on a recent study of increased risk of cancer in the vicinity of german nuclear power plants

kikk450.jpg

There had been quite some discussions about a recent case control study (in german) where cases were children younger than 5 years (diseased between 1980 and 2003) registered at the german childhood cancer registry (GCCR). According to a statement by Minister Gabriel of the Federal Ministry for the Environment, Nature Conversation and Nuclear Safety

“In this study by the German Childhood Cancer Registry (Deutsches Kinderkrebsregister) in Mainz, increased cancer incidences were ascertained statistically for children living in the vicinity of the 16 sites of German nuclear power plants.

Experts (see below) consider the study to be the world-wide methodically most elaborate and comprehensive study regarding the correlation between a higher risk of childhood cancer and proximity to a nuclear power plant. So what were the discussions about?


The discussions arose due to the fact that by nowadays scienitifc knowledge the official (!) radiation emitted by german nuclear power plants can’t explain this increased risk.

However the involved scientific knowledge has mainly been gained by investigations with adult cases from Hiroshima and Nagasaki which started in the fifties. There are new investigations which could indicate e.g. that infants may be much more sensitive to low long-term radiation like as it was e.g. indicated by an article about Childhood leukaemia following the Chernobyl power station accident (listed in the study).

In other words: there is a statistically evident higher risk of childhood cancer in the vicinity of german nuclear power plants but mainly due to a lack of scientific investigations there exists no scientific explanation for this risk. In particular it is not possible to prove that radiation is the reason for this increased risk.

But of course likewise it is also not possible to prove that radiation is NOT the reason for this increased risk

In the formulation of the ministry this reads as follows:

According to current scientific knowledge, the observed increase in cancer cannot be explained by radiation exposure from a nuclear power plant. The radiation exposure of the population resulting from the operation of nuclear power plants in Germany would have to be at least 1000 times higher in order to explain the observed increase in the risk of cancer. The statistical study and known causal connections between the risk of cancer and radiation are thus not compatible.

which is right in itself, but which can be misinterpreted by people lacking the background information which is sofar mostly in german. Hence for example a sentence like “a causal connection between the risk of cancer in infants and radiation can’t be excluded by current scientific knowledge” (which is also true!! (see the Chernobyl study)) would be helpful for a correct interpretation.

The press release of the study can likewise be misinterpreted. The study itself can hardly be understood by non experts, however for a probability theorist or statistician the risks should be evident.

Another delicate issue besides the “unlucky formulation” of the above is that the press releases and summaries contain only information about the risk in the 0-5km radius.

However the study conducted also investigations in a zone of a 50 km radius around a nuclear facility, which leads to a much higher incidence of additional childhood cancer- or in simple words: the higher risk does not stop at a radius of 5 km (see e.g. table 3.17a in the study were the estimated odds ratio for radii bigger than 5km is given)!

This unscientific concealment of results to the public was critizised by a board of external experts, which was in charge for evaluating the study.

Below I translated the report of this board (no guarantee for the translation!!!) since it is sofar in german only: (see also an interview with expert Eberhard Greiser in the newspaper TAZ (also in german))

translation of the Statement of the board of external experts of the Federal Office for Radiation Protection (BfS) to the KiKK study (Stellungnahme des externen Expertengremiums des BfS zur KiKK-Studie)

Nach einer internen Diskussion des Abschlußberichtes zur KiKK-Studie am 9. und 10.12.2007 kommt das externe Expertengremium zu folgenden Bewertungen

After an internal discussion of the final report on the KiKK study on Dec. 9 and 10. 2007 the board of external experts concludes with the following appraisals

A. Analysen im Rahmen des Studienauftrages

A. Analyses within the framework of the study mandate

1) Hauptergebnis der Studie ist eine kontinuierliche Zunahme des Erkrankungsrisikos für Krebserkrankungen und Leukämie bei unter 5jährigen Kindern mit zunehmender Wohnnähe zum nächstgelegenen Atomkraftwerksstandort.

1) Main result of the study is a continous increase of the risk to fall ill with cancer and leukaemia among under 5 year olds depending on the increasing proximity of the place of domicile to the next nuclear power plant.

2) Das abgestimmte Design entspricht dem Stand der epidemiologischen Wissenschaft.

