Medication Overuse Headache: Biobehavioral Issues and Solutions

Authors

  • Alvin E. Lake III PhD


  • From the Michigan Head-Pain and Neurological Institute, Ann Arbor, MI.

Address all correspondence to Alvin E. Lake III, PhD, Michigan Head-Pain and Neurological Institute, 3120 Professional Drive, Ann Arbor, MI 48104.

Abstract

This article reviews current research on medication-overuse headache (MOH), and provides clinical suggestions for effective treatment programs. Epidemiological research has identified reliance on analgesics as a predictive factor in headache chronicity. MOH can be distinguished as simple (Type I) or complex (Type II). Simple cases involve relatively short-term drug overuse, relatively modest amounts of overused medications, minimal psychiatric contribution, and no history of relapse after drug withdrawal. In contrast, complex cases often present with multiple psychiatric comorbidities and a history of relapse. Although limited, current research suggests that comorbid psychiatric disorders are more prevalent in MOH than in control headache conditions, and may precede the onset of MOH. There appears to be an elevated risk of family history of substance use disorders in MOH patients, and an increased risk of MOH in patients with diagnosed personality disorders. Current studies suggest a high rate of relapse at 3 to 4 years after drug withdrawal and pharmacological treatment, with most relapse occurring during the first year of treatment. Relapse is a greater problem with analgesics than ergots or triptans. The addition of behavioral treatment to prophylactic medication may significantly reduce the risk of relapse over a period of several years. Clinical recommendations include assessment and modification of psychological factors that may underlie MOH, provision of detailed educational information, and combining behavioral treatment with the current standard of drug withdrawal and use of prophylactic pharmacotherapy.

Abbreviations: 
BPD

borderline personality disorder

CDH

chronic daily headache

CTTH

chronic tension-type headache

DSM-IV

Diagnostic and Statistical Manual of Mental Disorders-4th ed.

ETTH

episodic tension-type headache

FISC

Family Informant Schedule and Criteria

ICHD-II

International Classification of Headache Disorders-2nd ed.

MCMI-III

Millon Clinical Multiaxial Inventory-3rd ed.

MHNI

Michigan Head-Pain and Neurological Institute

MIDAS

Migraine Disability Assessment

MINI

Mini International Neuropsychiatric Interview

MOH

medication-overuse headache

PMOH

probable medication-overuse headache

OR

odds ratio

RR

relative risk

SCID-II

Structured Clinical Interview for DSM-IV Personality Disorders, Version 2.0

SCID-CV

Structured Clinical Interview for DSM-IV Personality Disorders-Clinician Version

To use or not to use, that is the question.

Whether 'tis nobler in the mind to suffer the slings and arrows of outrageous migraine,

Or to take drugs against the sea of pain, and by opposing, end it.

To narcotize, to sedate…

To end the headache and the billion natural synapses to which the flesh is heir

To drug, intramuscular or intravenous, no more?

Perchance massage? Aye, there's the rub!

For in that drugged sleep what dreams may come,

When we have shuffled off this migrainous coil,

Must give us pause…

There's the rebound once again,

That makes calamity of so long life.1

Hamlet's soliloquy (in Shakespeare's original version) expresses his inner debate over suicide (“to be or not to be”) as he confronts significant family stressors and copes with major depression—significant psychiatric conditions, whose comorbidity with chronic headache disorders is discussed throughout this supplement. Here, a migrainous Hamlet grapples with a significant clinical conundrum confronted by many if not most of those who suffer from frequent and severe headaches. Although medication overuse may not technically be considered a psychiatric comorbidity, it is a biobehavioral problem that compounds treatment of those with chronic headache.2

Naïve headache patients may unwittingly fall into a pattern of inadvertent overuse of certain medications that offer short-term relief at the expense of increasing severity and intractability of headache. The naïve patient is confronted with an apparent paradox, that over reliance on drugs that relieve pain—ranging from simple over-the-counter analgesics to prescription opioids—can render the brain more excitable and prone to inflammation, enhance sensory input, lower pain thresholds, and lead to a prolonged state of hyperalgesia and increased sensitization to pain.3–8 Once this concept is understood, such patients often successfully stop their analgesics, triptans, ergots, or barbiturates, and experience relief after a reasonable period of time.

