Volume 2: Science
3. The nature and cause of BSE
The investigation of the BSE epidemic
Identification of BSE as a TSE
Epidemiological investigations
Assessment of the initial epidemiological conclusions
(i) Does BSE have an extended common source?

Identification of BSE as a TSE

3.6 The circumstances surrounding the identification of a new disease in cattle and the evolution of the BSE epidemic between 1986 and 1988 are discussed in detail in vol. 3: The Early Years, 1986-88. As described there, in 1985 MAFF's 'passive surveillance system' had drawn to the attention of the Central Veterinary Laboratory (CVL) in 1985 an outbreak of disease in Pitsham Farm in West Sussex starting in December 1984. Several animals on the farm showed unusual clinical signs and then died. Although a number of samples of body tissues were submitted to the CVL for pathological analysis, they did not include brain material. Eventually, a live affected animal from this farm was submitted to the CVL in September 1985 and slaughtered so that the brain could be examined histopathologically. At that time, the histopathology observed in the brain was thought to be due to a toxicity of some description. (It was not until July 1987, following a review of the CVL Pathology Department archives, that this case was confirmed as BSE.) Volume 3 also describes how between October 1983 and May 1985 the passive surveillance system missed five cases on a farm in Malmesbury, Wiltshire which with hindsight are believed to have been classic cases of BSE in adult dairy cows. ¹ The limitations of the passive surveillance system are noted in volume 3 and we discuss them further in Chapter 7 below.

3.7 By December 1986, however, it had become apparent to the pathologists at the CVL that they had discovered a new disease in cattle. Several more brains from affected animals had been referred to them from farms in Kent and Bristol, and histopathological examination showed spongiform change in the brains which closely resembled scrapie in sheep. This suggested the possibility that, like scrapie, the disease could be a TSE.

3.8 In the following weeks, government scientists sought evidence that this new disease was indeed a TSE, and found it in the demonstration of scrapie-associated fibrils (SAFs) from electron microscopic studies of brain homogenates. ² Around the same time, in February 1987, scientists at the CVL inoculated hamsters with BSE-infected bovine brain in an attempt to test for transmissibility. ³ Further transmissions to hamsters were attempted at the CVL in January 1988. In the event, hamsters proved not to be susceptible to BSE, although they were readily susceptible to sheep scrapie. ⁴ Proof of transmissibility had to await inoculations into mice, and these were initiated in November 1987 in collaboration with the Neuropathogenesis Unit (NPU) in Edinburgh. ⁵ The result, which confirmed that BSE was a TSE, was evident in September 1988, 290 days post-inoculation, some weeks earlier than might have been expected from scrapie inoculation. ⁶

3.9 Meanwhile, the presence of protease-resistant prion protein (PrP^Sc) in treated homogenates of cattle brain was demonstrated in October 1987 by Dr James Hope at the NPU, providing further confirmation that the disease was a TSE. ⁷ Dr Hope used the western blotting technique to detect PrP^Sc in extracts of infected brain material (the western blotting technique is described in Chapter 1). He had originally adapted this technique in 1986 for the detection of SAFs from scrapie-infected material. ⁸

3.10 MAFF scientists submitted a brief paper describing the clinical features and neuropathology of BSE for publication in the Veterinary Record in August 1987. It was published in October 1987. ⁹ The paper included a description of SAFs from brain homogenates, and concluded that the disease was probably due to an unconventional infectious agent as in other well-recognised TSEs.

Epidemiological investigations

3.11 BSE was first brought to the attention of Mr John Wilesmith, Head of the Epidemiology Unit at the CVL, in late May 1987, when Dr William Watson, Director of the CVL, asked him to investigate the epidemiology of the new disease. ¹⁰ Data on the disease collected by Veterinary Investigation Centres were passed to Mr Wilesmith in early June. At this time, six cases on four farms had been confirmed by neuropathological examination. A detailed account of the early epidemiological investigations carried out at the CVL is given in vol. 3: The Early Years, 1986-1988.

