Sex-related differences in migraine clinical features by frequency of occurrence: a cross-sectional study

Participants and setting

A cross-sectional design was conducted. The recruitment method used was stratified sampling since migraines are two or three times more prevalent in women than in men [22]. Participants were recruited from primary care centres in Valencia (Spain) for one year. Individuals diagnosed according the ICHD-3 criteria [23], with four or more episodes per month, and at least one-year migraine history were included. Patients with another type of primary or secondary headache, psychiatric disorders, or other chronic pain diagnosis different than migraines were excluded.

The assessments were carried out face-to-face by a researcher blinded to the objectives of the study. The study was approved by the Ethics Committee of the University of Valencia (H1509655117217).

Measurements

Migraine diagnosis (episodic or chronic) was provided by a neurologist. Migraine characteristics about family history, migraine duration, pain intensity, pain location, associated symptoms, triggers and relieving factors were collected.

Pain, frequency, and disability relating to migraines were assessed through the Migraine Disability Assessment (MIDAS). It includes questions about days lost or less productive days in terms of paid work, household chores and social activities, pain intensity and number of days with pain in the last three months [24]. The MIDAS has good internal consistency (Cronbach’s α=0.73–0.76) and high reliability and validity (Pearson’s correlation coefficient (CCI) test-retest r=0.80) [25].

Quality of life was evaluated with the Short Form 36 Health Survey (SF-36) [26]. It consists of 36 items assessing physical function, physical role, bodily pain, and general health (for the physical health subscale), and vitality, social function, emotional role, and mental health (for the mental health subscale). This questionnaire has a good internal consistency (Cronbach’s α=0.70), and a CCI of 0.99 on the physical subscale and 0.58 on the mental subscale [27].

Depression level through the Beck Depression Inventory II (BDI-II) which collects psychological and somatic manifestations of depressive episodes in the last two weeks [28]. Internal consistency is rated with a Cronbach’s α of 0.90, and test-retest reliability from 0.73 to 0.96 [29].

Anxiety level with the State-Trait Anxiety Inventory (STAI) [30]. It consists of state-anxiety subscale (i.e. emotional state that can vary in intensity over time), and trait-anxiety subscale (i.e. tendency to be anxious and to perceive situations as threatening). It has good internal consistency (Cronbach’s α=0.93) [30], and reliability from 0.87 to 0.93 [31].

Data related to Health resource use. Medication intake in the last 4 weeks (i.e., acute or preventive medications), and medical assistance (i.e. frequency of visits to a headache specialist, general practitioner over the last year, and other health care professionals) were collected.

Statistical analysis

Statistical analysis was performed using SPSS software v.24 (SPSS Inc., Chicago, IL, USA). The Kolmogorov-Smirnov test was carried out to assess normal distribution. A two-way analysis of variance (ANOVA) was performed with the fixed factors of sex (i.e., men and women) and frequency of occurrence (i.e., EM and CM), and quantitative dependent variables regarding migraine characteristics, psychophysical health, and health care resources. Post hoc analyses were performed using the Bonferroni tests on each set of groups regarding sex and frequency of occurrence. The assumption of homoscedasticity using Levene’s test and the sphericity using Mauchly’s test were analysed. Chi-square tests were performed for ordinal variables. Size effect was determined using Cohen’s d and contingency coefficient (CC) only when statistically significant differences were found. The level of the significance was set at p<0.05.

Sex-related differences in sociodemographic characteristics by migraine diagnosis

From the 218 participants initially recruited, 34 participants were excluded because they did not meet the inclusion criteria (i.e. 20 had less than four per month, and 14 suffered from another type of headache). Finally, a total of 184 participants were assessed, of whom 113 participants (61.41 %) were diagnosed with EM and 71 (38.57 %) with CM. Regarding sex, 133 participants were women (72.28 %) and 51 were men (27.72 %). Sociodemographic characteristics by diagnosis and sex are presented in Table 1. The mean age (SD) of the sample was 38.70 (10.39) years, and the average BMI was 23.84 (3.87). Most of the participants were employed (69.57 %). More than half of the participants were physically active (55.43 %). There were no significant differences when stratifying by diagnosis in age, occupation, and level of physical activity (p>0.05) between men and women, except for mean BMI in the EM group (p=0.003), thus most of the men included were in the overweight range (42.86 %), while the women belonged to the healthy weight range (70.51 %).

Sex-related differences in migraine clinical characteristics by diagnosis

As shown in Table 2, participants reported being 18.96 (9.39) years old on average when they experienced their first migraine attack, and 19.79 (11.34) years of migraine evolution. Furthermore, 89.67 % of the participants reported having a family history of migraines.

