The Course of Frequent Episodic Migraine in a Large Headache Clinic Population: A 12-year Retrospective Follow-up Study

Carl G. H. Dahlöf, MD, PhD; Maria Johansson, PhD; Susanne Casserstedt, PhD; Tina Motallebzadeh, PhD


Headache. 2009;49(8):1144-1152. 

In This Article


The results of this study suggest that in a substantial proportion of patients with frequent episodic migraine, the disorder eventually remits. These results are consistent with those from previous studies wherein 15–44% of adolescent subjects (at 3–18 years of follow-up) and 35–63% of adult subjects (at 10–20 years of follow-up) had become migraine-free.[11–15,17,18,20,23–25]

We found no gender differences in the liklihood of migraine remission: 30% of all men and 29% of all women studied no longer were experiencing migraine attacks at follow-up. In a French cohort study from 2005 which included 623 subjects with MwoA or MwA, the investigators similarly found no significant differences between the genders in regard to long-term prognosis, but the Danish study referred to earlier suggested a more favorable prognosis for men.[18,23] The latter study, however, involved only 53 subjects, of whom 11 were men, and the difference did not reach significance. In addition, this study only included subjects diagnosed with migraine with aura, which may make it difficult to compare the results of the 2 studies.

We found that past and persistent migraineurs differed in 2 areas: age and blood pressure. Mean ages at baseline were higher in past migraineurs than in those with persistent migraine; twice as many past migraineurs were over the age of 55 at baseline (23.6% vs 12.5%, P < .01). Average blood pressure also was significantly higher in past migraineurs; 34% of past migraineurs vs 20% of persistent migraineurs had a diastolic blood pressure over 90 mmHg at baseline (P = .01 for an association between migraine remission and a diastolic blood pressure over 90 mmHg). As past migraineurs were not interviewed in depth and thus not queried as to current medications, it is not possible to ascertain the prevalence of the use of antihypertensive medication in that group. Elevation of body mass index (BMI) did not influence the likelihood of remission, and mean BMIs for persistent and past migraineurs with a diastolic blood pressure over 90 mmHg were comparable.

One of our a priori hypotheses was that the younger the age of migraine onset, the longer the duration of disease at the time of follow-up, which would have increased the likelihood of becoming migraine-free. We instead found the opposite: with an older age of migraine onset, the proportion of patients free of migraine increased. One possible explanation for this finding is that individuals who begin to experience migraine early in life thus may become free of the disorder at an early age without ever presenting for evaluation at a specialized center. Because of this selection bias we thus could have missed those patients with a younger age of migraine onset who were destined to enjoy a more favorable prognosis.

No differences were detected regarding alcohol consumption. In men we found smoking to correlate with migraine remission; almost twice as many past migraineurs as persistent migraineurs were smokers at baseline. Persistent migraineurs seemed to have a smoking frequency comparable to that of the general population, whereas the baseline prevalence of smoking in males who became migraine-free was higher than that reported for the general population of Sweden.[33] Lacking information as to whether those who smoked at baseline continued to smoke over the follow-up period, the significance of this finding, if any, is unclear.

In women we found that the absence of a family history of migraine correlated with migraine remission. This result contrasts with that from an Italian study involving subjects with MwA, wherein a parental history of MwA correlated with migraine remission.[22] Another Italian study demonstrates the contrary and supports our results. On the other hand, Monastero et al found that the only factor associated with a 10-year persistence of migraine was a family history of the disorder.[17] In a Spanish study examining the natural history of migraine in children, no significant association was found between a family history of migraine and either persistence or remission of the disorder.[14]

We found no differences in our male patients related to the presence vs absence of aura. In women, however, aura was associated with migraine persistence. Most studies similar to ours that have included subjects diagnosed with MwoA or MwA but have involved no comparisons between these 2 groups.

Alcohol was the only migraine trigger which correlated with persistent vs resolved migraine in both males and females (albeit less robustly in females). To lack alcohol as an historical trigger was predictive of migraine remission. Cologno et al found absence of nutritional and visual trigger factors to be predictors of a favorable evolution.[22]

Differences at baseline in associated clinical features were found for men (as regards nausea and osmophobia) and for women (with worsening of pain consequent to physical activity). In all cases the associated feature was found to correlate with migraine persistence.

In the 264 subjects who continued to suffer from migraine attacks, there was evident a tendency for clinical improvement over time. Almost two-thirds reported a decrease in attack frequency, and over half reported milder pain (findings that may reflect the advances in clinical management that occurred over the follow-up period). Only 6 of the 374 subjects developed chronic migraine.

Our study has some methodological shortcomings. First, it was retrospective and thus susceptible to recall bias. In addition, we enrolled only patients who presented to a specialized clinic, and that group may not be representative of the migraine population at large. Differences between migraineurs who have vs have not sought treatment have been reported.[34] Finally, our statistical methodology did not include a control for multiple analyses, and the possibility that some correlations recorded were spurious (or the converse) cannot be discounted.

Our results suggest that the large majority of migraineurs have a favorable long-term prognosis. One caveat: how long a patient must have remained migraine-free in order to qualify as a "past migraineur" was not decided prospectively. When subjects were asked for how long they had been in remission, however, 91% reported that they had not experienced a migraine attack for 2 years or more. Unresolved is whether a remission of migraine of that length indicates final resolution of the disorder.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.