2) The concerted design is up-to-date with the current standing of epidemiological science.

3) Die Methodik der Modellierung der kontinuierlichen Entfernungsvariablen ist adäquat. Beide in der Studie verwendeten Modelle zeigen eine gute Anpassung an die erhobenen Daten. Die Modelle erlauben eine Schätzung der Erkrankungsrisiken in Abhängigkeit von der Entfernung des Wohnortes vom nächstgelegenen Atomkraftwerksstandort.

3) The method of the modelling of the continous distance variable is adequate. Both in the study used models show a good adaption to the surveyed data. The models allow for an estimation of the risk to fall ill depending on the distance of the place of domicile to the next nuclear facility.

4) Die Anregungen des externen Expertengremiums hinsichtlich Analysen zur Daten- und Ergebnisqualität wurden mehrheitlich umgesetzt. Aus diesen Analysen ergeben sich keine Hinweise auf bedeutsame Verzerrungen der Ergebnisse.

4) Proposals of the board of external experts regarding the analyses data and result quality were majoritarian implemented. By these analyses no indications of a significant bias of the results could be inferred.

5) Die Autoren konstatieren zu Recht, dass das Erkrankungsrisiko an kindlichen Krebserkrankungen und Leukämie mit zunehmender Wohnnähe zu einem Atomkraftwerk signifikant und stetig zunimmt. Die Studie ist weltweit die methodisch aufwendigste und umfassendste Untersuchung dieses Zusammenhanges. Der Zusammenhang zwischen Wohnnähe und Erkrankungsrisiko ist damit für Deutschland hinreichend belegt.

The authors rightly state that the risk to fall ill with infantile cancer and leukaemia with increasing proximity to a nuclear power plant raises significantly and continously. The study is world-wide the methodically most elaborate and comprehensive study regarding this correlation. The correlation between proximity of domicile and risk to fall ill is therewith sufficiently substantiated for Germany.

6) Die Qualität der Datenerhebung, -aufarbeitung und –analyse kann derzeit nicht abschließend beurteilt werden, weil die dafür erforderlichen Informationen dem externen Expertengremium nicht zur Verfügung gestellt wurden. Diese Fragen hätten durch das vom externen Expertengremium vorgeschlagene Audit geklärt werden können. Das Expertengremium empfiehlt, die Frage des Audits weiter zu verfolgen.

6) The quality of data inquiry, elaboration and analysis can’t be conclusively assessed, since the for this purpose necessary information was not made available to the board of external experts. These questions could have been clarified in an audit, which was proposed by the board of external experts. The board of external experts recommends to further pursue the matter of the audit.

B. Weitere Analysen durch die Auftragnehmer

B. Further analysis

7) Die Berechnungen zum Attributivrisiko waren im Auswerteplan nicht vorgesehen. Für die Kommunikation der Ergebnisse an die Politik und Öffentlichkeit ist eine Angabe des der Wohnortnähe zum Reaktor zuzuschreibenden Risikos und zum bevölkerungsbezogenen Risiko unverzichtbar.

7) Calculations concerning the attributable risk were not scheduled for the examination plan. For the communication of results to the range of politics and to the public a specification of the risk, which is attributed to the vicinity between domicile and nuclear facility, is indispensable.

8 ) Im vorliegenden Fall wurden die Berechnungen nicht korrekt durchgeführt:

In the cases at hand the calculations were not correct:

– Es wurde lediglich die 0-5 km-Region um die Atomstandorte berücksichtigt, während die übrigen Anteile des Untersuchungsgebietes außer acht blieben, obwohl auch dort signifikant erhöhte Risiken berechnet wurden.

-Solely the 0-5km region around nuclear facilities were accounted for, while the remaining parts of the investigated region were disregarded, although significantly higher risks were also calculated for these regions.

– Die Bezugspopulation für die Berechung des Anteiles aller Krebs- und Leukämiefälle bei Kindern unter 5 Jahren ist nicht korrekt bestimmt. In der Konsequenz wurde der tatsächlich auf die Wohnnähe zu Atomstandorten zurückzuführende Anteil der Krebsfälle unterschätzt.

-The reference population for the calculation of the contingent of all cases of cancer and leukaemia of kids under 5 years was not correctly determined. As a consequence the realistic portion of all cases of cancer which can be attributed to the vicinity between domicile and nuclear facility was underestimated.

