Features of migraine--comparison of a questionnaire study and a neurologist-examined random sample

In a community survey on headache in which a questionnaire was used, the responders were asked to state if their headaches were accompanied by symptoms usually associated with migraine. A random sample of 200 out of the 3,067 who returned the questionnaire were examined by a neurologist and the answers compared with those given in connection with the case history. Migraine features occurred sometimes in a very large percentage of the 2,018 who suffered form headache. When the answers were compared, "warning" and "nausea" occurring sometimes were reported significantly more often in the questionnaire. No difference was found when these features were reported as occurring usually or always. When interviewed by the neurologist, 20% of those with a clinical diagnosis of tension headache reported two of the three features "unilaterality", "warning" and "nausea", also. It is concluded that the questionnaire is a satisfactory tool if the aim is to select a population with distinct features of migraine, but the reliability rapidly decreases if efforts are made to include more elusive forms of the condition.