Regarding associated symptoms, 77.13 % of the sample reported suffering from two or more per episodes, being women with EM those with higher number of symptoms (p=0.03, d=0.45) than men. In addition, they reported a higher prevalence of phonophobia than men (60.26 vs. 25.71 %, p=0.001, CC=0.30), and nausea and vomiting (62.82 % vs. 22.86 %, p<0.001, CC=0.35). However, there were no significant differences by diagnosis nor by sex either in the prevalence of photophobia (71.20 % of the sample), pericranial sensitivity (42.39 %), aura (33.70 %), or others (35.87 %, i.e., tears or dizziness) (p>0.05).

Regarding trigger factors, women with CM presented a greater number of triggers than men with CM (p<0.001, d=0.45). Stress was the most reported (74.46 % of the sample) and as expected, menstruation was significantly more reported in women compared to men in both groups (p<0.001, CC=0.44). Sleep deprivation was reported more frequently in women with CM than in men (58.18 % vs. 12.50 %, p=0.001, CC=0.36), yet also differed by diagnosis in men (42.86 % in the EM group vs. 12.50 % in the CM group, p=0.03, CC=0.29). No significant differences by diagnosis nor by sex either were found in other trigger factors (p>0.05).

When analysing sex-related differences by migraine diagnosis, women with EM showed a higher pain intensity (p<0.001, d=0.70) and pain duration (p=0.03, d=0.56) than men with EM. However, there were no significant differences in frequency of onset by sex (p>0.05). Furthermore, no significant differences by sex were found in people with CM in any of the variables (p>0.05).

In addition, regarding diagnosis-related differences by sex, women with CM showed higher frequency of onset (p<0.001, d=1.44), and duration of episodes (p=0.03, d=0.35) than women with EM. Similarly, men with CM had a higher pain intensity (p=0.02, d=0.59), frequency of onset (p<0.001, d=2.01), and duration of episodes (p=0.002, d=1.11) than men with EM.

There were no significant differences by diagnosis nor by sex, either in relieving factors or in pain location (p>0.05) (data not shown). The most common relieving factor was medication intake (82.61 %), rest or sleep (44.02 %) and staying in a dark and quiet place (60.87 %), whereas the most frequent pain locations were temporal (54.89 %), orbicular (44.02 %), frontal (38.58 %) and occipital regions (33.70 %).

Sex-related differences in disability, impact, quality of life and emotional status by migraine diagnosis

As shown in Table 3, when analysing sex-related differences by migraine diagnosis, women showed a higher disability (p=0.01, d=0.61) than men with EM. Furthermore, no significant differences by sex were found in people with CM in associated disability (p>0.05). In addition, regarding diagnosis-related differences by sex, women with CM showed a higher disability (p<0.001, d=1.14) than women with EM. Similarly, men with CM had a higher associated disability (p<0.001, d=1.18) than men with EM.

Regarding quality of life, no significant differences were found between males and females for any of the variables either in EM or CM, except for in regard to a physical role in people with EM, where males showed a higher score (p=0.004, d=0.58) than women (Table 3). When comparing by diagnosis, the quality of life was significantly reduced in women with CM when compared to women with EM in the following instances: physically functioning (p<0.001, d=0.71), bodily pain (p=0.004, d=0.54), general health (p=0.04, d=0.34), social functioning (p=0.01, d=0.47), and the physical health subscale (p=0.01, d=0.49). For men, there were significant differences between EM and CM in physical role (p=0.01, d=0.95), general health (p=0.02, d=0.76), and physical health subscale (p=0.002, d=1.02).

No significant differences by diagnosis and sex were found in emotional status, such as levels of depression and anxiety (p>0.05) (Table 3).

Sex-related differences in health care resource use by migraine diagnosis

Regarding medication intake, most of the migraineurs took medication (97.28 %), of whom 46.20 % took preventive medication and 89.10 % took symptomatic medication. Migraine-specific medicines divided by diagnosis and sex are specified in Table 4. Women with EM took significantly more symptomatic medication than men (p=0.04, d=0.45). Moreover, significant differences in triptans intake were identified (29.49 % among women and 17.14 % among men, p=0.003, CC=0.31). However, there were no significant differences by sex in other specific-migraine drugs (p>0.05), nor between people with CM either (p>0.05). When EM and CM groups were compared, significant differences in medication type were found. Thus, there was an increase in the use of preventive medication in both women (26.92 % with EM vs. 74.55 % with CM, p<0.001, CC=0.38), and men (34.29 % with EM vs. 68.75 % with CM, p=0.002, CC=0.44).

