Sir

In my capacity as chairman of the UK Food Standards Agency, I read Christl Donnelly's paper (Nature 408, 787 –788; 2000) while it was being prepared for publication, and would like to comment on issues of food safety and policy raised by Donnelly's conclusions. Although the study of the French BSE epidemic is made in the context of the situation in the United Kingdom, there are some points that may have broader implications for the European situation as a whole.

Following the 4 December meeting of the Agriculture Council, the European Commission has announced a series of new measures which will have a significant impact in reducing risk in the future. These include a complete ban on the use of mammalian meat and bone-meal for all livestock, the exclusion of all bovine intestine from the food chain and the introduction of an over-30-month (OTM) rule for all member states. Cattle over 30 months will only be allowed into the food chain if they have been tested for BSE and found negative.

This year, the United Kingdom has had ten times as many cases of BSE as France, and yet, according to Donnelly, the current risk of eating an animal close to developing clinical BSE is substantially higher in France than in Britain (52 such animals going into the food chain in France versus 1.2 in the United Kingdom). The explanation for this is that Britain, uniquely, has not allowed any animals over the age of 30 months to go into the food chain, and virtually no animals are close to clinical BSE by this age. The studies of tissue infectivity carried out so far show that most infectivity occurs in animals within the last 12 months before clinical symptoms appear. Almost all infectivity in these animals is removed in the “specified risk material” that is excluded from the food chain.

Donnelly's conclusions depend on three key assumptions: (1) that the age distribution for the onset of clinical BSE in France, reflecting a combination of the incubation period and the age at infection, is the same as in the United Kingdom; (2) that French and UK cattle populations have similar survivorship curves; (3) that 100% of clinical cases were reported in France in 2000, with under-reporting in earlier years, as was the case in the UK. The assumption of complete reporting this year may prove to be optimistic, in which case, risk from French beef will increase roughly in proportion to the degree of under-reporting this year (in other words, 50% reporting would double the risk).

Because of the uncertainties about these three assumptions, especially the third, it is better to think of the relative risks as being an order of magnitude higher in France, rather than to focus on precise numbers. It is also important to keep the risks in perspective. Under Donnelly's assumptions, the risk of consumer exposure to animals close to clinical infection in the food chain in France today is similar to the risk in Britain, say, three or four years ago. The United Kingdom's risk is going down as the epidemic declines.

For consumers of French beef in other European Union countries, the risks will be the same as for French beef in France. However, in order to look at the risks to British consumers from eating imported French beef, one must make an assumption about how effective enforcement of the OTM rule is for imports.

If the OTM rule is fully enforced, Donnelly estimates that the risk from French beef imported to Britain is essentially zero (no cattle within twelve months of developing clinical BSE). If the rule is not fully enforced, the risk scales in proportion to the frequency of breach. Given informal indications to the Food Standards Agency from local authorities that the rule is generally followed in Britain, Donnelly's example of 75% enforcement seems reasonable, and yields a relative risk from French imports of the same order of magnitude as the estimate for Britain's domestic beef.

Donnelly's study (accepting its assumptions) indicates that there is no risk case for a ban on beef imports from France to the United Kingdom — even though this would be a populist move in Britain, given that the French banned UK beef in spite of their own higher risk. It also highlights the importance of effective enforcement of the OTM rule for protecting consumers in Britain, and shows why the European Commission's new plan for an OTM-like rule (the exemption being tested animals) throughout Europe is a key step forward.

Although Donnelly focuses on French carcass meat, comparable if not greater risks to UK consumers could arise from the import of meat products (such as pâtés and salamis) which usually contain beef, as it is difficult to ascertain the age and provenance of any cattle-derived contents.

The precautionary principle, unfortunately, has come to mean all things to all lobbyists. But in the European Commission's view it should start from the available evidence, acknowledging uncertainties; it should be consistent; it should be proportionate to the risk; and it should require frequent reassessment of the risks in the light of new evidence. Against this background, the UK Food Standards Agency will continue to update its assessment of risk and consequent need for action. However, the European Commission's decision to introduce the new EU-wide measures (including an OTM-like ban) will, when implemented on 1 January 2001, enhance protection against BSE risks for consumers in all member states.