| Abstract: |
kground: Obesity affects both physical and mental health, and bariatric patients often show high levels of psychological distress. Cardiac adipose tissue, which includes epicardial and pericardial fat, is an active fat depot linked to inflammation and cardiovascular risk. Its relationship with anxiety has not been well studied, especially in bariatric candidates. Methods: This cross-sectional study included 29 adults undergoing preoperative evaluation for bariatric surgery. All participants completed the Hamilton Anxiety Rating Scale and underwent CT imaging to measure epicardial and pericardial adipose tissue thickness. Additional adiposity measures included BMI, waist circumference, abdominal wall thickness, and adipose tissue density. Correlations and simple linear regressions were used to examine associations between anxiety severity and adiposity markers. Group differences across obesity grades were assessed with one-way ANOVA. Results: Higher anxiety scores were strongly associated with greater pericardial fat thickness (r = 0.621, p < 0.001), epicardial fat thickness (r = 0.667, p < 0.001), BMI (r = 0.840, p < 0.001), waist circumference (r = 0.748, p < 0.001), and abdominal wall thickness (r = 0.494, p = 0.007). Both pericardial and epicardial fat thickness significantly predicted Hamilton total score in regression models. Conclusion: Anxiety severity in bariatric patients is closely related to several markers of adiposity, especially cardiac adipose tissue thickness. These findings suggest that cardiac adipose tissue may play a meaningful role in the psychological profile of individuals with severe obesity. Integrating both biological and psychological factors may improve the assessment and care of bariatric candidates. Artificial intelligence, especially deep learning techniques, is starting to play an increasingly important role in the assessment of epicardial and pericardial adipose tissue. It allows for automated segmentation and quantification based on CT images, providing high accuracy and reducing the time required for data processing. Recent artificial intelligence models, such as convolutional neural networks and U-Net architectures, have demonstrated significant agreement with manual measurements performed by specialists, which supports the possibility of their integration into routine clinical assessment. In the case of bariatric patients, these technologies can increase the accuracy of cardiac adipose tissue assessment and facilitate a broader understanding of the relationship between obesity, cardiovascular risk and the associated psychological impact. [ABSTRACT FROM AUTHOR] |