Abstract
This study investigated the intensity of stress, anxiety and depression in a sample of 141 migraineurs compared with a control group of 109 non-migraine workers matched for age and sex. Stress was measured using the Perceived Stress Questionnaire, and anxiety and depression using the Hospital Anxiety and Depression Scale. Results indicated that stress and anxiety were higher in the migraine group than in the control group and above the clinical level. Depression scores remained low in both groups, under clinical relevance. Stress is a primordial factor in the triggering and perpetuation of migraine attacks. The high score of the items ‘morning fatigue’, ‘intrusive thoughts about work’, ‘feeling under pressure’, ‘impatience’, and ‘irritability’ of the stress questionnaire in the migraineurs is particularly significant in the intensive stress response. It seems necessary to manage stress to improve the daily life of migraineurs and to study the link between stress, anxiety and migraine.
Introduction
Migraine is often associated with psychopathological disorders, particularly anxiety and depression (1–3). Anxiety disorders may appear in early childhood or adolescence, preceding the onset of migraine and later followed by the development of depression (4, 5). These disorders are often highlighted as trigger factors of migraine attacks, but may also appear as a psychological reaction to recurrent and severe migraine attacks (6, 7).
The link between these disorders remains unclear. Possible causal relationship, genetic or environmental aetiology, and psychological vulnerability have been proposed in the implication of these disorders in migraine (8–11).
Migraine can be influenced by different psychological factors; among them, stress is frequently considered to precipitate, exacerbate and maintain migraine (12, 13). Stress is a specific adaptive and defensive physiological reaction to a great variety of physical or psychological stimuli. This personal reaction depends on the nature and intensity of the stressor, on the social context, and on the subject's ability to appraise and to cope with the events, according to his threshold of vulnerability.
De Benedettis (14) investigated the role of stressful life events in the persistence of migraine attacks. The onset of migraine can be preceded by a sudden increase in frequency or density of major stressful life events such as death, separation, divorce, associated with a cognitive emotional appraisal leading to a negative impact on life. However, minor daily hassles can contribute to the persistence of migraine; their impact depends not only on their recurrence but also on how they are judged by the individual. In the persistence of migraine, there is a tendency to appraise cognitively and emotionally any daily hassle as being frightening. This preexistent psychological vulnerability induced by stress and the outcomes of attacks, keeps migraineurs in constant adaptive efforts.
The aim of the present study was to evaluate the level of perceived stress in a group of migraineurs with a specific questionnaire: the Perceived Stress Questionnaire. It was used for the first time to study cardiovascular risk in a working population. It has never before been used in the migraine domain (15).
Materials and methods
The aim of this study was to evaluate the stress level and the intensity of anxiety and depression disorders in migraine patients attending a specialized centre, compared with a control group of healthy workers.
Subjects
The study was conducted in Laennec Hospital (Paris) in the Cephalalgia and Migraine Centre, and in the Department of Occupational Medicine of the same hospital, for the control group. The studied population consisted of two groups of adults: one of migraineurs visiting the centre for the first time (International Headache Society (IHS) criteria (16) were used to establish the diagnosis of migraine) and a control group of non-migraineurs, healthy volunteers matched for age and sex, working in Laennec Hospital and examined for their annual preventive medicine visit. Informed consent was obtained from all the subjects.
Methods
In this study, two different scales were used: a Perceived Stress Questionnaire to measure level of stress in each patient, and the Hospital Anxiety and Depression (HAD) scale. A directive interview was led by the same person to confirm the IHS inclusion criteria and to investigate the psychosocial characteristics of the migrainous population.
The Perceived Stress Questionnaire is a self-assessment instrument of perceived stress. It was elaborated by Consoli et al. (15) to evaluate daily stress routinely, it's impact on health, and to understand the links between stress and disease, especially cardiovascular ones. The questionnaire consists of nine items with four possible answers, rated from 0 to 3. The content of the nine items covers the perceived stress and its consequences: feeling of being under pressure, impatience, irritability, intrusive thoughts about work, inability to entertain, discouragement, morning fatigue, food compensation, smoking compensation. The global stress score is defined as the sum of the nine elementary scores. The score of clinical significance is 12 and over.
The HAD scale (17) is a self-assessment instrument of 14 items to evaluate anxiety and depression in physically ill populations. Each item has four possible answers, scored from 0 to 3. The score of clinical significance for the two subscales is 10 and over.
