Current Pain and Headache Reports

, 18:405

Psychiatric Comorbidity of Chronic Daily Headache: Focus on Traumatic Experiences in Childhood, Post-Traumatic Stress Disorder and Suicidality


    • Department of PsychiatryTaichung Veterans General Hospital
    • School of MedicineNational Yang Ming University
  • Chin-Yi Yang
    • Division of PsychiatryTon Yen General Hospital
Chronic Daily Headache (SJ Wang, Section Editor)

DOI: 10.1007/s11916-014-0405-8

Cite this article as:
Juang, K.D. & Yang, C. Curr Pain Headache Rep (2014) 18: 405. doi:10.1007/s11916-014-0405-8
Part of the following topical collections:
  1. Topical Collection on Chronic Daily Headache


The fifth edition of the Diagnostic and Statistic Manual (DSM-5) reclassified some mental disorders recently. Post-traumatic stress disorder (PTSD) is in a new section termed “trauma- and stressor-related disorder”. Community-based studies have shown that PTSD is associated with a notably high suicidal risk. In addition to previous findings of comorbidity between chronic daily headache (CDH) and both depressive disorders and anxiety disorders, recent data suggest that frequency of childhood maltreatment, PTSD, and suicidality are also increased in CDH. CDH patients with migraine aura are especially at risk of suicidal ideation. Research suggests that migraine attack, aura, frequency, and chronicity may all be related to serotonergic dysfunction. Vulnerability to PTSD and suicidality are also linked to brain serotonin function, including polymorphisms in the serotonin transporter gene (5-HTTLPR). In the present review, we focus on recent advances in knowledge of traumatic experiences in childhood, PTSD, and suicidality in relation to migraine and CDH. We hypothesize that vulnerability to PTSD is associated with migraine attack, migraine aura, and CDH. We further postulate that these associations may explain some of the elevated suicidal risks among patients with migraine, migraine aura, and/or CDH. Field studies are required to support these hypotheses.


Chronic daily headachePsychiatric comorbidityTraumatic experiences in childhoodPost-traumatic stress disorderSuicidality


Psychiatric comorbidity of headache has attracted a great deal of interest from clinicians and researchers for decades [13, 4•]. The presence of psychiatric disorders may affect prognosis and treatment of headache disorders [5, 6•, 7]. Furthermore, patterns of psychiatric comorbidity may elucidate common mechanisms underlying headache and psychiatric disorders [1, 4•, 6•, 8•]. Psychiatric comorbidity may also have implications for classification of headache disorders. For instance, a type of headache that is highly associated with suicide may warrant a distinct diagnostic category to raise awareness of this risk among clinicians. Different patterns of psychiatric comorbidity between two headache conditions may reflect differences in their underlying mechanisms, and hence may indicate a need for separate classifications [9••].

Chronic daily headache (CDH) is a disabling and distressing disorder [10]. The third edition of the International Classification of Headache Disorders (ICHD-3) defines CDH as headache occurring an average of at least 15 days per month for more than 3 months [11]. The most frequent subtypes of CDH are: 1) chronic tension-type headache (CTTH); and 2) chronic migraine (CM) [12], which requires at least 8 days of migrainous features per month for diagnosis by ICHD-3 [11]. Approximately 4–6 % of the population has CDH [8•, 10, 12]. Not surprisingly, CDH has a high burden on society. Patients with CDH suffer from persistent pain and may leave their job or school [7, 10, 13]. Comorbid psychiatric disorders, such as depressive and anxiety disorders, further complicate quality of life and increase suicidal risk [1, 7, 10, 14]. Depressive and anxiety disorders involve closely related monoamines in the brain, including serotonin, noradrenaline and dopamine [15, 16•], which also play important roles in migraine [17] and in migraine chronization [4•, 10, 17]. It is postulated that migraine, CDH, and psychiatric disorders may share common monoaminergic mechanisms.

Earlier reviews have described possible roles of monoamines in the associations among migraine, CDH, and psychiatric disorders [4•, 18]. More recently, a great deal of attention has been paid to childhood traumatic experiences, which appear to be associated with both psychiatric disorders [19•], such as post-traumatic stress disorder (PTSD), and chronic pain [2022] in later life. PTSD is no longer classified as an anxiety disorder in the fifth edition of the Diagnostic and Statistic Manual of Mental Disorders (DSM-5) by the American Psychiatric Association [23]. Instead, DSM-5 includes a new classification of “trauma- and stressor-related disorder” for PTSD and related disorders. In the last decade, pain experts have observed the importance of PTSD in chronic pain [24]. If childhood traumatic experiences and subsequent PTSD are associated with migraine and CDH, they may at least partially explain comorbidity between migraine/CDH and anxiety/depression, since traumatic experiences in childhood and PTSD are also associated with anxiety and depressive disorders [19•, 22].

