Quality of life and psychiatric co‑morbidity in Indian migraine patients: A headache clinic sample
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.117584
Source of Support: None, Conflict of Interest: None
Background: There is a lack of data from India on the impact of migraine on health-related quality of life (HRQoL) and the extent of psychiatric co-morbidities in migraine. Objective: The objectives of the study were to quantify the impairment in HRQoL in migraine patients compared to healthy controls, to compare the prevalence of clinically significant anxiety and depressive symptoms in these groups, and to identify patient and headache characteristics that may predict health-related quality of life. Materials and Methods: We interviewed 71 consecutive newly diagnosed migraine patients seen in the headache clinic of a tertiary referral center between September and December 2008. Age- and sex-matched healthy subjects (n = 71) were used as controls. Short Form-36, Migraine Disability Assessment Score, and Hospital Anxiety and Depression Scale were administered. Predictors of HRQoL were identified using regression analysis. Results: Migraineurs were significantly impaired in all subscales of the SF-36 compared to controls, with greatest impairments in role physical, general health, and role emotional subscales. Prevalence of clinically significant anxiety (48%) and depressive (41%) symptoms in patients was higher than in healthy controls. Female gender, headache-related disability, and severity of anxiety predicted worse Physical Component Summary scores, while severity of both anxiety and depressive symptoms predicted worse Mental Component Summary scores. Conclusion: HRQoL is significantly reduced in Indian migraine patients compared to healthy controls. Incidence of clinically significant anxiety and depressive symptoms is also much higher in these patients. These findings corroborate well with studies from other parts of the world and suggest that cultural differences do not significantly alter the subjective impact of migraine on quality of life.
Keywords: Anxiety, depression, HADS, migraine, quality of life, SF-36
K. Sharma and R. Remanan contributed equally to this work
Migraine headache is one of the commonest neurological disorders in the world.  Migraine has been placed 19 th in the World Health Organization (WHO)'s list of diseases ranked by years lived with disability.  An episodic disorder with minimal physical sequelae, this disease's outcomes have traditionally been difficult to define. Over the past two decades, health-related quality of life (HRQoL) measures have been widely accepted as indices of disease impact on an individual. These measures use subjective ratings of the individual in a variety of areas to assess the overall impact of illness and therapy on a patient.  The Short Form-36 (SF-36) questionnaire is one such measure of HRQoL that has been extensively used in migraine patients. ,,,,,,
There is little or no representation of Indian patients in international studies of migraine-related QoL. ,,,,,, Such a study is important in India for a number of reasons. The pooled prevalence of migraine varies from 5% in Africa to 15% in Europe. Preliminary data from an ongoing population-based study from Karnataka, India suggest that this number could be as high as 23% for Indians.  Migraine may thus be a significant public health concern and defining the effect of migraine on HRQoL measures would further quantify the impact of this condition. Such measures are somewhat comparable across other chronic medical conditions and may serve as a guide to healthcare policy making.  They may also serve as a baseline and subsequent measure of effectiveness of therapy.
Migraine is known to be associated with psychiatric comorbidities, with reported prevalence of 20-40% for anxiety and 10-15% for depression. ,,, Presence of these comorbidities reduces the HRQoL of migraineurs, independent of headache characteristics and headache-related disability. , Any attempt to establish an association between HRQoL in migraine and headache characteristics is likely to be confounded by coexisting anxiety or depressive symptoms. There is no published literature from India highlighting this aspect of migraine.
The present study was therefore designed to perform a case-control comparison of HRQoL in migraine patients from a tertiary care center using SF-36 questionnaire to quantitate the extent of anxiety and depressive symptoms and to identify patient and headache characteristics as the possible predictors of HRQoL.
Ethical clearance from the institute ethical committee at the All India Institute of Medical Sciences (AIIMS), New Delhi, India was obtained.
Consecutive patients aged 18-65 years, attending the headache clinic of a tertiary referral center were studied between September and December 2008. Patients were diagnosed with migraine (with or without aura) according to the International Classification of Headache Disorder, 2 nd edition (ICHD-2) by the same neurologist.  Subjects with duration of headache of at least 6 months and age at onset of migraine of less than 50 years, with no prior history of migraine prophylactic medication were included in the study. Exclusion criteria were recent history of antipsychotic, antidepressant, antianxiety medications in the last 3 months, or past alcohol or recreational drug abuse.
