The goal of this review is to outline advances in the behavioural and neurobiological bases of fear learning that contribute to the optimizing of exposure therapy in fear and anxiety disorders. A paradigmatic shift from fear reduction throughout exposure as the primary index of successful outcomes to enhancement of inhibitory learning and regulation can be observed (the latter independent of fear reduction).The Emotional Processing Theory was seen as one of the primary underlying processes of exposure therapy. More recently, instead of weakening of the original fear memories, it is now believed that the establishment of new memories that effectively compete with the original fear memories is central to extinction. Patients who fail to achieve symptom relief from traditional exposure-based therapies might have deficits in inhibitory learning and regulation. Optimizing inhibitory learning during exposure therapy might compensate deficits that are present in these individuals.

Learning based models of exposure therapy: historical overview

Wolpe (1958) used counter-conditioning or reciprocal inhibition to reduce anxiety symptoms. In systematic desensitization, individuals progress through increasingly more anxiety induced by exposure, while relaxation exercises are used to compete with the anxiety and to reciprocally inhibit this. Level of fear elicited by exposure was critical to the counter-conditioning throughout treatment. Later on, graduated imaginal exposure was shown to be equally effective whether combined with relaxation training or not. Other studies showed that relaxation was working against providing a physiological response that was antagonistic to anxious arousal.

In addition, reciprocal inhibition was challenged by the fact that flooding therapy seemed to be as effective as graduated exposure. The effectiveness of flooding challenges the premise of reciprocal inhibition, that anxiety should be kept at a sufficiently low level through which counter conditioning can take place.

Habituation, a reduction in response strength with repeated stimulus presentations, provided a descriptive framework for responses during systematic desensitization and in vivo exposure therapy because fear often disappeared as people engaged more in exposure. Habituation models gradually replaced reciprocal inhibition. However, habituation could not explain long lasting fear reduction from exposure therapy.

The Emotional processing theory states that exposure therapy is effective because the fear structure is activated and integrated information that was incompatible with it, resulting in the development of a non-fear structure that replaces or competes with the original structure. Emotional processing theory guidelines are focused on the elevation of fear followed by within-and between-session reductions of fear. However, there is not much evidence for within-and between-session habituation.

Exposure therapy: outcomes

Exposure therapy is more effective than wait-list or attention placebo controls, however, almost half of the patients remain symptomatic. Therefore, there is a need to optimize exposure therapy.

Inhibition model of extinction

It is now thought that inhibitory learning is central to extinction, although additional mechanisms, such as habituation may play a role as well. According to the inhibitory learning models, the original CS-US association learned during fear conditioning is not erased during extinction, but is left intact and a secondary inhibitory learning about the CS-US is developed.

4 processes that may lead to resurgence of fear and anxiety following treatment

First, spontaneous recovery. Bouton and colleagues propose that after extinction, the CS possess two meanings, namely, its original excitatory meaning as well as an additional inhibitory meaning. Conditional fear shows often a spontaneous recovery. Simultaneously fear often returns when a previously feared stimulus is re-encountered after treatment has been completed. Therefore, patients with fear of flying who had successful treatment will probably report a return of fear if they do not continue to practice once the treatment is completed.

Second, renewal. Fear extinction effects appear to be specific to the context in which extinction occurred. The extinction context does not become a general inhibitor or safety signal, because non-extinguished stimuli retain their value when tested in the extinction context. Thus, renewal during clinical intervention is highly relevant in order not to limit the effects of the treatment to solely one context.

Third, reinstatement of conditional fear occurs if unsignaled or unpaired US presentations occur between extinction and retest. Reinstatement means that the effect depends on the fearfulness of the context in which the CS is tested. A clinical implication is that adverse events following exposure therapy may lead to a return of fear even when this event is unrelated to the feared stimulus.

Fourth, if the CS-US pairings are repeated following extinction, it is called rapid reacquisition of the CR. The clinical application is that fears that have subsided may be easily and rapidly reacquired with re-traumatization.

Neurobiology of fear extinction: evidence for inhibitory regulation

The amygdala plays a primary role in fear conditioning. During conditioning the amygdala is activated. Also the hippocampus is involved, it processes the contextual cues of conditioning. In addition, the insular cortex is involved in interoception and awareness of and sensitivity to visceral activity. Furthermore, the dorsal and rostral ACC appear to play a role in anticipation of the CS and US. The vmPFC mediates extinction. The PFC serves as the neurobiological basis for inhibitory learning. It is also suggested that the hippocampus creates an unique representation of the context in which extinction took place modulates extinction by providing information regarding safe versus dangerous contexts.

Deficits in Extinction Learning in anxiety disorders

Individuals with anxiety disorders respond stronger to both the CS+ and the CS- during conditioning and extinction. One explanation is that individuals with anxiety disorders have impaired inhibitory learning, they do not learn to view cues as safe. A potential mechanism of this impaired inhibitory learning include an attentional bias toward threat, difficulty disengaging from threatening stimuli, interpretation of ambiguous stimuli as threatening, elevated expectancies for threat, and overgeneralization of fear responding to stimuli that resemble the threat cue.

