Abstract
In this review, we underline the importance of linking migraine to reproductive stages for optimal management of such a common disease across the lifespan of women. Menopause has a variable effect on migraine depending on individual vulnerability to neuroendocrine changes induced by estrogen fluctuations and on the length of menopausal transition. Indeed, an association between estrogen ‘milieu’ and attacks of migraine is strongly supported by several lines of evidence. During the perimenopause, it is likely to observe a worsening of migraine, and a tailored hormonal replacement therapy (HRT) to minimize estrogen/progesterone imbalance may be effective. In the natural menopause, women experience a more favourable course of migraine in comparison with those who have surgical menopause. When severe climacteric symptoms are present, postmenopausal women may be treated with continuous HRT. Even tibolone may be useful when analgesic overuse is documented. However, the transdermal route of oestradiol administration in the lowest effective dose should be preferred to avoid potential vascular risk.
Introduction
Athena the goddess was born from the skull of Zeus, who was suffering from severe recurrent headaches. Greek legend clearly exemplifies the evidence that migraine is considered a ‘female disease’ since ancient times. Even Sigmund Freud in the 19th century made significant observations on the nature of migraine, ‘a disease less common in the healthy man, more present in the active phase of sexual life, occurring following a certain amount of internal and external stimuli, characterized by periodic attacks of severe pain and with a complex etiology’. 1
Medical literature has linked gender to migraine, not only because its preponderance in women from puberty to menopause but also because both neuroendocrine events related to reproductive stages (menarche, pregnancy and menopause) and menstrual cyclicity, as well as the use of exogenous sex hormones, such as hormonal contraception and replacement therapy, may cause significant change in the clinical pattern of migraine itself. 2,3
Geographical differences in migraine prevalence are not marked. More recent studies of the general adult population indicate lifetime prevalence rates between 6% and 10% for men and between 15% and 26% for women. 4 In European studies, one-year prevalence of migraine ranges from 6 to 15% in adult men and from 14 to 35% in adult women. 5 The epidemiological profile of migraine has remained stable over the years in both sexes. 6–9 Migraine prevalence in women displays a striking increase between 10 and 12 years, around the time of puberty, with a female-to-male ratio of 3:1 during the fertile age, and a gradual, progressive decline after the age of 40 years, towards menopause. 10,11
In spite of the concept that ‘the femaleness of the migraine condition is inescapable’, 12 the role of ovarian function remains to be fully elucidated. Endogenous sex hormone fluctuations throughout the female life cycle significantly affect individual migraine history, but it is likely that the endocrine milieu is only one of the multiple triggers of migraine in susceptible women. 13,14 However, estrogen variations are highly implicated in modulating the threshold to stressful challenges by altering neuronal excitability, cerebral vasoactivity, pain sensitivity and neuroendocrine changes throughout the menstrual cycle and not only at menstruation, a time in which migraine attacks may be more severe, long-lasting and refractory to both acute and prophylactic treatment. 15,16 Indeed, an association between estrogen ‘withdrawal’ and attacks of migraine without aura, as well as evidence for an association between high estrogen states and attacks of migraine with aura, is strongly supported by several lines of research, 17–20 and practitioners treating women with migraine need to have a clear understanding of this special link. 21,22
That being so, we performed a literature review related to the course of migraine during menopausal transition in order to provide information relevant to the clinical management of such a common disease. In particular, we aimed to facilitate decision-making for hormonal treatment in migraine sufferers with climacteric symptoms.
Migraine and menopause
Migraine incidence generally decreases with advancing age and about two-thirds of patients no longer have migraines by the age of 65. The prevalence of migraine in the elderly population has been estimated around 5% and women continue to be affected twice as much as men. 23 More or less 90% of women enter menopause between 45 and 55 years of age and life expectancy after menopause is, therefore, longer than 30 years. 24
Menopause has a variable effect on migraine depending on the individual vulnerability to the neuroendocrine changes induced by estrogen fluctuations and on the duration of the menopausal transition. Erratic estrogen secretion and unbalanced estrogen exposure due to anovulatory cycles and/or progesterone deficiency may worsen or even initiate migraine during the perimenopausal period and such endocrine aberrations often precede by several years the stable and low plasma levels of estrogen, typical of the postmenopausal period. 25 In addition, the intensity of climacteric symptomatology such as hot flushes, palpitations, night sweats, disturbed sleep, negative emotions and panic attacks may more or less contribute to trigger or aggravate migraine attacks. 26
MacGregor 27 has recently reported that the perimenopausal years are extremely critical for first consultations, meaning that migraine becomes a problem for women later in fertile life. On the other hand, a low prevalence of migraine at menopause emerged from a retrospective study, indicating that only about 12%, out of 1300 women suffering from migraine, were referred to a headache centre during the postmenopausal years. 28
The controversy on the real role played by menopause in the natural history of migraine may also be ascribed to the fact that neurologists and gynaecologists have often carried out studies from different points of view. Indeed, migraine sufferers referred to headache centres are not representative of the general population and are probably the ones who worsen after menopause, while patients recruited in clinics for menopause are generally evaluated during menopausal transition and may lack accurate diagnostic criteria, since the ‘headache’ symptom is commonly included in the majority of scales assessing menopausal wellbeing. 29 The effect of the menopausal transition on the frequency of migraine has been recently explored in a cross-sectional community-based survey conducted in Chinese women aged 40–54 years. 30 The prevalence of migraine was similar in premenopausal and early perimenopausal women (16.7%) and much higher in the late perimenopausal group (31%). The spontaneous menopausal group had the lowest prevalence (7%), while women who had had a hysterectomy reported the highest migraine prevalence (27%), particularly in those who had suffered from the premenstrual syndrome (44%). The presence of low oestradiol (<50 pg/mL) and high follicle-stimulating hormone levels (>30 mIU/mL) was associated with lower migraine prevalence, even in the premenopausal and early perimenopausal women.
