Headache classification based on questionnaire data: which symptoms are especially suitable?

In replication of two recent studies, it was intended to show that headache symptoms obtained by means of questionnaires fit a categorial model, provided that appropriate methods of data analysis are used. In addition, the questions which are best posed to obtain a succinct classification should be determined. Configural frequency analysis (CFA) was applied to 7 answers for headache symptoms in 2 samples (n = 602 and 606). In both samples classification became more succinct when the symptoms taken into consideration were reduced to 5. Questions for the quality of pain (pulsating vs not pulsating) did not supply much information nor did the question about aggravation during physical activity in Sample I and for photophobia in Sample II enhance the succintness of the classification. Based on 5 symptoms, however, namely: (1) pain occurring in attacks; (2) unilaterality; (3) visual prodromi; (4) nausea/vomiting and (5) photophobia in Sample I, aggravation during physical activity in Sample II, CFA clearly revealed a few headache syndromes. They could easily be interpreted as migraine with aura, migraine without aura, and tension headaches. Combinations of both migraine and tension headache symptoms did not occur more often and sometimes less often than expected by chance.