Migraine as a risk factor for cardiovascular events in the Veterans Health Administration (VHA)

Editor’s note: The research described below comes from a recipient of a 2024 MSC Travel Grant supporting travel to the 66th Annual Scientific Meeting of the American Headache Society. These grants reimburse travel expenses for those who have had their abstract for a presentation or poster accepted at a meeting.

By Alexandra Schwartz, PhD student, Yeshiva University, US.

What is the research gap that your study addresses?

The current study addresses gaps in our understanding of the risk of cardiovascular events including cerebrovascular events, myocardial infarction, and cardiovascular mortality, conferred by migraine. Much of current literature investigating links between migraine and cardiovascular disease (CVD) focuses on women under the age of 50. As such, the association between migraine and CVD in men and individuals over the age of 50 remains comparatively under-examined. Men are overrepresented in the VHA, such that approximately 2/3 of those diagnosed with migraine in the VHA are men, providing a unique opportunity to understand the risk of cardiovascular events in an understudied group.

What is your research hypothesis?

It was hypothesized that migraine patients would demonstrate higher rates of cardiovascular outcome events, compared to non-migraine matched controls, during the study baseline period.

What methodology did you use to address your research hypothesis?

The current study followed a retrospective cohort design. The VHA Headache Center of Excellence (HCoE) Administrative Headache cohort includes every veteran diagnosed with migraine in the nationwide VHA integrated healthcare system between FY (fiscal year) 2008 and FY2021 (N=681,788) as well as controls matched 2:1 on age, gender, race and ethnicity, and location of care during the year of cohort entry (N=1,076,950). The current study describes the associations between medically coded migraine and cardiovascular events captured via International Classification of Diseases – 9/10 (ICD-9/10) codes entered into the electronic health record during the baseline period (18 months surrounding the initial visit for migraine coded in the VHA electronic health record). Chi-squared tests of independence evaluated differences in rates of outcomes among the migraine and matched control groups. Outcome variables include cerebrovascular events (i.e., ischemic stroke, hemorrhagic stroke, transient ischemic attack) and cardiovascular event (myocardial infarction, coronary/peripheral artery disease, congestive heart failure) at baseline. We also examine associations between migraine and known cardiovascular event risk factors, including cigarette smoking.

What are the main results of your study?

Rates of cardiovascular disease overall were higher among migraine patients than matched controls (2.63% vs. 1.21%, p < .001). Rates of cerebrovascular events were higher among migraine than matched controls, including ischemic stroke (1.41% vs. 0.81%, p < .001), hemorrhagic stroke (0.40% vs. 0.22%, p < .001), transient ischemic attack (0.63% vs. 0.14%, p < .001). We observed a very small effect such that rates of acute myocardial infarction were slightly higher in migraine than matched control (0.18% vs. 0.16%, p = 0.039). Rates of congestive heart failure were lower in migraine than in matched control(1.32% vs. 1.08%, p< .001).

What conclusions did you reach based on your results?

Results suggest migraine is consistently associated with higher rates of cerebrovascular events assessed cross-sectionally, with more complexity in relationships between migraine and cardiovascular events. Future studies should longitudinally evaluate associations between migraine and cardiovascular events, and evaluate understudied subgroups (i.e., men, individuals under the age of 50, migraine with/without aura).

What are the limitations of your study?

Patients were limited to veterans presenting for care in the US Veterans Health Administration (VHA) and may not generalize to other health systems. This study used administrative data; therefore, questions were limited to those that can be answered by using structured data routinely gathered in the treatment of people with migraine in the VHA. Administrative data have greater risk of errors and lack of specificity compared with data collected specifically for research. Finally, the VHA only cares for veterans after they have concluded their military service, limiting our ability to capture migraine onset earlier in life.

What is the relevance of your study to migraine?

Findings from this study should improve utility of the Framingham Cardiac Risk Score (FCRS) and Framingham Stroke Risk Score (FSRS). Improved risk scores will immediately impact the largest integrated healthcare system in the US, would readily translate into other large healthcare systems, and will set the stage for future cluster randomized clinical trials that can evaluate the efficacy of implementing these tools within a healthcare system. A greater understanding of associations between migraine and cardiovascular events would allow clinicians to make more informed lifestyle and intervention recommendations.