Press Releases

2003 Releases

Reanalysis of Mad Cow Disease Confirms Risk is Low in the U.S.

For immediate release:  October 31, 2003

Boston, MA— A study by the Harvard Center for Risk Analysis (HCRA) at the Harvard School of Public Health, assessing the likelihood of mad cow disease spreading in the United States cattle population, confirms the findings of the initial HCRA analysis done in 2001;  that even if infected animals or contaminated ruminant feed material entered the American animal agriculture system from Canada, the risk of mad cow spreading extensively within the American herd would be low, and that any possible spread would by now have been reversed by controls put in place in the late 1990s.

The new study was initiated at the request of the United States Department of Agriculture following discovery in the summer of 2003 of a Canadian cow infected with bovine spongiform encephalopathy (BSE). The reanalysis also finds that any disease that might have been introduced would eventually be eliminated from the United States.

The reanalysis, done by George Gray, executive director and Joshua Cohen, senior researcher, both at HCRA, specifically examined scenarios for the likely introductions of BSE from Canada into the U.S. These hypothetical introductions included both infected animals (the study assumed 5 infected animals imported even though Canada has only identified one case to date) and contaminated animal feed.  These scenarios were evaluated using the HCRA computer model that simulates conditions in the American agricultural system.  The analysis found that if BSE infected animals had been introduced as early as 1990, up to 500-600 cattle in the U.S. might have become infected, and approximately 20-25 percent would have demonstrated signs of BSE.  Such an outbreak was never detected, though it would have been below the prevalence level that surveillance systems in place at that time would likely have found.  If the introduction took place later, the total number of animals infected in the U.S. would have been smaller.

The HCRA study found that the 1997 U.S. imposition of a ban on feeding rendered ruminant protein back to other ruminants essentially chokes off and then reverses any possible spread of the disease.  Even accounting for incomplete compliance with that feed ban, the HCRA analysis found that had infected animals or contaminated feed come in from Canada or elsewhere, the spread of BSE in the American cattle population would have been reversed by now and that human exposure to contaminated animal tissue would have been very low.HCRA has also delivered to the USDA the revised version of the November, 2001 BSE report following extensive peer review by both American and European experts.

The revised document is available on the HCRA website at and the HCRA BSE computer model is available by contacting Joshua Cohen at HCRA (
For further information, please contact:

George Gray
Executive Director
Harvard Center for Risk Analysis

Robin Herman
Office of Communications
Harvard School of Public Health