HEADACHE EPIDEMIOLOGY: Emphasis on Migraine

DEFINITIONS OF PRIMARY HEADACHE

Many case definitions for migraine and TTH have been proposed.16, 22, 30, 35, 36 In defining migraine, Waters⁷¹ stressed three features: unilateral headache, a warning, and nausea or vomiting. He grouped prodromes and auras under the rubric of warning, treating them as a single phenomenon; however, this is unlikely to be true. Vahlquist,⁶⁸ and later Bille,³ defined migraine as paroxysmal headaches separated by headache-free intervals with two or more of the following features: nausea, scotomata

INCIDENCE OF PRIMARY HEADACHE DISORDERS

Studying the incidence of episodic disorders poses unique challenges.¹¹ For example, if five attacks are required to establish a diagnosis of migraine without aura, is the time of onset the first attack or the fifth? What if these attacks occur 5 years apart? What if people cannot report their total number of attacks accurately? These issues are problematic in population-based research; in clinical practice headaches usually occur with sufficient frequency to circumvent these problems. Similar

PREVALENCE

Estimates of migraine prevalence vary from less than 1% to more than 25% of the total population. Stewart et al⁶⁷ reviewed the 58 published population-based headache prevalence studies and identified 24 studies that met the inclusion criteria for a meta-analysis (Table 1). Only population-based studies with available information on age- and gender-specific migraine prevalence were included. Case status had to be determined by ascertaining migraine symptoms directly and then assigning a

AGE AND GENDER

Migraine prevalence varies with age and gender. The American Migraine Study revealed increasing prevalence in both male and female subjects from age 12 (when data collection started) until approximately age 40, after which time decreasing prevalence was noted.27, 52, 66 These dramatic age effects account for some of the variation in previous studies.⁶⁷ Rasmussen et al⁵¹ reported that migraine prevalence and age were not correlated. This may be an artifact based on their statistical methods,

INCOME AND SOCIOECONOMIC STATUS

Physician- and clinic-based studies suggest that migraine is associated with high intelligence and social class. Bille³ used school classes as a measure of intelligence and found that this was not true in school children. In adults, epidemiologic studies that used intelligence testing and occupation as measures of socioeconomic status revealed no evidence that connected migraine with social class or intelligence.⁷² In the American Migraine Study, migraine prevalence was found to be lower at

IS PREVALENCE INCREASING?

The Centers for Disease Control reported that migraine prevalence in the United States increased 60%, from 25.8/1000 to 41/1000 persons, between 1981 and 1989.³⁸ As the survey asked respondents if they have migraine headache, it missed the migraine sufferers who are not aware of their diagnosis. Because migraine is defined by self-reported diagnosis, improved recognition of migraine could produce this result. For example, if increased medical consultation produced increased rates of physician

PUBLIC HEALTH SIGNIFICANCE

Headache is a public health problem of enormous scope that has an impact on both the individual sufferer and on society. The American Migraine Study estimates that there are 23 million US residents with severe migraine headache.⁶⁶ Twenty-five percent of women experience four or more severe attacks per month; 35% experience one to three severe attacks per month; 40% experience one, or less than one, severe attack per month. Similar frequency patterns were observed for men.

In the American

DISABILITY FROM PRIMARY HEADACHE

When the number of lost work days due to headache is used as a measure of the societal impact of primary headache, the impact of TTH is much greater than that of migraine. Estimated annual lost work days per 1000 persons is 820 for TTH, almost triple that of migraine (270).²⁶ Migraine affects individual patients to a greater degree, however, as demonstrated by Rasmussen's finding. Forty-three percent of migraineurs missed work within a year of the clinical evaluation; in contrast, only 12% of

PROGNOSIS

Although cross-sectional data provide detailed information about prevalence, longitudinal epidemiologic data are relatively sparse. Bille² followed a cohort of children with severe migraine for up to 37 years. As young adults, 62% were migraine-free for more than 2 years, but after 30 years of age, only 40% continued to be migraine-free, suggesting that migraine is often a lifelong disorder. Hockaday²¹ reported similar long-term remissions. For 15 years, Fry¹⁷ collected information on migraine

SUMMARY

Through the use of IHS criteria, large population-based epidemiologic studies in Denmark, the United States, France, Canada, and elsewhere have shed light on the descriptive epidemiology of migraine. Although migraine is a remarkably common cause of temporary disability, many migraineurs, even those with disabling headache, have never consulted a physician for the problem. Prevalence is highest in women, in persons between the ages of 25 and 55, and, at least in the United States, in