Prevalence of menstrual migraine: A population-based study

Abstract

Aim

To present data from a population-based epidemiological study on menstrual migraine.

Material and methods

Altogether, 5000 women aged 30–34 years were screened for menstrual migraine. Women with self-reported menstrual migraine in at least half of their menstrual cycles were invited to an interview and examination. We expanded the International Classification of Headache Disorders III beta appendix criteria on menstrual migraine to include both migraine without aura and migraine with aura, as well as probable menstrual migraine with aura and migraine without aura.

Results

A total of 237 women were included in the study. The prevalence among all women was as follows: any type of menstrual migraine 7.6%; menstrual migraine without aura 6.1%; menstrual migraine with aura 0.6%; probable menstrual migraine without aura 0.6%; probable menstrual migraine with aura 0.3%. The corresponding figures among female migraineurs were: any type of menstrual migraine 22.0%, menstrual migraine without aura 17.6%, menstrual migraine with aura 1.7%, probable menstrual migraine without aura 1.6% and probable menstrual migraine with aura 1.0%.

Conclusion

More than one of every five female migraineurs aged 30–34 years have migraine in ≥50% of menstruations. The majority has menstrual migraine without aura and one of eight women had migraine with aura in relation to their menstruation. Our results indicate that the ICHD III beta appendix criteria of menstrual migraine are not exhaustive.

Keywords

Migraine menstrual migraine menstruation prevalence classification

Introduction

Migraine affects approximately 20% of women in the general population (1,2). The majority has migraine without aura (MO), while one-third has migraine with aura (MA) (3). Few have co-occurrence of MO and MA in the general population (3,4). Migraine is equally frequent in boys and girls, but after puberty the prevalence is 2–3 times higher in women than men (3,5,6). Some of the female preponderance is likely to be caused by female sex hormones.

Prospective diary studies have shown that the incidence of MO in women is significantly higher peri-menstrually than between menstruations (7 –9). The cyclic changes in female sex hormones, in particular the premenstrual drop in oestrogen level, are hypothesized to trigger these attacks (10,11).

The first edition of the International Classification of Headache Disorders (ICHD I) published in 1988 noted:

‘Migraine without aura may occur almost exclusively at a particular time of the cycle—so-called ‘menstrual’ migraine. Generally accepted criteria for this entity are not available. It seems reasonable to demand that 90% of attacks should occur between 2 days before menses and the last day of menses, but further epidemiological knowledge is needed’ (12).

The revised criteria from the 2nd edition of the International Classification of Headache Disorders (ICHD II) published in 2004 were added in the Appendix (13). Menstrual migraine without aura is defined as MO occurring in the peri-menstrual period, i.e. on day –2 to +3 of menstruation in at least two out of three menstrual cycles (13). Menstrual MO is subdivided into pure menstrual MO and menstrually-related MO. Pure menstrual MO is characterized by MO attacks occurring exclusively in the peri-menstrual period, while menstrually-related MO also have MO attacks outside the peri-menstrual period. Lately, the ICHD III beta was published with diagnostic criteria that was unchanged from ICHD II (14).

The aim of the present study was to provide prevalence data on menstrual migraine from the general population. In order to explore the full spectrum of menstrual migraine and its boundaries, we investigated whether the ICHD III beta was exhaustive, since we also included women with probable MO, MA and probable MA in relation to menstruation.

Material and methods

Population

A random sample of 5000 women aged 30–34 years residing in eastern Akershus County, was drawn from the National Personal Registry by Statistics Norway in January 2005. Akershus County has both rural and urban areas and is situated in close proximity to Oslo. The sample constituted 38.6% (5000/12,948) of all women aged 30–34 years in the sampling area and 3.0% (5000/168,752) in Norway (15). Data from Statistics Norway show that the sample was representative of the total Norwegian population regarding age and marital status. Regarding employment sector, trade, hotel/restaurant and transport were over-represented and industry, oil and gas and financial services were under-represented (15).

