Talk:Passive smoking

From Wikipedia, the free encyclopedia
Jump to: navigation, search

Enstrom and Kabat POV and bad sources[edit]

It says:-

Notably, the study had failed to identify a comparison group of "unexposed" persons.[117]

Source 117 has a some warped logic.

1. You cannot generate an RR without a comparison group.

2. The author of this source claims there is no comparison group because smoking was allowed in public places at the time so therefore everyone was exposed to ETS, i.e. impossible to find someone not exposed. The guy is crazy, the control group is good. Even if the groups are slightly redefined as this guy suggests; one group is exposed to ETS at home and outside the home, the control group is just exposed to ETS outside the home from time to time.

3. We have to remember that the American Cancer Society commissioned this report and only withdrew funding when they realized they were not going to get the results they wanted.

4. Published in the British Medical Journal which provoked a barrage of condemnation in which the then BMJ editor Dr Richard Smith was accused of every failing from naivety to active promotion of evil. His accusers demanded that he withdraw the article. To his credit, Smith refused, pointing out that the BMJ exists to publish science not polemic, and that the American study was proper, peer-reviewed science. A robust and persuasive anti-smoker, he replied that although the BMJ was "passionately anti-tobacco" it was not "anti-science". He went on to explain that "the question [of whether passive smoking kills] has not been definitively answered." "I found it disturbing that so many people and organisations referred to the flaws in the study without specifying what they were. Indeed, this debate was much more remarkable for its passion than its precision."

Source 118 - Is a letter asking for funding in a study in low levels of active smoking no doubt because Enstrom believed as the European study (Bjerregaard, Raaschou-Nielsen, Sørensen et al 2006 - see Richard Doll and Occasional smoking below) found that there was no relationship between occasional smoking and lung cancer.

Source 119 - Has the subtitle "Did the tobacco industry skew results of survey?" probably not! since the results are consistent with the WHO study on passive smoking. "Criminal racketeering" please.

The other sources are irrelevant Enstrom admits that after the ACS pulled out the tobacco industry funded the end of the report.

What is notable is that no-one has pointed out specific problems with the research and the findings are consistent with other studies.

The WHO study (Boffetta 98) found:-

Social study Male & Female RR = 1.03

Effect on children RR = 0.78 (note that they found ETS beneficial to children)

Workplace RR = 1.17

Spousal RR = 1.16

(NONE OF THESE ARE SIGNIFICANT ASSOCIATIONS, RR is less than 2. For the positive effect in chrildren to be significant it should be under 0.5).

Enstrom and Kabat

Lung Cancer Men RR = 0.75

Lung Cancer Women RR = 0.99


This is all consistent.

Here is a list of 57 partner studies (one partner smoked, the other didn't) from International Agency for Research on Cancer (IARC)

RR's over 3 = 0

RR's 2.5 or higher = 2

RR's over 2 = 7

RR's under 1 = 8

RR's 0.5 and under = 2 (i.e. passive smoking is beneficial)

RR's marked NR = 12 (i.e. RR was not reported)

With RR's like this any epidemiologist worth his salt will see this as inconclusive. That is an average of 1.39 of the 45 studies not marked NR.

Occasional smoking and Richard Doll[edit]

I suggest both of these are important to this subject.

Sir Richard Doll's "British Doctors Study" the study that proved the increased risk of lung cancer, concludes that "on average those who smoke until age 30 have no excess mortality, those who smoke until age 40 lose 1 year, those who smoke until 50 lose 4 years, and those who smoke until age 60 lose 7 years".

On Desert Island Discs in 2001, Sir Richard Doll, the man who proved the incontrovertible causal link between active smoking and lung cancer, said: "The effect of other people smoking in my presence is so small it doesn't worry me."

The risk of lung cancer according to Doll's groundbreaking study 1950, table 14.

