Abstract
BACKGROUND: Claudin-5 is an important cell adhesion molecule of tight junctions in brain endothelial cells and plays an important role in the permeability of the blood-brain barrier (BBB). Panic disorder (PD) is a disorder characterized by increased neuroinflammatory processes that may result in increased BBB permeability. This study aimed to examine claudin-5 levels in PD at baseline and after a 6-week follow-up and to compare them with those of the healthy control (HC) group. METHODS: Twenty-seven PD subjects (17 women and 10 men) and 25 HC subjects (15 women and 10 men) were included in this prospective cohort study. RESULTS: Pre-treatment claudin-5 (p = 0.021), C-reactive protein (CRP) (p < 0.001), CRP/albumin ratio (p < 0.001), and neutrophil count (p < 0.001) were higher in the PD group than in the HC group. Claudin-5 levels (p = 0.001) and Panic Disorder Scale (PDS) scores (p < 0.001) of the PD group decreased significantly after 6 weeks of follow-up compared to those at baseline. Post-treatment claudin-5 levels of the PD group were similar to those of the HC group (p = 0.230). In the PD group, partial correlation analysis was performed by controlling for the effects of age, gender, body mass index (BMI), smoking status, antidepressant and benzodiazepine status, and a significant relationship was found between the pre-treatment PDS score and pre-treatment claudin-5 level (r = 0.474, p = 0.030). After various modeling attempts, a hierarchical model (controlling for age, gender, and BMI) was created using pre-treatment claudin-5 level and the aggregate index of systemic inflammation (sensitivity = 70.4%, specificity = 76.0%; Nagelkerke R(2) = 0.468). The area under the receiver operating characteristic (ROC) curve of pre-treatment claudin-5 level for PD was 0.687 (p = 0.021). CONCLUSION: In subjects presenting with PD symptoms, increased parameters that may be associated with increased inflammation (such as CRP, CRP/albumin ratio, and neutrophil count), along with claudin-5 levels and the association between them, and the decrease in claudin-5 levels after a 6-week follow-up following a decrease in PD symptom severity, suggest that alterations in claudin-5 level in PD may be related to PD symptomatology. The high level of claudin-5 at the initial clinical presentation may be explained as a direct consequence of neuroinflammation in PD and/or as a compensatory change that occurs secondary to the neuroinflammation in PD. It can be assumed that as PD symptom severity decreases, neuroinflammation decreases, and claudin-5 production slows down.