Clinical manifestations

Presentation and course

The psychiatric conditions consistently demonstrated to coexist with migraine can be classified into three main categories, namely mood disorders, anxiety disorders, and personality disorders.

Mood disorders. Both unipolar and bipolar mood disorders have been reported to be comorbid with migraine. Regarding depression, symptoms such as lack of energy, poor concentration, sleep disturbance, loss of appetite, psychomotor agitation or retardation, weight changes, feelings of worthlessness, etc. are commonly observed in persons with migraine. The syndrome of depression can range from mild dysthymia to life-threatening major depression. The Zurich Cohort Study showed that 1-year prevalence of dysthymia and major depression in persons with migraine was 6.6% and 14.7% (70). Comparatively, findings from the National Comorbidity Survey-Replication in the U.S. showed 1-year prevalence of dysthymia and major depression was 8.8% and 18.8%, respectively (99). Lifetime prevalence of major depression is approximately three to four times higher in persons with migraine compared with controls (07; 13). A large-sample, case-control study in patients with depression in the UK also showed that the strength of association between migraine with aura and depression was high (OR=5.6, 95% CI=3.5 to 9.0) (98). A historical cohort study using the National insurance database in Taiwan showed that the risk of depression in chronic migraine was almost two times (RR = 1.88) as high as that of other migraines (18). The increased risk of depression was also evident in adolescents with migraines (02). Comorbid depression was more frequently observed in patients in the tertiary care or special headache clinics. Stress related to health issues significantly increased risk of depression (OR = 4.6, 95% CI = 1.6 to 13.5) (81). The findings support the idea that the risk of having depression is greater in persons with migraine with aura and also with chronic migraine. Other factors shown to be associated with mood disorders include frequency and duration of migraine attack (93). However, this study did not find statistically significant correlation between mood disorders and the total duration of migraine or the presence of aura, vomiting, and phono and photophobia. The presence of moderate to severe allodynia was found to be an independent risk factor for clinically significant depression (46). Another study also showed that allodynia, together with high migraine attack frequency, was associated with depression (59). A prospective diary study demonstrated that patients with migraine scored higher on acute depressive symptoms during the migraine headache phase, but not in the prodromal phase (p < 0.001) (24).

One study exploring the association between subtypes of depression (melancholic, atypical, combined, and unspecified) and subtypes of migraine (with or without aura) found that all but atypical depression were associated with an increased frequency of migraine with aura; however, only melancholic depression was associated with an increased frequency of migraine without aura (90). These findings supported the hypothesis that different mechanisms underlie each subtype of migraine and depression.

The onset of depression can either precede or follow that of migraine. Using survival analysis of the lifetime data, which covered the respondents’ history up to the time of the baseline interview, Breslau and colleagues demonstrated that the association between migraine and major depression is bidirectional (10; 13). The sex-adjusted hazard ratio of the first onset of major depression in persons with migraine was 2.35 [95% confidence interval (CI) 1.84 to 3.01], whereas the hazard ratio for the first occurrence of migraine in persons with prior major depression was 2.75 (95% CI 2.17 to 3.48). This bidirectional association was later confirmed by the same investigators using a different approach. In 2003, Breslau and colleagues compared the incidence of first-onset major depression in the 2-year follow-up period across persons with migraine, persons with severe headache not fulfilled migraine diagnosis, and nonheadache controls (11). The results revealed the incidence of major depression in the migraine group to be 10.5%, in the severe headache group 5.1%, and in the controls 2.0%. Based on these rates, the odds ratio for major depression in persons with migraine was 5.8 (95% CI 2.7 to 12.3) and 2.7 (95% CI 0.9 to 8.1) in those with nonmigraine severe headache. Preexisting major depression also increased the risk of having new-onset migraine. The incidence of new-onset migraine in the 2-year follow-up period in persons with history of major depression was 9.3%, compared to 2.9% in those without. The odds ratio of migraine associated with prior major depression was 3.4 (95% CI 1.4 to 7.6). A similar degree of strength of association between depression and the risk of developing migraine was also demonstrated in a study, which suggests the role of depression assessment in preventing and treating migraine and migraine-related burdens (25). It is interesting that, in contrast to migraine, prior depression did not increase risk of having severe headaches that did not fulfill the criteria for migraine diagnosis.

