Clinical characteristics and social burden of menstrual migraine in Japan: an online-based cross-sectional questionnaire study.

This was a cross-sectional, population-based web survey conducted among patients with migraine across Japan. Participants were randomly selected from a registry administered by Macromill Carenet Inc. (Tokyo, Japan), which includes over 1.3 million individuals (aged 20–59 years: 83%; female: 66%). Random sampling was carried out using a computer-generated random number algorithm. Eligible participants who met the inclusion criteria and did not meet the exclusion criteria, based on a screening assessment questionnaire, were invited to complete the survey through an online platform provided by Macromill Carenet Inc. To ensure an adequate number of responses for MM, the sampling process was controlled to limit the proportion of respondents aged 60 years and older to less than 20%. Data were collected from November 14 to November 25, 2024. Only data from female participants were included in the final analysis. The final sample of Japanese females aged 20–59 years in the Macromill Carenet panel appeared broadly comparable to the general Japanese female population in terms of age and region, which could help support the relevance of analyses based on this sample. The study enrolled females with migraine aged 18 years or older at the time of the survey who met the modified ICHD-3 and provided electronically documented informed consent to participate. In this study, individuals with comorbidities including psychiatric or neurological disorders for which medications could affect migraine severity (including major depressive disorder or epilepsy), underlying medical conditions, or concomitant use of medications that could potentially influence the pathophysiology or pain characteristics of migraine (e.g., cancer-related pain, fibromyalgia, chronic pelvic pain, or complex regional pain syndrome) were excluded. Accordingly, individuals were excluded if they were pregnant, were undergoing sex hormone therapy (e.g., for endometriosis or dysmenorrhea), experienced frequent migraine attacks (defined as ≥ 15 migraine days per month), were diagnosed with secondary headaches or cluster headaches. For the primary analysis comparing the MM and non-MM groups, only patients with regular menstruation were included, thereby excluding those approaching menopause who might have experienced headaches as a manifestation of perimenopausal symptoms. The patients were categorized into the MM group—defined as those experiencing migraine attacks during the perimenstrual period (days −2 to + 3 of the onset of menstruation) in at least two out of three menstrual cycles—and the non-MM group, which included all other women with migraine [ 1 ]. Menstrual status was classified as regular, irregular, or postmenopausal, and only participants with regular menstruation were included in the MM group. The MM group was further categorized into PMM or MRM subgroups. The PMM subgroup comprised patients who met the criteria for MM and did not experience migraine attacks at any other time during the menstrual cycle [ 1 ]. The MRM subgroup included patients who fulfilled the criteria for PMM and also experienced migraine attacks at times unrelated to menstruation. Within the non-MM group, a subgroup designated as insufficient-MRM was defined to include patients who experienced a migraine attack during only one of three menstrual cycles and also reported attacks occurring outside the perimenstrual period. The study assessed baseline characteristics, including age, body mass index (BMI), medical history, and employment status, as well as other clinical features such as migraine type (migraine with aura or migraine without aura), monthly migraine days (MMD), monthly headache days (MHD), age at disease onset, disease duration, smoking status, alcohol consumption, family history, migraine triggers, aura, associated symptoms, the relationship between menstrual cycle and headache, and the frequency of MM during menstrual cycle. Primary outcomes included pain intensity, pain duration, and migraine frequency during the menstrual period, as well as the social burden of migraine. Pain intensity was assessed using the Visual Analog Scale (VAS). The social burden of migraine was assessed using MIDAS, Migraine Interictal Burden Scale (MIBS-4), Migraine-Specific