Timothy C. Hain, MD, Chicago IL. Return to Migraine main page. Page last modified: May 4, 2018
You may also be interested in our many other pages on migraine on this site
Migraines are defined as recurrent headaches separated by symptom-free intervals and often accompanied by nausea and light sensitivity. Migraines are often accompanied by visual symptoms and are relieved by sleep; furthermore there is usually a throbbing quality.
|Migraine prevalence vs age in women (dashed) and men (solid). (Adapted from Stewart et al, 1994). The study that provided this data was according to age rather than time of menopause. This is probably the reason that this plot does not show the secondary peak in migraine that occurs perimenopausally (Macgregor and others).|
Women have more migraine about 3 times more commonly than men. This is attributed to hormonal fluctuations. Boys and girls prior to puberty have equal rates. However, starting at roughly the age of 12, the prevalence of migraine in women takes an abrupt increase compared to men. The prevalence of migraine peaks in women in the middle of their reproductive years -- about the age of 35. There is also a second peak of migraine at the time of menopause (see below), where migraine in women has a prevalence of roughly 30%. To our knowledge, migraine and hormonal influences has not been studied in individuals taking estrogen that are not women (i.e. men on estrogen).
There is good evidence that the increased incidence of migraines in women is caused by going from a high to a low estrogen state (Somerville, 1975; MacGregor et al, 2007; MacGregor 2015). As can be seen below, there are two points that this happens -- just after ovulation, and a few days prior to menstruation. Kiesner and Martin (2013) reported that there is a mid-cycle pattern (16%), a menstrual pattern (51%), and a noncyclic pattern (33). Other pieces of information supporting the estrogen withdrawal conjecture is that migraines typically diminish in women after roughly 4 years of menopause, and migraines also typically diminish in pregnancy, when estrogen levels are high but fairly constant. On the other hand, high stable estrogen levels may provoke aura.
Loder, Rizzoli and Golub published a very useful review in 2007. According to Loder et al (2007), "The most plausible explanation for estrogen withdrawal as a trigger for migraine is the hypothesis put forward by Welch et al of a “mismatch” between the timing of estrogen effects on gene regulation in the central nervous system, and its effects on cell membranes. He suggests that under ordinary circumstances estrogen-mediated gene regulation “modulates inhibitory peptide function in the trigeminal nerve.”
|Estrogen levels throughout the menstrual cycle, adapted from Ibrahimi et al, 2015|
Menstrual migraine (MM)
Maasumi et al (2017) recently provided an excellent review of Menstrual migraines and treatment options. The core idea is that increased headaches around menses are correlated with fluctuations in estrogen level (a withdrawal effect, see above). Estradiol drops from about 300 to 50 normally, around the time of menstruation. Strangely though, women with menstrual headaches are reported to have LESS fluctuation of both estradiol levels and the trigeminovascular vasodilator system than normal subjects (Ibrahimi et al, 2015). One would expect the opposite.
Martin et al (2006) stated that 'The overall prevalence of menstrual migraine in the general population is approximately 3%, but it is much higher within populations of migraineurs; 35% to 51% of female migraineurs have “menstrually related MWoA,” while 7% to 19% have “pure menstrual MWoA”'.
Women with menstrual migraines have similar ovarian hormones as women without menstrual migraines (Stewart, 1975). This rather old study looked at only 6 migraine patients however, and may have missed some of the fine points. Nevertheless, it is generally thought not very productive to check "hormone levels", as the difference in the migraine frequency appears to be related to personal susceptibility rather than estrogen, per se.
Ovarian hormones and migraine Menstrual migraine provocation during estradiol treated and native menstrual cycles (from Martin et al, 2006). The original data is from Somerville, 1972
It should be noted here that no hormonal treatment regimen has FDA approval for migraine or headache indication. Thus all of the use of these treatments is "off label".
The general goal of hormonal treatment is to stabilize estrogen and avoid or blunt the drop in estrogen around menstruation. There are many ways that estrogen can be administered -- basically pills, patches, and implants. One can also "shut off" estrogen using a gonadatrophin agonist such as lupron, or provide higher doses of estrogen. High doses of estrogen are generally now avoided due to worry about increasing the risk of blood clots and breast cancer. Nevertheless, there is a role for clinical judgment in situations where the benefits are large.
Additionally, the use of oral contraceptives is generally contraindicated in women with migraine with aura (Allais, Gabellari et al. 2009). The risk of stroke is roughly 2-fold increased (Schurks, Rist et al. 2009). Non-estrogen containing birth control pills can be used safely in migraine with aura, according to Dodick (Dodick 2009). In other words, estrogen is to be avoided when possible in women with migraine with aura.
Fertility clinics often use strong medications to manipulate hormones. These are prone to cause headaches. Lupron "cycling" therapy, associated with drastic drops in estrogen, often aggravates migraine, and if practical considering the entire health picture, it should be stopped.
