BMJ links cancer to alcohol consumption

[SterlingHopper]

So I see that there is a very similar thread running around right now, but I just saw BMJ's article linking alcohol consumption so some cancers in western Europe.

I think the tagline is:

If we assume causality, among men and women, 10% (95% confidence interval 7 to 13%) and 3% (1 to 5%) of the incidence of total cancer was attributable to former and current alcohol consumption in the selected European countries.

This seems to be a respected, peer-reviewed medical journal. Why haven't we learned that CORRELATION DOES NOT IMPLY CAUSATION?!

 

[SterlingHopper]

Okay, I guess in medical research it is pretty difficult and unethical to demonstrate causal links to cancer.

But, attributing 10% of the incidence of cancer to previous alcohol consumption is kind of a big deal, right?

 

[Wingy]

The article is not an attack on drinking. From the discussion section:

"We found that a substantial part of this incidence was associated with consumption above the recommended upper limit, indicating the potential for cancer prevention merely by adhering to the current recommendations."

"In contrast, for all cause mortality alcohol consumption is often shown to be associated with a lower risk for up to four drinks a day in men and two drinks a day in women."

(Source)

In other words, more than 2-4 drinks per day in men is associated with some forms of cancer, particularly in the upper GI tract. The interaction between smoking and alcohol in esophageal cancer is well known, and this study extends those findings in a larger cohort. It's not super groundbreaking, but it's a nice new piece of information. The general consensus remains that 1-2 drinks per day is healthy and decreases overall risk of death.

If your drinking level is currently at or below the recommended guidelines, there is really no cause for worry. If it is above the recommended guidelines, there is a small increase in the absolute risk of developing cancer in addition to other health problems.

The authors are very clear that the study does not prove causation. Before anyone says that statistical analyses aren't "real" science, unless you think it's ethical to force randomize a group of 30,000 people such that you force 10,000 people to drink heavily for 20 years and compare them to 10,000 people forced to abstain from alcohol, with another group of 10,000 forced moderate drinkers, there's no other way to get this type of information. Is it perfect? Hell no. But it's not meaningless either. Attack the methodology (in particular no information on H. pylori) rather than subscribe to a knee-jerk reaction because you might not like the results.

As for me, I'll keep drinking

 

[Inodoro_Pereyra]

So I see that there is a very similar thread running around right now,

Which begs the question: if you saw there's already another (several, actually) thread on the same subject, why did you open a new thread, instead of contributing to the thread already being discussed?

 

[toxick]

http://www.theregister.co.uk/2011/04/08/giving_up_booze_causes_cancer/

 

[JefeTheVol]

The article is not an attack on drinking. From the discussion section:

"We found that a substantial part of this incidence was associated with consumption above the recommended upper limit, indicating the potential for cancer prevention merely by adhering to the current recommendations."

"In contrast, for all cause mortality alcohol consumption is often shown to be associated with a lower risk for up to four drinks a day in men and two drinks a day in women."

(Source)

In other words, more than 2-4 drinks per day in men is associated with some forms of cancer, particularly in the upper GI tract. The interaction between smoking and alcohol in esophageal cancer is well known, and this study extends those findings in a larger cohort. It's not super groundbreaking, but it's a nice new piece of information. The general consensus remains that 1-2 drinks per day is healthy and decreases overall risk of death.

If your drinking level is currently at or below the recommended guidelines, there is really no cause for worry. If it is above the recommended guidelines, there is a small increase in the absolute risk of developing cancer in addition to other health problems.

The authors are very clear that the study does not prove causation. Before anyone says that statistical analyses aren't "real" science, unless you think it's ethical to force randomize a group of 30,000 people such that you force 10,000 people to drink heavily for 20 years and compare them to 10,000 people forced to abstain from alcohol, with another group of 10,000 forced moderate drinkers, there's no other way to get this type of information. Is it perfect? Hell no. But it's not meaningless either. Attack the methodology (in particular no information on H. pylori) rather than subscribe to a knee-jerk reaction because you might not like the results.

As for me, I'll keep drinking

+1

Some people have a hard time understanding causation, correlation and increased risk. EtOH INCREASES RISK

 

[SterlingHopper]

@Wingy Thank you. I feel a little foolish here; I read the BBC news article, then the abstract, without making it down to the results/discussion. Thanks for pulling out the relevant conclusion.

