Evidence of What?

This morning the Health section of the BBC News Web site came out with one of those stories that now attracts more of my attention in the context of my medical history. Here is the basic lead information:

About 10,000 cases of breast and bowel cancer could be prevented each year in the UK if people did more brisk walking, claim experts.

The World Cancer Research Fund scientists say any moderate activity that makes the heart beat faster should achieve the same.

For example, data suggest 45 minutes a day of moderate exercise could prevent about 5,500 cases of breast cancer.

Physical exercise helps prevent obesity, which is a cancer risk factor.

The WCRF team stress in their report that it is the total time spent being active that is important. You do not need to set aside half an hour each day to exercise. Shorter bouts of activity will be just as beneficial as long as they add up to the same, the charity says.

Alongside brisk walking, other activities that would count include cycling or swimming at a leisurely pace, dancing, gardening and vacuuming combined with other housework, says the WCRF.

Since this is pretty much consistent with my current life style, reading this was quite assuring. I have reached an age at which my joints feel a bit stiffer every morning, but I do not find it a strain to persist in my morning routine swim or my habit of walking to just about any destination within the San Francisco city limits. I may be a bit more discriminating in choosing routes whose hills are not at their most extreme, but I see that as simply knowing my own limits. Nevertheless, where California Avenue is concerned, I still tend to opt for the straight-line path, even if it means a few more stops to catch my breath.

However good this news from the BBC may appear, there remains the question of just what the news actually is. The most informative word in this report is the last one, coming from an interview with Henry Scowcroft, Science Information Manager at Cancer Research UK:

There's solid evidence that certain cancers - including breast and bowel cancer - are less common in people who do regular, moderate exercise such as brisk walking.

Given the trend to prefer flamboyant rhetoric in the interest of attracting more attention, I admire people like Scowcroft who still see the value of properly chosen words. From the title of his position, I assume that he chose the adjective "solid" because a large portion of the BBC audience does not know the meaning of the adjectival phrase "statistically significant;" and that just means that he worded the statement with his audience in mind. More important, however, is that he phrased his sentence to make it clear that the result involves a correlation, from which, at least at the present time, no conclusive inferences regarding causal connection can be drawn.

It has been my experience that, when science deteriorates into pseudo-science, the specious reasoning of the latter tends to involve assertions of causality based on evidence of correlation. Friedrich Nietzsche even made fun of this genre of fallacious reasoning in Twilight of the Idols when he stands the proposition that eating well makes you healthy on its head. In other words he refutes the assertion that proper diet is the cause of good health with the counter-proposition that healthy people eat the right food because their bodies reject the bad stuff. To be fair, Nietzsche may have been going for the joke regardless of the logic. In his "Preface" to Twilight, he lets the reader know immediately that he is interested in a "revaluation of all values;" but he also hits the reader with the following sentence (English translations by Walter Kaufmann):

Nothing succeeds if prankishness has no part in it.

To be fair the operative German noun in that sentence is Übermut, whose primary translation in my New Cassell's is "high spirits," which may be about as good a way to unpack that compound noun as any. However, because I tend to think that Nietzsche had a deep appreciation of the distinction between logic and rhetoric, I sometimes feel that "hyperbole" might be a more effective translation in this particular setting. Whatever the case may be, however, his "logic" about the relationship between diet and health was probably meant as little more than a joke; but the joke still nicely illustrates the abuse of logic by fallacy, which may well have been one of the values he wished to "revalue."

The real point, however, is that Scowcroft issued his own statement from a vantage point similar to Nietzsche's, just without the intention of making a joke or indulging in hyperbole. He wanted to choose his words to minimize the possibility that the results being reported support any hypothesis about causality. They do not; and, at best, they may help us to refine such hypotheses in our ongoing research to figure out just how cancer works and what we can do about it at a therapeutic level.