Effects of Neurofeedback on Emotional Regulation in Adolescent Females

Post by Elisa Guma

What's the science?

Adolescence is characterized by a wide range of neural changes that subserve higher order socio-emotional and cognitive function. During this transition period deficits in emotional regulation may emerge putting individuals at higher risk for developing psychiatric illnesses. The structural and functional connectivity of the amygdala and prefrontal cortex (PFC) are thought to play a key role in modulating emotional regulation, however, it is unclear whether we can intervene to actively shape these connectivity patterns to improve emotional and cognitive abilities during this critical time. This week in Neuroimage, Zich and colleagues aim to investigate if they can successfully use neurofeedback (a type of biofeedback aimed at teaching individuals self-control of brain activity via visual or audio feedback) from functional magnetic resonance imaging (fMRI) to help individuals regulate their emotions in real time.

How did they do it?

The authors conducted three different experiments to investigate whether neurofeedback based on an individual’s own PFC-amygdala connectivity could be used to modulate neural and emotional measures relevant for emotion regulation. In all experiments, adolescent females were evaluated on their emotional regulation abilities, as well as a variety of mood and anxiety measures. The first experiment was used to determine the best neurofeedback implementation. During the neurofeedback task participants were shown an image of a ten-segment thermometer while undergoing fMRI. During blocks in which no neurofeedback was occurring, the temperature was frozen at 6/10, whereas during neurofeedback blocks, the temperature of the thermometer was a direct real-time reflection of the PFC-amygdala connectivity. Participants were asked to try to control the temperature of the thermometer by controlling their thoughts and feelings and revisiting emotional reappraisal strategies. The three different implementations varied slightly in the way the thermometer displayed a change in connectivity.

Experiment 2 replicated experiment 1, except with a larger sample, and the best neurofeedback implementation (gleaned from experiment 1). Finally, in Experiment 3, the number of neurofeedback blocks was doubled, and the authors also collected Magnetic Resonance Spectroscopy data from the anterior cingulate cortex (implicated in reward processing) and PFC to measure Gamma aminobutyric acid (GABA) and glutamate, the major inhibitory and excitatory neurotransmitters in the brain. 

What did they find?

In their first experiment, the authors identified the best neurofeedback implementation and found a negative reinforcement of functional connectivity to be optimal (i.e. if the thermometer changes from 2/10 to 3/10 there was a more negative correlation of PFC-amygdala connectivity). In experiment 2, the authors aimed to assess the effects of one neurofeedback session on neural, emotional/cognitive measures, and their association in a larger sample. They did not find a significant change in emotional/cognitive measures or in the change in functional connectivity at the group level, but they did observe a practice-related change in connectivity, related to changes in thought control ability. Further, they found that state anxiety before the MRI session influenced the difference between functional connectivity in neurofeedback blocks relative to those with no neurofeedback.

Finally, in experiment 3 the authors replicated the findings from above in an independent sample. Moreover, they found that the concentrations of the inhibitory neurotransmitter, GABA ⁠— in both the PFC and anterior cingulate cortex ⁠— moderated the relationship between state anxiety before the MRI session and the effect of neurofeedback on PFC-amygdala functional connectivity.

What's the impact?

The study provides evidence for the feasibility of using neurofeedback in adolescent females to modulate functional connectivity measures between the prefrontal cortex and the amygdala. Further, the authors show that the relationship between state anxiety and the effect of the neurofeedback was modulated by GABA concentrations in the PFC and anterior cingulate cortex. Future studies may investigate the effects of longer training sessions and extend this work into populations with neuropsychiatric disorders.

Zich et al. Modulatory effects of dynamic fMRI-based neurofeedback on emotion regulation networks in adolescent females. Neuroimage (2020). Access the original scientific publication here

The Effects of Social Distancing on Body and Brain

Post by Anastasia Sares

What’s the science?

Humans evolved to be social with one another, and we function best when we have strong relationships and regular social contact. However, in many cities, half or more of the inhabitants live alone, and in the current COVID-19 pandemic, people are additionally deprived of in-person interactions at work and social gatherings. It is a good time to remind ourselves of the far-reaching impacts of loneliness and find ways to mitigate it. This week in Trends in Cognitive Sciences, Bzdok and Dunbar reviewed the consequences of social isolation and what we know about its neurobiology.

What do we already know?

Social connectivity is a huge factor in life expectancy. Social isolation increases the risk of dying within the next decade by 25%. The death of someone close, like a spouse, increases the likelihood of death in the immediate future by more than 15%. Severe social deprivation also shortens our telomeres, which are like caps on the DNA of every cell. Shortening telomeres are linked to aging.

Social connectivity is also related to immune function and physical health. In both humans and other primates, social belonging is related to stronger immune responses, faster wound healing, better regulation of stress hormones, lower systolic blood pressure, lower body mass index, and less inflammation. Finally, social connectivity protects against depression. People with a history of depression are 25% less likely to become depressed again if they belong to one social group (like a sports club, church, hobby group, or charity). If they belong to three social groups, their risk is decreased by around 67%.

One caveat for many of these large-scale human studies is that they involve correlation instead of causation. For example, if social isolation and body-mass index (weighing more for your height) are correlated, does it mean that social isolation leads to a higher body-mass index, or that having a higher body-mass index leads to social isolation? However, accumulating evidence from many different fields seems to indicate that loneliness is detrimental to our well-being.

What’s new?