3.12 CVL and NPU scientists met later in June 1987 to discuss research collaboration and possible directions for future research. (Further detail about the development of the research programme is given in Chapter 6.) Possible sources of infection from sheep and other species, especially animal protein in cattle feed, were considered, as well as genetic factors and semen used for artificial insemination. The prospect of scrapie having infected feed via MBM was specifically mentioned. ¹¹ This had previously been considered by Dr Alan Dickinson in relation to scrapie infection in sheep, but had not been investigated experimentally (see paragraph 2.123; Dr Dickinson became head of the NPU in 1981).

3.13 Mr Wilesmith was requested to take charge of the epidemiological investigation. He designed a questionnaire to assist in the systematic collection of data on factors which might be involved in the origin and spread of BSE. As the numbers of reported cases increased through voluntary notification, Mr Wilesmith visited farms with affected or suspected animals. By December 1987 he had compiled information on approximately 200 confirmed cases of BSE. ¹² The first paper produced as a result of the analysis of this data was published in December 1988. ¹³

3.14 The approach of this study was to determine the potential causative factors that were common to all affected animals. Already in the early months of the investigation the most likely factor was thought to be infection by a slow virus or unconventional prion protein agent, although toxic causes could not be excluded and so had to be investigated. Data were collected on the use of vaccines, hormones, organophosphorous (OP) pesticides, synthetic pyrethroid (insecticide) sprays and anthelmintics (used to treat parasitic worms). Weed killers and herbicides had not been used on 22.9 per cent of the farms and pesticides had not been used on 68.9 per cent of the farms. Vaccines, hormones, OPs, anthelmintics and other treatments were also excluded as common factors. ¹⁴

3.15 Genetic factors were also considered in 1987, given what was known about other TSEs and the possibility of transmission to offspring. It was known, for example, that autosomal dominant ¹⁵ forms of CJD existed in which the disease was transmitted to 50 per cent of offspring on average, and also that genetic susceptibility and resistance factors to sheep scrapie existed. Analysis of parental data on 501 BSE cases revealed the involvement of 239 sires. Search of the pedigree database for bulls revealed that the number of bulls connected to at least ten BSE cases was not indicative of a single gene mutation common to all affected cattle. The data also excluded autosomal recessive ¹⁶ inheritance of BSE, although genetic factors influencing susceptibility to infection were still a possibility. ¹⁷ See paragraphs 3.101-3.109 and 3.151-3.158 for an account of other genetic studies.

3.16 As scrapie in sheep and goats was the only TSE known to affect farm animals in the UK, it was the main candidate for the source of the epidemic in cattle. However, the epidemiology study revealed that sheep were not held on 20 per cent of the farms with BSE-affected cattle. Thus direct or indirect contact with sheep on affected farms was also excluded. ¹⁸

3.17 The one factor common to all affected farms was the use of commercial cattle feed, either as finished rations such as pelleted calf feed and dairy cow cake, or as protein supplements used in home-mix rations. These had been fed at some time to all of the cases for which accurate records were available.

3.18 For many years animal protein in the form of MBM had been incorporated into cattle feed to provide a rich source of protein. MBM was produced by rendering offal and other waste material produced during the preparation of carcasses of cattle, sheep, pigs and poultry for human consumption. (Vol. 13: Industry Processes and Controls contains a detailed account of the production of MBM and its use in cattle feed.) It was thus possible that MBM provided a vector for the transmission of scrapie from sheep to cattle. ¹⁹

3.19 Further anecdotal evidence that MBM was the vector for the agent responsible for the BSE epidemic was provided by the occurrence of a similar TSE in a nyala and a gemsbok from a wildlife park. Disease was diagnosed in these animals in 1986 and 1987 respectively, and it was established that both had been fed cattle feed containing MBM. ²⁰ Although the incubation times for both the nyala and the gemsbok were surprisingly short (the animals had only been exposed to MBM 3 months and between 3 and 15 months respectively before the onset of illness), pathological results from the experimental transmission of tissues from these animals to mice were similar to those found in mice inoculated with material from BSE-affected cattle. It was also noted that the early cases of BSE had occurred almost exclusively among dairy herds, an observation that correlated with the heavy feeding of cattle concentrates to dairy cows and their calves. (See vol. 12: Livestock Farming for more information on the different feeding regimes for dairy and beef cattle.)

