“Exercise can be as good a medicine as pills for people with conditions such as heart disease,” BBC News reports, while The Times urges doctors “to prescribe exercise rather than drugs”.
Both headlines are prompted by research comparing the relative benefits of exercise and drugs for people who have serious conditions such as heart failure. But while exercise can certainly help prevent many diseases, some of the headlines have overstated the evidence.
The researchers found that there are very few trials that directly compare exercise with drug therapy for any chronic condition. There were only enough trials to be able compare the following conditions:
- stroke rehabilitation
- coronary heart disease (specifically prevent heart disease after a heart attack)
- heart failure
- diabetes prevention in people with risk factors for that condition (“pre-diabetes”)
Exercise reduced the likelihood of death after a stroke, but drug therapy with diuretics reduced death rates for people with heart failure. They did not find any difference between exercise and drug therapy for death rates after a heart attack or in people likely to develop diabetes.
However, the main problem faced by the researchers is that most of these studies did not directly compare exercise with drug therapy. They want any future studies to directly compare the effects of exercise against the effects of drug therapy so that people will be able to make a more informed choice about the benefits and risks.
In the meantime, it is strongly advised that you continue taking any medication as prescribed.
Where did the story come from?
The study was carried out by researchers from the London School of Economics and Political Science, Harvard Medical School and Harvard Pilgrim Health Care Institute and the Stanford University School of Medicine and reports that it did not receive any funding.
There have been a number of inaccurate reports of this study in the media. The Daily Mirror reports that “exercise is better than drugs for heart failure” despite the fact that the research showed that diuretics were more effective than exercise for heart failure. Meanwhile, The Times rather enthusiastically reports that “doctors urged to prescribe exercise rather than drugs”. However, the researchers recommend exercise as well as medication until further direct comparisons between the two have been made.
What kind of research was this?
This research combined results of existing studies looking at the effects of exercise or physical activity compared to drug therapy on death rates in a statistical study. This is an example of a network meta-analysis.
It aimed to see whether exercise was as good as or better than drug therapy in reducing the risk of death. In order to prove cause and effect, the original studies were only included if they were randomised controlled trials.
What did the research involve?
The researchers first conducted a search of all previous meta-analysis of trials that directly assessed the impact of exercise on death rates for any type of illness up to December 2012. For each illness, they then found a meta-analysis that looked at the effectiveness of recommended drug treatments on death rates. Finally, they searched for any new randomised controlled trials that compared exercise with the drug treatments up to May 2013 that may not have been included in meta-analyses.
They chose to only include the most recent meta-analysis for each condition or intervention. In total, they included 16 meta-analyses that covered 305 trials and 339,274 people. The conditions and their relevant drug treatments were:
- coronary heart disease – statins, beta-blockers, angiotensin converting enzyme (ACE) inhibitors and antiplatelets
- stroke – anticoagulants and antiplatelets
- heart failure – ACE inhibitors, beta-blockers, diuretics and angiotensin receptor blockers
- “pre-diabetes” – alpha-glucosidase inhibitors, thiazolidinediones (also known as glitazones), biguanides (such as metformin), ACE inhibitors and glinides
None of the studies directly compared exercise with drug therapy for people who have suffered from a heart attack (secondary prevention of coronary heart disease). It is not clear how many of the few potential studies for stroke or heart failure compared exercise with drug therapy directly.
For pre-diabetes, one trial directly compared exercise with alpha glucosidase inhibitors, and two trials compared exercise, biguanides and control.
The data was analysed thoroughly using established statistical techniques for direct and indirect meta-analysis.
What were the basic results?
When the researchers looked at trials of drugs to prevent heart disease after a heart attack, they found that mortality was reduced, compared with control, in groups of patients given:
- statins (odds ratio (OR) 0.82, 95% credible intervals (CI) 0.75 to 0.90) – a credible interval is based on estimates and is not the same as a confidence interval which is based on actual data
- beta-blockers (OR 0.85, 95% CI 0.78 to 0.92)
- ACE inhibitors (OR 0.83, 95% CI 0.72 to 0.96)
- antiplatelets (OR 0.83, 95% CI 0.74 to 0.93)
Exercise interventions had a similar average improvement but a wider credible interval range, which included the possibility of there not being a statistically significant effect (OR 0.89, 95% CI 0.76 to 1.04).
When compared head-to-head using a statistical method called network meta-analyses, there were no statistically detectable differences among any of the exercise and drug interventions in terms of their effects on mortality outcomes.