Interview, physical and neurological examination

A screening questionnaire was mailed to all women in January 2005. It contained six questions about headache, migraine and its relation to menstruation. The question ‘How often do you have migraine in relation to menstruation?’ was used to screen for menstrual migraine. Responders with self-reported migraine in at least half of their menstrual cycles were included and invited to a semi-structured interview and a physical and neurological examination. Those with ≥180 headache days within the last year were excluded. The invitation letter stated, apart from ensuring confidentiality, that the objective was to study headache in relation to menstruation.

The interviews and examinations were conducted by a neurologist (K.G.V.) between September 2011 and May 2012 at the Research Centre, Akershus University Hospital, Norway. At that time, the study population was 36–40 years old. Participants unable to meet at the clinic were interviewed by telephone. The interview focused on headache and migraine. The headaches were classified according to the ICHD III beta including the menstrual migraine appendix criteria (14). We also included MA in order to explore whether it was associated with menstruation, applying the appendix criteria for menstrual MO, except that the migraine was MA and not MO. Furthermore, we added probable menstrual MO and probable menstrual MA, i.e. if the menstrual migraine appendix criteria were fulfilled except for one criterion, similar to that of probable migraine in the ICHD III beta version, but attacks were as a minimum required in 50% of menstrual cycles.

Statistics

The data were analysed using the Statistical Package of Social Sciences 20. Lifetime prevalence is presented with 95% confidence interval. Data from two Danish epidemiological surveys were used in the adjusted prevalence calculations of migraine (2,16).

Ethics

The study was approved by the Regional Committee for Medical Research Ethics and the Data Protection Authorities. The participants received written and verbal information about the project and inclusion was based on informed consent.

Results

Screening questionnaire

Figure 1 is a flow chart of the study. The sample size was reduced to 4802 due to error in the address list (n = 178), emigration (n = 15) and multi-handicap (n = 5). The response rate of the screening questionnaire was 73.2% (3514/4802). A total of 399 responders fulfilled the inclusion criteria, of whom 39 (9.7% 39/399) were excluded due to chronic headache.

Figure 1.

Flow chart of the study.

Interview

The study population of 360 was reduced to 308 eligible women, since four emigrated, nine had insufficient Norwegian language skills and 39 did not reply to five telephone calls or more. Altogether, 71 declined to participate due to lack of time, no interest or acute illness.

The participation rate of the interview was 76.9% (237/308), of whom 63.3% (150/237) were interviewed at the clinic and 36.7% (87/237) by telephone. Table 1 shows that participants and non-participants were not significantly different regarding self-reported frequency of menstrual migraine.

Table 1.

Comparison of self-reported menstrual migraine among participants and non-participants.

Participants N = 237 % (n)	95% confidence interval	Non-participants N = 123 % (n)	95% confidence interval
Frequency of menstrual migraine
Every time	35.9 (85)	30.0–42.2	33.3 (41)	25.6–42.1
2 out of 3 menstruations	37.1 (88)	31.2–43.4	35.8 (44)	27.9–44.6
1 out of 2 menstruations	27.0 (64)	21.8–33.0	30.9 (38)	23.4–39.5

Prevalence of migraine and menstrual migraine

Questionnaire

The crude lifetime prevalence of self-reported migraine among women was 34.6% (1215/3514) and the adjusted prevalence was 34.9% ((550/590 × 1215/3514) × 3514) + ((24/581 × 2299/3514) × 3514/3514 = 0.349). The 590 had a questionnaire diagnosis of migraine, while 550 also had the diagnosis at the interview while 40 did not. The 581 replied they had not migraine in the questionnaire, but 24 had migraine and 557 did not have migraine according to the interview (2,16). The remaining numbers were from the screening questionnaire based on self-reported migraine and total number of replies.

