It is difficult to get a man to understand something, when his salary depends on his not understanding it.[Upton Sinclair]
The Gordian Knot preacher has messed up again. The single most important background radiation study is the Karunagappally cohort which compares cancer incidence in several very high dose rate villages on the Kerala coast with the cancer incidence in neighboring villages with normal background dose rates. In 2009, Nair et al published the results, of the 15 year study, Figure 1.\cite{nair-2009}
Figure 1. Karunagappally Cancer Incidence, 2009,
The high end group which averaged 0.16 mSv/day (60 mSv/y) had a slightly lower cancer incidence than their low end neighbors. According to LNT, the high end group should have a 6% higher cancer rate than the low end. To most of us, this would be an ugly fact killing a beautiful theory. To Doctor Nair, it only meant that LNT is not under-stating the risk "our cancer incidence study, together with previously reported cancer mortality studies in the High Background area of Yangjiang, China, suggest it is unlikely that estimates of risks are substantially greater than currently believed". The Oracle of Delphi would be proud of that wording.
In 2021, the same group updated the Karunagappally study.\cite{amma-2021} Somehow I missed this. Thanks to Ken Chaplin for alerting us to this work. In the update, they approximately doubled the size of the cohort to 149,585 people and increased the average time in the study from 10 years to 19 years. Figure 2 summarizes the results.
Figure 2. Karunagappally Cancer Incidence, 2021.
The relative risks have changed very little; but the error bars have been squeezed down a lot. The study is now up to nearly 3 million person-years of data, and that ignores the fact that most of these people have endured the same dose rates for their entire lives, not just the time they were in the study.
When LNTers are confronted with data such as Figure 2, the inevitable response is: we just got unlucky. The linear increase could be there; we don't have enough statistical power to rule it out. For these people, their speculative theory, LNT, which denies our bodies can repair radiation damage, is the null hypothesis, inverting one of the most basic rules of science. Since LNT is the null hypothesis, this negative result does not mean we must abandon LNT.
But even is we accept the preposterous idea that LNT should be the null hypothesis, this data is damning. Brenner, a strong supporter of LNT, points out that, according to the National Research Council, to be statistically confident of the impact of a 5 mSv difference in dose, we would need to study a population of 7.9 million exposed people, with the age distribution of Americans, for the remainder of their lives, roughly 40 years.\cite{brenner-2003} But he is done in by LNT's cumulative assumption. Under the same rules, the NRC estimated we would need to study a population of 20,000 for the rest of their lives to confidently detect the LNT effect of a dose of 100 mSv.\cite{nrc-1995}[Table 7-2] That's about 800,000 people-years. In the updated Karunagappally study, we have 919,000 people-years over 100 mSv of which over 200,000 people-years is over 200 mSv. The study now comfortably meets Brenner's requirement for shooting down LNT, even if LNT is fallaciously assumed to be the null hypothesis.
However, all the Nair group will concede is the updated study might be "suggesting a possibility that the solid cancer risk associated with the continuous exposure to low dose rate radiation is significantly lower than that associated with acute exposure." Imagine spending 20 years of your life staring at this data, and that is all you can come up with.
Sinclair was wrong. It's not difficult; it is nigh on impossible.
Hi, Jack - The key is in the error bars. For either of these datasets, the proper null hypothesis is that long-period exposure to low levels of radiation has no effect at all on cancer risk. Both datasets support that conclusion, but for the first one, any line from a +20% increase in risk over the dose range studied to a -20% protective effect would fall within all the (I would hope the standard) 95% error bars. For the later version, the fourth error bar barely creeps above zero. That's the one that looks like it falsifies LNT and comes close to suggesting a fairly high probability of a protective effect (>90% of the error bar is in negative territory, and it's relatively narrow compared to the higher-dose bars.
But I don't understand something about the error bars. The third (50-99) bar has nearly 10 times the person-years of the 4th, yet its error bar isn't that much narrower. I'm still of the mind the best conclusion from the data is no effect.
Could you explain this a bit simpler for those of us who are technically less gifted, including the null hypothesis bit? Thank you.