Passive Smoking - Real?

[GreyPilgrim]

Now that the ban on smoking in public / enclosed places has passed through parliament, I thought I’d have a look at the FOREST site (forestonline dot org) and see what their reaction would be. I find it quite a sad site at the best of times, and I fully expected to see a lot of nonsense, emotion-driven, illogical rants. I wasn’t really disappointed.

Don’t get me wrong – I’m an ex-smoker, but I try really hard no to be a holier-than-thou evangical ex-smoker (although for a time I admit that I may have been…). I think everyone should have the right to make an educated decision to smoke, but that that right should not damage the health of other people. A partial ban may have offered a little bit of choice to patrons on where they go to eat or drink, but would offer very little protection to staff who don’t always have the luxury of choosing where to work.

I can honestly say that even when I smoked, I still felt very uncomfotable lighting up in a restaurant where smoking was allowed…I suppose I gave a damn about the other diners, so more often than not I’d dart outside. I really don’t see why this is so much of a problem.

Anyway, sorry to waffle. The point of the post. There was a link from the forest site to an article written by a chap called Dave Hitt, called Name Three (sorry, I don’t seem to be able to post links yet, but if you google Dave Hitt Name Threee you’ll get it), where he quickly manages to dispell the myth of the effects of passive smoking and environmental tobbace smoke with the line “…you claim that 63,000 people die from second hand smoke every year. Could you please name three or four or them”

It seems like a reasonable question. But then another part of my brain is telling me it’s a stupid statement. He actually opens the article by telling us his dad used to get him to eat his greens by saying “there are millions of kids starving in Africa” (to which the young Mr Hitt would reply “Name Three”)…by his own logic then, if he is saying that the failure of health professional to “name three” people who have died from passive smoking proves passive smoking is a myth, then his fathers failure to name three kids who have starved to death in Africa proves that starvation / famine is also a complete myth. Opinions?

 

[kieran]

Asking your father to name starving children in Africa he has never met is likely to lead to no response. Asking a voluntary overseas worker or missionary in a famine area (i.e. someone with more of a direct link and who should be more knowledgeable) is more likely to lead to the question being answered. As noted, the lack of an answer is not an indication of the lack of the phenomenon - but if you couldn't get any knowledgeable person to answer you then you would have cause to question whether there really was a phenomenon.

I think that this Dave Hitt guy is quite within his rights to ask such questions. You and I are not likely to know an answer but "experts" in the field should be in a much better position to provide one. Reading this web site has made me wonder why all these authorities haven't just pooled data to produce such answers - even if only to shut up this debate.

I'm not arguing that passive smoking does not exist and is not a killer.

Are there legal reasons that make it difficult to pronounce passive smoking as the cause of death? If so, then the experts could have provided an overview of these reasons.

 

[richardm]

Are there legal reasons that make it difficult to pronounce passive smoking as the cause of death? If so, then the experts could have provided an overview of these reasons.

Yes - the cause of death is supposed to be what actually killed you. So in the case of passive smoking it would typically be something related to lung or heart failure; what actually caused the failure is (from the point of view of the death certificate) not usually relevant. Or so I have been told by a genuine doctor, who (I hope) presumably knew what she was talking about.

 

[steenkh]

I cannot help you out by dispelling or confirming the myth. I could google for the studies on passive smoking but I am not qualified to judge whether those studies are worth anything or not. But those studies I have heard about actually recorded passive smokers who had a shorter life span than pure non-smokers, whatever that is worth.

I myself is a non-smoker who has never smoked at all. In Denmark there is no general ban on smoking at public or any other places, but such bans are becoming more common. I have nothing against people enjoying whatever vices they like, but it is areal problem with smoking that I have to be involved in it also.

Just a few days ago I was at a restaurant with a group of people from my work, about half of which were smokers. I could not enjoy my food, and the athomsphere was so heavy that I felt some discomfort at breathing. I had to go out to take a breath of clean air at times. When I came home I had to wash my hair immediately to get the stench out, the clothes had to be put in a closed plastic bag in order for my wife and I to sleep, and even the next morning I could still feel that there was smoke in my lungs!

