Global burden of disease from secondhand smoke

[Viking Prince]

key findings

More than 600,000 death per year worldwide are caused by second-hand smoke (SHS) - this is more than 1% of all deaths. 165,000 of these deaths are among children.








Smokers are not only putting themselves at risk, but also 1.8 billion non-smokers. In 2004, 40% of children, 33% of male non-smokers and 35% of female non-smokers were exposed to SHS worldwide. This exposure led to:

• 379,000 deaths from ischaemic heart disease
• 165,000 deaths from lower respiratory infections
• 36,900 deaths from asthma
• 21,400 deaths from lung cancer

Women and children are disproportionally affected by exposure to SHS.

Of the 603,000 deaths, 47% occurred in women, 28% in children, and 26% in men. Women suffer more from the impacts of SHS as they are 50% more likely to be non-smokers than men. In terms of years of life lost, children are by far the most affected from SHS, as most of their SHS deaths occur from respiratory infections during the first few years of life.

Regional variations of exposure and disease burden from SHS are wide.

The highest exposures to SHS are found in Eastern Europe, the Western Pacific, and South-East Asia, with more than 50% of some population groups exposed. About 60% of child deaths occurred in Africa and South-East Asia combined. This is a result of respiratory infections being more common in children living with adults who smoke. Eastern Europe, South-East Asia and the Eastern Mediterranean region are the most affected from exposure to SHS in terms of total deaths per capita.

Policy makers should take action to protect the population against exposure to SHS. Effective policies are available. Only 7.4% of the world lives in jurisdictions with comprehensive smoke-free laws at present.








The authors' three key recommendations are:

• Immediate enforcement of WHO's Framework Convention on Tobacco Control to create complete smoke-free environments in all indoor workplaces, public places and public transport.
• The inclusion of complementary educational strategies, like voluntary smoke-free home policies, for countries which already have smoke-free laws.
• The need to dispel the myth that developing countries can wait to deal with tobacco-related diseases until they have dealt with infectious diseases. Together, tobacco smoke and infections lead to substantial, avoidable mortality and loss of active life years.

Original article

Lancet article [pdf 525kb]

WHO source: New study: global burden of disease from second-hand smoke



This is published in Lancet to give it the aura of a scientific journal. Do not be confused though. This is funded and published by WHO. This is driven by politics. Just another example of why the UN should not closely supervise academic research as they did with global warming.

If in doubt, there is the little icon attached to the article:

http://www.who.int/media/buffet/121206_3.jpg


From a BBC article (buried at the very end):

Writing in the Lancet, Dr Heather Wipfli of the University of Southern California and colleagues, said: "There are well acknowledged uncertainties in estimates of disease burden.

"However, there can be no question that the 1.2bn smokers in the world are exposing billions of non-smokers to second-hand smoke, a disease-causing indoor air pollutant."

So the more direct answer is: second hand smoke is considered bad by the researchers, but we have no clue to the actual number of people killed by second hand smoke.

[Phier]

Epidemiological studies on low concentrations are often complete crap.

Recall the early Bush years when the democrats wanted to claim his policies would kill people because they were changing the arsenic standards for the water, using a straight line to predict deaths. Biology works on a curve though. What they were saying was that if X kills 50%, Y would kill 1%, even if Y was very small. You can see the error in this logic thinking of alcohol. Alcohol has a lethal dose, and if you did a straight line graph one drink would therefore kill people, but oddly thats not how it works.

So I don't trust this in the least without a full methods section. My guess is these are only vague estimates based on too many variables and not a lot of data.

Mind you I think smoking is perhaps the stupidest activity one can indulge in.

[Adar]

Epidemiological studies on low concentrations are often complete crap.

I took a look at the article and I don't have time to check the meta-studies but I don't think they rely on extrapolating high doses. I think the disease data are taken from studies where you compare smoking and none-smoking households to calculate the increased risk of lower respitory diseases etc.

Mind you I think smoking is perhaps the stupidest activity one can indulge in.

Posting claims like the OP is however the second stupidest activity one can indulge in.

VikingPrince is claiming that this article must be bad science becuase WHO has an anti-smoking agenda. Therefore any new reports about the dangers of smoking supported by WHO must be biased. The problem is that this article doesn't adress whatever smoking is dangerous or not. They use data from other studies (that prove smoking is dangerous) and then try to estimate the global health impact caused by smoking.