Unlike any of the drug interventions, exercise was significantly more effective than control in reducing the odds of mortality among patients with stroke (OR 0.09, 95% CI 0.01 to 0.72).
When compared head-to-head, exercise interventions appeared to be more effective than anticoagulants (OR 0.09, 95% CI 0.01 to 0.70) and antiplatelets (OR 0.10, 95% CI 0.01 to 0.62). However, these results should be interpreted with caution as there were differences between patients in the exercise trials and patients in the drug trials, and very few deaths occurred among patients in the exercise trials.
In groups of patients with heart failure, fewer deaths occurred with diuretics (OR 0.19, 95% CI 0.03 to 0.66) and beta-blockers (OR 0.71, 95% CI 0.61 to 0.80) compared with control.
Diuretics were more effective than exercise (OR 0.24, 95% CI 0.04 to 0.85), ACE inhibitors (OR 0.21, 95% CI 0.03, 0.76), beta-blockers (OR 0.27, 95% CI 0.04 to 0.93), and angiotensin receptor blockers (OR 0.21, 95% CI 0.03 to 0.73). Angiotensin receptor blockers were associated with more deaths compared with beta-blockers (OR 1.30, 95% CI 1.02 to 1.61).
Neither exercise nor drug interventions were clearly effective in reducing the odds of mortality in pre-diabetes when compared to control. There were also no major detectable differences between any of the exercise or drug interventions.
How did the researchers interpret the results?
The researchers concluded that this study “highlights the near absence of evidence on the comparative effectiveness of exercise and drug interventions on mortality outcomes” and that their analysis “suggests that exercise potentially had similar effectiveness to drug interventions with two exceptions. In the case of stroke rehabilitation, exercise seemed to be more effective than drug interventions. In heart failure, diuretics outperformed all comparators, including exercise”.
They suggest that future drug trials should include a treatment arm that is an exercise intervention to make sure that people are able to weigh up the benefits of taking medication or exercise.
The researchers found that there are very few trials that directly compare exercise with drug therapy for any condition. They were only able to find enough trials to be able to analyse results for four major conditions.
They found that exercise reduced death rates for people after a stroke (although this analysis has limitations and should be interpreted cautiously), and that drug therapy with diuretics improved death rates for people with heart failure. They did not find any difference between exercise and drug therapy for death rates after a heart attack or in people with pre-diabetes.
There were several limitations to this study, which the researchers have pointed out themselves, including:
- Most of these studies did not directly compare exercise with drug therapy – most people in the identified studies were on the usual drug therapies with additional exercise and/or lifestyle modifications.
- Exercise trials often included some drug interventions (although the details of drug treatments were limited), suggesting that the observed effect of exercise possibly represented the added benefit of exercise over and above the benefit conferred by drug interventions.
- The exercise interventions varied across the four conditions, so can’t be generalised. These interventions were not the same as the recommended target of 150 minutes per week of moderate-intensity activity. The exercise in the studies included cardiac rehabilitation after a heart attack, cardiorespiratory and muscle strengthening exercises after a stroke and aerobic and resistance training for heart failure. The exercise/strengthening happened as inpatients, outpatients and in the community or home. Increased physical activity was part of the lifestyle modifications recommended for pre-diabetes.
- Disease severity differed across the stroke trials. For example, people taking part in the exercise trials after stroke were able to walk and take exercise up to five months after their stroke. On the other hand, participants in the drug trials were all within two weeks of having a stroke. This difference between the two sets of interventions increases the likelihood that the increased effectiveness of exercise in the network meta-analysis was a result of confounding due to disease severity.
- Additionally, very few deaths occurred in the exercise trials among stroke patients, which reduces the certainty of the estimated effect. Substantial differences in disease severity were not seen among the other three conditions.
The researchers recommend further studies looking directly at the effects of exercise against the effects of drug therapy so that people will be able to make a choice about whether to try exercise instead. In the meantime, it is strongly advised that you continue taking any medication as prescribed.
Much of the reporting presented the study as showing that exercise is “better” than drugs. This is an oversimplification. For many chronic diseases, a combination of exercise and drugs is the most effective way to either treat or prevent a condition. Both drugs and exercise may have a role: the drugs may help you get well enough to exercise and the exercise may help you get well enough for your doctor to be able you to review your medication.
Analysis by Bazian
Edited by NHS Website
Links to the headlines
BBC News, 2 October 2013
Daily Mail, 2 October 2013
The Independent, 1 October 2013
Daily Mirror, 2 October 2013
Metro, 1 October 2013
The Times, 2 October 2013
Daily Express, 2 October 2013
Links to the science
BMJ. Published online October 1 2013