Migraine, a primary headache disorder, is characterized differently from other types of headaches. According to the International Classification of Headache Disorders, migraine is characterized by headache with moderate or severe pain intensity accompanied by nausea/vomiting, and its worsening by physical activity [1]. This disorder has an impact on quality of life (QoL) [2], and its characteristics cause disability [3] to extents different from those of other types of headache. The disability during headache attacks, the so-called headache-related disability [4], is defined by WHO as the consequences of illness on work and function in other roles [5]. Disability comprises impairment of body function, limitation of activity, and restriction of participation that involves a range of activities from self-care to working with capacity, and moving around at home and being able to travel or participate in sports, and pursue household activities to voting [5].
Instruments have been developed to specifically measure headache-related disability and, except for a few, have been tested for validity, internal consistency, and reliability [6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23] (shown in
Instruments specific for measuring headache-related disability
Pain Disability Index [7] | Unspecified chronic pain | Ability to engage in 7 categories of life activities | – | Preliminary validity |
Henry Ford Hospital HDI [8] | Nonspecific types of headache | Ability to complete daily life tasks | Good construct validity, acceptable internal consistency, strong test–retest reliability | – |
HImQ [9, 10] | Migraine | 16-Item: pain intensity and lost time in work outside the home, in household work, and nonwork activities | Acceptable internal consistency, strong test–retest correlation | Weak to strong criterion validity, complexity in scoring |
MIDAS Questionnaire [11, 12, 13] | Migraine | Simple 5-item self-administered type derived in part from and solving the practical limitation of the HImQ, capturing lost time from work for pay, housework, and nonwork activities | Moderate to strong criterion validity, acceptable internal consistency, strong to very strong test–retest reliability | – |
HIT [14] | Nonspecific types of headache | 6-Item on pain, social-role limitations, cognitive functioning, psychological distress, and vitality | Good discriminant validity, acceptable internal consistency, excellent test–retest reliability | – |
IPDS [15] | Chronic daily headache, medication overuse | 20-Item assessing beliefs regarding autonomy/disability in different situations of life | Moderate to strong convergent validity, acceptable internal consistency | No reliability test |
HADLI [16] | Migraine, tension-type, cervicogenic headaches | 9-Item on specific activity of daily living | Acceptable internal consistency | Face validity, no reliability test |
FAIM questionnaire [17] | Migraine | 9 Mental functioning items measuring dimensions of attention/thought, perception, and relevant 5 items of a list of activities and participation | Good convergent validity, acceptable internal consistency | No reliability test |
BURMIG [18] | Migraine | 77-Item of burden and disability obtained from MIDAS | Acceptable construct validity, acceptable internal consistency, strong reliability test | – |
EUROLIGHT [19] | Primary headache disorders | 103-Item of burden and disability revised from BURMIG | Acceptable construct validity, excellent internal consistency, good reliability test | – |
HARDSHIP questionnaire [20] | Migraine and tension-type and medication overuse headache | Burden measurement | Epidemiological validation | No clinical validation |
HALT [21] | All headache disorders | Global burden measurement and for alleviation of the burden; direct and close derivative of MIDAS | Epidemiological validation | No clinical validation |
MIBS [22] | Migraine | Interictal migraine-related burden measurement in 4 domains: work/school, family/social life, making plans/commitments, emotional/affective/cognitive distress | Good validity, internal consistency and reliability test | – |
MPFID [23] | Migraine | 13-Item on daily physical functioning | Good convergent validity, internal consistency, and reliability | – |
HDI, Headache Disability Inventory; HImQ, Headache Impact Questionnaire; MIDAS, Migraine Disability Assessment; HIT, Headache Impact Test; IPDS, Italian Perceived Disability Scale; HADLI, Headache Activity of Daily Living Index; FAIM, Functional Assessment In Migraine; BURMIG, Burden of Migraine Instrument; HARDSHIP, Headache Attributed Restriction, Disability, Social Handicap, and Impaired Participation; HALT, Headache Attributed Lost Time; MIBS, Migraine Interictal Burden Scale; MPFID, Migraine Physical Function Impact Diary.
