Facilitated physical activity as a treatment for depressed adults: randomised controlled trial
BMJ 2012; 344 doi: https://doi.org/10.1136/bmj.e2758 (Published 06 June 2012) Cite this as: BMJ 2012;344:e2758
All rapid responses
I share the concerns of the other correspondents.
Briefly:
The Beck Depression Inventory is a poor way to measure change in these subjects.
The exercise level aimed at was probably inappropriate to treat depression. There is a big difference between the level aiming to improve long-term cardiovascular health and the level likely needed for the treatment of a depressive episode.
The difference in proportions of subjects actually reaching this modest level at 4 months was only 10% between treatment and control groups. So while the intervention of providing advice and support was ineffective we certainly don't know from this whether or not exercise itself would be effective.
A number of people have commented on the press reporting, which I agree has been very unhelpful. What has not been said is that the authors of the study should bear responsibility for this misreporting. One press release from their host institution attributes this quote to Prof John Campbell: "This carefully designed research study has shown that exercise does not appear to be effective in treating depression." The authors should be invited to state what action they will take to correct this claim.
Competing interests: No competing interests
Depression and allied disorders have pathological roots in yet unexplored neuron energy transmission defects due to high threshold neurosensory and neuromotor stresses, leaving the complex neuro molecular energy transmission pathways in inertia deranged. The physical sensory stimuli theoretically appeared to be of therapeutic significance even if trial results did not indicate cure.
Competing interests: No competing interests
I am shocked that BMJ readers need to go through such poor articles and even worse that such misleading conclusions get reported in national newspapers with potentially devastating effects in an already sedentary enough population.
I wonder why in this "summary" of the findings from Chalder et al (2012) not only the points raised by Mr Neumann in the previous comment were not discussed, but also:
- The original paper reported p-values for the difference in physical activity even higher when missing numbers were taken into account (p=.11), although this was not mentioned in the present paper;
- With a drop-out rate of 36.8% and 40.2% at 8 months, it is completely worthless even trying to publish OR of increased physical activity in the intervention group;
- The use of the the 7-day recall diary opened wide doors to incorrect data collection and this was evident with the poor match of the data with accelerometer measurements, but this point was not further developed;
- With no further details on the type of exercise (weight bearing, indoor, outdoor, etc...) and only a self-reported intensity and the duration of the exercise, it would be very difficult to have any valid objective measure. Physiological adaptations only take place at higher physiological intensities for sustained exercise, not at "self-reported" high intensities.
With all these weaknesses in the study, drawing any conclusion would be inappropriate and misleading.
If the aim of the paper was "to evaluate the cost-effectiveness of a physical activity intervention as an addition to usual care as a treatment for depression" (Chalder et al, 2012), then the conclusion should be that the TREAD protocol is not effective. This does not mean that exercise is not effective as it seems to transpire.
In addition, according to the findings, it is plainly wrong to state that "The intervention increased self reported physical activity and this effect was sustained for 12 months".
Competing interests: No competing interests
Like other writers here, I was disappointed that the press almost universally rushed out a de-motivating, but attention-grabbing, headline in response to this report. But having read the report, I can't blame the press.
It is clear that what the researchers found is that meetings and telephone conversations about exercise have no beneficial effect on depressive symptoms. So having the NHS provide that regime would not be productive.
However, it is a bit of a leap to conclude that this proves that exercise is not beneficial to depressive symptoms. The report gives scant detail of the actual exercise undertaken by either group, and giving them the benefit of the doubt, this was probably not in their remit. This, though, does not excuse the apparent assumption in the report that provision of information and encouragement is the same as actual physical exercise.
Competing interests: No competing interests
I find this a very good study, but I have one question concerning bias:
In tabel 5 there is report of percentages and odds ratio of reporting physical activity. The authors have not mentioned in the text the apparent homogeneity between intervention and usual care, when it comes to reporting PA.
4 months I(ntervention)- 52% U(sual care) - 43%
8 months I 63% vs. U 49%
12 months I 58% vs. U 40%
There seem to have been an increase in PA in the control group from 28% at baseline to 43% at 4 months. Are they receiving different treatment as long as the actual treatment measured is received by 40-49% of the control-group?