I would actually be very surprised if it turns out that passive smoking is not dangerous!

 

[Rolfe]

I find this subject very interesting. I've heard that the epidemiological evidence of harm from passive smoking is very very much weaker than the evidence for harm from active smoking. Nevertheless, it does seem to stand to reason that if you breathe in enough second-hand smoke, and that smoke contains substances known to be dangerous to the smoker, then there is a chance you will be affected. I wonder if it is, as you'd expect, a dose-dependent thing, and that clear evidence may only be found in a group of very heavily exposed people.

However, although harm from passive smoking makes it easier to present the case for a smoking ban, should that be a necessary component? I find other people's cigarette smoke grossly unpleasant. I try to avoid it as much as possible, but sometimes it just can't be avoided without seriously curtailing my social life. I've seen smoking gradually retreat over the past 30 years though, so that one now expects to be able to travel by public transport, visit the cinema, and eat at a restaurant without having to breathe smoke. These things would have been impossible dreams in the 1970s.

It seems to me to be a matter of social acceptability. There used to be signs on all the buses saying "no spitting". Now there aren't, but people seldom spit. Have the signs disappeared because tuberculosis is now rare and so the chances of being harmed by someone spitting are slim, or simply because they're not needed?

I've heard in defence of the smoker, "oh, he enjoys it, it's his only pleasure, don't be a killjoy." But supposing that were extended to spitting? "Oh, spitting in your soup is his only pleasure, he enjoys it, why are you making such a fuss?" But I could then decline to eat my soup, if I knew it had been spat in. I can't just decide not to breathe because the air is smoky.

I suspect that smoking will decline, and smoking bans become acceptable, as smoking gradually approaches spitting in some else's soup in the ranks of socially acceptable behaviour. It's not there yet, but it's moving in that direction. When it gets there, you won't need to present statistics on the incidence of tuberculosis and the number of people who have contracted it because someone spat in their soup to justify the bans. Maybe, in another 50 years, you won't need the no smoking signs any more, because nobody would dream of smoking in public any more than they'd dream of spitting in someone else's soup.

Rolfe.

 

[brodski]

The problem is that "passive smoking" is thought to increase the likelihood of suffering from certain medical problems, it is very difficult to give a single case, because individual cases could have a number of causes, of which only one may, or may not, have been exposure to "2nd hand smoke" . For a good discussion on this issue, look at any of Yrregs acupuncture threads, only replace "acupuncture" with "exposure to smoke" and "got better" with "died".(remember I only recomeded that you read the trheads, the decison to partcipate or not shoud be entirely your own )

 

[Rolfe]

I remember as a child hearing about the death of a man who had been a caretaker at a Miners' Welfare club. He died of a chronic lung disease, I don't remember what. A post mortem examination was carried out. Afterwards, the consultant told his widow that it was his smoking that had been the main contributory factor to his death. The widow protested that he'd never smoked a cigarette in his life. However, he'd worked long hours in an atmosphere full of other people's cigarette smoke. This had apparently resulted in his lungs at post mortem being indistinguishable from the lungs of a heavy smoker.

Now that's an anecdote, and a very second-hand one at that, remembered from many years ago. However, if a case like that were properly documented, then that might be sufficient to prove to a reasonable standard of proof that that particular individual had died of passive smoking.

Rolfe.

 

[Crispy Duck]

The author of www.numberwatch.co.uk has plenty to say on passive smoking - his site is well worth a read generally.

I believe the epidemiological evidence for this sort of assertion generally comes from analysing the life-spans of large numbers people, dividing them into smokers, passive smokers, non-smokers, etc, controlling for other factors (?), and then seeing if passive smokers die earlier than non-smokers. From that, one obtains a ratio like "passive smokers' life-spans are X% of non-smokers' life-spans". From that, and the number of people that die annually, one arrives at the conclusion that "Y,000 people die of passive smoking per year".

