Ovarian Hormones and Migraine Headache: Understanding Mechanisms and Pathogenesis--Part 2

Vincent T. Martin, MD; Michael Behbehani, PhD


Headache. 2006;46(3):365-386. 

In This Article

Menstrual Migraine

Menstrual migraine has recently been defined by the International Headache Society in the appendix of their diagnostic criteria for headache and has been divided into 2 subcategories: "menstrually related MWoA" and "pure menstrual MWoA." The criteria for "menstrually related MWoA" include: (1) predictable migraine attacks occurring during the perimenstrual time period (2 days before to 3 days after the onset of menstruation), (2) migraines also occur at other times of the month, and (3) the association with menses must be confirmed in 2/3 menstrual cycles. "Pure menstrual MWoA" is similar to the above criteria, except that migraine headaches are strictly limited to the perimenstrual time period and do not occur at other times of the menstrual cycle. For the sake of this review, we will refer to both types of migraine as "menstrual migraine."

The overall prevalence of menstrual migraine in the general population is approximately 3%, but it is much higher within populations of migraineurs[6]; 35% to 51% of female migraineurs have "menstrually related MWoA," while 7% to 19% have "pure menstrual MWoA"[7,8,9,10,11,12,13] ( Table 1 ).

Population-based studies of nonselected female migraineurs demonstrate little difference between the attack characteristics of menstrual and nonmenstrual migraine.[14,15] Studies within subspecialty-based clinics, however, suggest that menstrual migraine is more severe, disabling, and associated with greater abortive medication use than nonmenstrually related migraine.[16,17,18] The discrepancies between specialty- and population-based studies most likely can be explained by their selection of patients. Subspecialty-based studies may have included a greater percentage of "hormonally sensitive" patients than population-based studies, thus allowing a greater chance of identifying differences between menstrually related and nonmenstrually related attacks.

Interestingly only attacks of MWoA occur during the perimenstrual time period. Stewart et al[15] reported that attacks of MWoA were 2.04 times more likely during the first 2 days of menstruation, while attacks of MWA occurred with equal frequency throughout the menstrual cycle. Johannes et al[14] found that attacks of MWoA were 1.66 times more common during the first 3 days of menstruation. Mattsson et al[12] reported that 21% of patients experiencing MWoA reported >75% of attacks occurring during the perimenstrual time period as compared to only 4% of patients with MWA. These data could suggest that ovarian hormones differentially modulate these two subtypes of migraine headache.

Several studies have compared serum levels of ovarian steroids between patients with and without menstrual migraine. Davies et al[19] measured serum estrogen and progesterone for the 7 days before menses as well as during 1 mid-cycle day in menstrual migraineurs, nonmenstrual migraineurs, and controls and found similar levels of ovarian hormones between the groups. Epstein et al[20] demonstrated no differences between serum estrogen levels during the follicular and luteal phases of the menstrual cycle between women with and without menstrual migraine. They did, however, find significantly higher serum progesterone levels during the early follicular phase in menstrual migraineurs, but the clinical significance of this finding is questionable, since progesterone levels are typically quite low at this time of the menstrual cycle. Therefore, the predominance of evidence indicates that levels of ovarian hormones are similar between menstrual and nonmenstrual migraineurs. This would suggest that menstrual migraine is likely the result of an "abnormal response" of the central nervous system to normal fluctuations in ovarian steroids.

Menstrual migraine can be significantly improved by medical oophorectomy, which would imply that ovarian hormones are directly responsible for its development. Lichten et al[21] administered depo-leuprolide to create a medical oophorectomy in 29 women with severe menstrual migraine and noted that 17 of the participants had a >50% improvement in the headache index when compared to a placebo run-in phase. Murray et al[22] gave depo-leuprolide to 5 women with pure menstrual migraine and demonstrated a 74% decrease in the headache index during a 2-month treatment phase when compared to baseline. Therefore, hormonal therapies which minimize fluctuations in ovarian hormones can improve menstrual migraine.

