The platelet theory of migraine causation predicts that drugs inhibiting platelet activation will be effective in migraine prevention, but the literature indicates that this is only partly the case. Conversely, therapy achieving clinical benefit should be associated with reduced platelet activity. To test this concept, the β-adrenergic blockers propranolol (non-selective), metoprolol (β1-selective) and Li-32468 (β2-selective) were used in migraine therapy with assessments of platelet aggregation and release, plasma thromboxane A2 (measured as thromboxane B2) and clinical response.

Between propranolol and Li-32468, there was lack of correlation of clinical with platelet effects. Propranolol and metoprolol, whose established efficacy in migraine prophylaxis was reflected here, actually had opposite effects on platelet activity, which was increased with the former and inhibited by the latter. Yet both drugs gave elevated thromboxane B2 levels.

In view of this complete dissociation between drug effects on platelets of migraineurs and symptoms of migraine, the platelet theory of migraine causation is untenable.