Passive smoking versus Radon gas attributed deaths

[gjonesy]

I got into a discussion with a friend about the risk of second hand smoke, so I did some research and found the statistical data that is some what puzzling. according to what I have read passive smoke kills 7,000 people each year. Radon gas is responsible for 21,000 deaths per year. There are over 55 million "Direct smokers" in the US and there are several Radon hot spots but the cancer data doesn't seem to correlate with the Radon distribution and rate of cancer for the data.

any help would be appreciated

http://www.cdc.gov/cancer/lung/statistics/state.htm

http://www.radon.com/maps/

https://en.wikipedia.org/wiki/Passive_smoking

http://www.cdc.gov/tobacco/campaign/tips/resources/data/cigarette-smoking-in-united-states.html

http://kff.org/other/state-indicator/smoking-adults/

http://ash.org/resources/tobacco-statistics-facts/

http://www.cancer.org/cancer/cancercauses/tobaccocancer/tobacco-related-cancer-fact-sheet

 

[Bystander]

"Lies. Damned lies. And statistics."

 

[gjonesy]

I tend to agree with that assessment Bystander, but that's what's published. The numbers don't make a lot of sense to me. But take into consideration the significant heath risk that is "Direct smoking" and how common the exposure to second hand smoke must be, then compare the Radon hot spot data. seems like those numbers should be opposite. Don't get me wrong I believe both are harmful, but to what degree. How much exposure is a "minimum safe level"? What other kinds of smoke have they compared? How do they tell one cancer and its cause from another? Can they tell radon cancer from other sources like the dust from Nevada's nuclear test sites?

very confusing, but none the less important information to know.
.

 

[Choppy]

Which of those links correlates "radon hot spots" and cancer rates? Overall regional cancer rates can be affected by a lot of other factors than radon exposure (mean population age, regional diets, regional lifestyles, etc.)

 

[gjonesy]

Look at the radon map then look at the CDC cancer map... Places where you'd expect to see higher cancer rates because of the presence of radon don't seem to be higher. However there is this band right through the middle of the US approximately where the mountains are where there are unusually clustered and high lung cancer rates. Places like Ohio have higher amounts of radon but Utah claims to have 100 radon associated deaths per year. Don't make sense to me.

 

[gjonesy]

And BTW I'm not trying to debunk or prove any point... I just really would like to make sense of the data. Smokers for example, there are over 55 million in the US (I have no problem believing that number). Let's say over a life time they expose 3 people to significant amounts of second hand smoke, ( I think that's a gross under estimate) but that comes to 165,000,000 but only 7,000 deaths from second hand smoke, from that number what is it...around 0.007% kinda small don't you think? Then look at the stats for radon, deaths are 2x higher. Is there really that much decaying radioactive material under our feet.

 

[Ygggdrasil]

The 7,000 number is an estimate of the people who die per year from second-hand smoke. The number of people who die from smoking directly is probably much larger than the number who die of SHS or Rn.

 

[gjonesy]

The 7,000 number is an estimate of the people who die per year from second-hand smoke. The number of people who die from smoking directly is probably much larger than the number who die of SHS or Rn.

No doubt about that. In fact that makes an important point. That estimate really doesn't jive. What's more interesting to me than anything else is the CDC lung cancer chart by state. I find it more than odd that there is what appears to be this geographic clustering where cancer rates are so much higher. A cancer belt? That is suspicious to me. Another thing I noticed was Ohio according to the EPA it has significantly high Radon levels. It sits in a valley so not a big surprise but what is surprising is that if you look at the CDC map cancer rates are moderate compared to the mountainous regions on either side of the valley. What else seems suspicious is the reported deaths in Utah which isn't particularly high in radon but reports 100 radon cancer deaths a year in (never smokers). All that being said, I do have an idea about it. But for purposes of keeping my own view out of it and focusing completely on the numbers, the charts and the map data, is this not "fishy"? Personally I think the numbers are inflated for nefarious reasons.

 

[Choppy]

It's good to be questioning the numbers.

With radon deaths, I suspect those estimates are based on a linear no-threshold (LNT) hypothesis for the induction of cancer based on effective doses to the population. There is still a lot of debate about how well this model works. Alternative models suggest a net benefit for small levels of radiation (look up radiation hormesis). The issue is that a lot of the data is for high level exposures, and it is then extrapolated down to low dose exposures because data at the low doses is both rare and noisy. But from a public safety policy perspective, policy-makers try to err on the side of caution and so the LNT hypothesis is often the best model to adopt. In practice, this may lead to an over-estimate the number of deaths attributed to radon exposure. It's probably not an over-estimate by a factor of three though.