Quality-of-Life Questionnaire version 2.1 (MSQ), and Work Productivity and Activity Impairment for Migraine (WPAI-M) scores [ 12 – 16 ]. Medication use was analyzed based on prescription and over-the-counter (OTC) medication use, number of days the prescription medication was taken for acute treatment, and OTC medication frequency. Prescription medications were also analyzed based on types, including triptans, lasmiditan, acetaminophen, and non-steroidal anti-inflammatory drugs (NSAIDs). Headache frequency and medication use during the perimenstrual and nonperimenstrual periods, along with changes in medication usage and type, were also evaluated. In addition, within the subgroup of patients with MRM in the MM group and patients in the non-MM group who experienced migraine during both menstrual and non-menstrual periods (insufficient-MRM subgroup), pain intensity and migraine attack duration were compared between the menstrual and non-menstrual periods. Data from all eligible participants who completed the survey were analyzed. Baseline characteristics, primary outcomes, and medication status in the non-MM and MM groups were summarized descriptively. Pain intensity, pain duration, and headache frequency in both groups were also summarized descriptively by menstrual periods. For pain intensity (VAS), MIDAS, MIBS-4, MSQ, and WPAI-M scores, comparative analyses between the MM and non-MM groups were conducted by adjusting potential confounders using linear regression models with inverse probability of treatment weighting (IPTW), aiming to estimate the average treatment effect (ATE) [ 17 ]. The propensity scores for IPTW were estimated using a logistic regression model for the non-MM and MM groups with potential confounders including age, migraine onset age, smoking status, alcohol consumption, BMI, treatment status, working status, and MMD, and stabilized weights were used. No trimming of extreme propensity scores was conducted. These covariates were selected based on a previous study [ 18 ], and additional variables considered relevant to the severity of migraine from the clinical perspective, in order to adjust for potential confounders. As an exploratory analysis, in the MRM and insufficient-MRM subgroups, pain intensity and migraine attack duration were compared between the menstrual and non-menstrual periods within patients using a paired t-test. Statistical tests were conducted at a significance level of 0.05. No imputations for missing values were planned, as participants were required to complete all applicable questionnaire on the web survey to submit their responses. The study data were managed in accordance with Japan’s Personal Information Protection Law. The study was conducted following approval from the ethics committee of the MINS (a specified nonprofit organization; approval ID: MINS-REC-240229). Patient consent was obtained before study participation, and participants were allowed to withdraw their consent at any time by contacting the Macromill Monitor Support System without disclosing the reason.

Of the 266,392 patients screened, those with a history of medical conditions ( n  = 18,089), pregnant women ( n  = 3,085), patients on sex hormone therapy ( n  = 6,625), those not diagnosed with migraine based on the study criteria for migraine ( n  = 211,990), and individuals who did not provide informed consent ( n  = 7,837) were excluded from the study (Fig. 1 ). A total of 18,750 patients (including 4,161 men) provided consent and completed the survey. Among 14,589 female patients, 4,592 who had a diagnosis of MM, per the ICHD-3 criteria were included in the analysis as the MM group, and 5,174 who had non-MM with regular menstruation were included in the analysis as the non-MM group. Fig. 1 Patient disposition and group creation. ICHD-3, International Classification of Headache Disorders, 3rd edition; MM, menstrual migraine; MRM, menstrually-related migraine; PMM, pure menstrual migraine Patient disposition and group creation. ICHD-3, International Classification of Headache Disorders, 3rd edition; MM, menstrual migraine; MRM, menstrually-related migraine; PMM, pure menstrual migraine Of the 4,592 patients with MM, 2,114 (46.0%) were categorized into the MRM subgroup, and 2,478 (54.0%) were categorized into the PMM subgroup. Among the 5,174 patients in the non-MM group, 783 were further classified into the