Attachment Style and Brain Activity in Adolescents

This blog post first appeared on on July 31, 2014.



Adolescence is a critical time period during human social-emotional development. It is generally associated with changes in teenagers’ brain structure and function, as well as behavior. There is, however, also a major environmental change occurring during adolescence that is less frequently mentioned in this context, namely the separation of adolescents from their family. During teenage years, time spent with family members drops while time spent with people outside the family context (e.g., peers, work colleagues, romantic partners, etc.) increases. This social-adaptation process requires adolescents not only to integrate new and diverse experiences in relation to the world and themselves but to resolve many disagreements and interpersonal conflicts. In a nutshell, one of the central developmental tasks teenagers are confronted with is the establishment of an accurate understanding of their changing social environment.

Being a social neuroscientist particularly interested in human attachment behavior and attachment style, one of my research goals is to better understand the neural underpinnings of the above-described social-adaptation process in adolescents from an attachment-theory perspective. In other words, I am asking the question of whether a teenager’s attachment orientation (or attachment style) — which is established early in life through interactions between children and their primary caregiver(s) — could potentially explain individual differences in how adolescents’ brains process social-emotional information. Despite the fact that we already know quite a bit about the neural signature of attachment insecurity in adults, it still remains largely unknown what a neural signature of attachment insecurity may look like in teenagers. Together with a group of researchers from the University of Geneva — including David Sander, Brittany Anderson, Deborah Badoud, Stephan Eliez and Martin Debbane — we therefore measured social-emotional brain activity in 33 healthy adolescents ages 12 to 19 by means of functional magnetic resonance imaging (fMRI) and also obtained measures of their attachment style. The corresponding paper has just been published and is freely available here.

FIGURE_1The adolescents’ task in the fMRI scanner consisted of rapidly detecting on which side of a white line there were more dots (see figure, part a). More important, however, was the feedback adolescents obtained after each trial (see figure, part b). The latter feedback was composed of a word (either “won” in case of a correct answer, or “lost” in case of an incorrect answer) and an emotional facial expression (either smiling or angry). While two feedback combinations were congruent (smiling face and word “won,” angry face and word “lost”), two feedback combinations were incongruent (smiling face and word “lost,” angry face and word “won”). This setup allowed us to investigate adolescents’ brain responses while they were integrating objective performance feedback (personal success vs. failure) with the social evaluation of their performance (emotional faces representing social support vs. disapproval). We refer to this integration as social-feedback processing.

Our main findings indicate that attachment insecurities in adolescents may be associated with individual differences in brain activation during social-feedback processing. Furthermore, such individual differences appear to be distinct as a function of avoidant vs. anxious attachment orientations. We found that attachment avoidance was linked to increased brain activity during congruent social feedback processing. In other words, avoidantly attached adolescents appeared to preferentially process social feedback confirming their objective task performance. Conversely, we found that attachment anxiety was associated with increased brain activity during incongruent social-feedback processing. The higher adolescents scored on attachment anxiety, the more they seemed to focus on social feedback that did not match their objective task performance.

During adolescence, adaptively responding to one’s (changing) social context appears to involve a certain sensitivity and openness to socially conflicting information. Our new fMRI data suggests that, on a neural level, teenagers with an insecure attachment style may process social conflict differently from their securely attached peers. In the case of attachment avoidance, our data indicate that social conflict may be given less attention (i.e., social conflict is avoided). In turn, in the case of attachment anxiety, social conflict may be given too much attention (i.e., social conflict is overrepresented). Both of these social-conflict processing strategies have the potential to interfere with successful interpersonal-conflict resolution and social learning.

High levels of perceived social conflict are consistently associated with children, adolescents, and young adults manifesting psychological distress and maladjustment. We therefore believe that confirming and extending our findings is of high potential (clinical) relevance. Future investigations should also address the question of how such altered social-emotional brain-activation patterns as a function of attachment insecurity in adolescents are related to the neural signature of attachment insecurity emerging in adults. Combining such knowledge will hopefully support the development of new (early) detection and intervention strategies in the context of attachment insecurity.

Posted in Attachment, Brain Imaging, Neuroscience, Psychology | Leave a comment

The Neural Signature of Attachment Insecurity

This blog post first appeared on on October 16, 2013.




In one of my previous blog posts on human social brain evolution, I already briefly mentioned the concept of attachment and attachment style. Here, I would like to elaborate on these considerations by describing in more detail what we have learned about the neural mechanisms underlying attachment behaviors in humans, and how such information could help us maintain better relationships with others.

In the late 1960s and early 1970s, Mary Ainsworth and John Bowlby first described a developmental psychological framework related to social interactions that has since become known as attachment theory. It states that, in order to survive, infants form a strong bond to their primary caregiver(s); they become attached. As infants grow older, this attachment evolves from a primarily biologically driven “survival instinct” into a sophisticated “cognitive-emotional social relationship processor.” A person’s mode of attachment, also called attachment style, determines what he/she will expect from social interactions, and how she/he will use cognitive and emotional resources for self-regulation during times of social emotional stress.