Another study conducted in women aged 40–74 years attending a population-based mammography screening programme indicated a decrease in risk for migraine without aura in postmenopausal women. 31 Collectively, these data bring about the idea that migraine prevalence increases before menopause when a hormonal imbalance is present, particularly in women vulnerable to hormonal changes, and declines after menopause when oestradiol levels are low and no longer cyclical.
In a menopause clinic, the postmenopausal course of headache with a premenopausal onset differed according to the type of headache and the type of menopause. While migraine improved in almost two-thirds of cases, tension-type headache worsened or was unchanged in 70% of cases. On the other hand, women who had a natural menopause experienced a more favourable course of migraine than women who had a surgical menopause with bilateral oophorectomy, suggesting the abrupt estrogen withdrawal as a well-defined aggravating factor of migraine, probably coupled to the emotional impact of hysterectomy. 32 Even in a sample of postmenopausal women collected in a tertiary headache centre, surgical menopause was found to be significantly associated with a worsening of migraine, 28 confirming the clinical impression that the severity of climacteric syndrome related to the premature lack of ovarian estrogen and androgen may have a role in aggravating migraine condition. Moreover, a cross-sectional population questionnaire survey conducted in the Netherlands in women aged 39–60 years has shown that even hysterectomy with ovarian retention is significantly more present in women reporting moderate to severe migraine (15.1%) when compared with non-hysterectomized women (8.8%). 33
Migraine and exogenous hormones
Hormonally associated migraine
Loder et al. 34 very elegantly revised the issue of hormonally associated migraine and stated the difficulties of establishing a causal relationship between the use of exogenous hormones and migraine, especially in women already suffering from the disease. Indeed, even though it is of paramount importance for further research, the attempt to classify ICD-II hormonally associated headaches in the group of secondary headaches, 35 serious diagnostic inconsistencies and clinical variability are evident, as a consequence of inconclusive scientific findings. To fulfil the criteria for exogenous hormone-induced headache, it is required that the headache begins or ‘markedly worsens’ within three months of beginning exogenous hormones and ‘revolve or revert to his previous pattern’ within three months of stopping exogenous hormones. On the other hand, to fulfil the criteria of estrogen-withdrawal headache, headache should develop at least within five days of discontinuation of estrogen used for at least 21 days and should resolve within three days. Such a temporal association is, however, based on anecdotal and personal beliefs and it does not take into account many variables related to reproductive biology that may be relevant to a better understanding of the role of exogenous hormones in migraine. Indeed, the multitude of clinical gynaecological conditions and the large variety of exogenous hormones available in terms of biochemical nature, route, scheme of treatment and dose should be taken into account to produce scientific evidences.
It is also possible that by understanding the hormonal nature of some migraine attacks, hormonal compounds can be proposed as therapeutic options for migraine relief. Hormonally sensitive women may be recognized on the basis of their migraine history, i.e. onset of migraine during menarche, menstrual migraine, premenstrual syndrome, severe climacteric symptomatology at menopausal transition, and may be considered a special subgroup of patients. 36,37
Migraine and hormonal replacement therapy
A multitude of compounds are available on the market in order to personalize hormonal replacement therapy (HRT) on women's need and risk/benefit profile. According to the biochemical nature, estrogens and progestins can display different effects in target organs, especially at vascular and neuronal levels, 38,39 which may be relevant for the course of migraine at menopause. In clinical practice, it is very common to observe some benefits from HRT when women are in the menopausal transition because the treatment prevents erratic hormonal secretion, particularly when stable plasma estrogen levels are provided by the use of continuous regimen, and significantly improves quality of life. 40 On the other hand, the cyclic administration of progestins, which is mandatory in non-hysterectomized women, may induce migraine attacks. 41 In these cases, on the basis of their clinical experience Silberstein and Merriam 42 suggested the use of a progestin with low androgenic properties or a natural progesterone and the switch to combined estrogen–progestin continuous therapy, while Panay and Studd 43 proposed progesterone-releasing coils and vaginal progesterone to minimize systemic side-effects in progestogen-intolerant women. MacGregor 44 in a preliminary uncontrolled retrospective study suggested that transdermal oestradiol was associated with more improvement in migraine than oral conjugated estrogens. In addition, the same author stated that high doses of exogenous estrogens may induce migraine with aura, as it happens during pregnancy and hormonal contraception. 20 Aura may develop ‘de novo’ or estrogens may increase the frequency of pre-existing attacks. Higher dosages of estrogen replacement therapy may be more capable of inducing aura symptoms, while lowering the dosage or changing to another type of estrogen replacement may lead to the amelioration of aura symptoms. 45 Hodson et al. 46 found that headache is a substantial problem at menopause and in HRT users, since 259 women out of 1000 reported a worsening of the number of attacks. In addition, by using logistic regression models the same authors showed history of migraine and increasing difficulty coping with stress to be strong predictors for worse headaches at menopause and with HRT.
