Headache Drugs Logo Robbins Headache Clinic
Home | About Dr. Robbins | Archived Articles | Headache Books | Topic Index | Links  

Back to List



The Prevalence of Personality Disorders
in Migraineurs
Lawrence Robbins, M.D.
Posted: April 2007  



This screening study assessed 1000 consecutive migraine patients at one headache clinic for Axis II (Personality Disorder (PD)) pathology. The purpose was to determine the prevalence of moderate or severe PD within a headache population.


1000 patients with IHS diagnosis of migraine (with or without tension headache) were screened. The patients were 30 to 92 years old, 805 women and 195 men. All patients had been assessed via interviews, as well as an intake psychiatric assessment form adapted for headache patients. The diagnosis of PD was done in accordance with DSM-IV criteria. The patients identified as PD fulfilled the criteria, and were also considered moderate or severe; the PD characteristics were pervasive, greatly influencing functioning and social interactions.

RESULTS   (1000 patients)
CLUSTER A PD: Paranoid: 2 (M) .2% Schizoid: 1 (F) .1% Schizotypal: 0
CLUSTER B PD: Antisocial: 3 (2M, 1F) .3% Borderline: 12 (2M, 0F) 1.2%
Histrionic: 5 (F) .5% Narcissistic: 4 (2M, 2F) .4%
CLUSTER C PD: Avoidant: 11 (1M,10F) 1.1% Dependent: 7 (7F) .7%
Obsessive-compulsive: 7 (7F) .7%
Total PD spectrum (Moderate or Severe): 55/1000 (5.5%); women=46/805 (5.7%) men=9/195 (4.6%).


In this study 5.5% of migraineurs had a moderate or severe PD. The most commonly seen PD’s were borderline (1.2%) and avoidant (1.1%). The recognition of PD greatly affects our approach and management of the pain patient.


Personality disorders (PD) are currently categorized as Cluster A (paranoid, schizoid, schizotypal), Cluster B (antisocial, borderline, histrionic, narcissistic) and Cluster C (avoidant, dependent, obsessive-compulsive). (1) Many people possess a number of characteristics of several PD’s, and are not easily categorized into one specific PD. This study was undertaken to assess the prevalence of moderate or severe PD’s within a chronic headache population. It is crucial to recognize PD patients, as the approach and treatment needs to be modified. Recognition of the more ‘dangerous’ PD’s is important in order to protect the healthcare providers. Left unrecognized, certain PD’s (paranoid, antisocial, borderline(BPD), narcissistic) increase the risk for legal and regulatory problems. In addition, on rare occasions, there may be threats of physical violence as well. For both the patient’s and the healthcare worker’s wellbeing it is important to be cognizant of PD characteristics.


1000 long-term patients at our headache clinic were evaluated, all with the International Headache Society diagnosis of migraine.(2) The patients had been treated at the clinic for a minimum of 6 months; the only exceptions, where patients were treated for less than 6 months, were those (4 total) who were dismissed quickly from the practice. 805 women and 195 men, ages 20 to 92, were evaluated.

The patients had completed a psychiatric screening intake form, and were interviewed by the treating neurologist (and psychopharmacologist). Charts were reviewed for the purposes of this study.

The diagnosis of PD was done in accordance with the DSM-IV criteria.(3) For this study, patients with ‘PD characteristics’ not fully fulfilling DSM-IV criteria were excluded. Only the patients deemed moderate to severe with PD psychopathology were included. The PD characteristics were pervasive, longstanding, and they greatly influenced social and work functioning. The purpose of this was twofold: 1. to identify patients at risk to themselves (and the healthcare providers), and 2. to exclude those with marginal PD diagnoses.

