When I was buying a house, one of the environmental checks that had to be done was the level of radon buildup in the building. It turns out the house that I bought has just below the recommended safe level in this country (UK) of radon buildup, and I was advised to leave windows open to let the radon dissipate whenever possible.

Radon is the second most common cause of lung cancer, after smoking... and yet, when I looked into whether areas in the country with higher radon ground output had statistically higher incidences of lung cancer, they did not. Does radon gas really pose a significant threat in causing lung cancer?

  • 10
    You might actually find an inverse effect - where radon gas is expected, it's tested for and mitigated. Where it's not expected, one might not test for it, resulting in higher incidence of cancer. In other words, your basis for the question is interesting, but of little value other than to cause you to question the assumption. Simple faq sheet here: cancer.gov/cancertopics/factsheet/Risk/radon but, of course, someone will point directly to the studies in an answer here soon, I suspect.
    – Adam Davis
    Commented Jun 15, 2011 at 19:42
  • +1 Interesting: scientific (not pseudo-scientific), and not question that it would have occurred to me to ask - et (until I saw the posted answer) not a question for which I knew the evidence.
    – ChrisW
    Commented Jun 17, 2011 at 4:09
  • I'm going to have to dig for references here, but IIRC the main problem with radon is that there is no proof whatsoever for a linear dose-risk relationship, and research in fact showed a negative correlation at doses below the legal limit (although that effect was too small to be statistically significant) edit Cohen, 1990, 1995, Thompson et al. 2008, from WP's Talk page on Radon.
    – MSalters
    Commented Jun 17, 2011 at 8:18
  • BTW, the UK recommended safe (Target) level is 100 Bq/m3; the Action level is 200 Bq/m3. (from ukrandom.org)
    – MSalters
    Commented Jun 17, 2011 at 8:31
  • In response to the answer post: First off, pCi/l is an SI unit, just one that is only commonly used in the USA. Interesting that even when the USA tries to be international, they pick obscure SI units that the rest of the world no longer uses. Second: We need to divide radon risk groups into 3 categories: people living or working in environments with high concentrations (like uranium mines and a few homes), smokers, and everyone else (nonsmokers in homes of low concentrations, below 8 pCi/l or about 300Bq/m3). Evidence that radon increases the risk of lung cancer is conclusive for the first 2
    – user11872
    Commented Mar 10, 2013 at 16:47

1 Answer 1


It does indeed.

To start with, the Wikipedia page on radon is pretty great (it's rated a "good article").

Now for facts: Radon was discovered to be a problem (woohoo, 1960's Time article) through studies of miners back in the 1950's, when just a few too many non-smoker miners started getting lung cancer. These were all uranium miners, and as radon is a direct byproduct of uranium decay, one thing led to another (emphasis mine):

To estimate the risk of lung cancer mortality among nonsmokers exposed to varying levels of radon daughters, 516 white men who never smoked cigarettes, pipes, or cigars were selected from the US Public Health Service cohort of Colorado Plateau uranium miners and followed up from 1950 through 1984. Age-specific mortality rates for nonsmokers from a study of US veterans were used for comparison. Fourteen deaths from lung cancer were observed among the nonsmoking miners, while 1.1 deaths were expected, yielding a standardized mortality ratio of 12.7 with 95% confidence limits of 8.0 and 20.1. These results confirm that exposure to radon daughters in the absence of cigarette smoking is a potent carcinogen that should be strictly controlled.