However, as expressed in this revised version of Hamlet's poetic monologue, those with chronic headache may be well aware of the concept of rebound, and yet struggle with issues of how frequently to use their abortive medications. Some wait too long during a developing headache for the drugs to be of much help. Others preemptively treat at the first warning of an impending headache or headache-related anxiety—a potentially useful strategy for relatively infrequent headaches, but a paved path to medication overuse if practiced several times a week on a regular basis. And some feel helpless—that they have no choice but to continue their medication use despite knowledge that it feeds an ongoing cycle of pain.

This article will briefly review diagnostic criteria for medication-overuse headache (MOH); epidemiological research on the prevalence of MOH, including the relationship between frequency of analgesic overuse and headache chronicity; psychological issues and psychiatric comorbidity associated with MOH; relapse rates; and nonpharmacological aspects of developing an effective treatment program and minimizing relapse for this condition.

DIAGNOSTIC CRITERIA FOR MOH

MOH refers to a headache pattern attributed to the overly frequent use of abortive medications that may only remit when the offending medications are withdrawn. According to the most recent revision of the International Classification of Headache Disorders-2nd ed. (ICHD-II), the diagnosis of MOH is based on the following criteria:9

  • A. Headache present for 15 or more days per month fulfilling criteria C and D.
  • B. Regular overuse for >3 months of 1 or more drugs that can be taken for acute and/or symptomatic treatment of headache (10 or more days per month for ergotamines, triptans, opioids, or combination analgesics; 15 or more days per month for simple analgesics).
  • C. Headache has developed or markedly worsened during medication overuse.
  • D. Headache resolves or reverts to its previous pattern within 2 months after discontinuation of overused medication.

“Combination analgesics” include drug compounds that may include butalbital, which can pose significant medication overuse problems with high recidivism.10–12 It is of interest to note that butalbital has been banned in Europe. MOH replaces the previously used terms of rebound headache, drug-induced headache, and medication-misuse headache, although rebound remains a familiar term to American practitioners in particular.

Probable medication-overuse headache is technically the diagnosis that should be applied when medication overuse is suspected (ie, not yet demonstrated) to be a causal factor in the headache pattern.9 As noted above, the current ICHD-II criteria only allow the diagnosis to be confirmed if the headache returns to its previous episodic pattern within 2 months of drug cessation. While this brief article is not the place to debate the merits of a diagnosis that can only be confirmed following headache remission, it should be noted that improvement following drug cessation may take between 3 and 6 months or more, based on clinical studies and case observations.3,13,14 Although not conclusively demonstrated, some researchers suspect that chronic opioid use may lead to long-term neuroplastic changes in the brain that would not necessarily be expected to restructure within 2 months.8,15–17 Also, medication overuse may impede the efficacy of most pharmacological and behavioral prophylactic regimens, making it difficult to determine if given prophylactic regimens are effective until after the offending medications are withdrawn. Resolution of the headache may then depend on finding new preventative treatment regimens in combination with drug cessation, which realistically may take longer than 2 months.

For the remainder of this article, MOH will be used as the generic term to describe both definite and probable MOH, as it is the more commonly used term in published studies and facilitates discussion of improvement rates, which would always be 100% at 2 months if the ICHD-II MOH criteria were strictly followed.

