The Impact of Migraine History and Medication Use on Pregnancy Outcomes

Recent studies contribute valuable findings by looking to two unique datasets.

Migraine is common among women of reproductive age and has been linked to adverse pregnancy outcomes including hypertensive disorders, low birth weight, preterm birth, and, more recently, spontaneous abortion (SAB) – pregnancy loss before 20 weeks of gestation. Similar associations have been reported for women who take migraine medication prior to or during pregnancy. Migraine and adverse pregnancy outcomes are also associated with increases in long-term risk for cardiovascular disease and stroke.

But the literature on these associations is inconsistent, perhaps because the studies examining them have been designed in varying ways. A pair of recent studies, by relying on two unique sources of data, now add important information on how migraine history and medication use affect pregnancy outcomes.

In one study, researchers led by Holly Crowe, Boston University School of Public Health, US, report that women with a history of preconception migraine were not at an increased risk of SAB. However, use of migraine medication in the preconception period modestly increased SAB risk, particularly in women who took their medications daily or preventively, or who used combination analgesic/caffeine medications. A modest increase in SAB risk was seen for those who took aspirin for migraine.

Alexandra Purdue-Smithe, Brigham and Women’s Hospital and Harvard Medical School, US, and first author of the second study, called the work from Crowe and colleagues “one of the largest and methodologically strongest prospective studies available” in this research area.

Meanwhile, in their study, Purdue-Smithe and colleagues report that pre-pregnancy migraine was associated with higher risks of preterm delivery, as well as some hypertensive disorders of pregnancy including gestational hypertension and preeclampsia. These risks were reduced in women who reported regular pre-pregnancy aspirin use.

“This work makes important contributions to the field,” wrote Juliana VanderPluym, Mayo Clinic, Scottsdale, US, in an accompanying editorial about the study from Purdue-Smith and her fellow investigators. “It is a large prospective study of pregnancy outcomes in individuals with migraine, in contrast to prior studies, which were often case-control studies, prone to recall bias, or retrospective registry-based studies that may select for patients with higher migraine burden by including only those with hospitalizations or prescriptions for migraine. Compared to prior small prospective studies, this study adjusts for important confounders and explores the relationship between migraine with aura and adverse pregnancy outcomes.”

The Crowe study appeared online in the The Journal of Headache and Pain on December 20, 2022, whereas the Purdue-Smithe study and accompanying editorial were published in Neurology on April 4, 2023.

Hey PRESTO
In her previous doctoral work, Crowe used data from Boston University’s Pregnancy Study Online (PRESTO). Since 2013, this unique fertility study has been recruiting female residents from the United States and Canada, age 21 to 45, who do not use contraception or receive fertility treatment. The purpose is to explore demographic, lifestyle, and medical factors associated with fertility and pregnancy loss among women of reproductive age.

It was through this work that Crowe saw the impact of migraine and other headache conditions on reproductive-age women, which spurred her to pivot to migraine as a key part of her dissertation research. The PRESTO database was the perfect opportunity to better understand potential associations between migraine and pregnancy outcomes, in part because of its unique study population, Crowe told Migraine Science Collaborative.

“We use the Internet to recruit couples who are attempting to conceive, and because they’re trying for a pregnancy, they tend to test more often. So, we find out about miscarriages more often or earlier than other studies, which may rely on doctors’ visits or other methods to ascertain whether a miscarriage occurred.”

PRESTO participants complete a baseline questionnaire about their lifestyle and medical/reproductive history before completing follow-up questionnaires every eight weeks for the next year to report pregnancy status. Pregnant participants are asked to complete additional questionnaires at around eight and 32 weeks of gestation, representing early and late pregnancy, respectively.

Exploring SAB risk in PRESTO
More than 15,000 eligible participants completed the baseline PRESTO questionnaire between June 2013 and September 2021. The final analytic sample included 7,890 participants who indicated during a follow-up questionnaire that they had conceived.

Twenty-one percent of participants reported a preconception migraine diagnosis or migraine medication use, and 19% of participants experienced an SAB over the course of the study. The median timeframe for SAB was six gestational weeks.

Contrary to the researchers’ hypothesis, a history of preconception migraine was not appreciably associated with an increased risk of SAB, after accounting for factors such as baseline age, body mass index, smoking status, and a history of polycystic ovarian syndrome or endometriosis. The frequency of migraine was also not appreciably associated with SAB risk.

But the risk of SAB was modestly higher in those who reported taking any migraine medication prior to conception, compared to participants without migraine; overall, the former had a 14% increased risk versus the latter, though the result was not statistically significant. Forty-eight percent of those with migraine reported using migraine medication at baseline or in the follow-up questionnaire.

After adjustment, the increase in risk was highest in those who took their migraine medications each day (a 38% increased risk compared to those without migraine; 14 SABs were reported) and in those who were using the medications for preventive purposes (a 43% increased risk; 11 SABs were reported). However, neither association was statistically significant.

Further, 68% of all participants reported using pain medication in their most recent follow-up before conception, with 7% saying they used it for migraine. Here, in general, there was little association between using pain medications for migraine or other indications and SAB. However, when looking at specific types of pain medications, the researchers found a modest 21% increase in those who used aspirin for migraine, compared to those who did not take pain medication.

Overall, Crowe was surprised by the lack of a definitive association between medication use and an increased risk of SAB, given that previous studies had reported such an association, particularly with non-steroidal anti-inflammatory drugs.

Finally, Crowe and colleagues concluded that any suspected associations between preconception migraine medication use and the risk of SAB likely stemmed from more severe underlying vascular pathology rather than from the medication itself.

