Home

About Green Audit

Green Audit Services

Publications

Papers

Epidemiology

Science & Policy Making

Political Theory

Azande Science

Links

FROM SELLAFIELD TO CHERNOBYL AND BEYOND

EXPOSURE TO MAN-MADE IONIZING RADIATION AS THE PRIMARY ENVIRONMENTAL CAUSE OF RECENT CANCER ISSUES

Contribution to the ASPIS seminar:

 Is cancer predominantly an environmental disease?

 Kos Island, September 6th 2000

Chris Busby, PhD

September 2000

GREEN AUDIT

OCCASIONAL PAPER 07/00

August 2000

Note; We apologise for the absence of figures in this paper. Hopefully, they will be forthcoming.

1.   Introduction

The question of the environmental origin of cancer has been answered by research.

The origin of cancer is broadly agreed. Few would now dispute Sir Walter Bodmer’s 1992 statement that ‘Cancer is a genetic disease expressed at the cellular level.' And in the last fifty years, epidemiology has examined the relationship between the incidence of the disease, the genetic origin of its victims and where they live or have moved to, and shown conclusively that it is the environmental history of the cancer victim that determines their risk. Since it has been also known for fifty years that genetic damage can be inherited, it was always clear that the environmental history of the victim included the environmental history of their inherited genetic material.

 So what percentage of cancer is environmental? The arguments for the almost complete environmental origin of cancer were made very adequately by Doll and Peto in a report to the National Academy of Sciences of the United States of America in 1981 (Doll and Peto, 1981) and I will not address them further here. Instead I want to focus on the real cancer increases which the world has been experiencing in the last twenty years and show that the primary environmental cause of this has been exposure to ionizing radiation from the novel man-made radioisotopes which first appeared in the history of this planet, and that of the genetic material of its inhabitants, in 1945.

2.   Chernobyl and Risk Factor Errors.

In chemistry, it is possible to obtain results which are unequivocal. I can stand in front of you and mix the clear solution of phenophthalein with a clear solution of sodium hydroxide and there is an immediate bright pink colour. I cannot hold up a child and mix low-level radiation with it to demonstrate leukaemia. I believe, however, that this experiment has now been done as well as it is possible to do it. I will therefore begin at the end of my argument and show that the accepted statutory low level radiation exposure models are unsafe. Low level radiation I will define here as below average natural background levels of 2mSv or so.

There is a presently widely accepted belief that exposure to ionising radiation is well understood. This contemporary model predicts that for low level exposure, the resulting cancer yields are too small to be detectable. This belief has its origin almost entirely in the cancer yield of the Hiroshima lifespan cohort, and is incorporated into risk factors which are published by the International Commission on Radiological Protection (ICRP) and others, and used by national agencies to advise on dose limits to the public. Such dose limits have been recently applied across the EU by the incorporation of the Euratom 96/29 Basic Safety Standards Directive into each member state’s law. However, a particular and quite specific response to the radioactive fallout from the Chernobyl explosion in 1986 has shown that these risk models are wildly inaccurate.

The Chernobyl accident contaminated large parts of Europe. However, the levels of contamination and doses were such that the Hiroshima risk model predicted no excess leukaemia in greater Europe. Despite the many reports of leukaemia increases coming from eastern Europe, (e.g.Savchenko, 1996) research by Dr Parkin’s team in Lyon dutifully reported that there was no sudden leukaemia increase in the combined child population of Europe, including eastern Europe, that could be ascribed to the fallout (Parkin et al.,1996 ).  However, Parkin et al.did not examine individual national populations and only reported the general increase over the period, which they assumed was an endemic effect.

 In 1988, it was noticed in Scotland, one of two areas of the UK where the Chernobyl fallout rained out, that there had been a sudden increase in infant leukaemia in those children born nine months or more after the exposure. (Gibson et al., 1988). This first small red flag was followed by a report of an increase in infant leukaemia in Greece (Petridou et al., 1996). In 1997, the cancer registry in Germany published data showing the same  effect there, with fractionation of their cases into different exposure risk groups. (Michaelis et al., 1997).