@Inodoro They seemed to be getting off track. Also, the study was examining a specific gene.

To summarize: <2 drinks a day may decrease risk, >4 drinks may increase risk of upper aerodigestive tract cancer.

I'm interested in how smoking impacts these results. Drinking and smoking go hand in hand for some people (or some cultures).

Given there is a causal association between alcohol consumption and risk of cancer in people who have never smoked, sensitivity analyses with the hazard rate ratios of never smokers indicated noticeable differences compared with alcohol attributable fractions that were based on hazard rate ratios adjusted for smoking from the total cohort, particularly for liver cancer, for which the alcohol attributable fraction in men who had never smoked (AAFSens) was 78% compared with 33% in the total population, and for upper aerodigestive tract cancer, for which the AAFSens was 14% compared with 44%.

 

[larrynoz]

Correlation does not imply causation. Children's growth is correlated to the size of their vocabulary but taller children do not have a larger vocabulary because both are caused by age.

I hate studies like this that get people on the news talking in order to scare you. There's so many other variables in a person's life that these studies don't control for (genetics, family history, etc.) Rant over.

Cheers

 

[dwarven_stout]

http://xkcd.com/552/

 

[Wingy]

@Wingy Thank you. I feel a little foolish here; I read the BBC news article, then the abstract, without making it down to the results/discussion. Thanks for pulling out the relevant conclusion.

@Inodoro They seemed to be getting off track. Also, the study was examining a specific gene.

To summarize: <2 drinks a day may decrease risk, >4 drinks may increase risk of upper aerodigestive tract cancer.

I'm interested in how smoking impacts these results. Drinking and smoking go hand in hand for some people (or some cultures).

Yeah I hate the way that the media characterizes most epidemiology studies, particularly those related to cancer. Most authors are very careful to clarify exactly what they mean, but the media only seems to take the big scary punchline and fails to put it in the proper context. I blame the lack of appropriately trained science journalists - it's difficult to look at studies like this unless you are familiar with the methods/jargon.

I just wanted to make one caveat to your summary statement: based on the paper, any alcohol consumption increases risk of cancer (please keep in mind that the absolute risk is still very low even though it increases 3-10%), the increase is just very mild for moderate drinkers. Moderate alcohol consumption decreases risk of cardiovascular disease, which is much more common, so the overall effect is to decrease risk of death (it's weird to think about it like this because it still increases the risk of developing cancer).

If you're interested in the relationship between alcohol and cancer you can find an excellent (and free!) review article here. It's dated, but the basic information has not changed dramatically. As said previously, this BMJ article is not breaking new ground.

Almost anything you do increases your risk of death somehow, and for me, the very small relative increase in the risk of developing cancer is offset by the fact that beer is delicious.

 

[Wingy]

http://xkcd.com/552/

This one is probably my favorite for most studies that find an association with cancer

http://xkcd.com/882/

Hooray for multiple comparisons!

 

[redcoat_or_rebel]

Correlation does not imply causation. Children's growth is correlated to the size of their vocabulary but taller children do not have a larger vocabulary because both are caused by age.

i don't think any part of that is correct. Physical growth is never linked to mental function and age is an arbitrary label.

 

[JefeTheVol]

OK, I believe I have had enough of the pitchforks and torches regarding this article. First of all, the British Medical Journal IS a scientific journal...not media in the Fox News/CNN sense of the word, and they definitely are no Anthony Bourdain twitter account. Second of all, we need to get our terms in order.

Odds Ratio: Odds of having cancer in a group exposed to EtOH (ie more than 2 drinks/day) NOT STUDIED HERE (the study is prospective)


Relative Risk: Probability of getting cancer in an exposed (EtOH) group as opposed to the unexposed. INCREASED


Attributable Risk: the difference in "risk" between the exposed (EtOH > 2/day) and the unexposed (EtOH < 2/day) INCREASED
according to the data

Absolute Risk: The reduction in risk associate with the unexposed (EtOH < 2/day) as compared to Placebo: NOT STUDIED HERE (because the study did not differentiate the 0 drinks/day group from the 1 or 2 drinks/day group)

This study was a Cohort Study...meaning that they took a group of people of comparable genetic backgrounds, both drinkers and nondrinkers, and followed them through time to see who developed CA of any etiology. What they calculated was RELATIVE RISK (the probability of getting CA was more in the exposed group than the unexposed group)

They found that the odds of the being a drinker (> 2 drinks/day) and getting CA was significantly more than the odds of being a drinker (> 2 drinks/day) and NOT getting CA. Thats all!!!