Concerning the impact of comorbid depression on the course of migraine, Breslau and colleagues showed that persons with this comorbidity experienced headache of greater severity than those without (11). They were not, however, at greater risk for headache persistence. A large population-based study revealed that the odd ratios for depression, as well as anxiety disorders, among both migraine and non-migraine patients increased with increasing headache frequency (123). A longitudinal study showed that although co-occurrence of migraine, personality changes, and depression in women does not appear to influence the results of treatment at short-term, it seems to be influential on headache history in the long term (78). Comorbid depression may have impact on brain structures. A study showed that total brain volume (white matter and gray matter volumes) was smaller in those with both migraine and depression, but not for those with migraine alone or depression alone (34).

Comorbid major depression has strong impact on patients’ quality of life. Based on a population-based, case-control study, Lipton and colleagues showed that patients with migraine and depression had lower scores on health-related quality of life as measured by the Short Form (SF)-12. Lower quality-of-life scores were observed in both mental health component scores (MCS-12) and physical health component scores (PCS-12) (58). Greater severity of depressive symptoms in patients with migraine were found to be associated with greater functional impairment, including migraine (related disability, work interference, pain interference, reduced career success) (84), as well as sexual dissatisfaction (23). Persons suffering from migraine with aura and coexisting major depression had higher rates of suicidal attempts and suicidal ideation compared with patients with neither migraine nor major depression (09). Compared with controls without history of severe headache, migraineurs and severe non-migraineurs had an increased risk of suicide attempt (OR= 4.43, 95% CI 1.93, 10.20; OR=6.20, 95% CI 2.40, 16.00, respectively) adjusted for sex, psychiatric disorder, and previous history of suicide attempts (12). The differences of OR between migraineurs and non-migraineurs were, however, not statistically significant. The result supported the theory that risk of suicide attempt was more likely be accounted for by pain severity than by migraine headache itself. Another cross-sectional study suggested migraine frequency as an independent risk factor for suicidal ideations and attempts in migraine patients with aura, but not those without aura (56).

A study in Taiwan showed that migraine with aura itself increases suicide risk score in adolescents with chronic daily headache, even after being adjusted for depression and anxiety disorders (117). The same investigators also showed that comorbidity with migraine was associated with more somatic symptoms in patients with major depression, and migraine was a strong and independent predictor for the somatic symptoms of this condition (41).

Apart from major depression, association between migraine and bipolar disorder has also been well documented. Persons with bipolar disorder had a relatively higher prevalence of migraine versus the general population (24.8% vs. 10.3%) (67). This association was found to be even more robust in adolescents (38.2% in bipolar disorder group vs. 3.4% in control group) (69). Moreover, patients with bipolar disorder and comorbid migraine were shown to have rapid cycling course (33). Another longitudinal study conducted by Breslau and colleagues showed sex-adjusted odds ratio was up to 4.7 (95% CI 1.4, 15.4) (09). An association between migraine and bipolar disorders has been evident consistently, especially for bipolar type II. A cross-sectional hospital-based study of 102 patients with major depression or mania found that, as compared to the patients without migraine (n = 49), the patients with comorbid migraine (n = 53) had a higher frequency of bipolar II disorder (43% vs. 10%), a lower frequency of bipolar I disorder (11% vs. 33%), and an approximately equal frequency of unipolar depressive disorder (45% vs. 57%) (26). Within the migraine group, those with chronic migraine had an increased risk of having bipolar disorder (RR = 1.81) compared to those with other migraines (18). Among the patients with bipolar disorders, prevalence of migraine was higher in persons with bipolar type II (77%) than those with bipolar type I (14%) (26). The comorbidity of migraine was correlated with poorer treatment outcome in bipolar disorder, and lithium was suggested to be associated with more severe manic symptoms in bipolar patients with comorbid migraine (102). Patients with bipolar disorder with migraine were younger, higher educated, more likely to be in work or study, and had fewer hospitalizations; the initial presentation for psychiatric treatment was more often depression, and these patients more likely to have a family history of migraine or psychiatric disorders (60). A cross-sectional study pointed out that migraineurs with bipolar disorder had a significantly higher percentage of family history of bipolar disorder (39.1% vs. 6.2%, P< 0.001), suicide attempt (30.4% vs. 5.2%, P< 0.001), and physical abuse (52.2% vs. 26.8%, P=0.019) than those without (50). Ortiz and colleagues also found that bipolar patients with migraine had more psychiatric morbidity such as social phobia, obsessive-compulsive disorder, panic disorder, and generalized anxiety disorder than those without migraine (82). Apart from psychiatric comorbidity, a comorbidity of migraine in patients with bipolar disorder was also shown to be correlated with an increased prevalence of pain and inflammatory conditions including fibromyalgia (OR=3.17, p< 0.01), psoriasis (OR=2.65, p=0.03), and asthma (OR=2.0, p< 0.01) (96). Despite consistent findings of bipolar disorder and migraine, comorbid association education and health care service use were inconclusive (60; 67). A retrospective study analyzing differences in medical and psychiatric comorbidities in bipolar patients by sex found that females with bipolar disorder were three times more likely to have migraines compared to males (83).