At the time this study was conducted, the most practical disability assessment specific for migraine was the MIDAS Questionnaire. This simple 5-item, self-administered questionairre is derived in part from and solved the practical limitations of the HImQ [12] capturing lost time from work for pay, house-work, and nonwork activities, and another 2 questions asking for total number of headache days in the past 3 months (question A) and mean headache severity (question B) [11]. It is short, easy to score and use, and provides meaningful information about lost days of activities within a 3-month timeframe [11]. The disability in people with migraine is potentially influenced by 2 major factors; one is directly related to the migraine itself and its treatments, and the other is indirectly related to the migraine such as comorbidities and their treatments. For example, depression comorbidity may cause more frequent migraine attacks, and thus cause more disability. For the migraine itself, a longer duration of headache, higher pain intensity, and more frequent migraine attacks may cause more days of lost activities per month. A MIDAS score of 0–5 days lost activity per month indicates little or no disability and is equal to MIDAS grade I; a score of 6–10 days lost activity indicates mild disability and is equal to MIDAS grade II; a score of 11–20 days lost activity indicates moderate disability and is equal to MIDAS grade III; and a score of 21 days or more indicates a severe disability and is equal to grade IV.
MIDAS also tests for sensitivity to change between episodic and chronic migraine [24] and changes between chronic migraine with a high frequency of medication use for headache (≥15 days/month) and a low frequency of medication use (<15 days/month) [25].
In Thailand, the Thai-version MSQ 2.1 has been used to assess QoL of people with migraine [26], but a translation of a questionnaire specific for assessing migraine-related disability has not been available. Two Thai versions of the MIDAS questionnaire have been published, one in 2013, and the other in 2018 [27, 28]. The first Thai questionnaire was only tested for content equivalence to the original English version by professional translators and medical specialists, but was not tested for language equivalence. In addition, the 5 items and the total score of that Thai questionnaire had only weak-to-strong criterion validity as measured by Spearman correlation coefficients and fair-to-excellent test–retest reliability as measured by intraclass correlation coefficient (ICC) [27]. Moreover, the questionnaire was not tested for internal consistency. The other Thai questionairre was only tested for test–retest reliability and internal consistency [28]. The present study fills the gaps in knowledge mentioned above and was primarily aimed at testing for concurrent validity and test–retest reliability, and secondarily at testing for internal consistency of the Thai-version MIDAS Questionnaire, and as a pilot study, for factors predictive of disability including duration of migraine history, migraine characteristics, and comorbidities for migraine-related disability as assessed by the Thai-version MIDAS Questionnaire.
The present study was conducted with a prospective design at the Chulalongkorn Comprehensive Headache Centre at King Chulalongkorn Memorial Hospital, Faculty of Medicine, Chulalongkorn University in 2011–2012. The Chulalongkorn Institutional Review Board approved the study (approval in 2554 (2011), endorsement No. 129/2560, IRB No. 418/54).
The Chulalongkorn Comprehensive Headache Centre at King Chulalongkorn Memorial Hospital was established in 2000. This center has registered patients with headache and recorded patient information with a standardized patient record form including socioeconomic data; comorbidity; cointervention; and history of migraine headache including the history of duration, location, pain quality, pain severity, pain score, pain duration, and frequency of headache. Patients also received a headache diary to prospectively record their headaches and treatments, if any.
