Understanding the untreated course of anxiety disorders in treatment-seeking samples: A systematic review and meta-analysis

Abstract

Background: Very little is known about the course of anxiety disorders when they go untreated, despite the significant theoretical and practical value of this information, such as for treatment planning and benchmarking purposes. This meta-analysis aimed to examine the course of anxiety disorders in treatment-seeking samples using the control groups of treatment studies for anxiety disorders.

Methods: Following pre-registration, we systematically searched the literature for RCTs of treatment for anxiety disorders. Studies were included if they randomised participants to a control arm, where treatment was not received (i.e. waitlist control or no-treatment control). Meta-analyses were conducted to determine the magnitude of symptom change over the control period (Hedges' g), and rate of response (pooled prevalence). Effects were compared between anxiety disorders, alongside other potential moderators.

Results: Following search and screening, 173 RCTs met criteria (n = 15,250) for data extraction. Overall, untreated participants demonstrated significant, but small improvements to anxiety symptoms (g = 0.17, 95% CI 0.14, 0.21). Significant differences were observed between anxiety disorders, and according to other methodological features of the included trials.

Conclusions: Results suggest that anxiety disorders are unlikely to remit without treatment, with some disorders remitting to a lesser extent than others. While this review is limited to a treatment-seeking sample, results provide theoretical and practical value for researchers and treatment providers.

Keywords: Anxiety disorders; Meta-analysis; Natural remission; Symptom change.