Comment by scottLobster | Hacker News Reader

Comment by scottLobster | Hacker News Reader

That study's introduction is a little disingenuous and frankly a little alarmist. To break it down:

Edit: Just to be clear, I'm not trying to disprove the study. I simply think their introduction is needlessly hyperbolic.

"Alcohol consumption is a well-established risk factor for a range of cancers, accounting for approximately 4% of new cancer cases annually [1, 2]."

The first citation is pay-walled, so I can't confirm, but the abstract states "The net effect of alcohol consumption on health is detrimental, with an estimated 3.8% of all global deaths and 4.6% of global disability-adjusted life-years attributable to alcohol." It says nothing about cancer specifically, but this could just be a vague abstract. Perhaps the data holds up, but a very large fraction of that 3.8% and 4.6% would have to be due to cancer. I'm skeptical.

The second citation however is blatantly misrepresented. It appears to be where the 4% number comes from, but it is UK specific (and almost certainly influenced by UK-specific drinking habits). The authors by contrast present it in a general context. Likewise it's worth noting that the source puts quitting smoking, keeping a healthy weight, and eating more fruits/vegetables as more important in preventing cancer than reducing alcohol consumption. And even at UK levels it's apparently only marginally more important than reducing sun exposure (also a UK-specific result).

They go on to effectively support the statement "alcohol increases cancer risk" in detail and I guess I can't blame them for wanting to grab the reader's attention, but scientific spin doctoring and lack of hard numbers irks me regardless of context.

As for alcohol companies telling their customers as little as legally required about the negative effects of alcohol, I'm glad they put some numbers on the issue but I think we can file that in the "no shit" folder. The policy makers/lesson planners/etc they appear to aim the study at are almost certainly well aware of the concept of conflict-of-interest, and if they aren't then I doubt they're literate enough to be swayed by this study.

> but it is UK specific (and almost certainly influenced by UK drinking habits).

You lost me here. What exactly is your hypothesis about UK "drinking habits" that causes brits to get cancer at significantly higher rates than the rest of the world? All studies have to study some accessible subset of humanity. This one picked the UK.

I mean, it's OK to ask for rigor and point out problems with research, but this is an awfully big whopper of your own. You don't get to just dismiss a fairly straightforward result because warm beer.

Uh... quantity for starters. Not that the UK drinks the most, that honor apparently goes to Belarus according to the WHO (https://en.wikipedia.org/wiki/List_of_countries_by_alcohol_c...), but the study itself states a dose-relationship response to some forms of cancer. Therefore the amount of alcohol a population consumes would affect the 4% number, and the UK is demonstrably not a representative sample of the rest of the world.

There are also biological/ethnic factors. Native Americans are more susceptible to alcoholism, and Asian populations are more likely to be allergic to alcohol. There are also regional differences in diet and the type of alcohol consumed, even when and how it's consumed. You can't just take a 4% number from one region and apply it generally.

Surely the study you cite was about a correlation to quantity and cancer, though, which would mean they've already corrected for that. To do otherwise would be laughably awful science. If you want to make that assertion, you need to show that the study is flawed, not just assert that you personally believe they are idiots who don't know how to get a paper published.

And the stuff about alcoholism and allergies[1] is just you flailing to find an excuse to discredit a result you don't like. That's not the way science works. At all.

[1] FWIW: it's not an allergy at all. East asians are more likely to carry a defective gene for alcohol dehydrogenase, meaning that ethanol metabolism goes through a different process in their bodies and tends to produce an inflamation response.

The study I cite, and the link you clearly didn't click, is simply a table of countries sorted by alcohol consumption per capita, showing the massive differences in consumption and types of alcohol consumed between countries as of 2010. That's just the largest factor of many that affect cancer rates due to alcohol consumption.

Also you seem to have the impression I'm trying to disprove the study. As I stated in my original comment I'm not, I simply think they were being hyperbolic in the introduction and that annoyed me. And assuming the UK study is the sole source of that 4% number, then it's wrong and alarmist as presented and should be qualified.

I have no trouble believing their results, which are wholly unrelated to this minor point. And I hope your defensive pedantry about the definition of an allergy makes you feel better, because it doesn't disprove my argument.

> The study I cite, and the link you clearly didn't click, is simply a table of countries sorted by alcohol consumption per capita

I meant the UK study. You say it can't be right because brits drink more. I say that it's clearly already written as a correlation between consumption and cancer rates (because what else would it be?), which indeed makes your point sort of specious.

> defensive pedantry about the definition of an allergy

You're pontificating on the internet about the validity of alcohol studies, yet dismissing a clear fact about the same subject as "pedantry". I think that says all that needs to be said about the argument. LOL, as it were.

Factors that (may) affect cancer rates, that vary between countries from the top of my head:

Amount of alcohol consumption, strength of the alcohol (I once read that only above ~15% alchohol gives you oral cancer), content of secondary alcohols like methanol, genetic factors like Alcohol flush reaction, which is known to also lead to higher cancer rates in combination with alcohol and is very common in Asia.

Go to Google Scholar and type in the paper's title - often there will be a link to a publically available copy of the pdf, such as is the case for reference 1.

And you sound like you made up your mind before even really digging into the literature. Your first job as a scientist is to not fool yourself.

I did exactly that actually. The only link to the study I could find in 5 minutes of searching is paywalled. I'm not going to pay $31.50 just to make a point on HN.

http://www.lunaliving.org/pdf/global-burden-of-disease-and-i...

The link is on the right next to the google scholar search results.

Based on the tables in this paper, alcohol-caused cancer is responsible for 0.8% of deaths. I don't see how the numbers in the paper could be used to calculate a value of 4% for any statistic related to alcohol+cancer.

I read it as 0.8% of all deaths, not of only the deaths caused by cancer. I think that is the most logical reading.

If that's correct, I get at that 4% this way: according to https://www.cancer.org/content/dam/cancer-org/research/cance..., about 20% of all deaths are caused by cancer.

If, of that 20%, 0.8% is caused by alcohol-induced cancer, 4% of all deaths caused by cancer are caused by alcohol-induced cancer.

I think the confusing part is trying to click on the "result". The PDF link is usually on the right, next to the title.

Reply on news.ycombinator.com