The chronization of episodic migraine (EM) is crucial to understanding CDH. Lipton summarized risk factors mediating the transformation from EM to CM, including obesity, stressful life events, and medication overuse, as well as frequent migraine attacks [25]. Childhood trauma is a risk factor for obesity in adulthood [26]. Abuse victims are frequently re-victimized in adulthood, resulting in increased stressful life events [27•]. They also tend to have more impulsivity [28], possibly increasing medication overuse due to a need for immediate therapeutic effects. Clearly, many factors related to transformation from EM to CM are associated with child abuse, suggesting that child abuse may contribute to chronization.

Traumatic experiences in childhood may affect brain development, causing disturbance in monoamine systems [29•]. PTSD is also related to monoamine systems and the hypothalamic-pituitary-adrenal (HPA) pathway of cortisol [30, 31]. In addition, estrogen may be involved in both child abuse and migraine [4•, 32]. Child abuse and PTSD may influence the brain to increase vulnerability to migraine or CDH [4•, 30, 32]. These conditions may intricately interact psychologically and biologically. Thus, investigation of the complicated relationship among child abuse, PTSD, migraine, and CDH may elucidate the mechanisms underlying these conditions.

The present review will focus on traumatic experiences of childhood and PTSD, along with suicidal risk in migraine and CDH. We will review data on CDH and EM, including community and clinic-based studies. For more general reviews on psychiatric comorbidity of EM and CM in community and clinical settings, please refer to recent articles by Buse et al. [8•] and Antonaci et al. [6•], respectively.

Traumatic Experiences in Childhood

Traumatic childhood experiences include sexual abuse, physical abuse and other maltreatment, such as neglect. These traumas strongly burden society as a whole [33••]. According to DSM-5 criteria of PTSD, trauma includes exposure to actual or threatened death, serious injury, or sexual violence, resulting from direct experience, in-person witness, learning from a close family member or friend, or repeated or extreme indirect exposure to aversive details [23]. While the DSM-5 criteria is very strict and is used only to diagnose PTSD, trauma is usually defined in a broader sense. For example, severe corporal punishment that does not lead to serious injury may not qualify as trauma according to DSM-5 PTSD criteria, but would be regarded as physical abuse by most experts. Some studies also define emotional abuse and neglect as a kind of child abuse [27•]. Great variation in definitions of traumatic experience makes comparison across studies difficult. In the following sections, we use the phrase “traumatic experiences” to include a wide variety of events that may be harmful to childhood development, including terms such as stressful life events, adversities, maltreatments, and abuse.

Another issue with studies on childhood trauma is that most studies identify traumatic experiences by retrograde self-report. Although some authors argue that current pain may bias recollection of painful experiences [33••, 34], the extent of bias in recall of previous abuse is unknown. Furthermore, prospective studies may have other limitations. Because most abuse is not reported during childhood, victims usually deal with traumatic experiences by themselves, which may induce feelings of loneliness and helplessness. Unreported victims may have vastly different experiences than victims whose abuse is documented and ceased by the court [35]. A prospective study of court-documented victims found no association between childhood abuse and pain in young adulthood [34], and a slight but statistically significant increase in pain compared to controls in mid-life [36•]. Another prospective study, the 1958 British Birth Cohort Study, which repeatedly assessed 17,638 children from childhood, demonstrated a relationship between chronic widespread pain and childhood adverse events, such as hospitalization after traffic accidents (OR = 1.5, 95 % CI: 1.05–2.1) and maternal death (OR = 2.0, 95 % CI: 1.08–3.7) after controlling for demographic confounders [37].

Community studies of childhood traumatic experiences in CDH are rare. We compared childhood adversities in a non-referred sample of CDH in young adolescents in public schools. Teachers reported each subject’s childhood adversities after they interviewed these subjects and their family members following a semi-structured form. Frequencies of physical abuse and parental divorce were significantly higher in the CDH group than in controls (10 % vs. 0 %, p = 0.012 and 17 % vs. 3 %, p = 0.015, respectively). The frequency of child neglect was similar between CDH (5 %) and control groups (2 %, p = 0.618), and no cases of sexual abuse were identified [38].