Statistical analysis
Results are given as means ± SD for the overall scores of each scale and for each item. The comparisons between the scores of each group were made with Student's t-test with unequal variance.
As there was no previous stress study in a migrainous population, the difference expected between controls and the migraineurs was difficult to forecast, therefore the size of the population was chosen in order to have a sufficient size of group taking into account the size of the population in previous studies concerning anxiety and depression (2, 8).
Results
The study lasted 18 months, from September 1997 to January 1999. The characteristics of the 250 subjects are listed in Table 1. It describes the average and standard deviation of the global scores in the migraine and control groups, as a whole and according to gender. The difference of age between groups as a whole or taking gender into account is not significant.
Comparison of the stress, anxiety and depression mean scores: migraine group vs. control group
Comparison between global migraineurs vs. non-migraineurs
∗P < 0.05; migrainous women vs. non-migrainous women
∗∗P < 0.05; migrainous men vs. non-migrainous men
∗∗∗P < 0.05.
Comparing global scores of each scale there is a statistical difference between the migraineurs and the control group, whatever the scale studied.
It appears that the global stress score is statistically higher in the migraine group (12.4 ± 3.5) than in the control group (8.3 ± 4.0), and clinically relevant (>12) while it is not in the control group. The global stress score in migrainous women (12.8 ± 3.6) is higher than in migrainous men (11.3 ± 3.0), and both are clinically significant. The controls (hospital workers without headache) may be considered as ‘normally’ stressed people: scores < 10.
Concerning anxiety, the relevant clinical score is not reached in migrainous men, nor in controls, whatever the gender. There is a statistical difference between global migraineurs and global controls (P < 0001) because of the high proportion of women and between migrainous women and matching controls.
None of the groups had a clinically significant depression score; the difference between men concerning depression is not statistically significant, while the difference is clearly significant for women, but none of the groups can be considered as depressed (depression score < 10).
Characteristics of the migrainous population
Among the 141 migraineurs, 95 patients had a familial previous history of migraine, 27 of depression, 27 of allergy, and 33 of hypertension. One hundred patients had a personal previous history of anxiety, 40 of depression, 29 of neuro-vegetative dystonia, 30 had used anxiolytics, seven had a psychiatric previous history, 15 had had psychological treatment, one had a personal previous history of vascular disease, and 31 had no medical past.
The majority of patients had on average 3.4 migraine attacks in a month, lasting from a few hours to 36 h. The age at the first attack was around 18 years. The onset of migraine may have been in puberty, but patients also emphasized changes in their life: new work, birth of a child, moving house, getting married, professional overwork, returning to work after a child's birth, and finally, life as a student (changing school, going to university or starting professional life). The migraine attack began at the end of the night for 97 patients, in the evening for 17, during the day for 40, and at any time for 47 patients. The prodroms were nausea for 60 patients, fatigue for 78, visual disturbance different from aura for 54 patients, irritability for 40, anxiety for 62, feeling of depression for 14, speech difficulties for 35, concentration problems for 84 patients.
Symptoms associated with migraine were nausea for 128 patients out of 141, vomiting for 65, intestinal disorders for 18, photophobia for 132, phonophobia for 134, anxiety for 43, unbearable smells for 50 patients.
Stress was considered a trigger factor for migraine for 122 patients of 141, anxiety for 67 patients, emotions for 47 patients, depression for eight patients, menstruation for 76 patients out of 106 women, ovulation for 12 women, physical effort for 33 patients, intellectual effort for 38, heat for 48, cold for 23, alcohol consumption for 68, chocolate for 30, and finally cheese for 10 patients. Forty-seven women of 106 were on oral contraceptives, and generally women had no migraine during their pregnancy. Of 141 patients, 116 still had migraine attacks during holidays. Finally, migraine sufferers considered that their quality of life was very impaired: 116 patients had a disturbed family life, 120 patients a disturbed professional life, 111 patients a disturbed sexual life, and only seven considered they had a good quality of life.
Discussion
Analysis of the comparison of the average scores of the HAD scale and the Perceived Stress Questionnaire confirms our hypothesis. Stress and anxiety disorders are higher in the migraine group when compared with the control group, and are above the clinical level. On the other hand, even though the depression score is significantly different in migrainous women compared with controls, the score is under the level of clinical relevance. In this study, depression is not associated with migraine profile; this is not surprising because it is usually noted that depression is more often associated with tension-type headache than with migraine (14, 18). This may indicate that when depression occurs in migraineurs it is more likely to be a consequence than a cause, although this is not a universally held opinion (7).
