Skip to content

Neuroscience and Psychiatry Module 2: Fear/Safety, Anxiety, and Anxiety Disorders


Anxiety disorders include a wide range of illnesses which affect an estimated 18% of adults in the U.S. While anxiety and fear can be beneficial and keep us safe in harmful situations, excessive anxiety and the inability to respond to safety cues can be disabling. This module explores the relationship between fear/safety, anxiety, and anxiety disorders.

Case Vignette

62 year old Caucasian male veteran who experienced incredible horror in Viet Nam. He has had a long course of severe PTSD. He is getting married on a rainy day and is walking out of a church in Vermont in a white tuxedo, with his new bride on his arm, a car drives by and backfires. He dives into the muddy bushes for cover.

In the face of a stimulus that reminded him of past dangers, why did he disregard the safety signals?

40 years since Viet Nam - now in a safe time

In Vermont not Viet Nam - now in a safe place

In a white tuxedo, not battle fatigues, with his new bride next to him - now in a safe context

Is Post Traumatic Stress Disorder An Inability to Respond to Safety Signals?

Video Presentation: What is the Relationship Between Fear/Safety, Anxiety, and Anxiety Disorders?

Let's start with some definitions. What is fear? Fear is a feeling of disquiet that begins rapidly in the presence of danger and dissipates quickly once the threat is removed. It is generally adaptive.

Anxiety, on the other hand, is uneasiness over the anticipation of less specific or predictable threats. It lasts longer than fear and can also be adaptive.

When fear and anxiety are greater than expected or last beyond what is adaptive, affecting well being and function, then an anxiety disorder is present.

Fear can be innate or learned. Examples of innate fear include the fear of scorpions, snakes, or heights. Learned fear stimuli such as guns would not have been frightening to someone who lived in the 12th century, for example. Neither would images associated with man-made disasters or destruction such as a car accident or mushroom cloud.

Fear is highly preserved across animal species, so many species exhibit fear conditioning.

We know that the classic fear response is "fright, freeze, flight" but where does that happen in the brain?

human brain showing amygdala, and face showing fear in response to seeing a snake.

We learned from animal models of fear that the amygdala is central to the fear response. Sensory input reaches the sensory thalamus and then very quickly the amygdala which allows for a fear response, via the so called low road. This direct pathway to the amygdala is necessary but not sufficient, it cannot differentiate a snake from a stick.

The cortex is needed for conscious awareness, context, and perspective. That process is slower and occurs via the so called high road.

Illustration of connection from sensory thalamus to amygdala demonstrating low road and connection from sensory thalamus through cortext to amygdala showing high road.

Here is a clinical example of how central the amygdala is to the fear response.

Patient SM has a very rare case of Urbach Wiethe disease, a congenital calcification of the amygdala bilaterally.

For many years, SM has repeatedly stated that she "hates" snakes and spiders. When asked to elaborate, SM reports that she simply does not like them and "tries to avoid them." In order to assess the validity of SM's claims, investigators took her to an exotic pet store that specializes in selling snakes and spiders.

Contrary to her verbally stated aversion to snakes and spiders, SM displayed a striking pattern of excessive approach-like behavior in concert with a lack of avoidance behavior; a pattern highly reminiscent of the behavior displayed by monkeys with focal amygdala lesions.

SM repeatedly approached all snakes, including holding and playing with a snake for over three minutes. She attempted to touch a tarantula, but had to be stopped due to the high risk of being bitten. She displayed a compulsive desire to want to "touch" and "poke" the store's larger and more dangerous snakes, even though the store employee repeatedly told her that these snakes were not safe and could potentially bite.

large, scary snakes and spiders

When asked why she would want to touch something that she knows is dangerous and claims to hate, SM consistently replied that she was overcome with curiosity.

Throughout all of this SM's reported experience of fear never surpassed a minimal level!

This case illustrates the essential role of the amygdala in fear, a human emotion.

Can We Use Animals to Model Human Emotions?

That may be difficult to do for some emotions.

But we can model fear in animals. This is useful because fear, anxiety, and anxiety disorders are likely to be biologically related. For example, neuroimaging studies show that there are brain circuits common to humans and animals that are involved in normal fear as well as anxiety disorders. And animals can learn to fear a neutral stimulus. Therefore animal models of fear are feasible and useful in informing our understanding of anxiety disorders in humans.

Let's begin by describing an animal model of fear developed by Dr. Michael Davis and his colleagues using fear potentiated startle.