Anxiety disorders. Anxiety disorder is more prevalent in patients with migraine than in the general population (09), including children (39). Among those with migraine, anxiety is considerably more common in chronic migraineurs than in patients with other migraines (RR = 1.48, p = < 0.0001) (18). Saunders and colleagues reported that 1-year prevalence of any anxiety disorder in migraineurs was 44.5%, which was higher than that of any mood disorder (24.7%) in the same household survey (99). Generalized anxiety disorder was found to significantly increase the severity of pain in patients with comorbid chronic pain (including chronic migraine) compared to those with chronic pain alone (20). A systematic review revealed that despite a higher prevalence of migraine in females, males with migraine are likely to report greater anxiety symptoms than females (47). A 14-month follow-up study revealed that patients with migraine at baseline had a significantly increased rate of panic disorder (odds ratio = 12.8, 95% CI 4.1 to 39.8) (08). Data drawn from a nationally representative sample in the United States revealed that persons with migraine, adjusted for demographic variables, had greater risk of having panic (odds ratio 2.37; 95% CI 1.42 to 3.99%) or generalized anxiety disorders (3.13; 95% CI 1.56 to 6.30) (68).

It should be noted that association between panic disorder and headache syndromes is not specific to migraine. In 2001, Breslau and colleagues showed that lifetime prevalence of panic disorder was significantly higher in persons with migraine as well as persons with other severe headaches. Both migraine and other severe headaches were associated with an increased risk of first onset of panic disorder [hazard ratios = 3.55 (95% CI 2.18 to 5.76) and 5.75 (95% CI 2.70 to 12.27), respectively]. Preexisting panic disorder was also associated with an increased risk of first onset of migraine and for first onset of other severe headaches, although the influence on this direction was lower [hazard ratios = 2.10 (95% CI 1.44 to 3.08) and 1.85, (95% CI 0.71 to 4.48) respectively] (14). In patients with coexisting migraine, treatment of panic disorder has shown to improve migraine in most cases (58%) (119). Association between panic disorder and migraine may vary depending on age group. For example, a population-based study in Sweden showed no significant association between migraine and panic disorder in women aged 40 to 74 years (66).

The association between frequency of migraine attacks and higher anxiety scores (using SAFA-anxiety subscales) was also observed in children. The same study demonstrated that being overweight was associated with higher frequency of migraine attacks as well as separation anxiety symptoms (111).

Coexistence of anxiety and mood disorders are frequently observed in persons with migraine. In a clinical sample of patients with migraine, 24% and 42%, respectively, had current and lifetime diagnoses of both mood and anxiety disorders (82). In the Detroit study, 88% of persons with a history of migraine and major depression also reported at least one anxiety disorder (09). Data from a prospective longitudinal study in Zurich indicated that age of onset of anxiety disorders generally preceded that of migraine and that the onset of affective disorders in the majority of comorbid subjects followed that of the onset of migraine (70; 72). The results implied a strong relationship between migraine and anxiety or depression, rather than their representing discrete manifestations.

