On Tuesday, May 6, at New York City’s cavernous Javitz Center, Dr. BJ Casey, PhD, took the stage to discuss why some of us are anxious — how we develop fear and anxiety responses neurobiologically, that is. Dr. Casey, a professor of developmental psychobiology and Director of the Sackler Institute at Weill Cornell Medical College, was speaking to a large room of psychiatrists, psychologists, and researchers at the American Psychiatric Association’s 167th Annual Meeting. “Thank you,” she announced, “I hope you’ve had some coffee. It’s a difficult time after lunch.”
As many as 18% of US adults and upwards of 30% of young people suffer from some form of anxiety, making it the most prevalent mental health disorder in the country. Depending on the type of anxiety and time course of symptoms, adults with anxiety are typically treated with various combinations of fast-acting anxiolytics (namely, benzodiazepines); other pharmacotherapies, such as antidepressants, and psychotherapy, particularly cognitive-behavioral therapy (CBT).
Similar strategies are used in children and adolescents, with CBT being the primary evidence-based behavioral therapy used in this population. The idea is to identify the cause of the anxiety and replace associated negative behaviors and thinking patterns with positive ones. Exposure-based therapy, a type of CBT, is considered one of the more effective approaches to treating pediatric anxiety, though it still only carries a 50%-60% success rate. The technique involves gradually and repeatedly exposing a patient to an anxiety- or fear-inducing stimulus or situation until they are able to overcome the negative associations and reactions — in other words, attempting to desensitize them to anxiety-causing cues.
Anxiety often goes undiagnosed, and therefore untreated, in children and adolescents. Left untreated, it can lead to chronic and debilitating mental and physical illness. Casey and a number of her colleagues feel that patient outcomes can be improved, not only through improved recognition but also by personalizing treatment. If researchers can untangle the web of connections occurring in the developing brain, and how certain developmental patterns increase or decrease anxiety risk, clinicians would then be able to better determine which patients are more likely to respond to a particular treatment.
The adolescent brain is frantic with activity and influences. “Regional synaptic pruning and changes in myelination are occurring at a time when the brain is being marinated in gonadal hormones,” Casey stated. She continued, “We also see peaks in neurotrophins and neurochemicals that are important to emotional and fear regulation.” Neurotrophins are proteins that control neuronal development and function.
Casey then pointed out that although most brain imaging studies in adolescents have looked at the cortex, the deep, more primitive brain structures also play a key role in development, particularly in terms of learning the emotional significance of environmental cues. By the early 2000s, numerous laboratories were looking at how different brain regions, both primitive and cortical, interact with each other during development.
Early Experiences and Later Life Anxiety
So how do the various developing brain regions contribute to our emotional development? And how do our adolescent experiences steer brain maturation and contribute to later-life fear and anxiety? The answer resides partially in the amygdala, part of the subcortical limbic system and a key player in emotion regulation. Also involved is the prefrontal cortex (PFC), the anterior part of our cerebral cortex responsible for complex thought and emotion processing.
Numerous studies have shown that fearful cues lead to heightened amygdala activity. And as Casey pointed out, the PFC relays inhibitory signals to the amygdala as part of our fear regulation circuit. With repeated exposures to an initially fearful stimulus that turns out to be unthreatening, increased signaling from the PFC dampens amygdala activity. As a result, output to the autonomic nervous system is dampened, and we feel less fear.
Of course, the neurobiology of human anxiety is far more complex than a signal circuit: Sensory inputs project to the amygdala’s lateral nucleus, where fear memories are maintained, and projections from the amygdala’s central nucleus also influence our autonomic and endocrine responses to fear. But it’s the relationship between the amygdala and the PFC that controls our ability to habituate to cues that might otherwise lead to anxiety and fear.
Work by Hare and colleagues (including Casey) found that in children who rate themselves as having high anxiety, a fearful stimulus results in initial recruitment of amygdala activity as expected. However with repeated presentations of the same cue, the activity doesn’t return to baseline, as it does in nonanxious control subjects. Their ability to habituate is impaired. Casey then explained the possible reasons why. Like so many mental and medical conditions, the culprit can be environmental, genetic, and oftentimes both.