To cut a long story short, the number is made up.

I have a genuine problem with the "controlling for other factors" step. Every epidemiological study I read about includes this step. This is absurd, logically speaking. It requires that, for any potential cause of death, one already has a complete understanding of all other causes of death, in order to control for them. Can anyone enlighten me as to how this is done?

 

[GreyPilgrim]

I think my whole problem with the FOREST site was one of their continual denial, really. All the cherry picked quotes and data, all of the nonsense arguments, all of the nostalgia and emotion they tried to employ to rouse the support of other smokers. FOREST was originally founded by an ex soldier who was asked to put out his pipe by a lady on a railway platform and Forest will continually remind it’s readers that this chap fought bravely for King, country and freedom, implying I guess that if we don’t support his organisation then he fought in the wars for nothing. They’ve also rolled out a list of celebrities who have contributed a signed letter of condemnation of the UK smoking ban, yet don’t give the same airtime to the British Medical Associations views on the subject. I actually like a couple of the celebs mentioned, but on health matters I would tend to believe the BMA more.

And here’s me saying I wasn’t an evangelical ex-smoker….sorry if it sounds that way, I think I’m just frustrated that forest would deny any danger from passive smoking regardless of how much evidence there was. I’m sure that decades ago, when the link between lung cancer and primary smoke was still unclear and not as concrete as it is now, forest would have been producing articles saying exactly the same thing, and would only even begin to acknowledge there was a problem when there was no possible way they could disagree with the facts.

 

[geni]

I have a genuine problem with the "controlling for other factors" step. Every epidemiological study I read about includes this step. This is absurd, logically speaking. It requires that, for any potential cause of death, one already has a complete understanding of all other causes of death, in order to control for them. Can anyone enlighten me as to how this is done?

Not really. You just have to make sure that the errors introduced by no totaly understanding all other causes of death are smaller than your measured effect.

 

[Crispy Duck]

Not really. You just have to make sure that the errors introduced by no totaly understanding all other causes of death are smaller than your measured effect.

Thanks, Geni. Does your "not really" mean "it isn't really logically absurd", or "you can't really explain how it's done"? Either way, I'm sure you can guess my next question....

How does one estimate the size of errors introduced by unknown factors?

I suspect that this is the exact reason that only significant effects (relative risk > 3.0) can be reliably detected.

 

[Beth]

To cut a long story short, the number is made up.

I have a genuine problem with the "controlling for other factors" step. Every epidemiological study I read about includes this step. This is absurd, logically speaking. It requires that, for any potential cause of death, one already has a complete understanding of all other causes of death, in order to control for them. Can anyone enlighten me as to how this is done?

The number is not 'made up', but computed. To truly understand what's being done, you'll need to take graduate level courses in statistics. It's not absurd at all, but the best way we have to understand whats happening.

On a simplistic level, to determine whether a particular factor (like passive smoking) has a statistically significant effect even though it's effect is likely to be very small in comparison to the effect of other factors (say diet and exercise choices) on the variable of interest (lifespan), first the effect of the other factors is computed and used to compute the expected lifespan for an individual. Prediction intervals are computed as well, so the amount of variation or mismatch between the expected and actual lifespans is known. Then the difference between the actual and the expected lifespan is used as the dependent variable of interest rather than the actual lifespan to determine if passive smoking has a statistically significant effect on those differences. The expected amount of uncertainty in the expected and actual lifespans is also taken into account, but the statistics get pretty involved in order to do all of that.

 

[ChristineR]

Any non-smoker who has spent time in a smoke filled room can tell you it is harmful to your health.

As far as whether there are long term serious health consequences, the best data I'm aware of comes from children of smokers, who do indeed get seriously sick.

http://www.ash.org.uk/html/factsheets/html/fact08.html

As far as people who work in smokey places, it's very hard to quanitize because if you're really having trouble, you'll just quit the job.

 

[Crispy Duck]

The number is not 'made up', but computed. ...<snip>... first the effect of the other factors is computed and used to compute the expected lifespan for an individual.

