|
|
|
Nuclear Safety > ... > Background information KiKK study
|
|
Background information on the KiKK study
|
|
Why carry out a new study?
The discussion about increased cancer rates in the vicinity of nuclear power plants has continued since the use of nuclear energy has been considered to be a problematic issue. For instance, in 1987 and 1989 British studies reported a statistically significant increase of childhood leukaemia within a radius of 10 miles around nuclear facilities in England and Wales. In 1992, an analogously performed ecological study of the German Childhood Cancer Registry (GCCR) for the period from 1980 to 1990 observed a statistically significant increased incidence rate for leukaemias among children below five years of age within the 5-km-zone around the sites.
Since these results were discussed very controversially and since at the same time a statistically significant increase in leukaemias occurred in the vicinity of the Krümmel nuclear power plant, a second ecological study was published in 1997. This study was again carried out by the GCCR, and it included data for the subsequent period of five years, i. e. from 1991 – 1995.
The discussion about a possible link between the occurrence of cancer in children and their residence in the vicinity of nuclear facilities in normal operation did not stop after the second study had been published. Moreover, the data of the GCCR were exploratively evaluated by other working groups.
The results of the GCCR studies and their re-evaluation by third parties were debated in the public and in the media. The question was if these results were chance findings or if their interpretation was governed by political motivation.
In 2001, upon invitation of the President of the Federal Office for Radiation Protection, Wolfram König, a „Round Table“ with various groups took place in Kassel. Here BfS decided to commission a study on the basis of these findings, although the study was to be methodically more ambitious. In order to gain more reliable results a so-called case-control-study was developed. This so-called KiKK-study (Kinderkrebs in der Umgebung von Kernkraftwerken – Childhood Cancer in the Vicinity of Nuclear Power Plants) started in 2003. The study design was developed by a multidisciplinary panel of experts consisting of 12 members. BfS acted on that suggestion and, following an invitation to tender, charged the German Childhood Cancer Registry (GCCR) in Mainz with the implementation of this study. After four years of research and altogether five meetings of the panel of experts, the GCCR submitted the final report in November 2007. At the same time, two scientific publications concerning the study will be published in mid-December.
The result now published in the study of the German Childhood Cancer Registry (GCCR) in Mainz gives clear evidence that the risk of leukaemia for children under the age of 5 increases with decreasing distance of their home to a nuclear power plant site.
The study is thus the third one in a series of corresponding investigations of the German Childhood Cancer Registry. However, with regard to its strength of evidence it significantly stands out from the two preceding ecological studies which were frequently referred to as the so-called “Michaelis”-studies , in the general public. For the first time not incidence rates in certain regions were compared, but exact data on the distance of the home from a reactor could be taken into account in a so-called case-control study – in fact for both children with cancer and children without cancer. This is the major advantage compared to previous studies.
The two aforementioned ecological studies carried out by the German Childhood Cancer Registry showed an increased occurrence of leukaemias in children under five years living within a radius of 5 km around reactor sites. Evaluation of data on cancer incidences in children in Bavaria showed an increased incidence rate of childhood cancer in the districts neighbouring one of five nuclear power plant sites and in the district neighbouring in the east.
How was the study carried out?
In the preceding ecological studies the incidence rates in defined areas (within a radius of 5 km to 10 km, and to 15 km around a nuclear power plant) were compared with the incidence rates in defined control areas.
Due to the case-control approach chosen for this study individual homes could be taken into account, in fact for both children with cancer and children without cancer taken from random sample via registration offices. These so-called controls were randomly selected from larger controls and were matched by age and gender.
In detail, those children were considered in the study whose cancer was reported to the German Childhood Cancer Registry and who lived in defined areas around 16 sites of German nuclear power plants. The children had to be under five years of age at the time of diagnosis. There are 1,592 cases of children with cancer. For each individual case, controls were determined as described above from the same area. Thus, 4,735 controls were chosen. The vicinity of the home to the respective nuclear power plant site could be exactly determined for all 6,327 children up to 25 metres on average. As a result, it could be tested if the children with cancer on average lived in a closer vicinity to the nuclear power plant site than their respective controls did.
The objective of the study was to answer three questions:
• Does cancer occur more frequently in children under five years of age living in the vicinity of nuclear power plant sites?
• Is there a negative trend, i. e. does the risk increase with decreasing distance to the nuclear power plant site?
• Are there confounding factors that could explain the gained result?
To answer the questions the study was divided into two parts:
• Part 1: Case-control-study without interview
• Part 2: For a sub-group of Part 1 a case-control study with interview in order to be able to take into account possible confounders which might explain the results gained in part 1.
What this case-control study cannot answer, is what causes cancer.