No smoking- 50 in 1 million = 0.00005% (1 non smoker in 20000)

1 to 4 cigarettes per day- 307 in 1 million = 0.000307% (1 light smoker in 3257)

50+ cigarettes per day- 3344 in 1 million = 0.003344% (1 heavy smoker in 300)


In a pan European study the effects of occasional smoking were studied with no significant results

The effect of occasional smoking on smoking-related cancers

Bjerregaard BK, Raaschou-Nielsen O, Sørensen M, Frederiksen K, Tjønneland A, Rohrmann S, Linseisen J, Bergman MM, Boeing H, Sieri S, Palli D, Tumino R, Sacerdote C, Bueno-de-Mesquita HB, Büchner FL, Gram IT, Braaten T, Lund E, Hallmans G, Agren A, Riboli E (December 2006). "The effect of occasional smoking on smoking-related cancers: in the European Prospective Investigation into Cancer and Nutrition (EPIC)". Cancer Causes Control 17 (10): 1305–9 "A 2006 European study on occasional smoking published findings that the risk of the major smoking-related cancers for occasional smokers was 1.24 times that of those who have never smoked at all but the result was not statistically significant. (For a confidence interval of 95%, this data showed an incidence rate ratio of 0.80 to 1.94.)"

Article Doesn't Reflect Latest Research[edit]

There is "No Clear Link Between Passive Smoking and Lung Cancer" per Journal of the National Cancer Research Institute, December 17, 2013. — Preceding unsigned comment added by (talk) 22:24, 24 December 2014 (UTC)

Why "causes" is better than "can cause"[edit]

I repeat here what has been said before about the two variants ("causes" vs "can cause") of the second sentence in the lead paragraph.

I have expressed the case for the first formulation very extensively in a previous Talk contribution (see [1]), which has not been challenged. Changing "cause" to "can cause" fails to make the point clearer - it rather obscures it, as it implies a meaning of "cause" which is different from the way the term is used in epidemiology and by the public health community (see definition of causality in epidemiology).

Moreover, the lead paragraph is a summary of the article and should be consistent with the body of the article. The "cause" formulation is consistent with all other references to causality in the body of the article, which have been around for a long time and have been well accepted by all editors:

  • "secondhand smoke ... causes the same problems as direct smoking" (first sentence of Effects section)
  • "the 3rd leading cause of preventable death" (last bullet point before Children section)
  • "Another research financed by the Swedish National Board of Health and Welfare and Bloomberg Philanthropies found that passive smoking causes about 603,000 death a year, which represents 1% of the world's death." (same paragraph)
  • "In France passive smoking has been estimated to cause between 3,000 and 5,000 premature deaths per year" (last paragraph of Evidence section)
  • "The International Agency for Research on Cancer of the World Health Organization concluded in 2002 that there was sufficient evidence that secondhand smoke caused cancer in humans" (first sentence of Risk level section)
  • "The governments of 168 nations have signed and currently 174 have ratified the World Health Organization Framework Convention on Tobacco Control, which states that "Parties recognize that scientific evidence has unequivocally established that exposure to tobacco smoke causes death, disease and disability." (last bullet point in list below Opinion of public health authorities section)
  • "In 1993, the United States Environmental Protection Agency (EPA) issued a report estimating that 3,000 lung cancer related deaths in the United States were caused by passive smoking annually." (first sentence of EPA lawsuit section)
  • "...the District Court for the District of Columbia found that the tobacco industry ... responded with 'efforts to undermine and discredit the scientific consensus that ETS causes disease'." (Tobacco industry response section, 1st paragraph)
  • "The U.S. District Court, in U.S.A. v. Philip Morris et al., found that '...despite their internal acknowledgment of the hazards of secondhand smoke, Defendants have fraudulently denied that ETS causes disease.'" (Tobacco industry response section, paragraph before Position of major tobacco companies)
  • "The ruling found that tobacco companies undertook joint efforts to undermine and discredit the scientific consensus that second-hand smoke causes disease, notably by controlling research findings via paid consultants." (under US racketeering lawsuit against tobacco companies)

All the instances of "causes" read very naturally, some being quotation of people (such as the judge of the Federal court) who are not specialized in epidemiology.

The body of the article contains not a single instance of "can cause".

