Migraine incidence in 5 years: a population-based prospective longitudinal study in Turkey

The incidence of migraine has been investigated only in a few studies worldwide and it is not known in our country. We, therefore, aimed to estimate the migraine incidence in a previously accomplished population-based prevalence study sample of 5323 individuals in the year 2008.

The former Turkish headache prevalence study has been completed as a nationwide, randomized, home-based study of face-to-face examination by physicians trained for headache diagnosis by using ICHD criteria. Five years after this study an optimized survey including 50 questions was performed to estimate the migraine incidence in migraine-free individuals in the previous study, with a 56.4 % responder rate. Two validation studies for this survey were performed prior and after the study each in 100 subjects by comparing the gold standard of expert diagnosis of headache, showing high rate of reliability (Crohnbach alpha: 0.911 and 0.706, respectively).

Migraine incidence was estimated as 2.38 % (2.98 % in women and 1.93 % in men) per year in 2563 migraine-free individuals; if the population at risk is defined as the group without any headaches, the migraine incidence decreased to 1.99 %. The chronic migraine (CM) incidence [without medication overuse (MOH)] was 0.066 % and that of MOH was 0.259 %. We found a significant burden of the disease on the occupational functionality as well as on social and family life, even in the early years of the migraine. The family history of headaches especially in the fathers could be useful to predict new cases of migraine, besides the well-known risk factor, diagnosis of depression, whereas income and education did not seem to relate to migraine onset.

Our study with a large population-based nation-wide sample, using ICHD-II criteria, with structured headache interviews as well as blinded re-validation of the questionnaire diagnoses showed a 2.38 % incidence rate of migraine in Turkey, higher than most of the other previous reports; a finding which could be related to genetic factors and also to the methodological differences in the study designs. Moreover the incidence of CM was found to be 0.066 %.