It has been well-documented that women could reduce their risk of being diagnosed with breast cancer by eating vegetables and fruits, exercising, and keeping a non-obese BMI (>25). But the pervasive feeling among physicians (yours truly included) was that once cancer is diagnosed, changes in lifestyle are a little too late. But a recent study (Journal of Clinical Oncology, vol. 25, pp. 2345-2351, 2007) looked at this issue, and the results were totally unexpected.


How the study was done

A prospective study was performed of 1,490 women diagnosed and treated for early-stage breast cancer between 1991 and 2000. Enrollment was an average of 2 years postdiagnosis. An analysis of the effect of interaction between different factors on survival (called multivariate analysis) found that a combination of exercise, equivalent to a 30-minute walk six times a week, and a diet rich in vegetable and fruits (five or more daily servings) had the biggest effect on survival. There was no effect of obesity per se on survival; as long as the woman exercised and ate well, her chances of survival were significantly increased.


How big was the effect?

Very big! Women who adhered to the diet and exercise regimen had a 50% reduced risk of dying, compared to women who did not.

Mind you, these were women with end-stage disease. To keep things in perspective, when we get say a 20% increased survival with a new chemotherapy drug, this is grounds for accelerated approval by the FDA. Here is a completely natural and healthy way of prolonging survival, with the only “side effect” being…feeling much better. Not bad.


The usual caveats

The study was prospective, which is the gold standard for good clinical trials. It was also multi-institutional (UCSD, Stanford, Kaiser Permanente, Arizona Cancer Center in Tucson, and the M.D. Anderson Cancer Center in Houston, TX ), which avoids the possibility of institution-specific confounding factors. The patients were all end-stage, which avoids the “contamination” of the sample with early stage, curable cancer.

On the downside, the study was relatively small (1,490 patients), although the results were statistically significant, because the effect was so big. If the effect was smaller, a much larger sample would have been necessary. Still, the authors rightly conclude that further study is needed.

I agree. A larger population size may have canceled out confounding factors that have not been considered. For instance: The women were entered into the study 2 years post-diagnosis and after receiving chemotherapy. If the diet/exercise regimen did not make any difference in mortality, then it would not have made any difference in the outcome. But if we assumed that it did, then during the 2 years preceding enrollment, the women who did not follow the regimen got sicker and entered the trial in poorer shape, increasing their chances of dying earlier. The net effect would have been to exaggerate the magnitude of the effect, not necessarily negate it.

Having said that, I think this study is of tremendous importance because it showed the role of lifestyle not only in the prevention of cancer, but also in therapy. This is novel, encouraging, and requires a major research effort in order to incorporate it in the treatment of breast and other cancers (for instance colon cancer) that have shown a relationship to diet, exercise, and obesity.