Interview

The clinical interview confirmed the questionnaire self-reported migraine diagnosis in 96.7% of the participants. A total of 176 of the 237 participants had menstrual migraine, corresponding to 7.6% of all women and 22.0% of all female migraineurs (Table 2). The prevalence of menstrual MO according to the ICHD III beta criteria was 17.6% in female migraineurs and 6.1% in all women. The 13 women with probable menstrual MO did not fulfil the ICHD III beta appendix criteria due to attack frequency between ≤1/2 and <2/3 of the menstrual cycles (n = 6), the attacks started on day 4 of the menstruation (n = 4) or headache characteristics fulfilled all but one of the criteria for MO (n = 3).

Table 2.

Prevalence of menstrual migraine in the general population and among migraineurs. CI denotes confidence intervals.

General population (N = 3514)			Among migraineurs (N = 1215)
n	Crude % (95% CI)	Adjusted¹ Adjusted^a % (95% CI)	Crude % (95% CI)	Adjusted¹ Adjusted^a % (95% CI)
Migraine without aura (MO)
Pure menstrual MO	19	0.5 (0.4–0.8)	0.8 (0.6–1.2)	1.6 (1.0–2.4)	2.4 (1.7–3.4)
Menstrually-related MO	122	3.5 (2.9–4.1)	5.3 (4.6–6.1)	10.0 (8.5–11.9)	15.2 (13.3–17.4)
All menstrual MO (ICHD III beta version criteria)	141	4.0 (3.4–4.7)	6.1 (5.4–6.9)	11.6 (9.9–13.5)	17.6 (15.6–19.9)
Probable pure menstrual MO	3	0.1 (0.0–0.3)	0.1 (0.1–0.3)	0.3 (0.1–0.7)	0.3 (0.1–0.7)
Probable menstrually-related MO	10	0.3 (0.2–0.5)	0.4 (0.3–0.7)	0.8 (0.5–1.5)	0.8 (0.5–1.5)
All probable menstrual MO	13	0.4 (0.2–0.6)	0.6 (0.4–0.9)	1.1 (0.6–1.8)	1.6 (1.1–2.5)
Total menstrual MO	154	4.4 (3.8–5.1)	6.7 (5.9–7.5)	12.7 (10.9–14.7)	19.3 (17.1–21.6)
Migraine with aura (MA)
Pure menstrual MA	1	0.0 (0.0–0.2)	0.1 (0.0–0.2)	0.1 (0.0–0.5)	0.2 (0.0–0.6)
Menstrually-related MA	13	0.4 (0.2–0.6)	0.6 (0.4–0.9)	1.1 (0.6–1.8)	1.7 (1.1–2.5)
All menstrual MA	14	0.4 (0.2–0.7)	0.6 (0.4–0.9)	1.2 (0.7–1.9)	1.7 (1.1–2.6)
Probable pure menstrual MA	1	0.0 (0.0–0.2)	0.1 (0.0–0.2)	0.1 (0.0–0.5)	0.2 (0.0–0.6)
Probable menstrually-related MA	7	0.2 (0.1–0.4)	0.3 (0.2–0.6)	0.6 (0.3–1.2)	0.9 (0.5–1.6)
All probable menstrual MA	8	0.2 (0.1–0.5)	0.3 (0.2–0.6)	0.7 (0.3–1.3)	1.0 (0.6–1.7)
Total menstrual MA	22	0.6 (0.4–1.0)	0.9 (0.7–1.3)	1.8 (1.2–2.7)	2.7 (1.9–3.8)
All menstrual migraine	176	5.0 (4.3–5.8)	7.6 (6.8–8.5)	14.5 (12.6–16.6)	22.0 (19.7–24.4)

CI: confidence intervals; ICHD: International Classification of Headache Disorders.

