Fig. 1From: Identifying menstrual migraine– improving the diagnostic criteria using a statistical methodSimple Markov chain model of the progression of migraine attacks. The transition probabilities μ and δ represent the probability of onset of a migraine attack – conditioned on being susceptible – and the probability that an ongoing attack will continue, respectively. We assume that each patient may have individual transition probabilities. The non-clustering assumption excludes μ≠δ in general. MM – interpreted within this simple model – postulates that women with MM deviate from the base-model above, and instead have two distinct migraine probabilities: μNM which represents her probability of migraine attack-onset outside her menstrual window, and μM which represents an elevated migraine attack-onset probability during their menstrual window. That is, MM is present when μM>μNMBack to article page