In a cross-sectional study of young adults (21 to 30 years old), the last-year persistence rate of migraine was defined as the proportion of those meeting lifetime criteria for migraine who reported a migraine attack in the preceding year. In men the persistence rate was 48% and in women 79%, suggesting a higher rate of remission in young adulthood among men compared with women (8). In general, men had an earlier average age of onset than women, a finding that was confirmed in another American study using reconstructed cohort methods (20), and also in a Danish twin study (22). In this large twin study (age 33 to 41 years), it was reported that males with migraine without aura had a significantly younger mean age at onset than females with migraine without aura (16.5 years versus 21.5 years), whereas males and females with migraine with aura were of similar age at onset (22). About 20% of the twins had ceased having attacks of migraine with or without aura for at least the previous 3 years. Almost half of the males with onset of migraine without aura before age 15 had ceased having attacks, compared to one-fifth of the females (Table 46-1). This is in accordance with the 40-year follow-up study by Bille (6); however, in this study the gender difference in cessation of migraine disappeared later in life (age 47 to 53), because males had recurrence of their migraine.

Whitty and Hockaday (24) conducted follow-up of 92 migraine patients for periods of 15 to 20 years and found cessation in about one third of patients and improvement in more than two thirds of those with continuing attacks. Some patients showed a long-term picture of migraine in which attacks ceased completely in early adult life and recurred later. Even after migraine remits, in many patients the capacity to have migraine remains, as demonstrated by reactivation of the disorder after a headache-free period (5, 6, 7,24). No age factor (age itself, age at onset, time since onset) was of prognostic value (24). Also, Fry (10), by collecting information for 15 years from his general practice, found a tendency for the severity and frequency of migraine attacks to become less over the years and found cessation in 32% of males and 42% of females after 15 years. Thus, according to several cross-sectional studies, the condition seems to be self-limiting although lifelong among some.

Longitudinal prospective epidemiologic data with periodic contacts and examinations are necessary to draw valid conclusions about the prognosis and natural history of migraine. To date, few studies of representative migraine populations include follow-up evaluations. Bille (2, 3, 4, 5, 6) conducted follow-up studies of 73 children (aged 7 to 15 at study onset) 6, 16, 22, 30, and 40 years after an initial assessment. In the migraine group, 34% were free of migraine at the 6-year follow-up and 62% were free of migraine for at least 2 years at the 16-year follow-up. By the 22-year follow-up, about one-third of this group were having migraine regularly. Thus, only 40% of the migraine children were free of migraine at the age of 30 years (Fig. 46-1). In the 30-year follow-up, 53% had migraine and 47% were
migraine free. The 40-year follow-up showed that 29% of the original 73 migraine children had suffered from migraine the whole time since childhood. Another 22% still had migraine, although they had experienced one or more migraine-free periods of 2 years or longer. Thus, over a 40-year period, more than half the migraineurs continued to report migraine attacks (see Fig. 46-1) (6). Bille found an unfavorable prognosis in girls compared with boys and also in children with visual aura (3,5).