New Hope for Battling Depression

For the 18 million American adults who suffer from depression in any given year, the highly effective antidepressant medications come with a dark side: Unless patients continue taking the drugs, they have a considerable risk of suffering a relapse in the year after they stop.

On the other hand, a growing body of research shows that patients completing a course of psychotherapy -- without medication -- relapse far less frequently.

As scientists struggle to explain the gap between the lasting effects of pills and psychotherapy, a new study offers an intriguing clue. Neuroscientists in Canada have found that a form of psychotherapy called cognitive-behavior therapy causes changes in brain activity that are the polar opposite of the changes caused by antidepressants.

The study is the first to show that a depressed brain responds differently to the two kinds of treatment. If the results hold up, they could point the way to combining antidepressants and psychotherapy more effectively to give patients a better chance of loosening the grip of depression for good.

The study found that antidepressants reduce activity in the brain's emotion centers, called the limbic system. Cognitive-behavior therapy quiets overactivity in a different region of the brain -- the cortex, which is the seat of higher thought. The study was published in the January issue of Archives of General Psychiatry.

Cognitive-behavior therapy seems to work by a top-down route, teaching patients not to ruminate endlessly about minor setbacks. Such therapy "gives patients an 'override' capacity," said psychiatrist Richard Shelton of Vanderbilt University, in Nashville, Tenn., who was not involved in the current study, so that when sad feelings bubble up from the brain's emotion centers patients can resist being sucked back into the pit of depression.

Despite the more lasting benefits of therapy, the use of antidepressants is rising steadily. Researchers at Columbia University reported in 2002 that the percentage of depression patients who were treated with antidepressant medication rose to 74.5% from 37.3% from 1987 to 1997. The percentage treated with psychotherapy fell to 60.2% from 71.1% during that time. (Some patients are treated with both.)

The growing preference for pills reflects the push by insurers to have patients treated by primary-care physicians rather than specialists. Primary-care doctors can write prescriptions but are seldom trained in psychotherapy.

The cost of psychotherapy varies by region of the country, but the national average for cognitive-behavior therapy is $100 per session for the standard course of 15-20 sessions, or about $1,500 to $2,000 for the full course of treatment. (In some cities, the price could be as much as twice that.) Branded antidepressants generally cost about $1,000 per year plus the bill for the doctor's appointments needed to continue the prescription.

Unrealistic Pessimism

In the new study, researchers led by neuroscientist Helen Mayberg of the Rotman Research Institute and the University of Toronto had 14 clinically depressed adults undergo 15 to 20 sessions of cognitive-behavior therapy. In this type of therapy, which was developed in the 1960s, patients are taught to identify aberrant beliefs -- "no one will ever love me," "nothing will ever go right for me" -- and their tendency to magnify disappointments into calamities and tragedy. Patients learn to test the accuracy of those beliefs -- by, for example, starting a friendship or applying for a job -- and see that they are often unrealistically pessimistic.

Thirteen other patients were treated with paroxetine (sold as Paxil by GlaxoSmithKline PLC). Both groups had depression of about the same severity, as measured by a standard assessment, and both experienced comparable improvement after their respective treatments. That mirrored the results of the many studies finding that cognitive-behavior therapy and antidepressants are about equally effective in treating mild, moderate and even severe depression.

Using PET (positron emission tomography) imaging, the Toronto scientists scanned the patients' brains before and after the full course of psychotherapy. "Our hypothesis was, if you do well with treatment for depression, your brain will have changed in the same way no matter which treatment you got," says Toronto's Zindel Segal, a co-author of the study. "To our surprise, the treatments operate on different regions of the brain."

"The findings make sense," says Steven Hollon, a psychology professor at Vanderbilt, who wasn't involved in the study. "Antidepressants damp down activity in the lower limbic regions, where stress and negative emotions come from. Cognitive-behavior therapy teaches the brain to respond to those signals in a healthier way, and that has a more enduring effect."

80% vs. 25%

That might explain the chief difference between antidepressants and cognitive-behavior therapy, namely, that pills exert their effect only as long as a patient takes them. A large 2001 study found that the risk of relapse in patients taking antidepressants only, in the year after they stop, is 80%. In contrast, patients receiving only cognitive behavior therapy in that study had a relapse rate of 25% in the year after ending treatment. Relapse and recurrence are measured by whether patients seek treatment again and (because only about half of patients with depression do) by standard psychological assessment of patients previously treated.

The study was funded by the Rotman Research Institute, the University of Toronto and the Canadian Institute for Health Research, a government agency. One weakness of the study is that patients were not randomly assigned to receive cognitive-behavior therapy; they expressed a preference for it over medication. That raises the possibility that the self-selected patients were better candidates for cognitive-behavior therapy than the average person with depression.

Write to Sharon Begley at sharon.begley@wsj.com

Updated January 6, 2004