The need for standards in the design of differential fear conditioning and extinction experiments in youth: A systematic review and recommendations for research on anxiety

Abstract

Fear extinction studies in youth have yielded mixed results due to developmental processes and variations in design, methodology and dependent measures. This systematic review focused on studies with healthy youth between 2 and 17 years of age to identify experimental parameters of studies documenting extinction effects. Thirty-five studies met inclusion criteria and the following themes emerged (a) some studies employed parameters and task demands that are complex and require active participant involvement whereas others involved simple stimulus configurations and passive participant involvement, and (b) variation exists among dependent measures in units of measurement, timing and type of measurement. The review identified that studies using geometric shape conditioned stimuli (CS) paired with a tone unconditioned stimulus (US) (e.g., metal scraping on slate), as well as face CSs with a scream US produced the most reliable extinction effects, although the latter combination may be associated with higher drop-out than shape CSs and a tone US. The most commonly used and effective dependent measures for revealing extinction effects were skin conductance responses (SCR) and subjective ratings (SR) of CS valence, fearfulness and arousal. Fear potentiated startle (FPS) blink reflexes were also an effective but less commonly used measure. It is recommended that future studies use shape CSs and the metal scraping on slate US in studies involving children and either shape CSs and the metal scraping on slate US or face CSs paired with a scream US with adolescents. It is also recommended that US expectancy ratings and CS evaluations are assessed trial-by-trial and between-phase, and that startle-eliciting stimuli to measure startle blink reflexes are delivered on every second trial per CS so that SCR and FPS can be examined. However, further research is required to determine whether increased participant involvement due to providing trial-by-trial and between-phase ratings of the CSs and US differentially influences responding, particularly in children relative to adolescents and adults.