Claimed evidence for the LNT model
claim low-dose radiation increases cancer risk
receive tremendous publicity in popular media
But, on close scrutiny,
major flaws become evident
raise major doubts about the publication and/or invalidate their conclusions.
A few examples are given below:
15-country study of radiation workers (Cardis, 2005)
All combined data show statistically significant increase in cancers in radiation workers.
Conclusion of publication: low-dose radiation increases cancer risk.
But, please note that the Canadian data
shows much higher risk than other data
if removed, All combined risk would not be statistically significant
In 2011, problems identified in the Canadian data (CNSC, 2011)
invalidates conclusion of publication of low-dose radiation cancer risk
(Cardis, 2005) study was quoted by BEIR VII report (2005) to infer increased cancer risk from low-dose radiation and to confirm the radiation cancer risk factor.
Note: Cardis et al. have not yet (as of end of 2015) published a retraction of the conclusion of (Cardis, 2005) publication.
The publication continues to be quoted to justify low-dose radiation cancer concerns.
Increased cancers following childhood CT scans (Pearce, 2012)
Conclusion: Brain cancer risk increases with radiation dose
Major problem with the study:
Likelihood of reverse causation (Walsh, 2014), (Boice, 2013)
(Boice, 2015) "The critical problem is that the reasons for performing the CT examinations were not known, and these reasons (rather than the CT radiation dose) were the probable cause of the subsequent cancer diagnoses. This is called ‘confoundingby indication’ or ‘reverse causation’."
Thus, it is a poorly designed study, and the study conclusion is not valid.
Other studies, that have considered the reason for performing CT scans, have shown no increase in cancer risk with CT radiation dose, e.g. (Huang, 2014), (Journy, 2015).
However, (Pearce, 2012) is routinely quoted as evidence for cancer risk from low-dose radiation:
2014 Opening statement in the debate on health effects of low-dose radiation
2014 Presentation at BEIR VIII Scoping Meeting
2015 INWORKS study of leukemias in radiation workers
2015 Comments by NCI on the NRC petitions regarding the LNT model
2015 Comments by Ian Fairlee on the NRC petitions regarding the LNT model
2015 Comments by EPA on the NRC petitions regarding the LNT model
2015 Comments by ACMUI on the NRC petitions regarding the LNT model
Increased cancers claimed in irradiated Taiwan apartment residents (Hwang, 2008)
In an earlier publication, (Hwang, 2006), the authors had stated that overall cancers were significantly lower in the irradiated population, when compared to an age-matched control population.
Please note that they used 95% CI in the 2006 publication.
If they had used the same CI in 2008, their conclusion would be no increased risk of cancers for leukemia and breast cancers.
However, they used the 90% CI in the (Hwang, 2008) publication and made the claim of increased risk from low-dose radiation.
Also, in (Hwang, 2006), the most significant result was the 20% reduction in overall cancers, as seen in above table.
Let us now estimate what was the overall reduction in cancers in 2008 publication, using data on cancer incidence reported in the publication:
All Cancers expected in Hwang 2006 paper was 114.9,
All Cancers expected for Hwang 2008 paper, with 3 years additional follow-up period, would be 114.9*1.36 = 156
(Correcting for the change in average age from 33.3 to 36.3)
All Cancers observed (in Hwang 2008 paper) was 128
Hence, Observed/Expected cases would be = 128/156 = 0.82
Thus, there is a 18% reduction in cancers observed in the follow-up paper.
This was not mentioned in the publication.
The conclusions presented in (Hwang, 2008) are misleading.
INWORKS study was designed to answer the question:
Does low-dose radiation over extended period of time increase cancer risk?
For this they studied:
leukemias and solid cancers as a function of occupational radiational dose
Study would be valid if
occupational dose was the predominant dose to the workers and
other doses were small in comparison to occupational dose
However, during the period of study (1944-2004)
occupational doses decreased drastically
per capita medical radiation dose increased considerably
For later years,
occupational dose would not be a valid indicator of total radiation dose
This major design flaw makes the study conclusion not dependable.
A pattern is beginning to emerge for publications claiming increased cancer risk from low-dose radiation. Major faullts are identified after the studies have been published negating their conclusions.
Next Section: Weak defense of the LNT model