Understanding the nature of psychiatric comorbidity in migraine: a systematic review focused on interactions and treatment implications

Both migraine and psychiatric disorders are prevalent and burdensome conditions challenging the health care systems worldwide [1‐5]. These conditions show a large overlap [6, 7] and epidemiological studies suggest that patients with migraine – especially those with chronic migraine (CM) and migraine with aura – are at increased risk for major depression, anxiety, or suicidal behavior when compared to subjects without migraine (e.g., [8‐10]). Besides, according to a recent large genome-wide association study, when compared to other neurological disorders, migraine showed a higher genetic correlation with psychiatric disorders suggesting common genetic bases or pathways [11]. The comorbidity between migraine and psychiatric disorders is highly relevant in the clinical practice, as it might influence both the response to treatment and likelihood to achieve remission [6]. Therefore, an interdisciplinary approach using pharmacological and non-pharmacological treatment strategies aimed to manage both migraine and comorbid psychiatric disorder is essential.

The comorbidity between migraine and psychiatric disorders presumably implies multiple causes, including either unidirectional causal explanations, but even shared environmental and/or genetic risk factors, and their interaction at multiple levels [12]. All the mentioned aspects need to be carefully considered regarding the diagnostic and therapeutic implications related to migraine comorbidity.

This systematic review will provide an updated and comprehensive overview of the current literature focusing on the comorbidity of migraine with depression, bipolar disorder (BD), anxiety disorders, post-traumatic stress disorder (PTSD), and other psychiatric disorders, including personality, substance use, and somatoform disorders, outlining the general findings, potential mechanisms of association, and implications for migraine treatment.

To achieve a high standard of reporting, we selected the most relevant studies in accordance with the PRISMA guidelines [13]. We included observational studies that explicitly and clearly report the adopted definitions of migraine and its psychiatric comorbidities; we also included clinical trials, open label studies, systematic reviews, guidelines, commentaries, editorials, and letters to editors focusing on the review topic. When a title/abstract appeared to describe a study eligible for inclusion, the full-text article was carefully analyzed to evaluate its relevance for our systematic review. Eligible papers had to be written in English and published from 1 January 1998 to 15 July 2018 on the following 3 major electronic databases: PubMed, Scopus, or ScienceDirect. The following search string was used in all databases: (“Headache” OR “migraine”) AND “comorbidity” AND (“psychiatric disorders” OR “substance abuse” OR “personality disorders” OR “major affective disorders” OR “bipolar disorder” OR “unipolar disorders” OR “psychotic disorders” OR “psychoses” OR “suicid*” OR “anxiety” OR “schizophrenia” OR “depression”). The reference lists of the retrieved articles were also screened to find eligible studies not covered by the above-mentioned search string. Two Reviewers (MP and GSf) conducted the literature search and independently screened titles and abstracts; later, they retrieved and selected full-text articles. Disagreements among these two Reviewers were solved by consensus. The relevant steps and main results of the literature search are shown in Fig. 1.

Major depressive episodes are characterized by periods of at least two weeks with symptoms including depressed or irritable mood, decreased interest or pleasure in most activities, significant weight change, change in sleep (insomnia or hypersomnia), change in activity (psychomotor agitation or retardation), fatigue or loss of energy, guilt/worthlessness, diminished ability to think or concentrate or more indecisiveness, and suicidality [14].

Depression is almost twice as frequent in patients with migraine when compared to the general population. The prevalence estimates in migraine vary across countries, from 6.1% to 73.7% (e.g., [15‐17]) while the prevalence odds ratios vary from 0.8 to 5.8 (see Table 1). Such differences may be mainly due to different sex, age, and ethnic composition of study samples, as well as from different psychometric instruments aimed to assess migraine and depression [18]. Comorbid major depression is linked to more frequent and disabling headache [19]. Notably, the risk of suicide attempts is particularly higher in migraine patients with comorbid anxiety and depressive symptoms [20].

BD is characterized by a periodic course of depressive episodes and episodes with exceptionally increased mood (mania or hypomania). It is commonly divided into BD type I (at least one manic or mixed episodes) and BD type II (at least one hypomanic, but no full manic phase), with BD I usually impairing more severely the individual functioning [14].