Personality disorders. Several personality traits have been reported to be more prevalent in persons with migraine. Among those, neuroticism measured by Eysenck’s personality questionnaire showed a considerable degree of association with migraine, which suggested that migraine sufferers might be more vulnerable to psychopathology (05; 06; 106). No significant difference among subtypes of migraine has been documented. Other personality traits associated with migraine include high harm avoidance, high persistence, low self-directedness, and low extraversion (32).

Personality disorders most commonly associated with migraine are cluster B and C (53). A vast majority of studies available today normally concern borderline personality disorder. However, two studies reported that the personality disorder most frequently coexisting in migraine patients were obsessive-compulsive personality disorder (cluster C) (48; 120). Although the results from both studies reiterated Wolff’s concept of the “migraine personality,” which included perfectionism, orderliness, moralistic preoccupation, and rigidity (118), the authors observed that their findings might be accounted for by the subjects’ sociocultural backgrounds, which was more suitable for the development of cluster C, rather than cluster B: personality disorders. The association between migraine and neuroticism remained significant when sex, history of major depression, and history of any anxiety disorder were controlled. Elevated rates of neuroticism and somatization, as measured by the Freiburg Personality Inventory and the Symptom Checklist 90, have been observed in persons with migraine (75). A metaanalysis even found the evidence that neuroticism influenced the severity of depression in patients with migraine (32). Abnormal personality profiles are more prevalent in persons with either chronic headache, chronic migraine, or migraine with analgesic rebound headache. A case-control study showed that persons with chronic migraine as well as analgesic rebound headache had higher score in multiple subscales of MMPI compared to persons with episodic migraine. These subscales included hypochondriasis, depression, schizophrenia, and social introversion (04). Interestingly, cluster A personality traits (eg, paranoid, schizoid traits) are more common in patients with cluster headache than in migraineurs (80).

Comorbidity with a personality disorder was found to be significantly correlated with a higher frequency of chronic migraine, more severe migraine symptoms, as well as more disability (120).

Significant headache is a common complaint in patients with borderline personality disorder. Based on information gathered from 112 patients visiting a psychiatric clinic, Hegarty reported overall prevalence of severe headache was 60.4% in those with borderline personality disorder. Fifty percent of females and 24% of males were diagnosed as having migraine (37).

Other psychiatric comorbidities. Posttraumatic stress disorder (PTSD), history of childhood trauma, and physical abuse in adulthood are commonly associated with migraine. The prevalence ranges from 9% to 43% for PTSD (89) and 33% to 58% for physical abuse (113). Moreover, migraineurs with PTSD were found to have greater disability (8 work-off days for migraineurs with PTSD compared with 2.6 days for migraineurs without PTSD, adjusted rate ratio 2.77 [95% CI 2.45 to 3.14]). These patients experience more difficulties maintaining a social life than migraineurs without PTSD (adjusted OR 2.32 [95% CI 1.15 to 4.09]) (89; 94). Migraine frequency was also found to be associated with poor sleep quality (107). A higher prevalence of migraine was reported in children and adolescents with attention deficit hyperactivity disorder (ADHD), although such association was not demonstrated in young adult populations (38). A Canadian population-based prospective cohort study suggested migraine as a significant risk factor for Alzheimer disease (OR 4.22, 95% CI=1.59 to10.42) and all-cause dementia (OR 2.97, 95% CI 1.25 to 6.61), but not vascular dementia (79).

Prognosis and complications

Major complications of migraine with psychiatric comorbidities are substance abuse and suicide. The rate of attempted suicide has been reported to be highest in persons with migraine with aura with comorbid major depression. The risk of suicidal attempt in this group is 38.5 per 100 cases, which is more than double that of the risk of depression alone (16.5 per 100 cases). Female migraineurs of any subtypes have higher rates of attempted suicide than their male counterparts (09). A physician-based study in 48 countries showed that primary headaches cause one suicide per 1,000,000 population each year. Cluster headache and migraine account for 70% to 80%. Cluster headache has greater risk of suicide than migraine (115).

Sleep disturbances are common in depression and anxiety. Therefore, persons with these conditions are at risk of sedative or hypnotic abuse.

