Evaluating the incidence of migraine requires longitudinal follow-up studies. There is a general lack of population-based data on the incidence of migraine. However, two epidemiologic studies have assessed retrospectively (based on the reported age of migraine onset) the incidence of migraine in the United States and Denmark. Stewart et al. conducted a population-based telephone survey of over 10,000 people 12 to 29 years of age in Washington County, Maryland (7). Among the 1,410 respondents defined as migraineurs, the onset of migraine peaked in childhood and adolescence, then declined over time. Incidence peaked earlier for boys than for girls, a result that is confirmed in another American study (8). The overall incidence rates for migraine were 601 per 100,000 female person-years and 222 per 100,000 male person-years.

In a study of a representative sample of 25- to 64-year-old Danes with clinically confirmed migraine meeting IHS diagnostic criteria, the age-adjusted annual incidence of migraine was estimated to be 370 per 100,000 person-years—580 per 100,000 person-years in women and 160 per 100,000 person-years in men (9). Incidence rates for subjects younger than 30 years of age were comparable in these studies. Estimates of migraine incidence based on cross-sectional data on recalled age at onset suffer from the drawback of recall errors and ignores possible changes in the risk over time. Other studies have estimated the incidence of medically recognized migraine cases based on retrospective inspection of medical records (10, 11, 12). These data suggest that migraine incidence increased substantially from 1979 to 1990. However, these studies, which were based only on patients who had consulted physicians for migraine, do not elucidate the real incidence of migraine in the population, but rather illustrate the change in referral patterns and verified improved medical recording over time. Thus, American studies have demonstrated that migraine consultation and diagnosis actually has increased over the last decade (3,13).

The first longitudinal population-based study applying the IHS diagnostic criteria by clinical interviews has recently reported on the incidence of migraine in Danish adults (aged 25 to 64 years). The overall annual incidence rates for migraine were 8.1 per 1,000 person-years, being 2.6 per 1,000 in males and 16.0 per 1,000 in females. Incidence peaked in 25 to 34 year olds in both genders (6.5 males, 22.8 females) and decreased markedly with age (14). Breslau et al. (15) estimated migraine incidence among members of a Health Maintenance Organization, aged 21 to 30 years at baseline. Follow-up interviews were completed after 3.5 and 5.5 years. The annual incidence rate was 17 per 1,000 person-years (6.5 males, 23.6 females), which are in agreement with the Danish follow-up study (14) (Table 25-1).

The results of several recent prevalence studies from the industrialized part of the world are shown in Table 25-2. It is remarkable that most population-based studies using the operational diagnostic criteria of the IHS (4) have reported congruent prevalence figures for migraine in adults (see Table 25-2).

In Denmark, we conducted in 1989 the first epidemiologic study using the operational diagnostic criteria of the IHS. The study sampled randomly from the National Central Person Registry 1,000 men and women 25 to 64 years of age and residing in Copenhagen. Headaches were classified according to a physician-conducted diagnostic interview and a neurologic examination. The participation rate was 76%, and information about a further 20% was obtained by telephone interview (25). The survey showed, in agreement with another Danish epidemiologic study (26), a lifetime prevalence of migraine of 16 to 8% in men and 25% in women, a male:female ratio of 1:3. The 1-year period prevalence of migraine was 10 to 6% in men and 15% in women (25). The 1-year period prevalence of migraine without aura (previously called common migraine) was 6%, with a male:female ratio of 1:5. The prevalence of migraine with aura (previously called classic migraine) was 4%, with a male:female ratio of 2:3 (30,31). Gender-related factors seem to be of even more importance in migraine without aura than in migraine with aura.

In a U.S. migraine prevalence study in 1989, questionnaires were mailed to 15,000 households selected to be representative of the U.S. population. A questionnaire was completed by each household member with severe headache. The overall response rate was 63.4%, yielding 20,468 respondents 12 to 80 years of age. The basis for migraine diagnosis differed from the IHS criteria in that the lifetime number of previous migraine attacks and headache duration were not considered. The self-administered questionnaires using modified IHS criteria yielded a 1-year period prevalence of migraine of 6% in males and 18% in females (28). These results are consistent with another large U.S. population study (29). In France, Henry et al. conducted a national migraine survey in 1990 using IHS criteria (19). To study a representative sample of the French population with regard to age, gender, occupational category, and place of residence, a quota sampling survey was performed by a national public opinion poll agency in which subjects were asked specific questions about their headaches and associated symptoms. The study included 4,204 subjects aged 15 or older. Using a diagnostic algorithm based on the IHS criteria (including borderline migraine cases), the prevalence (previous few years) was 12; 6% among men and 18% among women. These results are similar to those of the Danish studies, the American studies, and another French study (32). Recently, the three prevalence studies of migraine from Denmark, the United States, and France have been repeated 10 to 12 years after the first survey (21,33,34). The aim of these replicate studies has been to describe if prevalence, sociodemographics, or burden have changed over the last decade of many new treatments and improved awareness of migraine. Historically, estimates of migraine prevalence have varied, in large part because of differences in study methods, migraine case definitions, and the demographic characteristics of the study population. A meta-analysis of 24 studies (35) showed that about 70% of the variance in migraine prevalence estimates was explained by differences in case definition and sociodemographic profiles (age and gender) among study populations. The methodologic similarity, uniform case definition, demographic comprehensiveness, and national scope of these replicate studies hold these factors constant. In the very recent replicate study among a young adult Danish population (25 to 36 years old), a widely unchanged migraine prevalence over a 12-year period is reported, whereas the prevalence of frequent migraine (migraine for more that 14 days a year) showed a significant increase (33). The two methodologically identical national surveys in the United States conducted 10 years apart showed that the prevalence and distribution of migraine have remained stable over the decade (21,28). Two other U.S. surveys (3,29) and the migraine repeat study from France confirms unchanged prevalence of migraine over the past 10 years (19,34) (Fig. 25-1). However, a trend toward nonsignificant increasing migraine prevalence was seen, especially in the Danish and French studies. In the French study it is explained mainly by varying definitions of migrainous disorder in the first and second studies.

In summary, mutually confirmed prevalence data from recent population-based studies in Denmark (25,26,33), France (19,34), as well as a number of other studies in Western Europe (16,18,20,22,27), and in the United States (3,8,21,28,29) all using the IHS diagnostic criteria, have shown a homogeneous picture regarding migraine prevalence. A reasonable estimate of the 1-year prevalence of migraine in adults is 10 to 12%—6% among men and 15 to 18% among women.