Even Philip Morris, in stating their official position on the issue on their website, say unambiguously:

  • "Public health officials have concluded that secondhand smoke from cigarettes causes diseases, including lung cancer and heart disease, in non-smoking adults, as well as conditions in children such as asthma, respiratory infections, cough, wheezing, otitis media (middle ear infection) and sudden infant death syndrome."
  • "The World Health Organization External reference (WHO) provides information on its website which states that secondhand tobacco smoke is dangerous to health and that it causes cancer, heart disease, and many other serious diseases in adults.

I see therefore no valid reason to change the formulation which has been in place for years now and has withstood the test of time. It is succinct, articulate, corresponds to the worldwide consensus of public health experts and authorities. What else do we want!

--Dessources (talk) 01:50, 20 January 2015 (UTC)

I tend to agree, but it isn't open and shut for individuals, just for populations. There are other causes after all. We should be able to express the distinction somehow. LeadSongDog come howl! 17:52, 20 January 2015 (UTC)
True. However, when talking about this subject, health authorities and other pertinent sources use the epidemiological definition of causation (the verb "causes" in the lead links to such definition to make this clear). In the context of causality used by such sources, replacing "causes" with "can cause" might introduces an ambiguity for some readers, who could interpret this as meaning that the causality is not established yet (and surely the tobacco industry would be pleased to exploit such ambiguity). While for some diseases we can only say that current evidence is suggestive but not efficient to infer a causal relationship (using the US Surgeon General terminology), there are many diseases - and not just lung cancer - where evidence is sufficient to infer a causal relationship with exposure to passive smoking. Wikipedia rules are that "we publish the opinions only of reliable authors, and not the opinions of Wikipedians who have read and interpreted primary source material for themselves." This makes the issue easy here: all the major reliable sources we cite say that exposure to secondhand smoke "causes" diseases and none of them use the "can cause" terminology. Even Philip Morris uses the "causes" terminology.
Interestingly, it seems to me that the "can cause" vs. "causes" debate is rather specific to passive smoking. The sentence "active smoking causes lung cancer" is well accepted and is not subjected to the same objections, while it actually makes use of exactly the same epidemiological definition of causation, as indeed active smoking does not cause all smokers to get lung cancer. Restricting the view to a particular individual, his/her active smoking can indeed cause him/her to have lung cancer.
Finally, the "causes" terminology seems also more widely accepted. On my side of the world, a Google search for "smoking causes lung cancer" produces 119'000 hits, while "smoking can cause lung cancer" gives only 21'000 hits, almost six times less. For passive smoking, the difference is not as large, but still in the same direction: "passive smoking causes lung cancer" gives 7'490 hits, while "passive smoking can cause lung cancer" produces 3'360 hits, i.e. less than half.
Dessources (talk) 00:56, 21 January 2015 (UTC)
The issue is apparently a difference between common-language English and specialist terminology. A link to the epidemiologist's definition of "cause" seems necessary, but our article epidemiology does not address it well, and the principle of least astonishment applies. LeadSongDog come howl! 16:13, 21 January 2015 (UTC)
Good point. It seems however that the Google search settles the issue of least astonishment, since "causes" far outnumbers "can cause" on the web. In the current article, I see no reason to compromise on the principle of verifiability. And also, should one invoke the least astonishment principle to justify the "can cause" formulation, one would also have to be consistent and apply it throughout the article.
Finally, regular editors of this article know well that it is the target of attacks by people whose position on the issue coincides strangely with the fraudulent view of the tobacco industry. Their first objective is to mitigate the degree of causality between exposure to passive smoking and diseases, in spite of the fact that it is perhaps by far the most reliably established of all links between environmental factors and diseases. There are tobacco denialists, and this page needs to be protected against their attacks (of course, not all Wikipedians who modify this article in a way that moves its contents away from established evidence are denialists, but some definitely are.)