How does this affect the possibility of drawing any conclusions?
Competing interests: We use PA as an alternative treatment for depression in a group- setting
This unfortunately is a well written article with a massive flaw. The concluding sentence sets it out nicely:
"Clinicians and policy makers should alert people with depression that advice to increase physical activity will not increase their chances of recovery from depression"
Why is it flawed? Because the article is evaluating one intervention type alone, finding that it is not effective, and concluding that all other possible exercise interventions will also be ineffective.
The authors must bear responsibility for the immediate and destructive effect this will have on people's attitudes to physical exercise. Whatever happened to the null hypothesis? Can the authors disprove that physical exercise has had a positive effect on mental health? I don't believe this is possible and challenge them to do this.
The introduction also unfortunately uses misleading language by describing the evidence given as "evidence on the effectiveness of physical activity" when it actually is "evidence on the effectiveness of the TREAD intervention".
Unfortunately the mass media are unlikely to read the whole article which is more balanced than the intro and conclusion, and are now going to jump to the wrong conclusions about depression and physical exercise. They also won't think about one of the main measures of success in the article, namely whether the patient reduced or stopped their antidepressants. Surely the guidance on antidepressant prescribing is that these drugs are most effective when taken over a sufficiently long period of time, say at least 6 months, therefore the cohort of patients described would have been unlikely to discontinue use even if they were feeling better following the TREAD intervention, because this would be against the accepted wisdom that treatment should continue until after depressive symptoms have subsided.
It would not have taken much editing, BMJ, to change the emphasis in these key parts of the article in such a way as to prevent the current media misinterpretation.
The media look to journals such as the BMJ to provide balanced and accurate information which will guide not only the scientific community but also, increasingly, the general public. I do not have confidence on this occasion that this has been achieved.
Competing interests: No competing interests
I was also disappointed by the way that this was reported by the media, particularly as it did not appear to take into account this study's limitations.
One of the most obvious limitations is the broadness of the definition of treatment as usual. To include exercise on prescription as one of these options appears to automatically invalidate the aim of this study. Although the intervention arm did show an increased level of exercise on the odds ratio, the level of exercise in the treatment as usual arm was still between 40-49% during the follow-up period.
The study seems to be focused more on how to encourage patients to do more exercise, rather than actually making sure that they do it. If patients were actually asked to attend a particular exercise class etc., it would be more convincing. There is no real mention about what type of exercise patients are engaging in, and by this the setting may be just as important as the level of exertion.
Finally, I am not sure exactly what the intended aim of this study is. While it might be said that this study does not show an increased improvement in depression in exercise, are we really going tell patients not to exercise? We already know that patients with depression have increased mortality for physical health problems, so if we took a holistic approach then surely exercise is still of benefit? Why do we need to try and justify it with whether it reduces their depression score more rapidly? This is not a study which looks at whether exercise should be substituted for other treatments and if a patient perceives there to be a benefit then that is still of relevance. After all, there is an acknowledged large placebo effect with antidepressants, but even this is still "an effect".
Competing interests: No competing interests
Great - another depressing headline
Those of you out there working tirelessly to get people ‘enjoying’ a more active lifestyle are, like me, unlikely to be overjoyed at today’s headlines on the BBC website’s health page.1 A nice example of bad PR in my opinion, particularly when you actually read the conclusions made by the authors of the headline source paper.2 I went a bit mad on twitter for an hour or so after I read the study but after calming down decided that a better way to get answers to my many questions would be via this letter. So, for me it would be useful to know the following:
1. How does the conclusion “advice and encouragement to increase physical activity is not an effective strategy for reducing symptoms of depression” translate to “Exercise ‘no help for depression’ research suggests’?1
2. Why does OR 1.58, 0.94 to 2.66 P = 0.08 translate to “some evidence” when OR of 0.66, 0.4 to 1.11 P = 0.12 translates to “no evidence”?
3. Why was the primary outcome of depression score based on responses given at 4 months and the success of the intervention in increasing PA levels based on combined responses given at 4, 8 and 12 months?