Central to the understanding of the concept of attachment is the assumption that all infants will become attached, but that such attachment can be either secure (i.e., “good”) or insecure (i.e., “bad”). It all depends on how the primary caregiver(s) respond to the infant’s attempts to establish a close social connection. Ideally, the caregiver(s) are available, responsive and caring in times of need. Infants consequently learn to trust in others and their own social emotional abilities; they can establish a secure attachment style. Unfortunately, such scenario does not always happen, allowing for the formation of an insecure attachment style in infants. If the primary caregiver(s) react in an unpredictable or inconsistent manner to the infant’s social approach signals, an anxious attachment style is bound to emerge. The latter is characterized by a constant fear of abandonment and need for reconfirmation due to a very low self-esteem. In turn, if the primary caregiver(s) is/are unavailable or unresponsive in times of need, infants will develop an avoidant attachment style. They learn to expect social rejection and fail to associate social interactions with positive experiences. Interacting with others may even be seen as futile or dangerous. Consequences are social-emotional closure and the denial for the need of social relationships. Some more information can also be found in blogs by Lisa Firestone (for example here).

Although the foundations of a person’s (secure versus insecure) attachment style are laid early in life, this process can have a strong influence on social emotional abilities and well-being throughout the lifespan, and even affect future generations. This comes from the fact that an individual’s attachment style is thought to remain rather stable from childhood to adolescence and adulthood, and may be transmitted from one generation to the next. Given that at least one third of people are believed to have an insecure attachment style and attachment insecurity is a known risk factor for the emergence of psychopathology (e.g., social anxiety disorder, or borderline personality disorder), attachment insecurity has a high potential for causing considerable problems in society.

In order to develop new detection, prevention and intervention strategies for social emotional difficulties associated with attachment insecurity, researchers have recently begun to investigate the neural basis of attachment style in humans. The main goal of such approach is to find neural markers of attachment insecurity in the human brain, allowing for understanding what may cause people to react with avoidance or anxiety to social emotional signals. Because it is easiest to establish such brain-behavior associations in adults, initial research has focused on adult populations.

Let’s first consider attachment anxiety. Converging evidence from adults suggest that this insecure attachment style is characterized by hyper-sensitivity to social emotional information in general, particularly so if negative. Anxiously attached individuals show high activity in brain areas that normally process social rejection and conflict, as well as maintain high arousal and negative emotionality. It looks as if anxiously attached people’s brains are constantly in a high alert state, monitoring the environment for potential signs of social emotional threats. Even if there are none.

Attachment avoidance, in turn, seems to considerably lower the brain’s responsiveness to social emotional information, especially so if positive. Avoidantly attached individuals show low activity in brain areas that usually process reward and prosocial motivation. This even appears to hold true for mothers seeing pictures of their own smiling infants, normally representing very potent inducers of reward-related social brain activity. It appears as if avoidantly attached individuals do not feel good when interacting with others. Even if these interactions are mutual.

Although admittedly reductionist (see here for more details), the above simplified description of the neural basis of attachment anxiety and avoidance nicely illustrates the strong influence the environment within which a person grows up in can have on social emotional brain responses during adulthood. These findings in humans furthermore nicely accord with data from animal studies (see here and here), the latter suggesting that such environmental effects can even override potential genetic predispositions for attachment insecurity.

The practical implications of the above-mentioned brain activations associated with attachment anxiety and avoidance are manifold. Firstly, we should be strongly aware of the fact that our behavior towards infants (as well as children and adolescents) can crucially influence their social emotional brain development, shaping how they themselves will behave during adulthood. Key elements of such awareness should be the maintenance of a stable, predictable and responsive environment within which children grow up in. Second, and associated with the first point, we should be more aware of our own attachment style to better understand our own reactions in the course of social interactions. This can help us to proactively change our social emotional behavior, or at least disengage from interactions early enough to prevent damage. And third, knowing more about the brain basis of attachment anxiety and avoidance, we can start developing new prevention and intervention strategies specifically targeting affected social emotional reward and threat behaviors and underlying brain circuits.

Ongoing research on the neural basis of attachment insecurity in adolescents and children will hopefully soon add more information, allowing for an early detection and prevention of, and/or intervention in the case of social emotional problems associated with attachment insecurity. The sooner we can act, the better.