This "causes" formulation has been around for years in the current article and seems very well accepted. I even claim it is the one which creates least astonishment, as common usage of the word "cause" is closer - and even perhaps identical - to the way epidemiogists use it. A good evidence of this is provided by the following search I conducted on Google: when typing "drunk driving causes car accidents" I get 12'400 results. When searching for "drunk driving can cause car accidents", I get only 8 results, while everybody knows that driving a car while drunk does not inevitably lead to an accident. So perhaps people are more educated about epidemiology than we think.
Dessources (talk) 17:10, 21 January 2015 (UTC)
A wp:GOOGLE test isn't verifiable. We need wp:Reliable sources to cite that will demystify "cause". LeadSongDog come howl! 01:43, 23 January 2015 (UTC)
Sure, and I wouldn't refer to a Google search in the article. But the wp:GOOGLE rules does not apply to discussions on the Talk pages, where Wikipedians are free to express their views with the entire palette of arguments they deem suitable, as long as the debate remains civil and respectful, like the one we have now.
Dessources (talk) 11:14, 26 January 2015 (UTC)
And, BTW, I thought it was clear that the whole point of my comment above was to draw attention to the numerous quotations of highly reliable sources already present in the body of the article where "causes" is used instead of "can cause". Do you mean we need more reliable sources than the WHO Framework Convention, a treaty ratified by 178 countries, the report of the US Surgeon General, the monograph of the International Agency for Research on Cancer, prepared by a panel of the world's best experts, the US Environment Protection Agency, the California Environment Protection Agency, the National Cancer Institute, the US District Court for the State of Columbia, and ... even Philip Morris? If there are other sources which can challenge these, it would indeed be important to consider them.
Dessources (talk) 11:29, 26 January 2015 (UTC)
I very much disagree. I challenge the earlier discussion on this after having read it, and will be glad to have that discussion further as may be desired on this talk page. The body of the article ought to be using "can cause" if sticking with WP:NPOV. That it does not, is violation of that pillar for it is placing opinion (however much that is met with consensus) over the epidemiological understanding of causal inference. Let us quote that here so there is no mistake about what that is saying.
Although epidemiology is sometimes viewed as a collection of statistical tools used to elucidate the associations of exposures to health outcomes, a deeper understanding of this science is that of discovering causal relationships.
"Correlation does not imply causation" is a common theme for much of the epidemiological literature. For epidemiologists, the key is in the term inference. Epidemiologists use gathered data and a broad range of biomedical and psychosocial theories in an iterative way to generate or expand theory, to test hypotheses, and to make educated, informed assertions about which relationships are causal, and about exactly how they are causal.
Epidemiologists Rothman and Greenland emphasize that the "one cause – one effect" understanding is a simplistic mis-belief. Most outcomes, whether disease or death, are caused by a chain or web consisting of many component causes. Causes can be distinguished as necessary, sufficient or probabilistic conditions. If a necessary condition can be identified and controlled (e.g., antibodies to a disease agent), the harmful outcome can be avoided.''
The smoking data (via "causes") is written as if "one cause - one effect" and in doing so is misappropriating the term. As noted here, the diseases and death attributed to smoking are caused by a chain consisting of many component causes. If we linking to the epidemiological understanding of "causes" then we are misrepresenting the term by leaving it as is. Your entire argument for this is based on the opinions of others and deeming that as "authority." I grant that there is consensus in these opinions, but not that none of these opinions align with the usage of the term as stated in the epidemiology link which I have quoted above. If truly aligning with what is stated in that link, the lead would read, "Exposure to second-hand tobacco smoke has been found to have a inferred causal relationship to cases of disease, disability, and death." By going with simply "causes" this is clearly changing the way in which epidemiologists use the term. Gw40nw (talk) 23:03, 25 January 2015 (UTC)
It's not possible to change "causes" into "can cause" in the body of the article, since these occurrences of "causes" are in quotations of the most authoritative sources on the subject.
Dessources (talk) 11:36, 26 January 2015 (UTC)
Clearly you aren't reading what I'm writing, or for that matter what I am not writing. All I am saying is that we need to provide readers a link to a properly referenced explanation of what "cause" means in an epidemiological context. Doing so would largely avoid the need for the disputation which other editors seem to have found necessary. Is that so complicated? We wikilink things all the time! LeadSongDog come howl! 18:03, 26 January 2015 (UTC)