A report came in 1997 also from the USA where the doses were very low, but Joe Mangano still found a significant infant leukaemia effect due to the large base population at risk. (Mangano, 1997). Molly Scott Cato and I had been looking at the leukaemia and cancer yield after Chernobyl in Wales and Scotland and I reported in 1996 that there had been a statistically significant step-like increase in the age group 0-4 in both countries (Bramhall, 1996).  We had established also that there had been an increase in infant leukaemia age 0-1 in both countries and I reported this to the STOA workshop of the European Parliament in 1998 (Assimakopoulos 1998) and published them in June of this year (Busby and Scott Cato 2000). Taken together, these observations define a crucial experiment upon which the accepted model for radiation risk has shattered.

The difference between Wales and Scotland and the other countries where the effect had been found, was that the doses  in the UK had been published by the UK National Radiological Protection Board. The 1-year dose from Chernobyl fallout to the infants was 88 microSieverts (mSv), about one twentieth of the average natural background radiation level in the UK. However, the exposure route was quite different from the large Hiroshima external flash that informed the risk factors and from natural backgroud radiation also. The infants and their parents were exposed to Tellurium-132, Iodine-132 and Iodine-131, Caesium-134 and Caesium-137, Ruthenium-103, Ruthenium-106, Plutonium-239, Strontium-90 and many other novel isotopes. These substances were ingested or inhaled so that doses were internal and delivered in many instances by ‘hot particles’ i.e. micron and sub-micron sized aggregates capable of imparting very large doses to local tissue. (Hohenemser et al., 1986)

The doses, the predicted leukaemia yield and the observed leukaemia yield were thus separately  available  for a group of individuals who were as specified as it is possible to be, a group whose only possible cause of any statistically significant increase in the disease must be the exposure to the Chernobyl fallout. It was thus possible to examine the predictions of the NRPB/ICRP risk model.

The results indicate that the models are in error by a factor of upwards of 100-fold. The results for the effect in the European countries are given in Table 1.

The effect discovered in the USA was significant at the p = 0.05 level and thus we are able to calculate the probability that the increases were separate, unconnected, chance events in each country by simply multiplying the Poisson p-values. The resultant p-value for the hypothesis that the findings were due to chance is 3.75 x 10-10.

This is a real effect. There is no other possible explanation. It demonstrates an error in the accepted model for radiation induced leukaemia.  Before considering the implications of this I will look briefly at the shape of the dose response relationship shown by the findings.

Doses in the USA, Germany and Greece had been assessed (Savchenko 1995) and it was thus possible to construct a dose/response curve for the combined results. This is shown in Figure 1. The response is far from the linear/no threshold one assumed by the risk agencies but is biphasic. I will return to this below. 

3. Sellafield, Dounreay and La Hague

The studies on which the present risk factors are based are all studies of external irradiation. In the case of the largest, the Hiroshima study, this was also a very large acute dose. The possibility that the radiation risk factors based on these sources break down for low-level, internal, chronic exposure to man-made radioisotopes placed a large question mark over nuclear power and weapon policy options in the 1980s. This followed the discovery by a TV company of a highly statistically significant 10-fold excess of childhood leukaemia at Seascale, the closest village to the Sellafield reprocessing plant in the UK. This discovery was quickly followed by epidemiological confirmation and further reports of childhood leukaemia excess near the two other European reprocessing plants at Dounreay in Scotland and la Hague in France. All three plants have in common that they are sources of discharges of man-made fission-product radioisotopes to the environment and those living close to them are the most internally contaminated. Examination of the nuclear weapons sites at Aldermaston, Burghfield in west Berkshire and the research site Harwell in Oxfordshire revealed smaller but still significant childhood leukaemia clusters.  (Beral et al, 1993, Busby and Scott Cato1996, Viel and Poubel 1997).

These discoveries caused considerable alarm and led to a great deal of disputation and research. The official line became that the cause of the leukaemia excess (and in the case of Sellafield, a cancer excess also) could not be radiation because the risk factors would have to be in error by about 100- to 300-fold for this to be the case. Since it was assumed that this could not be possible, other causes were sought for the leukaemia excesses. The affair was apparently laid to rest following the untimely death of the epidemiologist Martin Gardner, and a court decision in 1993 to award for the defendants in the appropriately named case Reay and Hope vs BNFL in which two leukaemia victims attempted to show that releases from the Sellafield plant were responsible for their illnesses. Various unconvincing alternative explanations for the leukaemias were advanced.