When analysing medical assistance, 51.09 % of the participants visited a headache specialist in the last year, with an average number of visits of 2.01 (3.31) times (Table 4). 36.96 % of participants also visited other health professionals, of whom physiotherapists were the more frequently visited (17.93 %). A higher percentage of women visited a headache specialist in the last year due to migraine than men with EM (55.13 % vs. 22.86 %, p=0.001, CC=0.37), as well as other health professionals (44.87 % vs. 20.00 %, p<0.001, CC=0.37), especially physical therapists (20.51 % vs. 8.57 %, p=0.003, CC=0.34). When comparing by migraine diagnosis, a significantly higher percentage of men with CM visited a headache specialist in the last year than those with EM (22.86 % vs. 68.75%, p=0.002, CC=0.43).

Main findings

The main finding of this study was that women with EM presented a greater number of associated symptoms, pain intensity, disability, and duration of episodes, as well as a worse physical role than men with EM, yet those differences are lost in people with CM. Additionally, the use of preventive and symptomatic medication was similar in both men and women. However, women with EM visited migraine-related health professionals more frequently than men.

Interpretation of results in relation to current knowledge/literature

Changes in brain structure are described in migraine patients, such as reduced microstructural integrity in the thalamus, corpus callosum, longitudinal fasciculus, and cingulum [32]. Recent studies suggest that there are structural brain differences between EM and CM according to cortical surface area, cortical thickness, and regional volumes [33]. For example, more streamlines have been found between the left caudal anterior cingulate and right superior frontal gyri in CM compared to EM [34], as well as decreased axial diffusivity in white matter regions, that might reflect a set of maladaptive plastic changes after the onset of CM [35].

Previously it has been highlighted that females display greater associated disability [7], [8], [9, 36], more frequency (i.e., days per month) [9], longer attack duration and more associated symptoms [36] than men, which is consistent with our results among people with EM. In contrast, others did not find significant differences by sex [37]. In addition, it is interesting stratifying by frequency of occurrence as CM is much more disabling [6, 25, 38], [39], [40], [41], [42], [43], [44]. Surprisingly, we found that the differences found in our study between men and women with EM were lost in the CM group, since the impact of migraine chronicity affects both men and women equally. This may be explained because the aforementioned changes in CM in brain structure and function may occur in men and women indistinctly. In contrast to our results, previous studies reported that female migraineurs with CM experience higher levels of headache-related disability [45]. However, further analysis in men are warranted to elucidate disability levels since men are less likely to receive a diagnosis than women with migraine [46]. In addition, data obtained in previous studies are self-reported and not verified physician diagnoses or medical records, which could be misleading. For example, in the study by Buse et al. [47], disability values were lower than usually experienced in people with CM (i.e., 41.30 % of women with CM reported having severe headache-related disability vs. 28.10 % of men). In this regard, in the International Burden of Migraine Study [40], 78 % of the people with CM were classified as MIDAS grade IV (i.e., severe disability), even with an adjustment for sex, which is more similar to the values reported in the present study.

The increase in disability may influence quality of life [44]. In the present study, poorer physical functioning, physical role, bodily pain, general health, and social functioning was observed in people with CM compared with people with EM. The worse physical condition may be due to the fact that migraine chronification leads to difficulties in physical activity engagement, mainly because physical exertion is a trigger factor for migraines [23], and some people avoid physical activity as a prevention strategy [48]. It is worth noting that women with EM reported lower scores in the physical role item than men with EM. This could be explained precisely by the fact that women have traditionally exercised less, and consequently are in worse physical condition. Despite all this, the increase of beta-endorphin, endocannabinoid and brain-derived neurotrophic factor levers released in plasma after exercise may have prophylactic effect on migraine frequency [49]. It is worth noting that previous studies have not only highlighted the impairment of quality of life at the physical level, but also have found significant differences between diagnosis at the emotional level [44], specifically in the vitality, role emotional and mental health items [43, 50].

Regarding emotional status, the participants included in our study reported low levels of depression and anxiety, with no significant differences between men and women either in EM or in CM. This is consistent with the results obtained by Song et al. [37], since almost 30 % of the sample presented anxiety and 16 % presented symptoms of depression, without significant differences by sex. However, regarding potential differences by migraine subtypes, we did not find any differences between EM and CM either in men or in women. The disagreement with previous studies showing a higher prevalence of mental disorders in people with CM than people with EM [43] may be due to the high heterogeneity of the data found in our study. Thus, these results must be taken with caution. Overall, emotional status-related conditions should be considered in the clinical practice as they may increase the burden in patients with episodic and chronic headaches [51].