In the present study, we included patients who visited the center and who were diagnosed as having migraine by an attending neurologist with experience of migraine treatment according to the International Classification of Headache Disorders, second edition (ICHD-II). The inclusion criteria were patients aged 20 years or older with episodic or chronic migraine, with or without aura, who were literate in the Thai language and gave their written informed consent to participate in this study. The diagnostic criteria of migraine in this study used the criteria from the ICHD-II that was published in 2004. The present study was conducted in 2011–2012 before the ICHD-III beta version was released in 2013, and then as the ICHD-III in 2018. However, the diagnostic criteria for migraine without aura were not changed except for the diagnostic criteria for migraine with aura and chronic migraine. Migraine with aura in ICHD-II would have to fulfill one of the subforms of 1.2.1–1.2.6 that were as follows: typical aura including visual, sensory, and with dysphasic speech with migraine/nonmigraine or without a headache, familial/sporadic hemiplegic or basilar-type migraine. In the ICHD-III, migraine with aura includes one or more of the following fully reversible aura symptoms: visual, sensory, speech and/or language, motor, brainstem, and retinal. The retinal aura in the ICHD-III was not included in the migraine with aura in the ICHD-II, but it was separated into another entity. The chronic migraine in the ICHD-II would have to fulfill the migraine without aura on ≥15 days/month for >3 months. In the ICHD-III, the headache days and the duration remained the same as in the ICHD-II, but the headache characteristics could be migraine-like or tension-type-like, and on ≥8 days/month for >3 months, the headache would have to be migraine without aura.
We received permission to use the original English MIDAS questionnaire from AstraZeneca Pharmaceuticals. It was translated into Thai by the first author (TA), a bilingual neurologist with back-translation into English by a bilingual physician with a high degree of English proficiency who was blinded to the original version and was independent of the first author. This process was repeated until the original and the back-translation were identical in content. Finally, the language equivalence of the Thai-version MIDAS Questionnaire was examined by an active informant experienced with migraine headache at the Chulalongkorn Comprehensive Headache Centre.
The Thai-version MIDAS Questionnaire and a questionnaire asking about migraine characteristics including headache frequency, mean pain duration, and mean pain intensity in the past 3 months, 2 months, 1 month, and 2 weeks were administered to patients at our headache center. Participants were asked to complete the first questionnaire at the center and to return them before leaving. They were also asked to use a headache diary to help them complete the following questionnaires. Then, before leaving the center they received the second questionnaire with to be completed at home within the following 2 weeks and requested them to mail the completed questionairre to the center in the addressed-and-stamped envelope provided for their convenince. The patients were telephoned to remind them to complete the questionnaire and return it based on the assigned date. With 2-week intervals for the mail delivery process, any second questionnaires that were missing or incomplete would be replaced with other eligible patient questionnaires until meeting the assigned the sample size. For concurrent validity, within the first questionnaire, the second, and a combination of both, the MIDAS total score in the Thai-version questionnaire was validated with the total number of headache days (question A), mean headache severity (question B), and migraine characteristics in the past 3 months. The test–retest reliability for the individual question, MIDAS total score, total number of headache days (question A), mean headache severity (question B), and Thai-version MIDAS grade were evaluated. Migraine characteristics between the 2 occasions were also evaluated to measure whether any differences might affect the test–retest reliability. The migraine characteristics included frequency of headache, mean attack duration, and mean pain score measured as a numerical rating scale (NRS) score. The internal consistency test of the Thai-version MIDAS Questionnaire was evaluated for each occasion.
For the combination of both occasions of completing the questionnaires, duration of migraine history, comorbidity, and migraine characteristics in the past 3 months, 2 months, 1 month, and at 2 weeks were evaluated for independent predictive factors for migraine-related disability as assessed by the Thai-version MIDAS grade. The Thai-version MIDAS score is categorized into 4 grades according to the original version; grade I: little or no disability, grade II: mild disability, grade III: moderate disability, and grade IV: severe disability.
The baseline demographic and socioeconomic data and migraine characteristics between eligible and ineligible patients were compared with a Pearson χ2, Fisher exact, or Mann–Whitney
Sample size for test–retest reliability with a 2-tailed hypothesis tests was based upon ICC with a type I error of 0.05, power of 0.80, a minimally acceptable level of reliability of 0.40, and a proposed level of reliability for this study of 0.75 [29] yielded a required sample size of 27 participants. The level of reliability of 0.40 and 0.75 was based upon guidelines described by Cicchetti [30]. The sample size for internal consistency and predictive factors was not calculated.