Example ^aExample of calculation of adjusted prevalence of menstrual migraine: ((122/237) × 123) + 122/3514 = 0.053. The number of women with menstrually-related MO among participants (122) divided by all participants (237), multiplied by number of non-participants (123), gives the assumed number of women with menstrually-related MO among non-participants (63), adding the 122 women with menstrually-related migraine among participants, gives the number of affected women in the study population (185). Divided into all responders, this gives an adjusted prevalence 5.3.

Eight women had probable menstrual MA due to attack frequency between ≤1/2 and <2/3 of the menstrual cycles (n = 5) or the attack started on day 4 of menstruation (n = 3).

Among the 36 women with menstrual migraine and co-occurrence of MO and MA, 31 had menstrual MO and five had menstrual MA.

Altogether, 23 women reported headache meeting the criteria for tension-type headache in relation to their menstruation. Among these, 18 had tension-type headache in at least 2/3 menstruations on days 1 ± 2, while the remaining five had it in 1/2 of menstruations on day 1 ± 2. Sixteen of the women had co-occurrence of migraine (13 MO, two MA and one co-occurrence of MO and MA), but their migraine was not related to menstruation.

Demographic data

Data from Statistics Norway show that the women with menstrual migraine were representative to the age–gender matched Norwegian population regarding income, employment rate and marital status. Regarding profession, office jobs were over-represented, while sales and service were under-represented. Participants with menstrual migraine had a higher educational level compared to the general population.

Discussion

Our main findings are a 7.6% prevalence of any menstrual migraine in women and 22.0% in female migraineurs. The vast majority had menstrual MO according to the ICHD III beta appendix criteria, but one of five women had probable menstrual MO, menstrual MA or probable menstrual MA not encompassed in the ICHD III beta appendix criteria.

Methodological considerations

The sampling area was chosen due to its proximity to Akershus University Hospital. The age range was chosen in order to include fertile women at an age when the prevalence of migraine is high (1,6).

Strengths of our study are that participants were recruited from a large population-based sample and that we applied the ICHD III beta appendix criteria on a general population. The spectrum of the ICHD III beta appendix criteria for menstrual MO was additionally expanded to include probable menstrual MO, menstrual MA and probable menstrual MA. This expansion was performed because the criteria remained in the appendix of the recently published ICHD III beta and further research is warranted. Our study describes the whole spectrum of menstrual migraine and, if such data are not explored, the diagnostic criteria will never be revised based on the full scientific body of knowledge, but only on the already existing data.

Furthermore, the screening question ‘Have you ever had migraine?’ has proved valid due to a high chance-corrected agreement rate, kappa 0.87, and it ascertain approximately 80% (90/114) of all migraineurs from the general population (2,16). Our screening questionnaire encompassed all those who replied yes to the migraine question, but also included some who replied no to migraine, but yes to the menstrual migraine screening question in order to further increase ascertainment of migraineurs from the general population.

A limitation of retrospective studies is recall bias. The headache diagnoses followed the ‘gold standard’, i.e. a semi-structured interview by a physician with expertise in headache diagnostics. The temporal relationship between menstruation and migraine should, according to the ICHD III beta appendix criteria for menstrual migraine, be based on prospectively recorded evidence over at least three consecutive cycles. Diary cards are especially helpful in clinical settings, but in epidemiological studies it provides some challenges. The menstrual relation may change over a woman’s reproductive lifetime and some may cease to have menstrual migraine in the period filling in diary cards. Furthermore, compliance is low (17). In order to have information from all women with potential menstrual migraine, we based our study on retrospective information and the exact timing may in some cases not be accurate. One way or the other, neither is perfect for an epidemiological survey. Another limitation is the fact that our study population was interviewed 6 years after questionnaire screening. Thus, we missed women who had onset of menstrual migraine after the screening and before the clinical interview. However, our prevalence data are still valid for women aged 30–34 years, but represent a minimum prevalence figure for women age 36–40 years. The time delay between the screening questionnaire and the clinical interview may also have caused a lower participation rate, although it was 77%.