EPIDEMIOLOGICAL RESEARCH ON ANALGESIC USE AND HEADACHE CHRONICITY

Epidemiological research (the Head-HUNT study of >51,000 participants reported by Zwart et al) found the prevalence of daily or near daily use of analgesic medication for 3 months or more in those with chronic daily headache (CDH) (15 or more headache days per month) to be 0.9% in adults18 and 0.5% in adolescents.19 Chronic headache was more than 7 times more likely among those with analgesic overuse than those without (odds ratio [OR]= 7.5). When different chronic pain subgroups were analyzed separately, the association with analgesic overuse was strongest for chronic migraine (OR = 10.3), intermediate for chronic nonmigrainous headache (OR = 6.2), and weakest for chronic neck (OR = 2.6) and chronic low back pain (OR = 3.0). The association became stronger with increasing duration of analgesic use for all groups, and was most evident among those with headache, especially those with migraine.18

In a longitudinal 11-year follow-up of over 30,000 people from this study, individuals who reported use of analgesics daily or weekly at baseline had significantly elevated risk for chronic pain at follow-up.20 The relative risk (RR) was greatest for chronic migraine (RR = 13.3), followed by chronic nonmigrainous headache (RR = 6.2). RR was significantly lower for chronic neck pain (RR = 2.2), and chronic low back pain (RR = 2.3). It is possible that reliance on analgesics plays an etiological role in the transformation and maintenance of chronic headache conditions in particular, although it remains possible that those who relied on analgesics had more severe or disabling headaches at baseline. In either case, these observations underscore the potential importance of a careful examination of analgesic use patterns in the treatment of those with headache in particular.

PSYCHOLOGICAL ISSUES AND PSYCHIATRIC COMORBIDITY IN MOH

Psychological issues that can contribute to the overuse of medication include the following:

  • • the belief that drugs are the only solution to the headache problem;
  • • anticipatory fear of pain—pain panic or cephalagiaphobia;
  • • difficulty tolerating discomfort;
  • • sedation seeking—soporophilia;
  • • outside pressures—need to function;
  • • Axis I clinical syndromes; and
  • • Axis II personality disorders.

Our group has found it helpful to distinguish between Simple MOH (Type 1) and Complex MOH (Type II).2 While these types are not currently operationalized in detail, simple MOH is straightforward, relatively short term (eg, 3 months to a year), with relatively modest doses of drugs (eg, 2 triptans per day, or low doses of analgesics at appropriate intervals, such as 1 hydrocodone-containing mediation at 8 hour intervals on a daily basis), with minimal psychiatric contribution as noted above, and no history of relapse after withdrawal. In contrast, complex MOH involves a long-term history (>1 year to multiple years) of daily opioids or combination medications, often with more than 1 doctor prescribing pain medication. These complex patients may present with multiple psychiatric comorbidities including personality disorders, and have a history of relapse.

Although there exists relatively little research on this issue, psychiatric comorbidity may be an important factor in the transformation of episodic headache to MOH. Atasoy et al recently reported a comparison of 58 patients with MOH evolving from episodic migraine, 31 patients with MOH evolving from episodic tension-type headache (ETTH), and 31 patients with chronic tension-type headache without medication overuse (CTTH).21 The probability of a comorbid psychiatric condition was 68% in the MOH with prior ETTH group and 54% in the MOH with prior migraine group, in contrast to 35% in the CTTH group with no medication overuse. Mood disorders were significantly higher in the MOH with prior ETTH group, and mean depression scores were significantly higher in this group than in the CTTH without medication overuse patients. Obsessive-compulsive personality disorder (defined as a dysfunctional, pervasive pattern of preoccupation with orderliness, perfectionism, and mental and interpersonal control, at the expense of flexibility, openness, and efficiency) was the most common personality disorder diagnosis, and statistically more significant in the MOH with prior migraine group than the CTTH group with no medication overuse.

The strength of this study lies in the use of psychiatrists employing comprehensive, validated interview formats to make the psychiatric diagnosis. These instruments included the Structured Clinical Interview for DSM-IV, Clinical Version, for clinical syndromes (SCID-CV) and Version 2 for personality disorders (SCID-II),22–26 as well as the Beck Anxiety27–29 and Beck Depression Inventory.30,31 In contrast, most studies to date of psychiatric comorbidity in headache disorders have examined clinical syndromes only, neglecting the personality disorders, despite their prevalence in the general population and the extent to which they can confound clinical treatment. Also, the use of the nonmedication overusing CTTH group for comparison allows some control over the frequency of headache (rather than medication overuse per se) as a factor contributing to psychiatric comorbidity. However, the study does not allow a determination of what came first, MOH or psychiatric condition, and represents a clinic-based rather than population-based sample that may not generalize to those with headache as a whole.