My apologies for the misunderstanding. Do you mean that the explanation provided by the link to "Causal inference" when clicking on "causes" is not clear? Upon re-reading that article, I guess you are probably right. Unfortunately, I don't know of any better explanation. It would be nice if you, as an expert editor, would find one.
Dessources (talk) 18:12, 26 January 2015 (UTC)
Just note though that in my note above I was addressing Gw40nw proposal to amend the body of the article...
Dessources (talk) 18:15, 26 January 2015 (UTC)
Apart from the occurrences of "causes" in quotations, it is possible to change "causes" to "can cause" or something closer to the accurate interpretation of epidemiological causal inference. These alleged most authoritative sources are misusing the term. In their quotes, I am glad that we have it the way they are stating it. Makes them look like they lack clear understanding on how causation works within science and reason. Yet, on Wikipedia, the "can cause" is presentation of NPOV content that aligns with understanding of terms, not popular, albeit misguided, interpretations.
I know the body of the article was previously amended to "can cause" and then I discovered yesterday that it was reverted back to "causes" which apparently is based on your recent arguments put forth on Talk Page. I strongly believe, and will continue to argue for "can cause" despite all the popular people's infallible interpretation to see it as "causes." I presented the exact text that "causes" is currently linking to. In that, it says, "the key is in the term inference." It also says, "the 'one cause – one effect' understanding is a simplistic mis-belief. Most outcomes, whether disease or death, are caused by a chain or web consisting of many component causes." Apart from the lead, but found in this article and in the many sources, it is routinely presented as: one cause (smoking/passive smoking) leads to the effect (death, disease, disability). I don't dispute that this is the popular interpretation, but do dispute that understanding of epidemiological causal inference. That it is linked in the article is good, otherwise I'd be disputing it as anti-science. That it is linked and not changed to "can cause" is leading me to dispute it as anti-epidemiological. Gw40nw (talk) 19:33, 26 January 2015 (UTC)
Dear Gw40nw: "These alleged most authoritative sources are misusing the term." Well, whether we like it or not, and in the absence of better sources, we have to stick to what these authoritative sources say - that is the rule on Wikipedia (see: WP:SOURCES). If you think these sources are misusing the term "causes", I would then invite you to share with us the authoritative sources on which you base this judgement. I for my part would be very interested in knowing them and learning about their arguments. Then we can decide whether it is legitimate to change the present term. In the mean time, I see no reason to change it.
Dessources (talk) 21:45, 26 January 2015 (UTC)
Looking closer at the question, it's not that they are "misusing" the term, so much as that the term does not have a single agreed-upon definition. Rather there are ongoing debates, as seen in PMC 3888277, PMC 3219814, and PMC 2706236. Indeed, the first of these (Morabia) makes it clear that the philosophers have yet to get their act together fifty years after the Bradford-Hill criteria were published (1965): "Besides Hume and Mill, the epidemiologic literature is clueless about a plausible, pre-1965 philosophical origin of Hill's viewpoints. Thus, Hill's viewpoints may be philosophically novel, sui generis, still waiting to be validated and justified." For all the time, money, brainpower and effort that goes into identifying the "causes" of disease, we remain unclear on the precise meaning of the word. How crazy is that? Still, Hill's usage clearly predominates in medicine. As he famously put it: “What I do not believe is that we can usefully lay down some hard-and-fast rules of evidence that must be obeyed before we accept cause and effect. None of my nine viewpoints can bring indisputable evidence for or against the cause-and- effect hypothesis and none can be required as a sine qua non.” In other words, look at all the criteria, then make the most reasonable judgement call: is there some better explanation of the evidence alternative to causality? LeadSongDog come howl! 22:04, 26 January 2015 (UTC)
Thanks a lot for the explanation and for the fascinating references. This is indeed an interesting subject. It seems to me that one way to approach the question would be via experimental philosophy as some of our cognitive biases (no judgment implied) are probably at play in our understanding of the concept of causality and different groups of people may understand it differently. In the mean time, the safest is probably to stick to the language used in authoritative sources.
Dessources (talk) 01:28, 29 January 2015 (UTC)

Use of Tobacco and Passive Smoke Equates to Lower Risk of Dementia[edit]

There have been strong suggestions that dementia is a deficiency of acetylcholine in studies. Since literally all nicotine does is produce acetylcholine, does that not suggest a therapeutic use? — Preceding unsigned comment added by (talk) 05:34, 4 March 2015 (UTC)