A history of avoidance does not impact extinction learning in male rats

Abstract

Pervasive avoidance is one of the central symptoms of all anxiety-related disorders. In treatment, avoidance behaviors are typically discouraged because they are assumed to maintain anxiety. Yet, it is not clear if engaging in avoidance is always detrimental. In this study, we used a platform-mediated avoidance task to investigate the influence of avoidance history on extinction learning in male rats. Our results show that having the opportunity to avoid during fear acquisition training does not significantly influence the extinction of auditory-cued fear in rats subjected to this platform-mediated avoidance procedure, which constitutes a realistic approach/avoidance conflict. This holds true irrespective of whether or not avoidance was possible during the extinction phase. This suggests that imposing a realistic cost on avoidance behavior prevents the adverse effects that avoidance has been claimed to have on extinction. However, avoidance does not appear to have clear positive effects on extinction learning nor on retention either.

Fig. 1. Graphic representation of Experiments 1–3.

a Graphic representation of the different phases of Experiment 1. b Graphic representation of the different phases of Experiment 2. c Graphic representation of the different phases of Experiment 3.

Fig. 2. Avoidance training results for Experiments 1–3.

The box plots represent the average of the first 3 CSs on each day. The bold lines in the trial-by-trial plots represent the mean and the surrounding shaded area the standard error of the mean. Results are expressed in % of time during CS presentations. a Graphic representation of the avoidance training sessions in the three experiments. b In the Avoider group, avoidance increased significantly from day 1 to 2 (V = 1, p < 0.001, r = 0.861). Across both groups, freezing increased significantly from day 1 to 2 (F(1, 22) = 59.80, p < 0.001, η_p² = 0.731). c In the Avoider group, avoidance increased significantly from day 1 to 2 (V = 1, p < 0.001, r = 0.861). Across both groups, freezing increased significantly from day 1 to 2 (Q(1, 10.70) = 23.33, p < 0.001, η_p² = 0.477). d In the Avoider group, avoidance increased significantly from day 1 to 2 (t(11) = −2.74, p = 0.019, d = −0.791). Freezing differed significantly between Avoiders and Yoked animals (F(1, 22) = 6.66, p = 0.017, η_p² = 0.232). *p < 0.05, ***p < 0.001.

Fig. 3. Extinction results for Experiments 1–3.

The box plots represent the average of the first 3 CSs on each day. The bold lines in the trial-by-trial plots represent the mean and the surrounding shaded area the standard error of the mean. Results are expressed in % of time during CS presentations. a Graphic representation of the extinction training sessions. b Across both groups, freezing decreased significantly over the extinction sessions (Q(3, 11.33) = 87.44, p < 0.001, η_p² = 0.764). A further investigation of the interaction between group and extinction session (Q(3, 10.61) = 7.14, p = 0.007, η_p² = 0.13) revealed that on day 1 the Avoider group suppressed lever pressing more than the Yoked group (V = 135, p < 0.001), whereas on day 2 the Yoked group shower more suppression of lever pressing than the Avoider group (V = 32.5, p = 0.024). c In the Avoider group, avoidance decreased significantly over the extinction sessions (χ²(3) = 28, p < 0.001, W = 0.779). Freezing decreased significantly across both groups (Q(3, 9.6) = 30.56, p < 0.001, η_p² = 0.694), as did lever pressing (Q(3, 10.98) = 258.58, p < 0.001, η_p² = 0.77). d In the Avoider group, avoidance decreased significantly from day 1 to 2 (t(10) = 5.9, p < 0.001, d = 1.78). An interaction between group and test day (Q(1, 11.76) = 8.55, p = 0.013) revealed that freezing was significantly decreased in the Avoider group (V = 52, p = 0.014, r = 0.476), but not the Yoked group. *p < 0.05, **p < 0.01, ***p < 0.001.

Fig. 4. Reinstatement test results for Experiment 2 and 3.

The box plots represent the average of the first 3 CSs on each day. The bold lines in the trial-by-trial plots represent the mean and the surrounding shaded area the standard error of the mean. Results are expressed in % of time during CS presentations. a Graphical representation of reinstatement and the reinstatement test sessions in Experiment 2 and 3. b Suppression of lever pressing differed significantly between the Avoider and the Yoked group (Q(1, 13.94) = 5.58, p = 0.042, η_p² = 0.175). c Freezing decreased significantly between last day of extinction training and reinstatement test (Q(1, 11.33) = 6.72, p = 0.024, η_p² = 0.286), as did suppression of lever pressing (F(1, 21) = 5.31, p = 0.031, η_p² = 0.202). *p < 0.05.

Fig. 5. Spontaneous recovery test results for Experiment 3.

The box plots represent the average of the first 3 CSs on each day. The bold lines in the trial-by-trial plots represent the mean and the surrounding shaded area the standard error of the mean. Results are expressed in % of time during CS presentations. a Graphical representation of the last extinction session and the spontaneous recovery test session. b Freezing increased significantly from the last extinction session to the spontaneous recovery test session (Q(1, 9.72) = 25.61, p < 0.001, η_p² = 0.600). Similarly, suppression of lever pressing increased significantly between the last extinction session and the spontaneous recovery test session (Q(1, 10.65) = 10.99, p = 0.007, η_p² = 0.463). **p < 0.01, ***p < 0.001.