Anxious attachment improves and is predicted by anxiety sensitivity in internet-based, guided self-help cognitive behavioral treatment for panic disorder.

The purpose of this study was to examine whether anxious and avoidant attachment styles improve during guided internet-based cognitive behavioral treatment (ICBT) for panic disorder, and if so, to identify potential theoretically driven mechanisms related to the change. We examined changes in anxious and avoidant attachment and their time-lagged (1 week), longitudinal relationship with panic-related constructs in patients participating in ICBT (n = 79) in an open trial. Anxious attachment scores improved significantly with a medium effect during ICBT, d = 0.76 [0.45, 1.08]. According to benchmark analyses, changes were similar to the magnitude of change in face-to-face CBT and final scores to values of a nonclinical sample. Additionally, similar to findings in face-to-face CBT for panic disorder, longitudinal time analyses revealed that anxiety sensitivity scores predicted later improvement in anxious attachment scores, but not vice versa. Counter to our hypothesis, avoidant attachment did not significantly change during treatment, d = 0.15 [0.02, 0.46]; however, pretreatment level of avoidant attachment in ICBT was similar to the nonclinical sample. Also counter to our hypotheses, agoraphobic avoidant behaviors when alone did not predict changes in anxious attachment. These results suggest that anxious attachment can improve in ICBT for panic disorder even though the focus of the treatment is not on interpersonal relationships. These changes appear to follow changes in anxiety sensitivity. (PsycInfo Database Record (c) 2022 APA, all rights reserved).