Radat and colleagues performed a comparative study of psychiatric comorbidity in 41 patients with episodic migraine with 41 patients with MOH evolving from prior migraine.32 Unlike the Atasoy study described above, this research design does not control for the frequency of headaches as a factor in psychiatric comorbidity. However, psychiatric diagnoses were assigned according to criteria established by the American Psychiatric Association in the Diagnostic and Statistical Manual of Mental Disorders-4th ed. (DSM-IV),33 using the Mini International Neuropsychiatric Interview (MINI).34 The MINI permits determination of age of onset for psychiatric disorders, so that the investigators could make a retrospective determination of what came first—MOH or the onset of the psychiatric condition. Family history of psychiatric disorder was investigated with the Family Interview Schedule and Criteria,35,36 a standardized family history diagnostic interview that has been used to study family history of substance dependence and abuse, with modifications to gather additional information on the family history of migraine and MOH.

Overall, the presence of a comorbid psychiatric disorder was significantly more prevalent in the MOH group. Also, the onset of the psychiatric disorder was significantly more likely to precede the onset of MOH. These findings appear robust, and are depicted in Tables 1 and 2 (modified from Radat).32

Table 1.—.  Odds Ratios for Presence of Comorbid Psychiatric Disorder in MOH and Migraine Without Medication Overuse32
DisorderMOH (%)Migraine (%)Odds Ratio
  1. MOH = medication-overuse headache.

All mood disorders85514.5
Major depressive episode39 221.8 
All anxiety disorders83543.5
Panic disorder24 212.1 
Generalized anxiety disorder42106 
Social phobia34124.3
All substance-related disorders44157.6
Table 2.—.  Chronology of Onset of Psychiatric Disorders in Relation to Chronicization of Headaches in MOH Patients—ie, Whether Disorder Precedes or Follows CDH32
DisorderPrecedes CDH (%)Follows CDH (%)
First major depressive episode 7624
Panic disorder 7921
Generalized anxiety disorder 8020
Social phobia100 0
Substance-related disorders 8911

Family history data uncovered a significantly elevated risk of mood disorders (OR = 1.7) and substance use disorders (alcohol or illicit drugs, OR = 2.8) in the families of MOH patients. They raise the question of whether MOH may result in part from broad family based vulnerability to addiction that increases the risk of acute migraine drugs and analgesic abuse, and take care to underscore that their use of the term “addiction” does not refer to antisocial behavior but rather drug-dependent behavior characterized by a lack of control and by its expression despite the patient's awareness of its negative consequences.

Radat et al appropriately cite their conclusions as preliminary, acknowledging that the headache clinic sample may limit the external validity of their findings, as well as the limitations of retrospective data. Nevertheless, these results are intriguing and should lead to further research in this area.

At Michigan Head-Pain and Neurological Institute (MHNI), we recently reported initial treatment outcome for our comprehensive, multidisciplinary inpatient unit, including the presence of MOH and psychiatric comorbidity as outcome predictors. Based on a retrospective chart review of 284 consecutive inpatient admissions over a 6 month period, 158 (59%) of the 267 program completers had MOH based on prescription medication overuse, with almost half (48%) using daily opioids.37 The diagnosis of a comorbid psychiatric disorder was reached by consensus of experienced clinical psychologists and program physicians using DSM-IV criteria. Data for the diagnosis were based on clinical interview, with psychometric testing in most cases using the Millon Multiaxial Clinical Inventory-3rd ed. (MCMI-III),38–40 which is specifically oriented toward the assessment of personality disorders along with mood and anxiety disorders.

