Abstract
The aim of the study was to study seasonal variation in migraine headache in a group of women with menstrually-related migraine (MRM) compared with non-menstrual migraine. Via newspaper advertisement, women with migraine living in North Norway were invited. The patients were included by questionnaire and telephone interview. We prospectively recorded migraine attacks from a 12-month headache diary performed by a group of 62 women with a mean age of 36.0 years (range 16-46 years), who fulfilled the criteria of migraine without aura. Of these, 29 had MRM and 33 non-menstrual migraine. Mean ratio between number of attacks in the light arctic season (May-June-July) divided with total number of migraine attacks during 12 months was 0.24 (9.4/38.4) in the group of MRM compared with 0.25 (5.6/22.1) in others (confidence interval -4.2, 6.3, P = 0.84). Nor were there more migraine attacks in the dark season in an arctic area (November-December-January) in any group. We found a higher migraine attack rate in those with MRM, but no indication of more or less frequency of attacks during the bright arctic season. These findings support the assumption that MRM and seasonal variation of migraine are due to different mechanisms.
Introduction
Headache is a common complaint in the general population. The 1-year prevalence of migraine is approximately 5% in men and 15% in women (1–3). Before menarche, the prevalence of migraine is similar in both sexes. The gender difference in adults may be due to cyclic hormonal changes in women. Migraine is frequently associated with menstruation (4, 5), especially migraine without aura (6). Menstruation-related migraine attacks are also more severe and prolonged (5). Previous studies have shown that the occurrence of migraine attacks may be delayed by oestrogen, whereas progesterone delays menstruation but not migraine (7, 8). The complex interplay between the hypothalamic–pituitary axis and ovarian function seems to play a role in migraine related to menstruation (9). A decline in oestrogen concentration may precipitate migraine attacks, but the mechanism of how migraine attacks are initiated is unknown (10). Dysfunction in the suprachiasmatic nucleus (SNC) in the prodromal phase may play a role in the symptom production of migraine headache (11). When melatonin was measured in migraine subjects during a symptom-free interval and compared with controls, suppression of melatonin levels by exposure to light was more pronounced in migraineurs, supporting the assumption of SNC involvement (12). North Norway comprises Nordland, Troms and Finnmark counties with 460 000 inhabitants. Almost the entire region is located north of the Arctic Circle, and accordingly variations in light conditions during the year are huge. During the winter there are long periods of low sun and in the period around midwinter the sun never rises above the horizon. Equivalent to this period with polar night (dark season) there are several months with bright daylight (light season) even at midnight in the summer period. In North Norway, the sun is completely absent from about 20 November until 20 January depending on location, whereas the midnight sun is present from mid May until mid July. It is a common clinical experience that patients with migraine suffer more from headache in the light season. We have previously shown that up to one-third of patients with migraine living in North Norway report more migraine headaches in the arctic light season (13, 14). Seasonal variation of migraine headache occurs especially in patients with aura, but may also appear in patients without aura (14). Since MRM is largely a non-aura phenomenon, we proposed the hypothesis that the seasonal pattern of MRM attacks is not different from the seasonal pattern of non-menstrual migraine attacks.
Methods
Via advertisement in several local newspapers in North Norway and the internet site of ‘Nordlandssykehuset Bod⊘’, we invited women with migraine living in North Norway, an arctic area of extreme light conditions, to participate. One hundred and sixty-nine women participated in a questionnaire study about seasonal variation of headache and other migraine characteristics (14). The patients were then invited to keep a headache diary through 12 months; 89 women responded positively. From the diary, we recorded aura symptoms, duration of headache (time at onset and time when the attack ceased), character of the headache (throbbing, pressing), intensity of pain on a visual analogue scale from 0–10, additional symptoms (nausea, etc), use of medication and effect of medication (good, intermediate, no effect). The patients also recorded onset and duration of menstruation. Of these, 27 were patients with ‘pure aura’ migraine excluded. From this questionnaire, we obtained results from headache impact measurement by using Headache Impact Test (HIT)-6 (15). In the present study, the following inclusion criteria were presented in the advertisement: (i) female sex, (ii) age 16–45 years, (iii) mean one to six migraine attacks per month during the last 3 months before study entry, and (iv) settlement for at least 1 year in North Norway prior to study entry. At the time of telephone interview, we selected patients by using a complete list of criteria which also included: (v) migraine without aura or combined migraine with < 50% attacks with aura and (vi) lack of other serious medical conditions. After performing a 12-month diary from March 2004 to April 2005 that contained information about time of migraine attacks and menstruation periods, we completed the patient selection process by defining those with (vii) menstrually-related migraine (e.g. attacks occurring on day 1 ± 2 in at least two out of three menstrual cycles) according to the International Headache Society (IHS) criteria (Appendix) (16). Thus, the diagnosis ‘menstrual migraine’ was obtained from the patient's diary which contained information of onset of migraine attacks as well as menstruation. Patients with probable medication overuse headache (headache > 14 days per month) and patients who were not able to distinguish between migraine and other types of headache were excluded. We classified the migraine via telephone interview (K.B.A., S.I.B., R.S.) and recorded data according to the protocol using the revised IHS criteria (16). All patients who responded to the screening questionnaire about migraine and other headache characteristics were interviewed.