The most plausible theory to explain the pathophysiology of menstrual migraine is that of "estrogen withdrawal." The estrogen withdrawal theory was advanced by Somerville, who demonstrated that the intramuscular injection of estradiol valerate administered shortly before menstruation could delay the onset of menstrual migraine by artificially raising serum estradiol levels during the late luteal and early follicular phases of the menstrual cycle[23,24] (Figure 2). Intramuscular administration of progesterone prior to menstruation did not affect the time of onset of menstrual migraine.[25] He later administered an intramuscular injection of a short-acting estrogen preparation to menstrual migraineurs during the mid-follicular phase of the menstrual cycle, but this did not trigger an attack. This experiment suggested that several days of estrogen priming prior to estrogen withdrawal was necessary in order to provoke a migraine headache.[26]

Onset of menstrual migraine during estradiol-treated and native menstrual cycles. In a representative patient, menstrual migraine breaks through on day 1 when the serum estradiol levels fall below 50 pg/mL during a normal menstrual cycle. In the estradiol-treated patient, the serum estradiol levels are artificially increased during the perimenstrual time period and therefore menstrual migraine is delayed until day 6 of the menstrual cycle. Day 1 of the menstrual cycle is the first day of menstrual bleeding. (Reproduced with permission from Somerville B. JAMA 1972;221:845-846.)

The estrogen withdrawal theory is further supported by studies that demonstrate perimenstrual treatment with estrogen can prevent the development of menstrual migraine. de Lignieres et al[27] reported that menstrual migraine occurred in 96% of a placebo group and 31% of a group treated with percutaneous estradiol gel during the perimenstrual time period. In an open label study, Calhoun[28] administered 0.9 mg of conjugated estrogens during the placebo week of OCPs to menstrual migraineurs and noted a 77% reduction in the number of headache days per cycle. Pradalier et al[29] showed that a 100-mcg transdermal estradiol patch applied perimenstrually was effective in the prevention of menstrual migraine, while a 50-mcg patch was ineffective. These data suggested that it may be necessary to maintain serum estradiol levels above the 45 pg/mL range during the perimenstrual time period to prevent menstrual migraine because the 100-mcg estradiol patch maintains serum estradiol levels in a range of 45 to 75 pg/mL.

Prostaglandins are released into the systemic circulation by a shedding endometrium during the perimenstrual time period secondary to the withdrawal of progesterone. Several lines of evidence suggest that "prostaglandin release" plays a role in the pathophysiology of menstrual migraine.[30] First, migraine-like headaches can be triggered by injections of prostaglandin E2 in nonmigraineurs.[31] Second, serum taken from women during menstruation and later infused back to them at a later time can induce headache.[32] Third, medications that are prostaglandin inhibitors have been used to prevent menstrual migraine.[33]

Magnesium ions are essential to a number of functions within the cerebral arteries and central nervous system. They regulate the tone of cerebral arteries, modulate release of nitric oxide (NO) from blood vessels, control the release of serotonin, block influx of calcium through N-methyl-d-aspartate (NMDA) receptors, and are essential cofactors for a number of enzymes.[34] Ramadan et al[35] measured brain magnesium using 31-phosphorus nuclear magnetic resonance spectroscopy and demonstrated that magnesium levels were low during a migraine attack. Therefore, magnesium deficiency could theoretically play a role in migraine.

Some authors have proposed that the low magnesium levels may be a trigger for menstrual migraine. Mauskop et al[36] reported a deficiency in ionized magnesium in 45% of attacks of menstrual migraine, while only 15% of nonmenstrually related attacks had a deficiency. They also demonstrated that attacks associated with low ionized magnesium could be aborted by intravenous magnesium infusions.[37] Facchinetti et al[38] demonstrated that menstrual migraine could be prevented by administration of oral magnesium during the last 15 days of the menstrual cycle.

Hormone withdrawal, prostaglandin release, or magnesium deficiency during the late luteal and early follicular phases of the menstrual cycle could alter the function of neurotransmitter systems relevant to migraine pathophysiology. Opiatergic and serontonergic neurotransmitter systems as well as the production of NO within platelets have been demonstrated to be modulated by the phase of the menstrual cycle in menstrual migraineurs.