3.20 Results of the initial epidemiological investigations were published in the Veterinary Record in December 1988. In the paper, Mr Wilesmith concluded that, on the basis of current evidence, it appeared that MBM was the vehicle of infection. Although he had discounted autosomal dominant inheritance of the disorder, Mr Wilesmith was unable to comment on the occurrence and incidence of maternal transmission of the disease (we discuss maternal transmission later in this chapter). Given the ages of the BSE cases at the time, and therefore the number of their offspring which would survive the minimum incubation period, he did not expect maternal transmission to have an effect on the annual incidence until at least 1990.

3.21 Analysis of the 1987 BSE data also revealed the spatial pattern of the emerging disease and provided clues as to where the infectivity might have originated. Mr Wilesmith noted that there was no evidence of an association between the occurrence of BSE and any single compounder of proprietary feedstuffs. A number of compounders were involved and they, in turn, received MBM from many different renderers. As cases of BSE occurred simultaneously in geographically separate regions supplied by different compounders, Mr Wilesmith suggested that each case was an index case, ie, the first case in a defined population. ²¹ He also concluded that there was no evidence of cattle-to-cattle transmission, and that the distribution of cases seemed characteristic of an extended common source epidemic. Although this assessment was widely accepted at the time, and indeed still is by some, more recent analysis has suggested a different pattern for the BSE epidemic, and has raised questions over certain assumptions made at the time regarding the origin of the disease. See paragraphs 3.34-3.47 for a more detailed discussion.

3.22 The analysis also looked at when the first cases of BSE might have arisen. All cases occurred in adult animals with an age range of 2 years 9 months to 11 years of age, with the highest incidence in 4-year-olds. As the first cases in the study were identified in 1985, this suggested that it was most likely that most of the cattle were first exposed to infection in 1981-82. Mr Wilesmith suggested a number of factors that might have combined to initiate the epidemic. These included:

1. a dramatic increase in the sheep population in Great Britain;
2. a probable (but unproven) increase in the prevalence of scrapie-infected flocks;
3. a greater inclusion of sheep heads in material for rendering;
4. a greater inclusion of casualty and condemned sheep in material for rendering;
5. the introduction of continuous rendering processes during the 1970s and 1980s; and
6. the decline in the practice of using solvent extraction of tallow in rendering since the mid-1970s.

3.23 Mr Wilesmith regarded the most important factor in the exposure of cattle to the scrapie agent as being the introduction of continuous rendering processes that might have resulted in the rendering of animal material at lower temperatures and/or for shorter periods, and therefore in the possible failure to inactivate the scrapie agent. ²² It is appropriate to interrupt our narrative at this point to note that changes to the rendering processes were not driven by changes in, or relaxation of, legislation, but rather by new technology and market forces. Meanwhile, a requirement was introduced in 1981 that processed protein should reach the 'required biological standard' of being salmonella-free. ²³ Such an end result was considered to be sufficiently rigorous to deal with other common micro-organisms. This requirement as to the end result replaced a MAFF proposal which would have specified procedures designed to ensure inactivation of common micro-organisms. We now know that neither the achievement of a salmonella-free product, nor the adoption of the original proposed procedures would have inactivated the BSE agent: see our discussion of inactivation below. A discussion of rendering processes and the Regulations controlling animal waste processing can be found in vol. 13: Industry Processes and Controls and vol. 14: Responsibilities for Human and Animal Health, respectively.