We assumed that participants and non-participants were equally affected by menstrual migraine due to a similar reply pattern regarding self-reported menstrual migraine and frequency of menstrual migraine in the questionnaires. These data were used for the adjusted prevalence calculations for menstrual migraine.

We excluded 39 women with chronic headache, i.e. ≥180 headache days/year, since it is impossible to judge whether an attack of migraine in the peri-menstrual period occurs by chance or represents a ‘true’ attack of menstrual migraine. If we assume that the women with chronic headache had menstrual migraine similar to the other 74% (176/237) with self-reported menstrual migraine, we would identify an additional 29 persons with this diagnosis and the prevalence would increase 9.7%.

Discussion of results

Our 34.9% lifetime prevalence of self-reported migraine among women in their 30 s is comparable to findings in Denmark (30.8%), Canada (31%) and the Netherlands (39.2%)(2,18 –20). Prior to our study, the prevalence of menstrual MO has only been estimated in an American clinic-based study applying the ICHD II appendix criteria (20), which are similar to the ICHD III beta version. The study was based on a questionnaire and diary cards. However, due to selection bias, prevalence data from clinic population are less valid than prevalence from the general population.

Table 3 shows the prevalence of menstrual migraine in different studies. Our prevalence data are similar to Danish and Swedish population-based studies as well as British and American clinic populations, while three Italian clinic populations and a Serbian population-based questionnaire found higher figures (3,20 –25).

Table 3.

Prevalence of menstrual migraine in previous studies.

Prevalence among migraineurs
Country	Year	Study population	Age (year)	Participants N	Design	Definition of PMM and MRM	PMM	MRM	MM
Turkey (29)	2012	General population	Mean 36	2600	Physician interview	PMM: migraine from 3 days before and 5 days after the beginning of menstruation	4.4%		54.3%
Norway (28)	2010	General population	30–44	10,176	Questionnaire	ICHD 2, Appendix	7.7%	13.2%	20.9%
USA (20)	2008	Clinic population	Mean 32	610	Questionnaire and diary cards	ICHD 2, Appendix	12.1%	10.0%	22.1%
The Nether- lands (36)	2003	General population	13–55	1181	Questionnaire	PMM: Migraine attack exclusively on days −2 –+2 MRM: Attacks on days −2 to +2 and at other times	0.85% General population	3 % General population	3.85% General population
Sweden (21)	2003	General population	40–74	728	Physician interview	>75% of attack of MO within the days 1 ± 2			MO: 21.2% MA: 4.0%
Serbia (25)	2002	Female students	18–28	1943	Questionnaire	PMM: Migraine attack exclusively on day 1 ± 2 MRM: Migraine aggravated during menstruation	12.2% Among all women: 1.5%	48.6% Among all women: 6.1%
Italy (24)	2000	Clinic population	18–55	300	Case–control study	PMM: Migraine attack exclusively on day 1 ± 2 MRM: attacks mostly during menstrual period	MO: 3.5% MA: 4.0%	MO: 53.5% MA: 15.0%	MO: 57.0% MA: 19.0%
Italy (27)	1995	Headache clinic	Unknown	268	Prospective diary	PMM: Migraine attack exclusively from 2 days before to 3 days after menstruation MRM: not defined	MO: 13.6% MA: 3.8%	MO: 56.4% MA: 33.3%	MO: 70.1% MA: 37.2%
Denmark (3)	1996	General population	40	1000	Questionnaire and physician interview	Attacks occurring ±48 h from the start of menstruation in ≥90% of the cycles			MO 24.8% MA 8.1%
Italy (23)	1993	Headache clinic	18–70	1300	Retrospective study	PMM: Migraine attacks on day −3 to 3 days after menstruation in ≥90% of the peri-menstrual periods MRM: attacks occurred predominantly in this period	9.1%	50.8%	59.9%
UK (35)	1992	Headache clinic	26–46	14	Physician interview and diary cards	PMM: Migraine attack exclusively on day −3 to +3 MRM: The mean of the total headache activity was greater during the premenstrual and menstrual phases compared to the ovulatory and luteal phases	7.1%	71.4%	78.6%
UK (26)	1990	Headache clinic	17–50	55	Prospective diary	PMM: Migraine attack exclusively on day 1 ± 2 MRM: Increased number of attacks within this period	7.2%	34.5%	41.7%
UK (22)	1975	Headache clinic	≥14	92	Questionnaire	PMM: Migraine occurring regularly just before or during menstruation only MRM: Migraine occurring regularly just before or during menstruation and at other times	14.1%	12.0%	26.1%