The reason for drawing attention here to the nuclear site leukaemia clusters is twofold. First, note that the error factor in the risk model required to find that the nuclear site clusters were not caused by discharges from the plant is about the same as that we have shown exists in the case of the Chernobyl infants.    Second, I would like to point out that the principal method of Science, i.e. induction, was not used in the analysis of the nuclear site clusters. A scientific investigation of the health effects of low-level exposure to man-made fission product isotopes would look at all the cases where this occurred, and see what was common, mechanistically feasible and antecedent  to the cases, i.e. childhood leukaemia.

 It would not employ the false deductive method of (1) predicting the leukaemia yield in such cases on the basis of an entirely separate type of exposure, external, large and acute, and (2) arguing that the observed number of cases was too large for radiation to be the cause (see Mill 1879, Harre 1984, Scott Cato et al. 2000).

4.   Global weapons fallout 1959-63

If the Chernobyl infant leukaemias were caused by low level exposure to internal fission product isotopes then the error in the risk factors  also predicts the nuclear site leukaemia and cancer clusters. This assumption also has the advantage of scientific parsimony. What else should we expect if we follow this reasoning? The argument that follows is highly condensed from Busby, 1995.

            We know that cancer is a genetic disease expressed at the cellular level and we believe that it is almost wholly environmental. We also know that there is a time lag between the genetic damage event and the clinical expression of the disease. This time lag varies for the type of cancer but for all malignancies combined we can take it to be between 15 and 25 years. This is largely a function of the base cell replication rate for the organ or system involved and defines a period of evolution of damaged cells and their probabilistic acquisition of various critical mutations. In the UK, age standardised cancer rates began to increase in the late 1970s in Wales, the part of the UK where there is highest rainfall and therefore also the highest level of fallout deposition.  This cancer increase began in England, where fallout doses were about one half as high as Wales, about 5-years later. These observations would lead us to look for the appearance of an environmental carcinogen in the late 1950s. Because of the rainfall connection, the most suitable suspect is the global weapons fallout, which peaked between 1959 and 1963. The trend in the cumulative dose from this source is exactly mirrored by increases in cancer in Wales twenty years later, even to the discontinuities resulting from the partial test ban treaty of 1959. These are shown in Figure 2. Other effects of the global fallout are reviewed in Busby 1995.

            If the weapons fallout had caused sufficient genetic damage at the time to produce increases in cancer twenty years later, surely there should have been immediate effects on health also. There were. Infant mortality, which had been falling in the post war period, began to slow its fall or even to rise over the peak years of weapons testing. Strontium-90, Caesium-137, Plutonium-239 all were becoming incorporated into the food chain and were routinely being measured in food and in milk. The radiation origin of the infant mortality increases was pointed out by Sternglass (Sternglass 1969) and more recently by Whyte (Whyte 1990). But what happened to the fallout isotopes? Where did they go to? What is the current exposure risk from this source and from  the nuclear site discharges and what mechanism is involved? I hope to provide some answers to this question by reference to the work we have been doing in the last three years on cancer incidence in small areas near the Irish Sea and elsewhere.

5.   Cancer in small areas near the Irish sea

 Small area studies of cancer have become an area of interest since the discovery of the Seascale leukaemia cluster. However,since then, the discipline appears to have become frozen emasculated and unable to function due to the excessive and intensive development of esoteric mathematical procedures. It is almost as if epidemiologists in this area were searching desperately for ways in which they could explain away as artifact any effect that might seem real and politically dangerous. See, for example, Elliott 1992. However, the statistical methods and epidemiological reasoning needed to interpret small area cancer data are quite straightforward, and the appearance in the last five years of cheap, powerful computers and statistical packages has made the processing of enormous amounts of data relatively straightforward. Since 1997, Green Audit has been engaged in the analysis of cancer data from small areas of Wales, Ireland and parts of England in support of Irish litigation against BNFL Sellafield.  We have been able to examine cancer incidence in the entire three million population of Wales by sex and site and five year age group for each of the years between 1974 and 1990 in 193 small areas. The Irish National Cancer Registry ,which began in 1994 has given us their three first  years of similar data for 250 small areas of Ireland. In addition we have obtained ward level cancer mortality data for England and Wales from the UK Office of National Statistics. In addition we have focused on one small area near the sea and by use of survey and questionnaire have been able to place cases on the map over a ten year period.  The results of all this work show the effects of the radioactive pollution and suggest a mechanism for its biological action which I will now outline.