Acute medication was prescribed in 89.13 % of the patients, and preventive medication in 46.20 % of them. This is consistent with the results reported by Ford et al. [52], who described that >90 % of migraine patients were prescribed an acute treatment and >50 % a preventive one. In addition, triptans were the most commonly used drugs, as they were in our study after NSAIDs. However, they reported that opioids were significantly more frequently prescribed in patients with CM than in EM, whereas we found only significant differences between diagnosis subgroups in Botox injections prescription. It should be considered that the utilisation of migraine-specific medications varies between countries [53]. Nevertheless, since this pathology continues to affect people’s daily lives, it is claimed that migraine drugs are used inadequately [52], [53], [54]. Difficulty in finding an appropriate migraine treatment may be justified because the drug response differs among individuals. In this regard, we observed that the use of triptans was significantly higher in women than in men with EM. According to Van Casteren et al. [55], even though response rates of triptans are similar in males and females regarding headache and pain-free responses, men had a lower risk for headache recurrence, thus the need for triptan intake may be reduced. Furthermore, we observed that the amount of medication used per month was 28.30 % higher in women than in men. This may be due to the higher frequency of episodes in women, and in some cases it may lead to the phenomenon of medication overuse [56]. Therefore, more tailored and specific non-pharmacologic treatments such as those proposed by our group [57], [58], [59], are needed to lower polypharmacy among patients with migraine.

Regarding health care resource use, one half of the participants reported having visited a headache specialist in the previous year. Interestingly, we found that men with EM visited the doctor significantly less often than women with EM, while they experienced an increase in medical visits up to three times when they get to CM. This is consistent with the well-known higher health care resource use and direct and indirect costs associated to CM [60, 61]. Despite this, many people with migraine have unmet needs related to consulting for migraine, migraine diagnosis, and getting potentially beneficial migraine treatment [16], which may be due to the lack of accuracy in migraine diagnosis, probably due to sex-related underdiagnosed differences.

The novelty of our study is that multiple areas of clinical interest (i.e. migraine characteristics, psychophysical health and health care resource) are analysed by sex and frequency of occurrence in individuals that suffer from four or more migraine episodes per month. In this regard, most of the published articles that compare EM vs. CM included people with at least one migraine day per month [8, 20, 40], [41], [42], [43, 51, 61]. It is of interest to study this population since associated disability is higher than in people with less frequency of occurrence [52], and there is still a need to find effective prophylactic treatments for them to lessen the migraine burden [54]. In addition, in contrast to previous cross-sectional studies [7, 9, 40, 41, 43, 52, 54], individuals with other comorbidities (i.e., another type of primary or secondary headache, psychiatric disorders, or other chronic pain diagnosis) were excluded in our study. Thus, a more specific analysis of the impact of migraine on patients’ daily lives is performed. Notably, most studies assessing gender differences assess migraine patients without distinguishing between episodic and chronic migraines [7, 36, 37]. Furthermore, in some cases respondents did not have a medical diagnosis [8, 9, 20, 40, 41, 54, 60]. In this regard, correct diagnosis of migraines according to the ICHD-3 criteria [23], as well as understanding the patient’s needs, is important to design individualised treatment protocols. In addition, an early therapeutic approach, as well as control of risk factors for migraine progression (i.e., increased daily headache frequency, acute medication overuse/high frequency use and depression) will prevent the transition from EM to CM [21]. This, in turn, will improve the patient’s quality of life, and will result in reduced costs to the health care system. Finally, the use of reliable, validated, easy-to-use and low-cost assessment tools, such as those presented in this study, are easy to implement in clinical practice, thus allowing health professionals to improve both diagnostic accuracy and response to treatment monitoring.

Limitations

The main limitation of the study was that the sample included much more women than men, yet migraines occur much more often in women than in men. Another limitation was that our sample comprised of mostly people between 18 and 50 years old, thus these results may not be generalisable to other populations. However, it is important to study the young population since migraines are the leading cause of disability in this age group [2, 3]. In addition, the fact that the participants were young and relatively healthy decreased the possibility of comorbidities that could influence the analysis. In this regard, although the groups were comparable with respect to age, sex, and race, significant differences in BMI between men and women with EM were considered as a potential confounder. However, when this variable was controlled for in the analyses, there was no significant changes in the results for any of the main outcomes. Finally, the evaluations consisted in questionnaires, so more objective measures would be desirable.