Table 2 describes in detail the average scores of the different items of the Perceived Stress Questionnaire in the migraine and control groups, and separately for women and men.
Mean scores of the different items of the Perceived Stress Questionnaire in both groups
Four levels of answer from 0 to 3: maximum global score, 27; never, 0; sometimes, 1; often, 2; very often, 3.
∗P < 0.05; migrainous women vs. non-migrainous women
∗∗P < 0.05; migrainous men vs. non-migrainous men
∗∗∗P < 0.05.
In the migraine group, it is noted that the highest stress score for both genders concerns item 7, ‘morning fatigue’, and then item 4, ‘intrusive thoughts about work’ with a higher score for men; and finally items 2, ‘impatience’, and 3, ‘irritability’. On the other hand, item 8, ‘smoking compensation’, has the smallest score for both women and men. This low score might be explained by the fact than 60 patients out of 141 had nausea just before the crisis and 128 patients during the attack. There were few smokers among migraineurs in our group.
In the control group, the highest score were for item 4 and then item 7. They were highest for both groups, with higher scores for the migraine group. Item 1, ‘feeling of being under pressure’, had a high score for the migraine group, but had one of the lowest scores in the control group. Item 5, ‘inability to entertain’, had the lowest score in the control group. Globally the results show that the migraineurs had a more intensive stress response compared with non-migraineurs. In 97 patients out of 141 the beginning of the migraine attack occurred in the night or when waking up; this may explain the highest score of the stress scale item, ‘morning fatigue’, in the migraine sufferers.
For many patients, stress is underlined as one of the most common trigger factors with the menstrual cycle and alcoholic beverages (12). In our study, 122 patients estimated that stress was a primordial factor in the triggering and duration of migraine attacks. Stress seems to influence the course of migraine disease, and it causes relapse, a recrudescence of attacks so that the once effective pharmacological treatment becomes ineffective.
Stress influences migraine negatively, and recurrent migraine attacks can produce stress in the end and affect the migraineur's quality of life; the impact of migraine on the patient's quality of life is determined by the severity of the disease, the way it is handled by the patient and by the environment (19).
Quality of life is impaired not only during migraine attacks, but also in the symptom-free periods between attacks, through fear of the next one. Migraineurs perceive more subjective symptoms and have an increased anxiety and greater emotional stress, as well as disturbed contentment, vitality and sleep (20).
As a consequence of the disease's duration, it appears that migraine patients seem to be more inclined to use internal defences and to seek less social support in coping with problems (18).
Migraineurs indicate that to avoid recognized triggers, they are ready to adjust their life style and schedules. This has an impact on their quality of life because it has personal, professional and socioeconomic implications (21).
In different studies of quality of life in migraine patients with generic (SF36) and specific scales, it appears that a severe decrease exists in quality of life during the course of their disease. The use of these instruments has increased the awareness of the burden of migraine and its impact on health economic costs, and has shown the improvement obtained with therapeutic interventions (22, 23).
Our study concerns a particular clinic sample of migraine sufferers in whom pharmacological treatment has failed, justifying their coming to a specialized centre. They are not representative of all migraineurs, because about 40% of migraineurs never visit a doctor for this specific purpose. Our patients must then have had a more severe headache then the average migrainous population.
Our study underlines the links between anxiety, stress and migraine. Special attention should be given to the management of stress in the overall care of migraineurs, and specific studies are necessary.
There is an increasing interest in the evaluation of stress and its effects on health. In the field of migraine, general well-being and quality of life are increasingly taken into account as judgment criteria of drug efficiency.
Migraine, which mostly affects subjects from 17 to 55 years old, constitutes an important social handicap which has gained attention from medical professionals and health economists. Our study shows that migraine patients seem to have a greater emotional stress response; stress is considered a trigger factor of migraine attacks. It influences the quality of life and the general well-being of migraine patients, so it seems necessary to manage stress to improve the daily life of migraineurs. In a multifactorial treatment perspective of migraine, one should try to exclude situations of conflict or aggression, and also to influence factors increasing anxiety. The treatment strategies must take each case into account, patients’ usual behaviour and their environment; a psychotherapeutic treatment (psychological aid therapy, relaxation, biofeedback, familial therapy) associated with pharmacological treatment allows some patients to recover their stability, and to avoid the anxious anticipation of the next attack.