Most radioisotopic pollution, whether from weapons or Chernobyl fallout in rainfall or from direct discharges ends up in the sea. In the UK, where rainfall is highest in the west, weapons fallout has been washed into the Irish sea and Bristol Channel, where it has joined the large quantities of radioactive discharges from the civil nuclear sites which also discharge into the sea. By far the largest of these sites is the reprocessing plant at Sellafield which has discharged more than 117 million Petabequerels of beta-emitting and 1.35 million petabequerels of alpha-emitting radioisotopes to the sea since 1952.

In terms of beta- deposition density, Sellafield has contaminated the Irish Sea more than ten times more than the weapons fallout. In terms of the alpha emitters, Plutonium and Americium, the plant’s releases contribute more than a thousand times the fallout (Busby, 1995: 96). In addition, the radioactivity is not uniformly distributed. Research shows (Assinder,1983,  Baxter, 1989) that radioisotopes from Sellafield become concentrated on the finest particle grain sizes and where these particles become concentrated then so does the radiation. This means that  it is local tidal energy that decides the level of pollution.

 Tidal patterns in the Irish Sea are such that, apart from inlets and river estuaries there are two main offshore areas of weak or low tidal energy. One is on the north east of Ireland and includes Dundalk Bay and Carlingford Lough. The other is along the coast of north Wales and particularly the northern entrance to the Menai Strait near the town of Bangor where there is a large 50km2 silt bank called the Lavan Sands.  These are shown in Figure 3. As a control area we also looked at small areas in Somerset near the large offshore mud bank known as Steart Flats which is contaminated by discharges from the local Hinkley Point nuclear power station as well as forming a bank of material washed from the River Parrett which drains the low land in Somerset.

6. Small area results

6.1 Wales

We obtained Relative Risk values for each small areas shown on the map in Figure 4 for each of 15 cancer sites plus leukaemia and lymphomas for each year between 1974 and 1989, the period of peak releases from Sellafield. The total population of the study area was just over one million. The relative risks calculated were the observed incidence numbers in the small area divided by the expected number. This was calculated by applying the national England and Wales rates for 1979 to the 1981 census populations. Intercensal variation of the population was examined and so was the effect of deprivation using both Carstairs index and Welsh Index of Deprivation values. By using the Welsh Health Survey data for 1996 we were also able to use regression methods to examine the effects of smoking, eating green vegetables, exercise, overweight and other parameters over the whole of Wales by Unitary Authority Area.  Two methods were used to examine the trend in cancer by distance from the sea or some part of the coast. Both involved establishing the centroid of population of the area. The first method was to group the areas into seven distance bands, 0-800m, 900m-2km, 2.1-5km, 5.1-11km, 11.1-20km 21-40km and >41km. The effect of distance could then be examined by Local Regression  (LOESS) Plots or by a variation of Chi-square for linear trend in proportions (Schlesselman 1982). The second method used logarithmic regression of sea distance on Relative Risk weighted by expectation. This latter method allowed also the examination of other covariates by multiple regression. Besides the deprivation indices already mentioned we also looked at the effect of mean rainfall, plutonium in soil, plutonium in air and radon levels in homes. I will summarise the results:

·    The trend in cancer risk with distance from the coast was the same for most cancer sites and described a curious and unexpected pattern. It was significantly high very close to the coast, falling sharply in the first 2km and flattening out to about 20 km thereafter rising again more gently over the eastern watershed to fall towards the English border.

·    The effect was driven by high levels of cancer in small areas in north Wales closest to the radioactively contaminated mud banks.

·    The effect was greatest in children aged 0-4.

·    The effect worsened significantly over the 16 years of the study.

·    There were significant regression coefficients only for the covariates SEADIST, representing distance from the sea and SEAPU which represented the trend in plutonium in air across Wales as a result of sea-to–land transfer of particulate matter.

·    The deprivation covariates gave confusing and opposite effects and could not explain the general trends in cancer found. Deprivation variation over the study area was quite small.

