PET Theories

Researchers at UCLA have reported that psychotherapy for obsessive-compulsive disorder, or OCD, results in changes in nervous system functioning so marked that they are visible on PET scan, a form of radiograph that measures energy usage in different regions of the brain. This finding extends a 1992 study comparing psychotherapy to the antidepressant Prozac in the treatment of OCD; the study found that the two treatments, when effective, result in similar PET scan changes.

The press has responded to this news as if it were proof that psychotherapy really works for mental illness. But we already knew that. We have known it for decades from studies that look not at brains but at people. The therapy used at UCLA employed behavioral techniques--exposing patients to anxiety in settings in which they cannot enact their compulsive rituals--along with "talk" therapies, individual and group, that appeal more directly to patients' reason. That bundle of therapies was chosen because it is known to ameliorate OCD, based on observations of symptom reduction--the same criteria used to measure the efficacy of medications.

Indeed, the thrust of the UCLA study can be seen as going in the opposite direction: Psychotherapy sheds light on the biological aspects of illness. Because we already know that behavioral therapy works for OCD, the changes it produces on radiographs serve as suggestive evidence that certain parts of the brain are involved in the illness, and in recovery from illness. Even if, in time, those parts of the brain prove to be only incidentally implicated in OCD, the PET scan abnormalities, and their reversal, stand as 'biological markers' of the illness and of a dimension of its response to treatment.

The PET scan study is part of a useful blurring of boundaries within psychiatry. In the same journal in which the UCLA results appeared, researchers from the University of Pittsburgh looked at a different biological marker, brain-wave activity of depressed patients during sleep. The researchers found that abnormal brain-wave profiles identify patients who, when treated, respond only infrequently to a variety of psychotherapy that is often effective for depression. Such patients might be candidates for medication sooner rather than later.

So, in limited areas, we may now have biological markers that either predict or confirm the outcome of psychotherapy. These preliminary results point to a future psychiatry in which physiology and psychotherapy are seamlessly integrated, each aspect informing the other. Biological psychiatry will increasingly put psychotherapy to use. Psychotherapy will take biological markers into account. Diagnosis will involve assessment of physiological and psychological strengths. Treatment will entail skill in combining modalities.

As for whether psychotherapy works--whenever this question arises I think of the behaviorist B. F. Skinner and "Project Pigeon." Early in World War II, Skinner proposed that pigeons help guide bombs or missiles. Harnessed and wired birds would be trained to peck a transparent plate as a bomb fell, to administer mid-course corrections. The Pentagon quickly abandoned pigeons in favor of electronics. But at the war's end, electronic guidance was less accurate than the birds had been when the military canned Skinner's research.

Psychotherapy is like the pigeons. It works fine, but it gets no respect, because it involves flesh and blood, not hardware that resonates when you kick it. There is plenty of evidence that psychotherapy is effective. There is even reason to believe, on technical grounds, that outcome research, positive though it is, underestimates the effects of psychotherapies, particularly less structured varieties. But it seems that no amount of research will give psychotherapy status as a medical technology-unless that research itself employs high-tech tools, like PET scanners. Psychotherapists have long feared that biological psychiatry would be their ruination; if they can be open to the new research, it may be their salvation.

Beyond providing a reason to believe, there are grounds for hoping that biophysics will help to refine psychotherapy. Sensitively handled, information about biological markers might help answer questions about anomalous responses to treatment, or give hints about expectable outcomes. To practice psychotherapy in good faith is to embrace the science of one's time. Still, it is hard not to harbor a romantic resistance, one that continues to hope that aspects of psychotherapy will remain beyond the reach of measuring instruments, that art will inform science more than the reverse.