The study that never existed: cancer edition

A couple of years ago on this blog, we shared the unfortunate story of Iced tea, kidney stones, and the study that never existed.

Now, history seems to be repeating itself with another phantom study about vitamin supplements and cancer risk.

As Michael McBurney writes on the TalkingNutrition blog of supplement manufacturer DSM, a variety of media outlets (Sunday Express, The Guardian, Daily Mail, CBS News) reported on a study supposedly presented at the American Association for Cancer Research Annual Meeting by Dr. Tim Byers of the Colorado Cancer Center.

McBurney is not too happy about it.

Byers is reported to have presented evidence, after studying thousands of patients for ten years, that some people get more cancer while on vitamins. My problem is the apparent lack of peer-reviewed evidence with hazard ratios (HR) or odds ratios (OR), or even an abstract to partially substantiate the claim.

A session entitled “Dietary Supplements and Cancer Risk and Prognosis” was apparently held Monday, April 20 from 5-6:30pm. Dr Tim E Byers is listed as an invited speaker but there is no abstract for his presentation. The University of Colorado Cancer Center has promoted his research but an abstract, peer-reviewed paper, or supplemental data cannot be found. The question is, did journalists ask to see the scientific evidence? Certainly, I cannot make an evaluation without some scientific evidence beyond expert opinion.

In a follow-up post, McBurney wrote that there was, in fact, no presentation of new study data by Byers at the AACR meeting, and that “The entire news cycle linking multivitamin/mineral supplements with cancer risk seems to have been stimulated by the university press release alluding to a commentary published in 2012.”

I confirmed McBurney’s account with Garth Sundem in the University of Colorado media relations department this morning. He told me that there is no new meta-analysis, and that the “study” referred to in these news accounts is indeed the 2012 paper cited by McBurney (apparently a narrative review of the evidence and not study per se) from the Journal of the National Cancer Institute.

Sundem said he was surprised to see that his news release was “immediately and aggressively sensationalized” by the media, and described a “ripple effect, almost like a game of telephone tag, where news outlets, especially in the UK, seemed to give increasingly more sensational accounts of the study without ever going back to the original source.” He described Byers as being “just as horrified as you’d expect any academic researcher would be” by the hype surrounding the non-study.

This is a situation that’s truly surreal.

On the one hand, we have news stories — and there are lots of them — ginning up a cancer scare about vitamins on the basis of a news release that actually refers to a paper that was published 3 years ago. (That news release, by the way, bears some culpability in this mess for suggesting that there was actually some “news” to cover in Byer’s presentation, when that doesn’t seem to be the case. The release was also scant on evidence and other specifics that we consider essential elements of any good health news communication.)

But now we also have a supplement industry trade group, the Natural Products Association, robotically objecting to that imaginary new research and complaining that the studies in it were “cherry-picked”!

In this case, Dr. Tim Byers selected 12 trials but failed to include other studies that may have evaluated negative outcomes in a long-term prospective study. Meta-analyses include only a small fraction of available evidence and it is really at the researcher’s discretion – and therefore bias – as to which studies he or she decides to include in the analysis. We question why Dr. Byers cherry-picked these specific 12 studies and whether or not there was other research available that could have resulted in a different outcome. We simply cannot draw accurate conclusions from this sample of studies without understanding which were chosen and why.”

Talk about leading the media like sheep! This is a black eye for everyone involved in the health news food chain,  and another prime example of why HealthNewsReview.org is holding the players in that chain — including both news release writers and journalists — to a higher standard.

Hat tip to Ivan Oransky, MD, publisher of Retraction Watch, for pointing us to McBurney’s post.