Of the 58 participants finishing the first questionnaire, 32 mailed back the second. Of the 32 returned second questionnaires, 1 had some missing data. Therefore, 31 participants were eligible for this study (
Migraine characteristics in the eligible patients before completing their first Thai-version MIDAS Questionnaire
Frequency of headache, median attack (interquartile range) | 8 (14) | 6 (9) | 4 (7) | – | – | – |
Mean frequency of headache per month, median attack (interquartile range) | 2.7 (4.7) | 3 (4.5) | 4 (7) | 0.34 | 0.21 | 0.79 |
Mean pain duration per attack, median hours (interquartile range) | 2 (2) | 2 (2) | 2 (3) | 0.22 | 0.18 | >0.99 |
Mean pain score, median NRS score (interquartile range) | 5 (3) | 5 (2) | 5 (5) | 0.75 | 0.39 | 0.39 |
MIDAS, Migraine Disability Assessment; NRS, numerical rating scale.
Between 3 and 2 months.
Between 3 and 1 months.
Between 2 and 1 months.
Migraine characteristics before completing their first and their second Thai-version MIDAS Questionnaire
In the past 3 months (interquartile range) | |||
Frequency of headache, median attack | 8 (14) | 9 (11) | >0.99 |
Mean pain duration per attack, median hours | 2 (2) | 2 (2) | 0.34 |
Mean pain score, median NRS score | 5 (3) | 5 (4) | 0.58 |
In the past 2 months (interquartile range) | |||
Frequency of headache, median attack | 6 (9) | 6 (9) | 0.55 |
Mean pain duration per attack, median hours | 2 (2) | 2 (2) | >0.99 |
Mean pain score, median NRS score | 5 (2) | 5 (3) | 0.75 |
In the past 1 month (interquartile range) | |||
Frequency of headache, median attack | 4 (7) | 4 (6) | 0.61 |
Mean pain duration per attack, median hours | 2 (3) | 2 (2) | >0.99 |
Mean pain score, median NRS score | 5 (5) | 5 (4) | 0.77 |
In the past 2 weeks (interquartile range) | |||
Frequency of headache, median attack | 1 (3) | 2 (4) | 0.22 |
Mean pain duration per attack, median hours | 1 (2) | 1 (2) | >0.99 |
Mean pain score, median NRS score | 4 (1) | 5 (3) | >0.99 |
MIDAS, Migraine Disability Assessment; NRS, numerical rating scale.
The MIDAS total score of the Thai-version MIDAS Questionnaire in the first completion, the second completion, and a combination of both completions was tested for validity with the total number of headache days (question A), mean headache severity (question B), and migraine characteristics in the past 3 months. The results demonstrated that the mean headache severity (question B), the mean pain duration per attack and the mean pain NRS score had a moderate correlation with the MIDAS total score with the Spearman correlation coefficients of 0.49, 0.58, and 0.49; 0.42, 0.56, and 0.42; and 0.45, 0.57, and 0.45, respectively with the exception of the total number of headache days (question A) and frequency of headache with Spearman correlation coefficients of –0.07 to 0.20 (
Concurrent validity of the total number of days from 5 questions of the Thai-version MIDAS Questionnaire, and question A, question B, and migraine characteristics in the past 3 months
Total number of headache days (question A) | –0.07 | 0.10 | 0.20 |
Average headache severity (question B) | 0.49* | 0.42* | 0.45* |
Frequency of headache | –0.07 | 0.10 | 0.20 |
Mean pain duration per attack | 0.58* | 0.56* | 0.57* |
Mean pain score, NRS | 0.49* | 0.42* | 0.45* |
MIDAS, Migraine Disability Assessment; NRS, numerical rating scale score.
