Over the Edge: Extending the duration of a reconsolidation intervention for spider fear

Fear, while adaptive, can become maladaptive, contributing to anxiety disorders. Reactivating fear memories makes them vulnerable to reconsolidation disruption by amnesic agents like propranolol, leading to reduced fear behavior. Reconsolidation interventions offer a promising therapeutic approach, potentially reducing relapse rates. However, the success of these interventions is inconsistent, partly because the reconsolidation window is small and easily missed. Research on conditioned fears shows that the amount of prediction error (mismatch between expected and actual outcomes) during reactivation is crucial; too little leads to no change, too much to extinction learning, and a moderate amount to reconsolidation. In clinical practice, the learning history is unknown, making it challenging to control prediction error. This study aimed to find a behaviorally-controlled boundary condition for reconsolidation in naturalistic spider fear by manipulating reactivation procedure duration. Previous research successfully reduced spider fear using a ~2-minute tarantula exposure followed by propranolol. This study extended the exposure duration to investigate whether longer exposure would trigger the 'limbo' state, rendering the intervention ineffective. The hypothesis was that a brief exposure would trigger reconsolidation and subsequent fear reduction, while a longer exposure would not.

The reconsolidation of fear memories has been extensively studied using Pavlovian fear-conditioning paradigms. These paradigms involve associating a neutral stimulus (conditioned stimulus, CS) with an aversive experience (unconditioned stimulus, US), leading to fear of the CS. Reactivation of the CS, followed by an amnesic agent, can disrupt reconsolidation, reducing fear. Prediction error is a critical factor: the amount of unexpected information during reactivation determines whether reconsolidation, extinction, or the 'limbo' state occurs. In the laboratory, the amount of prediction error is manipulated by varying the number or duration of CS-alone presentations. However, translating these findings to clinical practice is challenging due to the unknown learning history of clinical fears. While research suggests that shorter reactivation procedures (seconds to minutes) trigger reconsolidation of weaker memories and longer procedures (hours) trigger extinction in exposure therapy, the exact duration for clinical fears remains unclear. This study leveraged the success of a previous reconsolidation intervention for spider fear to investigate the boundary conditions of reconsolidation in a naturalistic fear context.

This pre-registered study used a between-subjects design. Participants (N=43; 23 in the brief group, 20 in the extended group) with clinical or subclinical spider phobia were recruited. Inclusion criteria ensured sufficient fear of tarantulas and the suitability of the standardized reactivation procedure. Participants underwent either a brief (~3 min) or extended (~14 min) reactivation procedure involving a tarantula. During the procedure, the tarantula was repeatedly provoked to move. The longer procedure included more repetitions of the provocation. Following the reactivation, participants ingested 40mg of propranolol. Outcome measures included tarantula and house spider Behavioral Approach Tasks (BATs) (primary outcome was tarantula BAT), Spider Phobia Questionnaire (SPQ), and Fear of Spiders Questionnaire (FSQ), assessed before treatment (t0), two days after (t2), and at 3-day (t3), 3-month (t4), and 1-year (t5) follow-ups. Bayesian statistics were used for analysis, with a focus on Bayes Factors (BF) to assess evidence for or against hypotheses. Physiological measures (heart rate, blood pressure) were collected to assess propranolol's physiological effects. The study was pre-registered on the Open Science Framework.

Contrary to the hypothesis, no significant differences were found between the brief and extended reactivation groups on any outcome measure. Bayesian analyses revealed: 1) No significant difference in tarantula BAT scores between groups (BF01 = 4.80). 2) No significant difference in distress during the final tarantula BAT step (BF = 2.91). 3) Strong evidence that both groups showed improved house spider BAT scores after treatment (BF10 for the effect of time >1). 4) Strong evidence of a reduction in SUDs on the house-spider BAT final step from pre- to post-treatment in both groups (BFinclusion_Time = 7.73e7, BFExclusion_Group = 2.10, BFExclusion_Time*Group = 1.47). 5) Significant decrease in both SPQ and FSQ scores from pre- to one-year post-treatment, with no significant interaction effect for group (SPQ: BFinclusion_Time = ∞, BFExclusion_Condition*Time = 9.51; FSQ: BFinclusion_Time = 3.00e15, BFExclusion_Condition*Time = 16.28). Propranolol was shown to be physiologically active. All participants showed improvement on at least one outcome measure.

The unexpected finding of similar treatment effects across both exposure durations challenges the initial hypothesis. Several scenarios are considered: (A) Neither procedure triggered reconsolidation; (B) Reconsolidation was triggered in one group, and an alternative process caused a similar effect in the other; (C) Both procedures triggered reconsolidation. Scenario (A) is supported by the earlier reduction in self-reported fear compared to previous research, suggesting other mechanisms may be involved. The certainty of taking an active drug might have increased participant engagement, or the perceived effectiveness of the treatment might have influenced long-term results. Scenario (B) is plausible if reconsolidation occurred in the brief group, and other factors like increased self-efficacy or placebo effects contributed to the extended group's improvement. Scenario (C) is possible if both procedures triggered reconsolidation, possibly due to a wider reconsolidation window for strong, pre-existing fears than previously suggested by research on conditioned fears. The lack of a placebo group limits causal conclusions. This study highlights the limitations of directly translating findings from fear-conditioning research to clinical interventions for complex, naturally occurring fears.

This study unexpectedly found that varying the duration of a tarantula exposure during a reconsolidation intervention did not affect the outcome, with both groups showing significant and long-lasting reductions in spider fear. The lack of a placebo group and the absence of a clear marker for reconsolidation prevent definitive conclusions about the underlying mechanism. Future research should investigate alternative approaches to manipulate prediction error and include placebo controls. Nevertheless, the substantial fear reduction observed suggests the considerable potential of brief reconsolidation-based interventions for clinical anxiety disorders, warranting further investigation into optimal protocols and underlying mechanisms.

The lack of a placebo control group is a major limitation, making it difficult to isolate the specific effects of propranolol and the reactivation procedure. The strict inclusion criteria might have limited the generalizability of the findings. The study focuses on spider phobia, and findings may not generalize to other anxiety disorders. The standardized reactivation procedure might not have been equally suitable for all participants. The high exclusion rate due to various factors could have affected the results.