Elevated fear CR tot the CS- or the CS during extinction could also be due to over-excitation. Evidence for this is based on adapted paradigms that view inhibition apart from excitation during fear learning. This was found to be the case in combat veterans with PTSD symptoms. In a student sample, state anxiety induction prior to fear conditioning reduced subsequent inhibitory processing compared to individuals who did not undergo anxiety induction. At neural level there is limited evidence for decreased orbitofrontal and medial PFC during extinction and at extinction retests.

Enhancing inhibitory learning

One approach is to design exposures in such a way that the experience maximally violates the negative, excitatory expectancies regarding the rate with which aversive outcomes occur and the intensity of the outcomes. This should enhance the development of inhibitory expectancies. However, when tested, mixed findings were found. Exposure durations longer than the expected duration were equally effective as standard exposure therapy, although fewer trials were conducted.

Exposure therapy is designed in such a way that it provides experiences that disconfirm expectancies. Goals are adjusted during therapy to be violating expectancies all the time. Exposure to improve inhibitory learning is partly based on cognitive models of exposure.

Different methods to improve inhibitory learning will be discussed. (1) Super-extinction means that simultaneous presentation of multiple conditioned excitors throughout extinction training. Danger expectancies are stronger disconfirmed in this way than when only one conditioned excitor would have been used. In deepened extinction multiple fear stimuli are extinguished separately before being combined. From animal and human studies it is known that this form of extinction can lead to spontaneous recovery and reinstatement effects. Using extinction as a form of exposure, the goal is to teach the client that the feared event happens less often or is less bad than expected. In extinction this is accomplished by letting the client experience more than one ‘predictor’ of the feared events.

(2) If the task on ‘what has to be learned’ varies a lot, the retention of learned non-emotional stimuli will improve. A task could be a random practice task in which success is not always occurring. Variability induced by these kind of tasks increase storage capacity of the learned information and makes the information better retrievable later on. In variable exposure, exercises are not performed according to their hierarchy but after the easiest exercises the others will follow in a random order. Variability could be applied in used stimuli, duration and intensity. (3) Also variability in fear levels experienced during the exposure session is positive. Variability of fear levels throughout exercises improves generalization because the individual learns that he/she is able to cope with the fear in different contexts. (4) Removing of safety signals and safety behaviours is important because for some individuals the perceived fear is a safety signal. If fear is experienced, perceived safety is lower, it has become dependent on fear reduction and not on the actual threat of the situation. Also, when the safety signals are removed the fear will probably return. Some clients are afraid of feeling fear, for these individuals high fear levels should be kept during the exposure session.

Enhancing inhibitory Regulation

Several conducted studies have suggested that drug agonists of the d-cycloserine receptor can increase extinction in animal studies. D-cycloserine is thought to play a role in the consolidation of newly formed extinction memories. Other studies suggest that it is about preventing reinstatement and it is also shown that d-cycloserine may augment fear in some situations. Nonetheless, d-cycloserine has shown to lead to greater reduction in symptom severity and to lead to better maintained effects after treatment. D-cycloserine is primarily correlated with long-term outcomes that are not yet assessable during treatment.

Linguistic processing is an alternative method to improve inhibitory regulation. According to disruption theory, linguistic processing activates the PFC which leads to deactivation of the amygdala. Therefore, the anxiety response will be decreased. The underlying processes are not yet fully understood but are in this direction. Research conducted on images paired with word labels found that affective labelling of stimuli during the exposure session, reduced skin conductance and increased approach behaviour one week later in another context. Thus, linguistic processing can enhance inhibitory regulation during extinction or exposure. This in contrast to traditional cognitive therapy that tries to modify appraisals.

Weakening the fear memory

Recent research suggests that the retrieving of already stored memories induces a process of reconsolidation. The memory will be written again in long term memory, therefore new neurochemical processes are needed. Research suggests that propranolol (beta-blocker) might be effective in blocking this reconsolidation of memories and thereby also the fear that would have been stored again. Another method to weaken the fear memory itself is extinction during reconsolidation. Results are mixed but one recent study indicates that the findings might be related to enhancing the retrieval of the extinction learning instead of removing acquisition learning. It might be difficult to use this paradigm in clinical practice since factors such as time of presented stimuli are hard to control.

Enhancing retrieval of inhibitory learning

A possible way to enhance retrieval of extinction learning and offset context renewal is to include retrieval cues of the CS- no US association during extinction training in other contexts once extinction is over. Mentally reinstating an instructional retrieval cue was effective in anxiety patients. Clients had to wear a wrist band, as a cue, to remind them of what the learned in the therapy when they are experience fear in other contexts. Important is that these cues should not become safety signals. In context renewal it is problematic to determine which contexts are exactly relevant. Until now, studies on the use of different contexts to induce context renewal produced mixed results.