Psychiatric Comorbidity in the Evolution From Migraine to Medication Overuse Headache

Abstract

We set out to study the role of psychiatric comorbidity in the evolution of migraine to medication overuse headache (MOH) by a comparative study of 41 migraineurs (MIG) and 41 patients suffering from MOH deriving from migraine. There was an excess risk of suffering from mood disorders [odds ratio (OR) = 4.5, 95% confidence interval (CI) 1.5, 13.5], anxiety (OR = 5, 95% CI 1.2, 10.7) and disorders associated with the use of psychoactive substances other than analgesics (OR = 7.6, 95% CI 2.2, 26.0) in MOH compared with MIG. Retrospective study of the order of occurrence of disorders showed that in the MOH group, psychiatric disorders occurred significantly more often before the transformation from migraine into MOH than after. There was no crossed-family transmission between MOH and psychiatric disorders, except for substance-related disorders. MOH patients have a greater risk of suffering from anxiety and depression, and these disorders may be a risk factor for the evolution of migraine into MOH. Moreover, MOH patients have a greater risk of suffering from substance-related disorders than MIG sufferers. This could be due to the fact that MOH is part of the spectrum of addictive disorders.

Introduction

Ten to 15 per cent of the general population suffer from migraine (1) and a certain percentage of these individuals will develop chronic daily headaches with medication overuse. This syndrome, referred to by the International Headache Society 2nd edition (2) as medication overuse headache (MOH), affects 4% of migraine sufferers overall and approximately 30% of patients treated in specialty clinics (3, 4).

Two main factors are assumed to contribute to the expression of this syndrome: medication overuse and psychiatric comorbidity (5). However, no specific evidence exists concerning their specific roles. The fact that withdrawal can suppress the chronicity of headaches appears to underscore the role of medication overuse, but the literature concerning psychiatric comorbidity is far more scarce. MOH has been shown to be more strongly associated with anxiety and depressive disorders than is migraine without MOH (5–7). Thus, the comparison between migraine (MIG) and MOH should be informative concerning the identification of potentially distinct psychiatric risk factors. The possibility that such comorbidity plays a causal role in the transformation of migraine into MOH (5) must be examined through chronological studies of the onset of both disorders. Furthermore, although both anxiety and depression are associated with this disorder, MOH may be more strongly predicted by other unexamined forms of comorbidity, and notably by substance use disorders.

A final issue concerns the lack of information concerning the nature of psychiatric risk factors themselves, and particularly the importance of distinguishing between individual and familial based origins of risk (8). The present study examines these issues based on individual diagnostic and family history interviews in MOH patients compared with those with MIG.

Methods

Eighty-two patients seeking treatment at a neurology service of a large university hospital were consecutively included in the study sample if they met the International Headache Society (2) criteria for migraine (code 1.1 and 1.2) or MOH (code 8.2, with the exception of criteria D ‘headache reverts to its previous pattern within two months after medication withdrawal’ due to the cross-sectional design of this study) and provided written informed consent. The sample was then divided into two groups: those suffering currently from migraine without a history of MOH (MIG group, N = 41), and those having suffered from migraine and currently suffering from MOH (N = 41). The age of onset of chronic daily headache (more than 15 days per month for at least 3 months) as well as the age of onset of medication overuse (drug intake more than 18 days per month for at least 3 months) was assessed by retrospective recall questions. Psychiatric comorbidity was assessed prior to medication withdrawal according to the DSM-IV criteria (9), using the MINI diagnostic interview (10) that permits determination of age of onset for psychiatric disorders. To assess tobacco dependence, questions concerned whether the patient smoked at least 20 cigarettes a day (currently or in the past). In the case of a positive response, patients were asked if it had lasted for at least 1 year, and if so whether they tried to reduce their consumption or to stop smoking, or if smoking less or trying to stop resulted in any withdrawal symptoms. Patients responding affirmatively to all questions were considered to be dependent or as having been dependent in the past on nicotine.

Family history of psychiatric disorder was investigated with the FISC (11), a well-validated family history diagnostic interview, modified to acquire information concerning a family history of migraine and MOH. Logistic regression models were used to identify personal or family history variables associated with MOH (adjusted for gender and age).

