Headache classification has evolved through clinical experience because there are no objective findings which define causation or headache type. The resulting groupings are inconsistently defined, and may not be optimal for the study or treatment of headache patients. To determine whether the same headache types would have resulted if today's sophisticated statistical methods had been applied to standard data collected on unselected patients, 21 symptoms collected from 726 patients with acute headaches were analyzed. Five 'natural' groupings or syndromes of symptoms were found in the data. Three were similar to traditional groupings: tension, migraine and cluster. Approximately one fourth of the patients had both tension and migraine symptoms, which suggests that the division of headaches into 'tension' vs 'migraine' has a large overlap. An 'OCULAR' and a 'URI' syndrome were also detected. Contrary to expectations, vascular headaches were not usually unilateral and patients with cluster headache symptoms were not usually males. The discrepancy between these findings and current perceptions may have been caused by differences in the populations or from biased patient selection in previous studies. © 1982.
Diehr, P., Wood, R., Wolcott, B., Slay, L., & Tompkins, R. K. (1982). On the relationships among headache symptoms. Journal of Chronic Diseases, 35(5), 321–331. https://doi.org/10.1016/0021-9681(82)90003-0