The KiKK-study covers the period from 1980 to 2003, divided into two study periods: the first 11 years of operation of a power reactor and the remaining years. This was necessary since one result of the so-called Michaelis-studies was that the risk was higher in the first study than it was in the second one. Accordingly, one question in the KiKK-study was if there was a difference in the distance trend possibly to be found between the early and late period of operation. The study comprises the environment of the 16 German nuclear power plant sites. The distance of the home to the nearest nuclear power plant site on the day of diagnosis (for cases), or, respectively, to the analogue reference day (for controls) was determined as measure of the distance.
Radiation exposure could not be taken into consideration since no measured results are available nor is a modelling of radiation exposure reasonably possible. The distance between home and reactor was taken as an alternative to radiation exposure.
Which findings were obtained?
Study population: 1,592 cases and 4,735 controls
Overall, the study confirmed the correlation between the vicinity of the home at the day of diagnosis and the risk to contract cancer or leukaemia before the 5th birthday. However, the study cannot produce evidence, as to which risk factors cause this relationship.
The distance of the home to the nearest nuclear power plant site was determined within an accuracy of 25 m on average.
• It was found that all types of cancer as well as leukaemia occurred significantly more frequently in the vicinity of nuclear power plants (within a radius of 5 km) than in further distant areas. The findings for all tumours can be essentially attributed to the findings for leukaemia.
• This results in a negative downward trend; this means the cancer risk increases with the increasing vicinity to the reactor site.
• It was found that the willingness of the cases or controls to participate in the study strongly depended on the distance from the home to the reactor. Thus, there is a self-selection in Part 2 (case-control study with interviews) which does not allow a transfer of findings from this part of the investigation to the first part (without interviews). However, this had no influence on the overall study findings.
In detail, the following trends were found:
A statistically significant monotonously decreasing trend of risk by distance was found:
– For all the diseases under study
– Largely caused by leukaemia
A negative correlation was found for CNS (tumours affecting the central nervous system) and no correlation at all was found for embryonic tumours.
To verify the findings, further examinations were carried out. Excluding a single reactor site from the analysis did not change the results. This means the findings cannot be attributed to a single reactor site, but are valid for all 16 nuclear power plant sites in total. The previously alleged, so-called “Krümmel” effect (the average increased incidence rate of leukaemia cases around the Krümmel site) has therefore been eliminated. To verify the results further, the innermost radius of a reactor (0 to 5 km distance) was omitted to exclude a possible systematic error due to the vicinity. Even so, the negative downward trend was confirmed.
How was the study scientifically assessed?
The study findings are reliable. No error or mistake which could explain the observed effect can be found in the development of the study design or in obtaining or analysing the data.
The study marks a significant progress concerning the questions of health effects in the vicinity of reactors discussed for the past 30 years, as an epidemiologically more ambitious approach was used. Thus far, only so-called ecological studies had been carried on this subject out in Germany, where incidence rates in defined areas (5, 10 and 15 km distance to a nuclear power plant) were compared with the rates in control areas. Through the case-control study it was possible to consider individual places of homes for cases as well as controls (persons not diseased). The controls were obtained randomly from the Residents Registration Offices. They were matched to each case by age and gender.
The result of the study is not surprising in the light of already existing findings from the German Childhood Cancer Registry, namely the two ecological incidence comparative studies mentioned above (KKW1 and KKW2). The increased risk is comparable to the risk found in these studies, and is still present in a period after those studies (1996 - 2003), though it is less obvious.
Unexpected is the significant result of the regression analysis, which shows a continuous increase of risk at an increasing vicinity of the home to the reactor.
Attributive risks
– Within a radius of 5 km of the 16 sites of the 22 nuclear power plants, a statistical average of 46 cases of malign tumours of all types was to be expected. In reality 77 cases occurred. This means that an additional 29 cases can be attributed to living within this radius. This equals 1.2 additional cases per year within a radius of 5 km to the nuclear power plants. In Germany a total of 13,373 malign tumours were diagnosed between 1980 and 2003.
– Looking at the contraction of leukaemia, 17 cases were to be expected in line with the statistical average. In reality 37 cases occurred. This means that an additional 20 cases can be attributed to living within this radius. This equals 0.8 additional cases per year within a radius of 5 km to the nuclear power plants. In Germany a total of 5893 cases of leukaemia were diagnosed between 1980 and 2003.
The difference in the findings between the two periods - first half and second half of the operation of a power reactor - reflects roughly the results obtained in the studies KKW I and KKW II. The first study showed a statistically significantly increased risk (1980-1990, RR=3.01) in children up to the age of 5 within a radius of 5 km, which was markedly lower in the second period and did not reach a statistically significance (1991-1995, RR=1.39).
The negative result regarding CNS-tumours (central nervous system) does not support the result described for the Bavarian locations, which are regarded as hypothesis-generating. At that time it was found that the observed rates for malign tumours can be attributed mainly to increased rates in CNS-tumours .