We now understand a little better what’s going on in the brain. Advances in neuroscience have shown that social cognition recruits areas such as the default mode network (related to identity, reflection, etc.) and the limbic system (involved in emotion, motivation, and threat processing). Social isolation affects the brain just as much as the body—the shape and size of the limbic system change with our level of social isolation, and it also affects communication within the default mode network, and between the default mode network and the limbic system. 

Meanwhile, our social lives have gone digital. Research shows that people’s social tendencies are similar online. We seek out social interaction with the same frequency and have a similar social network as in real life. The problem with online interaction is that it is lower quality: until the rise of video chats, we couldn’t even pick up on facial expression and body language, which are important nonverbal cues. Synchronous behavior is still a challenge because of short delays in communication—as anyone who has tried to sing “Happy Birthday” in a conference call will know. Synchronous activities like team rowing or singing in a choir promote bonding in ways similar to physical touch and grooming and can help to prevent or reduce feelings of isolation. In short, nothing can fully replace face-to-face interaction, but digital communication does help to alleviate loneliness to some degree.

What's the bottom line?

We must take our social connections seriously, individually, and as a society. During times of social isolation like the current pandemic, this is especially important, but trends of urban living and aging populations mean that it is an issue we will be dealing with for years to come. Community organizations and hobby groups are crucial to preserving social interaction and community in this regard, as they can help to protect against social isolation.

Bzdok and Dunbar. The Neurobiology of Social Distance. Trends in Cognitive Sciences (2020). Access the original scientific publication here.

Loneliness Distorts Neural Representations of Social Connection

Post by Cody Walters

What’s the science?

Social connection is a key component of well-being. Social isolation and loneliness, on the other hand, are known to pose significant health risks. Despite the important role that social relationships play in one’s overall wellness, it remains unclear how the brain represents relationships between oneself and others and whether those representations are modified by loneliness. This week in The Journal of Neuroscience, Courtney and Meyer show that there are distinct neural representations stratified along social-closeness categories, with lonelier individuals having representational distortions between themselves and others.

How did they do it?

The authors used both univariate (i.e., average activity across voxels) and multivariate (i.e., multi-voxel patterns of activity; a voxel is like a 3D pixel in an image of the brain) functional magnetic resonance imaging (fMRI) analyses. Multivariate analyses typically involve training a classifier (machine learning model) to distinguish between multi-voxel activation patterns that correspond to specific stimuli. The authors used two multivariate methods: representational similarity analysis (RSA) and whole-brain searchlight analysis: RSA is a method for comparing patterns of blood-oxygen-level-dependent (BOLD) brain activity between distinct stimuli to quantify how similar (or dissimilar) they are, whereas whole-brain searchlight analysis is an approach for identifying voxelated neighborhoods that exhibit specific patterns of brain activity.

Prior to placing the participants inside the MRI scanner, the authors had participants list out and rank the names of five close others as well as five acquaintances. While in the scanner, participants fixated on a screen that displayed these target names (either their own name, one they supplied or one of five well-known celebrities) along with traits (e.g., polite, amusing, etc.). Participants then had to indicate how well the trait described that person on a scale from 1 (‘not at all’) to 4 (‘very much’).

What did they find?

The medial prefrontal cortex (MPFC) is known to represent information about the self as well as close others, thus the authors examined the activity of a predefined MPFC region of interest. Specifically, they constructed a representational dissimilarity matrix in order to test whether there was any meaningful structure in how self-other relationships are categorized in the MPFC. The authors identified that there were three representational clusters corresponding to self, social network members (i.e., close others and acquaintances combined), and celebrities. The authors then employed a whole-brain searchlight analysis to look for other brain regions that shared a similar clustering profile as the MPFC. They found that regions commonly implicated in social cognition — the posterior cingulate cortex (PCC), precuneus, middle temporal gyrus, and temporal poles — also exhibited a three-cluster structuring of self-other representations. Next, the authors investigated whether ranked closeness to the targets influenced neural responses. Restricting their analysis to the predefined MPFC region of interest, they found that mean MPFC activation linearly increased with perceived closeness to the target. The authors examined the extent to which representations of the self overlapped with representations of others. While self-other overlap did not linearly increase by target category (close others, acquaintances, and celebrities), they did find greater overlap between representations of the self and close others relative to acquaintances and celebrities. They identified the PCC/precuneus, as well as the MPFC as regions where the representations of others, were more similar to the representation corresponding to the self.

To examine whether loneliness modulated self-other representations, the authors used an established loneliness questionnaire. Between target categories, they found that the MPFC of individuals who reported higher loneliness represented adjacent (e.g., close others and acquaintances) and distal (e.g., close others and celebrities) targets as being more similar to one another. Furthermore, they found that within categories, acquaintances were represented more similarly to one another in both the MPFC and PCC of lonelier individuals. These data suggest that there is a blurring of representational similarity within and between social groups in lonely individuals. The authors also found that loneliness was negatively correlated with self-other similarity across all categories (close others, acquaintances, and celebrities) in the MPFC, whereas loneliness was positively correlated with self-other similarity across all categories in the PCC. These findings suggest that lonelier individuals might perceive others as being dissimilar from themselves owing to a lack of self-other representational similarity.

What’s the impact?

The authors provided evidence indicating how the brain might map out subjective social closeness in terms of representational similarity and how these representations are blurred and skewed in lonelier individuals. Developing a better understanding of how the brain processes interpersonal ties and how that processing is disrupted as a result of social isolation has implications for advancing our scientific understanding of happiness and well-being.

Self-other representation in the social brain reflects social connection. The Journal of Neuroscience, (2020). Access the publication here.