ICHD: International Classification of Headache Disorders; PMM: pure menstrual migraine; MRM: menstrually-related migraine; MM: menstrual migraine. MO: migraine without aura; MA: migraine with aura.

The variation reflects different definitions, ascertainment and study populations. The lack of uniform criteria, particularly for the definition of menstrually-related migraine (MRM), results in large prevalence differences, while the prevalence of pure menstrual migraine is reported more uniform. MRM has been defined as self-reported aggravation of migraine during the peri-menstrual period (25), an increased number of attacks in the peri-menstrual period (26), attacks occurring regularly just before or during menstruation (22), migraine attacks occurring mostly during the peri-menstrual period (24), attacks occurring predominantly in the peri-menstrual period without consideration of the exact frequency (23) or definition is lacking (27). Similarly, the peri-menstrual period is defined from 3 days before the first day of menstruation until the last day of menstruation, expanding the peri-menstrual period by several days as compared to ICHD III beta appendix criteria. Several studies exclusively analyse unspecified migraine (22,25,28,29), while other studies calculate the prevalence of menstrual MO and menstrual MA among women with MO and MA, rather than all migraineurs (3,21,23). Assuming that two-thirds of all migraineurs have MO (2), the expected prevalence of menstrual MO would be 26% among women with MO in our sample and the corresponding figures of pure menstrual MO and menstrually-related MO would be 3.6% and 22.8% respectively.

Another issue is different study populations. Clinic populations have more frequent migraine attacks than the general population. A study from the general population found that only 7% had more than 14 days with migraine within the last year, while 16% had 8–14 days with migraine within the last year (30). Co-occurrence of MO and MA is about 50% among migraineurs from clinic populations (3), while in the general population it is similar to the prevalence of MO and MA, i.e. about 5–13% among migraineurs (3,4,18). However, we found that 20.4% of those with menstrual migraine had co-occurrence of MO and MA, which may explain the high attack frequency in some of those with menstrual migraine.

The golden standard of a precise headache diagnose is a clinical interview by a physician. Thus, ascertainment is an important aspect. Questionnaires based on ICHD II headache characteristics in order to provide a diagnosis are not valid, while the single question do you have migraine ‘yes’ or ‘no’ has proved to be valid (2,16). Lay interviews are less valid than interviews by a physician. Diary cards are especially helpful for the precise diagnosis of menstrual migraine but, as discussed above, diaries provide challenges in epidemiological studies. For the purpose of epidemiological surveys, one must rely on the clinical or lay interview or questionnaire, since menstrual migraine may change over a woman’s reproductive life and diary cards will not identify those who ceased to have menstrual migraine (17,31).

The ICHD III beta appendix criteria for menstrual migraine only include MO. We identified 14 women with menstrual MA, representing approximately 5% of all women with MA. In comparison, 26% of all women with MO have menstrual MO. Previous studies have found menstrual MA in 4%, 8% and 37% of women with MA (3,21,27). A prospective diary study found an increased risk of MA on days 1–3 of menstruation (32). Twelve of 14 women with MA reported menstruation to trigger attacks of MA (33). Studies exploring MO and MA in relation to hormonal changes show different patterns with an increase of MA and decrease of MO during conditions associated with high plasma concentrations of oestrogen, such as pregnancy (34). Whether menstrual MA is a real phenomenon or if these attacks occur by chance still remains unclear. If the first is true, there might be other underlying pathophysiological mechanisms than in menstrual MO.

