Exercise Blog Series, Part 2: Feeling Depressed? Ask your doctor about exercise.

Part 2 of a Series By: Jamie Foehl and Zachary Zenko, PhD


You’re probably familiar with the notion that exercise might help with depression. Maybe you’ve seen an article or headline that starts off with something like: “There’s evidence suggesting that exercise can help with depression.” Have you ever wondered what that “evidence” is? What did researchers DO to learn more about this idea? The goal of this post (and blog series) is to dig into some of the research that led to these claims in the first place.


Exercise for the Treatment of Depression


Let’s look at a seminal[1] study on exercise and depression – conducted by Blumenthal and colleagues in 1999 (b). The type of depression we’re talking about for this study is a major depressive disorder (MDD) which is different from the occasional blues or being sad about a loss[2]. Some people refer to MDD as “clinical depression”.

The researchers wanted to compare the therapeutic effect of exercise compared with the more standard treatment of antidepressants. They also wanted to see the therapeutic effect when people took antidepressants and exercised.

What do we mean by “therapeutic effect”? It means that we wanted to see if people’s depression got better, worse or stayed the same depending on if they were treated with a) antidepressants, b) exercise c) antidepressants + exercise. In other words, did the depression change at all?

To be able to say that people experienced a change in depression there needed to be a measure of depression. Everyone in the study met the criteria for a diagnosis of MDD. The researchers needed a measure that would help them assess the severity of the depression. Fortunately, there were two valid tools for measuring depression severity: the HAM-D and the BDI.

A valid tool refers to something that measures the things it’s trying to measure. For example, a basic bathroom scale is a valid measure if you want to find out your weight. The scale is not a valid measure if you wanted to determine how many days are left in the year.

The Hamilton Rating Scale for Depression (HAM-D) is a clinical rating scale. It’s a list of seventeen items and a clinician must determine how to rate the person with MDD on each item. The “items” look psychological symptoms (e.g. feelings of guilt) and somatic (having to do with the body, not the mind) symptoms. Each item is rated on severity. For example, when it comes to “agitation” someone who is “fidgety” will get a lower score than someone who is hair-pulling and lip-biting.

Using the HAM-D requires some judgment from the clinician. For this reason, two clinicians use the tool independently and their assessments are compared. This is done to ensure reliability. In a research measurement context, reliability refers to a tool that provides consistent and reproducible. Imagine you step on a scale and it shows you weigh 150 pounds. Then you step on it again and you weigh 200 pounds. That scale would be unreliable. On the other hand, if it consistently shows that you weigh 150 pounds, it is reliable.

The other measurement tool they used to measure depression is the Beck Depression Inventory (BDI). It’s similar to the HAM-D except it’s self-reported: It’s filled out by the person with depression. The researchers look at (and compare) the self-report (BDI) and the clinician report (HAM-D) as a check. Other measures were taken, but changes in the BDI and HAM-D were the main outcomes researchers looked at.

Everyone was randomly assigned to one of three groups: receive pills, exercise or pills + exercise. “Randomly assigned” means that there were no criteria or input into what group the person was put in. This way, nobody can say that one group had a different outcome because of the people who were put in that group were different to begin with.

The people in the “pills” group took the same antidepressant for 16 weeks. The people in the exercise group attended three supervised exercise sessions each week for sixteen weeks. The combination group did both. Every week, HAM-D and BDI scores were measured and recorded.

Four months is a long time, and not everyone stayed in the program, attended all the exercise sessions, took a pill every day. The reality of field research is that other things will get in the way of participation and adherence. Researchers keep careful track of these behaviors and they have ways to account for them in the analysis.

At the end of four months the researchers compared results across the groups. How did the groups compare?

There were no differences between groups.

In other words, when you looked at outcomes (remember, outcomes for this study are HAM-D and BDI scores) from everyone in the study, you couldn’t tell the difference between someone who took pills only from someone who only exercised! Furthermore, across all groups at least 60% of people went into remittance – that means they no longer had the MDD diagnosis.

That’s pretty remarkable. Most people know that exercise is a good idea when it comes to a healthy lifestyle. The idea, however, that exercise is just as effective as medication and could put depression into remission? That probably comes as a bit more of a surprise.

Keep in mind, however, that the results of the study are based on where the people were at four months. That’s just one point in time. What happened in the different groups during the study? And, importantly, what happens to people after four months, when they’re no longer in the study? During the study, people were provided with exercise support, check-ins and were closely monitored. Those are important conditions for research but don’t exactly mirror the circumstances of the broader population.

Let’s start with what happened during the four months of the study. There were some differences in how rapidly the different groups responded to different treatments during the sixteen weeks. People in the “pills only” group did have a more rapid initial response. 


Exercise for the Treatment of Depression: Persistent Effects


Six months after the study ended, the researchers followed up with the participants. They were particularly interested in cases where the MDD had gone into remission (remember, after four months, roughly 60% of people in the study were in remission). The authors refer to this group as the “remitted subjects.” Overall, they saw more people had recovered than relapsed. That said, they found much lower relapse rates from “remitted subjects” in the exercise group compared to the medication group (a).

If exercise might be comparable to medication, and might be even better to prevent relapse, why aren’t more people exercising? For one thing, it is much easier to take a pill than it is to exercise (this may be particularly relevant for people who are depressed, though some studies indicate that exercise is well tolerated by people with depression and exercise should be a feasible treatment (d)). Secondly, exercise isn’t quite (yet!) viewed as something that a doctor could prescribe the same way that medication is prescribed. After all, you can’t pick up a gym membership at the pharmacy. For many people knowing that exercise is “healthy for you” isn’t enough reason enough to do it.

Furthermore, the exercise doesn’t have the same marketing budget as the pharmaceutical industry! Our message to the public, however, is the same as that of the pharmaceutical industry – ask your doctor about exercise as a treatment option.

The notion (of exercise being comparable with medication) is gaining traction (c), particularly in light of the rising costs of healthcare. Getting people to ask their doctors about exercise is a powerful step towards considering exercise on par with medication.



  1. Babyak M, Blumenthal JA, Herman S, et al. Exercise treatment for major depression: Maintenance of therapeutic benefit at 10 months. Psychosom Med 2000;62(5):633–638.
  2. Blumenthal JA, Babyak MA, Moore KA, et al. Effects of exercise training on older patients with major depression. [Internet]. Arch Intern Med 1999;159(19):2349–2356.
  1. Ekkekakis P, Murri MB. Exercise as antidepressant treatment: Time for the transition from trials to clinic? Gen Hosp Psychiatry (in press).
  1. Stubbs B, Vancampfort D, Rosenbaum S, et al. Dropout from exercise randomized controlled trials among people with depression: A meta-analysis and meta regression [Internet]. J Affect Disord 2016;190:457–466.

[1] “seminal” refers to studies that really influenced research in a field.

[2] The exact symptoms of MDD can be found in the Diagnostic and Statistical Manual of Mental Disorders. There have been five versions! People refer to the manual as the DSM-5.