All patients met criteria for at least 1 Axis I psychiatric disorder, most commonly stress-related response affecting a medical condition (82%), depressive disorders (70%), and anxiety disorders (49%). Seventy (26%) of the program completers met criteria for an Axis II personality disorder, with 42 (15%) showing evidence of significant cluster B traits (borderline, histrionic narcissistic, or antisocial). The percentage of patients with personality disorders in this tertiary inpatient sample appears higher than expected, based on recent epidemiological studies of representative community samples in the United States, where the rate appears to range between 13.4%41 and 14.8%.42

Although there was no relationship between Axis I disorders and MOH, those with personality disorders were significantly more likely to be opioid dependent than the nonpersonality disorder patients (62% vs 38%) and less likely to attain moderate-significant improvement in head pain (based on consensus judgment of medical staff and patient headache data) by the end of the 14-day hospitalization (68% vs 81%). There are several possible interpretations of these findings. Patients with Axis II disorders may be more prone to developing substance dependence patterns. It also seems likely that those with headache with personality disorders may be more demanding of their physicians, and convince them to prescribe more opioids. It is also possible that there are underlying pathophysiological factors leading to increased intractability in the combination of a personality disorder with MOH. Personality disorders were statistically associated with a less positive outcome by the conclusion of hospitalization, regardless of the presence of MOH. In fact, those patients with personality disorders without medication overuse had the lowest likelihood of a positive outcome of any diagnostic condition, although 53% of this group still reached moderate-significant improvement. Longer-term outcome data on these hospitalized patients are not currently available.

Despite the clinical problems posed by comorbid personality disorders in patients with chronic migraine and MOH,43 there is very little published research in this area. One recent exception to this dearth of studies was reported by Rothrock and colleagues.44 They compared headache-relevant features and treatment outcome for 50 consecutive patients with migraine and psychiatrically diagnosed borderline personality disorder (BPD) with 50 patients with migraine of all types and no history of BPD, and 50 additional patients with migraine but no BPD who were matched to the BPD group on age, gender, and headache frequency. They found coexisting BPD to be associated with female gender, more pervasive headache, more migraine-related disability, a higher prevalence of MOH, more unscheduled visits for acute migraine treatment, active self-reported depression, and a lower likelihood of responding to pharmacologic therapy intended for headache management. One puzzling finding in this study, however, was that a validated screening measure for identifying BPD (the McLean Screening Instrument for Borderline Personality Disorder45) was insensitive in detecting BPD, which may raise some question about the validity and reliability of the BPD diagnosis in the studied patients.

RELAPSE RATES IN MOH

In our inpatient sample, all MOH patients who completed hospitalization were successfully withdrawn from their triptans, ergots, opioids, or other analgesics. The therapeutic challenge is to maintain appropriate limits on abortive or analgesic medication after discharge, and maintain control over the headache. For some patients, drug withdrawal can be accomplished on an outpatient basis. However, Bigal and others found, in a retrospective chart review of 456 MOH patients with transformed migraine, that 30% either failed to discontinue the drugs of concern, or returned to medication overuse within 1 year of initiating treatment.12

Relapse after drug withdrawal was studied prospectively in 96 MOH patients by Katsarava et al.46 They defined relapse as frequent use of any pain medication, more than 15 days per month, over more than 3 months. About 3 out of 10 patients (31%) relapsed within the first 6 months, with relapse rates increasing to 41% at 1 year and 45% at 4 years. The 4-year relapse rate was significantly higher for analgesics (71%) than ergots (27%) or triptans (21%). Among diagnostic groups, 4-year relapse rates were significantly higher for tension-type headache (91%) or combined tension-type headache and migraine (70%) than what was observed for migraine only patients (32%).

EFFECTIVE MANAGEMENT OF MOH AND PREVENTION OF RELAPSE

At MHNI, we have utilized a 3-step approach to the management of MOH:47

  • 1Identify the problem, including the individual patient's psychological issues specific to MOH.
  • 2Provide specific, detailed education about MOH, the expected time for improvement after withdrawal, and the risk of relapse.
  • 3Combine medical management (drug withdrawal and prophylactic pharmacotherapy) with teaching and underscoring the importance of cognitive-behavior (nondrug) coping skills.