General characteristics of each PD are as follows: paranoid (suspicious, rigid, thin-skinned, angry, dangerous, doubt loyalty); schizoid (self sufficient, indifferent, detached); schizotypal (eccentric, grandiose); antisocial (exploitative, condescending, opportunistic, deceitful, impulsive, irritable, lack of remorse); borderline (instability, poor self image, abandonment fears, impulsive, suicidal, splitting, angry, depressed, dramatic); histrionic (expressive, attention-seeking, emotional, shallow, theatrical, needy, seductive, childlike, exaggerated, body-oriented); narcissistic (grandiose, lacks empathy, needs admiration, condescending, sense of entitlement, exploitative, envious, arrogant, selfish, uncaring, rageful, insensitive); avoidant (socially inhibited, inadequacy, hypersensitive, fearful, sad, hypervigilant); dependent (submissive, clings, fear of separation, need reassurance, manipulative, naive, lack of confidence); obsessive-compulsive (rigid, preoccupied with orderliness, miserly, perfectionistic, indecisive, cold, hoarding, controlling, anxious, workaholic, demanding).

RESULTS   (1000 patients):
CLUSTER A PD: Paranoid: 2 (M) .2% Schizoid: 1 (F) .1% Schizotypal: 0
CLUSTER B PD: Antisocial: 3 (2M, 1F) .3% Borderline: 12 (2M, 0F) 1.2%
Histrionic: 5 (F) .5% Narcissistic: 4 (2M, 2F) .4%
CLUSTER C PD: Avoidant: 11 (1M,10F) 1.1% Dependent: 7 (7F) .7%
Obsessive-compulsive: 7 (7F) .7%
Total PD spectrum (Moderate or Severe): 55/1000 (5.5%); women=46/805 (5.7%) men=9/195 (4.6%).


In this screening study, 5.5% of migraineurs had a moderate or severe PD. The most commonly observed PD’s were borderline personality disorder (BPD)(1.2%) and avoidant(1.1%).

Personality disorders are not discussed as extensively as Axis I disorders (anxiety/depression/substance abuse). The reasons for this include the lack of funding for research (there are no specific drugs for Axis II), and the difficult nature of the patients. However, PD in pain or headache patients is an important arena, as the psychiatric illness often goes unrecognized or misdiagnosed. This leads to inappropriate treatment, and also may place the healthcare provider at risk. The increased risks associated with dealing with certain PD patients include legal (malpractice), regulatory (an increase in letters to hospital boards and state departments of regulation), and physical (threats or acts of violence).

The differential diagnosis usually includes separating Axis II from bipolar disorder, post-traumatic stress disorder, major depression with anxiety, or dysthymia. PD may certainly be comorbid with these conditions.

We currently have a rather rigid categorical listing of PD’s in DSM-IV. DSM-V will probably be more dimensional and less categorical, possibly with a ‘spectrum approach’. Many PD’s do not easily fit into one category, but are a combination of several.

The etiology of PD’s remains controversial, but certainly has a biological basis. Genetics plays a role. A number of theories, such as early over-attachment or detachment with parents, or childhood trauma, have not been validated. Most likely, PD is a genetically based, biological illness that may be expressed after exposure to environmental influences.(4) A genetic make-up that leads to a higher degree of resilience, in particular 2 long arms of the serotonin transporter gene, may be protective against developing a severe PD.

One previous study on borderline personality disorder(BPD) concluded that BPD comorbid with migraine is associated with increased disability.(5) In addition, there were more women with BPD, a higher degree of medication overuse headache, and headaches were more pervasive among those with BPD. There was a higher degree of depression reported among those with BPD, more unscheduled visits for acute headache treatment, and a lesser chance of adequate response to headache medications. Those with BPD were more severely affected by the headaches, and more likely to be refractory to treatment.(5) Another study indicated that the incidence of BPD was increased in migraineurs.(6) There is ample evidence that transformed migraine is associated with more prevalent psychopathology, including PD, than is episodic migraine. BPD itself is the mental health equivalent of chronic pain. The two most important prognostic indicators for those with a PD are impulsivity and substance abuse.