As estrogen drops trigger migraine, and most birth control pills (OCP) contain 21 days of estrogen and 7 days of a "blank", it follows that birth control pills might reasonably increase migraine. Birth control pills also contain a progestin. Accordingly, in women with migraine that occurs just pre-menstruation, ideally birth control pills that cause estrogen to fluctuate (i.e. with spacers) should be stopped. If hormones cannot be stopped, say because of endometriosis, then they should be changed to a constant amount every day. It often takes 2-3 months for the beneficial effects of hormonal manipulations to take effect.
Our impression is that this strategy is generally not effective, perhaps due to the smaller amounts of estrogen in current birth control pills.
In the 1970s OCPs contained 50 to 100 mcg of ethinyl estradiol, while more recent pills contain 15 to 35 mcg. (Martin et al, 2006). Variant OCP pills include monophasic, triphasic and extended release. Triphasic OCP change the dose of hormone on a weekly basis followed by a placebo. Extended release OCPs contain a fixed dose of hormones for 3 months followed by a placebo week. Progestin-only pills have a fixed dose of progestin for 4 weeks.
Menstrual periods can also be stopped entirely or reduced in number. As of 2017, the most common method of doing this is to use estrogen containing birth control pills that "omit the spacer" for 2 out of every 3 months. This strategy has roughly the same health risk as estrogen containing birth control pills, but can theoretically reduce hormonal migraines by a factor of 3. As current formulations of birth control pills now contain less estrogen than in the past, this strategy may also be not be entirely effective. There have also been experiments where the spacer was reduced in days (i.e. from 21/7 to 24/4). This seems reasonable enough, but less likely to work.
Calhoun (2004) reported an open trial of an oral contraceptive in 11 women, containing 20 ug ethinyl estradiol on days 1-21, supplemented with 0.9 mg of conjugated equine estrogens on days 22-28. Calhoun reported a 50% reduction in headaches days/cycle, and also stated that migraine could be prevented if "the decline in estradiol is limited to the equivalent of 10 ug". While we are willing to consider this, we are doubtful that this conclusion can be reasonably established from such a small study.
A more drastic method of stopping hormonal fluctuations is to combine Leuprolide (Lupron) to stop internal estrogen with HRT, eventually followed by oopherectomy. There are reports of excellent responses (Loder et al, 2007). Obviously this is not a rational treatment option if one is hoping to have children. Leuprolide (Lupron) without HRT causes a precipitous drop in estrogen, and would be likely to trigger migraines rather than reduce them.
Use of transdermal preparations: Use of oral estrogen supplements has not generally shown as large a benefit as transdermal gels (Loder et al, 2007). The reason for this is not entirely clear. A trial of 50 ug estrogen patches during the pill-free interval of oral contraceptives was modestly effective (MacGregor and Hackshaw, 2002). This may have been too small a dose. Larger doses of patches seem a bit more successful. Estrogen patches (100 ug) was reported helpful (Pradalia et al, 1994).
Macgregor (2015) suggests using 1.5mg patches starting between days -5 and 02 of menstruation for 7 days (a total of 7 days). A 1.5 mg patch is a relatively large amount. Patches used as replacements for low dose estrogen pills may provide relatively more estrogen causing other risks and side effects.
According to Martin et al (2006), a 100 mcg estradiol patch results in serum estradiol levels of 45-75.
Estrogen implants (the most popular IUD is a progesterone device), were reported as very effective by Magios et al. (1983).
It is our impression from our clinical practice that the progesterone containing implants are a "wash" with respect to migraine. Implants, like "ablations", add to the difficulty of diagnosis as it is no longer easy to tell when the woman is cycling (or even if she has any estrogen fluctuations).
Medications used in a special way to prevent MM include:
- Diamox 250 mg: Twice a day, 3-4 days prior to menses. We are dubious. Tingling in the fingers and toes can be a side effect as well as increased urination.
- Magnesium (360 mg/day) was reported helpful in a small study (Facchinetti et al, 1991), and is generally accepted as useful in migraine in general. More recently, 500-600 mg is recommended. We think this is a good idea. For menstrual migraine, it need only be taken during the last 15 days of the menstrual cycle (i.e. counting day 1 as the last day of bleeding).
- naproxen (550 mg) can be started on day -5 and continued for 5 days. This approach obviously runs the risk of some stomach irritation
- Aspirin or Motrin, use a small dose for 3 days prior to anticipated menses (Sances et al, 1990)
- mefenamic acid (Ponstel) : (250 every 6 hours, not to exceed one week)
- Triptans -- most are taken twice or 3 times/day, starting 4 days before the anticipated headache. These were reviewed by Hu et al (2013).
- Frovatriptan 2.5 mg-- this 24 hour triptan is also used similarly to Amerge. Recently, extremely high pricing has made frovatriptan unaffordable to many. Most protocols involve 6 days of medication, starting 4 days prior to flow. Twice/day treatment is more effective than once/day, but why not just use 5 mg once/day as the drug lasts a very long time. We think this is often a good idea.
- Naratriptan (Amerge) -- this triptan is sometimes used off-label to prevent menstrual migraine. Safety is not known. Twice/day is the usual frequency.