·    Wales-wide examination of the health survey results showed that none of the factors normally associated with ill health or cancer correlated with the cancer risks by area. This was even true for smoking. The risk was driven by proximity to the Irish sea

By the end of the study period, cancer risk in children in the north Wales coastal areas near the contaminated mud banks was significantly and startlingly high and comparable with Seascale. In Bangor from 1984- 88 the all malignancy RR for children 0-4 was over 10; the brain tumour risk was also similarly very high. Over the whole 16-year period, the leukaemia RR for Bangor was 2.44 (15.9 expected, 37observed p =.00005). Other north Wales towns of similar population along the contaminated coast showed the same coastal effect. Prestatyn and Rhyl and Abergele, along the coast to the east of Bangor had 16 year RRs of  1.7, 1.8 and 2.0.  

            The all malignancy results illustrate the trend in cancer risk over the whole period and are shown in Table 2 and Figure 5 below. As an example of another site, Figure 6 shows the effect for colon cancer. Childhood cancer trend is shown by Table 3 and in Fig 7.

            The results shown here are samples from a very large study of all of the main cancer sites and types in which trends are analysed by distance from the coast and through time. One examplke which is interesting is adult leukaemia and its trend in space and time relative to the peaks in discharge from Sellafield in the mid 1970s. In Figure 8 is shown a regression surface plot of leukaemia risk in adults simultaneously by distance from the sea and through time. A clear increase in leukaemia is seen near the coast (but not inland) developing some five to eight years after the radioisotopes appeared on the coast.

Overall, these results are persuasive and suggest that the releases from Sellafield, somehow were able to cause increases in cancer in people living close to the shores of the Irish Sea and inlets or estuaries where the sediment had been contaminated. It suggested that it was the radiation that was causing cancer, not smoking, not agrochemicals, not disadvantagement or lack of exercise, or eating too many chips, or living in the highly polluted atmospreres of the coal-mining areas of South Wales, where cancer rates were far lower. But could it be some other cause that was affecting the sea coast groups? We decided to look at a control area where there was also a contaminated offshore mud bank, the North Somerset Coast.

6.2 Hinkley Point and the Steart Flats

Cancer data for small areas is very difficult to obtain. However, the UK Office for National Statistics recently released mortality data for electoral ward areas in England and Wales for the years 1995-1998 and we used this data to examine the relative risk of cancer mortality in 104 wards near the nuclear power station at Hinkley Point in Somerset, which has been discharging fission-product radioisotopes to the sea and air since the late 1960s. The station is on the coast and there is low tidal energy in the sea area next to the power station and at the mouth of the tidal River Parratt. This has resulted in the existence of a large 50km2 mud bank, the Steart Flats, which has absorbed most of the releases from the plant or from fallout washed into the river. The closest town to the mud bank is Burnham on Sea, population about 10,000, slightly smaller than Bangor. Age Adjusted Relative Mortality Risk in the wards was calculated in the same way as for Wales except that the base used was the average 1995-1998 England and Wales mortality from the cancer considered. In addition we adjusted for Social Class using census data and data on the correlation between Social Class and cancer, although this did not affect the results by more than a few percent in our study area. We looked at ‘All malignancy’, Breast, Prostate, Lung and Stomach Cancer and published our results in three reports (Busby et al. 2000a, 2000b, 2000c). These received wide press coverage and interest in the region, where there was already anecdotal evidence of high levels of cancer. The results showed the same sea-coast effect that we found in Wales. Relative risk was again highest in the town closest to the contaminated mud, Burnham on Sea where breast cancer was as high as twice the national average (8.7 expected, 17 observed; RR = 1.95, p = 0.02). Prostate cancer and all malignancy mortality were also highest in Burnham. The trends in effect by distance are shown in Figure 9 and maps of the all malignancy and the lung cancer risk in the wards is given in Fig 10 and Fig 11. Two other factors emerged in these studies. First, the measured gamma ray background was highest near the coastal mud banks at 60-100nGy/hr versus 35nGy/hr inland. Second, cancer risk was significantly higher on the low land near the tidal rivers than on the high land of the Mendip and Quantock hills ( for all cancers RR=1.09 p=.04, breast cancer RR = 1.64; p=0.02, prostate 1.31; p= .08, lung cancer RR=1.40, p=.001). This is clearly seen in the lung cancer map of Fig 11 where the higher risk follows the valley of the tidal river Parrett. Indeed, in a similar pattern to Wales, lung cancer in this area was not as well correlated with social class as it was with location near the river, a finding similar to that of Haviland in 1888.