Clinical vignette

A 52-year-old married female was referred to a neurologic clinic for management of headache. She suffered from recurrent headaches since the age of 15 years. The attacks were characterized by recurrent, severe, unilateral or bilateral throbbing headache with accompanying nausea and sometimes vomiting. The frequency of headache attacks was two to three times a month, and each attack usually lasted 1 to 2 days. The headache was triggered by menstruation, mental stress, and lack of sleep. Her general physician advised her to avoid trigger factors and take painkillers when attacks occurred. No prophylactic medication was prescribed. After the age of 47 years, her headache attacks became more frequent and eventually occurred in a daily basis for six months. The pain was diffuse, nonthrobbing, and pressure-like in character. She took analgesics every day to control her symptoms. Her sleep became fragmented. She felt depressed and had asthenia. Her symptoms fulfilled depression criteria, and her depression was eventually diagnosed and treated.

Biological basis

Etiology and pathogenesis

Ample evidence has indicated that the association between migraine and psychiatric comorbidities is not a matter of chance. Two explanations for this association are possible (74; 105). First, migraine and depression or anxiety are causally related, either migraine causing depression or anxiety or the reverse. In this model, an index disorder causes or predisposes to the development of the comorbid disorder. Therefore, this association must be unidirectional in nature. However, epidemiological and clinical evidence confirms that the association between migraine and psychiatric disorders, especially major depression, is bidirectional.

The second explanation is that migraine and depression or anxiety share common underlying pathologic mechanisms. The index and comorbid disorders may represent alternative manifestations of the same underlying factor or factors, or different stages of the same disease. A population-based study from Taiwan found that siblings of migraine patients had a higher propensity to develop depression. Siblings of patients with depression were at an increased risk of migraine as well. This familial coaggregation of these two conditions may suggest shared genetic and/or environmental etiologies (16). Franchini and colleagues reported that mood disorder and migraine familiarity in first-degree relatives was significantly related to the risk for comorbidity (29). Clinical evidence also supports this common underlying hypothesis. Drugs that act on serotonin such as SSRIs and SNRIs can be used effectively, both in the treatment of depression and in the prophylaxis of migraine (112). In 1989, Hudson and Pope proposed a hypothesis of an “affective spectrum disorder” to explain the comorbidity among eight disorders that included fibromyalgia, irritable bowel syndrome, major depression, panic disorder, obsessive compulsive disorder, bulimia, cataplexy, migraine, and attention deficit disorder with hyperactivity (40). They suggested that all these disorders exhibited shared phenomenology, family history, and treatment response to antidepressant medications (73). This model is in accordance with the observed bidirectional association. This might be the result of a common, albeit unknown, pathophysiology.

Psychosocial factors have also been mentioned. External locus of control is associated with higher level of depression, poor pain coping strategies, and greater disability. Concept of locus of control, which was introduced by Rotter, describes an individual’s approach to interpreting and attributing events (97). Scharff and colleagues found that external locus of control was significantly related to headache intensity as well as to the patient’s perception of the extent to which pain interfered with many domains of their lives (100). Variances explained that headache-related disability is accounted for independently by locus of control and self-efficacy belief (30). Concept of learned helplessness has also been proposed. Such helplessness caused by uncontrolled migraine attacks brought about chronic recurrent migraine and also depression (104).