Due to lack of normative data for SF-36 for the Indian population, age- and sex-matched healthy subjects were selected as controls. This was done by frequency matching for males and females separately after completion of enrolment in the patient group. To maintain similar socioeconomic profiles for both the patient and control groups, the control subjects were selected from among the family members of patients attending the general neurology clinics in the same hospital. "Healthy" controls were defined as individuals feeling subjectively healthy, and specifically denying any chronic illness including headaches, psychiatric disorders, hypertension, diabetes, chronic pain condition, skin disorders, chronic obstructive lung disease, or past history of major surgery requiring general anesthesia.
After obtaining informed consent from subjects, the following study instruments were administered by the authors.
Diagnosis of migraine was confirmed with an interview tool developed at AIIMS that uses ICHD-2 criteria.  Questions on various headache characteristics including duration of symptoms, presence and type of aura, precipitating factors, localization and frequency of headache, and maximum severity using a visual analog scale with a range of 0-100 were recorded on the questionnaire based on clinical history.
This validated 36-item questionnaire assesses HRQoL in eight domains: Physical functioning (PF), role physical (RF), body pain (BP), general health (GH), vitality (VT), social functioning (SF), role emotional (RE), and mental health (MH). The SF-36 also furnishes two summary scales for overall mental and physical health - the Mental Component Summary (MCS) score and Physical Component Summary (PCS) score, respectively. A difference of more than 5 points on any of the subscales is considered clinically significant according to the SF-36 survey manual.  Due to lack of a validated Hindi language version, the authors (K. S. and R. R.) agreed upon a semantically equivalent conversational Hindi framework for the interview.
Migraine disability assessment score (MIDAS)
This is a validated headache-specific tool that assesses the degree of migraine-related disability.  It captures information on lost time from work for pay, housework or chores, and leisure activities. The MIDAS score is simply a sum of the number of lost days in these three domains.
Hospital anxiety and depression scale
This is a validated 14-item questionnaire used to assess the severity of depressive and anxiety symptoms in separate scales. In accordance with empiric evidence, scores ≥8 on the anxiety (HADS-A) and depression (HADS-D) scales were taken as cut-offs for the presence of clinically significant anxiety and depressive symptoms. , The standardized English version was translated into Hindi language for the study and then back-translated to verify semantic equivalence with the assistance of the Hindi language department at AIIMS, New Delhi.
Statistical analysis was done using SPSS version 15 for Windows. Score comparison between patients and controls was done by Student's t-test. Multiple linear regression analyses were used to identify the predictors of physical and mental health summary scales in migraineurs while adjusting for psychological co-morbidity as determined by HADS scores. These variables included: Patient demographics (age, gender), headache characteristics [headache frequency, average headache intensity (0-100 on a visual analog scale)], and headache-related disability (MIDAS score). Significance was assumed at a value of P < 0.05.
Out of 71 subjects in each group, 56 (79%) were females. There was no statistical difference in the age distribution of patients (Mean 30.6 years, SD 9.1, range 18-50 years) and controls (Mean 31.3 years, SD 9.1, range 18-54 years) due to frequency matching employed in the study [Table 1].
Selected headache characteristics of the migraine group are summarized in [Table 1]. Chronic migraine patients were not treated as a separate group for most of the analysis since we wanted to evaluate the linear relationship of HA frequency with HRQoL.
The prevalence of migraine with aura in our study group was 23.9%, which is similar to the result of prior large-scale population-based studies (15-30%) from outside India, , with a higher prevalence among males (40%) than females (20%). Indian studies have reported the prevalence of aura as low as 2.25% in migraine patients.  Reason for the higher rate of aura in our study group may be due to the inclusion of only newly diagnosed cases and exclusion of patients on preventive migraine therapy. Another reason may be over-reporting of symptoms prompted by use of a migraine proforma in the study.
Results of the SF-36 questionnaire revealed both clinically (>5 points) and statistically significant (P < 0.05) reduction in all eight subscales and both summary scales [Table 2] as compared to the healthy controls.
Migraineurs had significantly higher anxiety (P < 0.001) and depressive (P = 0.021) symptoms than healthy controls [Table 2]. The prevalence of clinically significant anxiety and depressive symptoms according to original cut-offs proposed by Zigmond and Snaith is given in [Table 3]. 
