clinical depression and keeping the condition from returning, researchers say.

Scientists at Duke University Medical Center tested exercise against Zoloft, and found the ability of either -- or a combination of the two -- to reduce or eliminate symptoms were about the same. But they found exercise seemed to do a better job of keeping symptoms from coming back after the depression lifted.

"The present findings suggest that a modest exercise programme ... is an effective, robust treatment for patients with major depression,'' said the report in the October issue of the journal Psychosomatic Medicine.

The findings go well beyond the idea that exercise fights the blues. "It was not just that they had a bad day -- or a bad several days,'' said James A.

Blumenthal, lead researcher.

The patients in this study and others in a continuing series of experiments at Duke had been diagnosed with major depressive disorder. Their loss of pleasure, and feelings of worthlessness and guilt, had to exist for at least two weeks, Blumenthal said. The severity of their symptoms was generally mild to moderate, but in some cases was severe, he said.

The October report followed earlier research in which 156 adult volunteers had taken part in a four-month comparison of exercise, Zoloft or a combination.

The exercise primarily consisted of brisk walking, stationary bike riding, or jogging for 30 minutes, plus a 10-minute warmup and five-minute cooldown, three times a week.

After four months, patients in all groups had comparable improvements, the researchers said. About 60 percent of the exercisers had vastly improved or had no symptoms, compared with about 66 percent of the medication group and 69 percent of the combination group.

The new study was an attempt to see if the benefits continued after the earlier study ended. At this point, 10 months from the beginning of the project, researchers checked 133 of their old research subjects who had stuck with the programme.

Exercisers who had been in remission after four months were far less likely to see their depression return after 10 months, compared with people taking the drug or a combination therapy, the study found. Eight percent of exercisers saw symptoms come back, compared with 38 percent of those taking drugs and 31 percent getting both.

It will take more research to discover why exercise should work better than prescription drugs. But it's possible that exercisers gained a sense of control over their lives that drug-takers could not match, Blumenthal said.

It's also unclear why combination therapy did not work better than either alone, he said.

How exercise could create the benefit is not known either, Blumenthal said.

The once-popular idea that natural mood-enhancing chemicals called endorphins go up because of exercise is hard to prove because the experiment would require samples to be drawn from the central nervous system, he said. The usual method, taking endorphin levels from blood in the circulatory system, can be considered only a second-best way to find out, he said.

Zoloft helps the body to regulate levels of another brain chemical, serotonin, which is also believed to affect mood.

Exercise and depression The studies do not prove exercise relieves depression, in part because the exercisers worked out in a group, so group dynamics may have played a role, Blumenthal said. A new study will attempt to find out, by comparing supervised group exercisers with people who were simply given an exercise plan and sent home, he said.

The published findings may not apply to all depression patients, because the study worked with volunteers, Blumenthal said. Volunteers presumably would be more likely to stick with exercise than would patients who simply were told by their doctors to exercise, he said.

"It sounds like very interesting work,'' said Dr. Robert Pohl of Wayne State University in Ohio, who was not connected with the Duke project.

Still, it's possible some of the improvements in the Duke study came from patients' attempts to please the researchers, said Pohl, a psychiatry professor who treats patients with depression.

And getting people to exercise is not easy. Most Americans, depressed or not, don't work out enough. "I always try to get my patients exercising, and I can't do it,'' he said.