From the Division of Toxicology and Clinical Pharmacology, Headache Centre, University Centre for Adaptive Disorders and Headache, Section Modena II, University of Modena and Reggio Emilia, Modena, Italy (Drs. Ferrari, Leone, Bertolini, Coccia, Pinetti, and Sternieri); Headache Unit, IRCCS C, Mondino Institute of Neurology Foundation, University Centre for Adaptive Disorders and Headache, Section Pavia, University of Pavia, Pavia, Italy (Dr. Sances); and Department of Biomedical Sciences, Section of Pharmacology, University of Modena and Reggio Emilia, Modena, Italy (Drs. Vergoni and Ottani).
Address all correspondence to Dr. Anna Ferrari, Division of Toxicology and Clinical Pharmacology, Headache Centre, University of Modena and Reggio Emilia, Largo del Pozzo, 71. 41100 Modena, Italy.
Objective.—To quantify and characterize the similarities and the differences between chronic migraine (CM) patients with medication overuse and episodic migraine (EM) patients with only occasional analgesic use.
Background.—Population-level epidemiology, characteristics, mechanisms of chronic daily headache, and medication-overuse headache have been widely studied but patient characteristics have received less attention.
Methods.—We compared sociodemographic data, family history, physiological and medical history, health services utilized, drugs taken/prescribed, and outcome of 2 groups of subjects: 150 patients, suffering from CM, complicated by probable medication-overuse headache (CM group), consecutively admitted during 2005 to the inpatients' ward of the Headache Centre of the University Hospital of Modena and Reggio Emilia, Italy, to undergo withdrawal from their overused medications; 100 patients suffering from EM, uncomplicated by medication overuse (EM group), consecutively referred to the outpatients' ward of the Headache Centre during November and December 2005.
Results.—All sociodemographic characteristics were significantly different between the 2 groups. As a whole, the CM group began to suffer from migraine earlier than the EM group. Drug and/or alcohol abuse was significantly higher among first-degree relatives of CM (19%) than of EM (6%) patients. The most frequent comorbid disorders were psychiatric (67%) and gastrointestinal diseases (43%) in the CM group, and allergies in the EM group (31%). Seventy percent of CM patients and 42% of EM patients were taking daily at least another drug, besides those for headache treatment. Most overused medications in the CM group were triptans (43%); the EM group used above all single NSAIDs (56%). At 3-month follow-up, prophylactic treatments reduced, by at least 50%, the frequency of headache in about three-fourths of patients of both the groups; however, headache remained significantly more frequent in the CM than in EM group: only a minority (15%) of CM patients reverted to a headache frequency comparable to that of the EM group.
Conclusions.—CM patients present more multiple comorbid disorders, polypharmacy, and social impediments than EM patients. These associated conditions complicate CM clinical management. Even after withdrawal from medication overuse, CM could not be completely reverted by current prophylactic treatments.
Population-level epidemiology, characteristics, mechanisms of chronic daily headache, and medication-overuse headache have been studied widely,1–3 but patient characteristics have received less attention.
The objective of this study was to quantify and characterize the differences between consecutive outpatients with episodic migraine (EM), uncomplicated by medication overuse; and consecutive inpatients whose initially similar episodic headaches had evolved into chronic migraine (CM), complicated by probable medication-overuse headache.
PATIENTS AND METHODS
Subjects.— We compared 2 groups of subjects.
Chronic Migraine Group.— All patients, males and females of all age, suffering from CM complicated by probable medication-overuse headache, according to the revised ICHD-II criteria,4,5 consecutively admitted during 2005 to the inpatients' ward of the Headache Centre of the University Hospital of Modena and Reggio Emilia, Italy, to undergo withdrawal from their overused medications.
In Italy, inpatient treatment is funded fully by the National Health Service, independent of their socioeconomic status. Only subjects with migraine both with and/or without aura at the onset were included. According to these criteria, out of 183 subjects admitted, 150 (139 [93%] with the diagnosis of migraine without aura at the onset and 11 [7%] with the diagnosis of migraine with and without aura) were included in the study, and 33 patients (23 chronic tension-type headaches and 10 episodic tension-type headaches at the onset) were excluded. During 7 to 10 days of hospitalization, patients received standardized treatment: abrupt withdrawal of the overused medications; intravenous infusion of delorazepam (daily doses: 0.2 to 0.5 mg) in 250 mL normal saline, as baseline therapy; intravenous infusion of ketoprofen (daily doses: 50 to 100 mg) in 250 mL normal saline for the first 2 to 3 days and then only as needed, for the treatment of rebound headache; sumatriptan 6 mg subcutaneously for the treatment of severe migraine attacks; intravenous alizapride (dose: 50 to 100 mg) to control nausea and vomiting, as needed. A prophylactic treatment was usually started after 5 to 7 days of stay.
