Abstract
Selective serotonin reuptake inhibitors (SSRIs) are the first choice of treatment for anxiety-like disorders such as panic disorder, generalized anxiety disorder and post-traumatic stress disorder. Fear learning plays an important role in the etiopathology of these disorders. However, it is unclear if and which fear learning processes are affected by
... read more
SSRIs. This systematic review investigated the effect of six clinically effective SSRIs on the fear learning processes acquisition, expression and extinction. Since SSRIs have been shown to effectively treat anxiety-like disorders, the results of this systematic review could provide insight into which fear learning processes are important to include in future research regarding the development and treatment of anxiety-like disorders. A systematic search in the Medline and Embase databases yielded 128 articles that met the inclusion criteria. Of these articles 120 were eligible for the meta-analysis. Data regarding the study subjects, intervention, experimental design and size and direction of the effects were extracted. Meta-analysis was conducted in R, the R-package metafor [1] was used to estimate the overall effect size, using a random-effects model. Five categorical predefined moderators were coded to account for between-studies heterogeneity (type of SSRI, duration of treatment, disease induction, species, type of test). The effect of these moderators was analysed with a Bayesian penalized meta-regression (BRMA) which is a new method. This analysis was carried out with the pema R-package [2]. The meta-analysis showed that SSRIs significantly reduced contextual fear expression and facilitated extinction learning to cue. Bayesian penalized meta-regression further suggested that chronic treatment with SSRIs is associated with stronger anxiolytics effects on cued fear expression than acute treatment. Other variables, including type of SSRI, species, disease induction and type of test, did not seem to moderate the effect of SSRIs This systematic review suggests that the clinical efficacy of SSRIs may be specifically related to their effects on fear expression and extinction, rather than fear acquisition. It could be that the effects of SSRIs on these fear processes are due to general inhibition of fear-related emotions. Therefore, it would be interesting to investigate how SSRIs affect other forms of anxiety, such as unconditioned fear responses. In addition, studies aimed at explaining the sources of the high levels of heterogeneity observed in our meta-analyses could help to optimise the experimental set-up to further investigate the mechanisms underlying fear learning. In order to gain more insight in how the effects of SSRIs on these processes contribute to the anxiolytic effects seen in the clinic, it would be valuable to conduct studies that use experimental designs that allow us to selectively evaluate the effects of SSRIs on fear extinction. This experimental data is interesting to obtain since various exposure therapies in patients are based on promoting extinction. Furthermore, we want to encourage fellow researchers to consider using BRMA when working with a dataset with small sample sizes containing high levels of multicollinearity to avoid overestimation of effects.
show less