insufficient-MRM subgroup, which comprised females with migraine who experienced migraine attacks during only one of three menstrual cycles and also had attacks outside the menstrual period (Fig. 1 ). The mean±standard deviation (SD) age was 36.8 ± 8.0 years (median [first quartile, Q1–third quartile, Q3]: 38.0 years [31.0–43.0]; range: 18.0–58.0 years) in the MM group ( N  = 4,592) and 35.5 ± 8.3 years (median [Q1–Q3]: 36.0 years [30.0–42.0]; range: 18.0–74.0 years) in the non-MM group ( N  = 5,174) (Table 1 ). The mean±SD of MHD was 5.9 ± 4.5 in the MM group and 4.5 ± 3.9 in the non-MM group, respectively. The proportion of patients with a history of gynecological conditions and sleep disorders was higher in the MM group than in the non-MM group. Premenstrual syndrome was also more frequently reported in the MM group than in the non-MM group. Similarly, the prevalence of uterine fibroids and dysmenorrhea was also higher in the MM group than in the non-MM group. Most patients in both the MM (65.1%) and non-MM (69.3%) groups never received a formal diagnosis of headache or visited a doctor for their headache symptoms (Supplementary Table S1 ). Overall, 55.7% of patients in the MM group reported 3 episodes, and 44.3% reported 2 episodes of migraine during their 3 consecutive menstrual cycles (Table 1 ). Among patients in the non-MM group who experienced headache during perimenstrual period, the majority of patients (89.6%, 1,780/1,986) had 1 episode of migraine, while the remaining patients (10.4%; 206/1,986) had no migraine episodes during their 3 consecutive menstrual cycles. The proportion of patients with aura (MM; 26.6%, non-MM; 22.0%, p  < 0.001) and associated symptoms (nausea; 63.4% in MM and 59.2% in non-MM, phonophobia; 58.3% in MM and 50.0% in non-MM, photophobia; 58.2% in MM and 51.3% in non-MM, osmophobia; 29.3% in MM and 21.1% in non-MM, all p  < 0.001), was higher in the MM group than in the non-MM group (Table 1 ). Differences in detailed characteristics such as employment status, age at migraine onset, smoking status, alcohol consumption, hospital visits due to headache, and family history between the non-MM and MM groups are presented in Supplementary Table S1 . Table 1 Patient demographics and clinical characteristics Characteristics Non-MM group (N = 5,174) MM group (N = 4,592) p-value* Age (years)  Mean±SD 35.5±8.3 36.8±8.0  < 0.001 BMI (kg/m 2 )  Mean±SD 21.4±3.9 21.4±3.9 0.600   < 25 4,444 (85.9) 3,996 (87.0) 0.234  ≥ 25 to < 30 551 (10.6) 457 (10.0)  ≥ 30 179 (3.5) 139 (3.0) MHD  Mean±SD 4.5±3.9 5.9±4.5  < 0.001 MMD  Mean±SD 2.7±2.3 3.4±2.6  < 0.001 Aura  Yes 1,137 (22.0) 1,221 (26.6)  < 0.001 Associated symptoms  Nausea 3,062 (59.2) 2,910 (63.4)  < 0.001  Vomiting 882 (17.0) 775 (16.9) 0.824  Phonophobia 2,587 (50.0) 2,678 (58.3)  < 0.001  Photophobia 2,654 (51.3) 2,671 (58.2)  < 0.001  Osmophobia 1,094 (21.1) 1,345 (29.3)  < 0.001 Medical history  Sleep disorder 173 (3.3) 212 (4.6) 0.001  Anxiety 238 (4.6) 232 (5.1) 0.297  Gastric ulcer/gastrointestinal bleeding 15 (0.3) 15 (0.3) 0.743  Diabetes 61 (1.2) 38 (0.8) 0.084  Hypertension 132 (2.6) 145 (3.2) 0.072  Uterine fibroids 318 (6.1) 396 (8.6)  < 0.001  Endometriosis 156 (3.0) 182 (4.0) 0.010  Adenomyosis 48 (0.9) 67 (1.5) 0.015  Gynecological cancer 59 (1.1) 52 (1.1) 0.971  Dysmenorrhea 171 (3.3) 263 (5.7)  < 0.001  PMS 1,054 (20.4) 1,509 (32.9)  < 0.001  None of these 3,321 (64.2) 2,410 (52.5)  < 0.001 Triggers for migraine  Stress 2,364 (45.7) 2,635 (57.4)  < 0.001  Relief from stress 480 (9.3) 530 (11.5)  < 0.001  Menstruation 1,523 (29.4) 3,539 (77.1)  < 0.001  Emotional ups and downs (overexcitement or mood swings) 645 (12.5) 919 (20.0)  < 0.001  Fatigue and malaise 2,679 (51.8) 2,885 (62.8)  < 0.001  Change in the season 1,628 (31.5) 2,027 (44.1)  < 0.001  Change in climate 2,861 (55.3) 3,015 (65.7)  < 0.001  Poor posture 1,124 (21.7) 1,278 (27.8)  < 0.001  Overexercise 209 (4.0) 324 (7.1)  < 0.001  Irregular eating habits 174 (3.4) 239 (5.2)  < 0.001  Alcohol consumption 467 (9.0) 667 (14.5)  < 0.001  Caffeine consumption 230 (4.4) 297 (6.5)  < 0.001  Consumption of specific foods 66 (1.3) 87 (1.9) 0.014  Exposure to light 769 (14.9) 929 (20.2)  < 0.001  Noise 364 (7.0) 580 (12.6)  < 0.001  Medication 44 (0.9) 51 (1.1) 0.191  Irregular sleep 1,768 (34.2) 1,965 (42.8)  < 0.001  Certain smells (e.g., gasoline or perfume) 493 (9.5) 650 (14.2)  < 