Identify the Medication Overuse Problem.— In his article on confronting behavioral disturbances headache patients in this supplement, Saper identifies several strategies for identifying patterns of medication overuse.48 The obvious first step is to ask the patient directly about the frequency and quantity of abortive or analgesic use. However, complex MOH patients are not always forthright about their use. In cases of intractable CDH, the possibility of a more significant MOH pattern than first appears to be the case must be considered. In individual interviews, it is helpful to inquire about use patterns in a nonjudgmental tone of voice, and rephrase the same question two or three times. As the patient develops trust over the course of the interview, he or she may choose to reveal information that was initially minimized or obscured. However, the individual interview may be insufficient to fully assess use patterns. In intractable headache cases, it is often quite helpful to interview significant others and family members, who may identify problems in drug use that the patient minimizes. Urine drug screens (routine at the time of our initial evaluations) can help identify drugs that the patient is using but not reporting. It can also be useful in complex cases to contact the patient's primary doctor as well as other prescribing physicians. Records from insurance companies that pay for medication can also help identify patterns of multi-sourcing.

Some states now have registry systems that can provide doctors with information on all filled prescriptions for controlled substances over a defined period of time. For example, Michigan physicians can request a report from the Michigan Automated Prescription System, which documents all prescriptions for controlled substances that were filled by the patient over the past year. Again, headache intractability can serve as the impetus for ordering these types of reports. At MHNI we continue to be surprised on many occasions when patients who we would be least likely to suspect of a medication overuse pattern are revealed to be multi-sourcing—in most cases due to their fears of the headache and felt need for pain control. These reports can then be reviewed with the patient and other prescribing physicians. To put it simply, complex cases of MOH cannot be effectively treated if the treater is unaware that a medication overuse problem exists.

Educate the Patient.— The patient needs to “buy into” the concept of MOH. For example, in our inpatient program group education provides ample time for discussion and interaction on recent science related to analgesic and opioid-induced hypersensitivity to pain.6–8,15 The educational message emphasizes that “rebound” can occur with use patterns of 3 days per week or more—daily use is not required—and that the patient needs to maintain appropriate expectations, since significant headache improvement following drug withdrawal may require as long as 12 weeks13 or in some cases over 6 months.14 Although empirical research on what patient factors are predictive of the recovery interval, patients are informed that the time course for significant improvement may depend on the frequency and amount of drug taken, the duration of overuse, as well as individual differences in susceptibility to MOH. Patient information emphasizes that drug withdrawal from the overused medication may not be sufficient in itself for improvement, but is necessary in order to open the door for effective prophylaxis, which may take some time to evaluate. The primary intent is to engage patients in a collaborative treatment model based on mutual understanding, rather than simply to impose what be perceived as authoritarian limits on analgesics that the patient may rebel against based on his or her own belief system.

Rothrock et al recently published a 6-month follow-up study of 100 consecutive headache patients, of which half were randomly assigned to “headache school”—3 90-minute standardized classes of didactic instruction regarding migraine pathogenesis and management, taught by lay migraineurs who had all received intensive training in the topics covered.49 A neurologist managed all patients with prescription abortive medication, and prophylactic medication when appropriate, and botulinum toxin.

The results clearly favored those who attended headache school. These patients had significantly lower scores on the primary outcome measure, the Migraine Disability Assessment questionnaire (MIDAS) at 6 months. The headache school patients used significantly less abortive therapy, had less analgesic overuse, reported less headache days per month, experienced less headache-related functional incapacitation, and made fewer headache-related calls to the clinic or unscheduled visits. These results are preliminary, and the authors acknowledged some limitations including significantly lower mean MIDAS scores at baseline in the headache school group. As with many educational interventions, the treating neurologist was not blinded to whether patients had been enrolled in headache school. Nevertheless, this study does underscore the potential value of detailed educational information as part of the treatment program.

The complex MOH patient with opioid dependence may hold dysfunctional beliefs, such as “more opioids are necessary to make the pain go away—this is the only drug that can help me” or “if my analgesics are losing effectiveness that must mean the pain is getting worse and overpowering my drugs.” Such beliefs can lead to increasing pain-related anxiety, and need to be assessed and addressed. The goal is to help the patient develop more accurate beliefs based on current scientific understanding, such as “sustained opioids can reduce pain thresholds and actually increase pain,” and “analgesic tolerance is a red flag for opioid-induced hyperalgesia” as alternatives to “the pain is getting worse on its own and I need more pain relievers.”

