Scientists are to test a hypothesis that young people who have died from the human form of BSE were infected by contaminated baby foods as far back as 1970.

The controversial idea supposes that some meat products were harmful to people 16 years before BSE in cows was even recognised, and 25 years before young adults began dying from its dreadful human equivalent.

Should this prove true, it will mean rethinking the likely future course of the disease, which is predominantly British, although cases have occurred in other countries.

Variant CJD here appears to be on the wane. Only nine people died in 2004, the fewest since 1995, its first recorded year, giving rise to the hope that no more than a few hundred may eventually succumb to it. Since 1995, 154 Britons have been identified with the disease, a handful of whom are still alive.

But the hypothesis advanced by Stephen Dealler, a microbiologist at Lancaster Royal infirmary, suggests that only the "first wave" is declining.

He argues that there were further infections in the mid- to late-1980s, when teenagers and others ate contaminated meat, including burgers. By then hundreds of thousands of cattle were carrying BSE and the tissues most likely to contain infection were not banned in food until 1989.

Babies are more susceptible to infection because their gut walls are more permeable, Dr Dealler said yesterday. But even in them the disease took about 25 years to take its course.

People infected later would take far longer, up to 40 or 50 years, to develop the clinical disease, indeed might never do so at all, but could still be in fectious; a nightmare for blood transfusion services, which depend on the under 40s for donations.

Dr Dealler put his ideas to the Spongiform Encephalopathy Advisory Committee, a government advisory body, which greeted them with scepticism.

But even doubters are concerned that the average age of victims at death is still in the late 20s, an average which ought to be getting higher as more years pass since the food controls introduced in the late 1980s.

Extrapolation from studies of otherwise healthy appendixes have suggested that as many as 3,800 people may be carrying the infection.

Moreover, all those who have died from the disease so far have been from one genetic group, but evidence of vCJD infection in the spleen has been found in a patient who died from another cause and had a different genetic make-up.

This raised the fear that far more people may yet go down with the disease while displaying different symptoms.

Dr Dealler claims that his hypothesis fits the evidence from animals with similar diseases, and from cannibals in Papua New Guinea acciden tally infected with a brain disease.

"It has been shown that neonatal animals are more easily infected, and with lower doses of disease, than older animals," he said. "The real epidemic of BSE in humans has not actually started. What we are just seeing is the beginning with young children."

Proving his ideas will be difficult, and food manufacturers have refused to give him data from the 1970s and 1980s.

The possible drawbacks to his hypothesis include the fact that many of the 15 people infected with vCJD recorded abroad had never been to Britain, and only one, from the US, was a baby in Britain.

Other scientists question his assumptions about the incubation periods in animals and humans.

Professor James Ironside, of the CJD surveillance unit in Edinburgh, was cautious, but admitted: "Exposure to baby food is indeed a possibility."

Professor Chris Higgins of Imperial College London, who chairs Seac, was blunter: "There is a lot of anecdote there, rather than hard and fast data.

"We really need to go away and assess that before anyone jumps to any conclusions. I think we would all accept there is some age range during which infection probably occurs. But I am not at all convinced at the moment, until we have looked at all the details, that the idea that it is first the very young, and secondly pre-the main epidemic is likely to be right at all."

The Infant and Dietetic Foods Association, representing baby food manufacturers, insists on its website that manufacturers "have never used any of the high risk materials banned as a result of the controls on BSE".