Episodic Migraine Group.— All patients suffering from migraine with and/or without aura according to the ICHD-II criteria,6 only occasionally using analgesics, consecutively referred by their general practitioner to the outpatients' ward of the Headache Centre (for the first time) during November and December 2005. Patients who suffered from migraine with aura alone were excluded. According to these criteria, out of 144 patients referred, 100 (90 [90%] with diagnosis of migraine without aura and 10 [10%] with diagnosis of migraine with and without aura) were selected, and 44 excluded (8 migraines with aura, 4 cluster headaches, 9 tension-type headaches, 7 chronic tension-type headaches, 16 secondary headaches or nonheadache disorders).
Procedures.— We collected sociodemographic data, family history, physiological and medical history, health services utilized, drugs taken/prescribed, and outcome, from each patient. Written informed consent was obtained from each subject, following an exhaustive description of the study's objectives and procedures. The study was approved by the Ethical Committee of Modena. Medication overuse and headache/migraine were determined frequently by the patients' diaries kept during the months prior to the visit at the Headache Centre. Study physicians registered drugs prescribed at discharge and the outcome at the follow-up visit, after 3 months. Prophylactic treatment was considered effective if it had reduced, by at least 50%, the frequency of headache when compared with the 4-week prior frequency. All patients' data were recorded by means of a specially devised medical chart, developed in collaboration with the Headache Centre of Pavia, University Hospital, Italy.
Data Analysis.— All data were checked by a trained physician (who contacted authors in order to clarify ambiguous data) and inserted into a specially prepared database. All statistical analyses were performed with the help of the SPSS 8.0 (SPSS Inc., Chicago, IL, USA) statistical software. A complete descriptive analysis of the patients' sampled parameters was carried out, as well as a comparison between CM and EM groups. The frequency distributions were compared by the chi-square test. When appropriate, Student's t-test for unpaired data was performed. A significance level of at least <.05 was chosen for all tests.7
All sociodemographic characteristics (Table 1) were significantly different between groups: females were significantly more in the CM than in EM group. Overall, EM patients had a higher education level (92% had attended secondary school) than CM patients (nearly half had not attended secondary school). The majority (88%) of CM patients were or had been married; instead, more than one-third of EM patients had never been married. The CM patient group featured significantly more retired persons and fewer workers than the EM patient group.
Table 1.—. Demographic Data
Chronicn Migraine: n = 150 (%)
Episodic Migraine: n = 100 (%)
*P < .001 versus episodic migraine (Student's t-test); **P < .001 versus episodic migraine (chi-square test).
Mean age ± SD (years)
49 ± 13.4 *
36 ± 10.2
(21 to 75)
(22 to 61)
Distribution by age (years)
21 to 31
32 to 41
42 to 51
52 to 61
62 to 75
Half of the CM patients (Table 2) began to suffer from migraine before 20 years of age, while the majority of the EM patients began later, between 20 and 40 years of age. Considering female patients, migraine remitted during pregnancy significantly more in the EM than in CM group. The majority of patients had 3 to 6 migraine attacks per month in the EM group and daily headache in the CM group.
Table 2.—. Headache Characteristics
Chronic Migraine: n = 150 (%)
Episodic Migraine: n = 100 (%)
*P < .05 and **P < .001 chronic migraine versus episodic migraine (chi-square test).
†Only female subjects were considered.
‡Females who had pregnancy were 96 of the chronic migraine group and 31 of the episodic migraine group.
Age at onset (years)*
10 to 20
21 to 40
Onset at menarche†
Remission during pregnancy‡
1 to 2 per month
3 to 5 per month
6 per month
>15 days per month
Fifty-one percent (n = 76) of the CM patients had suffered from CM and overused medication for more than 5 years and the remaining 49% (n = 74) from 1 to 5 years. All patients of this group started to take medication more frequently due to the worsening of their migraine concurring with particular events (familial stress 35%; working stress 8%; menopause 10%; hypertension 5%; surgical operation 3%; trauma 2%; others 37%). Nearly 60% (n = 88) of these patients had previously tried to discontinue drug overuse without success, and 44% (n = 66) of them had tried at least twice.
Migraine was the most recurrent disorder (Fig. 1) in the history of first-degree relatives in both the groups, either CM (n = 128; 85%) or EM (n = 68; 68%) group. Overall disorders, but depression and cerebrovascular diseases, were significantly more frequent among CM than EM relatives. Notably, drug and/or alcohol abuse was significantly higher among first-degree relatives of CM (n = 29; 19%) than of EM (n = 6; 6%) patients.