3.24 From late 1987 onwards, Mr Wilesmith thought it most likely that BSE was originally derived from existing strains of scrapie. The alternative explanation that the emergence of BSE might have been due to a new mutant strain of scrapie, transmissible to cattle, was also considered by Mr Wilesmith, but discounted. He suggested that 'the form of the epidemic and the geographically widespread occurrence of BSE would require the simultaneous emergence of this mutant scrapie strain in a large number of flocks (or cattle herds) throughout the country', ²⁴ and that this theory was inconsistent with the data. Nonetheless, he observed some geographical variation in incidence at the start of the epidemic with a 'greater incidence in Southern England . . . especially in Kent'. He suggested that this might be due to variation among compounders in the use and inclusion rate of MBM in cattle feedstuffs. This geographical variation, and the latest thinking on its implications for the possible source of the disease, is discussed later in the chapter.

3.25 The ruminant feed ban, which banned the use of ruminant-derived protein in ruminant feed in Great Britain, came into force on 18 July 1988 (see vol. 3: The Early Years 1986-88 and vol. 5: Animal Health, 1989-96 for more detail on the ruminant feed ban). Following the introduction of the ban, Mr Wilesmith produced estimates on the likely course and extent of the epidemic. ²⁵ His predictions depended on a number of assumptions, which he recognised might not necessarily be correct. He first assumed that maternal transmission did not occur (see paragraphs 3.101ff for discussion on possible routes of transmission), and then that the effect of the inclusion of infected cattle in the cattle food chain, ie, recycling of infection, was minimal. ²⁶ Given these assumptions, he concluded that the annual incidence would continue at a constant rate of the order of 2 cases per 10,000 adult cows per year until 1992, when the ruminant feed ban would result in a steady decline in the epidemic. ²⁷ It was these assumptions that Mr Wilesmith relayed to the Southwood Working Party (see vol. 4: The Southwood Working Party 1988-89).

3.26 The ruminant feed ban did result in a dramatic reduction after 1992 in the number of cases of BSE, indicating that Mr Wilesmith's conclusion that feed was the vector for the spread of the disease was correct. This was a major breakthrough in the understanding of BSE, and one that was instrumental in controlling it.

Changes in the thinking on rendering

3.27 After 1988 Mr Wilesmith's ideas on how rendering changes might have led to the emergence of BSE evolved, since several facts were found to be inconsistent with his first conclusions. A survey of all rendering plants in Great Britain which he carried out at the end of 1988 established, for example, that none of the rendering processes used was actually capable of completely inactivating the scrapie agent, and that the change to continuous rendering had not in fact generally resulted in lower time/temperature combinations. In addition, the introduction of continuous rendering took place gradually over a number of years and was not universal. ²⁸

3.28 However, the same survey also revealed that the use of solvents to extract fat from greaves had ceased in all but two plants, and that the cessation was consistent with the estimated time of onset of exposure to the infective agent. These two plants were in Scotland, where the incidence of BSE was lower. The survey also indicated a geographical correlation between the incidence of BSE and the proportion of MBM produced as a result of reprocessing greaves, which involved a double-heat treatment.

3.29 Mr Wilesmith thus discounted the introduction of continuous rendering as a major cause for the emergence of BSE and postulated that the abandonment of solvent extraction of tallow from greaves was the more critical factor. The first suggestion of this change in thinking on rendering was published in May 1990, ²⁹ followed by a full description of the survey results in 1991. ³⁰ The impact of the solvent extraction procedure was thought to be two-fold: firstly, the use of the solvents themselves; and, secondly, the application of superheated steam to remove the traces of solvent from the defatted products. Mr Wilesmith suggested that the second of these two factors was the more important. ³¹

3.30 In Northern Ireland, a slightly different pattern was seen, suggesting that it might have been the combination of the cessation of solvent extraction and a shift from batch to continuous rendering which led to conditions in which the infective agent was not inactivated. ³² However, since this combination of factors did not coincide with the pattern in the rest of the UK, it was still thought most likely that solvent extraction cessation alone was the cause.