Posted in Attachment, Brain Imaging, Neuroscience, Psychology | Leave a comment

Tickling the Child Brain

This blog post first appeared on on October 2, 2013.


laughing child


Not too long ago, there was a vivid debate on the evolutionary origins and function of humor in humans (HuffPost TEDWeekends; see herehere and here). I do agree that these are important questions to ask. I do also agree that, in order to answer these questions, we need more direct evidence on how humor is processed by the human brain. Luckily, thanks to modern neuroscience methods like functional resonance imaging (fMRI), we can nowadays look into the human brain right at the moment when it is exposed to something (deemed) funny. And that’s exactly what we did in a study yielding two recently published papers (paper 1 and 2): we tickled the child brain!

The experimental setup we used was simple. Our participants, who were 6 to 13 years old children, watched different kinds of short movies while their brains were scanned with fMRI. Movie clips were funny, positive (enjoyable to watch but nonfunny) or neutral, and funny movies contained scenes typically shown in the TV show America’s Funniest Home Videos. Our humor condition thus contained what Peter McGraw and Joel Warner would probably call “benign violations“: unsettling and/or threatening, but still acceptable and safe scenes.

The results of our manipulation revealed that the child brain processes humor very similarly to the adult brain, and thus mainly in two separate (but interconnected) networks. One set of areas is most likely involved in detecting and resolving the “benign violations”, or in our words “incongruity” that arises during humor processing. Such incongruity detection and resolution then likely entails a sense of amusement or mirth, reflected in increased activation of another set of brain structures associated with reward. This activation pattern reflects the latest “neural circuit model of (verbal) humor processing“ proposed by Chen and colleagues, which builds on Wyer and Collins’ original “comprehension and elaboration theory of humor“.

Besides confirming and further extending our knowledge of the neural substrates of humor processing in humans, our two child studies also tested for the influence of four additional variables. These were sex, intelligence, age, and temperament/personality.

The first two of these four variables, sex and intelligence, are particularly relevant from an evolutionary point of view, because one humor theory suggests that humor in humans might have evolved to serve as a tool for mate choice. More extensively explained in another blog, this theory proposes that women may use humor as a tool to evaluate men’s intelligence, resiliency, and/or social skills. Combined with previously acquired data from our laboratory in adults, our first paper in children may provide support for this mate choice theory. We found significantly stronger activity to humor in incongruity detection and resolution as well as reward areas in girls as compared to boys. These results imply that the female brain may be better prepared to perceive humor. One possible underlying mechanism for such humor predisposition in girls and women could be a weaker reward expectation, the latter appearing to be stronger in boys and men.

Although interesting, such sex-differences in humor processing related with the evolutionary aspect of mate choice do not provide any clues on a possible association with intelligence. That is why, in our second paper, we looked at neural response to humor as a function of IQ. Our findings again seem to support the mate choice theory, because we observed higher brain activity in both above-described networks with increasing IQ scores (across both sexes). It therefore appears that a higher IQ may not only facilitate incongruity detection and resolution, but also enhance the experience of amusement following it. As preliminary as our findings regarding intelligence are, they offer a good starting point for future scientific ventures, and represent a first putative link between humor theories and neural mechanisms associated with humor processing.

Apart from such evolutionary considerations related to sex and IQ, we also probed for the effects of age and temperament on humor processing in the second paper. Our results revealed that humor appears to undergo developmental changes, and that its neural signature is subject to individual differences in personality. It therefore seems that, although our brains may be to some degree predisposed to process humor, such humor processing can vary. For example, combined with previous data from children, we found that increasing age was associated with less reward-related, but more incongruity detection and resolution related brain activity. It could therefore be that older children process “benign violations” in a more elaborate way, less dependent on the fundamental safety and play nature of humor. Such findings corroborate a suggested differentiation between more basic, so-called “protohumor”, and more complex “formal humor”, the latter depending more strongly on theory of mind and language skills.

Finally, in our second paper, we observed that humor is susceptible to differences in personality. Of particular interest was the finding of weaker brain activity to humor in both incongruity processing and reward-related areas in shy children. Shy children are generally described as taking a long time to warm up to strangers, having difficulties with making friends, and not being very sociable. Such a personality trait may therefore hinder children to learn understanding and experiencing humor through social interactions. Consequently, shy children may have difficulties in detecting and resolving “benign violations”, because they do not have enough opportunities to learn what a “benign violation” actually is. In addition, shy children may experience social interactions as uncomfortable or “unsafe”, which could prevent shy children from interpreting them as “benign” from the start. These mechanisms might partially explain the increased risk for fearfulness and anxiety disorders, as well as proneness to peer rejection and victimization of shy children.

Overall, our findings on the neural basis of humor processing (in children) may provide some preliminary clues informing theories regarding humor evolution and function in humans. More fMRI and related neuroimaging research is, however, needed to reproduce and elaborate such considerations. This process may not only keep advancing humor theories, but could also be of clinical relevance for conditions involving social disturbances.

Posted in Brain Imaging, Humor, Neuroscience, Psychology | Leave a comment

Evolution of the ‘Social Brain’ in Humans: What Are the Benefits and Costs of Belonging to a Social Species?