To identify possible environmental influences on anxiety, Casey and her group turned to the often traumatic experience of an orphanage upbringing. They wanted to assess how adversity early in life impacts emotion regulation, in a sense using an orphanage childhood as a proxy for disorganized parenting.
They looked at children raised in overseas orphanages who were adopted and relocated to the New York City area. Study subjects were between 5 and 15 years old and had been in the United States for at least 2 years, to ensure that they had had time to adjust to the new environment.
First, they were presented with a series of neutral visual cues on a screen. Next, a fear-inducing face was flashed, to which the subjects were asked not to pay attention. A higher percentage of those who grew up in orphanages exhibited increased amygdala activity in response to the fear cue; in controls, increased activity in the PFC was seen, presumably suppressing attention toward the fear cue.
An important part of emotional regulation is our ability to suppress inappropriate reactions — in other words, to not waste our attention on emotional cues that have no significant relevance to a situation. Those who grew up in orphanages were unable to redirect their attention and ignore fearful cues.
The study also revealed how these findings might relate to real-world functioning. When the child participants came into the laboratory, they were asked to leave their adoptive parent to play a game with an experimenter for 10-15 minutes. When they came back, Casey’s team would monitor their interactions with their adoptive parents; those with greater amygdala activity made significantly less eye contact following this separation and reunion with their parent.
Casey cautioned that a drawback of human naturalistic experiments is the lack of control over prenatal history and genetic background. So the findings of dysregulated fear could be due to factors other than the “disorganized parenting” of the orphanage experience.
To help disentangle causality, her group ran a parallel study mimicking the orphanage experience in mice. Nesting material was removed from cages, and maternal behavior was monitored. The mothers spent more time foraging for nesting than with their pups. Mothers whose nesting wasn’t removed spent more time grooming their litter.
This model was then used to assess fear habituation, using a similar approach as in the human experiment. “How do you get mice to ignore a potential threat?” asked Casey. “As it turns out, they really like sweetened condensed milk,” she followed. The mice were put in a “home” cage equipped with a milk nozzle. When moved to a new cage with a light, considered a threat by the rodents, they froze for a period of time before going after the milk.
This slower latency time in the presence of stress was correlated with heightened amygdala activity. This pattern of increased latency and amygdala activity remained even after the stressor was removed 20 days later and even after development of the PFC in adulthood, suggesting long-term dysregulation of emotion with early adversity.
Genetic influences also appear to be a major contributor to fear and anxiety. Of particular importance appears to be the gene encoding for brain-derived neurotrophic factor (BDNF), a neurotrophin responsible for neuronal growth, differentiation, and survival.
Casey’s Cornell colleague Dr. Frances Lee has developed a new strain of mouse carrying a BDNF polymorphism in which methionine is substituted for valine. This BDNF Val66Met knock-in model results in decreased neurotrophin activity, biologically recapitulating the effects of the human polymorphism. The mice mimic the “wallflower at the school dance,” according to Casey: When tested in an open field task, they spent more time near the walls, an indication of anxiety-like behavior.
Fatima Soliman, Casey and Lee’s joint MD, PhD, student, performed a related study in the mice and humans with the BDNF Val66Met polymorphism, looking at Pavlovian conditioning to fear cues and their ability to extinguish negative associations. The mice received a minor foot shock paired with a neutral audible tone, whereas the humans were played an intrusive sound — Casey likened it to that “obnoxious alarm clock” — paired with innocuous images of colored squares. Next, the conditioned stimuli were presented repeatedly without the noxious stimulus, and subjects were monitored for fear responses (in mice, freezing; in humans, sweating).
Wild-type mice were found to freeze at first, but not after numerous tone presentations; BDNF Val66Met mice exhibited freezing behavior that did not extinguish with subsequent trials. Similar findings were seen in humans with the polymorphism, present in roughly 30% of those of white persons.