Statistical analysis
Data were analysed with SPSS software (version 12.0 for Windows; SPSS Inc., Chicago, IL, USA). Most of the data were ordinal and categorical. To compare frequencies, χ2 tests were performed. Independent t-test was used for comparison of means between the groups. Confidence intervals were also calculated. The level of significance was set at 5%. The regional ethics committee of Northern Norway reviewed the protocol and recommended the study.
Results
Demographic and clinical characteristics of patients are listed in Table 1. A history of previous stroke, epilepsy, hypertension or other serious disorders was not reported in any case, but 12 patients reported previous consultation with a neurologist for headache (six in either group). Forty-six patients (74%) used triptans for migraine and eight (13%) reported use of contraceptive pills. The mean number of migraine attacks in the total group was 32.1 (
Number of migraine attacks per month in menstrually related migraine (n = 29) and non-menstrual migraine (n = 33).
Demographic characteristic in patients with menstrually-related migraine and others
Two patients received work compensation due to migraine.
P = 0.36.
P = 0.68.
Migraine attacks in women with menstrually-related migraine and others as related to the arctic light and dark season; a comprehensive diary registration during 12 months
November, December and January.
May, June and July.
Ratio between number of migraine attacks in the light season divided by the total number of migraine attacks.
Discussion
In this study we could not demonstrate any relationship between MRM and seasonal variation of migraine headache in women who had been living in North Norway for at least 12 months prior to the study. Whether the experience of more migraine attacks during the bleeding period than during other phases of the menstrual cycle in women with migraine is due to dysfunction of the hypothalamus or to more direct effects of female sex hormones is not clearly proven. Recent reports have shown that MRM and pure menstrual migraine are conditions generally not associated with aura (17, 18). Furthermore, MRM lasts longer, is not associated with aura and is less responsive to therapy (17). Interestingly, we found that menstrual migraineurs had more frequent headache compared with others, suggesting that those with a hormonal trigger may have more frequent migraine attacks. The rhythmic secretion of melatonin occurs mainly during the night and is reinforced during winter (19). Since the number of migraine attacks in MRM was almost identical in light and dark seasons, the data do not support an important role for melatonin in explaining increases in frequency of migraine attacks in those with migraine without aura. MacGregor et al. confirmed the old assumption that migraine may be triggered by a fall in oestrogen level (7, 20). The cyclical changes manifesting in menstruation in fertile women are presumed to be driven by the hypothalamic–pituitary axis (21). One might hypothesize that the preponderance of migraine attacks during menstruation is linked to biological rhythms induced from the SCN. Lack of seasonal variation in migraine attack frequency in women with menstrual-related migraine compared with menstrual migraine may support the assumption that MRM may be due to abnormal response to hormonal variation rather than abnormal hormonal variations. The question of a possible common hormonal mechanism for both MRM and seasonal variation in migraine living in an arctic climate appeared after the observation that patients with migraine reported more headache ailments in the light season, as reported in previous studies (13). In a follow-up study, we found that 47% of patients with aura reported more frequent attacks during the arctic light season, compared with 17% in cases without aura (14). Seasonal variation of migraine regardless of light or dark periods of the year was reported in 75% in the aura group and 46% in the non-aura group (14). The findings that MRM seems to occur in patients without aura while seasonal variation of migraine is far more common in aura patients suggest that those two phenomena are triggered by different mechanisms. There are some critical annotations to the study, including a limited sample and possible confounders such as use of contraceptive pills and exposure to other external factors and a spectrum of social conditions known to play a role in symptom production in migraineurs. On the other hand, the patients completed the diary comprehensively during 12 months, which enabled us to study the relationship between MRM and seasonal variation quite closely. Seasonal migraine variation may of course be a response to external stimuli without any underlying dysfunction of the hypothalamic–pituitary system. Light is known to be a trigger factor for migraine attacks, especially in aura patients (22). Also, interictal light hypersensitivity is common in migraine (14, 23). Hypersensitivity to light may theoretically be related to enhanced melatonin suppression interictally in migraineurs (12), and some have suggested that melatonin is efficacious in migraine prevention (24). Furthermore, melatonin secretion is prolonged in winter and may be more biologically effective in periods where the sun is completely absent (19). This is one interesting hypothesis derived from this study that needs attention in future studies. The hormonal role in the production of cyclic symptoms in a condition that by its nature is cyclical is probably complex and needs further exploration to clarify the underlying mechanisms.
Conclusion
In this study we did not detect any relationship between menstrually-related migraine attacks and seasonal variation of migraine. These findings support the assumption that menstrually-related migraine and seasonal variation of migraine in a geographical area with extreme light and dark exposure are independent phenomena.