RSH Data & Documents

"Low Level
Radiation Health
Effects: Compiling
the Data"

Revision 2
March 30, 1999
by Radiation, Science, and
Health, Inc.,
Edited by J. Muckerheide

1.2.6.3
Natural Activity and Radiation - Radon

"So far, only few data can be found in the international literature on lung cancer epidemiology in one of the highest residential radon areas of the world, the former Soviet uranium mining districts in southern Saxony. The observed rates among (essentially never-smoking) females have been found to be substantially lower than in the population average. Thus various ecological and case control studies in other countries, showing no detrimental effects of in-creased residential radon levels, can be confirmed in an area which has been closely associated with the history of ionizing radiation and it’s biological effects."

"There has been a trend towards lower dose limits over the last half century, with the evaluation of the Japanese bomb survivors as main scientific evidence. It remains, however, questionable that the high dose-rate, high-level exposures to a mixed gamma-neutron radiation field under extreme stress conditions can be extrapolated to the low-dose, dose-rate exposures under which all species, including man, evolved. There is evidence that the exposure of large populations to substantially increased natural radiation levels over many generations may provide a more realistic better test of the LNT hypothesis. Various investigations in areas with high external radiation such as in Kerala, India, or parts of Brazil and Iran, did not show any detrimental health effects. At least equally important is the epidemiology of high internal exposures due to residential radon. Some of those studies in various countries showed indications of a lung cancer increase with increasing radon levels in particular among heavy smokers (e.g. in Sweden). However, in others no effect of radon (e.g. in Canada and Finland), or even bio-positive effects (e.g. in China), instead of a doubling of ‘natural’ lung cancer for approximately 100 Bqm-3 as predicted by the ‘official’ ICRP recommendations have been found in ecological as well as in case control and cohort studies (for compilations, see Becker and Schuttman 1998; Lubin 1998).

"Since the 1989 German unification, previously secret data from the large former Soviet uranium mining operations in the Ore Mountains (Erzgebirge) in Southern Saxony and Thuringia became available. In particular, the old mining towns around Schneeberg are of interest because of dense population and very high exposure levels."

"Indoor radon levels in Germany vary between 20 and more than 100,000 Bqm-3, with an average around 50. Some of the highest concentrations occur in the former Kingdom (now Free State) of Saxony, where silver and other metals have been mined since the Middle Age, and where the famous physician Paracelsus already described the high rate of lung diseases among miners in 1537.

"Also regarding biological effects, this area has been a focal point of relevant research (Schuttmann 1988; Schuttmann 1996), in particular because its radon spas such as Oberschlema, Bad Brambach, and Joachimsthal have been highly successful in curing Morbus Bechterew, arthritic, rheumatic and other diseases, as confirmed in double-blind studies (see Pratzel 1993), for most of this century.

"Other studies demonstrated that residential radon concentrations in the area were high, sometimes extremely high by contemporary standards, with intervention levels around 150-250 Bqm-3. Nevertheless, lung cancer rates among the residents of such houses had not increased, as noted already in a pathology textbook by a professor of Leipzig university in 1882 (Cohnheim 1882), in which he states: 'One can observe that those inhabitants of Schneeberg who never enter a mine will never get this lung disease'. The obvious reasons were the much lower concentrations in comparison with those in the mines, and the absence of the many confounding factors mentioned above.

"This was also noticed by radiologists from Dresden, the capital of Saxony, who studied 70 y ago the lung cancer mortality in the high radon areas and found, in a time when the diagnostics of this disease had already been well-established, only two cases of lung cancer in the region (Saupe 1928). This was surprising because in other areas of Saxony, due to the fact that Germany's first German cigarette production had started in Dresden in 1862, lung cancer rates had clearly increased.

"After WW II, when the former Soviet Union started to extract eventually over 200,000 tons of uranium from one of the world's most important uranium mining regions to create its nuclear potential, all matters related to radiation effects were considered highly secret. Nevertheless, first estimates in East Germany (the former GDR) seemed to suggest that, based on ICRP and other publications, up to 50% of all female lung cancers of the country might be radon-related (Schüttmann 1983). This estimate implied that the rate in the high-radon areas should be substantially higher than the average.

"However a detailed analysis of the very comprehensive GDR cancer register clearly showed no indication of increased female lung cancer in this region (Mehnert et al 1992). In fact, in one of the high-radon mining districts (Gera) the values were among the lowest of the whole country. Nevertheless, extremely sophisticated remediation efforts have been initiated there after the German unification in 1990, with a radon-related cost component of about 2,000 mill $ U.S.(Becker 1996).