Several clinic-based studies compared frequencies of traumatic experiences between patients with EM and CM. In a large sample of 1,348 migraineurs from 11 headache clinics, Tietjen et al. found that frequencies of physical and emotional abuse and neglect, collected retrospectively using the Childhood Trauma Questionnaire, were significantly higher in patients with CM. After adjusting for sociodemographic data, current anxiety, and current depression, emotional abuse remained significantly associated with CM (OR = 1.77, 95 % CI: 1.19–2.62) and transformation from EM to CM (OR = 1.89, 95 % CI: 1.25–2.85) [39••]. These results suggest that childhood maltreatment may play a role in chronization of migraine. In support of this hypothesis, Peterlin et al. found a higher frequency of self-reported physical and/or sexual abuse in a series of new patients at a headache center with CM compared to EM [40].

Community studies of childhood trauma in headache subjects often utilize self-report of headache/migraine instead of headache diagnosis. Our study in adolescents is an exception. Using ICHD diagnosis, we found associations between physical maltreatment and migrainous features, headache frequency, and headache severity. Frequencies of physical maltreatment were similar between migraineurs with and without aura [41•]. Self-reported diagnosis of migraine by a health professional was also associated with self-reported childhood physical abuse in a community study in Canada (OR = 1.77, 99 % CI: 1.39–2.25), even after controlling for confounders (OR = 1.36, 99 % CI: 1.04–1.79) [42]. In a community study in New Zealand, self-reported headache/migraine was associated with adult physical abuse, but not childhood physical or sexual abuse [43]. A meta-analysis of three studies found a large effect size of 0.94 (95 % CI: 0.89–0.99) for the association between child abuse and headache/migraine, which is comparable to the sequelae of psychiatric disorders after child abuse [44].

A prospective study showed that frequency of major depression was increased in victims of child abuse (OR = 1.51, 95 % CI: 1.06–2.14) [19•]. While both psychiatric disorders and migraine headaches are associated with childhood traumatic experiences, it is possible that comorbidity of psychiatric disorder and headache may be due in part to childhood trauma. In a clinical setting, Tietjen et al. showed associations among childhood trauma, psychiatric disorder, and migraine [45], suggesting a possible role of childhood trauma in psychiatric comorbidity of migraine.

Post-Traumatic Stress Disorder

DSM-5 defines PTSD as having at least one symptom of intrusive re-experiencing, one symptom of avoidance, two symptoms of negative cognition or mood, and two symptoms of arousal that last for more than one month after exposure to trauma. A specifier of “with dissociative symptoms” (depersonalization and/or derealization) was also added [23]. National Comorbidity Study found the prevalence of PTSD to be 7.8 % [46]. An earlier study by Davidson et al. reported lifetime and 6-month prevalence of PTSD in the community to be 1.30 % and 0.44 %, respectively. They also found PTSD to be strongly related to suicidality, as suicide attempts occurred in 19.8 % of subjects with PTSD, 3.9 % of subjects with other psychiatric diagnosis, and 0.8 % of subjects without psychiatric diagnosis [47]. Raphael et al. found that PTSD exerts a moderator but not mediator effect on the relationship between child abuse and chronic pain. They argued that current cobmorbid PTSD is more important than history of abuse in patients with chronic pain [36•].

PTSD and chronic pain share some common psychological mechanisms, such as anxiety sensitivity and fear avoidance [48, 49]. PTSD is also closely related to serotonergic function, especially 5-HTTLPR polymorphisms [50••]. Both migraine and PTSD are related to serotonin and 5-HTTLPR. Estrogen and the HPA axis may also play a role in the comorbidity of PTSD and migraine [4•, 30, 32]. Despite many shared mechanisms that suggest a close relationship between PTSD and migraine, PTSD has seldom been explored in headache, with the notable exception of post-traumatic headache. The first report of PTSD in headache patients, published in 2005 by Leeuw et al., suggested that PTSD symptoms may be frequent in individuals having migraine and CTTH [51].

In a community sample from the National Comorbidity Survey Replication (NCS-R), Peterlin et al. found that lifetime prevalence of PTSD was 4.5 % in subjects without headache, 12.6 % in subjects with episodic non-migraine headache, 21.5 % in subjects with EM, and 19.2 % in subjects with CDH. Twelve-month prevalence of PTSD was 2.1 % in subjects without headache, 7.8 % in subjects with episodic non-migraine headache, 14.3 % in subjects with EM, and 11.3 % in subjects with CDH. Although prevalence of illicit drug abuse was slightly increased in subjects with migraine and CDH, these differences disappeared after adjusting for comorbid PTSD and/or major depression, suggesting that PTSD and/or major depression accounted for more frequent drug abuse in headache patients [52••].