They then calculated the attributable risk of EtOH exposure by the following formula:

Drinkers with CA/(Total drinkers with and without CA) - Nondrinkers w CA/ (Total Nondrinkers with and without CA)

Ya'll are correct...these are only statistical data so they are NOT PREDICTIVE...the only thing they calculated was that EtOH is ATTRIBUTABLE to CA, NOT CAUSATIVE.

Nowhere in the article is causation even mentioned, because the authors were not investigating cause, they were investigating attributable risk.

I understand that everyone wants to defend their favorite hobby, but attacking a study that took data from 360,000 people is pretty hard at which to raise a valid pitchfork!!!

-Jefe-

 

[rico567]

Oh, not this crap again.

Lewis Black says it best:

"The people who told us about sun block were the same people who told us, when I was a kid, that eggs were good. So I ate a lot of eggs. Ten years later they said they were bad. I went, "Well, I just ate the eggs!" So I stopped eating eggs, and ten years later they said they were good again! Well, then I ate twice as many, and then they said they were bad. Well, now I'm really ****ed! Then they said they're good, they're bad, they're good, the whites are good, th-the yellows - make up your mind! It's breakfast I've gotta eat!"

 

[redcoat_or_rebel]

attacking a study that took data from 360,000 people is pretty hard at which to raise a valid pitchfork!!!

-Jefe-

have you seen the medical industry's track record ?? "peer reviewed" should be synonymous with "back scrathing".

 

[SterlingHopper]

XKCD ftw!

I'm willing to eat my words about the validity of this study, and focus negativity back at the press coverage that made the results sound so bad.

But, I'm still confused about the Alcohol attributable fractions. Isn't AAF like saying, "of the X people who got Y in their lifetime, Z percent of them can attribute it to increased alcohol consumption"? I see the AAF calculation, but it seems like they're just adding probabilities, not calculating them.

 

[JefeTheVol]

have you seen the medical industry's track record ?? "peer reviewed" should be synonymous with "back scrathing".

Well, in some sense you are correct, that does happen, as it does in any profession. But I feel that such a wide and absolute statement about the thousands of MD's and PhD's that dedicate their lives to improving our knowledge base is detrimental to science in general. Half of all medical research is not done by MD's anyway.
-Jefe-

 

[redcoat_or_rebel]

Well, in some sense you are correct, that does happen, as it does in any profession. But I feel that such a wide and absolute statement about the thousands of MD's and PhD's that dedicate their lives to improving our knowledge base is detrimental to science in general. Half of all medical research is not done by MD's anyway.
-Jefe-

I agree that all the scientists and MD's aren't bad but it is like any other system - sure there are good journals and good research going on but if you think medical journals established fields of science are above censorship and bribery you are mistaken.

i like to use global warming as a prime example. everything about the science of it was completely made up, the weather model data was purposely altered to reflect a warming trend when we were actually on a cooling trend. the emails came out of scientists at univeristies in different parts of the world discussing suggesting different techniques to manipulate the weather models.
there were also certain journals only publishing papers of pro-global warming research.

crappy people seek power, good people tend to want to live their lives and let be people live theirs. i find its true in most things.
like Thomas Jefferson says, beer softens the temper, cheers the spirit and promotes health hombrew research!