"It was found, for example, that in the old mining town of Schneeberg, with an average radon concentration in homes of 290 Bqm-3, 13% exceeded 1000, 11% 15,000, and over 1% more than 15,000, with a maximum of 115,000 - in essentially the same houses in which no lung cancers had been detected in the careful studies published 70 y ago (Saupe 1928)."

"This area appears, therefore, to be an interesting location for testing the currently dominating assumptions about low-dose radiation health hazards in general, and indoor radon hazards in particular. A recent study claimed that the relative residential radon risk of never-smokers substantially exceeds that of smokers, amounting to 14% at 50 Bqm-3 among non-smoking women (Steindorf et al 1996). With at least 90% of the women in this area being non-smokers in the tobacco deficient post-war years (as late as the nineties, 88% of the females in smaller communities of the area were never-smokers), and the average radon levels in the houses exceeding the East German average by factors between 3 and 10, a substantial increase in lung cancers should be observed. The actual results, comparing the expected (based on the country average) and the actual numbers of female lung cancers in the high-radon areas between 1983 and 1987 (Arndt 1992) show exactly the contrary, of what should be observed based on the relevant ICRP recommendations (Table 1).

Table 1

"This result is confirmed by the observation that the lung cancer rate in the southern parts of the former GDR, in which all the uranium mining and processing took place, has consistently been lower than the GDR average, in agreement with findings of other ecological and case-control studies, that the residential lung cancer rate appears to be inversely related to residential radon concentration.

"Similar results have also been reported recently by others (Conrady et al 1996), who carefully examined the lung cancer register of (essentially non-smoking) females in communities in Saxony between 1961 and 1988. The communities have been classified, based on extensive measurements by the German Federal Radiation Protection Institute, into four radon exposure categories between less than 100 and more than 500 Bqm-3 (Table 2). In this table, the observed values are compared with the expected values for a standard population in Saxony, taking into account the smoking habits (third column). Column 4 presents the hypothetical number based on the LNT hypothesis, assuming an equilibrium factor f=0.4. The discrepancy between those columns, increasing with higher radon levels, is obvious.

Table 2

"The authors (Conrady et al 1996) came to the conclusion, that 'even in the highest exposure category, there is an obvious deficit in (lung cancer) cases. An additional risk, as expected on the basis of the LNT hypothesis, could not be detected...There are differences between reality and the hypothesis about a linear relation between radiation dose and effect without threshold.' In summary:

1. The observed cases at the high radon levels are below the statistical expectation.

2. As the radon levels increase, the observed cases decrease, thus clearly indicating the contrary of current ICRP assumtions about LNT.

"Numerous other recent publications, of which only a few are mentioned here, provide direct or indirect supporting evidence for these findings. The most recent publication is from Hungary, in which, with a probability not less than 98%, a lower risk of all types of cancer has been found among women younger than 61 y in the 110-165 Bqm-3 residential radon category (Toth et al 1998). Also, the so-called ‘leukemia clusters’ which have been claimed to exist in the proximity of nuclear installations in Germany, France, the U.K., etc., are much less likely to occur in the areas with high natural radiation levels. In particular, no clusters of childhood leukemia have been found in the high radon areas of Saxony. (Conrady et al 1997).

"A recent survey of extensive studies of low-LET radiation on radiogenic lung cancer (Rossi and Zaider 1997), in which a threshold of approx. 2 Gy has been found, also supports such results. As the authors of this study concluded: ‘The evidence for no, and probably a negative, risk of lung cancer at small doses not only conflicts with, linearity’, but also invalidates risk estimates based on non-uniform irradiation.

"It should also be noted that studies on lung cancer caused by inhaled Pu-239 in workers of the large Mayak nuclear enterprise in Russia (Tokarskaya et al 1997) also indicate a threshold around 0.8 Gy. Another important evidence for the correct assessment of radiogenic lung cancer risks has been provided by studies with mice, which have been exposed to identical aerosols spiked with an alpha-emitter (Cm-242) or a beta-emitter (Ca-45) with the same half-life of 163 days (Kellington et al 1997). The authors conclude that ‘the study was unable to confirm the appropriateness of a weighting factor of 20, and that the RBE determined for lung tumor induction over the range of doses employed was approximately two.’ In view of the numerous studies showing adaptive response and hormesis (for compilations see UNSCEAR 1995, Muckerheide 1998, Becker in press), such results are not particularly surprising."