Exercise may be as effective as drugs for some mortality outcomes, study finds

Exercise may be as effective as drug therapy in providing mortality benefits for certain conditions, according to a new study.

Researchers performed a meta-epidemiological study of meta-analyses involving randomized, controlled trials that compared the effect of exercise and drug interventions on mortality with each other or with a control (defined as a placebo or usual care). A random-effects network meta-analysis was used to combine study-level death outcomes; the current study's main outcome measure was mortality. The results were published online Oct. 1 by BMJ.

An initial search of Medline and the Cochrane Database of Systematic Reviews found 4 exercise and 12 drug meta-analyses. A separate Medline search found 3 recent exercise trials that had been published between 2008 and 2012 and were not included in any of the 4 meta-analyses. When these meta-analyses and trials were incorporated, the current review included 305 trials involving 339,274 participants. Of these, 14,716 participants were randomly assigned to a physical activity intervention in 57 trials. There was evidence of the effectiveness of exercise on mortality for 4 conditions: secondary prevention of coronary heart disease, stroke rehabilitation, heart failure treatment and diabetes prevention. The authors found no statistically detectable differences between exercise and drug interventions for secondary prevention of coronary heart disease or diabetes prevention. Physical activity was more effective than drugs for stroke rehabilitation (odds ratios, 0.09 [95% credible interval, 0.01 to 0.70] for exercise vs. anticoagulants and 0.10 [95% credible interval, 0.01 to 0.62] for exercise vs. antiplatelets), but diuretics were more effective than exercise for heart failure treatment (odds ratio, 4.11 for exercise vs. diuretics [95% credible interval, 1.17 to 24.76]).

The authors noted that their findings regarding stroke rehabilitation and heart failure should be interpreted with care because data were scarce and from different study settings. They also pointed out that exercise interventions were defined differently in different trials and that some of the meta-analyses of drug interventions were outdated, among other limitations. However, they said that their study was the first to compare the effects of exercise and drugs on mortality and that it “highlights the need to perform randomised trials on the comparative effectiveness of exercise and drug interventions.” They suggested that since funding for additional exercise trials may be limited, regulators might consider requiring pharmaceutical companies to test new drugs against an exercise intervention. “In cases where drug options provide only modest benefit, patients deserve to understand the relative impact that physical activity might have on their condition,” the authors wrote.