Thanks for the info, Beth. I have a basic knowledge of statistics and statistical methods - I'm a programmer in the pharmaceutical industry. Of course, I wasn't claiming that the figure was a complete fabrication, but rather that it is the result of some heavy weight stats rather than a simple count of people who've died due to passive smoking.

I'm prepared to accept that I just don't fully appreciate the power of stats, but still... It seems to me that the two steps I've highlighted above are somewhat circular, or at least that there must be some sense of diminishing returns. Given the starting point that the effect of passive smoking is unknown, it follows that this factor cannot have been controlled for when calculating the effects of more important factors like diet and exercise. Similarly, the effect of [diet / exercise] can't have been controlled for when calculating the effect of [exercise / diet] (delete as applicable). And that's only speaking of the factors that have been deliberately controlled for - how many others might be relevant, but completely ignored?

It all seems a bit like voodoo to me. And all those concerns are just with proving the correlation, never mind proving causality...

 

[Crispy Duck]

Any non-smoker who has spent time in a smoke filled room can tell you it is harmful to your health.

Hmmm. Well, I'm a non-smoker, and until changing jobs recently I used to spend two or three nights a week in a smoky pub, playing pool (all my team-mates were smokers). I specifically cannot tell you that it was harmful to my health. It did make my clothes smell, I'll give you that. I'm not aware of any effects on my health, whether short- or long-term, trivial or serious. I trained for and ran a marathon during that period.

My anecdote doesn't necessarily trump your sweeping generalisation, but certainly your 'any' should be 'less than 100% of'...

I'm not pro-smoking, and I wouldn't necessarily seek out a smoky pub over a non-smoky one, but I'm very suspicious of the use of equivocal statistics to force a political agenda. The UK smoking ban is a result of the government not wanting people to smoke, pure-and-simple. The claims about protecting people working in smoky environments are completely spurious - are they going to legislate against all hazardous workplaces?

Although not a smoker, I do enjoy another hobby that is potentially dangerous to my health, and that some others find objectionable - I fly light aircraft. Should I be allowed to continue?

 

[ChristineR]

Hmmm. Well, I'm a non-smoker, and until changing jobs recently I used to spend two or three nights a week in a smoky pub, playing pool (all my team-mates were smokers). I specifically cannot tell you that it was harmful to my health. It did make my clothes smell, I'll give you that. I'm not aware of any effects on my health, whether short- or long-term, trivial or serious. I trained for and ran a marathon during that period.

My anecdote doesn't necessarily trump your sweeping generalisation, but certainly your 'any' should be 'less than 100% of'...

I'm not pro-smoking, and I wouldn't necessarily seek out a smoky pub over a non-smoky one, but I'm very suspicious of the use of equivocal statistics to force a political agenda. The UK smoking ban is a result of the government not wanting people to smoke, pure-and-simple. The claims about protecting people working in smoky environments are completely spurious - are they going to legislate against all hazardous workplaces?

Although not a smoker, I do enjoy another hobby that is potentially dangerous to my health, and that some others find objectionable - I fly light aircraft. Should I be allowed to continue?

I have no objection to you flying aircraft. Why should I?

If you were spending two or three nights a week in a smokey pub, either you are naturally less sensitive to smoke or you became less sensistive because you were spending so much time around smoke. I suspect the latter. Me and most people of my personal acquaintance could not spend several hours in a smokey pub without becoming ill. I don't mean seriously ill, just things like watery eyes and sore throat--although I do know some asthmatics who have to use an inhaler to prevent becoming seriously ill.

I don't know to what extent you were harming your health in these pubs, although I don't think it was huge.

I live in the U.S., and here this argument is usually framed around the idea that people like me should be able to go to a pub and enjoy themselves. Only rarely do you hear the argument that the government should be forcing people not to smoke. I also think that there's a lot less smoke in the air here than in the U.K. My information on this may be outdated, though.

