Vitamin pills commonly taken by millions of people are doing them more harm than good, an analysis of the evidence has concluded.

Three supplements — vitamins A and E and beta carotene — appear to increase death rates among those taking them. Vitamin C and selenium have no effect.

The results, published in The Journal of the American Medical Association, suggest that money spent on vitamin supplements is wasted. In response, the British Heart Foundation said people should not take supplements but should concentrate instead on eating a healthy diet.

The new study is a meta-analysis; a procedure in which many earlier studies are collected together. It was carried out by a team led by Goran Bjelakovic, of Copenhagen University Hospital, and colleagues, using methods developed by the Cochrane Collaboration, the leading international group specialising in the analysis of what works in medicine.

Supporters of vitamin supplements, which are consumed regularly by up to ten million Britons, believe these can act as antioxidants, preventing highly active oxygen radicals in the body-damaging molecules such as cholesterol causing heart disease. The theory seemed plausible, and some initial trial results appeared to lend it support. But as better trials were done different results emerged. The Copenhagen team considered 68 randomised control trials, involving 232,606 people, and published by October 2005. Of these, they rate 47 trials as being of the best quality, with the rest more prone to bias of one sort or another.

Taking only the 47 low-bias trials, involving 180,938 people, they found that supplements as a whole increased the death rate by 5 per cent. When the supplements were taken separately, beta carotene increased death rates by 7 per cent, vitamin A by 16 per cent, and vitamin E by 4 per cent. Vitamin C gave contradictory results, but when given singly or in combination with other vitamins in good-quality trials, increased the death rate by 6 per cent.

Selenium was the only supplement to emerge with any credit. It appears to cut death rates by 10 per cent when given on its own or with other supplements in high-quality trials, but the result is not statistically significant. The team concludes: “Our findings contradict the findings of observational studies claiming that antioxidants improve health.”

They say there are several possible explanations. One is that oxidative stress is not the cause of conditions such as heart disease for which it has been blamed, but may be a consequence of such conditions.

Alternatively, by eliminating free radicals we may interfere with essential defensive mechanisms. But the team adds that they examined only the use of synthetic vitamins, and their findings should not be translated to fruit and vegetables.

Ann Walker, of the Health Supplements Information Service, a body funded by the industry, said that the study combined the results of trials on two different classes of people: those with no known ill-health, and those already suffering from conditions such as heart disease. “The results of these mixed-sample metaanalyses are worthless,” she said.

Ellen Mason, a cardiac nurse at the British Heart Foundation, said: “There are good scientific reasons for believing that antioxidant supplements might protect against heart disease but a number of clinical trials have failed to provide any robust evidence in favour of this"

Times recommends

• Live chat: Thrifty ways to update your winter wardrobe
• Is anybody out there really listening?
• Nutrition Q&A: Will eating late make me fat?