In the “shared mechanism model,” pleiotropic effects of the same genes could lead to different clinical manifestations (migraine, depression, or anxiety) depending on the background genes and the intrinsic and extrinsic environment in which they are expressed (74). Shared genetic vulnerability is the most likely explanation for migraine and comorbid depression. A study of 758 monozygotic and 306 dizygotic female pairs showed that heritability was estimated to be 58% (95% CI 48% to 67%) for depression and 44% (95% CI 32% to 56%) for migraine (101). Using bivariate structural equation modeling, the authors estimated that 20% of the variability in depression and migraine headaches was due to shared genes, and 4% was due to shared unique environmental factors. The shared genetic factor hypothesis is also supported by a study in genetic isolate population. Stam and colleagues investigated the contribution of shared genetic factors in migraine and depression by comparing heritability estimates for migraine with and without adjustment for symptoms of depression, and by comparing the heritability scores of depression between migraineurs and controls (109). The results showed that the heritability estimates were significant for all migraine (0.56), migraine without aura (0.77), and migraine with aura (0.96) and decreased after adjustment for symptoms of depression or use of antidepressant medication, especially in migraine with aura. Comparison of the heritability scores for depression between patients with migraine and controls showed a genetic correlation between Hospital Anxiety and Depression Scale score and migraine with aura. A large metaanalysis published in 2018 also supported the shared genetic factors underlying migraine and depression. Results for 8,045,569 single-nucleotide polymorphisms (SNPs) from a migraine genome-wide association study (GWAS) (comprising 30,465 migraine cases and 143,147 controls) and the top 10,000 single-nucleotide polymorphisms from a major depressive disorder genome-wide association study (comprising 75,607 major depressive disorder cases and 231,747 controls) implicated three single-nucleotide polymorphisms (rs146377178, rs672931, and rs11858956) with novel genome-wide significant association to migraine and depression. Combining results across migraine and major depressive disorder, two genes, ANKDD1B and KCNK5, produced Fisher combined gene-based P values that surpassed the genome-wide significance threshold. Pathway analysis of genes suggested several pathways involved in migraine and depression, most notably neural-related pathways of signalling and ion channel regulation (121). Other genome-wide association studies also identified several single-nucleotide polymorphism variants associated with insomnia and subjective cognitive decline in migraineurs (01; 122).

Derangement in aminergic activity, especially serotonin and dopamine, in the central nervous system is the most likely hypothesis for explaining the comorbidities. It is known that this central aminergic system plays a pivotal role in controlling various behaviors, ie, sleep-wake, feeding, emotional level, etc. Change in the amine system in persons with migraine with comorbid depression has been documented. In 1995, Merikangas and colleagues reported lower tyramine sulfate excretion values among persons with migraine and depression compared to those in migraine alone or depression alone (76). Based on these findings, the authors suggested that comorbid migraine with depression may represent a more severe form of migraine than migraine alone.

Serotonin and its subtypes play major roles in pathogenesis of depression, anxiety, and migraine. Decreased central serotonergic activity is a main theory in pathogenesis of depression. Depletion of platelet serotonin and increased level of urinary 5-hydroxyindole acetic acid have been reported during the attack of migraine. It is interesting that the reduction of serotonin concentration is more prominent in persons with migraine complicated with medication overuse headache, the condition in which the risk of having comorbid depression is greater (108). The hypothesis of hyposerotonin is supported by the clinical observation that drugs that enhance serotonergic function, such as tricyclic antidepressants or selective serotonin reuptake inhibitors, are effective in treatment of both migraine and depression.

A growing number of studies have suggested the effect of the microbiota and gut-brain axis on the serotonergic system, which underlies migraine and depression. Dysbiosis of the gut microbiota can impact serotonin production. The use of probiotics has been found to help restore microbial balance, which could potentially lead to improved outcomes of migraine and depression treatment (21).

A great number of inflammatory biomarkers (eg, TNF-α, CRP, and IL-6), cortisol, and ACTH are shown to be elevated in patients with major depressive disorder. These biomarkers are decreased in treatment responsive patients. In contrast, the level of these biomarkers remains elevated in SSRI/SNRI resistant patients (31).

Dopamine has been noted with its effect on yawning, mood change, nausea, and vomiting, which are common in the prodromal phase of migraine. Moreover, antidopaminergic compounds are used effectively in helping to relieve these symptoms. Patients with migraine have an increased density of dopamine receptors on peripheral lymphocytes, which reflects hypofunction of the dopaminergic system (03). Migraine with aura, anxiety disorder, and major depression can be components of a distinct syndrome associated with allelic variations within the DRD2 gene (87). However, one genetic study does not support the role of the dopaminergic system in migraine and comorbid panic disorder. Stochino and colleagues showed that the allele frequencies of DRD1, DRD3, DRD5, and DRD2 in persons with migraine with aura with comorbid panic disorder did not differ from that of parental nontransmitted chromosomes (110). GABA was another neurotransmitter found to be associated with both migraine and depression. Vieira and colleagues examined CSF GABA levels and found that chronic migraine patients with depression had significantly lower CSF GABA levels than those without (116).

