Statistical testing of association between menstruation and migraine

Area of Science:

• Neurology
• Epidemiology
• Biostatistics

Background:

• Menstrually related migraine (MRM) impacts approximately 20% of female migraine sufferers.
• The precise mechanisms linking menstruation to migraine are unknown, with hormonal fluctuations being a leading hypothesis.
• Current diagnostic criteria rely on the concurrency of migraine attacks with menstruation, necessitating robust statistical methods to confirm true associations.

Purpose of the Study:

• To refine and correct a previously proposed statistical method for analyzing the association between migraine and menstruation.
• To develop a reliable statistical framework for identifying genuine cases of menstrually related migraine.
• To facilitate further research into the pathophysiological links between menstruation and migraine.

Main Methods:

• The refined statistical method utilizes a two-parameter null model for MRM, incorporating simulation modeling.
• Fisher's exact test with mid-p correction is applied to 2x2 contingency tables derived from patient headache diaries.
• This approach corrects a previously published flawed framework, offering a novel statistical means for establishing menstruation-migraine association.

Main Results:

• The probabilistic methodology demonstrated strong performance in receiver operating characteristic curve analysis.
• Tabulated cutoff values are provided for clinical settings to assess the menstruation-migraine association based on headache history.
• The refined method offers improved accuracy in identifying true associations compared to previous approaches.

Conclusions:

• The corrected statistical method provides a more reliable means to establish the association between menstruation and migraine.
• The study recommends extending the standard 3-cycle observation period with at least one perimenstrual window for adequate statistical analysis.
• This refinement aims to improve the diagnostic accuracy of menstrually related migraine.