3.31 One particularly puzzling aspect of the BSE story was the fact that the disease was confined to Great Britain and the Channel Islands until 1989. MBM was used in cattle feed in both Europe and North America and was prepared by processes similar, if not identical, to those in the UK. If BSE were derived from existing strains of scrapie, one would expect the disease to have emerged in these countries too. Although it has been suggested that differing calf-feeding practices may have been responsible (dairy calves are offered concentrates containing MBM in addition to milk at a very early age - see vol. 12: Livestock Farming), the same practices were used across Europe and thus could not explain the variation. It was not clear why BSE had not occurred elsewhere as expected. Although cases subsequently appeared in continental Europe, the numbers remained comparatively small and could be explained by the export of animals with pre-clinical disease or by the export of MBM. The evidence increasingly pointed to an origin of the disease in the UK alone.

Assessment of the initial epidemiological conclusions

3.32 Mr Wilesmith's initial conclusions can be summarised as follows:

1. The simultaneous emergence of BSE cases in geographically separate regions suggested that each case was an index case, with the distribution of cases being characteristic of an extended common source epidemic.
2. Pathological evidence suggested that the origin of BSE was the scrapie agent. Evidence did not support direct transmission of scrapie from sheep to cattle; rather, that cattle were exposed to the scrapie agent via sheep offal present in cattle feedstuffs.
3. Cattle became infected following changes in rendering methods which resulted in either a cessation or a reduction of inactivation of the scrapie agent.

3.33 These conclusions can be re-evaluated with hindsight, using what we now know about the cause of the disease and its biology. Evidence is now available to suggest that the epidemic might not have had an extended common source. It may have had a single point source with limited local spread, followed by widespread distribution through the recycling of infected cattle waste in MBM. In addition, data suggest, firstly, that BSE does not have a scrapie origin and, secondly, that changes in rendering procedures were not responsible for the emergence of BSE. These points are discussed below.

(i) Does BSE have an extended common source?

3.34 The theory that each of the early cases of BSE was an index case with a distribution typical of an extended common source has been challenged recently, since several features of the disease have been found to be inconsistent with it, or equally consistent with other theories. At the outset it is important to note a feature whose substantial impact was not appreciated until after the Report of the Southwood Working Party had been published. This was that tissues from animals incubating the disease, but not showing clinical signs, were highly infective. Failure to appreciate this fact affected the appraisal of the expected numbers of cases of BSE up to 1992, the incubation period of the disease and its geographical spread. This section discusses the observed features of the epidemic and describes the latest analyses that suggest that BSE did not have an extended common source.

Estimates of case numbers

3.35 The possibility that subclinically affected animals could be a significant source of infection was not unknown at the time of the BSE outbreak. Mr Wilesmith was aware of the potential for such infectivity (which had been demonstrated by the louping-ill incident in 1936, see paragraphs 2.128-2.130, above). But at the time he thought (he now accepts, wrongly) that scrapie-infected MBM continued to be the vector of disease, and he did not feel able to assess the likely impact of MBM from infected cows. Mr Wilesmith qualified his calculations on the likely number of BSE cases in the way noted by the Southwood Working Party:

There are however three factors which may have affected the exposure of cattle to the BSE agent since 1985 . . . Thirdly, the inclusion of infected cattle into the cattle food chain until 18 July 1988 could have increased the exposure. If this is the case, and given the incubation period distribution, then a constant number of cases, of the order 350-400 per month, can be expected; this is an incidence of 1 case per 1,000 adult cows per year. If the age structure of the national adult herd remains constant, as the usual life span of a milk cow in Great Britain is at present 5 to 6 years this rate of presentation of the disease will continue until 1993, a cumulative total of about 17,000-20,000 cases from cows currently alive and subclinically affected. ³³

3.36 In fact, at the time this view was recorded, a dramatic increase in the BSE epidemic (hidden in cattle incubating the disease) had already been fuelled by recycling of infected cattle tissue to the point at which the total number of cases could have turned out to be ten times greater than Mr Wilesmith had estimated.