4. Why is a 25% drop out from an individualised treatment programme not considered worthy of further discussion?
The answer to questions 1-4 above will help me:
1. Rationalise how a study not actually assessing EXERCISE per se can produce the above headlines
2. Understand the circumstances when a non-significant effect can be evidence of some evidence or no evidence. In the paper I took P = 0.08 for meaning that there was no difference in self-reported physical activity between the two groups at 4 months, therefore meaning the intervention was not successful in increasing PA at 4 months.
3. To decide if the authors chose to use the combined PA responses as opposed to the 4 month PA as the former showed significant difference and the later did not.
4. Respond to the journal editors/reviewers why I don’t have to explain the reasons for levels of drop out/loss to follow up in my studies.
That aside, surely the headlines stemming from this are counterproductive. Even the study authors found “…some participants in the trial reported that physical activity helped to improve their mood” and concluded “...there is still a possibility that physical activity itself might have some benefits for depression”. I fear, however, that the damage may have already been done.
1. Facilitated physical activity as a treatment for depressed adults: randomised controlled trial. BMJ 2012;344:e2758
2. Jeffreys B. Exercise ‘no help for depression’, research suggests, media release, 6 June 2012, BBC news, viewed 6 June 2012, http://www.bbc.co.uk/news/health-18335173
Competing interests: No competing interests
Like the previous response (David Nunan), I despaired when reading the headlines today and the way in which this research has been reported in the media.
As a long time sufferer of depression, I can honestly say that the only thing that has always worked for me, without fail, is to get out and exercise. Noting that this is about having personal freedom and a feeling of getting back in control of one's life. I'm not sure the outcome would be the same if I had been nagged or "given encouragement" to exercise.
It would be deeply regrettable if the findings of this research were incorporated into the NICE guidelines as media reports are suggesting. Thus setting in motion a trend for exercise as a treatment option being taken away from those who genuinely benefit from it.
A few questions that I would like to ask:
1) How was the accuracy of patients' self-reporting measured?
2) Was any attempt made to distinguish between different forms of physical activity? e.g. walking or yoga may have been more relaxing and easier to do than doing sit ups and someone who plunged head first into a harder activity may not have found it enjoyable or helpful.
3) What were the effects of exercise in the short term? i.e. immediately after exercise on a day to day basis?
4) How was consistency ensured for the reporting of symptoms?
There are potentially lots of external factors which would affect how someone reports their own symptoms. People may exaggerate, underplay, or believe they feel better or worse than they did at an earlier point in time, without this being the case. There would appear to be quite a large margin for error when relying on an individual's own assessment of how they feel.
In my opinion, it would be more scientifically rigorous to have used and studied brain scans and responses to certain stimuli, or even to use the patients' (former) lives as benchmark and how well they were able to get back into their pre-depression routines.
5) What about the impact of other activities that would affect how a person feels? e.g. drinking, smoking and drugs which could numb or alleviate a person's mood. Were these taken into account and was this uniform across both groups? Is it possible that the group undertaking less physical activity spent more time doing these unhealthy activities and so felt like they were doing better than they actually were?
6) What are the effects of anti-depressants on endorphins released during exercise?
7) Were any distinctions made between different types of depression and how well each type responded to treatment or exercise?
8) Why was the fact that the group undertaking more physical activity felt slightly better, than the other group, not investigated further?
9) Is 361 people really a large enough sample to account for all the variables in something as complex as depression?
10) What was "usual care" in each of the 361 cases? One of the biggest anomalies in the study is that "usual care" may well have included exercise on prescription.
Quote: Participants in both groups were asked to continue to follow the healthcare advice of their general practitioner for their depression and were therefore free during the trial to access any treatment usually available in primary care, including the use of antidepressants, counselling, referral to “exercise on prescription” schemes, or secondary care mental health services.
Depression is an enigma for many people and I don't think this study, with so many loose ends, has helped matters. I would accuse the press coverage surrounding this article as being over-confident, ill-informed and treating depression like a one size fits all illness.
I really think this particular study should have come with a disclaimer suggesting that further research would be required to investigate other factors which were not addressed.