The test–retest reliability is shown in
Test–retest reliability of the Thai-version MIDAS Questionnaire
1. “On how many days in the last 3 months did you miss work or school because of your headaches?” | |||
Mean (SD) | 1.4 (3.93) | 1.5 (4.00) | 0.94 |
Median | 0 | 0 | |
Min–max | 0–21 | 0–21 | |
2. “How many days in the last 3 months was your productivity at work or school reduced by half or more because of your headaches? (Do not include days you counted in question 1 where you missed work or school.)” | |||
Mean (SD) | 1.8 (4.30) | 1.5 (3.90) | 0.95 |
Median | 0 | 0 | |
Min–max | 0–21 | 0–21 | |
3. “On how many days in the last 3 months did you not do household work (such as housework, home repairs and maintenance, shopping, caring for children and relatives) because of your headaches?” | |||
Mean (SD) | 1.4 (3.97) | 1.0 (3.83) | 0.95 |
Median | 0 | 0 | |
Min–max | 0–21 | 0–21 | |
4. “How many days in the last 3 months was your productivity in household work reduced by half or more because of your headaches? (Do not include days you counted in question 3 where you did not do household work.)” | |||
Mean (SD) | 1.7 (3.95) | 1.7 (4.07) | 0.94 |
Median | 0 | 0 | |
Min–max | 0–21 | 0–21 | |
5. “On how many days in the last 3 months did you miss family, social, or leisure activities because of your headaches?” | |||
Mean (SD) | 1.5 (3.94) | 1.1 (3.83) | 0.95 |
Median | 0 | 0 | |
Min–max | 0–21 | 0–21 | |
Total number of days from 5 questions | |||
Mean (SD) | 7.8 (19.50) | 6.8 (19.03) | 0.98 |
Median | 0 | 0 | |
Min–max | 0–105 | 0–105 | |
A. “On how many days in the last 3 months did you have a headache? (If a headache lasted more than 1 day, count each day.)” | |||
Mean (SD) | 18.4 (25.73) | 17.7 (24.99) | 0.97 |
Median | 8 | 9 | |
Min–max | 1–90 | 1–90 | |
B. “On a scale of 0 to 10, on average how painful were these headaches? (where 0 = no pain at all, and 10 = pain as bad as it can be)” | |||
Mean (SD) | 5.5 (2.15) | 5.6 (2.23) | 0.89 |
Median | 5 | 5 | |
Min–max | 1–10 | 1–10 | |
Grade I | 22 (71%) | 23 (71%) | 0.66† |
Grade II | 3 (10%) | 4 (13%) | |
Grade III | 2 (7%) | 4 (13%) | |
Grade IV | 4 (13%) | 1 (3%) |
The Thai-version MIDAS score is categorized into 4 grades according to the original version, grade I: little or no disability, grade II: mild disability, grade III: moderate disability, and grade IV: severe disability.
ICC, intraclass correlation coefficient; MIDAS, Migraine Disability Assessment.
Weighted κ.
The median duration of migraine history and migraine characteristics in the past 3 months, 2 months, 1 month, and 2 weeks including frequency of headache, mean pain duration per attack and mean pain NRS for the combination of both occasions were significantly associated with migraine-related disability (
Factors predictive of migraine disability
Median duration of migraine history | <0.001 | 0.35 |
Having comorbidities | 0.30 | – |
In the past 3 months | ||
Frequency of headache, median attack | <0.001 | >0.99 |
Average pain duration per attack, median hours | <0.001 | 0.73 |
Average pain score, median NRS score | <0.001 | 0.007 |
In the past 2 months | ||
Frequency of headache, median attack | <0.001 | 0.76 |
Average pain duration per attack, median hours | <0.001 | 0.95 |
Average pain score, median NRS score | <0.001 | 0.13 |
In the past 1 month | ||
Frequency of headache, median attack | <0.001 | 0.33 |
Average pain duration per attack, median hours | 0.001 | 0.14 |
Average pain score, median NRS score | <0.001 | 0.56 |
In the past 2 weeks | ||
Frequency of headache, median attack | 0.23 | – |
Average pain duration per attack, median hours | 0.73 | – |
Average pain score, median NRS score | <0.001 | 0.86 |
MIDAS, Migraine Disability Assessment; NRS, numerical rating scale.
