Abstract text (incl. figure legends and references)

The objective was to evaluate the symptoms of the migraine premonitory phase prospectively.

Methods: We used Migrebot headache diary database to select subjects with migraine features and headache frequency 3-8 days per month. Selected subjects proceeded to complete a specially designed version of the diary (ProdromaBot) to assess the characteristics of migraine attacks and interictal symptoms for at least 30 days. Participants completed 3 time points (TP) daily (9am,15pm,21pm). At each TP, participants answered 51 questions about potential triggers, overall wellbeing, premonitory symptoms, and presence of a headache and its characteristics.

Results: 98 subjects entered the study, 71 subjects completed at least 30 days period with at least 80% compliance. 59 subjects visited the clinic to confirm migraine. Patients completed not all TPs, so we selected only TP with new headache episodes (N=682) which were preceded or followed by the fully completed TP. Thus, we had 581 premonitory TPs and 640 postdrome TPs for further analysis. We analyzed the frequency of premonitory symptoms depending on the headache attack characteristics.

Phonophobia was the symptom which has the greatest number of premonitory symptoms (excess of light p<0,04, light sensitivity p<0,0001, excess of noise p<0,001, sound sensitivity p<0,0001, odor sensitivity p<0,047, hunger p<0,003, dehydration p<0,0001, feeling anxious p<0,0001 or depressed p<0,004,yawing p<0,0001, eye strain p<0,014,scalp allodynia p<0,001,unilateral lacrimation or nasal congestion p<0,007,frequent urination p<0,005). Allodynia was the most frequent premonitory symptom followed by the light sensitivity, feeling anxious, dehydration, unilateral lacrimation or nasal congestion and frequent urination.

Conclusions: This is the first study in which migraine premonitory symptoms were analyzed prospectively for at least 30 days 3 times daily. Premonitory symptoms vary significantly depending on the migraine attack characteristics.