Treating Anxiety:Thoughts from Neuroscience

Fyodor Dostoyevsky’s Crime and Punishment, published in 1866, contains a wonderful passage about psychiatric treatment:

“There have already been some serious experiments relating to the possibility of treating the insane by means of the simple influence of logical reasoning. . . A certain professor. . .believed that insanity is, as it were, a logical error, an error of judgement, a mistaken view of things. He would refute the arguments of his patient step by step and, would you believe it, it’s said he achieved results that way!”

I’ve tried this method—persuading an anxious patient that his worries don’t make logical sense—and I’m sorry to say it doesn’t work. The patient usually feels better during the session and for a short time afterwards, but at his next visit he’s just as anxious, and he reports he’s continued to worry and avoid whatever he’s afraid of. Patients with enough confidence to speak up soon tell me their treatment isn’t helping.

My patients benefitted when I learned about cognitive therapy. I began to teach them how to argue with their own irrational beliefs rather than doing it myself. Self-reassurance turns out to work better and last longer than calming words from anyone else, no matter how great an authority. And if the patient can be taught to take apart and disable his negative thoughts, they trouble him even less. Cognitive therapy, particularly when coupled with behavioral techniques such as exposure and relaxation, is one of our most effective treatments for anxiety disorders.

We have made progress since Dostoyevsky’s time, but cognitive-behavioral therapy, even when combined with anti-anxiety medication, is of limited efficacy, and many patients continue to suffer from paralyzing anxiety despite treatment.

One of the elements common to all effective treatment of anxiety is what behavior therapists call exposure: the therapist persuades her patient to put himself in a situation he fears. When, contrary to his expectations, he experiences no harm, his anxiety diminishes. Toward the end of the book I drew from in last week's post, Anxious: Using the Brain to Understand and Treat Fear and Anxiety, Joseph LeDoux discusses exposure therapy from a neuroscience perspective. He begins by explaining the complex brain phenomena which underlie behavioral extinction, the deconditioning of basic threat responses. Extinction is widely considered to be the basis of exposure therapy. As I noted last week, LeDoux breaks down conditioning and extinction into several distinct components, each with its own neurobiological mechanism. Some occur outside conscious awareness (nonconsciously, in LeDoux’s terminology.) And, importantly for treatment, there is evidence that the conscious aspects of therapy we all value so much may actually interfere with the nonconscious processes of extinction. These include not only language-based interventions such as disputing negative thoughts, reframing, and interpretation, but many other consciously based aspects of conditioning treatment as it is usually practiced, such as stepwise desensitization.

Because language-based interventions can interfere with nonconscious aspects of desensitization, LeDoux argues that cognitive treatments should take place after conditioning exercises which address nonconscious elements of anxiety. As a nonclinician, he does not say how to do this, but techniques have been developed to minimize conscious elements of desensitization in experimental subjects, and the psychologist Michelle Craske has translated some of this work into clinical recommendations.

He makes a good case that because conditioning is context-dependent, exposure should take place in as many different situations as possible—in the presence and absence of the therapist and other supportive people, using both imagination and real stimuli, in safe situations such as home and the therapist’s office as well as the real world, and at different times of the day.

He also shows that extinction is not simply the elimination of memories: as trauma therapists know well, even after extensive treatment, traumatic memories can be reawakened by stress or by stimuli which were not addressed during therapy. Extinction requires creation of new memories—memories that feared situations are not actually dangerous.

Here we come to some molecular mechanisms which were new to me: generation of new long-term memory requires gene expression and protein synthesis in nerve cells. For extinction of a conditioned stimulus, this takes place in the intercalated cells of the amygdala, which use the inhibitory neurotransmitter GABA (gamma-amino butyric acid), and the infralimbic region of the medial prefrontal cortex, which projects to those cells and thus inhibits emotional responses such as fear. This is one of the simplest instances of emotional regulation.

At the molecular level, this requires CREB (cyclic AMP response binding element protein,) a transcription factor which initiates the process of gene expression and protein synthesis. Exposure training uses CREB to generate new memories of safety, and extensive training can exhaust the supply of CREB so no further learning can take place. I suspect this may explain the mental fatigue patients sometimes report in sessions involving a lot of new learning. LeDoux recommends careful pacing of training sessions. And since protein synthesis takes four to six hours, he also suggests that after therapy sessions patients should avoid activities which interfere with consolidation of learning, although it is not fully clear what those activities are. Since sleep facilitates memory consolidation, a nap after a therapy session might be a good idea, and there is some clinical data to support this.

LeDoux also cites newer research on acid-sensing receptors in the brain, which respond to low pH in the cerebrospinal fluid. When the blood contains a high concentration of carbon dioxide (CO2,) the resulting acidity stimulates pH receptors on neurons in the amygdala and the bed nucleus of the stria terminalis. These neurons become more excitable and responsive to threats. That suggests to me that activities such as exercise, which increase blood CO2, might interfere with the process of learning to be less anxious, but the neurobiology of extinction is so complex that clinical data would be needed to support any recommendations based on this.

LeDoux presents findings, then, which may be clinically important: anxiety treatment needs to work at more than one level, and the top-down language based treatments may actually interfere with the nonconscious learning necessary to reduce conditioned responses to threatening stimuli. And since learning requires protein synthesis, pacing of therapy is important, and quiet time or sleep after intense therapy sessions may help.

In Persuasion and Healing, his classic study of how psychotherapies work, Jerome Frank shows that regardless of the therapist's technique, effective psychotherapy for anxiety disorders requires that the patient actively face the situations he fears. LeDoux’s work on conditioning advances our understanding how such mastery of fear takes place. Frank identified other common elements in effective psychotherapies, including the expectation of help and the therapeutic relationship. LeDoux’s work does not address these important aspects of treatment, and his findings, while important, need to be integrated with other neuroscientific findings, such as those about how hope and empathy contribute to therapy.

Further reading: Joseph LeDoux, Anxious: Using the Brain to Understand and Treat Fear and Anxiety, 2015.

Michelle G. Craske, Michael Treanor, Chris Conway, Tomislav Zbozinek, and Bram Vervliet, Maximizing exposure therapy: an inhibitory learning approach. Behav Res Ther 2014 July; 58:10-23.

Jerome D. Frank and Julia B. Frank, Persuasion and Healing: A Comparative Study of Psychotherapy, 3rd Edition, 1993.

The Dostoyevsky quote is from the David McDuff translation of Crime and Punishment, Penguin, 1991, page 506.

Recent Posts