A greater use of antimigraine preparations by estrogen users than by non-users has been reported 47 and a recent cross-sectional study found that current HT use was associated with higher rates of migraine headache than non-use. 48 In the same study, migraine was experienced by 11.2% in one year and was significantly associated also with a younger age, a younger age at menopause and surgical menopause. 48 A very recent large, cross-sectional population-based study in Norway regarding use of HRT and headache in posmenopausal women demonstrated that headache, especially migraine, was more likely in women using HRT. 49 The Postmenopausal Estrogen/Progestin Interventions Trial showed an improvement in headache with oral conjugated estrogens when compared with placebo in those with a history of headache at baseline, whereas if there was no history of headache they were more likely to develop headache as a side-effect. 50 Having a history of premenopausal menstrually related migraine predisposes to develop a migraine attack following estrogen deprivation in postmenopausal women. 51 These data imply that HRT worsens the clinical picture of migraine when menopause is already well established.
Some prospective studies based on the use of headache diaries, which make it possible to record prospectively the characteristics of every attack, 52 have been conducted to investigate the impact of various HRT regimens on the course of postmenopausal headache. HRT significantly affects the course of migraine, but not of episodic tension-type headache, in postmenopausal women. 53 This observation fits with the common knowledge that migraine is more sensitive to ovarian hormones in comparison with tension-type headache, which is likely more affected by psychological distress and coping strategies. 54 Moreover, it was interesting to observe different effects exerted by the different routes of administration of HRT on the course of migraine. In particular, outcome measures worsened in women receiving an oral conjugated estrogen and medroxyprogesterone combination, while they did not change as compared with baseline in those receiving a 50-µg transdermal oestradiol patch and medroxyprogesterone. These data could suggest that the best way to hormonally replace a postmenopausal woman suffering from migraine is by using the transdermal route of oestradiol administration because it maintains a more stable hormonal milieu. 53 Indeed, in a subsequent study, Facchinetti et al. 55 suggested that the oral route of HRT administration progressively increases the attack frequency, days with headache and analgesic consumption over six months. However, women receiving regimens of daily continuous combined HRT had lower outcome measures than those receiving HRT regimens with intermittent dosing regimens (continuous sequential oral cyclic sequential), supporting the concept that giving a fixed dose of exogenous hormones on a daily basis is superior in the replacement of postmenopausal women with migraine. The dose of the progestin apparently does not make any difference in adverse events, including headache. 56 However, well-conducted studies to clarify the role of progestogens in the clinical expression of migraine in postmenopausal women are still lacking.
A recent experience in a sample of postmenopausal women referring to a tertiary headache centre with a long-lasting history of disease proposed the use of tibolone, a tissue-selective steroid with estrogenic, progestogenic and androgenic properties for alleviating climacteric symptoms, as a valid alternative to low-dose conventional HRT with estrogens and progestins. Indeed, such unique versions of HRT may offer some benefits in terms of analgesic responsiveness in long-term migraine sufferers reporting analgesic overuse. 57 Very recently, MacGregor 58 suggested that in menopausal women who have contraindications to estrogen therapy or do not wish to use it, compounds that inhibit serotonin reuptake, such as venlafaxine, fluoxetine, and paroxetine, may be used for the control of hot flashes and prevention of migraine. Even gabapentin, another non-hormonal option with some effectiveness in treating hot flashes and reducing the frequency and severity of migraine attacks, may be proposed. No conclusive data are available on the use of selective estrogen receptor modulators for osteoporosis prevention in postmenopausal migraine sufferers.
Conclusions
The present short review underlines the importance of linking migraine to reproductive stages for an optimal management of such a common disease across the lifespan of women. During the perimenopause, it is likely to observe a worsening of migraine, and a tailored HRT to minimize estrogen/progesterone imbalance may be effective. But during the postmenopausal years, women with severe climacteric symptoms may be treated with continuous HRT. Even tibolone may be useful when analgesic overuse is documented. However, the transdermal route of oestradiol administration and the lowest effective dose should be preferred to avoid potential vascular risk.
Competing interests
None declared.