The Role of Combined Hormonal Contraceptives in Migraine
In 2006, a consensus statement of the American College of Obstetrics and Gynecology (ACOG) recommended against use of combined OCs in women with migraine and focal neurologic deficits. They cited three reasons:
Some of the concerns raised were:
1. Are all women with migraine at increased risk of stroke with OCs?
There has been a concern for decades that if the stroke risk of OCs was combined with the stroke risk of migraine, the resultant hazards ratio would be unacceptably high. Evidence for this remains elusive. In the 1975 article that reported a relative risk of 4.4 with high-dose OCs, the risk in migraineurs taking those pills was not significantly different at 4.6. More recently, a population-based case-control study compared 386 reproductive-aged women who had suffered a first ischemic stroke with 614 controls. Subjects were classified as having no migraine, probable migraine without visual aura, or probable migraine with visual aura (PMVA). Although PMVA, in combination with smoking, was a significant risk factor for stroke (and that risk was synergistically increased by the addition of OC use), there was no increased risk for women with PMVA who used OCs but did not smoke.
A recent meta-analysis reviewed 14 studies to determine the relationship between migraine and risk of ischemic stroke. They reported significant relative risks of 2.27 for women with MwA and 1.83 for those with MwoA, while migraineurs who took OCs had a relative risk of 8.72. A more critical look, however, reveals that only 3 of those 14 studies were conducted in this century, indicating that results reflect older OC formulations. Of the 3 recent studies, one showed no risk associated with OC use in migraineurs, while the other two showed more modest risk ratios (1.6 and 2.3, respectively). Furthermore, one of these 2 studies clarified that among stroke cases, the majority used high-dose pills. The study's authors concluded that MwA and high frequency of aura were major risk factors for ischemic stroke but "in no case did correction for OC usage significantly alter these odds ratios."
2. A look at the 2-fold increased risk of ischemic stroke in the pooled analysis of 2 large US studies
A pooled analysis of 2 large US population-based case-control studies showed no increased risk of stroke with OC use. Reflecting today's lower doses and current prescribing habits in the US, fewer than 1% of cases and controls used high-dose OCs, and only 17% of OC users were smokers.
In that study, the reported "2-fold increased risk of ischemic stroke" among migraineurs using low-dose OCs was based on only 4 cases. Raw prevalence of migraine was actually identical in cases and controls – 7.8% vs 7.7%, respectively (4/51 cases and 14/182 controls). Thus, the relative risk of 2.08 was attained only after adjustment for other factors in 4 cases.
The study's authors urged caution in interpreting these data as "imprecise methods," which differed between the 2 study sites, were used to assign migraine diagnosis. Given the overlapping confidence intervals and the results of formal tests of heterogeneity, the authors' final statement was, "Taken together, no firm conclusions can be drawn… ."
3. A look at the 3-fold increased stroke risk in the Danish case-control study
A large Danish population-based case-control study also found no increased risk of stroke with low-dose OCs. Progestin-only pills were used by 0.6% of stroke cases and 0.7% of controls (NS); CHCs containing 20 μg EE were used by 2.9% of cases and 4.0% of controls (NS); 30–40 μg EE pills were used by 24.0% of cases and 23.7% of controls (a difference which was found to be significant only after adjustment for other factors; odds ratio 1.7); but high dose 50 μg EE pills were used by 4.2% of cases and only 1.1% of controls, giving an adjusted odds ratio of 4.7 – similar to the risk associated with high-dose pills since 1975.
The 3-fold increased risk in migraineurs using OCs is suspect, however, as only 6% of controls were identified as migraineurs in a population where 19% of women are reported to have migraine (17% of cases were identified as migraineurs). Although it is possible that migraineurs in Denmark were selectively excluded from receiving OCs, a similar study in France found an equal distribution of migraine diagnosis between stroke cases and controls – with both groups actually exceeding the published prevalence of migraine in French women.
Exploring American College of Obstetrics and Gynecology's Concerns
In 2006, a consensus statement of the American College of Obstetrics and Gynecology (ACOG) recommended against use of combined OCs in women with migraine and focal neurologic deficits. They cited three reasons:
Concerns remain that all women with migraines are at increased risk of stroke if they take OCs.