Results

Concerning demographic differences between groups, MIG patients were more likely than MOH patients to be male (χ² = 6.011, P = 0.01), but no differences were found in age, family status, education or profession. Age of onset of migraine was 16.32 ± 8.30 years for the MIG group, and 16.63 ± 6.47 years for the MOH group (NS). In the MOH group, the age of onset of chronic daily headache was not different from the age of onset of medication overuse (35.9 ± 10.35 vs. 36.7 ± 8.8; NS). The number of days with headache during the last month was significantly higher in the MOH group than in the MIG group (26.7 ± 5.1 vs. 6.5 ± 3.2; t = 21.6, P < 0.000). Acute migraine medications used by MOH patients were: paracetamol (n = 31), caffeine compounds (n = 22), codeine compounds (n = 21), triptans (n = 19), non-steroid anti-inflammatory drugs (NSAIDs) except acetylsalicylic acid (ASA) (n = 12), noramidopyrine (n = 11), ergot compounds (n = 9), and ASA (n = 7). Some patients used a combination of these different medications.

Compared with MIG patients, MOH were more likely to suffer from the aggregate categories of any mood disorder [odds ratio (OR) = 4.5, 95% confidence interval (CI) 1.5, 13.5], and any anxiety disorder (OR = 3.5, 95% CI 1.2, 10.7). Analyses by individual disorder revealed that MOH patients more frequently met criteria for a current major depressive episode (OR = 21.8, 95% CI 2.8, 177.5), current generalized anxiety disorder (OR = 6.0, 95% CI 1.7, 20.8), current panic disorder with or without agoraphobia (OR = 12.1, 95% CI 1.4, 104.4), and current social phobias (OR = 4.3, 95% CI 1.3, 14.5). In addition, MOH was associated with dependence on/abuse of psychoactive substances other than analgesics or acute migraine drugs (OR = 7.6, 95% CI 2.2, 26.0) (see Table 1). Comparing the MIG group with the MOH group, different substances used were: tobacco (n = 2 vs. n = 6), benzodiazepines (1 vs. 7), alcohol (2 vs. 4), illicit drugs (1 vs. 1).

Table 1

Relative risk of suffering from psychiatric disorders for migraine (MIG) compared with medication overuse headache (MOH) (odds ratios and confidence intervals)

	MOH (n, %)	MIG (n, %)	OR	95% CI	P
Major depressive episode	16 (39)	1 (2.4)	21.8	(2.7, 177.5)	0.004
All mood disorder	35 (85.4)	21 (51)	4.5	(1.5, 13.5)	0.007
Generalized anxiety disorder	17 (41.5)	4 (9.8)	6	(1.74, 20.9)	0.004
Panic disorder	10 (24.4)	1 (2.4)	12.1	(1.4, 104.4)	0.02
Social phobia	14 (34.1)	5 (12.2)	4.3	(1.3, 14.5)	0.02
All anxiety disorders	34 (82.9)	22 (53.7)	3.5	(1.2, 10.1)	0.02
All substance-related disorders	18 (43.9)	6 (14.6)	7.6	(2.2, 26)	0.001

In the MOH group, the mean age of occurrence of the various disorders was 14.8 ± 4.6 years for social phobias, 23.6 ± 6.9 years for substance-related disorders, 27.6 ± 12.4 years for generalized anxiety disorders, 28.2 ± 10.7 years for first major depressive episode, 28.9 ± 11.9 for panic disorder, and finally 35.9 ± 10.35 years for chronic headache. The percentage of cases in which psychiatric disorders preceded chronic headache was always significantly higher than the number of cases where it followed chronic daily headache (see Table 2).

Table 2

Chronology of onset of psychiatric disorders in relation to chronicization of headaches (χ² test for comparision of percentages)

	The disorder preceeds onset of chronic daily headache (%)	The disorder follows onset of chronic daily headache (%)	P
Social phobia	100% (n = 17)	0% (n = 0)	–
Substance-related disorders	88.9% (n = 16)	11.1% (n = 2)	0.001
Generalized anxiety disorder	80% (n = 20)	20% (n = 4)	0.007
First major depressive episode	76% (n = 26)	23.5% (n = 8)	0.002
Panic disorder	78.9% (n = 15)	21.1% (n = 4)	0.01

Concerning family history of disorder, the risk of suffering from MOH was greater in the MOH relatives group than in the MIG relatives group (OR = 10.3, 95% CI 1.2, 87.3). In addition, relatives of MOH patients were more likely than MIG relatives to have a mood disorder (OR = 1.7, 95% CI 1.0, 2.7), or substance use disorder (alcohol or illicit drugs) (OR = 2.8, 95% CI 1.0, 7.4). Specific disorder comparisons revealed increased rates in MOH relatives of major depressive episodes (OR = 2.6, 95% CI 1.4, 4.7), and panic disorder (OR = 9.4, 95% CI 1.1, 79.1).