0.001  Not sure/no trigger 676 (13.1) 220 (4.8)  < 0.001 Relationship between menstrual cycle and headache  It is related 2,192 (42.4) 4,276 (93.1)  < 0.001  It is not related 2,982 (57.6) 316 (6.9)  < 0.001 MM episodes during 3 menstrual cycles  No episode 206 (10.4) 0 (0.0) -  1 episode 1,780 (89.6) 0 (0.0) -  2 episodes 0 (0.0) 2,034 (44.3) -  3 episodes 0 (0.0) 2,558 (55.7) - *The t-test for continuous variables and exact test for categorical variables were performed at a significant level of 0.05. Data are presented as n (%), unless otherwise specified. BMI, body mass index; MHD, monthly headache days; MM, menstrual migraine; MMD, monthly migraine days; PMS, premenstrual syndrome; SD, standard deviation Patient demographics and clinical characteristics *The t-test for continuous variables and exact test for categorical variables were performed at a significant level of 0.05. Data are presented as n (%), unless otherwise specified. BMI, body mass index; MHD, monthly headache days; MM, menstrual migraine; MMD, monthly migraine days; PMS, premenstrual syndrome; SD, standard deviation The mean±SD VAS score for pain intensity during both perimenstrual and non-perimenstrual periods were higher ( p  < 0.001) in the MM group versus the non-MM group (Fig. 2 A). The greatest difference in pain intensity between the MM group and non-MM group was at the start of the menstruation (mean score difference: 7.7; 95% confidence interval [CI]: 6.0, 9.4, p  < 0.001). Pain intensity in the MM group peaked 1 day before the start of the menstrual period, with a mean±SD VAS score of 57.5 ± 19.7. In the non-MM group, migraine pain intensity peaked 1 day after the start of the menstrual period, with a mean±SD VAS score of 51.8 ± 19.1 (Fig. 2 A). After IPTW, standard mean differences (SMD) for all covariates between the compared group were < 0.1, except for age (SMD for age, 0.11). Fig. 2 ( A ) Pain intensity and ( B ) headache frequency during menstrual periods in the non-MM and MM groups. Error bars represent standard deviation. The difference (95% CI) and p-value are presented for the outcome of the comparison between MM and non-MM groups following IPTW adjustment. CI, confidence interval; IPTW, inverse probability of treatment weighting; MM, menstrual migraine; VAS, Visual Analog Scale ( A ) Pain intensity and ( B ) headache frequency during menstrual periods in the non-MM and MM groups. Error bars represent standard deviation. The difference (95% CI) and p-value are presented for the outcome of the comparison between MM and non-MM groups following IPTW adjustment. CI, confidence interval; IPTW, inverse probability of treatment weighting; MM, menstrual migraine; VAS, Visual Analog Scale The mean±SD pain duration (in hours) during both perimenstrual and non-perimenstrual periods were longer in the MM versus the non-MM group (Supplementary Figure S1 ). Pain duration in the MM group peaked at the start of the menstrual period, with a mean±SD pain duration of 7.0 ± 8.0 (Supplementary Figure S1 ). In the non-MM group, pain duration peaked 2 days before the start of the menstrual period, with a mean±SD pain duration of 5.0 ± 7.5 (Supplementary Figure S1 ). Two days before the start of menstruation, 73.1% (2,001/2,737) of patients in the MM group experienced “extremely frequent” or “frequent” headaches compared with those in the non-MM group (37.2%; 412/1,109) (Fig. 2 B). Similarly, a higher proportion of patients in the MM group (77.3%; 2,324/3,007) experienced “extremely frequent” or “frequent” headaches 1 day before menstrual period compared with those in the non-MM group (41.2%; 381/924) (Fig. 2 B). A higher proportion of patients reported migraine triggers in the MM group than in the non-MM group (Table 1 ). The top 3 migraine triggers (excluding menstruation)—climate change, fatigue/malaise, and stress—were similar between groups. However, menstruation was the most frequently reported trigger in the MM group (77.1%), whereas it ranked sixth among the 18 triggers studied in the non-MM group, with a frequency of 29.4% (Table 1 ). Moreover, the MM group reported a higher frequency of each trigger compared with the non-MM group. The majority of patients (93.1%) in the MM group felt that their migraine was related to menstruation, compared with 42.4% of patients in the non-MM group (Table 1 ). Overall, the proportion of patients prescribed acute