Best of 2023, part 2: Neurodiversity and compulsive behaviour. How we code Reward Prediction Errors, and more.

Nature has published a list of the top 25 behavioural science articles of 2023. And (with the help of Consensus AI), I’ve been through the papers to see what could inform our work in the future.

Key takeaways

1. Neurochemical balance affects decision-making and impulsivity, relevant to understanding gambling behaviours.

2. Reward prediction errors are processed asymmetrically, influencing risk assessment and reward-seeking in gambling.

3. Cognitive and emotional control mechanisms differ in anxious individuals, potentially affecting risk-taking and decision-making in gambling.

4. The prefrontal cortex plays a crucial role in addiction behaviours, suggesting targeted interventions might mitigate gambling addiction.

5. Understanding brain responses to rewards and losses can inform strategies to address gambling addiction by rebalancing neural pathways involved in reward processing.

Cortical glutamate and GABA are related to compulsive behaviour in individuals with obsessive compulsive disorder and healthy controls

Utilises 7-Tesla proton magnetic resonance spectroscopy to explore the neurochemical basis of compulsive behaviour in individuals with and without Obsessive-Compulsive Disorder (OCD). It found a relationship between compulsive behaviour and glutamate levels in the supplementary motor area, and between habitual control and the glutamate:GABA ratio. Participants with OCD showed elevated glutamate and lower GABA levels in the anterior cingulate cortex, indicating a neurochemical imbalance related to compulsivity and habitual control.

Utility: neurodiverse individuals – those with OCD, on the autism spectrum or with schizoid characteristics – may be more susceptible to developing compulsive behaviours than neurotypicals. This might inform our use-warnings in future.

Neural circuit selective for fast but not slow dopamine increases in drug reward

Investigates the neural circuitry response to the speed of dopamine increases from methylphenidate (MP) administration in humans, comparing oral (slow) and intravenous (IV, fast) deliveries. Utilising PET-fMRI, it identifies a corticostriatal circuit, including the dorsal anterior cingulate cortex and insula, that is selectively activated by fast dopamine increases and associated with subjective ‘high’ experiences. This provides insights into the role of the salience network in drug reward and potential targets for addiction treatment. 

Utility: dopamine is a reward chemical, but also a learning one. Neural circuits ‘learn’ and are shaped by the reward signal dopamine provides. While drugs can provide vastly more dopamine than normal activities (and thus show clearer clinical results) we should be aware that there is a strong possibility that ‘faster paced’ games may be more addictive than slower paced ones. Faster paced games may create learned ‘response and reward’ pathways in the brain that are actually solely due to chance, leading to users to make poor decisions.

Personalised functional brain network topography is associated with individual differences in youth cognition

Examines the relationship between personalised functional brain network (PFN) topography and cognitive performance in children, utilising data from the Adolescent Brain Cognitive Development study. It finds that the cortical representation of fronto-parietal PFNs positively correlates with general cognition. Moreover, PFN topography can predict cognitive performance across domains, with prediction accuracy increasing along the cortex’s sensorimotor-association organizational axis. These results highlight the importance of individual differences in functional brain network organization for cognitive performance in children. 

Utility: photographic representations of kid’s brains can predict cognitive performance and prediction success. That’s just a bit scary to be honest.

Behavioural representational similarity analysis reveals how episodic learning is influenced by and reshapes semantic memory

Investigates the interaction between semantic and episodic memory during learning. It shows that episodic learning can reshape semantic memory, making related concepts more similar in semantic space. This process improves recall accuracy, especially for semantically related pairs, by enhancing the predictability of cues for targets. The study employs a novel behavioural representational similarity analysis approach, demonstrating systematic changes in semantic representations through learning, contributing to our understanding of memory integration and retrieval processes. 

Utility: onboarding users in episodic chunks can improve overall coherence of learned information and improve recall.

Transcranial focused ultrasound-mediated neurochemical and functional connectivity changes in deep cortical regions in humans

Explores the effects of transcranial ultrasound stimulation (TUS) on neurochemical and functional connectivity in deep cortical regions in humans. It demonstrates that TUS can selectively reduce gamma-aminobutyric acid (GABA) levels in the posterior cingulate cortex (PCC) and increase functional connectivity in targeted brain regions. These changes, observed to last at least 50 minutes after stimulation, suggest TUS influences overall brain excitability by modulating GABAergic inhibition, highlighting its potential for neuroplasticity induction. 

Utility: could we rewire problematic learned pathways in the brain? Maybe!

Association between vmPFC gray matter volume and smoking initiation in adolescents

Explores the relationship between the ventromedial prefrontal cortex (vmPFC) gray matter volume and smoking initiation in adolescents. It finds smaller left vmPFC volume is associated with smoking initiation, while changes in right vmPFC volume are linked to the maintenance of smoking behaviour. This suggests that vmPFC volume could be a biomarker for early stages of nicotine addiction.

Utility: we know smoking and gambling behaviours often appear together. We know activity in vmPFC affects confidence. It’s likely this area of the brain plays a role in both enjoyment (or at least confidence in) gambling behaviour and addiction. Could be a good area for continued study.

Asymmetric coding of reward prediction errors in human insula and dorsomedial prefrontal cortex

Explores asymmetric coding of reward prediction errors (RPEs) in the human insula and dorsomedial prefrontal cortex, using intracranial EEG recordings. It finds these brain areas encode positive and negative RPEs in a valence-specific manner, with a bias towards positive RPEs. This supports the idea of asymmetric coding across neural populations for RPE processing, influencing theories on the roles of these brain regions in reinforcement learning and cognitive control. 

Utility: positive RPEs are more likely to be encoded. That’s when you don’t expect much (or anything), but are positively surprised by the outcome.