The Benefits of Being Wrong: Using Positive Prediction Error to Treat PTSD

    Positive prediction error is a term used to describe a situation where an organism that has previously associated a certain context with a negative outcome now experiences that same context with a positive outcome. By providing the organism with a prediction error (experiencing something in the present they were not expecting), that prior associated memory is now subject to reconsolidation. This means the memory is malleable and allows for a window of opportunity to update that memory.
    The idea of positive prediction error is especially important in the context of post-traumatic stress disorder (PTSD). Dr. Stephanie Grella was able to update fear memories and disrupt their reconsolidation in mice models. She used optogenetics to artificially activate a previously experienced positive memory once the mice were placed in a box where they had previously experienced an electric shock. This resulted in a diminished fear response in the mice. One important discovery is how the original fear memory—not just the emotional aspect of the memory—was altered.
    One common question, though, is how this research could be applicable to humans. While optogenetics is not something that can be done in humans, there are other potential routes that can be taken to produce a positive prediction error in humans. One such technique—the “flash technique” (FT)— is currently being used. Sik-Lam Wong writes about the use of this technique and its connections to eye movement desensitization and reprocessing (EMDR) therapy. EMDR works by therapists asking clients to think of unpleasant or discomforting memories (usually traumatic) while also having the clients watch a light flash back and forth (bilateral stimulation). This dual-attention focus keeps the client in the present moment without fully “flashing back” to their traumatic memory. In doing so, the traumatic memory is remembered without the expected stress-induced behaviors (like a panic attack). The combination of being in a safe environment and a client's working memory being stretched between two tasks allows for a decrease in the vividness and intensity of the recalled memory. Over many sessions, this can lead to a decrease in their PTSD symptoms in the long run.
    There are cases, though, in which the recalled memory is too intense for traditional EMDR to work because the client will almost instantly enter a dissociative state. This is where, as Wong notes, the flash technique can be beneficial. Here, a client identifies the target memory to be processed and then focuses on a positive experience. A flash then prompts the client to recall the memory and instantly return to the positive experience. Unlike EMDR, clients are not focusing on the traumatic memories or their feelings towards said memories. There has been clinical efficacy in groups of clients who experienced an earthquake and a group of healthcare providers impacted by the stress of COVID-19. The positive prediction error introduced by Dr. Grella in her talk can be seen in both EMDR and FT as a way to treat those with stress-related disturbances in memory.
    The proposed neuroscience for how these therapies works is also fascinating. Two main structures of interest are the prefrontal cortex (PFC) and the amygdala. The amygdala is responsible for eliciting a fear response, and the PFC—specifically the vmPFC—plays an integral role in fear extinction. Models of PTSD propose that there is a disconnect between these two structures, which is why those with PTSD frequently relive traumatic moments and generalize the fear onto other stimuli. As a result of this imbalance, driven by high concentrations of the neurotransmitter norepinephrine, traumatized individuals stay in a bottom-up mode and respond intensely to innocuous stimuli. One proposed mechanism for how EMDR and FT works is by allowing the PFC to “stay online.” This eventually elicits top-down regulation in clients whose now active PFC can shut down fear responses to harmless stimuli.
    Secondly, the use of working memory is also important. Working memory is spread across cortices and is especially prevalent in the PFC during goal-orientated tasks. By directing clients with a goal (follow the flashing light in EMDR), the PFC remains online instead of being shut off during the recollection of their traumatic memories. Additionally, working memory has the ability to designate more attention to some tasks than others leading to different states of activation for each item being held. Salient items—which are ones that hold emotional significance—are by default given more attention. It is thought that FT is able to bypass this automatic process (and circumvent the almost instant dissociation) through the client’s brief thought of the memory without any details of it. For those having trouble with this, some therapists will have the client visualize putting the memory into a book. Upon recall, the therapist will instruct the client to look at the metaphorical book instead of thinking of the memory itself. It is also suggested that this is where the positive prediction error occurs: the PFC is no longer being shut off by an overactive amygdala despite the traumatic memory being held in working memory. The positive aspect comes from the co-occurance of a positive memory (in the client’s working memory) in a safe environment (the therapist’s office). This unexpected and new experience provides an opportunity for memory reconsolidation.
    A separate study by Baek et al. sought to analyze the neural circuits in alternating bilateral sensory stimulation (ABS)—the key component in EMDR therapy (the flashing lights) in mice models. Using single-unit recording in the basolateral amygdala (BLA), researchers found decreased activity in the neurons responsible for the fear state and unchanged activity in the neurons responsible for the extinction state when presenting ABS to mice that previously underwent fear conditioning. This change in firing was sustained one week later, which suggests ABS can lead to long-lasting attenuation of fear responses.
    Overall, the work of Dr. Grella, Sik-Lam Wong, and Baek et al. highlight how positive prediction error can lead to memory reconsolidation of fear memories which produces a decrease in maladaptive behaviors in both humans and animal models.



References


Baek, J., Lee, S., Cho, T., Kim, S., Kim, M., Yoon, Y., Kim, K.K., Byun, J., Kim, S.J., Jeong, J., and Shin, H. (2019). Neural circuits underlying a psychotherapeutic regimen for fear disorders. Nature, 566 339-343. https://doi.org/10.1038/s41586-019-0931-y

Grella, S.L., Fortin, A.H., Ruesch, E., Bladon, J.H., Reynolds, L.F., Gross, A., Shpokayte, M., Cincotta, C., Zaki, Y., and Ramirez, S. (2022). Reactivating hippocampal-mediated memories during reconsolidation to disrupt fear. Nature Communications, 13(4733), 1-19. https://doi.org/10.1038/s41467-022-32246-8

Wong, S.L. (2021). A Model for the Flash Technique Based on Working Memory and Neuroscience Research. Journal of EMDR Practice and Research, 15(3), 174-184. DOI: 10.1891/EMDR-D-20-00048