medications was 33.9% in the MM group and 27.1% in the non-MM group (Table 2 ). Among acute medications, the most frequently prescribed drugs were NSAIDs (MM group: 51.4%; non-MM group: 45.4%), followed by acetaminophen (43.9%; 43.2%), triptans (26.2%; 24.0%) and lasmiditan (3.7%; 1.9%). The proportion of patients prescribed oral and injectable preventive medications was 36.4% and 1.6% in the MM group, respectively, compared with 32.5% and 1.5% in the non-MM group. More patients in the MM group (54.0%) reported an increase in medication usage during the perimenstrual period compared with those in the non-MM group (24.0%) (Table 2 ). Moreover, 13.2% of patients in the MM group reported an increase in the types of medications used during the perimenstrual period, compared with 7.1% in the non-MM group. Table 2 Medication status and change in medication status Table 2(a). Medication status Medication status Non-MM group (N = 5,174) MM group (N = 4,592) Prescription of acute medication  Yes 1,402 (27.1) 1,557 (33.9)  No 680 (13.1) 484 (10.5)  I have not been to the hospital 3,092 (59.8) 2,551 (55.6) Description of the prescribed acute medication  n 1,402 1,557  Triptan 336 (24.0) 408 (26.2)  Lasmiditan 26 (1.9) 58 (3.7)  Acetaminophen 606 (43.2) 684 (43.9)  NSAIDs 636 (45.4) 801 (51.4)  Prescribed but not sure of its classification 187 (13.3) 161 (10.3) Prescribed acute medication (days per month)  n 1,402 1,557  Mean±SD 3.0±2.8 4.0±3.5 Use of OTC medicines  Yes 4,160 (80.4) 3,813 (83.0)  No 1,014 (19.6) 779 (17.0) OTC medicine use (days per month)  n 4,160 3,813  Mean±SD 2.9±2.9 3.9±3.5 Prescription of oral preventive medication  n 2,082 2,041  Yes 676 (32.5) 743 (36.4)  No 1,406 (67.5) 1,298 (63.6) Prescription of injectable preventive medication  Yes 31 (1.5) 33 (1.6)  No 2,051 (98.5) 2,008 (98.4) Table 2(b) Change in medication status Change in medication status Non-MM group (N = 705) MM group (N = 2,015) Change in usage of medication during perimenstrual period  Increased usage of medicines during perimenstrual period 169 (24.0) 1,088 (54.0)  Decreased usage of medicines during perimenstrual period 51 (7.2) 81 (4.0)  No change 485 (68.8) 846 (42.0) Change in medication type due to menstruation  Used more types of medicines 50 (7.1) 265 (13.2)  Used fewer types of medicines 28 (4.0) 36 (1.8)  No change 627 (88.9) 1,714 (85.1) Data are presented as n (%), unless otherwise specified. MM, menstrual migraine; NSAID, nonsteroidal anti-inflammatory drug; OTC, over-the-counter; SD, standard deviation Medication status and change in medication status Data are presented as n (%), unless otherwise specified. MM, menstrual migraine; NSAID, nonsteroidal anti-inflammatory drug; OTC, over-the-counter; SD, standard deviation MIDAS score (mean±SD) (8.1 ± 10.8 vs 5.3 ± 8.0) and MIBS-4 score (mean±SD) (3.7 ± 3.3 vs 3.1 ± 3.3) were higher (all p  < 0.001) in the MM group than in the non-MM group (Fig. 3 A). MSQ scores (mean±SD) for each domain were higher in non-MM group than in the MM group ( p  < 0.001): restriction (77.3 ± 17.8 vs 72.5 ± 17.3), interference (85.3 ± 18.6 vs 82.8 ± 18.3), and emotion (82.2 ± 20.2 vs 77.6 ± 20.6) (Fig. 3 B). The WPAI-M score demonstrated that MM group had higher (all p  < 0.001) overall work impairment (36.8 ± 24.2 in MM vs 30.1 ± 23.9 in non-MM), activity impairment (37.4 ± 24.1 vs 30.6 ± 23.4), and presenteeism (35.4 ± 23.2 vs 28.8 ± 22.9) than the non-MM group (Fig. 3 C). However, absenteeism remained similar between the MM and non-MM groups (3.5 ± 11.8 vs 3.2 ± 12.7; p  = 0.590) (Fig. 3 C). Fig. 3 Burden of MM ( A ) MIBS-4 and MIDAS scores ( B ) MSQ score and ( C ) WPAI-M score. Error bars represent standard deviation. The difference (95% CI) and p-value are presented for the outcome of the comparison between MM and non-MM groups following IPTW adjustment. CI, confidence interval; IPTW, inverse probability of treatment weighting; MIBS-4, Migraine Interictal Burden Scale; MIDAS, Migraine Disability Assessment; MM, menstrual migraine; MSQ, Migraine-Specific Quality-of-Life Questionnaire version 2.1; WPAI M, Work Productivity and Activity Impairment for Migraine. Burden of MM ( A ) MIBS-4 and MIDAS scores ( B ) MSQ score and ( C ) WPAI-M score. Error bars represent standard deviation. The difference (95% CI) and p-value are presented for the outcome of the comparison between MM and non-MM groups following IPTW