Adam Davis' commented link to the National Cancer Institute is certainly informative, as are a few pages created by the Environmental Protection Agency (EPA) (both links share a certain amount of information). The first (both, in fact) EPA link has a nice "risks chart" for smokers and non-smokers, showing the levels of radon exposure and the correlation to cancer occurrence. This chart will be better understood, though, if one knows something about measurement meanings (units are in this case, as in most cases when dealing with radiation, diverse and un-unified at best). Taken from the Wiki article above:

  • Radon is often measured in picocuries per liter (pCi/L).
  • The more "standard" (read: SI) measurement is becquerels per cubic meter, where 1 pCi/L = 37 Bq/m3
  • The mining industry uses something entirely different, of course, called a working level (WL). Since this is where all the data initially came from, what we now use has been translated from it. But for consistency's sake, one month's worth of exposure to a WL, a working level month (WLM), is "roughly equivalent to living one year in an atmosphere with a radon concentration of 230 Bq/m3."

It'll be even better understood once one knows that, from the second EPA article, "The average radon concentration in the indoor air of America’s homes is about 1.3 pCi/L" (~48 Bq/m3). But some are going to be more, and it should absolutely be tested on a case by case basis (a figure the Wiki article claims in the Concentration Scale section is a universal rating of 100 Bq/m3 (2.7pCi/L) for a house, which is certainly higher than the EPA's recommended amount).

To wrap this up with some more modern statistics, take a look at the EPA fact sheet from a pooled study from 2005 (emphasis mine, extra parenthesis and spacing issues theirs):

Overall, the odds ratios for lung cancer increased with increasing radon exposure categories, with an odds ratio of 1.37 (0.98-1.92)) for concentrations exceeding 200 Bq/m3 [5.4 pCi/L] relative to concentrations under 25 Bq/m3 [.68 pCi/L].The overall estimate of the excess odds ratio for lung cancer per 100 Bq/m3 [2.7 pCi/L] was 11% (0%-28%).

That info appears to meld smokers and non-smokers, but that's fine. The standard claim (from almost all of the citations here) seems to be that about 21,000 Americans die every year from radon exposure, which is also a combined number (about 2,900 non-smokers, according to the second EPA page). One can quickly see that this 21,000 number is far less than 11% (which is good, especially since that's only the increased odds of getting cancer, not dying from it) - it's actually 0.007% of the overall population, or about 0.001% for non-smokers (given an American population of about 300 million). This shows that our houses, on average, are pretty good about having low radon levels. If we take the 1.3 pCi/L (48 Bq/m3) EPA estimate to be true, then we're at about half of the "universal number."

So even if you're a smoker, radon isn't going to enormously change your risk of dying from cancer (unless you live in a particularly radioactive area), but there is a definite and defined increase, and you'll certainly be more likely to develop the disease. It's a very easy "better safe than sorry" call to make.

(I apologize that all of this data primarily concerns the US and not the UK, but I'm sure this all holds roughly true anywhere in the world.)

  • 1
    Staying skeptical, who's to say that the dust during mining uranium isn't enough to explain the increased rate of lung cancer? Commented Mar 11, 2013 at 14:32
  • 1
    Miners have much higher levels of radon exposure than homeowners, and it's sporadic, so every once in a while, they get a huge blast of radon. The extrapolation to homeowners is using linear-no-threshhold model which is false. In all likelihood, due to repair mechanisms in the body, radon poses zero risk in environmental doses, and this is a surprising statistical discovery of the 1990s which is still not completely settled, but is supported by the predominance of the evidence.
    – Ron Maimon
    Commented Dec 4, 2014 at 11:50
  • @0xC0000022L - One would have to demonstrate that they were subject to exposure. If they have safety equipment that can be objectively shown to protect against that exposure, then there's no reason to think that is the mechanism. Or, if the levels of radioactivity and concentration of uranium, itself, in the mined ore is low enough to not present a risk, then that would also be a reason. One would expect that, if they are trying to pinpoint radon effects, they'd have thought to account for potential confounding uranium factors. Commented Sep 8, 2017 at 13:03
  • Your idea isn't that novel or unexpected that no one conducting the study or people who evaluated the study afterward would have neglected to think of it.. Commented Sep 8, 2017 at 13:04
  • @PoloHoleSet While the correlation is established by the study, I don't see it establishing causation. Commented Sep 8, 2017 at 14:29

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .