Since the emergence of a new human disease linked to eating beef from cattle with "mad cow disease" in 1996, the British public has been gripped by one question: How bad will the epidemic be? In the hopes of clearing up the confusion, two teams of scientists have been refining their projections--and they are arriving at different answers.
Although humans consumed roughly 750,000 cattle infected with bovine spongiform encephalopathy (BSE) between about 1980 and 1996, no one knows how many people became infected, nor how long it takes for an infected person to fall sick. By late September 2001, 107 people living in the United Kingdom had died from variant Creutzfeldt-Jakob disease (vCJD), an invariably fatal neurodegenerative malady. How many more are at risk remains unclear; the most authoritative estimate to date, by epidemiologist Roy Anderson's group at Imperial College in London, predicts from a few hundred to up to 136,000.
A new mathematical analysis by researchers at the London School of Hygiene and Tropical Medicine provides some encouraging news. The study, published online by Science on 25 October, concludes that the epidemic might be nearing its peak. The team used an approach, called "back-calculation," which focuses on a much smaller number of assumptions, including guesses about how many people BSE infected, when they were infected, and the length of incubation. No matter how those parameters varied, the upper limit of cases is "less than 10,000," says London School statistical epidemiologist Simon Cousens, a co-author.
But these hopes are challenged by Anderson's group. Its newly completed but still unpublished analysis uses different mathematical techniques and comes up with maximum estimates that are "substantially higher," says team member Neil Ferguson. Both teams agree that their models may be overly optimistic if one key assumption turns out to be wrong: that all victims of vCJD will share the same genetic profile. While this has been true for all vCJD cases so far, a number of researchers believe that other genetic profiles may also be susceptible.
Given the absence of reliable tests for BSE infection in humans, it's unsurprising that different mathematical models would produce different results, researchers say. Modelers are relying on arbitrary assumptions, says epidemiologist Peter Bacchetti of the University of California, San Francisco. But because the London School predicts that the epidemic may soon peak, their projections may be more testable in the short term. Says veterinary epidemiologist Mark Woolhouse of the University of Edinburgh in Scotland: "We will soon know if they are right."