adjustment. CI, confidence interval; IPTW, inverse probability of treatment weighting; MIBS-4, Migraine Interictal Burden Scale; MIDAS, Migraine Disability Assessment; MM, menstrual migraine; MSQ, Migraine-Specific Quality-of-Life Questionnaire version 2.1; WPAI M, Work Productivity and Activity Impairment for Migraine. In the MRM subgroup, VAS score was higher during the perimenstrual period compared with the non-perimenstrual period (63.1 ± 19.5 vs 56.0 ± 20.1; p  < 0.001) (Table 3 ). In the insufficient-MRM subgroups, there was no significant difference in VAS score between perimenstrual period and non-perimenstrual period (52.9 ± 19.4 vs 53.6 ± 19.4; p  = 0.322). Pain duration was similar within the MRM subgroup during the perimenstrual period compared with the non-perimenstrual period (8.3 ± 9.1 vs 8.2 ± 11.4; p  = 0.438) (Table 3 ). Table 3 Comparison between the perimenstrual and non-perimenstrual periods within patients in the MRM and insufficient-MRM subgroups Pain intensity and pain duration Comparison within the MRM subgroup (N = 2,114) Comparison within the insufficient-MRM subgroup (N = 783) Pain intensity (VAS score) Perimenstrual period, mean±SD 63.1±19.5 52.9±19.4 Non-perimenstrual period, mean±SD 56.0±20.1 53.6±19.4 Difference (95% CI) 7.1 (6.2, 8.0) −0.7 (−2.0, 0.7) p-value  < 0.001 0.322 Pain duration (hours) Perimenstrual period, mean±SD 8.3±9.1 6.0±7.7 Non-perimenstrual period, mean±SD 8.2±11.4 6.5±9.2 Difference (95% CI) 0.2 (−0.2, 0.5) −0.5 (−1.0, −0.0) p-value 0.438 0.033 CI, confidence interval; MRM, menstrually-related migraine; SD, standard deviation; VAS, Visual Analog Scale Comparison between the perimenstrual and non-perimenstrual periods within patients in the MRM and insufficient-MRM subgroups CI, confidence interval; MRM, menstrually-related migraine; SD, standard deviation; VAS, Visual Analog Scale

According to the International Classification of Headache Disorders, 3rd edition (ICHD-3), menstrual migraine (MM) is defined in the appendix as migraine attacks that occur during the perimenstrual period in at least two out of three consecutive menstrual cycles [ 1 ]. MM is classified into two subtypes: pure menstrual migraine (PMM), which occurs exclusively during the perimenstrual period—defined as two days before to the third day of menstruation where day 1 reflects the first day of bleeding—and menstrually-related migraine (MRM), which occurs both during the perimenstrual period and at other times throughout the menstrual cycle [ 1 ]. Both PMM and MRM may present with or without aura. The reported prevalence of MM in Japanese female varies widely, ranging from 10.6% to 41.4% [ 2 , 3 ]. Previous studies have demonstrated an association between migraine and menstruation, particularly in relation to fluctuations in sex hormone levels [ 4 , 5 ]. Silberstein and Merriam reported that estrogen withdrawal could be a major pathogenic trigger for MM [ 4 ]. Raffaelli et al. demonstrated that hormonal fluctuations, a decline in estrogen level, are associated with increased susceptibility to migraine during menstruation [ 5 ]. Episodes of MM are typically more debilitating, longer in duration, and less responsive to acute or preventive treatment than non-MM episodes [ 6 – 8 ]. A previous study reported that 56.2% of women with MM had moderate-to-severe disability associated with their migraine (mean Migraine Disability Assessment [MIDAS] score: 24.8 ± 34.8) [ 8 ]. Additionally, a cross-sectional study conducted in China reported that patients with MM had poorer health-related quality of life (QOL) in the majority of scores measured by SF-36 compared with non-MM patients [ 9 ]. In the study, they also reported that headache frequency, the impact of headaches on daily life, depressive symptoms, social support, and suicidal ideation were associated with health-related QOL in patients with MM [ 9 ]. Importantly, there is a significant lack of evidence on the clinical characteristics and burden of MM in real-world clinical practice in Japan, despite the expected increased burden associated with menstruation on females with migraine. Additionally, the diagnosis and treatment of MM are often inadequate in clinical settings [ 9 – 11 ]. Clarifying the clinical characteristics of MM could facilitate individualized treatment and reduce the burden on patients, ultimately improving their QOL. Therefore, this study aimed to describe the clinical characteristics of patients with MM compared with those with non-MM and to evaluate the social burden of MM in Japanese females.

