Comparison of factor structure models for the Beck a nxiety i nventory among cardiac rehabilitation patients

Abstract Objective Individuals with cardiovascular disease (CVD) experience greater rates of distress symptoms, such as anxiety and depressive symptoms, than the general population. These psychological outcomes have been linked to greater risk for negative outcomes following a cardiac event; however, research examining the relationship between specific components of anxiety and outcomes in CVD is limited. Further, prior research has not investigated the structure of anxiety symptoms in CVD. This study sought to compare previously established one, two, and four-factor models of the Beck Anxiety Inventory (BAI) in individuals enrolled in cardiac rehabilitation (CR). Methods Our sample included 208 individuals with CVD recruited during enrollment in a phase II CR program. Participants completed the BAI at enrollment in CR (Time 1) and again 12 weeks later at CR completion (Time 2, n = 151). Results Consistent with prior literature, 41% of our sample reported at least mild symptoms of anxiety (BAI > 8), and the BAI proved to be a reliable measure within this sample (α = 0 .89). Confirmatory factor analysis (CFA) results indicated that a second-order model with four first order factors, consisting of cognitive, autonomic, neuromotor, and panic components, fit our data well. A multi-group CFA approach supported measurement invariance across time. Conclusion These results suggest that anxiety following CVD can be evaluated based on cognitive, autonomic, neuromotor, and panic components as well as the encompassing anxiety construct.