Central opiatergic tone may change during different phases of the menstrual cycle in menstrual migraineurs. The release of β-endorphin and corticotropin-releasing factor is under negative control from the opiatergic system. Therefore, administration of naloxone, an opioid antagonist, should increase levels of β-endorphin and cortisol levels. Facchinetti et al[39] found little increase in the serum levels of cortisol and β-endorphin after administration of naloxone during the late luteal phase of the menstrual cycle in menstrual migraineurs, while their response during the follicular phase was similar to controls. This would suggest a failure of central opioid tonus during the late luteal phase in women with menstrual migraine.

Serotonergic function both within the central nervous system and platelets may be influenced by the phase of the menstrual cycle. Cassidy et al[40] found a hypersensitivity of 5-HT1A receptors during the early follicular phase (eg, first 5 days of the menstrual cycle). This hypersensitivity of 5-HT1A receptors was reduced or absent in those with chronic migraine or menstrual status migrainosis.[41,42] D'Andrea et al[43] showed that platelet serotonin levels were highest during the mid-cycle phase (late follicular and early luteal) and lowest during early follicular and late luteal time periods in migraine patients with and without aura. Fioroni et al[44] demonstrated that platelet monoamine oxidase-B (MAO-B) activity and 5-hydroxyindole acetic acid levels are increased and platelet serotonin levels are decreased during the late luteal phase as compared with the follicular phase in menstrual migraine patients. If platelet serotonin and MAO-B levels reflect serotonergic activity within the central nervous system, then patients with menstrual migraine could have decreased serotonergic tone during the late luteal and early follicular phases secondary to decreased synthesis, increased release, and/or increased catabolism of serotonin.

Sarchielli et al[45] found increased amounts of collagen-induced NO formation and cGMP production within platelets during the luteal phase of the menstrual cycle as compared to the follicular phases in women with menstrual migraine. Controls and patients with nonmenstrually related migraine experienced the greatest amounts of NO and cGMP production during the late follicular time period. They postulated that enhanced NO production during the luteal phase of the menstrual cycle could play a role in menstrual migraine if platelet abnormalities reflect events within the CNS or vascular endothelium.

Neurotransmitter systems could be directly affected through "withdrawal" of ovarian hormones or indirectly modulated through other chemical mediators (eg, prostaglandins, magnesium, NO). When compared to other phases of the menstrual cycle, the serotonergic/opiatergic systems seem to be suppressed during the late luteal and early follicular phases. These systems have been theorized to represent important inhibitory neurotransmitter systems for trigeminal pain pathways.

One could hypothesize that menstrual migraine develops secondary to a delicate balance of excitatory and inhibitory neurotransmission. As mentioned during part 1 of this review, animal studies suggest that estrogen enhances excitatory neurotransmission of glutamatergic synapses secondary to sprouting of dendritic spines and formation of NMDA receptors.[46,47,48] We postulate that the presence of estrogen induces similar structural changes within the trigeminal nucleus caudalis (TNC), but that neuroexcitation is dampened during the "high estrogen milieu" of the late follicular and early to mid-luteal phases of the menstrual cycle secondary to activation of inhibitory neurotransmitter systems (eg, serotonergic, opiatergic, GABAergic) in particular the GABAergic system. Recently, it has been established that high levels of sex hormones cause formation of GABA-A receptors that contain δ subunits.[49] These receptors are located in the extrajunctional sites and are more efficient in producing inhibitory responses to GABA. However, during the perimenstrual time period (late luteal and early follicular periods), a decline in serum estradiol levels leads to inactivation of these neuroinhibitory systems. These changes along with a delayed recovery of the neuroexcitatory glutamatergic system (because the structural changes within the TNC induced by estrogen may not reverse as fast) could render the perimenstrual time a particularly vulnerable period for migraine headache. (Figure 3; see part 1 of review for further details.) Obviously these hypotheses are speculative and will need to be confirmed in future studies.

Theorized pathogenesis of menstrual migraine. Estrogen and progesterone withdrawal, prostaglandin release, or magnesium deficiency during the late luteal and early follicular phases of the menstrual cycle lead to effects on excitatory and inhibitory neurotransmitter systems that increase the vulnerability to menstrual migraine.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.