How Does Exercise Treat Depression?

Doctors know it helps, but it's hard for researchers to prove.

Posted Jul 25, 2014

These people must be happy...right?

During my residency days in the 1980s, every psychiatrist’s office came outfitted with two ashtrays, one for the patient and the other for the doctor, and therapy was often conducted in a nicotine haze. Sometimes a cigar was just a cigar--though a pipe or cigarette would do just as well.  

At the end of the day I’d come home reeking of smoke. It was hard to fight that culture: psychoanalytic supervisors would chew you out if you asked patients not to smoke in session!

Now, thirty years later, the ashtrays have disappeared. Therapists--even psychoanalysts--commonly teach their patients meditation techniques, and steer their patients toward exercise, yoga, and healthy food. Smoking in your therapist’s office? It’s almost unimaginable!

It’s hard to argue with clean living.  

Yet, one has to wonder: do exercise, meditation, yoga and the like really make a difference in the outcome of psychiatric disorders? Or are therapists just piggybacking onto the zeitgeist, so to speak, copycatting the advice of cardiologists and sports physicians?

Pick one of these common recommendations, exercise. Does exercise work as an antidepressant?  Is it something that is just good for you or can it be a useful antidepressant treatment modality, either by itself or combined with other treatments? And if it does work, who is it best for?

Beginning in the 1980s and 1990s a few small studies suggested that exercise had significant antidepressant effects. But it has only been recently that there have been enough well-designed outcome studies to begin to answer this question. And interestingly, in our New Neuropsychiatry era, brain imaging and other biological studies are also beginning to explain the ways in which it might work, and for whom.

And the verdict? Is exercise an antidepressant?

Yes, sort of, and for some people. Similar to studies of antidepressant medications, studies of exercise often, frustratingly enough, don’t show as much of an effect as seems to be apparent in clinical practice.

There are many negative studies. As reviewed by Schuch and de Almeida Fleck, large randomized trials (TREAD, TREAD-UK, DEMO, DEMO-II) have often “failed to find any antidepressant effects of exercise.” They say, “considering these recent results the answer to the question ‘is exercise an efficacious treatment for depression?’ the answer appears to be ‘No.’”

On the other hand, some meta-analyses have also failed to show that antidepressant medicines are effective either.  [A meta-analysis is a type of analysis of the results of multiple studies which attempts to determine whether a particular type of treatment is effective].

Schuch and de Almeida Fleck speculate that the negative results, rather than proving exercise isn’t helpful, may instead reflect problems with how the studies are conducted. They may result from the “heterogeneity of depression”--that is, the many different types of depression which may arise from different causes; exercise might help some but not others. The lack of a strong finding may result from using outcome measures like the Hamilton Depression Rating Scale--a widely used scale that was developed over 50 years ago to measure symptoms among people hospitalized for depression. People living in the community don’t necessarily have the same symptoms as those who are hospitalized!

Plus, exercise is complex, involving many different types of activity, including aerobic and resistance exercise. It also can include increased socialization--and exposure to sunlight and fresh air. All of these are possible ‘confounds’ that may make it difficult to measure a true effect of exercise, and to prove that such effects actually result from exercise as opposed to other factors

For people assigned to get more exercise, it may be difficult to actually get them to do the increased exercise, especially over extended periods of time.

Finally, it’s difficult to do a truly randomized study since people applying to join an exercise study may not agree to be assigned to a non-exercise condition. They may start exercising more on their own--outside of the study.

All of these factors could contribute to studies that show less effect than therapists have long observed. In a way, the difficulty proving the effectiveness of exercise in depression is similar to that encountered in proving the effectiveness of SSRI antidepressants! There too, doctors and patients observe significant improvement as a result of treatment, and yet many studies come out negative.

Schuch and de Almeida Fleck conclude, “exercise may not be more efficacious than conventional treatments; however, it is not less efficacious.”

Hardly a ringing endorsement!

And yet...pretty much any therapist, psychiatrist, or other health care worker can tell you that exercise does seem to help many of their depressed patients.

There are a few emerging hints that particular subgroups of people may benefit more from exercise. In particular, depressed people with systemic inflammation may particularly benefit. Systemic inflammation can be measured by levels of cytokines, small molecules that are produced by white blood cells.  The recent TREAD study (TReatment with Exercise Augmentation for Depression) evaluated the benefit of exercise for people who had a partial response to SSRI antidepressant medication. Those who had elevated cytokines (circulating proteins that are measures of immune activation) were more likely to improve with exercise than those without elevated cytokines.

The proinflammatory cytokines such as TNA-alpha are elevated when people have immune activation--and can also cause symptoms of depression. C-reactive protein (CRP) is another marker of systemic inflammation, and has been linked to heart disease and a host of other problems; depressed people with elevated CRP also have a greater chance of responding to exercise.

So, in summary, exercise often does seem to help depression, though it’s often been difficult for researchers to prove! And there may be subgroups of people who have a greater likelihood of responding to exercise--particularly those with signs of systemic inflammation. In my next post, I’ll describe how researchers have worked on the issue of getting people with depression to start--and continue--to exercise, as well as evidence for the ‘dose’ and frequency of exercise needed to get an antidepressant effect.