3.37 Mr Wilesmith explained to the Inquiry that he had initially underestimated the effect of the recycling of infection, ³⁴ considering that cases up to mid-1989 were most likely due to infection from the sheep scrapie source and thereafter from recycling of cattle materials through MBM. In 1995 he altered his view to that which he currently holds, which is that most of the cases infected around 1989 were the result of recycling. ³⁵ This implies that unidentified, probably subclinical, cases of BSE were entering the human and animal food chains earlier, cases which would have been infected earlier still.

3.38 This possibility had been considered by Mr Keith Meldrum in 1990, when, as Chief Veterinary Officer, he was commenting on a draft of Mr Wilesmith's rendering paper, ³⁶ later published in 1991. In a minute to Dr Thomas Little of the CVL, Mr Meldrum said that he had 'great difficulty in understanding how we had such a widespread distribution of the disease circumstantially in 1985, if a strain of scrapie jumped the species barrier in 1981/1982'. ³⁷ He felt that Mr Wilesmith should re-examine his paper to see if the data were consistent with the possibility that the agent might have been present in the population for some time prior to 1981-82. Mr Wilesmith addressed this possibility in his paper. He said:

Epidemiological results obtained so far are consistent with the hypothesis that the occurrence of BSE in cattle has been a result of an increase in exposure to the scrapie agent, via ruminant derived protein in feedstuffs. This implies that cattle have always been susceptible to the scrapie agent but their exposure has, in the past, been insufficient to result in a detectable incidence of clinical disease. ³⁸

3.39 He also explained the contemporaneous and widespread emergence of the disease as being due to the reduction in the use of solvent extraction nationwide over a very short period. ³⁹

3.40 The recognition by Mr Wilesmith that his initial thinking was wrong raises the question of the number of recycled subclinical cases. Epidemiological modelling studies carried out later, in 1996, by Professor Roy Anderson and colleagues at Oxford University, estimated that as many as 54,000 BSE-infected cattle could have entered the human food chain before the emergence of the BSE epidemic in 1985, ⁴⁰ depending on the amount of under-reporting assumed. Overall, it was calculated that 480,000 infected animals entered the food chain before the introduction of the Specified Bovine Offal (SBO) ban. Many of these animals would have been slaughtered before the onset of clinical disease, and their infected offal and waste material would have been included in increasing amounts in MBM and recycled to other animals.

Incubation period

3.41 The increase in the incubation period of a TSE when it first crosses a species barrier has been well documented (see Chapter 2) and is a notable feature of these diseases. If BSE resulted from scrapie in sheep passing to cattle, one would have expected a longer incubation period initially than for cattle infected later in the epidemic from a bovine source (ie, through recycling of infected cattle). Indeed, in a 1993 paper, Mr Wilesmith suggested that the cattle infected later through recycling 'would not be protected by a species barrier' because the strain was cattle-adapted. He stated that the incubation time might therefore have been shorter, as seen in within-species transmission experiments in rodents. ⁴¹ The potential consequences of such a reduction in incubation period were judged by Mr Wilesmith to be considerable (eg, an increase in national incidence due to more young animals affected; and adaptation of the agent for cattle, resulting in a greater attack rate). ⁴² However, even though considerable attention had been paid to detecting such changes, none was identified. ⁴³

3.42 Since one of Mr Wilesmith's assumptions was that cattle were first exposed to a novel agent in 1981-82, ⁴⁴ and another that recycling only became important in 1989, ⁴⁵ he was looking for a change in incubation time in the early 1990s. But if cattle were first exposed to a novel agent earlier, perhaps in the 1970s, the shift in incubation period would also have occurred earlier. The absence of any shift in incubation period from 1987 onwards leaves open the possibility that it had in fact happened earlier. (There are other hypotheses as to the origin and spread of the disease which would be consistent with the observed epidemic - see paragraphs 3.68-3.71).