The risk of suffering from a substance-related disorder in the MOH relatives group was greater than in the MIG group even after adjustment for the presence of a substance-related disorder in the patient (OR = 2.7, 95% CI 1.0, 7.4), but a parallel analysis for anxiety or mood disorders in the MOH relatives was not significant. The opposite relationship was nearly significant: the risk of suffering from MOH was greater in the substance-related disorder relative group than in the relatives of patients who were not suffering from substance-related disorder, even after adjustment for the presence of MOH in the patients (OR = 8.48, 95% CI 0.9, 76.9).

Discussion

The present findings confirm the high rates of anxiety and depressive disorders in MOH patients reported by past investigations (6, 13–15). The observed CIs are large for individual disorders due to the small number of patients, but aggregating by broad disorder categories allowed for more precise OR estimates. In a study with a similar design (6), an increased risk of suffering from major depressive disorder, panic disorder and social phobia was also observed. However, the discrepant results concerning generalized anxiety disorder may be related to the use of DSM-IIIR criteria in the former study, and DSM-IV in the present one. The results also demonstrate for the first time a strong comorbidity with substance use disorders.

Concerning the mechanisms of association underlying the different forms of comorbidity, the fact that all psychiatric conditions were more likely to precede than to follow MOH lends support to their conceptualization as risk factors. However, the absence of a clear time sequence between onset of daily headaches and onset of medication overuse does not allow us to determine if psychiatric comorbidity is best seen as a risk factor for the worsening of headache frequency (followed by medication overuse) or rather as a risk factor for medication overuse (followed by worsening of headache frequency). The findings nonetheless underscore the notion that some migraine patients may be vulnerable to developing MOH due either to a personal history of substance use disorders, or to other markers of addiction vulnerability.

Analyses of family history data indicate that substance use disorders are more frequent in the families of MOH patients, whether or not the patient has a personal history of substance abuse (other than analgesic medication). The opposite relationship could also be tenable: MOH may be more frequent in the families of patients suffering from substance-related disorders, whether or not the patient has a personal history of MOH. These results suggest a cotransmission of these two conditions, due to a common underlying vulnerability. For this reason, MOH may result in part from broad family-based vulnerability to addiction that increases the risk of acute migraine drugs and analgesic abuse, as well as of substance use disorders in general. Addiction is a general term which concerns all behaviour characterized by a lack of control and by its expression despite the patient's awareness of its negative consequences. This term does not imply any antisocial behaviour, as often observed in certain forms of drug addiction. Indeed, such behaviour is seldom present in headache patients. The average percentage of MOH patients exhibiting an addictive behaviour, rather than misuse or overuse behaviour secondary to a worsening of their headache, remains unknown. While it has been suggested that such patients are not a majority (16), clarification of this issue should be a priority for future investigations.

The conclusions of this investigation are preliminary and are limited by the small number of patients as well as by the use of indirect, albeit well-validated, measures of family psychopathology (mood disorders, anxiety disorders, alcohol and illicit drugs-related disorders). Moreover, the assessment of family history specifically of MIG and MOH was not previously validated. Certain characteristics of the sample, such as its recruitment from a university specialty clinic, may also limit generalizability to other migraine populations. Concerning other methodological issues, the high psychiatric comorbidity rates in both the MOH group and the MIG group may be attributable to the fact that the study sample was recruited from a headache clinic. The retrospective assessment of the age of onset of disorders can also be considered as a limitation, although the focus on substance use disorder comorbidity is in itself novel. The inclusion of migraine without MOH as a comparison disorder and family history assessments underscore the salient individual and family risk factors for MOH patients. Additional research is needed to further understand psychiatric comorbidity in MOH, but the development of close consultation between neurologists and psychiatrists already appears to be clearly warranted in the treatment of this disorder. It is of considerable importance that signs of addiction as well as anxiety and depressive symptomatology be monitored in MOH patients when medication withdrawal is proposed.