Peterlin also published two similar clinic-based studies of PTSD [53, 54]. The larger study from six headache centers showed a greater frequency of PTSD in CDH than in EM (30.3 % vs. 22.4 %, p = 0.043), which was no longer significant after adjusting for demographic data and depression. Comorbid major depression and PTSD was significantly more frequent in CDH than in EM (24.6 % vs. 15.8 %, p < 0.002) [53]. PTSD is also comorbid with depressive and anxiety disorders [46], and thus PTSD may explain some of psychiatric comorbidities of EM and CM. However, there is a need for community-based studies in this area.


Suicidality includes suicidal ideation, suicide attempts, and suicide, defined as death from suicidal behavior. Both suicide and migraine are associated with dysfunction of serotonergic neural activity [55••], and pain and suicidality share several other common mechanisms [56••]. It is possible that patients with CDH, suffering from chronic pain, are more vulnerable to suicidality. However, little is known about suicidality in CDH. In a community study of young adolescents, we found that 20 % of young adolescents with CDH had high suicidal risk, defined as a suicidal score of 10 or higher on the MINI-Kid. The prevalence of major depression and panic disorder was 21 % and 19 %, respectively. Frequencies of high suicidal risk were 7.5 % in CDH subjects without migraine, 20.9 % in those with migraine without aura, and 50 % in those with migraine with aura. The association between migraine and high suicidal risk (OR = 3.0, 95 % CI: 1.1–8.1) disappeared after controlling for demographics, depression, and anxiety. However, the presence of migraine aura remained a strong predictor of high suicidal risk relative to subjects without migraine, even after controlling for age, gender, anxiety disorders, and major depression (OR = 7.8, 95 % CI: 1.4–44.6) [9••]. Due to time limits on the psychiatric interview, this study did not assess PTSD or trauma history.

The association between migraine aura and suicidality has been observed for decades. In a community sample of young adult migraineurs, Breslau et al. reported that migraine aura was associated with suicide attempts (OR = 3.0, 95 % CI: 1.4–6.6) after adjusting for psychiatric disorders [57]. We confirmed these findings in young adolescents in a community setting. Frequencies of suicidal ideation were 4.5 % in subjects without headache, 7.9 % in subjects with non-migraine headache, 13.7 % in those with probable migraine, 15.9 % in those with migraine without aura, and 23.9 % in those with migraine with aura. After controlling for demographics and depression score, associations with suicidal ideation remained for migraine with aura (OR = 1.79, 95 % CI: 1.07–2.99) and high headache frequency (more than 7 days per month, OR = 1.69, 95 % CI: 1.12–2.56), but not for the diagnosis of migraine per se. Frequencies of suicide ideation were not different between migraine and non-migraine headache after controlling [55••].

Breslau et al. investigated whether suicidal risk differs between migraine and non-migraine headache in a 2-year follow-up community-based study. Subjects with migraine (OR = 4.43, 95 % CI: 1.93–10.2) and with severe non-migraine headache (OR = 6.20, 95 % CI; 2.40–16.0) showed similarly higher frequencies of suicide attempts relative to controls. The authors concluded that headache severity, rather than headache diagnosis, might account for suicide attempts [58•]. In contrast, Ilgen et al. found a possible difference between migraine and non-migraine headache in suicide. Using clinical records from the Department of Veterans Healthcare System, they showed that rates of suicide from the National Death Index in a 3-year follow-up were increased in both patients diagnosed with migraine headache [by International Classification of Diseases, Ninth Revision, Clinical Modification, (ICD-9-CM), codes 346.0 to 346.9] and those diagnosed with headache/tension headache (ICD-9-CM codes 784.0 and 307.81). Significantly higher rates of suicide were found for migraine (OR = 1.68, 99 % CI: 1.28–2.20) and for headache/tension headache (OR = 1.38, 99 % CI: 1.17–1.64), but after adjusting for age, sex, a medical comorbidity score, and psychiatric disorders, only migraine remained as a significant predictor of suicide (OR = 1.34, 99 % CI: 1.02–1.77) [59].