– Statt der von den Autoren allein für die 0-5 km Region angegebenen zusätzlichen 29 Krebsfälle bei Kindern unter 5 Jahren muss von mindestens 121-275 zusätzlichen Neuerkrankungen im Umkreis von 50 km um alle westdeutschen Atomstandorte im Zeitraum zwischen 1980-2003 ausgegangen werden. Dies entspricht 8-18 % aller im 50 km Umkreis um Atomanlagen aufgetretenen Krebserkrankungen bei unter 5jährigen Kindern.

-Instead of the by the authors indicated 29 additional cases of cancer of children under 5 years for the 0-5 km region one has to assume that at least 121-275 additional new cases within a radius of 50 km around all westgerman nuclear facilities occurred in the period of 1980-2003. This amounts to 8-18% of all incidences of cancer for under five years olds living within a radius of 50km around a nuclear facility.

– Bezogen auf alle im Deutschen Kinderkrebsregister gespeicherten Erkrankungsfälle im gleichen Zeitraum entspricht dies einem Anteil von 1,03 – 2,35 %. Bei dieser Zahl muss davon ausgegangen werden, dass es sich hierbei um eine Unterschätzung handelt, weil designbedingt nicht alle betroffenen Kinder erfasst werden konnten. Dieses Risiko liegt erheblich über dem von den Autoren berichteten 0,22%.

-Regarding all cases which are registered within the Deutschen Kinderkrebsregister (German Childhood Cancer Registry) within the given time period this accounts for a share of 1,03-2,35% [of additional cases of cancer]. Due to the design not all affected children could be recorded thus it has to be assumed that this number is an underestimate. The risk is conspicuously above the by the authors reported 0,22%.

9) Die Autoren schreiben, dass „… aufgrund des aktuellen strahlenbiologischen und strahlenepidemiologischen Wissens die von deutschen Kernkraftwerken im Normalbetrieb emittierte ionisierende Strahlung grundsätzlich nicht als Ursache interpretiert werden kann.“
Im Gegensatz zu den Autoren ist das externe Expertengremium einhellig der Überzeugung, dass aufgrund des besonders hohen Strahlenrisikos für Kleinkinder sowie der unzureichenden Daten zur Emissionen von Leistungsreaktoren dieser Zusammenhang keinesfalls ausgeschlossen werden kann. Darüber hinaus sprechen mehrere epidemiologische Kausalitätskriterien für einen solchen Zusammenhang. Es ist jetzt Aufgabe der Wissenschaft, einen Erklärungsansatz für die Differenz zwischen epidemiologischer und strahlenbiologischer Evidenz zu finden.

9) The authors write, that: “… based on the actual radiation-biological and radiation-epidemiological knowledge the in normal operation emmitted amount of ionizing radiation of german nuclear power plants cant be -as a matter of principle – interpreted as being the cause.” In contrast to the authors the board of experts remains unanimously of the conviction that due to the high radiation risk of infants, as well as due to unadequate data concerning the emissions of the reactors this correlation can’t be excluded. Furthermore several causality criteria speak for such a correlation. It is now a scientific task to find an explanation for the deviation between epidemiological and radiation biological evidence.

10) Die Autoren der Studie führen zur Erklärung des von ihnen nachgewiesenen Risikos um Atomkraftwerke noch unbekannte Faktoren (sog. Confounder), nicht näher beschriebene Selektionsmechanismen oder den statistischen Zufall an. Alle drei Erklärungsansätze hält das externe Gremium angesichts der Studienergebnisse für unwahrscheinlich.

10) As an explanation of the by the authors confirmed risks around nuclear facilities, the authors of the study list unknown factors (socalled confounders) , unspecified selection criteria or statistical chance. The board of experts renders all three criteria as unlikely.

Leave a Reply


The below box is for leaving comments. Interesting comments in german, french and russian will eventually be translated into english. If you write a comment you consent to our data protection practices as specified here. If your comment text is not too rude and if your URL is not clearly SPAM then both will be published after moderation. Your email adress will not be published. Moderation is done by hand and might take up to a couple of days.
you can use LaTeX in your math comments, by using the [latex] shortcode:
[latex] E = m c^2 [/latex]