Competing interests: No competing interests
Re: Facilitated physical activity as a treatment for depressed adults: randomised controlled trial
The recently reported trial by Chalder et al (BMJ 2012;344:e2758) and the accompanying editorial by Daley and Jolly (BMJ 2012;344:e3181) raises questions about the value of physical activity in the treatment of depression. The trial has also been widely reported in the media (1) as demonstrating that physical exercise does not help to improve mood and that the current recommendation in the NICE guideline on Depression (2) for the use of physical exercise may not be correct. Further, it explicitly advises doctors to no longer recommend exercise to depressed patients overriding NICE guidance. The trial however has significant limitations in relation to implications for the treatment of depression that suggests considerable caution in relation to overriding current guidance.
The NICE guideline recommends the use of structured group physical activity for the treatment of sub-threshold and mild to moderate depression. The TREAD trial did not evaluate the NICE recommended intervention but aimed to increase the frequency and intensity of physical activity through the use of specially trained physical activity facilitators. The trial results have little bearing on the NICE recommendations but rather are focused on the ability of the activity facilitator to increase the frequency of a range of physical activities, including walking, to such a level that it may bring about improvements in mood. Although there was a 9% self-reported increase in such activity in the intervention arm the absence of improvement in mood may be due to the limited change in the nature and intensity of the physical activity by the facilitators relative to the control group. The relatively high levels of physical activity in the control arm (43%) may support such an explanation. Given that the facilitated increase in physical activity had no benefit, it could be argued that the relatively frequent contact (up to 12 contacts over 9 months) with the facilitator might be better focused on the direct provision of structured group exercise as recommended by NICE.
The NICE guideline, based on a review of the higher quality studies in the area (2), recommends that physical activity should be used for a population with subthreshold depressive symptoms or mild to moderate depression. The average score on the Beck Depression Inventory on entry to the TREAD trial was 32.1, which would place the majority of the trial population within the moderate to severe depression range (2,3). Evidence reviewed in the NICE guideline suggests that individuals with moderate to severe depression are unlikely to benefit from physical activity (2) and therefore, according to the guidance, the intervention was not appropriate for probably the majority of TREAD participants. The NICE guideline is also clear that physical exercise should be offered as a ‘low intensity intervention’ to people for whom anti-depressant medication would not be the first choice treatment. In TREAD 59% in the intervention arm and 53% in the control arm received antidepressant medication which was most likely responsible for a significant proportion of the improvement seen in both arms of the trial.
In considering the implications of the trial the authors conclude that when referring to people with depression that an ’increase in physical activity will not increase their chances of recovery from depression’ (see section on “What this study adds’)). In our view, this is a premature and potentially risky conclusion that could deprive patients of effective modes of intervention. TREAD findings suggest that for moderate to severe depression the combination of the facilitation of physical activity and antidepressants may be no more effective than antidepressants alone. For this group of patients NICE recommends the combination of antidepressants with cognitive behavioral therapy (CBT) or interpersonal therapy. Such combination treatments are very likely to produce a significant improvement (typically with recovery rates of over 50% (2)) on the poor ‘recovery’ rate of 28.2% at four months reported in the intervention arm of the TREAD trial. Antidepressants alone are also likely to produce a significant improvement on the results of TREAD (typically 35-45% recovery (2)).
In summary, expecting para-professional staff using a previously untested approach to increasing physical activity to bring about significant improvements in mood in a moderate to severely depressed group of patients was unrealistic. They should have been offered treatments, such CBT combined with antidepressants, which are effective. We believe the evidence still supports the use of exercise in subthreshold and mild to moderate depression, despite claims to the contrary, TREAD adds little to that evidence.
References
1. http://www.bbc.co.uk/news/health-18335173
2. Anderson, I. Pilling, S, Barnes, A. et al (2009) “Clinical Practice Guideline No.90: Update: Depression in Adults in Primary and Secondary Care (Update)” Gaskell/British Psychological Society. London
3. Beck, A. T., Steer, R. A., & Brown, G. K. (1996). Manual for the Beck Depression
Inventory–II. San Antonio, TX: Psychological Corporation.
Competing interests: SP receives funding from NICE for the development of clincial guidelines. IA chaired the upadte of the NICE Depression Guideline published in 2009.