Central to this model is fear conditioning or cue conditioning, which is essentially learning to associate a neutral stimulus such as a light with an aversive event, for example, an electric shock. One can conceive of that as learning about danger.

This animal model relies on the observation that animals and people startle more when they are afraid. For example, if the phone rings suddenly, you may startle a bit. If you are alone at night watching a scary movie and the phone rings suddenly, you would startle far more severely. So how is this applied to animals.

The experimental device consists of a box with a grid that is capable of delivering an electric shock.

During training, a light is paired repeatedly with an electric shock such that the animal learns that the light will be always followed by a shock and becomes afraid of the light. That is classical conditioning.

In the testing phase, the animal is startled by a noise while the light is off. No more shocks are given. The extent of the startle is measured by how high the animal jumps.

The light is then turned on which induces fear in the animal. The animal is then startled by the same noise. Because it is afraid, the animal startles much more, and jumps higher. The difference in the height of the jump between startle when the light is off and when it's on is a measure of fear potentiated startle.

So what happens in the brain as conditioned fear is acquired?

New excitatory connections are thought to be formed within the amygdala as fear learning takes place through long term potentiation. These connections are formed between the lateral nucleus of the amygdala and the central nucleus of the amygdala which is thought to be the main amygdaloid output structure sending projections subcortically.

part of the amygdala with lines to show where connections are formed using conditioning

Therefore, fear producing cues through the basal nucleus and the central nucleus of the amygdala activate multiple brain regions that then produce a variety of signs and symptoms of fear and anxiety.

For example, activation of the lateral hypothalamus can lead to changes in blood pressure and heart rate, and sweating. Activation of the parabrachial nucleus can lead to panting and respiratory distress, and so on.

The neurochemistry of fear learning has been elucidated using animal models. It is known that glutamate acting at the NMDA receptor is critically involved in learning and memory as well as in fear learning. For example, blockade of NMDA receptors in the amygdala interferes with the acquisition of conditioned fear.

NMDA receptor showing the GLU, NMDA agonist-binding sites, APV, CPP antagonist sites, and Glycine site. Also shows blockade of  magnesium molecule.

Animals can also learn to respond to safety signals.

If a light is consistently followed by a shock, the light becomes a danger signal. If a light is repeatedly not followed by a shock, the light becomes a safety signal.

Can this animal model be adapted to humans?

Dr. Davis and his group were able to devise a fear potentiated startle paradigm in humans to study safety signal learning based on the animal model.

As you see in this picture, the acoustic startle probe is delivered through headphones. The EMG recordings of the eye blink muscle contraction are used as a measure of the startle response. The unconditioned stimulus is a puff of air to the throat which is mildly noxious. The conditioned stimulus is the viewing of a colored square.

wearing headphones, sensors to measure eye blink, and air puff tube attached at the base of the throat

During the habituation phase, the acoustic startle probe or the noise is delivered and startle reactivity is measured.

During conditioning, the conditioned stimulus or fear signal, in this case a green square is shown along with a neutral stimulus or yellow square and followed by an air puff.

Using the response key pad the subject indicates whether he is expecting an airblast, no airblast or does not know.

keypad with three buttons labeled with plus sign, zero and minus sign

The safety signal pink square is shown along with a neutral stimulus or yellow square and never followed by an air puff. Using the response key pad the subject responds.

The danger signal is then shown. Paired with the safety signal.

In this paradigm, several questions were asked:

Can fear potentiated startle be detected in people using this model?

Can subjects learn to respond not only to a fear signal but also to a safety signal?

Is the ability to respond to either fear or safety signals impaired in PTSD?

Here are the results in healthy controls. A, refers to the green square which represents danger and X refers to the yellow square which is neutral. Clearly the subjects show a fear potentiated startle response to the danger signal.

In contrast, the safety signal does not induce such a response. B refers to the pink square which represents safety.

When the danger and safety signal are shown together AB the response is intermediate. This indicates an ability to generalize or transfer the safety signal.

Finally, when the subjects are presented with the green square representing danger along with a black square or C - a neutral signal never before seen - they react to the danger signal as they have before. This is a controlled condition.

bar chart showing study results

Veterans of the Croatian war who do not suffer from PTSD showed the same results as healthy controls. However, veterans of the Croatian war who suffer from PTSD, showed a different set of responses.

They responded to the danger signal and the safety signal in the same way but responded to the mixed signal in the same way they responded to the danger signal. What that suggests is that the subjects were not able to generalize or transfer the safety signal.