Treatment for those with PD necessitates a different approach.(7) Limits must be set, including with telephone calls; one previous study indicated that those with personality disorders, and also chronic daily headaches, called the clinic much more often.(8) No abuse of the staff should be tolerated. Referral to other healthcare providers should be suggested, particularly mental health professionals. Psychotherapists and psychiatrists experienced with this population are vital if the patient is to be adequately managed; many of the PD patients do not do well with traditional ‘insight-oriented therapy treatment’, but are best ‘managed’ long-term with a dialectical behavioral approach. For therapy to be beneficial, it must be consistent and long-term. A psychoeducational approach may also help. Unfortunately, many PD patients will not continue in therapy. Our therapeutic goals in the PD patient are relatively modest and limited.

It is easy to become ‘hooked in’ by the drama displayed by certain types of PD’s, particularly those with BPD.(9) The patient with BPD may grant us ‘power’, but the patient then subverts the therapy. An example might be “Doctor, you are the greatest, these headaches ruin my life, only you can help me…….but I know that nothing is going to work!” Some physicians are able to manage these patients without becoming hooked in by the drama and countertransference, but most do not do well with these patients. At some point, referral and/or dismissal of the patient may be necessary. If there are signs of a ‘dangerous’ PD from the outset(the first phone call to the clinic/the first visit), with abuse and anger, at times it is best to refer the patient prior to becoming enmeshed in the relationship.

There are risks inherent in caring for those with certain PD’s. As compared with the general population, those with BPD have an increased risk for suicide, particularly as they progress into middle age. Risk factors for suicide among those with BPD include:1. repeated (at least 5 times) failed psychiatric treatment, 2. recent (within 5 days) psychiatric hospitalization, and 3. (possibly) birth trauma, particularly for adolescents with BPD.(10) Certain types of PD (paranoid, narcissistic, antisocial, borderline) are more likely to become angry with healthcare providers, with subsequent threats of lawsuits or letters to departments of regulation. Violence may be a real threat. Limits, documentation, and (if necessary) dismissal are important in these situations. A PD patient often enters as a ‘victim’, but may rapidly flip into the role of persecutor; they focus their anger intensely, creating a hazardous environment for healthcare workers.

Medications, while limited, may be beneficial for impulsivity, aggression, self-mutilation, anxiety, and depression.(11) While there are no specific medications indicated for those with a PD, the Axis I symptoms are more amenable to pharmacotherapy. Antidepressants, mood stabilizers, and antipsychotics may ameliorate certain symptoms. Some of these medications may also lessen the headaches as well. It is important to limit and closely monitor addicting medications. PD patients with severe pain present additional challenges for treatment. Particularly with BPD, opioids and benzodiazepines are best avoided, if possible.

The clinical stakes for not recognizing personality disorders are enormous. The diagnosis of a moderate or severe personality disorder alters both our goals and approach.


I would like to acknowledge Lindsay Roshto and Caroline Kaminski for research assistance.


  1. Lester G. Personality Disorders in Social Work and Health Care. Nashville: Cross Country University Press;2002:28-79.
  2. Headache Classification Committee of the International Headache Society. The International Classification of Headache Disorders, 2nd edition. Cephalalgia. 2004;24(suppl 1)1-160.
  3. American Psychiatric Association. Diagnostic and statistical manual of mental disorders, 4th edition. Washington DC: American Psychiatric Association:1994.
  4. Lester G. Borderline Personality Disorder. Treatment and Management That Works. Nashville: Cross Country University Press;2005:24-25.
  5. Rothrock J, et al. Borderline Personality Disorder and Migraine. Headache. 2007; 47:22-26.
  6. Hegarty AM. The prevalence of migraine in borderline personality disorder. Headache. 1993;33:271.
  7. Saper J, Lake A. Borderline personality disorder and the chronic headache patients: Review and management recommendations. Headache. 2002;42:663-674.
  8. Loder E, Geweke L. Volume and nature of telephone calls in a specialty headache practice. Headache. 2002;42(9):883-887.
  9. Lester G. Borderline Personality Disorder. Treatment and Management That Works. Nashville: Cross Country University Press;2005:117-128.
  10. Lester G. Borderline Personality Disorder. Treatment and Management That Works. Nashville: Cross Country University Press;2005:15-19.
  11. Lester G. Borderline Personality Disorder. Treatment and Management That Works. Nashville: Cross Country University Press;2005:88-91.