The majority of patients (Fig. 2), without differences between the 2 groups, were nonsmokers. Coffee and/or tea and alcoholic drinks were consumed significantly more by EM than CM patients. Bad sleep (difficulty in falling asleep, mid-sleep awakenings, early morning awakenings, and nonrestorative sleep) and constipation occurred significantly more in CM (in over half of the patients) than in EM patients. More female patients in the CM than in EM group had had at least 1 pregnancy, whereas more females in the EM group than in CM group were taking oral contraceptives.
Comorbid disorders (Fig. 3), except allergies and thyroid disorders, were significantly less common in the medical history of EM rather than CM patients. Among CM patients the most frequent disorders were psychiatric (n = 101; 67%), above all anxiety (n = 45; 44%) and depressive disorders (n = 35; 35%), followed by gastrointestinal diseases (n = 65; 43%). Allergy was the most frequent complaint in the EM group (n = 31; 31%).
Seventy percent (n = 105) of the CM patients (Fig. 4) and 42% (n = 42) of the EM patients were taking daily at least 1 other drug, besides those for headache treatment. CM patients were most commonly taking sedative-hypnotics (n = 45; 30%) and antihypertensive drugs (n = 39; 26%). Instead, EM patients were above all taking anti-ulcer agents (n = 15; 15%). All kinds of drugs, except for anti-ulcer agents, laxatives, and thyroid hormones, had been taken more frequently by CM than by EM patients.
CT scan (Fig. 5) was the diagnostic test most often carried out by patients, either in the CM (n = 86; 57%) or in the EM (n = 33; 33%) group. Significantly more patients in the CM group than in the EM group had been investigated by CT scan and MRI. Almost all patients in the EM group (n = 97; 97%), but only three-quarters in the CM group (n = 113; 76%), had consulted their family doctor for headache at least once during the previous 12 months. Patients of the CM group had sought care in emergency departments and consulted other headache centers significantly more often than patients in the EM group. No patient in the EM group had ever been hospitalized for headache. Less than 30% of patients in both the groups had consulted different specialists: EM patients had consulted neurologists (n = 26; 26%), whereas CM patients consulted otorhinolaryngologists (n = 33; 22%).
On arrival at the Headache Centre (Fig. 6), all patients in the 2 groups were using/overusing at least one medication for acute treatment: EM patients tended to use single NSAIDs (n = 56; 56%), while CM patients were much more likely to overuse triptans (n = 65; 43%). Combinations of analgesics with caffeine were taken more by CM (n = 53; 35%) than by EM (n = 7; 7%) patients. No patient used/overused combinations of analgesics with barbiturates or opioids. The majority (n = 85; 57%) of the CM patients, but only a minority (n = 20; 20%) of the EM patients, was taking a prophylactic headache treatment (X2= 31.626, P < .001). Prevailing treatments were antidepressant agents (n = 57; 38%) in the CM group, and flunarizine (n = 11; 11%) in the EM group.
Upon discharge from the Headache Centre (Fig. 7), an acute medication was prescribed to each patient: triptans were the most widely prescribed drugs either to CM (n = 93; 62%) or, more likely, to EM (n = 77; 77%) patients. After withdrawal of the overused medication, a prophylactic headache treatment was prescribed for at least 3 months to each patient in the CM group, antidepressants being the most prescribed drugs (n = 89; 59%). Prophylactic headache treatment was also prescribed to 86 out of 100 patients in the EM group, flunarizine being the most prescribed drug (n = 50; 58%).
No patient in either group was lost at the 3-month follow-up. Considering only those patients who had received a prophylactic drug, about three-quarters of patients in both the groups (Table 3) had the frequency of headache reduced by at least 50% with respect to prior frequency (and, in the case of the CM group, medication overuse was interrupted). Notwithstanding this improvement, headache remained significantly more frequent in the CM rather than the EM group; only a minority (n = 17; 15%) of the CM patients reverted to a headache frequency comparable to that of the EM patients (maximum 6 attacks per month).
Table 3.—. Outcomes at 3-Month Follow-Up Visit
Chronic Migraine: n = 150 (%)
Episodic Migraine: n = 100 (%)
**P < .001 chronic migraine versus episodic migraine (chi-square test).
Number of responders
<1 per month
1 to 2 per month
3 to 4 per month
5 to 6 per month
7 to 10 per month
11 to 14 per month
Even if a patient's migraine can improve in time,8 in a minority it evolves into a CM complicated by medication overuse.9 The migraine at the onset seems the same, but there are different characteristics that may predispose the patient to develop CM. In our study, the patients in the CM group (Table 1) were mainly females (83%), of adult-old age (mean age 49 ± 13.4 years), and a large number of them were retired persons or housewives (46%). In the EM group, the prevalence of females was lower (69%); they were especially young and adult workers (mean age 36 ± 10.2 years). As a whole, migraine started earlier and remitted less frequently during pregnancy (Table 2) in the CM than in the EM group.
