Intended for healthcare professionals


Does physical activity prevent cancer?

BMJ 2000; 321 doi: (Published 09 December 2000) Cite this as: BMJ 2000;321:1424

Evidence suggests protection against colon cancer and probably breast cancer

  1. David Batty, research fellow in epidemiology (david.batty{at},
  2. Inger Thune, associate professor of cancer epidemiology
  1. Epidemiology Unit, Department of Epidemiology and Population Health, London School of Hygiene and Tropical Medicine, London WC1E 7HT
  2. Norwegian Cancer Society Institute of Community Medicine, Faculty of Medicine, University of Tromsø, N-9037 Tromsø, Norway

    Physical activity has marked effects on several functions of the human body that may influence cancer risk. These effects vary according to the mode, duration, frequency, and intensity of the activity and include changes in cardiovascular and pulmonary capacity, bowel motility, endogenous hormones, energy balance, immune function, antioxidant defence, and DNA repair. Although a role for energy balance in cancer causation was advanced almost three centuries ago, it is mainly in the past decade that over 200 population based studies have linked work, leisure, and household physical activities to cancer risk. The most researched cancers are those of the bowel, breast, endometrium, prostate, testes, and lung.

    Cancer of the large bowel is the most commonly investigated cancer in relation to physical activity.14 Meta-analysis1 and systematic reviews 2 3 show an inverse dose-response association between activity and colon cancer such that physically active men and women experience around half the risk of their sedentary counterparts. This observation is seen across populations and study methods, with little indication of publication bias.1 Plausible mechanisms of protection include the favourable effect of physical exertion on insulin, prostaglandin, and bile acid levels, all of which influence the growth and proliferation of colonic cells. Moreover, physical activity reduces bowel transit time and thereby the duration of contact between faecal carcinogens and colonic mucosa, which may explain its inverse association with colon cancer risk and the absence of a relation with cancer at the rectum.

    Endogenous sex hormones are strongly implicated in the development of breast and endometrial cancer. Physical activity may modulate the production, metabolism, and excretion of these hormones, so an association with these cancers is biologically possible. Physical activity may also reduce the risk of cancer through its normalising effect on body weight and composition. Evidence from population based studies suggests that occupational, leisure, and household activities are associated with about a 30% reduction in breast cancer rates,5 with a dose-response relation reported. 3 6 7 Findings are, however, less consistent than for colon cancer, and the sizes of the reported associations are generally lower. This may reflect a genuinely weaker relation or the fact that the strength of the association may vary across the life course as it does for more established markers of risk such as reproductive factors and body mass index. Those studies that have explored the link between physical exertion and the risk of endometrial cancer suggest a negative association. 1 3

    The observation that athletes show lower levels of circulating testosterone than non-athletes, and that testosterone influences the development of prostate cancer, has led to the hypothesis that physical activity may protect against this cancer.3 Though most studies suggest an inverse association between activity and prostate cancer, null and positive associations have also been shown.3 These inconsistent findings may be explained by a variation in the detection of latent disease. Data are similarly discrepant for testicular cancer. 8 9

    Although physical activity improves pulmonary ventilation and perfusion, which may reduce both the concentration of carcinogenic agents in the airways and the duration of agent-airway interaction, the association of activity with lung cancer has received relatively little attention. Findings from most, but not all, studies suggest a negative relation, 1 3 with those of strongest design — prospective cohort studies relating repeated assessments of physical activity to subsequent lung cancer 10 11 — showing an inverse, dose-response association in men.

    In the absence of randomised trials, confounding could be an alternative explanation for the apparent protective effect of activity. Individuals who are physically active may be different from their sedentary counterparts in genetic predisposition, dietary habits, and tobacco and alcohol use. Although several investigators report inverse associations between activity and cancer that are robust to statistical adjustment for these potential confounders, genetic predisposition has been little studied and dietary characteristics have been inadequately assessed. Furthermore, physical activity itself is often measured crudely, so misclassification, albeit non-differential, is likely to result.

    In addition to the apparent role of physical activity in the primary prevention of some cancers, there is growing interest in its use in the treatment and rehabilitation of patients with cancer. 12 13 Physical activity may reduce the likelihood of recurrence and enhance survival through its capacity for improving bodily movement, reducing fatigue, and enhancing immune function. Studies are, however, hampered by small sample sizes, short follow up, selection bias, and variations in the stage of cancer at study induction. Thus, although initial results are promising, clearer conclusions depend on larger and better designed studies.

    How can the clinician interpret these data on physical activity and site-specific cancers? Overall the evidence supports a potentially important protective effect of activity against colon cancer and probably breast cancer, with no association with cancer of the rectum. Notably, physical exertion does not appear consistently to increase the risk of any cancer. Further data relating activity to cancers of the endometrium, prostate, testes, and lung and to haematopoietic cancer14 are required. The optimal permutation of mode, intensity, duration, and frequency of physical activity, and its association with cancer at different stages of life, is unclear. In the meantime, in light of the decreasing population prevalence of total physical activity, doctors should advocate moderate endurance-type activity, such as walking and cycling. As well as reducing the risk of chronic diseases such as coronary heart disease and non-insulin dependent diabetes, such physical activity does seem to protect against some cancers.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    14. 14.
    View Abstract