According to this:
http://www.cdc.gov/tobacco/data_sta...cts/tobacco_related_mortality/index.htm#women

Smoking generates about 130000 deaths due to lung cancer per year. By your numbers radon is responsible for 21000 and second hand smoke 7000 - so that's roughly 82% smoking, 13% radon and 5% second hand smoke. I'm sure there's significant error bars on each of these, but that would suggest that smoking is roughly six times the factor that radon is when it comes to inducing lung cancer. So when mapping out lung cancer rates by local region, the only way you would see a correlation with radon exposure is if smoking rates were equal across the board. But smoking rates are going to be influenced by tobacco tax rates, state laws and regulations, socio-economics, etc. which will vary from state to state. Hence, when mapping out lung cancer rates, smoking is likely to run the show. Background radon levels that fluctuate by a factor of ~ 2 or so would only be noise unless the smoking levels are corrected for.

There are other factors to consider as well. In regions that are known to have high radon levels, building regulations are likely to account for this. So despite a high background level, it's possible that people who live in the area just take more measures to mitigate exposure, resulting in a population with a lower than average effective dose.

Anyway, in the end I think it's good to ask questions when it seems like the statistics don't jive.

 

[gjonesy]

Yeah, that and the geography has me stumped as well... That cluster of states with higher than average cancer makes me think environmental exposure is a big factor. Studying SHS got me looking, the comparison to radon gas got me suspicious, the geography got me stumped. But thanks for all the information it help make a little more sense.

 

[Ygggdrasil]

Yeah, that and the geography has me stumped as well... That cluster of states with higher than average cancer makes me think environmental exposure is a big factor. Studying SHS got me looking, the comparison to radon gas got me suspicious, the geography got me stumped. But thanks for all the information it help make a little more sense.

According to the KFF link you posted (http://kff.org/other/state-indicator/smoking-adults/), the five states with the highest percentage of adults who smoke are WV, KY, AR, TN, and LA. All of these states have a high incidence of lung cancer (according to your CDC link http://www.cdc.gov/cancer/lung/statistics/state.htm). Similarly, the states with the lowest percetage of adults who smoke are UT, CA, HI, NY, and TX, and all of these have very low incidence of lung cancer. The incidence of lung cancer seems to correlate pretty well with the fraction of the population that smokes.

 

[gjonesy]

Yeah but isn't that "odd" the cluster? And it remains constant over several years? What I would expect to see, and even yet random distribution over several years, smoking is a factor don't get me wrong I totally agree on that point but it seem from a commonsense stand point that if you have certain risk factors then you don't want to be living in those areas...some hidden variable we don't quite yet understand.

 

[russ_watters]

For everyone's info, the OP and I had a PM conversation, so if you see any disembodied quotes, that's where they came from (by permission).

I got into a discussion with a friend about the risk of second hand smoke, so I did some research and found the statistical data that is some what puzzling. according to what I have read passive smoke kills 7,000 people each year. Radon gas is responsible for 21,000 deaths per year. There are over 55 million "Direct smokers" in the US and there are several Radon hot spots but the cancer data doesn't seem to correlate with the Radon distribution and rate of cancer for the data.
[PM conversation post]
Utah has a publicized high incidence of radon lung cancer death but not much radon according to the EPA, Kansas is eat up with radon but has a normal rate of lung cancer.

The way I'm understanding your issue is that you are trying to get the cancer rate totals to add-up and are having trouble. And in particular you think that radon hot spots should show up in the cancer data. I think you are over-emphasizing the risk from radon by not figuring-in the risk from smoking. Radon would only substantially impact the total number of cancer deaths if it is a substantial fraction of cancer deaths. Otherwise an area could have, for example, double the average number of radon deaths and still be lost in the noise of smoking cancer deaths. It appears to me that that's what is happening. So let's take a concise look at the key stats:

# of cancer deaths per year from:
Smoking: 130,000
Radon: 21,000
"Passive" smoking: 7,000
Other: 19,000
Total: 158,000
[from your sources]
Note that the sum of the causes exceeds the total by 12% due to overlap between the causes and uncertainty in the studies.

So based on these numbers, if, for example, you had a location with double the average radon deaths, that's right at the edge of the error margin of the total and has a relatively small (12%) impact on the total. Since the spread in death rates between the highest and lowest state is 300%, a variation in radon death rate would have to be pretty big to be visible in the data.