Patients with BD display an increased prevalence of migraine that can reach up to 55.3% (Table 1), although – as stated above in the case of major depression – rates may considerably vary across countries due to several factors. Migraine prevalence seems to be higher in BD II than BD I [43, 44], and mostly migraine precedes the onset of BD [45]. In a population-based study, the prevalence of migraine was higher in subjects with both manic and depressive episodes than in those with depressive episodes only [46]. The available data suggest that BD has a more severe course when it is comorbid with migraine [47].

Table 1 shows the quantitative data regarding the association between migraine and anxiety. Notably, the prevalence of anxiety increases with migraine frequency [72, 73], suggesting a ‘dose-response’ effect; the comorbidity between migraine and anxiety disorders is also enhanced by the presence of medication overuse [74] and concurrent depression [75]. As for major depression, the risk of suicide attempts is increased in patients with migraine and anxiety disorders [20].

The relation between stress and migraine is bidirectional and may be direct or indirect. Patients with migraine report higher stress levels when compared to healthy controls (e.g., [109]), and stressors are usually reported as migraine triggers [84, 110‐115]. Conversely, migraine itself acts as a stressor resulting in a vicious circle with a strong impact on important individual domains such as work and social functioning [116]. Stress exposure might also mediate the association between migraine and other psychiatric comorbidities, including major depression [112], and is implied in the transformation of headache into chronic headache [117, 118].

The development of PTSD mandatorily depends on the direct or indirect exposure to traumatic life events and is characterized by intrusive symptoms, avoidance, and negative alterations in both cognitions and mood [14]. PTSD is related to the development of pain disorders [119], with a higher prevalence in patients with migraine, and mostly CM, compared to healthy controls [120]. Notably, the available studies found a higher incidence of childhood abuse in migraine patients with BD or depression as compared to those with migraine only [110, 121]. There is also evidence suggesting that PTSD, but not the mere exposure to a traumatic event, is correlated with migraine [122], while in the absence of definite PTSD, only repetitive (≥3) traumatic events enhance the risk for migraine [123‐126]. Besides, CM is associated with a higher susceptibility for PTSD compared to EM; indeed, patients with CM were more influenced by traumatic events, as evident in more avoidant and re-experiencing symptoms when compared to patients with EM [127].

PTSD is more frequent among patients with CM when compared to those with chronic tension-type headache [128], suggesting that the association is specific to migraine. Moreover, PTSD with comorbid depression was associated with a higher risk of migraine chronification when compared to depression alone [120]. In line with those findings, an Italian study found that major traumatic events were associated with CM and MOH [129]. Not surprisingly, patients with migraine and comorbid PTSD report greater headache-related disability and quality of life impairment than those with migraine only [120, 122, 130].

Overall, our review underlines the consistent association between migraine and psychiatric disorders (e.g., major depression, bipolar disorder, anxiety disorders, PTSD, and other mental disorders including substance-related behavior and personality disorders) that might be attributed to common etiological (both environmental or genetic) factors or to the existence of bidirectional relations between the disorders, as for major depression and panic disorder. Although the actual nature of this complex association between neurological and mental disorders is difficult to determine given the available studies, for therapists it is important to recognize it and to include it into the diagnostic and therapeutic process [161] (see Table 2). This also requires a more intense collaboration between the disciplines of neurology and psychiatry, which are mostly taught separately in medical schools.

The present systematic review should be considered in the light of the following limitations/shortcomings. For instance, most studies included in the present review used cross-sectional designs, limiting the assessment of causal relations between phenomena (e.g., [161]). Besides, the distribution of sex, age, and migraine subtype, as well as methods to diagnose migraine and psychiatric disorders were largely heterogeneous among the different studies, limiting the possibility of quantifying the real impact of comorbidities in the general population (for more details, see Table 1). Moreover, there are few and mostly indirect data examining the neural mechanisms underlying the comorbidity between migraine and psychiatric comorbidities. Having more than only one psychiatric comorbidity (e.g., concurrent anxiety disorder and major depression) poses another caveat which is often not sufficiently considered in existing studies, but may have further consequences on the course and treatment of migraine [25, 120, 160]. All the mentioned criticisms once again enhance the complexity of migraine and psychiatric comorbidities regarding etiology, pathophysiology or interactions over time. However, the more consistent implications from the reviewed studies have been summarized in Table 2.

