Importance of proposed work
Brain oscillations are the rhythmic fluctuation in the excitability of a population of neurons, representing synchronised neural firing. This synchronisation is key for ‘binding’ together sensory elements of experience and memory. In particular, modulation of theta band oscillations (4hz~), a dominant signal in the hippocampus (Jacobs, 2014), has been shown to impact learning in animals (Hyman et al., 2003).
A non-invasive means to investigate the causal impact of brain oscillations on memory is via the sensory entrainment effect. Neuronal assemblies in sensory regions are known to synchronise to the frequencies and phase of external sensory ‘entrainment’ stimuli, such as a 4Hz auditory tone, or flickering pattern (Becher et al., 2015). Clouter et al., (2017) used two stimuli elements (auditory and visual), modulated such that they were either synchronous or asynchronous, in an associative learning paradigm. When auditory and visual stimuli were in a synchronous theta (4hz) phase, greater recall of the association between auditory and visual elements was observed. Further investigation of these findings will allow us to gain a greater understanding of how human memory represents associations, and offers great potential utility via non-invasive neural entrainment for regulating memory. This could include new low-intensity treatments for memory dysfunctions, or for those wishing to enhance learning and retrieval of new associations.
Two crucial questions are whether the theta effect is influenced by the memory content (e.g. emotional vs. neutral) and whether neural entrainment could be used to enhance the recall of previously consolidated memories. A further key unknown is whether entrainment could be used to more readily induce memory destabilisation and reconsolidation of long-term associative memories. This memory maintenance process (Lee, 2009) provides a mechanism through which problematic long-term memories could subsequently be updated or otherwise altered. Manipulating memories in this way could offer a breakthrough for the development of treatments for disorders where strongly encoded maladaptive memory associations putatively play a key role, such as substance use disorders and threat-related disorders (Paulus, Kamboj, Das, & Saladin, 2019 & Kamboj, 2018; Walsh, Das, Saladin, & Kamboj, 2018 & Saladin, 2019). A key roadblock in tapping this mechanism is that is that the memory destabilisation/reconsolidation process is not universally triggered upon memory retrieval. This has resulted in failures to observe reconsolidation-interference effects (Chalkia, Van Oudenhove, & Beckers, 2020 2020; Elsey, Van Ast, & Kindt, 2018), which may represent a failure to destabilise memories at retrieval. Since theta plays an established role in memory binding and prediction error signalling (Cavanagh, Frank, Klein, & Allen, 2010), a putatively key determinate of memory destabilisation (Exton-McGuinness, Lee, & Reichelt, 2015), I hypothesis that theta entrainment could be used to boost memory destabilisation; enhancing subsequent reconsolidation effects. I will address these key questions in a series of three experimental studies.
The experiments utilise altered versions of the multi-sensory entrainment paradigm developed by Clouter et al. (2017). During the ‘encoding phase’, participants are presented with a three-second audio clip and an image. They are required to remember the association between the image and the audio. In some conditions the luminance of the visual stimuli is oscillated, such that it ‘flickers’ at a frequency we wish to entrain brain oscillations to (e.g. theta). Non-flickering stimuli will act as a within-subjects control condition. The amplitude of associated auditory stimuli is manipulated in the same way, either in-phase, or out-of-phase, with visual stimuli. At a later test, recall will be assessed by presenting participants with the audio components and asking them to identify the matching image.
We will expand on the novel theta memory enhancement effect identified by Clouter et al. (2017) in key ways, and investigate possible clinical utilities of neural entrainment, through two small studies and one large study. In the first study, we partially replicate and extend Clouter et al.’s (2017) findings by investigating whether the content of the memory (emotive/neutral) influences the size of neural entrainment effects on memory, as emotional content is typically encoded more strongly. The second study will investigate the long-term effects of oscillatory entrainment on retrieval of previously consolidated memories (encoded 24+ hours prior); a key step for practical application of this approach. The final study will draw on the findings from the first two, and investigate whether putative memory destabilisation/reconsolidation effects can be boosted by oscillatory entrainment at retrieval.
Sample size and study cost
The first two studies will last approximately two hours. Based on past research showing an effect size of η2p= .08 (Wang et al., 2018), to obtain 95% power with 3x2 design we will need at least 40 participants, and will recruit 52 participants for this study. As such each study will cost £1092 (at a rate of £7.50 per hour plus service fee and VAT).
The third study investigates reconsolidation and thus requires a 3 session protocol (session 1: encoding; session 2: retrieval and manipulation; session 3: retrieval test), with 5 hours total testing time. Assuming a medium effect size, to obtain 90% power, with 4 between-subject groups, we will recruit 156 participants. As such this study will cost £7371. Therefore we estimate the total cost for this proposal to be £9555.
The first study has been pre-registered at aspredicted.org (#70691). The two following studies will both be pre-registered before data collection begins. These have not yet been pre-registered to allow our initial findings to guide subsequent designs.
Ethics approval is in the process of being obtained from our institution to conduct the study.