In a further analysis, the dependent variable was the presence of migraine type (MwA = 1, MwoA = 0). Independent variables CVR to l-arginine in the MCA, CVR to l-arginine in the PCA and FMD were transformed into attributive

variables. In step one, we evaluated a possible association of CVR to l-arginine in the MCA and the PCA with migraine, and also of CVR to l-arginine in the MCA and the PCA with MwA and MwoA. We found a significant negative association between CVR to l-arginine Gefitinib in vitro in the PCA and migraine (p = 0.01), but not between CVR to l-arginine in the MCA and migraine (p = 0.44). The results are summarized in Table 1. Similarly, we found a significant negative association between CVR to l-arginine in the PCA on MwA (p = 0.01) and between CVR to l-arginine in the PCA and MwoA (p = 0.02). Again we did not find any association between CVR to l-arginine in the MCA and MwA (p = 0.39) and also between CVR to l-arginine and MwoA (p = 0.47). The results are summarized in Table 2. In step two, we evaluated a possible association of FMD with NU7441 purchase migraine, and repeated the procedure separately with MwA and MwoA. The results are represented in Table 3. The binary logistic regression did not show any association between FMD and migraine (p = 0.96) and also between FMD and MwA (p = 0.99) and MwoA (p = 0.99). The main original finding of our post hoc study is that we have found a significant negative association

between CVR to l-arginine in the posterior cerebral circulation and migraine, and no association

between CVR SB-3CT to l-arginine in the anterior cerebral circulation and migraine. In recent years it has been proposed that migraine affects not only systemic but also cerebral circulation [1]. Namely, ischemic stroke can occur between or during migraine attacks, particularly in MwA and young women [21], [22], [23] and [24]. The territory of the posterior cerebral artery is preferentially affected [25]. In addition to clinical strokes, focal ischemic and hyperintensive, ischemic-like lesions have been found in the territory of the posterior cerebral circulation [22], [26] and [27]. In our previous study we showed a lower CVR to l-arginine in the PCA and normal CVR to l-arginine in the MCA in migraine patients without comorbidities compared to healthy subjects [9]. In such circumstances this could be applied to cerebral endothelial dysfunction localized only in the territory of the posterior cerebral circulation. However, a confirmation from another point of view was still missing. For this purpose we analyzed the association between migraine and parameters of systemic, as well as cerebral endothelial function. The findings of this study have shown that impaired posterior cerebral endothelial function could be associated with migraine, while intact anterior cerebral endothelial function could not be only associated with migraine but it could be also attributed to physiological conditions.