Conditioned fear extinction and reinstatement in a human fear-potentiated startle paradigm

Abstract

The purpose of this study was to analyze fear extinction and reinstatement in humans using fear-potentiated startle. Participants were fear conditioned using a simple discrimination procedure with colored lights as the conditioned stimuli (CSs) and an airblast to the throat as the unconditioned stimulus (US). Participants were extinguished 24 h after fear conditioning. Upon presentation of unsignaled USs after extinction, participants displayed significant fear reinstatement. In summary, these procedures produced robust fear-potentiated startle, significant CS+/CS- discrimination, within-session extinction, and significant reinstatement. This is the first demonstration of fear extinction and reinstatement in humans using startle measures.

Figure 1

A summary of the acquisition phase. (A) The partial schedule of reinforcement used during the Acquisition phase produced robust potentiation of the acoustic startle reflex during A (CS+) trials and significant discrimination between the CS+ (light A) and CS− (light B). (NA) Noise alone. (B) Subjects’ responses on the keypad indicated successful fear conditioning during the Acquisition phase. Expectancy ratings were scored as follows: DANGER = 1, UNCERTAIN = 0, and SAFETY = −1. Error bars, SEM.

Figure 2

(A) Subjects exhibited a within-session decrement in startle magnitude across trial types during the Extinction phase (Blocks 1–5). (B) In addition, subjects displayed significant within-session extinction of fear-potentiated startle as measured by Difference Score. There was an increase in potentiated startle to light B at the outset of the Extinction phase, yet subject startle responses during the second block of Extinction demonstrate significant retention of the Acquisition contingency. Difference Score = [mean startle magnitude in response to light A or B] − [mean startle magnitude to startle probe alone (NA)]. Error bars, SEM. (C) Subject expectancy ratings indicated significant retention of the Acquisition experimental contingency (first trial). In addition, subject ratings of light A as DANGER were significantly extinguished during Extinction. Expectancy ratings were scored as follows: DANGER = 1, UNCERTAIN = 0, and SAFETY = −1. Error bars, SEM.

Figure 3

Dissociation between startle measures and US expectancy during reinstatement testing phase. For startle measures, reinstatement was assessed in those subjects displaying an extinction decrement of greater than or equal to 50% (Extinguishers, A and B). The presentation of three unsignaled airblasts after Extinction elicited a significant reinstatement of fear potentiated startle (B). Error bars, SEM. For US Expectancy ratings, all subjects were included in the Reinstatement analyses. Expectancy ratings from all subjects in the (C) Control group (no unsignaled USs) did not change between Trials 20 and 21. Expectancy ratings from all subjects in the (D) Reinstatement group increased significantly following three unsignaled USs between Trials 20 and 21. Expectancy ratings were scored as follows: DANGER = 1, UNCERTAIN = 0, and SAFETY = −1. Error bars, SEM.