It has been reported that chronic daily headache carries a substantial genetic predisposition.10 Indeed, in our sample (Fig. 1), a family history for migraine and substance abuse was found significantly more often in the CM than in the EM group. Just these 2 characteristics, mutually interacting, might have been prejudicial to both migraine transformation and medication overuse. However, in the single patient, migraine evolution into a chronic pattern seemed to be a personal story-linked phenomenon, related in the majority of patients (96%) to particular events occurred in the course of private, relational, and working life.
A number of medical conditions (such as hypertension, allergy, ischemic stroke) have been associated with migraine,11 but psychiatric comorbidity, and in particular depression and anxiety, distinctively marks chronic headache.12,13 Our CM patients (Fig. 3) suffered above all from psychiatric disorders (67%) (anxiety and depression), followed by gastrointestinal disorders (43%), perhaps due to the overuse of NSAIDs, alone or combined. Polypharmacy is common in a Headache Centre population.14 In the case of CM group (Fig. 4), where 73% of the patients were daily taking nonheadache drugs, above all sedative-hypnotics (30%), followed by antihypertensive agents (26%) and antidepressants (23%), polypharmacy was probably a consequence of their comorbid diseases and the result of their frequent and different consultations. Indeed, the majority of CM patients (Fig. 5) had already carried out many diagnostic tests, and they had asked advice to their family doctor and to different specialists before being addressed to our Centre. Several patients of the CM group had already tried, even more than once, to withdraw overused medication, and 75% were taking a prophylactic treatment. It is understandable that during the search for an effective solution to bear the daily headache, they had taken acute medications. Among the different available symptomatic drugs (Fig. 6), patients in the CM group overused triptans above all (43%). This choice probably reflected the large number of consultations they had carried out, but, at the same time, it indicated that the effectiveness of triptans, even if symptomatic and transitory, persisted in spite of repeated assumptions. At variance with what has been reported in other studies,15,16 in our population there were few patients overusing ergot derivatives (9%) and no patient overusing barbiturates combinations. Triptans are not much prescribed by general practitioners; they are especially prescribed by headache specialists.17 On admission at the Headache Centre, only 31% of the EM patients was taking triptans; the majority (56%) was instead taking NSAIDs, as acute treatment. Upon discharge (Fig. 7), triptans were the most prescribed acute medications, either to the CM group (62%) or to the EM group (77%). Moreover, prophylactic treatment, especially with antidepressants (59%) (that present the clearest evidence of efficacy in chronic headache)18 was prescribed to every patient in CM group and to 86 out of 100 patients in EM group, in this case especially with flunarizine (58%). The effectiveness of currently used prophylactic treatments for transformed migraine is limited and a large number of patients, as a consequence, are at risk of relapse.19 In our study, at 3-month follow-up (Table 3), apparently the treatment resulted effective (it had reduced by at least 50% the frequency of headache) in the same proportion of both the CM (75%) and EM (71%) group. In fact, the majority of the CM group had still a frequency of headache much higher than that of EM group.
Our study has some limits: the population was heterogeneous as far as patients' characteristics, doses, and patterns of medication taken are concerned. Our findings apply to headache clinic populations and cannot be generalized to other patients and contexts.
From our results a picture of the typical CM patient emerges (Table 4). In this patient, CM does not present as an isolated disorder but coexists with multiple medical and psychiatric conditions, polypharmacy, and social impediments that complicate clinical management, represent a major challenge, and require long-term follow-up. Even if a considerable share of health-care resources is used by individuals with CM, the treatment of this disorder remains suboptimal. Several CM patients under prophylactic treatment may still manifest a frequency of headache higher than EM patients and therefore are at risk for future overuse. Triptan overuse may increase migraine frequency sooner than other medications20 but these drugs often are, in our experience, the only effective for the acute treatment in CM patients. Therefore, we think that in CM patients triptan use should be monitored in order to prevent overuse, but not banned.
Table 4.—. Summary of the Differences Between Chronic and Episodic Migraine Patients
More likely female
Less likely female
P < .001
Lower levels of educational attainment
Higher levels of educational attainment
P < .001
More likely to be married
More likely to be unmarried
P < .001
More likely to be unemployed
More likely to be employed
P < .001
P < .05
Migraine remission during pregnancy
P < .05
Disorders in first-order relatives
More likely (especially migraine, cardiovascular diseases, drug/alcohol abuse)
P < .05
More likely menopause, constipation
More likely oral contraceptives use
P < .001
Coexistence of other disorders
More likely (especially psychiatric, gastrointestinal, musculoskeletal disorders, and hypertension)
Less likely (possibly allergies)
P < .001
Use of health-care resources
Higher (especially neuroimaging, hospitalization, and other headache centers)
P < .001
More likely (especially sedative-hypnotics, anti-hypertensives)