Booze balderdash: alcohol, cancer and another nonexistent study

Kevin Lomangino is the managing editor of HealthNewsReview.org. He tweets as @KLomangino.

Photograph of human hands holding two cocktail shakers and pouring orange color cocktail into two martini glassesYou’ll be tempted to put your happy hour plans on ice after reading today’s alarming news stories about alcohol consumption, which claim that “even low to moderate amounts” of booze are now proven to cause cancer.

But you should raise a glass anyway, because it appears the editors on these stories have all had one too many.

Christopher Snowdon, writing at The Spectator, explains what’s wrong with stories about this “study” — and that’s that there was no “study.”

The study was published in the journal Addiction and the only problem with it is that it isn’t a study. It’s a comment piece. It doesn’t contain any new research, nor does it contain statistical analysis of previous research. It’s not a meta-analysis or a systematic review. It is a short essay in the ‘For Debate’ section of the journal in which one woman gives her opinion about whether correlation equals causation when it comes to the epidemiological evidence on alcohol and cancer.

To clarify: There’s no news here. No study. No reason for anyone to think alcohol is any more dangerous today than we thought it was yesterday.

What we have is a scientist making an argument that, based on existing observational studies, there’s enough evidence to conclude that alcohol causes specific types of cancer.

It’s an interesting argument. And you are entitled to agree with that opinion or disagree with it. But you are not entitled to label the opinion a “study” or a “meta-analysis” and cite it as new proof that alcohol causes cancer.

As Snowdon correctly points out, the issue here is not whether alcohol does or doesn’t cause cancer but how the news media reports on the question.

How have we got to the stage where the opinion of a single academic from New Zealand, writing in the commentary section of a specialist journal, becomes front-page news for two British newspapers? How can these newspapers justify describing an op-ed as a ‘study’ and someone’s subjective view as ‘proof’?

But Snowdon himself is wrong to suggest that health care news has sunk to some kind of new low with this story. Journalists have been reporting on studies that don’t exist for years now.

You might also like

Comments (10)

We Welcome Comments. But please note: We will delete comments left by anyone who doesn’t leave an actual first and last name and an actual email address.

We will delete comments that include personal attacks, unfounded allegations, unverified facts, product pitches, or profanity. We will also end any thread of repetitive comments. Comments should primarily discuss the quality (or lack thereof) in journalism or other media messages about health and medicine. This is not intended to be a forum for definitive discussions about medicine or science. Nor is it a forum to share your personal story about a disease or treatment -- your comment must relate to media messages about health care. If your comment doesn't adhere to these policies, we won't post it. Questions? Please see more on our comments policy.

Marc Beishon

July 22, 2016 at 5:15 pm

To be fair, the commentary author, Jennie Connor, has just published research:

Connor JL, Kydd R, Maclennan B, Shield K, Rehm J. Alcohol-attributable cancer deaths under 80 years of age in New Zealand. Drug and Alcohol Review 2016 DOI: 10.1111/dar.12443
http://www.ncbi.nlm.nih.gov/pubmed/27306121

And her commentary is a useful review of the latest epidemiology for alcohol and cancer and was worth the press release – or at least, I found its useful and it’s rather more than Snowden says (writing in a right wing publication that tends to sneer at such things).

    Janet Camp

    July 25, 2016 at 8:12 am

    Thanks for the balance on this. I saw it in the Guardian, and they do say it is “an analysis” of evidence, although they do not make that as clear as they should have and they also refer to it as a “study”. They also give other useful information at length, about UK gov’t guidelines on alcohol and why they were established. On balance, it’s a good article, though it could have been clearer on the source of the information. Also, thanks for being clear about Snowden–which is something HR should have done.

Janet Camp

July 25, 2016 at 8:17 am

I looked at the PubMed entry and now I am actually angry at Health News for the way they presented this. They are no better than those they criticize. Offering the Snowden criticism was nothing more than an example of what they are supposesedly against.

    Kevin Lomangino

    July 25, 2016 at 9:32 am

    Janet,

    Thanks for your comment, but I beg to differ. The Guardian piece is as clear as mud because the headline trumpets a “study” that does not exist. I believe that’s something worth pointing out to our readers. Perhaps you are conflating our criticism of the news stories with the issue of whether alcohol causes cancer or not. Maybe it does, maybe it doesn’t. I’m not commenting on that at all. The issue is that a journal opinion article that reflects one researcher’s review of the evidence is not a “study” and is not something that should generate definitive headlines such as “Alcohol is a direct cause of seven ​​forms of cancer, finds study.” That headline is absolutely wrong, false, and misleading because there is no study — only one person’s subjective summary and review of evidence. The entire premise of these stories is fundamentally incorrect and deceptive. The research that’s referenced by the other commenter is not what the stories generating headlines are based on. Just because this researcher happens to have published a study recently does not give news outlets license to call her separate opinion article a “study” and claim that it provides definitive new information.