A meta-analysis of ecological studies published in 2007 also comes to the conclusion that an increased incidence rate of childhood leukaemia in the vicinity of nuclear facilities can be observed, but that no hypothesis exists to explain this. A review by Grosche in 2006 also draws the conclusion that a slight increase can be observed in children and young adults with the greatest significance found within the youngest age-group living closest to the reactor.
What are the causes of the increased cancer rates?
When considering the hypothesis of an impact of radioactive discharge, it can be established that current knowledge suggests that the additional radiation exposure of the public through the operation of a power reactor is too small to cause the effect. The exposure would have to be 1000 to 10000 times higher. There currently is no plausible explanation for the observed effect, which has been showing an overall consistent picture with little fluctuations over the 24 year period.
9th December 2007
Supplement to the background paper (as of December 14, 2007)
Role of the panel of experts
In 2001, upon invitation of the President of the Federal Office for Radiation Protection (BfS), Wolfram König, a „Round Table“ with various groups convened in Kassel. To gain reliable results, BfS decided on the occasion of this Round Table to commission a methodogically more ambitious study, a so-called case-control study.
Prior to commissioning, BfS appointed a panel of experts who discussed the study concept and design in July 2002. BfS attached great importance to the fact that the scientists represented a broad and well-balanced spectrum regarding both the represented scientific disciplines and social positions. Appointed members were experts who had submitted relevant studies on this issue in the past (Dr. Kaatsch, Dr. Körblein, Dr. Küchenhoff, Dr. Scherb, Prof. Dr. Greiser, Prof. Dr. Hoffmann), designated epidemiologists (Prof. Dr. Jöckel, Prof. Dr. Wichmann, Dr. Straif, Dr. Brüske-Hohlfeld), a paediatrician (Prof. Dr. Walther) and representatives of NGOs of the Round Table (Dr. Pflugbeil, Dipl.-Phys. Wurzbacher).
The panel of experts submitted a proposal regarding the topic and type of the study. Taking this proposal into account, BfS publicly invited tenders for the performance of the study. The GCCR in Mainz filed a tender at BfS for carrying out the study (competing with a second tender) and was awarded the contract, since the tender was more cost-effective than the competing tender. Thereupon the panel of experts was reorganised such that Dr. Kaatsch of GCCR left the panel, since he was now a contractor. Furthermore, Prof. Dr. Wichmann left the panel on his own request due to the workload associated with membership, and another paediatrician (Prof. Dr. Wirth) was appointed to the panel.
The order was commissioned at the end of 2002. In the first two meetings held by the panel of experts the study design was further specified on the basis of the preceding discussions, the call for tenders and the tender submitted by GCCR for the performance of the study. After 4 years of research and altogether five meetings of the panel of experts the GCCR submitted its final report in November 2007.
Current state of evaluation of results
The study found that there is a significantly increased risk for children unter five years of age to contract cancer and leukaemia the nearer they live to a nuclear power plant results from the study itself. According to today´s radio-biological knowledge the increase in risk determined in the study cannot be explained through the emissions from the nuclear power plants. The additional radiation exposure of the population required for this explanation would have to be about 1,000 to 10,000 times higher than it has been observed. However, from this observation it cannot be concluded that radiation as a cause can absolutely be excluded. Currently there are no plausible explanations for the determined effect, which has shown an altogether consistent picture with little fluctuation over the period of 24 years under investigation.
With regard to the radiation induced risk for children the following should be supplemented: The „Committee to Assess Health Risks from Exposure to Low Level of Ionizing Radiation“ of the US National Research Council has, in 2006, determined that the data situation on the radiation induced risk of childhood cancer is insufficient for carrying out a reliable risk assessment.
The International Commission on Radiological Protection (ICRP) and the United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) state that the radiation related cancer risk for children is by factor 2 – 3 higher than the risk for adults. Additionally, the uncertainty in the current central risk estimate for adults might lead to an actual cancer risk that could be higher or lower by a factor 2-3 again. Furthermore, the dose coefficients (dose assessment due to incorporated activity) show considerable assessment uncertainties, in particular for children.
The study thus provides indications for possible causes for the increased number of cancer incidences around nuclear power plant sites but cannot furnish proof in terms of a causality.
The next steps
In November 2007, BfS charged the epidemiologist Prof. Dr. Jöckel with carrying out a detailed expert opinion on the study, which is expected by the end of 2007. BfS will then develop its final statement.
Furthermore – as was already planned at the beginning of the study – the Commission on Radiological Protection evaluates the study and its results on behalf of the Federal Ministry for the Environment, Nature Conservation and Nuclear Safety.
Dr. W. Weiss
For more details a summary of the study (German and English) and the full study (German) are available here
|
|
Printer safe version
|
|
|