This blog post first appeared on on September 16, 2013.




Humans are highly social beings. We like to be surrounded by friends and share our personal experiences with others. The recent appearance of various social networking tools, and their adoption at a virtually explosive rate, nicely illustrate the strong and fundamental human desire for social belonging and interpersonal exchange.

Not surprisingly, there is emerging evidence that evolutionary processes have favored the development of complex social behaviors in humans, along with the brain architecture that supports them. The human brain, and particularly the neocortex (which constitutes its outmost layer), is much larger in humans as compared to other primates and mammals of similar size. This is particularly interesting because the neocortex comprises many of the brain areas involved in higher social cognition, such as conscious thought, language, behavioral and emotion regulation, as well as empathy and theory of mind — the ability to understand the feelings and intentions of others. We are, so to speak, biologically hard-wired for interacting with others, and are thus said to be endowed with a “social brain.”

The reasons for the evolution of the human social brain are not yet completely understood. There is, however, growing consensus that two processes likely played key roles in triggering the observed dramatic increase in brain, and particularly neocortex, size. These were the development of (i) socially monogamous pair bonds, and (ii) paternal care / the involvement of the father in rising children (see here for additional information). Both of these processes offered additional defense mechanisms against infanticide and predation on offspring. In a nutshell: if the father stuck around long enough with his partner, and vice versa, the common children had a higher chance of survival, which in turn increased both the female’s and the male’s reproductive success. And that is what ultimately counts in evolution: promotion of survival of the fittest.

In terms of brain development, the emergence of socially monogamous pair bonds and bi-parental care imposed unprecedented needs for extensive social coordination and synchronization between the two partners and parents. For example, who is responsible for which aspects of childcare? Or how can both partners optimize the time used for searching food versus providing shelter to their offspring? These newly appearing requirements are thought to have laid the foundations for human social evolution as reflected in the development of advanced social cognition and skills.

It is believed that eventually, the cognitive mechanisms necessary for pair bonding and coordinated bi-parental care generalized to non-reproductive individuals. This generalization likely helped the formation of larger social groups, such as extended families, same-sex alliances, large coalitions, etc., again promoting higher rates of survival and reproduction. The increasing size of such expanded social groups in turn entailed further social brain development, ultimately giving rise to the present form of the human social brain.

The obvious benefits of social brain evolution are that humans are, today, equipped with a highly sophisticated social processing machine that enables us to engage in complex social interactions, and to maintain relationships to a great number of different individuals as well as groups. Our brains are further wired in a way that we experience reward during mutual social interactions, and feel sensations similar to physical pain when we are socially rejected or disapproved. It therefore looks like evolution has provided us with the perfect hardware for living in a world that is becoming ever more crowded.

Social evolution, however, does not only come with benefits. It also has its costs. Because the human brain has become so large and sophisticated in terms of the social computations it supports, it takes a very long time for it to develop fully. Compared to other mammals, human children have a very long developmental period and are highly dependent on care by adults. Human parents not only have to nurture their children until their brains are fully operational biologically, but they also have to provide an extended and stable context within which their children can safely acquire all the skills necessary for understanding their social surroundings. And this process continues far beyond childhood. For example, some social skills can only be learned by means of peer activities during adolescence, and throughout this period parents still have important protective and sheltering roles.

Recent research on attachment style in humans is starting to reveal some of the underlying behavioral and neural correlates of such guided social learning, and particularly the consequences if social learning does not occur in a sheltered environment. An individual’s attachment style is a measure for the quality of his/her social bonds with others. It is crucially shaped through interactions with caregivers in early life, such as a child’s parents. If others close to a child are responsive and caring, the child develops a secure attachment style. If they are unresponsive or inconsistent in their behavior towards the child, however, the child develops an insecure, either avoidant or anxious attachment style. Once acquired, the attachment style of a person is believed to remain rather stable throughout the lifespan, and to even be transmitted from one generation to the next. It is therefore likely to circularly influence many of the steps involved in social brain development and skill acquisition during childhood, adolescence, and even adulthood. We now have evidence that an insecure attachment style prompts people to respond either too strongly or too weakly to social information. This is especially striking in the case of attachment avoidance, which appears to considerably decrease both behavioral and brain responses within positive social contexts normally experienced as strongly rewarding. In other words, we know that a person’s behavioral and brain responses to social information can be importantly altered if social brain development and skill acquisition do not occur in a protected setting.

Becoming social has made us who we are today. Evolution has provided us with the best tools possible for successfully engaging in social interactions. We should, however, not forget our responsibilities to ensure that future generations can learn in a protected and sheltered environment how to properly use these tools for their own lives.

Posted in Attachment, Brain Imaging, Neuroscience, Psychology | Leave a comment