Increased risks in NHSII
Purdue-Smithe looked to another large dataset, the Nurses’ Health Study II (NHSII), for her research. NHSII is a prospective cohort study of more than 115,000 US female registered nurses age 25 to 42. Data collection began with a baseline questionnaire in 1982, with updated data on medical conditions and behavioral factors collected on a biennial basis since 1989, including new diagnoses or experiences of migraine, as well as if participants had become pregnant. Like Crowe, Purdue-Smithe felt her chosen dataset was a unique resource to rigorously evaluate her hypotheses.

Nurses who had received a physician diagnosis of migraine or who said they had experienced a migraine before their first or most recent pregnancy were classified as having a pre-pregnancy migraine. All pregnancies following a migraine diagnosis were deemed migraine-exposed pregnancies.

The 2007 iteration of NHSII also asked participants who indicated they had experienced a migraine whether they had an accompanying aura or not. Crowe said this was “a step forward in understanding how the reproductive implications of different migraine phenotypes may differ.”

Of the 19,694 participants included in the analytic sample, 2,234 (11.3%) had received a physician diagnosis of migraine in the 1989 questionnaire, 1,078 (5.5%) were classified as having migraine with aura, and 1,156 (5.9%) were classified as having migraine without aura.

Unlike the finding from Crowe and colleagues with regard to SAB, the results from Purdue-Smithe did often show an increased risk of other adverse pregnancy outcomes in people with a history of migraine.

Crowe told MSC that it was difficult to directly compare the two studies due to their nature and outcomes of interest. For instance, the SABs Crowe focused on typically occur before the 20th week of pregnancy, whereas Purdue-Smithe’s outcomes were contingent on the pregnancy lasting beyond this point. However, Crowe felt both studies highlighted that the risk of adverse pregnancy outcomes may differ across different migraine presentations, such as migraine requiring medication use, migraine with or without aura, and so forth.

For instance, after adjusting for relevant health and behavioral factors in their statistical models, Purdue-Smith and colleagues found that a pre-pregnancy physician-diagnosed history of migraine was associated with a 17% increase in the risk of preterm delivery (prior to 37 weeks of gestation), a 28% increase in the risk of gestational hypertension, and a 40% increase in the risk of preeclampsia, compared to those without a history of migraine. Migraine was not associated with the risk of gestational diabetes mellitus or low birth weight.

Study participants experiencing migraine with aura had a greater increase in the risk of preeclampsia (51% versus no migraine) compared to those with migraine without aura (30% versus no migraine), though this finding was not statistically significant. There were similar increases in the risk of preterm delivery and gestational hypertension for those with migraine with or without aura, compared to those without migraine.

Finally, regular pre-pregnancy aspirin use (twice or more a week) in those with migraine reduced the risk of preterm delivery by 45%, compared to those who did not have migraine but still reported regular aspirin use. Regular pre-pregnancy aspirin use also reduced the risk of preeclampsia in those with migraine, although that association was not statistically significant.

The findings about aspirin provide an interesting contrast to Crowe’s results, where aspirin increased the risk of SAB.

“It could be that aspirin use reduces the risk of a number of adverse outcomes later in pregnancy, and my findings reflected an earlier connection between those whose migraine necessitated aspirin use for pain relief and SAB, rather than the aspirin itself,” Crowe told MSC.

Purdue-Smithe and colleagues wrote in their paper that migraine and adverse pregnancy outcomes are both associated with an adverse cardiometabolic risk profile, which suggests a shared underlying pathophysiology. However, adjustment for these and other confounders had little effect on the positive associations they observed between migraine and adverse pregnancy outcomes.

Instead, they argue that other subclinical factors may be important. For instance, C-reactive protein could play a role, as this marker of systemic inflammation is higher in those with migraine and has also been implicated in preterm delivery and hypertensive disorders of pregnancy.

The research keeps moving forward
Crowe and Purdue-Smithe hope the pair of papers will increase awareness of the potential reproductive health implications of migraine, from both the patient and healthcare provider perspective, particularly with regard to managing health in the preconception period.

“If you’re someone who suffers from migraines, given the natural inclination to want to reduce migraine medication use during pregnancy, you don’t want to overlook the period of time prior to conception, because this could still have an impact,” Crowe told MSC.

Meanwhile, the editorial accompanying Purdue-Smithe and colleagues’ paper says their work “suggests a potential role for aspirin in reducing the risk of preterm delivery and preeclampsia. The role of aspirin in modulating obstetric risk among pregnant individuals with migraine requires further study.”

As for now, the research marches on: A third study, from Crowe and colleagues, also showed an increased risk of hypertensive disorders of pregnancy in people with a history of migraine. That study, published in the April 2023 edition of Cephalalgia, and based on data from the UK’s Clinical Practice Research Datalink GOLD, a large longitudinal database of patient records, also reported an increased risk of these disorders in those who took migraine medication, compared to those without migraine.

Lincoln Tracy is a researcher and freelance writer from Melbourne, Australia. You can follow him on Twitter @lincolntracy.

References
Pre-pregnancy migraine diagnosis, medication use, and spontaneous abortion: a prospective cohort study.
Crowe et al.
J Headache Pain. 2022 Dec 20;23(1):162.

Prepregnancy migraine, migraine phenotype, and risk of adverse pregnancy outcomes.
Purdue-Smithe et al.
Neurology. 2023 Apr 4;100(14):e1464-e1473.

Adverse pregnancy outcomes and migraine: What we know and what we can do.
VanderPluym JH.
Neurology. 2023 Apr 4;100(14):645-46.

Migraine and risk of hypertensive disorders of pregnancy: A population-based cohort study.
Crowe et al.
Cephalalgia. 2023 Apr;43(4):3331024231161746.