It isn't difficult to come up with good reasons why eating fibre might be good for you. Since 1969, when a prestigious researcher suggested in the Lancet that eating fibre reduced rates of bowel cancer, there has been an open market in loosely substantiated claims.

Fibre might move nasty toxins through your bowel more quickly. It might dilute and inhibit carcinogens. It might promote the production of anti-cancer acids within your guts. It might vibrate in tune with local ley-lines and promote spiritual harmony. But when it comes down to it, who cares what fibre might do? We want to know what it actually does. And for that we need evidence.

Over the past 35 years, a host of studies have tried to get to the bottom of it. The best have been trials randomly giving people diets containing different amounts of fibre. They have produced unmistakable outcomes: high-fibre diets do nothing to reduce the risk of getting bowel cancer.

But other studies have persistently and simultaneously shown the opposite. They have tended to be observational studies, ones that simply follow a large amount of people and track how much fibre they eat and how much cancer they wind up with. The trouble with these observational studies is that the two groups of people they are comparing aren't similar to begin with. People who eat fibre aren't the same as people who don't, and there's no way of knowing whether it's the fibre that makes the difference. In December 2005, the Journal of the American Medical Association (Jama) published the largest collection yet of this kind of data. It pooled the best of the observational studies: almost three-quarters of a million people in total. Those who ate more fibre got less bowel cancer.

But there was a catch. While there are bound to be factors we don't know about that affect bowel cancer rates, there are some we are aware of. The total amount you eat is important. So is consumption of milk, alcohol, red meat and vitamins. Once Jama's data was adjusted to take those factors into account, the link between fibre and bowel cancer vanished. Hopeful fantasy turns out not to be the same thing as well-researched fact and fibre doesn't appear to protect us against bowel cancer any more than it stiffens our moral character. Watch out for mindless food fascists who want to scare you into believing otherwise.