AND INCIDENTALLY

WHO would have thought it?

Vitamins are good for us - except when they aren't.

Scientists have just published a review of dozens of studies into the health effects of vitamin D and discovered that a daily dose could cut the risk of breast, ovarian and colon cancer by up to 50 per cent. Except, like so many such stories, it's really an example of science generating more heat than light.

The study might sound good for everyone who likes milk, meat and especially sunlight - which our skin turns into the most potent source of vitamin D. But no: researchers would prefer we stayed indoors popping supplement pills and eating oily fish instead. After all, research has shown just how lethal a stroll in the sun can be, and only a few months ago scientists revealed a link between meat- eating and an increased risk of bowel cancer.

But don't overdo the vitamin D pills and mackerel either: too much has been linked to liver and kidney damage.

Trying to make sense of nutritional research is like listening to Vicky Pollard in a lab coat: "Yeah, but no, but yeah, but " Over the past months, we've been told that coffee is really a potent source of heartprotecting antioxidants, and chocolate is actually good for arteries. Meanwhile, decaff coffee has been linked to increased heart-attack risk.

What's going on? In many cases the studies have fallen foul of so- called confounding variables - factors like social class, which can fool scientists into seeing links which aren't really there. For example, we're now told light-to-moderate drinking is bad for you after all.

Apparently, previous studies had failed to take account of the fact that light drinkers, for reasons no one understands, have a lower risk of heart disease in the first place.

When researchers do dig up something, the link is often only with hefty levels of consumption - and even then the extra risk is piffling. That supposed link between red meat and bowel cancer emerged from people wolfing down almost twice the average amount, and even then the cancer risk was increased by just one extra case per 2,000 people.

But scientists have careers to sustain, and scientific journals want to make headlines In the end, all these risible studies confirm the only health advice worth taking seriously: everything in moderation. Claims by scientists to know better should be greeted with a Pollardesque "Yeah, whatever."

.Robert Matthews is Visiting Reader in science at Aston University and author of 25 Big Ideas: Science that is Changing our World (Oneworld, Pounds 9.99).

(c)2005. Associated Newspapers Ltd.. Provided by ProQuest Information and Learning Company. All rights Reserved.