 

[rico567]

Well, in some sense you are correct, that does happen, as it does in any profession. But I feel that such a wide and absolute statement about the thousands of MD's and PhD's that dedicate their lives to improving our knowledge base is detrimental to science in general. Half of all medical research is not done by MD's anyway.
-Jefe-

I mean no disrespect, but good intentions (=dedication) are not a component in validating a theory. This is in line with Allan Bloom's assertion in Closing of the American Mind that "commitment" has become a major component in the hallucination that values can somehow become objective. This is in line with the blind assertion of some people that thimerosol in vaccines is the cause of autism. Thus, the process of finding an actual cause for a condition is short-circuited while parents take the hideous risk of not vaccinating their children, based on nothing but sheer belief, as multiple studies in the U.S. and Europe have found NO link between thimerosol and autism.

 

[JefeTheVol]

I mean no disrespect, but good intentions (=dedication) are not a component in validating a theory. This is in line with Allan Bloom's assertion in Closing of the American Mind that "commitment" has become a major component in the hallucination that values can somehow become objective. This is in line with the blind assertion of some people that thimerosol in vaccines is the cause of autism. Thus, the process of finding an actual cause for a condition is short-circuited while parents take the hideous risk of not vaccinating their children, based on nothing but sheer belief, as multiple studies in the U.S. and Europe have found NO link between thimerosol and autism.

Are we no longer talking about the validity of Western European statistics and now talking about inherent humanistic qualities that bias research?

I dont know anything about the "hallucination of values" but the whole idea behind scientific theory is that it is flawed but its the best we got.

 

[MikeRoBrew1]

Thanks Wingy, your post was very informative

 

[redcoat_or_rebel]

I mean no disrespect, but good intentions (=dedication) are not a component in validating a theory. This is in line with Allan Bloom's assertion in Closing of the American Mind that "commitment" has become a major component in the hallucination that values can somehow become objective. This is in line with the blind assertion of some people that thimerosol in vaccines is the cause of autism. Thus, the process of finding an actual cause for a condition is short-circuited while parents take the hideous risk of not vaccinating their children, based on nothing but sheer belief, as multiple studies in the U.S. and Europe have found NO link between thimerosol and autism.

There is a link between vaccines and autism - Andrew Wakefield's study's results have recently been reproduced by several others. Autism is actually a symptom of inflammation of the GI tract adn the gut in general. Vaccines use adjuvants to force an extreme immune reaction causing an auto-immune condition. the treatment (well, one aspect of treatment) for autism is diet. you take away the food allergies and allow the GI and gut to be able to absorb the nutrients it needs to heal itself.

 

[dwarven_stout]

There is a link between vaccines and autism

That's the ultimate in arguing that correlation proves causation.

 

[Wingy]

That's the ultimate in arguing that correlation proves causation.

Yeah... for someone who is eager to claim data fabrication for solid studies, it's ironic that they support one of the worst examples of fraud, bad ethics, and using science to promote a specific viewpoint.

Not trying to start a flame war here, but vaccines do not cause autism. Period. Full stop. "Dr." Wakefield (he was struck off the list of physicians in the UK for ethical violations and now runs a clinic in Austin, TX, where he is once again under investigation for dubious ethical practices) published a paper in 1998 where he linked gut inflammation to autism. Nowhere in the paper is a link between vaccines and autism mentioned (it would never have been published in The Lancet if he had made such an egregious claim); Mr. Wakefield claimed this link in a subsequent press conference (much to the surprise of his coauthors).

The findings of this study were never replicated (if they have, they haven't shown up in the peer-reviewed literature), the pathologist withdrew his name from the paper after he discovered Mr. Wakefield had "reexamined" specimens and changed the results to match what he wanted. He's the kind of person that doesn't think twice about subjecting children to horribly invasive and unnecessary procedures (taking a piece of someone's intestine is not something you can do in an office visit) and laughed about paying children at a birthday party for blood samples.

Coincidentally, Mr. Wakefield only claimed that the trivalent MMR vaccine was bad (and just happened to recommend a monovalent measles vaccine instead. He developed this vaccine and owned the patent on it). The study was financed by a drug company and this financial support was not officially disclosed when the paper went to press.

/rant off.

Jefe, I agree with you. The current scientific method is not perfect but it's better than everything else we've tried so far.

