Rationale
Selective serotonin reuptake inhibitors (SSRIs) are the first choice of treatment for anxiety-like disorders. However, which aspects of anxiety are affected by SSRIs is not yet fully understood.
Objective
We aimed to systematically review the effect of six clinically effective SSRIs on four aspects of unconditioned anxiety: approach-avoidance behaviour (elevated plus maze), repetitive behaviour (marble burying), distress behaviour (ultrasonic vocalization), and activation of the autonomous nervous system (stress-induced hyperthermia).
Methods
We identified publications by searching Medline and Embase databases and assessed the risk of bias. A random effects meta-analysis was performed and moderator effects were analysed with Bayesian penalized meta-regression.
Results
Our search yielded 105 elevated plus maze, 63 marble burying, 11 ultrasonic vocalization, and 7 stress-induced hyperthermia articles. Meta-analysis suggested that SSRIs reduce anxiety-like behaviour in the elevated plus maze, marble burying and ultrasonic vocalization test and that effects are moderated by pre-existing stress conditions (elevated plus maze) and dose dependency (marble burying) but not by duration of treatment or type of SSRI. The reporting quality was low, publication bias was likely, and heterogeneity was high.
Conclusion
SSRIs seem to reduce a broad range of unconditioned anxiety-associated behaviours. These results should be interpreted with caution due to a high risk of bias, likely occurrence of publication bias, substantial heterogeneity and limited moderator data availability. Our review demonstrates the importance of including bias assessments when interpreting meta-analysis results. We further recommend improving the reporting quality, the conduct of animal research, and the publication of all results regardless of significance.