To the best of our knowledge, this cross-sectional study is the first to clarify the clinical characteristics and burden of MM relative to non-MM in the Japanese population. Our findings revealed that individuals with MRM experienced higher pain intensity during the perimenstrual period compared to non-perimenstrual period. This increased pain intensity was accompanied by increased medication use, suggesting that patients with MM may face challenges in achieving adequate pain control during menstruation. Furthermore, various assessment scales, such as MIDAS, MIBS-4, MSQ, and WPAI-M, revealed that MM imposes a substantial burden on daily functioning and is associated with reduced work productivity. A case-control study from a population-based group of Danish individuals with migraine showed that patients with MM had higher pain intensity, measured using the VAS, during migraine attacks compared with those in the non-MM group [ 19 ]. Our findings also demonstrated higher pain intensity in the MM group compared with the non-MM group, which is consistent with the previous report [ 19 ]. Additionally, our findings showed that migraine pain intensity peaked around the onset of menstruation. A higher proportion of patients with MM experienced “extremely frequent” or “frequent” headaches, which was associated with increased usage and diversity of medication use. These findings align with a previous case-control study conducted among Danish individuals with MM ( n  = 1,532), which reported that MM was characterized by “more frequent” and “more severe” migraine attacks [ 19 ]. Similarly, another study reported that acute medications are often taken daily as a form of preventive treatment to manage headaches during the perimenstrual period [ 20 ]. Our study further supports these findings by demonstrating that patients in the MM group experienced a higher frequency of migraine headaches during the perimenstrual period. The pathophysiology of MM is complex and is thought to involve estrogen decline, which is associated with a threefold increase in prostaglandin, contributing to dysmenorrhea and MM [ 5 , 21 – 23 ]. Additionally, they can enhance calcitonin gene-related peptide release via transient receptor potential A1 (TRPA1) receptor activation [ 24 ]. In our study, we evaluated pain intensity during the perimenstrual and non-perimenstrual periods between the 2 groups: those who experienced headaches in 2 or more of 3 menstrual cycles (MRM subgroup) and those who experienced headaches in only 1 cycle (insufficient-MRM subgroup). The MRM subgroup exhibited higher pain intensity during the perimenstrual period, whereas no such association between menstrual timing and pain intensity was observed in the insufficient-MRM subgroup. These findings support the robustness of the ICHD-3 diagnostic criteria, which require at least 2 attacks within 3 menstrual cycles for the diagnosis of MM. As for the pain duration, the MM group exhibited longer durations compared to the non-MM group. In addition, there was no difference in pain duration during perimenstrual period compared to non-perimenstrual period in MRM group unlike the previous study [ 19 ]. One possible explanation for this discrepancy is that the use of acute medications during headache episodes may have masked differences in pain duration, as this study was not a prospective investigation, and participants were allowed to take medications at their own discretion. Regarding prescription patterns between the MM and non-MM groups, the MM group exhibited a higher frequency of lasmiditan use. In Japan, lasmiditan is considered a second-line agent, typically reserved for patients who show an inadequate response to triptans or intolerance such as CV diseases, to first-line agents [ 25 , 26 ]. The increased use or combination of acute therapies in the MM group may reflect the greater clinical challenge of achieving sufficient pain control in this population. Importantly, under the current Japanese national health insurance system, gepants are not reimbursed for acute migraine treatment. In addition, gepants, topiramate, and onabotulinumtoxin-A are not covered for preventive therapy, potentially limiting access to a broader range of