And seriously, finding it is suspicious that references are "hidden" at the end of an article. Where do you usually put the references?

[Phier]

I took a look at the article and I don't have time to check the meta-studies but I don't think they rely on extrapolating high doses. I think the disease data are taken from studies where you compare smoking and none-smoking households to calculate the increased risk of lower respitory diseases etc.

Meta analysis in my opinion should only be used to determine if a bigger study should be conducted. Meta analysis tends to be very dicey, normally I saw it as something for a grad student to lazy to do a survey study.

Posting claims like the OP is however the second stupidest activity one can indulge in.

VikingPrince is claiming that this article must be bad science becuase WHO has an anti-smoking agenda. Therefore any new reports about the dangers of smoking supported by WHO must be biased. The problem is that this article doesn't adress whatever smoking is dangerous or not. They use data from other studies (that prove smoking is dangerous) and then try to estimate the global health impact caused by smoking.

And seriously, finding it is suspicious that references are "hidden" at the end of an article. Where do you usually put the references?

Well hes right with this line...

Writing in the Lancet, Dr Heather Wipfli of the University of Southern California and colleagues, said: "There are well acknowledged uncertainties in estimates of disease burden.

"However, there can be no question that the 1.2bn smokers in the world are exposing billions of non-smokers to second-hand smoke, a disease-causing indoor air pollutant."

Reading between the lines I'm guessing Dr. Wipfli is saying the same thing I am, the data is suspect, but we think its bad to be exposed to second hand smoke.

I personally am VERY suspicious about the 165k child deaths as a result of second hand smoke.

[Adar]

Meta analysis in my opinion should only be used to determine if a bigger study should be conducted. Meta analysis tends to be very dicey, normally I saw it as something for a grad student to lazy to do a survey study.

The problem with meta-analysis of data is that it often is reliant on studies that are shaky. In this case they have created the best guesstimate available. I am not here to debate the merits of meta-analysis. Merely the fact that VPs initial post attack the study for something that isn't covered by the study he is attacking.

Well hes right with this line...

Reading between the lines I'm guessing Dr. Wipfli is saying the same thing I am, the data is suspect, but we think its bad to be exposed to second hand smoke.

I personally am VERY speciousness about the 165k child deaths as a result of second hand smoke.

I agree with your statement.

[Plant]

Alcohol has a lethal dose, and if you did a straight line graph one drink would therefore kill people

Could you tell me how you acheived this? How did you extrapolate this?

This is published in Lancet to give it the aura of a scientific journal.

The Lancet is a peer reviewed medical journal. How can you say it is not a scientific journal?

This is funded and published by WHO.

Funding

Swedish National Board of Health and Welfare and Bloomberg Philanthropies.

Also it is published by Lancet is it not?

BTW, Vikingprince, your links don't work apart from the logo one and the BBC one.
http://www.thelancet.com/journals/lancet/article/PIIS0140-6736(10)61388-8/fulltext

[Phier]

Could you tell me how you acheived this? How did you extrapolate this?

Lets say the black is the LD50 and the red is the LD80, the straight line would be the LD10 at the green spot and that just doesn't work that way.

[Viking Prince]

Could you tell me how you acheived this? How did you extrapolate this?



The Lancet is a peer reviewed medical journal. How can you say it is not a scientific journal?




Also it is published by Lancet is it not?

BTW, Vikingprince, your links don't work apart from the logo one and the BBC one.
http://www.thelancet.com/journals/la...388-8/fulltext

As to the funding soure you were not playing fair -- at least not if you were quoting from:

Acknowledgments
We thank the Swedish National Board of Health and Welfare and the
Bloomberg Philanthropies for their funding support. The authors alone
are responsible for the views expressed in this publication, which do not
necessarily reflect the decisions or the stated policy of WHO or of its

Member States. ©World Health Organization, 2010.

It is funded and published under the authority of WHO. Whether others also contributed funding as this footnote suggests does not change that this is a WHO funded publication. The disclaimer is not there without a reason. I would suggest that the disclaimor is also not sufficient to put me at ease. WHO did select the authors for the study writeup. I trust that WHO did not make a random selection of researchers, but chose carefully based upon previous work and probable bias towards a conclusion. As I stated, this is how the global warming mess started.

Btw, I never did state nor meant to suggest that Lancet is not a peer reviewed journal.