Using a magnetic resonance spectroscopy to examine the ratios of several brain metabolites to total creatinine in migraine patients with and without depression, Lirng and colleagues suggested that an increased myo-inositol level in the dorsolateral prefrontal cortex might be implicated in the comorbidity of depression in migraine patients, with myo-inositol/total creatinine ratio in the right dorsolateral prefrontal cortex positively correlated with the severity of depression, measured by the Beck Depression Inventory (57).

Association between mitochondrial dysfunction or depression and migraine comorbidity, together with irritable bowel syndrome were demonstrated by Burnett and colleagues (15). The investigators recruited 166 families with at least one member who had been diagnosed with mitochondrial disease by a physician. Probable maternal inheritance group and probable nonmaternal inheritance group were identified and compared. The results showed that those from the former group had a higher prevalence of depression, migraine, and bowel dysmotility with statistical significance. The hypothesis that mitochondrial dysfunction is a significant common factor underlying the association of these three conditions in the general population was proposed (15).

A theory of sinus hypoxic nitric oxide proposes that migraine is triggered by diffused paranasal nitric oxide through nasal mucosa (95). Trigeminal nerve impulse is increased by this fashion resulting in central neurotransmitter dysfunction and neurodegeneration. Such effects are expressed as aggravated psychiatric disorders.

In summary, mechanism of association has been so far discarded in the hypothesis of chance or direct causation, leaving the common factors hypothesis to be verified further. Common biological factors are by far the most likely explanation of the association between migraine and psychiatric disorders. However, the exact mechanism of such factors is still far from clear.

Epidemiology

Comorbidity between migraine and psychiatric disorders has been reported from several countries worldwide, including the United States, Switzerland, Greece, Turkey, Taiwan, Singapore, and Italy. The information obtained from these countries is rather consistent and shows the lifetime prevalence of major depression in persons with migraine to be three to five times higher than in those without (35; 13; 49). A case-control study revealed strong association between depression and migraine with aura with odds ratio of 5.6 (98). A nationwide survey in Canada showed that migraine was associated with major depressive disorder, bipolar disorder, panic disorder, and social phobia (42). The lifetime prevalence of these psychiatric disorders was more than twice as high in those with migraines compared with those without. The findings from this study suggested that migraine was not associated with alcohol, drug, or substance dependence. However, evidence from another study demonstrated some associations between migraine and alcoholism (43). The higher prevalence of psychiatric disorders in migraineurs was not related to sociodemographic variables (42).

Risk of having comorbid depression is higher in the persons suffering from chronic migraine and those with analgesic overuse headache (77). Mathew and colleagues reported that 46% of patients with transformed migraine who visited Houston Headache Clinic had clinical depression (65). Subsequent study from the same clinic showed that persons with transformed migraine had higher scores in Zung Depression Scale, Beck Depression Inventory, and Type A Behavioral Pattern compared with those with episodic migraine (64). The prevalence of depression is higher in persons with migraine with chronic substance-induced headache (odds ratio, 8.7; 95% CI 1.78 to 42.9) compared to persons with migraine without analgesic overuse (92). A study from Taiwan showed that 78% of persons with transformed migraine had psychiatric comorbidity, namely: major depression (57%), panic (30%), dysthymia (11%), and generalized anxiety disorder (8%). The authors also suggested that women and patients with transformed migraine were at higher risk of psychiatric morbidity (44). However, a study called the Eurolight project reported contradictory findings that migraine is associated with anxiety disorder than mood disorder. The ORs were 2.1 [1.3-3.4] (p = 0.002) for depression and 4.2 [2.8-6.3] (p < 0.0001) for anxiety (54). Moreover, Magnusson and Becker showed that depression scores obtained from persons with transformed migraine and episodic migraine were not significantly different (62). Psychological distress and impaired quality of life are associated with frequent headache and frequent disability but not with severity of headache (63; 84).

Prevention

Factors increasing psychiatric comorbidity include migraine with aura, chronicity of migraine, female gender, analgesic overuse, and coexisting painful conditions. Higher risk of having comorbid depression and anxiety disorders (panic or generalized anxiety disorders) has also been identified in migraine patients who also suffer from other pain syndromes such as arthritis or back pain (68). Other painful conditions in patients with migraine, therefore, need to be identified and properly treated to prevent possible comorbid anxiety and depression later on.

