Get access

Migraine Disability Assessment (MIDAS) Score: Relation to Headache Frequency, Pain Intensity, and Headache Symptoms


Address all correspondence to Dr. Walter F. Stewart, Johns Hopkins School of Public Health, 615 North Wolfe Street, Baltimore, MD 21205.


Objective.—To determine the extent to which variation in the Migraine Disability Assessment (MIDAS) score is associated with headache frequency, pain intensity, headache symptoms, gender, and employment status.

Background.—The MIDAS questionnaire is a 7-item questionnaire (with 5 scored items) designed to measure headache-related disability, to improve physician-patient communication, and to identify patients with high treatment needs.

Methods.—Data from 3 population-based studies (total sample, n  =  397) conducted in the United States and the United Kingdom were used to evaluate the relationship between headache features (attack frequency, pain intensity, pain quality, and associated symptoms) and MIDAS score. Data on headache features were collected by telephone using a standardized interview. The MIDAS questionnaire was completed shortly after the telephone interview. General linear models were used to determine the extent to which population variation in the MIDAS score was explained by headache features.

Results.—Using linear regression, variables for all headache features (ie, headache frequency, pain intensity, pain quality, and associated symptoms) and demographic characteristics explained only 22% of the variation in MIDAS scores. Almost all (19.9%) the explained variance was accounted for by average pain intensity (12.0%), number of headache days (6.1%), and exacerbation of pain with movement (1.8%). When pain intensity and headache frequency were included in the model, no statistically significant differences in MIDAS scores were observed by gender or employment status. Although explaining only 2.1% of the variance, age was significantly associated with MIDAS scores, with those under 25 years demonstrating higher MIDAS scores than other age groups. No other variables (ie, frequency of occurrence of associated symptoms and other measures of quality of pain) were associated with MIDAS scores.

Conclusions.—Challenges to the utility of the MIDAS as a measure include whether headache-related disability is largely a function of other routine headache features and whether MIDAS is inherently biased based on work status and gender. While the MIDAS score was associated with headache frequency and average pain score, these two headache features explain only a modest proportion of the variation in MIDAS scores. Additionally, gender and work status were not related to MIDAS scores. These findings suggest that the MIDAS score captures information about disability that is not inherent to other headache features and is independent of gender and work status.