This same experiment was repeated by Dr. Kerry Ressler and his colleagues in the Grady Trauma Project. This is an entirely different group of civilians from inner-city Atlanta who have experienced multiple traumas. Some suffer from PTSD and some do not.

In this graph, the purple column shows the response to the combined danger and safety signals. Subjects with PTSD responded to it in the same way they did the danger signal while subjects without PTSD were able to transfer or generalize the safety learning to the combined state.

Grady Trauma bar chart showing study results

These experiments demonstrate that it is possible to have an objective measure of safety signal learning not only in animals, but also in humans. And, as illustrated in the case vignette, people with post-traumatic stress disorder do not respond to safety signals in the same way as healthy controls.

How can we use this information to help us develop novel treatments for this disorder?

Currently, selective serotonin reuptake inhibitors are the mainstay of long-term pharmacological treatments of anxiety disorders and benzodiazepines are useful for short-term relief. While SSRIs have demonstrated efficacy in controlled clinical trials of PTSD, their universal effectiveness in the real world has been questioned. Novel therapies are needed, particularly ones that do not have to be administered indefinitely and that have fewer side effects.

In addition to pharmacological interventions, Cognitive Behavioral Therapy, or CBT is an effective treatment approach for anxiety disorders and a first-line treatment for PTSD. Exposure therapy is a CBT approach that is thought to lead to extinction of fear.

In exposure therapy, a patient is repeatedly exposed for increasing periods of time to a feared object or situation in the company of a supportive therapist. The lack of aversive consequences leads to extinction. This is thought to be due to new learning.

In turn, this new learning allows the patient to face feared cues or situations with less fear and avoidance.

NMDA receptors are known to play a role in fear learning and fear extinction. NMDA receptor antagonists impair extinction in rodent models and NMDA agonists facilitate it.

D-cycloserine is a compound that indirectly activates the NMDA receptor.

receptor showing the GLU, NMDA agonist-binding sites, APV, CPP antagonist sites, and DCS binding at the glycine site which activates the receptor

A single dose of D-cycloserine given 2-4 hours prior to the therapy session, produced an enhanced decrease of fear at 2 weeks and 3 months after this single treatment.

This was tested in a virtual reality model for people with acrophobia, or fear of heights, where they wear goggles and they feel that they are in a three-dimensional space.

They can virtually walk out on ledges of increasing height and look down.

The first observation in this placebo controlled study is that D-cycloserine did not increase or decrease anxiety by itself indicating that it is not an anxiolytic.

After only two doses of DCS each administered before an exposure session, the DCS group showed a significantly greater decrease in anxiety than the placebo group, and that change persisted at follow-up after three months, with no additional DCS given.

Since it is not an anxiolytic this data suggest that it worked by enhancing extinction.

In addition, three months after the treatment, the DCS group reported exposing themselves to heights significantly more than the placebo group, indicating that the treatment altered behavior in the real world.

DCS is not approved for clinical use, but remains a research finding. In addition to simple phobias, the enhancement of cognitive behavioral therapy with DCS has been shown to be effective in studies on panic disorder, obsessive compulsive disorder, and social anxiety. However, it is not clear whether DCS will enhance CBT in PTSD. Recruitment of subjects for PTSD clinical trials is underway at multiple centers.

Take Home Messages

Animal models of fear and safety learning are feasible and highly informative.

It is possible to have an objective measure of safety learning in animals and humans.

People with post-traumatic stress disorder do not respond appropriately to safety signals.

In research studies, D-cycloserine facilitates fear extinction and may play a role in enhancing the effects of CBT.

Understanding the neurobiology of fear and safety circuits can help us develop novel treatments.

Here is our newly wed PTSD patient describing his symptoms and their triggers.

"I can't get the memories out of my mind! The images come flooding back in such vivid detail, and they're triggered by the most inconsequential things, like a door slamming or the smell of stir-fried pork."

"Last night, I went to bed and I was having a good night's sleep for a change. And then in the early morning the storm-front passes through, there's this big bolt of crackling thunder. I woke up instantly, frozen in fear."

"And I'm right back in Viet Nam, I'm in the middle of the monsoon season at my guard post. And I am sure that I'll get hit in the next volley and I'm convinced I will die."

"My hands are freezing, yet sweat is pouring from my entire body. I feel each hair on the back of my neck standing on end. I can't catch my breath, my heart is pounding. And I smell a damp sulfur smell. The next bolt of lightning and clap of thunder makes me jump so much that I fall to the floor."

Several Questions Remain

Will this treatment work for PTSD?

Will our newly wed PTSD patient learn to generalize the safety signal:

Safe time, safe place, safe context despite the complexity of triggers that bring back his fear response?


