February 14, 2024

expert reaction to systematic review and network meta-analysis on the effect of exercise on depression

A study published in The BMJ looks at the effect of exercise on depression. 

Prof Mark Hamer, Professor of Sport and Exercise Medicine, University College London, said:

“Contrary to expectations, studies describing a level of participant autonomy (ie, choice over frequency, intensity, type, or time) tended to show weaker effects than those that did not. This suggests targeted exercise prescription is important for better outcomes.

“Also the overall effects were similar for individual and group based exercise – this is notable as in the past we’ve thought that anti-depressive effects of exercise can possibility be attributed to increased social contact during the interventions but this clearly isn’t the case. Effects are being driven by exercise itself.

“The effects are stronger for higher intensity exercise. However, at a population level we know that participation in higher intensity activity, such as sport/gym classes , is very low (<20% of population). Therefore, we need to think about how these types of intervention can be practically implemented so that people will actually do them.”

Prof Gavin Sandercock, Professor in the School of Sport, Rehabilitation and Exercise Sciences, University of Essex, said:

“Rather than ‘a’ study this paper is a meta-analysis (systematic review) which includes all of the relevant studies in this area.  The studies included in the analysis are of high quality and provide robust findings. When many studies are combined in this way the results represent the most robust evidence scientific research can provide.”

Prof Jonathan Roiser, Professor of Neuroscience and Mental Health and Neuroscience and Mental Health Group Leader, University College London (UCL), said:

“This paper is a review of over 200 clinical trials examining the effect of physical activity on depression. It reports an aggregate analysis (called a “network meta-analysis”) which averages over all relevant studies, including those comparing physical activity to other established treatments, as well as to different types of control condition or no treatment. The headline result is that all types of physical activity (especially aerobic exercise – which gets you sweaty and out of breath), cause a reduction in depressive symptoms; a similar conclusion to many reviews over the past decade. The authors report a particularly large effect of dancing, but this should not be taken too seriously owing to the low number of studies included (5) each of which was quite small. Smaller effects (although still clinically important) were observed for milder types of activity, such as stretching or tai chi. There were also some possible sex differences (e.g. men experiencing more benefit from tai chi) but this analysis was quite crude, as it was based on the average gender mix in each study and not individual characteristics. One striking finding was that the effect of aerobic exercise appeared to be numerically larger than for SSRIs, although there weren’t many head-to-head studies and so this result needs to be interpreted cautiously. It’s also important to bear in mind that blinding always poses a challenge in studies of physical activity as participants obviously know what activity they are doing, although the authors report that even taking this and other biases into account the results are robust overall. The authors conclude that exercise should be considered seriously by doctors when treating depression, and I would agree. Physical activity is actually currently included in the UK NICE guidelines for depression, but unfortunately it is rarely prescribed in practice, even though we know it works. What is needed is a cultural shift in the way we think about treating depression which emphasises the important role that physical activity can play, alongside standard treatments like medication and therapy. And as the authors say, we also need a lot more research in understanding how exercise works for depression, as this could potentially increase people’s willingness to consider it, and also provide important clues for the development of new treatments.”

Dr Paul Keedwell, Consultant Psychiatrist and Fellow of the Royal College of Psychiatrists, said:

“This is a welcome review of the evidence supporting exercise as part of a holistic programme of treatment for depression. The power of exercise to lift mood is often overlooked. 

“Social interaction might be almost as important as the physiological effects of exercise (with group activities such as yoga, dance and walking groups being particularly helpful), and context is probably important too, with additional benefits to be gained in green and natural environments. 

“Some caution is needed in interpreting the findings. Many studies had small sample sizes and were not conducted in real world conditions. Also, many depressed individuals find exercise very challenging. But, taken together, the evidence supports exercise being an important part of a package of treatments for depression, and it will help with physical health too. Depression is often associated with increased medical morbidity due to poor diet and inactivity.” 

Prof Michael Bloomfield, Head of Translational Psychiatry Research Group and Consultant Psychiatrist, University College London (UCL), said:

“There is lots of evidence that exercise and behavioural activation can be helpful for mental health and specifically for depression. Whilst this paper adds to the evidence that exercise is good for you, it is not possible to say at this stage that exercise on its own is better than existing treatments including medication and psychological therapies. As the authors acknowledge, many of the studies of exercise were at risk of bias. Furthermore, when people experience more severe forms of depression simply offering exercise may not be completely helpful, for example, when someone is struggling to get out of bed let alone get to the gym. Depression is often caused by a range of factors and it is more helpful to think about the pros and cons of a range of treatments including medication, talking therapy and exercise, that can be offered to patients, rather than an unhelpful division of medication vs. talking therapy vs. exercise.”

Prof David Curtis, Honorary Professor, UCL Genetics Institute, University College London (UCL), said:

“This seems to be a high quality study which brings together the findings of many clinical trials to conclude that different modalities of exercise are perhaps surprisingly effective as treatments for depression. It may be important to note that the trials compared people allocated to an exercise intervention against controls. Presumably, if one considered just the people within that group who actually complied with the exercise program the effects might be even greater.

“Exercise seems to be a helpful treatment for many different kinds of people with depression. The only slight caveat is that all the studies would have included only people who consented to be possibly randomised to receive exercise as a treatment. It would not have included people who were unwilling or unable to even contemplate exercise.

“Even moderate forms of exercise, such as walking, produced benefits, although the effects do seem to be stronger with more vigorous exercise. Overall, the effects of exercise seem as strong or stronger than for antidepressants or standard psychological treatments. Thus, the authors’ suggestion that exercise should always be recommended as part of the treatment package seems sensible.

“The report is clear and well written and contains many interesting features. It is available Open Access, which means that it is freely available for anybody to read on the internet, and I would recommend that anybody interested should look at it themselves.”

‘Effect of exercise for depression: systematic review and network meta-analysis of randomised controlled trials’ by Michael Noetel et al. was published in The BMJ at 23:30 UK time on Wednesday 14^th February.

DOI: 10.1136/bmj-2023-075847

Declared interests

Prof Jonathan Roiser: No relevant COIs

Dr Paul Keedwell: No conflict of interest.

Prof Michael Bloomfield: I’m a director of Bloomfield Health Limited.

Prof David Curtis: I have no conflict of interest.

For all other experts, no reply to our request for DOIs was received.