There have been 9 non-English versions of the MIDAS questionnaire, with Italian [31], Japanese [32], Turkish [33, 34], Taiwanese [35], Thai [27, 28], Arabic [36], Greek [37], German [38], and Spanish [39] versions that have been previously reported (
Summary translation process, validity, internal consistency, and test–retest reliability of non-English MIDAS questionnaires
Italian [31] | Two independent | Two different experts | Not mentioned | 0.7 | 0.54–0.77 |
Japanese [32] | Not mentioned | Not mentioned | 0.36–0.88 | Not mentioned | 0.59–0.83 |
Turkish 1 [33] | Not mentioned | Not mentioned | 0.47–0.63 | 0.73–0.87 | 0.68 |
Turkish 2 [34] | A physician | A neurologist | 0.54–0.89 | 0.78–0.90 | 0.66–0.81 |
Taiwanese [35] | Following standard technique, not detailed | Following standard technique, not detailed | 0.34–0.37 | 0.79 | 0.48–0.73 |
Thai [27] | Three; one professional | One professional | 0.32–0.74 | Not mentioned | 0.58–0.82† |
Thai [28] | Two; one neurologist, another native speaker | One fluent in English | Not tested | 0.95† | 0.36–0.94 |
Arabic [36] | One professional | Independent professional | 0.71–0.92 | 0.86, 0.96 | 0.97–0.99† |
Greek [37] | Two bilingual speakers | Two bilingual speakers | Correlation with RAND-36 | 0.71, 0.82 | 0.41–0.81† |
German [38] | Two bilingual independent different background | Two bilingual independent different background | Not mentioned | 0.69, 0.67 | 0.88–0.99† |
Spanish [39] | Two bilingual independent experts | Two bilingual experts | 0.53, 0.34 | 0.79 | 0.63–0.90† |
Present study | Bilingual neurologist | Another bilingual physician with blinding and independence | 0.45–0.57‡ | 0.98 | 0.89–0.98† |
MIDAS grade | 0.66§ |
MIDAS, Migraine Disability Assessment.
Intraclass correlation coefficient.
Mean headache severity (question B), mean pain duration per attack and numerical rating scale score.
Weighted κ.
In the present study, the process involving the forward and backward translation of the Thai-version language questionnaire was a limitation and this study conducted in 2011–2012 in the midst of the release of the new translation guidelines in 2011 [40]. However, the patients who participated were typical and would generally represent people with migraine. Most had episodic migraine with a median frequency 3–4 attacks of headache per month and were mostly stable in migraine conditions (
The concurrent validity of the Thai-version MIDAS Questionnaire between the MIDAS total score and mean headache severity (question B) was moderately correlated. The concurrent validity between the MIDAS total score and the mean pain duration per attack and the mean pain NRS was also moderately correlated. However, the concurrent validity between the MIDAS total score and the total number of headache days (question A) and the frequency of headache was very weak or weakly correlated. These results imply that even though the total number of headache days and the frequency of headache were higher, which includes the short pain duration per attack and the low pain score, the headache may not interfere with daily activities. By contrast, when total numbers of headache days were few, but were with long pain duration and high pain score, the migraine would interfere with daily activities causing lost days.
Interestingly, factors predictive for migraine-related disability in the present study were the median of the mean pain NRS score in the past 3 months as an independent predictive factor for migraine-related disability. The frequency of headache at any period was not predictive for migraine disability. This supports the hypothesis that the frequency of headache is not correlated with the MIDAS total score that counts missed daily activities and is not a predictor for migraine disability. To determine whether other factors that were not shown to be predictive for migraine-related disability, a study with a larger sample size should be conducted. Migraine prevention might have an effect on migraine-related disability and is a limitation of this study.
This study had some other limitations including the translation process for the Thai-version questionnaire, a lack of sample size calculation for the internal consistency test, not including migraine treatment in the analysis of factors predictive for migraine-related disability, and not adjusting
The present Thai-version MIDAS Questionnaire had moderate concurrent validity, acceptable internal consistency, and excellent test–retest reliability. It would be helpful to assess clinical outcomes. Future study with a standardized translation process for the Thai language questionnaire and larger sample size are required to confirm internal consistency and determine other factors predictive for migraine-related disability.