We identified 21 women with probable menstrual MO or menstrual MA. The 11 women with lower than the required attack frequency at menstruation probably represent the borderline of menstrual migraine, where menstruation alone is too weak to trigger an attack.

The seven women with attacks on the 4th day of menstruation may or may not represent the borderline of menstrual migraine. A delayed decline in premenstrual oestrogen levels in these individuals, or prostaglandin-release, may explain the late onset of menstrual attacks. A diary study is likely to clarify that matter.

Twenty-three women had tension-type headache related to menstruation. Since we did not screen for headache in relation to menstruation, our data are not suitable for menstrual tension-type headache prevalence calculation.

Conclusion

Migraine in at least 50% of menstrual cycles occurs in 7.6% of all women and in 22% of women with migraine. It does not exclusively occur in MO, but the ICHD III beta appendix criteria only include MO. Since the appendix criteria are novel entities that have not been sufficiently validated by research studies, it would be appropriate to add probable menstrual MO, menstrual MA and probable menstrual MA into the appendix criteria of the future ICHD III, in order to confirm or refute the association. The criteria can then be used to identify potential pathophysiological mechanisms for these attacks.

Clinical or public health relevance

1/5 female migraineurs have migraine in at least 50% of menstruations.

The vast majority have menstrual MO, while few have menstrual MA.

The ICHD III beta appendix criteria for menstrual migraine are not exhaustive.

Footnotes

Funding

This study is supported by grants from the South-Eastern Norway Regional Health Authority (grant number 2013001), Akershus University Hospital (grant number 2629014) and Institue of Clinical medicine, University of Oslo.

Conflict of interest

None declared.

References

Lipton

Stewart

Diamond

. Prevalence and burden of migraine in the United States: data from the American Migraine Study II. Headache 2001; 41: 646–657.

Russell

Rasmussen

Thorvaldsen

. Prevalence and sex-ratio of the subtypes of migraine. Int J Epidemiol 1995; 24: 612–618.

Russell

Rasmussen

Fenger

. Migraine without aura and migraine with aura are distinct clinical entities: A study of four hundred and eighty-four male and female migraineurs from the general population. Cephalalgia 1996; 16: 239–245.

Russell

Ulrich

Gervil

. Migraine without aura and migraine with aura are distinct disorders. A population-based twin survey. Headache 2002; 42: 332–336.

Bille

. A 40-year follow-up of school children with migraine. Cephalalgia 1997; 17: 488–491. discussion 7.

Steiner

Scher

Stewart

. The prevalence and disability burden of adult migraine in England and their relationships to age, gender and ethnicity. Cephalalgia 2003; 23: 519–527.

Johannes

Linet

Stewart

. Relationship of headache to phase of the menstrual cycle among young women: A daily diary study. Neurology 1995; 45: 1076–1082.

Stewart

Lipton

Chee

. Menstrual cycle and headache in a population sample of migraineurs. Neurology 2000; 55: 1517–1523.

MacGregor

Hackshaw

. Prevalence of migraine on each day of the natural menstrual cycle. Neurology 2004; 63: 351–353.

10.

MacGregor

Frith

Ellis

. Incidence of migraine relative to menstrual cycle phases of rising and falling estrogen. Neurology 2006; 67: 2154–2158.

11.

Somerville

. The role of estradiol withdrawal in the etiology of menstrual migraine. Neurology 1972; 22: 355–365.

12.

Headache Classification Committee of the International Headache Society. Classification and diagnostic criteria for headache disorders, cranial neuralgias and facial pain. Cephalalgia 1988; 8(Suppl. 7): 1–96.

13.

Headache Classification Subcommittee of the International Headache Society. The International Classification of Headache Disorders: 2nd edition. Cephalalgia 2004; 24(Suppl. 1): 9–160.