A pooled analysis of two large US population-based case-control studies identified a statistically significant 2-fold increased risk of ischemic stroke among current users of OCs who reported migraine compared with women with migraines who did not use OCs.
A large Danish population-based case-control study found that among women with a history of migraine, the risk of stroke was elevated approximately 3-fold.
Some of the concerns raised were:
1. Are all women with migraine at increased risk of stroke with OCs?
There has been a concern for decades that if the stroke risk of OCs was combined with the stroke risk of migraine, the resultant hazards ratio would be unacceptably high. Evidence for this remains elusive. In the 1975 article that reported a relative risk of 4.4 with high-dose OCs, the risk in migraineurs taking those pills was not significantly different at 4.6. More recently, a population-based case-control study compared 386 reproductive-aged women who had suffered a first ischemic stroke with 614 controls. Subjects were classified as having no migraine, probable migraine without visual aura, or probable migraine with visual aura (PMVA). Although PMVA, in combination with smoking, was a significant risk factor for stroke (and that risk was synergistically increased by the addition of OC use), there was no increased risk for women with PMVA who used OCs but did not smoke.
A recent meta-analysis reviewed 14 studies to determine the relationship between migraine and risk of ischemic stroke. They reported significant relative risks of 2.27 for women with MwA and 1.83 for those with MwoA, while migraineurs who took OCs had a relative risk of 8.72. A more critical look, however, reveals that only 3 of those 14 studies were conducted in this century, indicating that results reflect older OC formulations. Of the 3 recent studies, one showed no risk associated with OC use in migraineurs, while the other two showed more modest risk ratios (1.6 and 2.3, respectively). Furthermore, one of these 2 studies clarified that among stroke cases, the majority used high-dose pills. The study's authors concluded that MwA and high frequency of aura were major risk factors for ischemic stroke but "in no case did correction for OC usage significantly alter these odds ratios."
2. A look at the 2-fold increased risk of ischemic stroke in the pooled analysis of 2 large US studies
A pooled analysis of 2 large US population-based case-control studies showed no increased risk of stroke with OC use. Reflecting today's lower doses and current prescribing habits in the US, fewer than 1% of cases and controls used high-dose OCs, and only 17% of OC users were smokers.
In that study, the reported "2-fold increased risk of ischemic stroke" among migraineurs using low-dose OCs was based on only 4 cases. Raw prevalence of migraine was actually identical in cases and controls – 7.8% vs 7.7%, respectively (4/51 cases and 14/182 controls). Thus, the relative risk of 2.08 was attained only after adjustment for other factors in 4 cases.
The study's authors urged caution in interpreting these data as "imprecise methods," which differed between the 2 study sites, were used to assign migraine diagnosis. Given the overlapping confidence intervals and the results of formal tests of heterogeneity, the authors' final statement was, "Taken together, no firm conclusions can be drawn… ."
3. A look at the 3-fold increased stroke risk in the Danish case-control study
A large Danish population-based case-control study also found no increased risk of stroke with low-dose OCs. Progestin-only pills were used by 0.6% of stroke cases and 0.7% of controls (NS); CHCs containing 20 μg EE were used by 2.9% of cases and 4.0% of controls (NS); 30–40 μg EE pills were used by 24.0% of cases and 23.7% of controls (a difference which was found to be significant only after adjustment for other factors; odds ratio 1.7); but high dose 50 μg EE pills were used by 4.2% of cases and only 1.1% of controls, giving an adjusted odds ratio of 4.7 – similar to the risk associated with high-dose pills since 1975.
The 3-fold increased risk in migraineurs using OCs is suspect, however, as only 6% of controls were identified as migraineurs in a population where 19% of women are reported to have migraine (17% of cases were identified as migraineurs). Although it is possible that migraineurs in Denmark were selectively excluded from receiving OCs, a similar study in France found an equal distribution of migraine diagnosis between stroke cases and controls – with both groups actually exceeding the published prevalence of migraine in French women.