Psychiatric comorbidities are more prevalent in persons with frequent migraine attacks with or without analgesics or ergot overuse. The higher risk of having depression may be the result of chronicity of the disease. Early and effective treatment of migraine is then highly recommended. A report from secondary data analysis of the Treatment of Severe Migraine trial conducted in the United States showed that migraineurs with comorbid mood and anxiety disorders may benefit more from preventive therapy than those without (103). Comorbid mood disorders should be looked for, and prompt treatment should be employed.

Several drugs used in aborting or preventing the attacks of migraine may, however, influence the risk of having comorbid psychiatric symptoms. Clinical depression can be uncovered or worsened by some prophylactic medications, such as beta-adrenergic blockers or calcium channel blockers. Caffeine-containing compounds, which are used as abortive medications, can cause palpitation and trigger the panic attack. Therefore, physicians should prescribe those with caution, especially in the persons with higher risk.

Differential diagnosis

Mood changes can be observed during the prodromal period of migraine. Such mood changes should not be diagnosed as depressive or manic episodes.

Diagnostic workup

Both migraine and its comorbid psychiatric conditions so far have no specific biological markers. The diagnoses are made solely on clinical symptoms. Psychological tests for depression or anxiety are useful in epidemiological study but are less beneficial in clinical situations. Laboratory investigations including neuroimaging studies are useful in excluding the secondary causes that can mimic migraine--for example, vascular lesions such as arteriovenous malformation. They can also identify and exclude lesions in subcortical structures, such as the thalamus, that can cause depression.

Management

The poorly understood mechanism underlying the co-occurrence of migraine and psychiatric disorders poses a challenge in treatment. For instance, one study shows that migraineurs who suffer from depression are more likely to be refractory to migraine treatments, resulting in medication-overuse headache and disability (85). Furthermore, it was shown that migraine patients with higher degree of psychiatric symptoms were less likely to seek psychiatric help for fear of stigmatization, which could lead to underdiagnosis and undertreatment (51).

To provide effective management of patients with migraine and psychiatric disorders, therefore, a comprehensive approach of both diseases simultaneously is required. First and foremost, physicians must be aware of the high prevalence of such comorbidities. Physicians should elicit from the history any mood and anxiety symptoms when patients present their symptoms of migraine. Further exploration into past history is also beneficial. Second, before prescribing specific medication, history of multiple medical treatments or drug used should be explored. Generally, tricyclic antidepressants such as amitriptyline, nortriptyline, or doxepin work well in treating migraine and depression. Dosage of these medications should be high enough to reach antidepressant levels, which are usually higher than those for controlling migraine attacks. With the higher dosages, some adverse effects, especially anticholinergic effects, can be severe and limit the use of these medications. A selective serotonin reuptake inhibitor, such as fluoxetine, can be an alternative. This group of drugs is useful in treating depression and has fewer side effects (104; 91). Education about side effects of antidepressants is crucial, and emphasis on the benefits of regular and long term use of the medication is necessary. Explaining to patients that antidepressants take time to show their effects and that patients must tolerate the drugs’ side effects at the beginning of treatment helps increase compliance. In those not responding to conventional treatments, novel approaches, such as CGRP-targeting therapies, vagal nerve stimulation, or transcranial magnetic stimulation, could be considered (17). Third, supporting psychotherapy (which stresses an increasing internal locus of control, healthy lifestyle, and adhering to medication) benefits patients with long-term suffering from both migraine and mood disorders. Cognitive behavioral therapy has a prophylactic efficacy to migraine and treatment efficacy to mood disorders. The combination of behavioral therapy with prophylactic medication creates a synergistic effect, increasing efficacy beyond either type of treatment alone. Cognitive behavioral therapy has earned an important place in the comprehensive treatment of patients with episodic migraine (52; 55). Social support may also play an important role in the preventing or reducing migraine attacks, as it was found to be inversely correlated with the severity of depression and anxiety in migraine patients (22). Finally, in patients who are refractory to treatment, liaison with a psychiatrist is recommended.