Exercise, Prophets, and Survival Models

From Ashley Croft and Joanne Palmer in The Lancet:

Chi Pang Wen and colleagues claim that exercising for 15 min per day results in a 14% reduced risk of all-cause mortality (0·86, 95% CI 0·81—0·91). Further, they claim that every additional 15 min of daily exercise beyond the minimum daily amount of 15 min reduces all-cause mortality by an additional 4% (2·5—7·0).
This cannot be true, however, since the risk of mortality is an absolute. The risk can be postponed, but it cannot be reduced, as Wen and colleagues claim, and nor can it be eliminated. In other words you can run faster, but you arrive later.
We are aware of just two reported exceptions to this otherwise invariable rule and they are Elijah, who while still alive went up by a whirlwind into heaven (Kings 2:11), and Enoch, who was translated that he should not see death (Genesis 5:24). The details of Enoch’s final end are veiled in mystery but Elijah’s heavenward passage in a fiery chariot was witnessed by his servant Elisha, who picked up Elijah’s mantle as it fell to the ground and thereafter assumed his master’s role as prophet. In any event, neither episode occurred in the context of a randomised controlled trial.
We contend, therefore, that the risk of mortality for everyone—prophets included—is 1·0 (1·0—1·0).

h/t MG on Facebook.

4 Responses to Exercise, Prophets, and Survival Models

  1. Andrew Gelman March 6, 2012 at 11:28 am #


    I followed the link and the authors are Ashley Croft and Joanne Palmer. The Lancet is just the place where their article appeared. I hope that if you refer to my work, you won’t say, “The Journal of the American Statistical Association said . . .” but will instead use my name!

    P.S. Regarding the substance of the article, it is possible to reduce aggregate mortality by a year. Consider a simple example. Suppose everybody will live to the age of 70, and then you have an intervention that increases lifespan by 1 year. Then, in steady state, the aggregate death rate will go from 1/70 to 1/71, a decline of 1.4 percentage points. And the one-year death rate among 70-year-olds will decline by 100%. That would be a real decline (in this stylized setting), not merely a statistical illusion.

    • Erik Voeten March 6, 2012 at 11:37 am #

      I have added the authors’ names to the post.

  2. Robert March 7, 2012 at 9:54 am #

    I agree that it is essential to provide the authors’ names — especially in an instance such as this where it is only right that the authors receive the recognition they are due for a comment that is so stoopid.

  3. Moby Hick March 7, 2012 at 10:27 am #

    Medical journals need more jokes. This is the best one since the parachute study.