 

[NeilC]

You are hardly likely to notice your cancer risk going up are you?

I knew a window fitter who got lung cancer - never smoked but was a pub singer in the evening. He put it down to smokey pubs.

If you fly over my garden and annoy me and spray exhaust fumes over me and my washing then you should not be allowed to continue. People say we don't HAVE to go to the pubs but I believe practical situations dictates most of us do even if we don't want to.

Flying is not a dangerous addictive drug. You truely have a choice to do it or not. Smokers don't once addicted. So their "freedom" to smoke is illusory. It is hardly a right worth defending.

Flying is a pastime/hobby. Smoking is an addiction. Hobbies are a good thing to do in general. Feeding drug addictions is not generally a good thing.

So I don't think flying and smoking are parallel.

 

[ClusterBoy]

How does one estimate the size of errors introduced by unknown factors?

I suspect that this is the exact reason that only significant effects (relative risk > 3.0) can be reliably detected.

You can't estimate the size. Almost all good research papers will have a discussion section where they talk about other potential (unmeasured) factors. As for unknowns, you have to think about biological plausability. If it is biological plausible, chances are someone somewhere has done some research on it, and will tell you what the effect is.

What do you mean by relative risk > 3 being reliably detected? Repeatedly? Accurately? Statistically significantly?
The numberwatch website is a man with an axe to grind who thinks he understands epidemiology and clearly doesn't.

 

[Beth]

Thanks for the info, Beth. I have a basic knowledge of statistics and statistical methods - I'm a programmer in the pharmaceutical industry. Of course, I wasn't claiming that the figure was a complete fabrication, but rather that it is the result of some heavy weight stats rather than a simple count of people who've died due to passive smoking.

I'm prepared to accept that I just don't fully appreciate the power of stats, but still... It seems to me that the two steps I've highlighted above are somewhat circular, or at least that there must be some sense of diminishing returns. Given the starting point that the effect of passive smoking is unknown, it follows that this factor cannot have been controlled for when calculating the effects of more important factors like diet and exercise. Similarly, the effect of [diet / exercise] can't have been controlled for when calculating the effect of [exercise / diet] (delete as applicable). And that's only speaking of the factors that have been deliberately controlled for - how many others might be relevant, but completely ignored?

It all seems a bit like voodoo to me. And all those concerns are just with proving the correlation, never mind proving causality...

Well, I can understand that it seems a bit like voodoo. It seems that way to me too sometimes and I've studied statistics pretty extensively - if everything goes well, I should finish my Ph.D. in stats in 2007. And yes, it's only correlation, not causation because without a designed experiment with people randomly assigned to groups, analyzing the stats can only show correlation.

However, you might want to keep in mind that the whole smoking-cancer link was only correlation for the first several decades of research. Causation has been firmly established now, but it took decades of research to do so.

At any rate, the idea is that if the effects of certain factors (such as diet or exercise) are known, their effect can be removed when testing for correlations of lessor factors such as passive smoking. Then, if the results are consistent and repeatable, you can have a high level of confidence that the factor being tested really does have a significant effect. As far as other factors being relevant but ignored, generally all known relevant factors will try to be accounted for. Known factors are not ignored, but there can be unknown factors that are unaccounted for. At this point, however, I find it hard to believe that there are major unknown factors regarding the health problems associated with smoking that we have yet to discover.

 

[GreyPilgrim]

As far as people who work in smokey places, it's very hard to quanitize because if you're really having trouble, you'll just quit the job.

But thats not always possible surely? Not everyone has the abilities / experience to be able to drop a job and walk into another one straight away especially if the catering or bar sector is all they are really qualified for (and please don't think that was meant to be derogatory to bar workers...I would never offend the hundreds of people helped me get intoxicated in the past).

I've been lucky enough never to have to claim unemployment benefit so far but I'm wondering whether if I went to my local unemployment office, told them I had quit my bar job because of the risks of passive smoking, they would be sympathetic, or whether they would tell me I had made myself voluntarily unemployed and refuse to give me dole money?