In regard to opioids, it is also helpful to alert patients to the few published studies on long-term scheduled opioids for pain control, which highlight their limitations not only in headache control,17,50 but in other types of pain as well.51 For example, in a recent metaanalysis of randomized, double-blind, placebo-controlled studies of scheduled opioid efficacy in noncancer pain performed by Kalso and colleagues, the mean pain relief with opioid therapy was about 30%, and only a minority of patients appeared to benefit from long-term treatment.51 While there may be a very limited role for scheduled opioids in the treatment of some refractory patients, we believe that MOH should first be ruled out with a sufficient washout period and adequate trials of prophylactic medications, behavioral intervention, and procedural interventions in appropriate patients.17,50

Teach and Reinforce Nondrug Coping Skills.— MOH patients who only rely on withdrawal from the offending drugs with follow-up pharmacological management may be at significantly higher risk of relapse and failure to sustain improvement. While there is little research on this issue, Grazzi, Andrasik, and others reported long-term results that point to the potential value of adding nondrug coping skills to the treatment program.52 Sixty-one consecutive patients with transformed migraine and analgesic overuse were evaluated and followed over 3 years. All patients were initially hospitalized, withdrawn from analgesics, received intravenous and intramuscular drug therapies, and started on prophylactic pharmacotherapy. Nineteen received additional biofeedback-assisted relaxation therapy. Assignment to biofeedback was described as quasi-random, in the sense that patients who lived at great distance from the institute could decline biofeedback treatment if they wished.

At the 3-year follow-up, 42% of the 38 pharmacotherapy-only patients who were recontacted and responded to the follow-up request had relapsed, in the sense that their frequency of abortive use again met criteria for MOH. This percentage is remarkably close to the 45% relapse rate reported at 4 years in the Katsarava study review above. In contrast, only 2 of the 16 (12.5%) of the responding combined treatment patients had relapsed, a significant difference in favor of the combined treatment. The combined treatment patients also showed a trend toward fewer headache days at the 3-year mark, with a mean of 11.2 versus 18.1 for the drugs-only group. Most dramatic was the significant difference in the mean number of analgesics consumed monthly: 4.9 for the combined treatment, versus 20.1 for pharmacotherapy only. These findings imply that the biofeedback patients were choosing to take analgesics on only a minority of their headache days, where the drugs-only group appeared to be medicating most of their headaches. Despite the relatively small n, and admitted limitations in the random assignment process, this study clearly points to the potential value of adding nondrug treatment to pharmacology for MOH patients, given the high-risk of relapse over time.

Behavioral treatments do not necessarily provide the immediacy of headache reduction that may come from a triptan, ergot, or analgesic. However, patients undergoing biofeedback and behavioral therapy often learn to restructure their cognitive approach to pain, in essence learning how to tolerate discomfort, reduce pain-related emotional distress, stop the overly frequent pharmacological preemptive treatment of an impending headache, and reduce limbic escalation of the pain experience. If reinforced and maintained over time, these learned behaviors can help reduce the likelihood of overusing pain medication and MOH relapse.

COMMENTS

The literature on MOH is evolving from an initial identification of the extent to which overuse of analgesics and abortive medications can perpetuate headache and the necessity of drug withdrawal, to a broader exploration of the extent to which psychological issues, comorbid psychiatric disorders, and family history may compound the medication overuse problem. Relapse rates appear unacceptably high. For complex cases, optimal treatment programs may need to assess and modify psychological factors that may underlie medication overuse, provide detailed educational information about MOH and the hyperalgesic effects of opioids in particular, highlight the risk of relapse, develop relapse prevention strategies, and combine cognitive behavioral treatment with the current standard of drug withdrawal and use of prophylactic pharmacotherapy.

Conflict of Interest:  None

Ancillary