6.3 Ireland

The Irish National Cancer Registry gave us incidence data for 253 small areas which we defined using aggregations of District Electoral Divisions. A map of these is shown in Fig 12.  For legal reasons I am unable to report the results of our analysis of these small areas in any depth. Preliminary results suggest that in the three years of data we used, 1994-96, the same sea-coast effect existed in all malignancies incidence, all ages, on the east, or Irish Sea coast, but was only very slight on the south coast and was not found on the west coast control area of County Clare and County Galway. This is in the rank of the levels of contamination from Sellafield.  The trends for females are indicated in Fig 13.

6.4 High Resolution

In addition to the above, we have increased the resolution of our examination of small areas near the Irish sea coastal contaminated mud banks by using a survey questionnaire in an area of 2000 population near such a bank. Again, for legal reasons I cannot go into detail but these results will be published in due course. What they show is that in the area studied,  within 2km of the sea, the effect is greatest (RR = 2.5; p=.004) in those living less than 100m from the sea.

7. Mechanisms

I will briefly outline the mechanism which we propose to account for our findings. The Hiroshima-based models of radiation risk are based on external irradiation, averaged over the body. For averaged external dose,at natural background levels, individual cells receive one hit per year, which they can easily repair. However, there now exist many miltiple decay exposure systems which can cause double hits inside the repair replication period of cells. Oine such system is Strontium-90/Yttrium-90, a component of weapons fallout and nuclear discharges, which can hit a cell, set up a repair and then hit the repair cycle. Another is an inhaled ‘hot particle’ for example a micron sized aggregate particle of plutonium oxide, which gives off alpha particles again and again and can similarly bypass the repair systems that cells have evolved. The averaging of such doses over whole bodies or organs is a mechanistic error which has proved fatal to the present risk model. We have called this idea the ‘Second Event Theory’ (Busby, 1995, Edwards and Cox 2000, Busby 2000)

            The phenomenon of sea-to-land transfer of radioactive plutonium particles was established in the mid 1980s (Eakins and Lally, 1984) and the trend of plutonium in air measured near the Irish Sea. This trend, shown in Fig 14, is remarkably similar to that of the cancer effect we have found. In addition we see the same trend in measurements on sheep faeces (Eakins et al. 1984), plutonium in childrens teeth (Priest et al., 1997), and in the tracheobronchial lymph nodes of autopsy specimens which show without doubt that the plutonium (and other isotopes) comes ashore from the Irish Sea and is reaches the lungs of people across the whole of the United Kingdom (Popplewell et al., 1985).  Plutonium in soil and dust near Reading in the centre of England has been increasing consistently since the 1980s, much to the consternation and astonishment of those who are measuring it (Croudace et al, 2000). This is hardly surprising considering the large proportion (perhaps as high as 30%) of all particulates which originate in the sea (APEG 1999). The effect has been known since the 1960s. I show a map of seaspray penetration in the US, published in 1962 (Junge, 1962) in  Fig 15.

 This, we believe is the origin of the sea-coast effect. It is the area micro-distribution of this dust, and the factors that determine its air concentration, that also determine cancer risk. This may even explain the elevated cancer levels near high voltage power lines where recent work by Henshaw’s team at Bristol has shown concentration of alpha emitting dust particles. (Henshaw et al. 1999)

We are currently researching, within the limits of our budget and ability, the behaviour of such particles and their attraction to discontinuities in Earth electric field gradients. This may explain, for example, the attraction of such particles to river valleys or geological discontinuities along valley strata.

            I finish by returning to the Chernobyl infants and the curious biphasic dose-response curve shown in Fig 1. The considerations which led to the Second Event Theory led us to look closely at cell responses to radiation. Cells repair sub-lethal damage and replicate.  Once started this cycle has to complete. During this repair replication cycle cells are up to 600 times more sensitive to radiation (see references in Busby 1995). There will always be a small fraction of cells in such a sensitive phase,due to natural damage, senescence and so forth. It follows that as the dose is gradually increased from zero, these will be the cells that at first mutate and then, at higher dose die. As the dose increases again, the less sensitive non-replicating G0 cells become mutated and as the dose is made even higher, they too will die. The resulting dose response will look like that in Figure 16. Such a response is indeed seen in most studies of low level radiation, but a straight line is usually drawn through it. It was first pointed out by Burlakova, that a meta-analysis of all the published radiation leukaemia studies gave such a curve (Burlakova 1996) although she had a different explanation, and I reported this to the European Parliament in 1998 (STOA).

