Prevalence of migraine headache and association with sex, age, race, and rural/urban residence: a population-based study of Georgia Medicaid recipients.

The estimates of migraine headache prevalence vary widely and fluctuate with the population examined and the methodologic factors used in studies examining this condition. As an alternative to survey techniques, a retrospective review of Medicaid claims data from 22 continuous months (January 1, 1989, to October 31, 1990) was used to detect medical episodes and physician-initiated pharmacologic therapy indicative of migraine headache. Specifically, the objectives of this study were to measure the prevalence of migraine headache in Georgia Medicaid recipients, estimate the prevalence in the US population, and describe the relationships between migraine and sociodemographic variables including sex, age, race, and rural versus urban residence. Logistic regression was used to isolate the independent effects of age, race, residence, and length of Medicaid eligibility on the presence or absence of migraine. The data consisted of adjudicated claims for 847,453 Georgia Medicaid recipients. Medicaid profiles for 678,079 recipients (468,448 female and 209,631 male) aged older than 4 years were analyzed as persons at risk of migraine. Migraine was identified in 6518 (1.39%) females and 991 (0.47%) males. Adjusting for eligibility, age, and race, the projected 22-month period prevalence for the United States was estimated as 3.83% (females) and 1.33% (males). Females, whites, and individuals residing in rural counties were more likely to suffer from migraine headache than their respective comparison groups. For both sexes, the peak prevalence was in the fourth and fifth decades of life. Migraine headache in the United States is estimated to afflict 4.5 million females and 1.4 million males. This prevalence is lower than previously reported and indicates that migraine headache may not be as common as previously believed. An alternative explanation is that many Medicaid recipients self-treat the condition, thus circumventing physician care and subsequent diagnosis and treatment.