Marathons and Cancer

Division of Labour's Robert Lawson cites UC-Irvine economist Art DeVany's "Top Ten Reasons not to Run Marathons" (in which, among other reasons, he claims that running marathons might actually increases the risk of cancer) along with some literature regarding exercise and cancer. Go read both.

I wrote the following email to Lawson, which I forwarded to DeVany:

De Vany cites studies that link increases in "cancer markers" to running marathons. Specifically he cites increases in S100b and TNFa. What he leaves out, however, is extremely important. TNF is a marker of inflammation of all kinds. It was discovered and named for its ability in mice (and humans) to mount an immune response against malignant tumors. The body employs it for inflammatory responses of all kinds, to fighting cancers and knocking out infectious agents. It comes as no surprise, then, marathon runners, who essentially break down elements of their bodies as they run farther, have increased levels of TNF in their systems.

As to S100, this is a native protein to several different cells in the body, nerves and malanocytes among many others. I'm sure its circulating levels in the blood are next to nothing, but would increase under circumstances where either malignant cells proliferate wildly (as in cancer) or when normal cells are acutely broken down (as in running marathons).

The bottom line is that De Vany's "evidence" amounts to nothing more than confusing confounding associations and causation. That these two chemicals can be increased in the setting of cancer, they do not cause cancer (and in the case of TNF, fights cancer), and they are increased in the setting of heavy tissue breakdown and inflammation in general (which cancerous tumors can cause, but are certainly not the sole, or even main, reason). And you echo this point in the literature you cite: Studies, if anything, show decreases in cancer rates in the setting of heavy exercise, like marathon-running. Sharing your hesitancy in jumping headfirst into meta-analyses, I still think it safe to say there is little evidence that running marathons increases cancer risks. Especially if De Vany's best evidence is a badly confounded association, my confidence in this statement grows.

I can make my own list of ten reasons not to run a marathon. And then I can think of ten more. None of these reasons, however, would commit the original scientific sin of confusing correlation with causation.

While I disagree big time with DeVany on this, go read his take on steroids and professional sports (with which I agree) and just about everything else (ditto).

Share this

If you people would just

If you people would just cure cancer already, nobody would be doing these tea-leaf readings.