8. Conclusions

Considerable evidence now points to man-made ionizing radiation as the cause of the cancer increases in the last 20 years. The distribution of cancer in the  west of the UK and Ireland is largely the present and historical distribution of internal radiation dose from man-made radioisotopes, whose effects are not described by the external radiation risk models presently applied as the basis of statutory frameworks for the protection of the public.

References

Airborne Particles Expert Group (APEG) (1999) Source Apportionment of Airborne Particulate Matter in the United Kingdom London: DETR

Assimakopoulos, P. (ed.) (1998), Survey and Evaluation of Criticisms of Basic Safety Standards for the Protection of Workers and Members of the Public against Ionizing Radiations. Proceedings of the STOA workshop held in Brussels, 5 February 1998  (Brussels: European Commission)

Assinder, D.J, (1983), ‘Behaviour of plutonium in the intertidal sediments of the eastern Irish sea.’ In Ecological aspects of radionuclide release, Special Publication No 3 of the British Ecological Society, Eds P.J.Coughtrey, J.N.B.Bell and T.M.Roberts, 189-197

Baxter, M. S. (1989), An Assessment of Artificial Radionuclide Transfer from Sellafield to South West Scotland, DoE report No PECD 7/9/343, (Glasgow, East Kilbride: SURRC.

Beral, V,  E. Roman, and M. Bobrow (eds.) (1993), Childhood Cancer and Nuclear Installations (London: British Medical Journal )

Bramhall, R. (ed.), The Health Effects of Low Level Radiation: Proceedings of a Symposium held at the House of Commons, 24 April 1996 (Aberystwyth: Green Audit).

Burlakova, E.B,  A. N. Goloshchapov, N. V. Gorbunova, G. P. Zhizhina, A. I. Kozachenko, D. B. Korman, A. A. Konradov, E. M. Molochkina, L. G. Nagler, I. B. Ozewra, L. M. Rozhdestvenskii, V. A. Shevchenko, S. I. Skalatskaya, M. A. Smotryaeva, O. M. Tarasenko, Yu. A. Treshchenkova, `Mechanisms of Biological Action of Low Dose Irradiation in E. B. Burlakova (ed.), Consequences of the Chernobyl Catastrophe for Human Health (Moscow: Centre for Russian Environmental Policy, 1996).

Busby, C. C. (1995), Wings of Death: Nuclear Pollution and Human Health  (Aberystwyth: Green Audit)

Busby C.,(2000), ‘Reponse to Commentary on the second event theory by Busby’ International Journal of Radiation Biology  76 (1) 123-125

Busby C, Dorfman P, Rowe H (2000a) Cancer Mortality and Proximity to Hinkley Point Nuclear Power Station in Somerset: Part I Breast Cancer. Occasional Paper 2000/2 Aberystwyth: Green Audit

Busby C, Dorfman P, Rowe H (2000b) Cancer Mortality and Proximity to Hinkley Point Nuclear Power Station in Somerset: Part II Prostate Cancer. Occasional Paper 2000/3 Aberystwyth: Green Audit

Busby C, Dorfman P, Rowe H (2000c) Cancer Mortality and Proximity to Hinkley Point Nuclear Power Station in Somerset: Part III All malignancies, lung and stomach cancer. Summary Occasional Paper 2000/4 Aberystwyth: Green Audit

Busby, C. C. and Cato, M. S. (1997), ‘Death rates from leukaemia are higher than expected in areas around nuclear sites in Oxfordshire and Berkshire,’ British Medical Journal, 315: 309

Busby, C. C. and Cato, M. S. (2000), ‘Increases in leukaemia in infants in Wales and Scotland following Chernobyl: evidence for errors in risk estimates’ Energy and Environment 11(2) 127-139

Croudace I, Warick P, Cundy A, Warneke T, Oh J-S, Taylor R (2000) An assessment of radioactive contamination in the environment as a result of operations at the AWE sites in Berkshire Report 2 Soiuthampton Oceanography Centre: Geosciences Advisory Unit

Doll, R. and Peto, R. (1981), The Causes of Cancer (Oxford: University Press)

Eakins, J. D., Lally, A. E., Cambray, R. S., Kilworth, D., Morrison, R. T., and Pratley, F. (1984), ‘Plutonium in sheep faeces as an indicator of deposition on vegetation’, Journal of Environmental Radioactivity, 87-105.