Major affective and anxiety disorders have been shown to be the most frequent and disabling psychiatric comorbidities associated with migraine, influencing its clinical course, treatment response, and clinical outcome. Here, the comorbidity has important clinical and therapeutic implications, demanding specific attention by practitioners. For instance, having a comorbid major depression or anxiety disorder can increase the likelihood of suicide attempts in patients with migraine [20]. The comorbidity with psychiatric disorders is not uniformly increased in the different migraine subtypes, but generally more elevated in patients with CM or migraine with aura, suggesting the need for more specific care in those patients.

Genetic variants, dysfunction in neurotransmitters (especially 5-HT), and HPA axis dysregulation are among the most supported pathophysiological mechanisms underlying the comorbidity between migraine and depression. Specific neural network patterns overlap between both entities, which may be the result of the above hypothesized mechanisms. Unfortunately, no specific and valid biomarkers have been documented for the risk of comorbid migraine and depression. Still, it is not sufficiently clear, if the observed biological or chemical parameters are indicators of specific underlying etiological and pathogenic pathways or if they represent epiphenomena.

The relation between migraine and major depression is likely to be bidirectional. Regarding the bipolar spectrum, clinicians need to suspect and identify the possible comorbidity between migraine and BD, especially among female patients and subjects with BD II. Here, the possible misidentification of unipolar depressive subtypes should be avoided, as pharmacological indications for the two conditions widely differ or may be counterproductive.

Whether the comorbidity between migraine and anxiety is uni- or bidirectional is a matter of debate. Excess worry, fear, and other anxiety symptoms such as avoidance behavior are usually part of the clinical migraine presentation, while, conversely, headache symptoms may also be part of a clinical anxiety disorder. This implies the need to correctly identify the clinical characteristics of both these conditions, to derive a working hypothesis regarding which of the disorders may be primary or secondary, as the adequate treatment of the primary condition may be beneficial for the outcome of the secondary condition. PD appears to be more consistently associated with migraine than most other anxiety disorders. Comorbid PD is associated with greater health care costs, higher disability, and functional impairment as well as risk for chronification, medication overuse, and henceforth MOH. As with migraine and depression, the relation between PD and migraine appears to be bidirectional. The link between PTSD and migraine seems to be more evident in chronic forms, as the repeated exposure to stressful situations/factors causes cortical response modification with modulation of the vascular trigeminal system leading to a lower pain threshold.

There are few and sometimes conflicting studies in the current literature concerning the comorbidity between migraine and psychiatric disorders associated with minor incidence in the general population, including personality disorders, substance use disorders, somatoform disorders, and eating disorders. Here, the association may be also due to or mediated by concurrent major depression. Specific heed should be paid to MOH, which is more common in patients with comorbid psychiatric disorders [162]. In this condition, pain-relieving drugs for the acute treatment of headache symptoms may themselves lead to headache when taken frequently for long periods. Here, therapists should explain this phenomenon in detail to the patients in order to assure a proper use of prescribed or over-the-counter medication. In some cases of MOH, comorbid psychiatric symptoms might be considered an epiphenomenon of medication overuse – at least animal data imply such a mechanism [163].

Reviewing the current literature, we are still far from comprehensively understanding the biological origin/axis underlying the migraine and psychiatric disorder comorbidity. For most of the comorbid disorders, the commonly mentioned pathogenic players such as heritability, specific genes or neurotransmitter systems may play a significant role (Fig. 2). As influences seem to be generally complex and dynamic in their nature over time, simple therapeutic solutions are not available and combined approaches are required. Instead, the involvement of different disciplines is needed in order to carefully account for each patient’s disorder and treatment history in a multimodal treatment approach perspective.

Clarifying the comorbidity between psychiatric disorders and migraine is essential not only at the clinical diagnostic level but above all for the complex therapeutic implications of such comorbidity [164]. CBT has been shown in several population studies to be a valid alternative in addition to pharmacological treatments in patients with migraine and psychiatric comorbidity. Importantly, the pharmacological prophylaxis of migraine might be influenced by psychiatric comorbidities. Many biological and neural aspects related to the comorbidity still need to be clearly elucidated to better approach the real complexity of this issue.

Our intention is to conclude this review emphasizing three most relevant key points for clinicians: