Beefing up the debate
How a public health consultant challenged the orthodoxy on mad cows and CJD.
‘For too long I was reading studies and papers about the link between BSE and new-variant CJD and thinking to myself, “My arse!”. So I decided to do something about it, to start a debate, to inject some rational thinking into what was becoming an irrational stance.’
George Venters is a no-nonsense public health consultant from Hamilton, Scotland, who took a break from his grandchildren’s Halloween party on 31 October to tell me about his ‘little obsession’: the question of whether ingesting bovine spongiform encephalopathy (BSE) causes new-variant Creutzfeldt-Jakob disease (nvCJD) in humans. In short, is there a link between eating beef from mad cows and contracting a fatal neurological illness? Venters doesn’t mince his words: ‘No there isn’t. From the evidence available, it appears that BSE is not a cause of nvCJD.’
Venters first aired his views (‘my straightforward views, but which some see as a bit controversial’) in the British Medical Journal (BMJ) on 13 October 2001 – and was immediately propelled into the medical and media limelight. He was invited on to radio talkshows, his findings made the headlines (‘suddenly my news was on the news’), and he provoked an ongoing and sometimes heated debate in the BMJ’s letters pages. ‘Which is just what I wanted’, says Venters, with relish. ‘To kick off a debate and challenge the orthodoxy.’
So what did Venters do that was so groundbreaking? As he points out, ‘The idea of a link between BSE and nvCJD has now evolved into orthodoxy in the medical profession in the UK and throughout much of Europe’. Since December 1995, when leading UK neuropathologist Sir Bernard Tomlinson announced that he had stopped eating hamburgers for fear that he might contract CJD, concern about contaminated meat has been rife.
In 1996, a report in The Lancet announced the discovery of a ‘new variant’ of CJD (hence new-variant CJD) and named BSE as a possible cause (1). In March 2001, a report by the Leicestershire Health Authority attributed five cases of nvCJD in the North Leicestershire village of Queniborough to local butchers relying on small abattoirs supplied by local farmers – once again bumping up the idea that mad cow meat makes people ill (2). And in October 2000, the New Labour government endorsed The Report of the BSE Inquiry, which argued that, ‘The importance of precautionary measures should not be played down on the grounds that the risk is unproven’ (3).
After a decade of fear, speculation and plummeting beef sales, how did Venters set about ‘challenging the orthodoxy’?
‘I used basic epidemiological criteria to explore the likelihood of causality between predisposing cause and disease’, says Venters. ‘In other words, to explore whether there is a link between being exposed to BSE and contracting nvCJD. Sometimes, in some diseases, the link between cause and disease can be self-evident – but often it can only be established or refuted through extensive observation, by experimenting, and by testing all the hypotheses. I wasn’t convinced that there was a link between BSE and nvCJD, so I decided to test it as thoroughly as I could and then tell people what I found.’
What Venters found was that, ‘at every level, for every criterion’, the evidence for a link between BSE and nvCJD ‘just wasn’t there’. ‘Epidemiologists use certain criteria to work out whether there’s a link between cause and effect in a disease’, says Venters. ‘And when those criteria are used to look at nvCJD and whether it is caused by exposure to the BSE prion, the evidence is – to put it bluntly – weak, very weak.’
So what are the criteria used by epidemiologists to find out what does and doesn’t cause a disease? ‘First there is biological plausibility’, says Venters. ‘Basically you ask: is the link between cause and disease biologically plausible? Is there an accord between the current understanding of biological processes and the likelihood of the cause producing the effect? And I found that the biological plausibility of a link between BSE and nvCJD is highly unlikely.’
On the plausibility front, Venters points to the lack of evidence that the BSE prion is infectious to humans: ‘The idea that the BSE prion can cross the species barrier and cause disease in humans remains unproven.’
‘The BSE prion is known to produce prion encephalopathies when ingested by other species’, says Venters, ‘and, by analogy, it is just possible that such infection can also occur in humans. But it remains the case that there is no direct evidence that this prion is infectious to humans – if it was infectious it would have to survive cooking, digestion and on top of that the human immune system, which is pretty tough.
‘There is evidence, however, for a firm, robust species barrier between humans and prions from ungulate species [hooved animals] – in other words, there is evidence that it is bloody difficult for prions to cross from cattle into people. The thing is, prions produced in ungulates and humans have different sequences of amino acids. So people don’t get scrapie, do they? There were experiments done to prove that the BSE prion can cross the species barrier, and scientists did manage to infect mice and primates under experimental conditions. But – and this is a big but – when they injected the BSE prion into mice that were genetically modified to carry the gene for human prion protein, it didn’t work, it didn’t transfer – again failing to prove that the BSE prion is infectious to humans.’
‘Plus’, says Venters, ‘we are not operating in experimental conditions, are we? We are operating in the real world. And in the real world, there is no evidence that the BSE prion has ever crossed the species barrier into humans’.
After tests for biological plausibility came tests for ‘strength of association’ (how often exposure to the cause leads to the disease) and ‘consistency of findings’ (whether the findings are consistent with other studies among different populations at different times). For both, Venters found the evidence ‘wanting’.
‘The strength of association is simply unknown’, says Venters, ‘as details of nvCJD victims’ exposure to the BSE prion and then the occurrence of nvCJD afterwards is not known – it is not logged in any scientific way. With the consistency of findings, there is also uncertainty. Ask yourself: why, if the general British population of all different ages is exposed to the BSE prion, is nvCJD predominantly found among young people? And why is it that cases have been found and reported in France, even though there is a much lower level of possible exposure to the BSE prion for French people? There isn’t a consistency in the findings that suggests evidence for a link between BSE and nvCJD’.
Next came tests for ‘quality of evidence’ – where Venters examined ‘how robust and pertinent was the evidence provided’. His conclusion? That it was at best ‘variable and selective’ – at worst ‘irresponsible’.
‘Because it’s impossible to prove that the BSE prion is infectious to humans’, says Venters, ‘evidence for a link has to be indirect. So what has happened is that there’s been a process of hypothesis confirmation rather than hypothesis testing. And evidence that has got in the way of the process of confirmation, evidence that has been awkward or contrary, has either been played down or just outright ignored.’
Venters accuses those scientists and health experts who have presented evidence of a link between BSE and nvCJD of using the ‘faggot fallacy’ – ‘the belief that multiple pieces of suspect or weak evidence provide strong evidence when bundled together’. ‘Take the idea that this is a new disease, a new variant of CJD, for which the evidence is weak. There was an obvious and very big improvement in the detection and reporting of all prion encephalopathies following the establishment of the UK Creutzfeldt-Jakob Disease Surveillance Unit in 1990 – but this better way of detecting the illness was not even considered as an explanation for the discovery of a new disease and the concern about a new epidemic.
‘So in the Lancet article that first described the new variant, the authors noted that the 10 index cases “would not ordinarily have been referred to our Unit” – but they were referred to the Unit because of the widespread worry about BSE causing CJD. This meant there was a qualitative change in the type of the patients referred to the CJD Unit – but the evidence didn’t take this into account, it didn’t say “Maybe that’s why there were more cases, and talk of a new disease”. A lot of the evidence presented has this kind of character, which means the quality of the evidence just doesn’t stand up.’
A further epidemiological test is ‘reversibility’ – whether removing the supposed cause of the disease stops the disease from occurring. But how can Venters prove this, without single-handedly getting rid of the BSE prion and ensuring that all British beef is prion-free and above board? Venters laughs: ‘That’s not likely – but what the reversibility test means is that the hypothesis of a link between BSE and nvCJD will be falsified and proved wrong after the BSE prion has been eliminated from the human food chain in the UK. I am confident that, once that has happened, there will still be occurrences of CJD in humans.’
