Why Cognitive-Behavioral Therapy Helps
Like Drugs, Talk Therapy Can Change Brain Chemistry
by Richard A. Friedman, M.D.
After six years of twice-weekly psychotherapy sessions, Eric had plenty of insight.
But his anxiety level had barely changed.
He was still bedeviled by a ceaseless urge to wash his hands and shameful
and repetitive violent thoughts. Out of desperation and against the wishes of
his therapist, he visited my office to discuss the possibility of medication.
"I thought I could understand my way out of my obsessive compulsive
disorder," he recalled recently. "I wanted to be able to do it on my own,
What he did not remember was his vehement opposition to psychotropic
medication on the ground that it was not natural and would change his brain
Of course, he was right. Like Eric, many patients and therapists share the
view that psychotherapy (such as cognitive-behavioral therapy) is preferable to pharmacotherapy because it is more
"natural" and because it supposedly gets to the root of the patient's problem.
They are convinced that self-understanding will bring relief, whether the
problem is anxiety, depression or obsessional thinking.
Insight is a prerequisite of happiness, the theory goes, and well-being
achieved without the hard work of psychotherapy is artificial and inauthentic.
But new evidence suggests that the talking cure and psychotropic
medication have much more in common than had been thought. In fact, both produce
surprisingly similar changes in the brain.
Take Eric's obsessive compulsive disorder. It hobbles patients with
unwanted thoughts, often violent or sexual, that play in the mind like a broken
record. Owing to the sometimes lurid nature of the thoughts, the treatment
mainstay had for years been psychoanalytically-oriented therapy to unlock the
sexual and aggressive conflicts presumed to underlie the symptoms.
There was just one problem. That form of psychotherapy rarely, if ever,
worked for those patients, a point now widely accepted by most psychoanalysts
But two seemingly different treatments can be highly effective: a form of
talk therapy called cognitive-behavior therapy and a class of antidepressants
called selective serotonin reuptake inhibitor antidepressants, or S.S.R.I.'s,
drugs like Prozac and Zoloft. It is well known that patients with the disorder
have altered serotonin function compared with normal controls.
Brain imaging that uses PET scans, or positron emission topography, has
shown that the disorder is associated with functional hyperactivity of the
caudate nucleus, a structure buried beneath the cerebral cortex. Some
researchers hypothesize that the caudate is part of a subcortical circuit that
acts as a kind of filter, sifting out extraneous thoughts and impulses.
In obsessive compulsive disorder, they theorize, the subcortical filter
malfunctions, allowing the unwanted thoughts to reach the cortex and then on to
In a study by Dr. Lewis Baxter at the U.C.L.A. School of Medicine,
patients with the disorder who responded to either a reuptake inhibitor like
Prozac or cognitive behavior therapy over 10 weeks showed virtually the same
changes in their brains, decreases in the activities of the caudate nuclei and,
thus, changes toward normal function.
When patients improved, the changes in their brains, as shown in the PET
scans, looked the same regardless of whether they had received antidepressants for OCD
or psychotherapy for OCD.
An S.S.R.I. works, in part, by enhancing the neurotransmitter serotonin,
whose activity is often abnormal in people with obsessive compulsive disorder
and depression. Cognitive behavior therapy focuses on changing distorted
patterns of thinking.
The intriguing finding from the PET scans is not limited to O.C.D. Two
studies of patients with depression, reported last year in The Archives of
General Psychiatry, compared the effects of interpersonal psychotherapy with an
antidepressant on brain function, as observed in PET scans. In those studies,
the depressed patients received interpersonal therapy, a short-term talk
treatment that focuses on the effects of social relationships and major life
events on mood.
In one study, a 12-week trial that compared an S.S.R.I., Paxil, to
interpersonal psychotherapy, Dr. Arthur Brody, also at U.C.L.A., found that
depressed patients who responded to either treatment had nearly identical
changes in their brain function, a decrease in the abnormally high activity seen
in the prefrontal cortex before treatment.
In the second study, Dr. Stephen D. Martin at the research unit of Cherry
Knowle Hospital in Sunderland, England, reported that six weeks of Effexor, an
antidepressant that enhances both serotonin and norepinephrine, and
interpersonal therapy produced similar effects in those depressed subjects who
responded either to medicine or to psychotherapy. Each had shown an increase in
the activity of the basal ganglia, a subcortical brain structure.
Although the observed changes with psychotherapy and antidepressant were
similar in that study, they were not identical. Subjects with interpersonal
therapy but not Effexor also had activation of a brain area called the cingulate
gyrus, which responds to serotonin in the brain and has a role in regulating
The studies show that pharmacotherapy and psychotherapy can produce
remarkably similar effects on functional brain activity. But does that mean that
antidepressants and psychotherapy are really equivalent?
In a word, no. Psychotherapy alone has so far been largely ineffective for
diseases like schizophrenia, where there is strong evidence of structural, as
well as functional, brain abnormalities. So it seems that if the brain is
severely disordered, then talk therapy cannot alter it.
But it is clear that talk therapy can alter brain function. The reason may
come from the elegant work of a Nobel Prize-winning psychiatrist and
neurobiologist, Dr. Eric Kandel. Studying the simple and well-mapped nervous
system of a sea slug, Aplysia, Dr. Kandel showed that learning leads to the
production of new proteins and, in turn, to the remodeling of neurons.
Sea slugs exposed to the controlled-learning condition that produced
long-term memory ended up with double the number of neuronal connections as the
untrained animals. In essence, Dr. Kandel has proved that learning involves the
creation of new neuronal connections.
The clear implication for humans is that learning literally changes the
structure and function of the brain.
Now it may seem a big leap from a snail to a human. But if psychotherapy
is thought of as a form of learning, then when therapists talk to patients, they
cause them to learn, perhaps changing their brain function and, perhaps, for the
In the end, Eric chose cognitive behavior therapy and improved
drastically. Through exposure to those situations that he feared like messy
dirty places, he became desensitized to them and lost his compulsion to wash.
Had he chosen an antidepressant, chances are that he would also have
If psychotherapy produces nearly the same brain changes as
pharmacotherapy, then the boundary between mind and brain is purely artificial —