A single polymorphism is unlikely to fully account for the complexities of human fear and associated neurodevelopmental aberrations, and with twice as many presentations, they eventually learned. However, in subjects carrying a substituted methionine in the BDNF gene, fear doesn’t extinguish as readily in response to repeated nonthreatening cues; in those with a valine, it extinguishes more easily. Using neuroimaging, the authors showed that polymorphism carriers exhibited more amygdala and less PFC activity with subsequent cue presentations.
Clinical Implications and Manipulating Memory
The ultimate goal of this work is, of course, to treat patients suffering from anxiety. The data suggest that it might be possible to predict who will respond to certain therapies on the basis of their genetic profile and its correlation with fear extinction.
One study showed that adults with posttraumatic stress disorder who have the BDNF Val66Met polymorphism do not respond well to exposure CBT. Collaborative work between Casey and Lee looked at fear conditioning across all ages to determine whether certain age groups are better or less able to extinguish fear memories.
In response to the same negative cues as in their previous work, preadolescent and adult mice showed considerable fear extinction, whereas adolescent mice showed little to none. Humans showed a similar response pattern. Using expression of the c-Fos gene — an indirect marker of neuronal activity — the investigators showed that the PFC is not heavily recruited in adolescents; hence, their fear does not extinguish as well with repeated presentation of empty threat cues.
These data suggest that individuals of certain ages may not be as responsive to exposure-based CBT. Examination of existing clinical data by Casey and her colleague Dr. John Walkup indicate that this may be the case, but further research is needed, Collectively, this work provides evidence for whom and when exposure forms of CBT may be most effective.
Bringing it back to the neurobiological underpinnings of anxiety, Casey concluded by highlighting ongoing work in her laboratory looking at how anxiety might be alleviated by bypassing the reliance on the crucial projections from the PFC to the amygdala that are still developing during adolescence and altering fear memories themselves at the level of the amygdala. “Memory is not static, but dynamic,” said Casey. “When we learn something, we store it in memory. But every time we retrieve it, we update it with new information. We can attach new meaning to that memory.”
Is altering a memory that easy? Possibly. Casey and her colleagues wanted to attenuate fear memories by altering the memory during the so-called memory “reconsolidation window,” which appears to be between 10 minutes and a few hours after retrieval of an existing memory. Building on work in adults by Monfils and colleagues and Schiller and colleagues, they first applied their reliable unpleasant noise/colored square method to induce fear acquisition. Before extinction training of the fear memory, they presented the square that had been paired with unpleasant noise as a reminder.
After waiting 10-15 minutes — to coincide with the reconsolidation window — extinction learning was initiated. Those not exposed to the reminder cue before extinction learning showed an arousal response when retested the next day. Those reminded did not.
“This finding suggests that one way to work around exposure-resistant fear and anxiety is by taking advantage of this period of reconsolidation,” said Casey, nearing the end of her talk. It’s not that CBT doesn’t work across the board in adolescents, but rather that the correct type and timing of the CBT are essential.
Casey suggests that clinicians build on the reconsolidation window findings in the clinic. First, the patient comes in to the clinic and is reminded of why they are there (reminder cue). Then, clinicians establish a positive and safe rapport with the patient for 10-15 minutes — waiting for that critical window of apparent plasticity (reconsolidation window) — and then they initiate exposure therapy. “Many clinicians are already doing this,” Casey commented, “but [previously], we just didn’t have the evidence for why this timing may be so effective for some and not others.”
Following Casey’s talk, during the Q&A session, an audience member asked whether the data presented during the previous hour suggest that the “preadolescent anxious kid inevitably becomes the adolescent anxious kid.”
Casey’s response suggested that all hope is not lost — that intervening early before the circuitry becomes hard-wired may be the best hope for alleviating anxiety. This will require better and earlier identification of those at risk in order to intervene and ultimately prevent the anxiety from escalating. “I treat a lot of college students with social phobia who are terrified to speak in classes,” she responded, “If they’d gotten exposure therapy earlier, would they by more amenable to therapy now? It’s possible. But we just don’t know.”