In February 2024, Noetel and colleagues published a systematic review and network meta-analysis in The BMJ examining the effect of exercise on depression. The study encompassed 218 randomized controlled trials with 495 treatment arms and 14,170 participants. The findings were unambiguous in their clinical significance: walking or jogging produced a Hedges' g of -0.62, yoga -0.55, strength training -0.49, mixed aerobic exercise -0.43, and tai chi or qigong -0.42, all compared with active controls including usual care and placebo. These effect sizes are comparable to or exceed those reported for SSRIs and psychotherapy in the major depression treatment literature. The effects were proportional to prescribed intensity, and the authors concluded that exercise could be considered alongside psychotherapy and antidepressants as a core treatment for depression. This was not a preliminary finding. It was a network meta-analysis of 218 trials published in one of the world's highest-impact medical journals.
The cardiovascular evidence is older and, if anything, more robust. Cochrane systematic reviews of exercise-based cardiac rehabilitation, encompassing 85 randomized controlled trials and 23,430 participants, have found significant reductions in cardiovascular mortality, myocardial infarction risk, and all-cause hospitalization. The dose-response relationship is well-characterized: adults performing 150 to 300 minutes per week of moderate-intensity aerobic exercise — the current World Health Organization guideline minimum — achieve substantial mortality reduction, with those performing 300 to 600 minutes per week achieving a 26 to 31 percent reduction in all-cause mortality and a 28 to 38 percent reduction in cardiovascular mortality compared with inactive adults. No harmful cardiovascular effects have been documented at any studied level of physical activity, including volumes exceeding ten times the recommended minimum.
The metabolic evidence is equally strong. The Diabetes Prevention Program, published in the New England Journal of Medicine in 2002, demonstrated that a lifestyle intervention targeting 7 percent weight loss and 150 minutes of weekly physical activity reduced the incidence of type 2 diabetes by 58 percent in adults with prediabetes — nearly twice the 31 percent reduction achieved by metformin. The trial was stopped early because the lifestyle intervention was so clearly superior. Exercise independently improves insulin sensitivity, reduces visceral adiposity, lowers blood pressure, and improves lipid profiles through mechanisms that are distinct from and complementary to dietary intervention. The number needed to treat for the lifestyle intervention was 6.9 versus 13.9 for metformin — meaning half as many people needed to adopt the lifestyle intervention to prevent one case of diabetes.
The evidence base for exercise as medicine is not a collection of small or preliminary studies. It is a body of literature published across decades in the New England Journal of Medicine, The BMJ, The Lancet, JAMA, and the Cochrane Library, comprising thousands of randomized controlled trials and hundreds of thousands of participants. The effect sizes for depression, cardiovascular disease, type 2 diabetes, and all-cause mortality are large, replicated, and in many cases superior to the pharmacological comparators that constitute first-line treatment in standard clinical practice. The question this evidence poses is not whether exercise works. It is why the healthcare system is not organized to deliver it.