes during menopause leading to sympathetic reactions underlying VSM. MHPG: 3-methoxy-4-hydroxyphenyl-glycol.