So the question is, what is causing that "cancer belt" you are seeing in the national map? The obvious choice: differing smoking rates:

http://www.cdc.gov/mmwr/preview/mmwrhtml/mm5935a3.htm

Is there really that much decaying radioactive material under our feet.

Scary, but yes. I have a radon abatement system on my house.

 

[russ_watters]

Yeah but isn't that "odd" the cluster? And it remains constant over several years? What I would expect to see, and even yet random distribution over several years...

Smokers, particularly those with the highest cancer risk, smoke for a long time and perhaps their entire lives. And smokers give birth to smokers. So no, I would not expect the geographical differences in the smoking rates to change quickly.

 

[gjonesy]

Actually I'm questioning the radon data... But you are very "warm" in your thinking. Yes smoking kills smokers, SHS is bad, radiation for nonsmokers in certain areas seems pretty clearly bad. Biggest question in my mind is... Is this naturally occurring? With out over stepping PF rules simply because I think this is important ( please look really close at the data. ) there is a "smoking gun" variable that's never considered... I'll leave it at that.

 

[gjonesy]

I also discovered that some of this is based on metadata,... what's the margin of error? can that be calculated?

 

[indio007]

Remember this?
http://news.heartland.org/newspaper...eral-court-rejects-epa-secondhand-smoke-study
Federal Court Rejects EPA Secondhand Smoke Study

"In his blistering 92-page decision, Judge William Osteen of the Middle District of North Carolina essentially vindicated those who had accused EPA of manipulating data in order to reach a preconceived conclusion. Osteen ruled that EPA had violated provisions of the 1986 Radon Gas and Indoor Air Quality Act, under which the agency determined that exposure to ETS is hazardous."

"Osteen added that EPA’s findings were based on insufficiently rigorous statistical tests and were therefore invalid. EPA, he noted, “disregarded information and made findings based on selective information . . . ; deviated from its risk assessment guidelines; failed to disclose important [opposition] findings and reasoning; and left significant questions without answers.”"

 

[russ_watters]

Remember this?
http://news.heartland.org/newspaper...eral-court-rejects-epa-secondhand-smoke-study
Federal Court Rejects EPA Secondhand Smoke Study

That study is 22 years old. Quite a bit more research has been done in the meantime.

 

[DrDu]

I am working as an epidemiologist and have been working on radon, too. Most important points have been brought up by others already.
Indeed the risk estimates for radon exposure are based mainly on data from uranium miners exposed to radon in mining or processing. These occupational exposures are much higher than radon levels in homes. The cancer risk from radon in homes is hard to extract from ecological studies, and even less from looking at cancer incidence maps as the main confounding risk factors for cancer - age and smoking - vary too much.
In US, current legislation is mainly based on the BEIR VI report http://www.nap.edu/read/5499/chapter/1 which is based mainly on the results of an analysis of uranium miner studies in the US.
On an international level, important for current legislation are the UNSCEAR 2006 report of the WHO is important. However, it also relies heavily on the findings of BEIR VI. A more recent important review is the report 115 of the International Commission on Radiation Protection (ICRP) http://www.icrp.org/publication.asp?id=ICRP Publication 115
which includes also results from large other studies.

 

[shredder]

My understanding is that there at least two sources of radon - radon gas from geological sources, and radon adsorbed on to tobacco leaves
whilst they are growing. The latter decays through radon daughters to 210Polonium - a hazardous alpha emitter. Damage to the brush
border epithelium of the bronchial tree prevents clearance of the polonium loaded secretions in smokers, who have increased risk of malignancy due to exogenous radon in air, with further risk from inhaled radon products from tobacco, as well as the carcinogenic effects of polycyclic hydrocarbons and similar compounds from the tobacco itself. It is likely that the damage to DNA caused by chemical plus radiological insults is synergistic. To use a rather tired cliche, a "perfect storm" of carcinogenic factors.
Incidentally, cancer incidence is affected by other factors, notably the lag time for radiation induced carcinogenesis (~years) and population mobility.

 

[DrDu]

As I remember, the risk due to polonium adsorbed to the leaves of tobacco is rather small. Another problem with judging the risk due to radon from maps is that the radon concentration in buildings depends on a lot of factors besides concentriation in the ground: Ventilation behaviour, type of the building (stone or wood, cellar or not), the floor you are living in (usually, enhanced concentrations are only measured in the souterrain and ground floor, ...