4

I have been discussing smoking with a fellow who states that for all the decades of study of the tobacco leaf, still nobody has managed to find what part of the process of smoking the tobacco leaf causes lung cancer. He pointed out that so far most of the evidence is from epidemiological studies, which can only show correlations, but not causation, and he pointed out flaws in some of these. For example, he said in one of them - the original epidemiological study, I believe - that showed that 90% of people with lung cancer are smokers, it was the case that 90% of the population were smokers, so the study could with as much certainty have said that 90% of the people with broken legs were smokers therefore smoking causes broken legs.

He pointed out that already many cancers that were previously attributed to smoking, such as throat and cervical cancer, have actually been found to be caused by a virus (HPV). So given there has been no causal link discovered, it may very well be that something other than tobacco, such as a virus, causes lung cancer, and hence humanity is missing a huge opportunity to find the actual cause for lung cancer and thus avert untold suffering.


I found one New York Times that states:

The findings, published today in the journal Science, report the first evidence from the cell biology level linking smoking to lung cancer. The scientists say a chemical found in cigarette smoke has been found to cause genetic damage in lung cells that is identical to the damage observed in many malignant tumors of the lung.

The findings reported today establish the long-missing link, in the opinion of experts in the field of cancer genetics, and may also play a role in pending litigation about smoking illnesses and passive smoking.

If that part is true, this means that at the very least, no causal link had been discovered before the article was published, in October 1996.

The article states that:

[...] the scientific team exposed human lung cells to the chemical BPDE, a powerful carcinogen that is produced in the body from the breakdown of benzo$(a$)pyrene, a chemical in tobacco smoke.

Two researchers [...] examined the p53 gene and found that molecules of BPDE had bound chemically to the p53 DNA at precisely the three points that had already been identified as mutational hot spots in lung cancer.

If correct this means that benzo(a)pyrene is the chemical found in tobacco which, when smoked, causes lung cancer. Link to the paper itself: Preferential formation of benzo[a]pyrene adducts at lung cancer mutational hotspots in P53..


However the fellow said that they thought they had discovered the causal link but later had to retract it. The NY Times article is written that it is pretty cut and dried, but I have found indicators that it isn't. This rebuttal by Lauren A. Colby, published 9 days after the NY Times article, points out several flaws in the study (severely edited for length):

The authors of the article start out by pointing out that, in about 60% of lung cancer cases, there is mutational damage to the p53 gene [...] This is another way of saying, of course, that in 40% of lung cancer cases, there is no damage to the gene, meaning that people can get lung cancer even if their p53 genes are in perfect condition.

The authors did not study any actual human lung cancers. [...]

The authors of the study apparently concede that [Benzo(a)pyrene (BAP)] [...] is not terribly carcinogenic [...]; it must be converted to BAPDE. There is no evidence that the lungs, themselves, can metabolize BAP into BAPDE. Even if they could, the amount of BAP reaching the lungs from cigarette smoke is dwarfed by the amount reaching the lungs in the blood supply [...] from consumption of burnt food. [...] The study could just as well be said to prove an indirect link between the consumption of burnt food and lung cancer.

Further evidence: [This FAQ] about benzo(a)pyrene, last updated in 2009, states that "Benzo[a]pyrene is a probable cancer-causing agent in humans." Why 'probable' if it has been proven, as the 1996 study claims?


So, is he right? Is there no as-yet proven causal link between smoking and lung cancer?