‘There is a really telling point’, says Venters, ‘and it’s a strong one: if this really was a food-borne infection you would expect a parallel curve – you would expect the frequency with which cases occur in the population, to run parallel to the rate at which the population is exposed to the infection. Simple. A simple but effective measurement of whether this was a food-borne infection – the more people are possibly exposed to the prion, the more infections there are. But that is certainly not happening with BSE.
‘So the rate of BSE infection in cattle is thought to have risen exponentially between 1983 and 1988, and it reached its height at about 350,000 cases in 1988, I think. As these cattle entered the food chain they would have become a potential source of infection for people. But nvCJD cases have been appearing since 1994, and the rate of increase falls far short, far short, of what you’d expect if this was a food-borne infection.’
Venters goes even further than challenging the evidence for a link between BSE and nvCJD – he also questions just how new new-variant CJD is, and wonders whether scientists have been too quick to name this a new disease.
‘Proving that a disease is new’, says Venters, ‘means first and foremost rejecting all other possibilities – and I don’t think those who talk up new-variant CJD did that satisfactorily. There are two other diseases that are similar to nvCJD. First there is Kuru, a prion encephalopathy found among the Fore people in Papua New Guinea and spread by cannibalism. The clinical and neuropathological features of Kuru include those found in nvCJD, and both diseases involve the lymporeticular system.
‘The second is CJD itself, the original. Creutzfeldt’s original case, the person he first studied, died aged 23 in the 1920s, with the kind of clinical features and also the gross kind of neuropathology that is entirely consistent with nvCJD. So this idea of the disease’s newness is certainly open to question. My point is that Creutzfeldt himself took seven years and a lot of bloody hard work to establish the originality of the disease. We should surely apply the same sort of rigour before declaring a new variant of CJD and before talking about a link between BSE and nvCJD. Surely we should acknowledge Creutzfeldt’s primacy.’
But one question remains. If a link between BSE and nvCJD is biologically implausible, with an unknown strength of association, inconsistent findings, selective and variable evidence, and a lack of rigour, how did it become ‘orthodoxy’? How did it become accepted among the very people – scientists and health experts – you would expect to be precise, rigorous, questioning and certain? And how did it scare a population into fear of beef?
‘Because you don’t let the facts get in the way of a good story’, says Venters. ‘And because it’s been a process of hypothesis confirmation rather than hypothesis testing. It was not rigorously addressed, it was not rigorously tested, and that was the problem. Also, it is far easier to frighten a population than to make rational assertions about risk – just look at what’s happening with anthrax now.’
Venters also puts some of the blame on successive governments and ‘establishment antipathy’: ‘It is almost like they made up their minds about a link between BSE and nvCJD and so they set about confirming it. They would have saved a lot of time, effort and panicking if they had set about testing it instead.
Venters says the response to his BMJ article has been ‘largely supportive’ – but he also had people accusing him of being wrong (‘but me and my team soon dealt with them’); of being unconvincing (‘that’s just not good enough, unless they can scientifically convince me of something else’); and of being irresponsible (‘since when has testing things as best as one can been in any way irresponsible?’).
Now, Venters hopes the debate will move forward: ‘What I’ve tried to do is start a more rational appreciation of risk. To have behaviour – both our own behaviour and government behaviour – formed by what is likely to happen rather than by what is a distant, remote but extremely unlikely possibility.’
Brendan O’Neill is coordinating the spiked-conference Panic attack: Interrogating our obsession with risk, on Friday 9 May 2003, at the Royal Institution in London.
spiked-issues: Mad cow panic
(1) ‘A new variant of Creutzfeldt-Jakob disease in the UK’, Lancet, 1996, 347:921-5
(2) See The Queniborough CJD cluster, by Dr Michael Fitzpatrick
(3) Report of the BSE Inquiry, October 2000
To enquire about republishing spiked’s content, a right to reply or to request a correction, please contact the managing editor, Viv Regan.