- Sumatriptan 25 mg 3 times/day 2-3 days before expected headache and continued for 5 days (Newman et al, 1998). Interestingly, sumatriptan is a generic medication and far less expensive than frovatriptan for most. However, this regimen seems less convenient and also more "up/down" in blood levels than frovatriptan. We think this is a "fall back" for those whose insurance won't cover frovatriptan.
- Zolmitriptan (2.5mg times/day) also has been studied and is somewhat effective. We see this as similar to sumatriptan.
Abortive medications: These medications are used identically for migraine in general, look here for a discussion. Triptans on a one time basis may not be as durable in women with premenstrual headaches however (Bhambri et al, 2014).
|Migraine by age in women and men, in the author's "dizzy" clinic practice. This graphic clearly shows that women have far more migraine than men, and also that the peak age is not 35 as in the general population, but rather is 50 in patients who go to the clinic for help with their dizziness (in both men and women).|
There is evidence that while migraine diminishes with age, there is often a flare in migraine perimenopausally (MacGregor and Barnes, 1999; MacGregor 2006; Wang et al, 2003). According to MacGregor (1999), about 30% of women in menopause experience migraine headache. Menopause is defined as 12 consecutive months without menstrual bleeding. Some studies report much higher prevalence of migraine (e.g. see Loder et al, 2007).
The 30% prevalence figure is similar to the prevalence reported in 35 year old women, and thus represents a second "peak". Owada and Suzuki (2014) reported that menopause related dizziness is correlated with hot flashes. We agree.
Estrogen during perimenopause Urinary estrogen and progesterone in perimenopausal and younger women. From Martin et al, 2006. Original is from Santoro NJ, 1996.
According to Loder et al (2007), "Although one might expect that estrogen levels in premenopausal women decline smoothly and gradually during this transition, estradiol levels are in fact increased during the perimenopause, and often are higher than those of the premenstrual years. The figure above shows that estrogen is roughly 1.5 higher during ovulation than in women in mid-life. Progesterone, on the other hand, is less. Thus perimenopausal women have a steeper decline in estrogen when they menstruate than women in mid-life. In addition, in perimenopause, estradiol receptors also may be increased in tissues.
Estrogen declines markedly in the ﬁrst year after the last menstrual period and then remain low and stable. " Serum levels typically range from 10-20 pg/ml. (Martin et al, 2006).
In the author's practice, a pattern of more headaches for about 4 years following onset of menopause is common, and usually diagnosed as migraine. The duration of headaches is not studied in the literature at all as yet, but the author's experience is that there is roughly a 10% decrease in the number of patients troubled with each year, up to roughly 65., It is rare to encounter migraines in either gender after the age of 65. Medication management in the author's practice usually starts with a drug that reduces hot flashes (e.g. venlafaxine). If this doesn't work, one may (as a last resort) use drugs that suppress of hormonal fluctuation (Loder et al, 2007), see below.
Hormone treatment -- generally not used for migraine in Menopause
It should be noted here that no hormonal treatment regimen has FDA approval for migraine or headache indication. Thus all of the use of these treatments is "off label".
HRT -- hormone replacement therapy -- consists of a combination of low-dose estrogen and medroxyprogesterone. The estrogens in HRT are typically much lower strength than those in oral contraceptives. Although one would expect that HRT would help with perimenopausal migraines, the evidence for it helping is not compelling.
According to Loder (2007), "Thus, evidence suggests that all forms of HRT may have negative effects on migraine and headache complaints, but that the extent of worsening may depend on the particular regimen that is employed. " In essence, continuous regimens are less troublesome than intermittent regimens. This is puzzling - -perhaps the treatments have been too timid to stop migraine or perhaps migraines perimenopausally are triggered by something other than estrogen fluctuations. For example, perhaps the gonadatrophins modulate migraine as well.
The progesterone component is used in women who have an intact uterus to prevent uterine hyperplasia. Progesterone, per se, appears to have no effect on headaches (Loder, 2007), or perhaps a small effect (e.g. Martin et al, 2006).
HRT is generally not continued indefinitely due to the potential for long term adverse effects, including increasing the risk of breast cancer. There is an increase in risk of stroke in women with migraine and aura, and for this reason, estrogen containing drugs (including BC pills) are generally avoided. (Sheikh et al, 2018). Estrogen should be avoided in women who have a genetic predisposition to breast cancer.
Regarding the risks in menopausal women, according to a reasonably credible source, the USPSTF, HRT should not be used for "prevention of chronic conditions in postmenopausal women". By this, they mean prevention of osteoporosis (for example). The USPSTF did not consider disabling migraines, but nevertheless it does not seem prudent to use HRT unless other treatments have failed. We recommend stopping HRT for migraine after 5 years. Of course, clinical judgment must be used.
Other drugs that may help manage the hot-flashes and headaches that are often combined include SSRI's (such as paroxetine and fluoxetine), SNRI's (such as venlafaxine in low doses), low-dose clonidine, and gabapentin.
Migraine prophylactic medications are often helpful for controlling the headache symptoms. We particularly favor venlafaxine in low doses for this purpose. Venlafaxine has two roles -- prevent migraine and prevent hot flashes. Migraine abortive medications such as triptans are also very useful.