Adolphs R, Tranel D, Damasio H, Damasio A. 1994. Impaired recognition of emotion in facial expressions following bilateral damage to the human amygdala. Nature 372:669 - 672.

Davis M, Ressler K, Rothbaum BO, Richardson R. 2006. Effects of D-cycloserine on extinction: translation from preclinical to clinical work. Biol Psychiatry, 60(4):369-375.

Ekman P, Friesen WV. Pictures of Facial Affect. Consulting Psychologists Press, Palo Alto, CA, 1976.

Guastella AJ, Richardson R, Lovibond PF, Rapee RM, Gaston JE, Mitchell P, Dadds MR. 2008. A randomized controlled trial of D-cycloserine enhancement of exposure therapy for social anxiety disorder. Biol Psychiatry, 63(6):544-549.

Jovanovic T, Keyes M, Fiallos A, Myers KM, Davis M, Duncan, EJ. 2005. Fear potentiation and fear inhibition in a human fear-potentiated startle paradigm. Biol Psychiatry, 57(12):1559-1564.

Jovanovic T, Norrholm SD, Blanding NW, Davis M, Duncan E, Bradley B, Ressler KJ. 2010. Impaired fear inhibition is a biomarker of PTSD but not depression. Depression and Anxiety, 27(3):244-251.

Jovanovic T, Norrholm SD, Myers KM, Davis M, Duncan EJ. 2011. Conditioned inhibition in healthy veterans in Croatia. Manuscript in preparation.

Kessler RC, Chiu WT, Demler O, Walters EE. Prevalence, severity, and comorbidity of twelve-month DSM-IV disorders in the National Comorbidity Survey Replication (NCS-R). Archives of General Psychiatry, 2005 Jun;62(6):617-27.

Kim JJ, Jung MW 2006. Neural circuits and mechanisms involved in Pavlovian fear conditioning: A critical review. Neuroscience and Biobehavioral Reviews, 30:188-202.

Kushner MG, Kim SW, Donahue C, Thuras P, Adson D, Kotlyar M, McCabe J, Peterson J, Foa EB. 2007. D-cycloserine augmented exposure therapy for obsessive-compulsive disorder. Biol Psychiatry, 62(8):835-838.

LeDoux JE. 1994. Adapted from: Emotion, Memory, and the Brain. Scientific American, 270(6):50-57. Retrieved August 25, 2011 from New York University Center for Neural Science Web site:

LeDoux JE. Memory and Emotion. Retrieved August 25, 2011 from New York University Center for Neural Science Web site:

Myers KM, Davis M. 2007. Adapted from: Mechanisms of fear extinction. Mol Psychiatry, 12(2):120-150.

Nestler, Eric. Adaptation of illustration.

Norrholm SD, Jovanovic T, Olin IW, Sands LA, Karapanou I, Bradley B, Ressler KJ. 2011. Fear extinction in traumatized civilians with posttraumatic stress disorder: relation to symptom severity. Biol Psychiatry, 69(6):556-563.

Otto MW, Tolin DF, Simon NM, Pearlson GD, Basden S, Meunier SA, Hofmann SG, Eisenmenger K, Krystal JH, Pollack MH. 2010. Efficacy of d-cycloserine for enhancing response to cognitive-behavior therapy for panic disorder. Biol Psychiatry, 67(4):365-370.

Phelps EA ,LeDoux JE. 2005. Adapted from: Contributions of the amygdala to emotion processing: from animal models to human behavior. Neuron, 48(2):175-187.

Ressler K, Jovanovic T, Blanding, Norrholm S, Duncan E. Grady Trauma Project. Unpublished manuscript.

Ressler KJ, Rothbaum BO, Tannenbaum L, Anderson P, Graap K, Zimand E, Hodges L, Davis M. 2004. Cognitive enhancers as adjuncts to psychotherapy: use of D-cycloserine in phobic individuals to facilitate extinction of fear. Arch Gen Psychiatry, 61(11):1136-1144.

Wilhelm S, Buhlmann U, Tolin DF, Meunier SA, Pearlson GD, Reese HE, Cannistraro P, Jenike MA, Rauch SL. 2008. Augmentation of behavior therapy with D-cycloserine for obsessive-compulsive disorder. Am J Psychiatry, 165(3):335-341.


NIMH is grateful to Michael Davis, Ph.D., Emory University, for his significant contributions to the development of this module.

Rat brain image courtesy of Miles Herkenham, NIMH. Stock photos from iStockPhoto.