evidence-based treatment options. In our study, PMM accounted for 54.0% of the MM group. Previous studies have reported that the proportion of PMM among individuals with MM varies widely, ranging from 3% to 76% [ 27 , 28 ]. It is plausible that the more intense pain experienced during the perimenstrual period may obscure or attenuate the perception of pain during non-perimenstrual migraine attacks. Moreover, the absence of electronic diary use in our study may have introduced recall bias, particularly for mild migraine episodes. In our study, the proportion of migraine with aura was higher in the MM group compared to non-MM group (26.6% vs 22.0%). Previous reports on MM attacks have indicated that they frequently occur without aura. In contrast, Verhagen et al. reported no association between the presence of aura and MM [ 29 ]. Furthermore, it is well recognized that patients with migraine with aura do not necessarily experience aura with every attack, and that MM attacks often occur without aura. In our study, some individuals with MM were observed to experience visual aura; however, we did not assess the presence or absence of aura separately for menstrual and nonmenstrual periods. Therefore, prospective studies utilizing detailed migraine diaries are warranted to clarify the relationship between MM and aura. In our study, the MM group exhibited a higher prevalence of photophobia and phonophobia compared with the non-MM group, consistent with previous reports indicating stronger sensory hypersensitivity to light and sound during MRM attacks [ 30 ]. Furthermore, although a high prevalence of osmophobia has been reported among Japanese patients with migraine, our findings revealed that this tendency was even more pronounced in the MM group [ 31 ]. These most bothersome symptoms are thought to be associated with dysfunction in thalamic sensory filtering mechanisms implicated in migraine pathophysiology. Our results strongly suggest that in MM, not only pain but also accompanying symptoms contribute significantly to the overall disease burden. These findings underscore the need for a more intensive therapeutic approach in MM and support the development of treatment strategies specifically tailored to this subtype. This study was a self-reported, cross-sectional online survey. Given the nature of self-reported surveys, it was difficult to validate participants’ reports concerning their symptoms, treatment patterns, actual adherence, and factors associated with symptoms and treatment. In addition, MM was estimated by patients rather than objectively assessed using a validated headache diary. Although participants were randomly selected within the Macromill Carenet panel, this randomization does not completely represent the general population. Since this survey was not conducted prospectively using an electronic diary, and patients were recruited through the registry of Macromill Carenet Inc., there may be selection bias. However, these practices are consistent with other web-based patient surveys. This study excluded females with irregular menstrual cycles to clarify the clinical characteristics and social burden of MM. However, this exclusion may limit the generalizability of the findings to real-world clinical practice, where MM may also be observed in females with irregular menstruation. Although the study included planned statistical analyses to control for confounding factors, unmeasured confounders may still have introduced bias in the comparisons between groups.

The MM group demonstrated greater disability across multiple dimensions—including pain intensity, pain duration and frequency during the perimenstrual period, and social burden—compared with the non-MM group. Furthermore, within-individual analyses in the MM group, pain intensity was higher during the perimenstrual period compared to non-perimenstrual period, and more patients increased usage and type of medication during the menstrual period. Given the substantial clinical and social burden associated with MM, and the increased use of NSAIDs—which, with frequent use, may lead to medication overuse, and with long-term use, may cause cardiovascular adverse effects and renal impairment—these findings underscore the need for the development of safer and more effective treatment strategies specifically tailored to MM.

Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1