14.

Headache Classification Subcommittee of the International Headache Society. The International Classification of Headache Disorders: 3rd edition (beta version). Cephalalgia 2013; 33(9): 629–808.

15.

Statistics Norway (2012). Available at: http://statbank.ssb.no/statistikkbanken (accessed 15 January 2013).

16.

Rasmussen

Jensen

Olesen

. Questionnaire versus clinical interview in the diagnosis of headache. Headache 1991; 31: 290–295.

17.

Russell

Rasmussen

Brennum

. Presentation of a new instrument: The diagnostic headache diary. Cephalalgia 1992; 12: 369–374.

18.

Launer

Terwindt

Ferrari

. The prevalence and characteristics of migraine in a population-based cohort: The GEM study. Neurology 1999; 53: 537–542.

19.

O'Brien

Goeree

Streiner

. Prevalence of migraine headache in Canada: A population-based survey. Int J Epidemiol 1994; 23: 1020–1026.

20.

Tepper

Zatochill

Szeto

. Development of a simple menstrual migraine screening tool for obstetric and gynecology clinics: the menstrual migraine assessment tool. Headache 2008; 48: 1419–1425.

21.

Mattsson

. Hormonal factors in migraine: a population-based study of women aged 40 to 74 years. Headache 2003; 43: 27–35.

22.

Epstein

Hockaday

. Migraine and reporoductive hormones throughout the menstrual cycle. Lancet 1975; 1(7906): 543–548.

23.

Granella

Sances

Zanferrari

. Migraine without aura and reproductive life events: A clinical epidemiological study in 1300 women. Headache 1993; 33: 385–389.

24.

Granella

Sances

Pucci

. Migraine with aura and reproductive life events: a case control study. Cephalalgia 2000; 20: 701–707.

25.

Dzoljic

Sipetic

Vlajinac

. Prevalence of menstrually related migraine and nonmigraine primary headache in female students of Belgrade University. Headache 2002; 42: 185–193.

26.

MacGregor

Chia

Vohrah

. Migraine and menstruation: A pilot study. Cephalalgia 1990; 10: 305–310.

27.

Cupini

Matteis

Troisi

. Sex-hormone-related events in migrainous females. A clinical comparative study between migraine with aura and migraine without aura. Cephalalgia 1995; 15: 140–144.

28.

Vetvik

MacGregor

Lundqvist

. Self-reported menstrual migraine in the general population. J Headache Pain 2010; 11: 87–92.

29.

Karli

Baykan

Ertas

. Impact of sex hormonal changes on tension-type headache and migraine: A cross-sectional population-based survey in 2,600 women. J Headache Pain 2012; 13: 557–565.

30.

Rasmussen

Olesen

. Migraine with aura and migraine without aura: An epidemiological study. Cephalalgia 1992; 12: 221–228. discussion 186.

31.

Russell

Iversen

Olesen

. Improved description of the migraine aura by a diagnostic aura diary. Cephalalgia 1994; 14: 107–117.

32.

Salhofer-Polanyi

Frantal

Brannath

. Prospective analysis of factors related to migraine aura – the PAMINA study. Headache 2012; 52: 1236–1245.

33.

Hauge

Kirchmann

Olesen

. Characterization of consistent triggers of migraine with aura. Cephalalgia 2011; 31: 416–438.

34.

MacGregor

. Oestrogen and attacks of migraine with and without aura. Lancet Neurol 2004; 3: 354–361.

35.

Beckham JC, Krug LM, Penzien DB, et al. The relationship of ovarian steroids, headache activity and menstrual distress: A pilot study with female migraineurs. Headache 1992; 32(6): 292–297.

36.

Couturier EG, Bomhof MA, Neven AK, et al. Menstrual migraine in a representative Dutch population sample: Prevalence, disability and treatment. Cephalalgia 2003; 23(4): 302–308.