 

[pjj2ba]

Well, in some sense you are correct, that does happen, as it does in any profession. But I feel that such a wide and absolute statement about the thousands of MD's and PhD's that dedicate their lives to improving our knowledge base is detrimental to science in general. Half of all medical research is not done by MD's anyway.
-Jefe-

No offense to any Physicians, but why do so many think that their MD training means they are trained to be good scientists? With a PhD, it takes 4-7 yrs to get the degree, and the focus of your training the whole time is how to do good science (hopefully). Then you have another 2-4 years of training as a post-doc, before getting your own lab (hopefully). And some of us still do crappy science! I will occasionally peruse the medical journals and am dismayed to see what passes as good research, as in, we saw this ONCE, so this must be the way it is. To be sure, there is lots of great work being done, but also a lot that is a bit iffy. There certainly are MDs out there doing good research, but I tend to think of them more like they are naturals at it. Some folks have the knack, others don't. Unfortunately not having the knack for it doesn't stop some. This is fine if you are brewing beer, but not so good if people are making medical decisions.

I have a friend whose wife (PhD) works for a MAJOR pharmaceutical company. Her job responsibility is to oversee and make sure the MDs were using proper scientific methods and techniques, as in, they aren't competent enough to do it on there own. They certainly have useful knowledge, but just need a little help with conducting good research

I feel more comfortable (research wise) when the MD is also a PhD.

Now with the alcohol study. Meh. I bet one could find that most anything will cause cancer in some portion of the population. That is the quirk of genetics. Something can be totally benign to 99+% of the population, but for the unfortunate few, it can cause cancer. If one "carefully" chooses their test population this # can go up, which can translate into more research $$ for the lab. So does it cause cancer? Should it be banned or regulated? Lets let a bunch of politicians decide - they know what is best. Sorry for that, let's not get this all political

 

[redcoat_or_rebel]

Yeah... for someone who is eager to claim data fabrication for solid studies, it's ironic that they support one of the worst examples of fraud, bad ethics, and using science to promote a specific viewpoint.

Not trying to start a flame war here, but vaccines do not cause autism. Period. Full stop. "Dr." Wakefield (he was struck off the list of physicians in the UK for ethical violations and now runs a clinic in Austin, TX, where he is once again under investigation for dubious ethical practices) published a paper in 1998 where he linked gut inflammation to autism. Nowhere in the paper is a link between vaccines and autism mentioned (it would never have been published in The Lancet if he had made such an egregious claim); Mr. Wakefield claimed this link in a subsequent press conference (much to the surprise of his coauthors).

The findings of this study were never replicated (if they have, they haven't shown up in the peer-reviewed literature), the pathologist withdrew his name from the paper after he discovered Mr. Wakefield had "reexamined" specimens and changed the results to match what he wanted. He's the kind of person that doesn't think twice about subjecting children to horribly invasive and unnecessary procedures (taking a piece of someone's intestine is not something you can do in an office visit) and laughed about paying children at a birthday party for blood samples.

Coincidentally, Mr. Wakefield only claimed that the trivalent MMR vaccine was bad (and just happened to recommend a monovalent measles vaccine instead. He developed this vaccine and owned the patent on it). The study was financed by a drug company and this financial support was not officially disclosed when the paper went to press.

/rant off.

Jefe, I agree with you. The current scientific method is not perfect but it's better than everything else we've tried so far.

so you believe everything you read? people's blind faith in the medical INDUSTRY and medical journals is pretty astounding.

depending on my schedule tomorrow i may be going to hear him speak about his book.

 

[Wingy]

so you believe everything you read? people's blind faith in the medical INDUSTRY and medical journals is pretty astounding.

depending on my schedule tomorrow i may be going to hear him speak about his book.

Not blind faith. Background in molecular biology, epidemiology, and lots of time spent looking at the effect of vaccines in the population. I've read almost all of the studies and reporting that deal with the "controversy" and it's really the only solid conclusion. It's ONE study and a few fringe scientists vs. the general consensus.

The consensus (that vaccines are safe, effective, and do not cause autism) has a molecular mechanism, a solid theory, and decades of data to back it up. The contradicting viewpoint simply does not. It was an interesting idea when Mr. Wakefield first proposed it, but it has since been conclusively disproven. Totally normal for science.

Arguing about things like this is totally pointless (it's not like I'll be able to convince you that you are wrong and I am right and vice versa) and we've gone way off topic so no more posts from me about vaccines here.