March 2011
Reviewer: Jeff C. Huffman, MD
Exercise and pharmacotherapy in patients with major depression: one-year follow-up of the SMILE study
Hoffman BM, Babyak MA, Craighead WE, et al
Psychosom Med 2011; 73(2):127-133
Background: A prior report from the SMILE study found that exercise and sertraline were equally effective in treating major depressive disorder in sedentary adults over a four-month study period. Further assessment of exercise as an antidepressant during follow-up had not been assessed in this or other studies.
Methods: The SMILE study was a randomized trial of 202 sedentary persons (mean age 51) with MDD. In that trial, subjects were randomized to supervised exercise, home-based exercise, sertraline, or placebo antidepressant, for a four month period. The primary analysis of that study (i.e., outcomes at 4 months) found that exercise and sertraline were comparable in their effects on MDD and were both better than placebo; overall, 46% of subjects had reached remission. Subjects who completed the trial then were told of their study condition (e.g., if they got a placebo), and were offered an exercise prescription, consultation with a study psychiatrist for medication, or both. Subjects were then assessed after a 1 year naturalistic post-participation period, with depressive symptoms measured by HAM-D, and exercise/antidepressant prescription measured by self-report. Primary outcome variables were continuous HAM-D scores, and depression status (depressed/partial remission/full remission), compared by initial group assignment and ongoing treatment(s) over the follow-up period.
Results: A total of 172 subjects (85% of original cohort) provided follow-up data. By the end of the 1-year follow-up, 66% of subjects had reached remission, and there was no significant effect of initial treatment group on depression scores or remission status after 1 year follow-up. In addition, 36% of subjects reported taking antidepressants, but antidepressant use was not associated with remission at follow-up. However, 50% of subjects reported at least some weekly exercise, with increasing amounts of exercise (until reaching 180 minutes/week) associated with lower depression scores and greater chance of remission; of note, on a secondary analysis, the effects of exercise on depression scores was the most pronounced for subjects who had the highest levels of anxiety on study entry.
Commentary: Though somewhat limited by the naturalistic nature of the follow-up, the self-report nature of exercise assessment, and the single assessment point, these are interesting findings. It is surprising that use of antidepressants was not associated with greater rates of remission at follow-up, though agents (and presumably doses) varied considerably. The finding that exercise, in contrast, was associated with improvement of depressive symptoms and with a greater likelihood of remission is indeed the main finding here, and is made more interesting by the fact that exercise’s effect was most pronounced in the most anxious patients, a subpopulation for whom remission is generally harder to achieve. Regarding this connection between exercise and depression remission, it does seem difficult to make a causative link—what if patients who exercised more did so because their depression had improved (via antidepressants or spontaneous resolution), and exercise was more of a marker of recovery than an antidepressant intervention?
Overall, though, these findings may make sense clinically. It is hard for many of us as clinicians to prescribe exercise alone as an initial antidepressant, especially in depressed medically-ill patients who may have substantial problems with function, energy, and motivation as a result of having both depression and medical illness. This can make exercise (while doubly important in this cohort) doubly hard to initiate in this population initially. However, if patients can achieve remission or at least some improvement in depressive symptoms, or even have some distance away from their most acute medical symptoms, exercise as a second-step intervention (either to maintain remission or to achieve it once some symptoms may have improved with medications/time), may be a very nice intervention to improve mood symptoms and to provide the multitude of medical benefits associated with exercise. These findings, of course, are also another indirect plug for cardiac rehabilitation and related programs that can simultaneously address mood symptoms and health behaviors.