To identify the predictors of HRQoL in migraine patients, the PCS and MCS scores were modeled on age, female sex (males as reference group), headache frequency per month, mean headache intensity, MIDAS, HADS-A and HADS-D scores [Table 4]. Female gender, headache-related disability, and presence of anxiety were also associated linearly with worse PCS score. As for MCS scores, both HADS anxiety and depression scales accounted for most of the variance in MCS scores. None of the headache characteristics significantly predicted MCS scores. Regression coefficients did not change appreciably even after excluding the HADS scores from the analyses (models not shown).
This study demonstrates a pervasive impairment in all dimensions of HRQoL in migraine patients as compared to matched controls. The large role limitation due to physical and emotional problems (RP, RE scales) with relatively intact physical and social functioning (PF, SF scales) in our study is similar to population- and clinic-based studies done in other parts of the world. ,,, Also, in keeping with previous similar studies, our patients showed a greater impairment in the physical summary scale as compared to mental summary scale. , It thus seems that cultural differences do not significantly alter the impact of migraine on the QoL.
In terms of psychiatric co-morbidity, our migraine patients had much higher prevalence of depressive and anxiety symptoms compared to controls. This was higher than previous reports from clinic-based samples of migraine patients. Devlen reported a prevalence of 50% for anxiety and 20% for depression (HADS scores >8), while corresponding values in this study are 48% and 41%, respectively.  The prevalence of probable anxiety and depression (score ≥ 11) in a migraine clinic study by Jarman et al. was 20% and 8% as compared to the prevalence in our study of 31% and 20%, respectively. 
In our patient sample, females had significantly higher impairment in physical HRQoL scales compared to males even after adjusting for headache frequency, severity, disability, and psychiatric illness. This is consistent with the findings of Lipton et al., who also noted the same trend in a population-based study though they employed the SF-12 questionnaire (and corresponding summary scales), an abbreviated 12-item version of the 36-item SF-36 questionnaire.  Wang et al. also noted an association of female sex with decreased scores in two of the physical health subscales (PF and BP) and one mental health subscale (RE). 
Only headache-related disability assessed from MIDAS score was found to be a significant predictor of poorer PCS scores, probably because this score essentially quantifies physical role limitations by questions related to reduced productivity, which are also components of the PCS scale. Interestingly, monthly headache frequency was not a predictor of HRQoL scores in our study, while previous studies have shown a definite association between increasing frequency and poorer scores, especially in the context of chronic migraines (monthly frequency ≥ 15).  Our study had a relatively low percentage of chronic migraine sufferers (27%) due to selection bias from excluding patients already on preventive therapy. This may have caused an effect size reduction of headache frequency on SF-36 and HADS scores. Since the study was not powered to detect differences between patient subgroups like episodic versus chronic migraineurs, we have refrained from reporting these post-hoc subgroup analyses.
We also attempted to compare the health effects of migraine with other pain conditions. We found one report of an Indian clinic-based study by Aggarwal et al., which assessed QoL in rheumatoid arthritis (RA) patients using the SF-36 study instrument and a methodology similar to ours.  That study showed much higher impairment in physical functioning (mean PF score 50) and social functioning (mean SF score 56) compared to this study (74 and 76, respectively). Since RA is a physically disabling disease, this is quite intuitive. However, it is surprising that the impairments in physical and emotional role performance (mean RP 33, RE 48) were quite similar to those of our patients (32 and 50, respectively). The authors of that study also noted that their RA patients fared better on mental and emotional health scores compared to international patient populations. They attributed this to possibly better family support systems in India. In our study, however, migraine patients fared quite poorly on the all the mental and emotional health measures compared to controls, a finding that is ubiquitous in previous international studies as well. It thus seems that migraine-related impairment in QoL cuts across socio-cultural boundaries. We must point out that the study by Aggarwal et al. did not have a control group, and in the absence of a population standard for the SF-36, such direct comparison across studies could be inaccurate.
To our knowledge, this is the first study in India that attempts to quantify the health impact of migraine in terms of QoL. An age- and sex-matched control group is a definite virtue of this study. The study was done in a headache clinic at a tertiary referral center (AIIMS), which limits the ability to generalize these findings to the general population. The degree of impairment in our study may be expected to be higher than migraineurs in the general population. There is already a much needed population-based study underway investigating QoL in headache patients in Karnataka, India.  It shall be interesting to compare the HRQoL impairments of migraineurs in the general population and a headache clinic population.
[Table 1], [Table 2], [Table 3], [Table 4]