Eakins, J.D and Lally, A.E., (1984), 'The transfer to land of actinide bearing sediments from the Irish Sea by spray.'  Science of the Total Environment 35 23-32

Edwards, AA and Cox R (2000), ‘Commentary on the second event theory of Busby’ International Journal of Radiation Biology 76 (1) 119-122

Elliott, P, Cusick, J, English, D and Stern, R (1992), Geographical and Environmetal Epidemiology: Methods for Small Area Studies Oxford: University Press

Gibson, B. E. S., Eden, O. B., Barrett, A. et al. (1998), ‘Leukaemia in young children in Scotland’, Lancet,  630.

Harre R (1984) The Philosophies of Science ( Oxford: University Press)

Haviland A. (1888) ‘The geographical distribution of cancerous disease in the British Isles’ The Lancet Feb 25th 1888, 365-367; March 3rd :412-414; March 10th : 467-468

Henshaw, D.L, Fews, A, Keitch, P, Close JJ, Wilding, RJ (1999) ‘Increased Exposure to Pollutant Aerosols under High Voltage Power Cables’  International Journal of Radiation Biology 75/12:1505-21

Hohenemser C, Deicher M, Hofsass H, et al. (1986), ‘Agricultural impact of Chernobyl: a warning.’ Nature 26th June p 817

Junge C.E. (1963) Air Chemistry and Radioactivity New York: Academic Press

Mangano, J. (1997), ‘Childhood leukaemia in the US may have risen due to fallout from Chernobyl’,  British Medical Journal,  314: 1200.

Michaelis J, Kaletsch U, Burkart W and Grosche B, (1997) ‘Infant leukaemia after the Chernobyl Accident’ Nature 387, 246

Mill J.S (1879) A system of Logic (London: Longmans Green)

Papineau D (1996) The Philosophy of Science (Oxford: University Press)

Parkin, D. M. et al. (1996), ‘Childhood leukaemia in Europe after Chernobyl: 5 year follow up’,  British Journal of Cancer, 73: 1006-1012.

Petridou, E., D.Trichopoulos, N.Dessypris, V.Flytzani, S.Haidas, M.Kalmanti, D.Koliouskas, H.Kosmidis, F.Piperolou, and F.Tzortzatou, (1996) ‘Infant Leukaemia after in utero exposure to radiation from Chernobyl’ Nature, 382:25, 352

Popplewell, D.S, Ham G.J, Johnson, T.E and Barry, S.F (1985) ‘Plutonium in autopsy tissues in Great Britain’  Health Physics 49:304

Priest, N. D., O’Donnell, R.G., Mitchell, P. I., Strange, L., Fox, A., Henshaw, D. L., and Long, S. C. (1997), ‘Variations in the concentration of plutonium, strontium-90 and total alpha emitters in human teeth collected within the British Isles’, Science of the Total Environment, 201, 235-243.

Savchenko, V. K. (1995), The Ecology of the Chernobyl Catastrophe: Scientific Outlines of an International Programme of Collaborative Research (Paris: UNESCO).

Schlesselman J. (1982) Case Control Studies p200 Oxford: University Press

Scott Cato, M, Busby C and Bramhall R (2000) I don’t know much about Science: Political Decision Making in Scientific and Technical Areas (Aberystwyth: Green Audit)

Sternglass, E.J (1971) `Environmental Radiation and Human Health’, in Proceedings of the Sixth Berkeley Symposium on Mathematical Statistics and Probability, ed. J. Neyman (Berkeley, Calif.: University of California Press).

Viel, J.-F. and Poubel, D. (1997), ‘Case control study of leukaemia among Young People near La Hague Nuclear Reprocessing Plant: The Environmental Hypothesis Revisited’,  British Medical Journal, 14, 101-6.

Whyte R.K(1992), ‘First Day Neonatal Mortality since 1935: A re-examination of the Cross Hypothesis.’ British Medical Journal, 304: 343-6

Tables and Figures

Fig 1 Dose response relationship between exposure to the infants who were in utero at the time of the Chernobyl fallout and the subsequent risk of leukaemia.