The links were my mistake though.
Lancet Article: http://www.who.int/quantifying_ehimp...ns/smoking.pdf
WHO link intro to Lancet Article: http://www.who.int/quantifying_ehimp.../en/index.html


I give you one of three on this observation. Good enough for major league baseball hitting, but not good enough for an academic journal.

[Adar]

It is funded and published under the authority of WHO. Whether others also contributed funding as this footnote suggests does not change that this is a WHO funded publication. The disclaimer is not there without a reason. I would suggest that the disclaimor is also not sufficient to put me at ease. WHO did select the authors for the study writeup. I trust that WHO did not make a random selection of researchers, but chose carefully based upon previous work and probable bias towards a conclusion. As I stated, this is how the global warming mess started.

Btw, I never did state nor meant to suggest that Lancet is not a peer reviewed journal.

The links were my mistake though.
Lancet Article: http://www.who.int/quantifying_ehimp...ns/smoking.pdf
WHO link intro to Lancet Article: http://www.who.int/quantifying_ehimp.../en/index.html


I give you one of three on this observation. Good enough for major league baseball hitting, but not good enough for an academic journal.

I honestly do not understand your point. You want to claim that research funded by WHO is unreliable due to the fact that they are part of some conspiracy to bring down big corporations?

[Viking Prince]

I honestly do not understand your point. You want to claim that research funded by WHO is unreliable due to the fact that they are part of some conspiracy to bring down big corporations?

I believe that this particular report is the result of bad science. I believe that WHO funded the report and supported the choice of scientists to write up the results from the data with the intention of getting a predetermined result. I believe Lancet was selected to publish the report because they also would be inclined to review the report favorably. No conspiracy. Just judicious use of resources to achieve a desired outcome, which makes this a political report and not even close to being science. Just like how the UN approached global warming.

[Elfdude]

I'm sorry but there's only so much you can guess about secondhand smoke and lung disease by looking at incidence of X diseases in relation to incidence of smoking. There's so many other variables being tossed out of the window. Honestly given the diffusal rate of smoke in the air it simply doesn't make sense for second hand smokers to end up with more deaths than smokers.

[Adar]

I'm sorry but there's only so much you can guess about secondhand smoke and lung disease by looking at incidence of X diseases in relation to incidence of smoking. There's so many other variables being tossed out of the window. Honestly given the diffusal rate of smoke in the air it simply doesn't make sense for second hand smokers to end up with more deaths than smokers.

Far more than 650 000 die each year from smoking. WHO estimated that 5.4 million people die from smoking each year (source). The next step is of course to guesstimate if it is reasonable that smoking is 10 times more dangerous than 2nd hand smoking.

[Angrychris]

the air in america is all second hand from factories. noone cares

[Phier]

As a side note the Lancet has lost a lot of credibility lately.

They published the bogus 1 million Iraqis have been killed by the coalition forces nonsense which used methodology so flawed a first year bachelors degree student would have objected to it.

A journal is only as good as its editors.

[Ludicus]

As a side note the Lancet has lost a lot of credibility lately.

Equator network. The following guidelines are currently being developed:

Reporting guidelines under development

The SPIRIT initiative (Standard Protocol Items for Randomized Trials) aims to produce evidence-based recommendations for key information to include in a trial protocol

The SPIRIT initiative: Defining Standard Protocol Items for Randomized Trials. Executive summary (August 2010)

Read the BMJ blog about the third SPIRIT meeting held on 2 October 2010

Strengthening the credibility of clinical research .The Lancet 2010;375: 1225. doi:10.1016/S0140-6736(10)60523-5 Read the editorial

Chan, A-W; Tetzlaff, J; Altman, DG; Gotzsche, PC; Hrobjartsson, A, et al. The SPIRIT initiative: Defining Standard Protocol Items for Randomized Trials

[Viking Prince]

I also do not and have not smoked. As an infant, my father was told to make a choice between smoking and my health. (I was nearly a bubble boy.) I have a great sympathy with reducing secondhand smoke effects, but not with false science. If this be a political or moral question -- argue it as a political and moral question. Do not push bad science into the debate.

So this is a bit personal with me.

[xcorps]

I've tried and failed to quit about a dozen times. I've managed to cut down to about a pack a day, sometimes less. I think I'd be not to inclined to complain if tobacco were prohibited.