Several community-based studies have demonstrated a relationship between self-reported headache and suicidality. For example, in the Epidemiological Catchment Area Study, subjects who endorsed the item “ever had a lot of troubles with headache” were at higher risk of having suicidal thoughts and behaviors after adjusting for psychiatric disorders and demographics (OR = 1.48, 95 % CI: 1.04-2.11) [60]. In the National Comorbidity Survey-Replication, subjects who self-reported frequent or severe headache within the previous 12 months had higher risks of suicidal ideation (OR = 1.9, 95 % CI: 1.2-3.0) and suicide attempts (OR = 2.3, 95 % CI: 1.2-4.4) after adjusting for concurrent psychiatric disorders and other chronic medical conditions [61].

PTSD, Migraine Aura, and Suicide: A Hypothesis

What mechanisms underlie the association between migraine with aura and suicidality? A possible role of serotonin has been suggested [55••]. For example, the serotonin transporter may play an important role in the pathophysiology of suicide [62]. Polymorphisms of the serotonin transporter gene (5-HTTLPR) are associated with migraine with aura, but not migraine without aura [55••]. Recent studies suggest that 5-HTTLPR genotype is also related to the vulnerability of developing PTSD after stressful life events [50••]. While migraine, suicidality and PTSD are all closely related to brain serotonergic function, PTSD may mediate or moderate the association between migraine aura and suicidality.

As reviewed above, suicidal risk is very high in subjects with PTSD [47]. The National Comorbidity Survey found that after adjusting for demographics and all other psychiatric disorders, lifetime suicidal ideation (OR = 2.79, 95 % CI: 2.02–3.84) and suicide attempts (OR = 2.67, 95 % CI: 1.82–3.91) were increased in PTSD [63]. There are similarities between symptoms of PTSD and migraine aura. For example, flashbacks or vivid re-experiencing in PTSD may include intrusive visual or other sensory components, similar to migraine aura, which involves visual and other cortical areas. If there is an association between migraine aura and PTSD, then PTSD could partially explain the increased suicidal risk associated with migraine aura. Studies of the relationship between PTSD and migraine aura are needed.

What is the relationship between migraine aura and childhood trauma? In our community-based study of young adolescents, frequency of childhood maltreatment was similar between migraine with and without aura [41•]. Tietjen et al. showed similar results in a clinic-based study [39••]. However, not every childhood trauma leads to PTSD. A prospective study of a randomly sampled community population of children and adolescents aged 9, 11, and 13 years found that traumatic experiences were very frequent (68.2 %), but only 0.5 % developed full-blown PTSD by age 16 [64]. Vulnerability to PTSD is associated with 5-HTTLPR polymorphisms [50••]. Since serotonin is also important in the development of migraine with and without aura, chronic migraine, and migraine frequency [17, 55••], we hypothesize that vulnerability to PTSD after trauma is related to migraine attacks, migraine chronicity, and especially migraine aura, and that these associations account for some of the elevated suicidal risk in patients with these headache features.


Clinical and community studies have found frequent psychiatric comorbidities in CDH [6•, 7, 8•, 9••, 65]. Recent reports further suggest that CDH is also accompanied by frequent childhood traumatic experiences, PTSD, and suicidality. Hence, routine screening for PTSD, suicide, and major depression seems warranted in patients with CDH or those with migraine with aura. In managing CDH complicated by so many psychiatric and psychological conditions, clinicians may recommend psychotherapy. Holroyd et al. demonstrated that patients with CTTH were more likely to experience a clinically significant reduction (50 % or more) in headache index scores from psychotherapy (stress management, relaxation training, and cognitive coping therapy) combined with medication than medication alone (64 % vs. 38 %, p = 0.06) [66]. Future studies should investigate whether patients with CM may benefit from psychotherapy, what treatment plan is best for CDH patients with comorbid depression, anxiety, PTSD, and/or suicidality, and whether different treatment guideline should be implemented for patients with these comorbid psychiatric conditions. A diagnostic entity or specifier of “with aura” may be warranted in CDH or CM due to its high association with suicide [9••], to facilitate identification of at-risk patients.

Compliance with Ethics Guidelines

Conflict of Interest

Dr. Kai Dih Juang and Dr. Chin-Yi Yang each declare no potential conflicts of interest relevant to this article.

Human and Animal Rights and Informed Consent

This article does not contain any studies with human or animal subjects performed by any of the authors.

Copyright information

© Springer Science+Business Media New York 2014