Geographic spread

3.43 In 1987 it was recognised that clinical symptoms of the BSE epidemic had first appeared in Southern England (Sussex, Kent, Hampshire, Devon and Somerset) in 1985. The disease spread to the Midlands, Wales and Northern England, reaching Scotland in November 1987 and Northern Ireland late in 1988. ⁴⁶ This pattern tends to support the view that a new or modified agent might have arisen somewhere in Southern England, and then spread from there to other parts of the country.

3.44 Indeed, recent analyses support this idea. Work undertaken by Mr Wilesmith in collaboration with Professor Roger Morris and colleagues at Massey University in New Zealand, showed an intense focus of cases occurring in the eastern part of the South West of England, in the Devon/Somerset region, and the subsequent spread of disease from this point to the surrounding areas. ⁴⁷ The analysis used data not previously the subject of research and newly developed geographical information system (GIS) techniques, which permitted a detailed spatio-temporal analysis of BSE cases down to parish level. In oral evidence to the Inquiry, Professor Morris suggested that BSE could have first appeared as a 'unifocal epidemic', which started from a single point, or limited area in the South West, and which was derived from contamination of cattle by a 'novel TSE agent capable of infecting cattle' ⁴⁸ (see below).

3.45 Professor Morris put forward three illustrative propositions for how such a localised source could have resulted in the epidemic as eventually observed:

1. An extended high-intensity exposure from about 1981 to 1984 or later, the time of exposure that had originally been suggested, producing index cases in 1986-87. Professor Morris suggested that this was the least likely of the three options, since it would have required a large amount of BSE agent distributed over a short period.
2. A short moderate-intensity exposure to infection around the period 1975-77, providing around 100 index cases in 1981, not identified at the time. The 1986-87 cases would then be the 'second wave' produced by the recycling in feed of the index cases.
3. A brief low-intensity exposure in about 1970-72, which produced just a few unidentified index cases in 1975-77, which then precipitated the 1981 second wave, which in turn precipitated the third phase of cases from 1986.

3.46 Each of these three propositions would have resulted in 'waves' of cases, interspersed by periods when cattle were not showing symptoms but were incubating the disease. Because of the long incubation period associated with BSE, many of these cases would never have shown clinical symptoms, but might have entered the animal feed chain through the rendering process. According to Professor Morris, such mini-epidemics during the 1970s could easily have gone unnoticed. ⁴⁹ He highlighted the point that since BSE commonly affected only one animal in a herd, large numbers could have been affected over a significant period of time 'without [the disease] being differentiated from other neurological disorders' or from certain 'nutritional disorders'. ⁵⁰ This theory has been supported by anecdotal accounts from UK veterinarians, who apparently noticed BSE-like symptoms from at least 1983. ⁵¹ Dr Harry Coulthard, who had been a veterinary surgeon for the previous 40 years, described to the Inquiry how each year he would see at least six cases of cattle with nervous symptoms diagnosed variously as hypomagnesaemia, lead and plant poisoning, nervous acetonaemia, tetanus, listeriosis or 'G. O. K.'s' disease ('God Only Knows'). ⁵² Laboratory tests on these disorders were mainly inconclusive and the animals which did not recover were sent for hide removal and rendering or to local hunt kennels. Possible 'incomplete ascertainment' was also acknowledged by Mr Wilesmith himself in evidence to the Inquiry. ⁵³

3.47 Independent analysis by Professor Roy Anderson, an epidemiologist from Oxford University, is consistent with Professor Morris's analysis. In oral evidence to the Inquiry, he stated that his back-calculation method could be used to illustrate a single point source from a single infected cow for BSE. He also said that the spatial pattern could be explained by the distribution of infected feed, but concluded that this could probably not be resolved scientifically. ⁵⁴ We understand that Professor Morris is currently working with the CVL to re-analyse the epidemiological data, which will include an investigation of the distribution of MBM and cattle feed in the early days of the BSE epidemic, work that may provide clues as to where the initial source of infection arose, and how it spread through the animal feed system. These data on the distribution of MBM have not been made available to the Inquiry.