    Or is your objection perhaps more tied into Snowdon’s political leanings? I was not aware of his political stance and frankly see little relevance. The issue he’s bringing to the fore is something that people across the political spectrum can agree with: News outlets should not report on studies that don’t exist and claim that those non-existent studies provide definitive new information.

    Kevin Lomangino
    Managing Editor

      Marc Beishon

      July 25, 2016 at 1:31 pm

      Kevin, I work in this field too, in cancer in particular, and I am a little concerned that you’re not always tracking back to the source properly as I have noted a couple of mistakes recently, one of which you have corrected.

      In this case, have you read Jennie Connor’s paper? I read a lot of papers and I feel that this does qualify as a study. It’s free to air at:

      http://onlinelibrary.wiley.com/doi/10.1111/add.13477/full

      It is not a ‘short essay’ – it is an authoritative review of more than 3,700 words and while you might say ‘so what’ to doing a Medline review of existing meta-analyses the author says “more recent epidemiological studies not included in these publications were also reviewed.”

      I certainly take issue with Snowdon’s put down of a ‘one woman from New Zealand’ – as though he knows anything about her research and qualifications – but more than this it is supremely ironic for you that Connor says:

      ‘In this context, some confusion and scepticism about whether alcohol causes cancer may seem understandable, but in some cases doubt is also being generated by dissemination of misinformation, which undermines research findings and contradicts evidence-based public health messages.”

      My conclusion is that this study – as it surely is – provided the basis for correct reporting on harms that are not being widely reported in this overall fashion for a number of cancer types and certainly there is a widely held view that alcohol and cancer are not a serious concern.

      Kevin Lomangino

      July 25, 2016 at 2:51 pm

      Marc,

      Thanks for commenting. The Connor paper is a narrative review and comment, the purpose of which is to argue a point — that alcohol is causally linked to cancer. It is not a study. Here’s the description of the methods: “Recent epidemiological and biological research on alcohol and cancer was reviewed and summarized, drawing upon published meta-analyses identified from the Medline database and the archives of the International Agency for Research on Cancer.” In other words, “I looked at and summarized these other studies and this is what I think.” There’s no description of how the literature search was conducted. No discussion of inclusion or exclusion criteria. No attempt to address publication bias. Nothing that would distinguish this as a new and objective piece of research rather than the “For Debate” commentary that it is. Maybe this type of review should be described as a study in your book, but that’s not how such papers have traditionally been characterized. Journals typically do not issue press releases about such reviews, and newspapers typically do not report on them, because one person’s assessment of the literature — no matter how well reasoned and well-supported it may be — is not a study, nor is it news.

      Kevin

      Marc Beishon

      July 26, 2016 at 9:07 am

      Just to be clear Kevin – leaving aside whether it’s reported as a study or a review (and the distinction is not likely to be appreciated by a majority of laypeople) – are you saying the paper has no news merit whatsoever and should not be reported?

      Kevin Lomangino

      July 26, 2016 at 10:57 am

      It’s exactly because the distinction is not clear to lay people that researchers, journalists and PR professionals need to appreciate and respect it even more. As far as the paper having news merit or not, I personally don’t see the value, except perhaps if this were to be used as a jumping off point for a larger discussion of this contested issue. But the paper should be accurately described for what it is — one researcher’s opinion based on her analysis of existing studies. It should not be framed as definitive new evidence showing that alcohol causes cancer, because it’s not.

      Marc Beishon

      July 27, 2016 at 1:50 pm

      Well I still think you’ve come down too hard on this one given that the science itself is accurately reported in the review and mostly in the news stories (and the press release was careful to only use the word ‘review’). There is contention about the counter-cardio benefits and the extent of harm at low-moderate level, I agree. But I think the bigger picture, which you don’t address, is the lack of awareness of this particular harm from alcohol and whether the press stories are on balance doing a reasonable job, even as usual that the headlines are often over the top and not written by the reporters. And that there is a strong drink lobby that pumps out messages that deny harm.

      You’ll be pleased to see – and maybe it’s down to you – that at least two of the stories were amended:

      At the Huffington Post: “Correction: An earlier version of this article incorrectly identified Connor’s opinion article as a new study on alcohol’s link to cancer. In fact, Connor’s opinion piece reviews existing literature on the subject and does not present new data or conclusions. Language has been updated throughout.”
      http://www.huffingtonpost.com/entry/alcohol-cancer-risk_us_5792c2ede4b0d3568f83637e

      and: “Editor’s Note: This story was updated on July 25. The original story referred to the Connor’s article as a “meta-analysis,” the story was corrected to state that the article was a review, and an opinion article.”
      http://www.livescience.com/55497-does-alcohol-cause-cancer.html

      Music to your ears!

David Nisbett

July 25, 2016 at 12:27 pm

Are we forgetting the amount of alcohol this scientist states is required to increase the risk of various cancers? 50 grams….PER DAY!! Pretty sure these folks are going to die from many other related or unrelated issues long before the cancer takes them. I just hope they don’t take someone else with them.