[Viking Prince]

It is not even the lack of the raw data that is an issue. The report does not even bother to use a footnote or an apendix to provide the regression equations and variables. They allude to several methods being tested, but neglect to state which method was used. To even imply that a straight line regression was one of the methods is sufficient for me that was the method chosen. And then to go back to Dr Heather Wipfli of the University of Southern California and colleagues, said: "There are well acknowledged uncertainties in estimates of disease burden."

The burden is on the paper to prove that the risk factors are properly applied. The burden is on the paper to show that a straight line regression is a proper means of handling the data. I doubt there is a single economics or math department in the country that would support that methodology for a medical research data base analysis even for an undergraduate class.

btw -- reviewers of papers are seldom given access to raw data or the computer program codes used. Peer review is not replication of results. If you are lucky, you will have the abillity to request some minor data not included in the paper. Many footnotes and appendices result from those requests. In the end, what you see published is basicly what the peer review signed off on. To suggest that we need more to know than the peer reviewers had available simply condemns the report.

[Plant]

As to the funding soure you were not playing fair -- at least not if you were quoting from:


It is funded and published under the authority of WHO. Whether others also contributed funding as this footnote suggests does not change that this is a WHO funded publication. The disclaimer is not there without a reason. I would suggest that the disclaimor is also not sufficient to put me at ease. WHO did select the authors for the study writeup. I trust that WHO did not make a random selection of researchers, but chose carefully based upon previous work and probable bias towards a conclusion. As I stated, this is how the global warming mess started.

Btw, I never did state nor meant to suggest that Lancet is not a peer reviewed journal.

The links were my mistake though.
Lancet Article: http://www.who.int/quantifying_ehimpacts/publications/smoking.pdf
WHO link intro to Lancet Article: http://www.who.int/quantifying_ehimpacts/publications/shsarticle2010/en/index.html


I give you one of three on this observation. Good enough for major league baseball hitting, but not good enough for an academic journal.

It really isn't my fault that your links didn't work, and had to follow links from your BBC link as a source. In any case, I skimmed the lancet article;

Role of the funding source
The sponsors of the study had no role in study design,
data collection, data analysis, data interpretation, or
writing of the report. The corresponding author had full
access to all the data in the study and had final

responsibility for the decision to submit for publication.

I'm no medical practitioner, but it seems to be rather justifiable. Just because "there are well acknowledged uncertainties in estimates of disease burden."
should not mean that medical researchers should not bother. The burden of argument is on you to say why you beleive the report is deceitful. It does not use a straight line method from what I can see, rather it ses collected information from various sources and uses several equations to receive its results of premature deaths and DALYs. No, I am not here to argue about it's validity, only that it appears to have a valid methodolgy from what I can read and understand. It is up to fellow medical researchers to judge its merits. It details its methodology, with its sources, along with its references.

________________________________________________________________________________________________________________________

Lets say the black is the LD50 and the red is the LD80, the straight line would be the LD10 at the green spot and that just doesn't work that way.

I see what you mean, thank you. Though I am curious about the axis. One of them will have to be an inverse, otherwise, the graph will not make sense. It will be bad science to extrapolate, in such a way, without researching point lower dosages, even if your graph do suggest a straight line relationship, I agree. Do you have a source for such a report though?

[Phier]

I see what you mean, thank you. Though I am curious about the axis. One of them will have to be an inverse, otherwise, the graph will not make sense. It will be bad science to extrapolate, in such a way, without researching point lower dosages, even if your graph do suggest a straight line relationship, I agree. Do you have a source for such a report though?

The axis would be the % killed on Y and the dose (inverse) on X.

I don't have a source for this on hand no, it was a big topic in 2001 prior to 9/11 since the Bush EPA wanted to lower the standard for naturally occurring arsenic in water supplies. Its very expensive to remove and the allowed amount was very small. Being Bush this meant of course he wanted to poison the water supply to all the democratic pundits and such straight line graphs were used to show that X 1000's would die because of it.

Its a common political tactic for 'estimated' numbers of whatever malady you wish to politicize. Being they have absolutely no way of knowing how many children die of second hand smoke, and the number being as high as it is, I'm going to assume that they used such a relationship. Being the UN normally doesn't make its methods known for their reports, and just do summaries, its very difficult to find out their methods. Likewise being even Lancet was rather cautious to fully agree, I take that as further proof things were not very scientific.

I'm going to guess it was something like take the number of children dying of asthma, URI's etc, who have a parent who smokes, multiply by a fudge factor, claim thats how many were killed by second hand smoke.