DaleJohnston
  • 751
  • 1
  • 5
  • 12
  • 1
    Is this not a duplicate of [Does Smoking Cigarettes Cause Lung Cancer?](http://skeptics.stackexchange.com/questions/7883/does-smoking-cigarettes-cause-lung-cancer) – Oddthinking Feb 03 '15 at 23:47
  • 3
    Logical fallacy here: "cancers that were previously attributed to smoking, such as throat and cervical cancer, have actually been found to be caused by a virus". Just because a virus can cause those cancers doesn't mean that smoking can't also cause them. – DJClayworth Feb 04 '15 at 00:16
  • 1
    "that showed that 90% of people with lung cancer are smokers, it was the case that 90% of the population were smokers" Don't suppose you can source it? – NPSF3000 Feb 04 '15 at 00:43
  • @Oddthinking I don't think so. That question asks *does* it cause cancer; this question asks *why* it causes cancer (AFAICT). – Shokhet Feb 04 '15 at 02:42
  • Either this question asks *if* smoking causes cancer, and it's a dupe, or it is about *how* or *why*, and it's off topic. In either case it should be closed, I've opted for the second due to your comments. – Sklivvz Feb 04 '15 at 03:51
  • @Sklivvz: Hmm I didn't realize it may be off-topic. What if I rephrased it as: "Is it true that there is no causal link between smoking and lung cancer?" Then I can cite where I find the claim. – DaleJohnston Feb 04 '15 at 04:13
  • 2
    @Claudiu that would make it a duplicate. – Sklivvz Feb 04 '15 at 06:08
  • @Sklivvz I think the claim is that they haven't found any [causal agent](https://www.google.fr/search?q=causal+agent); and I tried to answer that they actually found several. – ChrisW Feb 04 '15 at 09:22
  • @ChrisW The question could be simply asking whether benzo(a)pyrene is such an agent, but it needs major fixing in order to do so, and furthermore it would make your answer obsolete. I would also have doubts about the notability of such a claim. – Sklivvz Feb 04 '15 at 10:20
  • @Sklivvz Well I think the claim is that there are no known causal agents: "So given there has been no causal link discovered, it may very well be that something other ..." and "they thought they had discovered the causal link but later had to retract it" and "Benzo[a]pyrene is a probable cancer-causing agent in humans" and finally, "Is there no as-yet proven causal link between smoking and lung cancer?" The sites I quoted seem to me to be saying that there's not a single causal link, and that rather there's about 70 of them. – ChrisW Feb 04 '15 at 10:41
  • 1
    @ChrisW In any case the question is still not answerable because it's not clear, it *specifically* asks for the wrong thing in the title, and needs major clarification work before it can be reopened. – Sklivvz Feb 04 '15 at 10:45
  • @Sklivvz: Ok I think I understand. It sounds like I should participate in the [other question](http://skeptics.stackexchange.com/questions/7883/does-smoking-cigarettes-cause-lung-cancer). What if I dispute the claims in that answer - would I post a separate answer? Post a comment? Or is skeptics.SE not the place to do this? – DaleJohnston Feb 04 '15 at 22:35
  • @Claudiu no. If you have an answer to the other *question* then post it, but do not answer another answer. The idea of this network is that voting takes care of disputes. – Sklivvz Feb 05 '15 at 09:33

1 Answers1

4

The modern, public information is that tobacco smoke contains many chemicals which cause cancer:

American Lung Association

There are approximately 600 ingredients in cigarettes. When burned, they create more than 7,000 chemicals. At least 69 of these chemicals are known to cause cancer, and many are poisonous.

Cancer Research UK

Some cancer-causing chemicals in tobacco smoke and their other common uses

  • Tar - a mixture of dangerous chemicals
  • Arsenic - used in wood preservatives
  • Benzene - an industrial solvent, refined from crude oil
  • Cadmium - used in batteries
  • Formaldehyde - used in mortuaries and paint manufacturing
  • Polonium-210 - a highly radioactive element
  • Chromium - used to manufacture dye, paints and alloys
  • 1,3-Butadiene - used in rubber manufacturing
  • Polycyclic aromatic hydrocarbons - a group of dangerous DNA-damaging chemicals
  • Tobacco-specific nitrosamines - a group of carcinogens only found in tobacco
  • Acrolein - used as a chemical weapon in World War 1 - and acetaldehyde - used in the industrial production of acid

American Cancer Society

Tobacco smoke is made up of more than 7,000 chemicals, including over 70 known to cause cancer (carcinogens).

ChrisW
  • 26,552
  • 5
  • 108
  • 141
  • 5
    None of this is very convincing. I do believe that smoking causes cancer but I don't think this answer shows it. Containing chemicals that are known to cause cancer could conceivably be OK if the chemicals were in small enough quantities in cigarettes. Being used in e.g. batteries or rubber is not evidence that a chemical is dangerous. – bdsl Feb 04 '15 at 00:09
  • @bdsl Go to the cancer Research UK page and read the details of why each substance is cancerous. – DJClayworth Feb 04 '15 at 00:18
  • If a person is diagnosed with cancer when they're [age 70](http://skeptics.stackexchange.com/a/24007/2703) after smoking a 20-pack per day since they were aged 20, that's ((70-20) * 20 * 365 =) more than three hundred thousand cigarettes. – ChrisW Feb 04 '15 at 00:18
  • I'm not looking evidence that smoking causes cancer - I already believe that. I was commenting on this answer. Perhaps this answer could be improved by including some of those details from the Cancer Research UK page. – bdsl Feb 04 '15 at 00:21
  • @bdsl This question wasn't about that: check out [this question](http://skeptics.stackexchange.com/questions/7883/does-smoking-cigarettes-cause-lung-cancer). This question was asking about *which parts* of cigarettes are carcinogenic, which I think was answered more than adequately here. – Shokhet Feb 04 '15 at 02:40
  • 1
    @Shokhet: I think this question doesn't ask about which parts of cigarettes are carcinogenic. That could be another question. That question, together with the question of weather smoking causes cancer leads to this question - are the carcinogenic parts of cigarettes enough to actually cause cancer? Having actually written software to predict cancer risk I can say that the answer to that has a lot to do with probabilities and what it means to "increase risk". – slebetman Feb 04 '15 at 03:43
  • 1
    So what I would ask here is: What is the proof that the tar in tobacco, when smoked, causes lung cancer? What is the proof that the arsenic in tobacco, when smoked, causes lung cancer? Etc. – DaleJohnston Feb 04 '15 at 04:29