Tuesday, October 01, 2019

One study on red meat does not invalidate decades of other studies

A recent health study in the respected journal Annals of Internal Medicine has received an awful lot - probably too much - of media attention. That's because it goes against pretty much every other related study over the last 20 years. To call it controversial is putting it mildly and, of course, controversial sells papers, TV time, and internet advertising.
The meta-analysis (combining many different published studies and reinterpreting the result) was led by researchers at McMaster and Dalhousie Universities in Canada, and was commissioned, produced and paid for by an organization called NutriRECS, a little known outfit that purports to be "an independent group with clinical, nutritional and public health content expertise". It has resulted in headlines like "Eat less red meat, scientists said. Now some believe that was bad advice", "A study says full speed ahead on processed an red meat consumption", "Is it time to put red meat back on the menu?", "No, beef isn't bad for you: scientists conclude there is no need to eat less red or processed meat", etc, etc
What the study actually concluded was that, "Most people can continue to eat red and processed meat as they do now. The major studies have found that cutting back has little impact on health". It essentially found that the evidence it looked at was too weak to say for sure that there was a link between eating red and processed meat and life-threatening conditions like cancer, heart disease, stroke and diabetes.
Now, this is not quite as dramatic as some of the headlines, but it still flies in the face of most other studies and the advice of organizations like the World Health Organization, the US Federal Government, the American Heart Association, the American Institute for Cancer Research, the Canadian Cancer Society, Harvard University, Public Health England, Health Canada, etc, etc. The very fact that it contradicts so many other studies and august scientific bodies is a good indication that something is probably wrong with it, and a little further research indicates that MANY things are probably wrong with the study.
A WebMD article quotes one major nutrition expert as saying, "It's the most egregious abuse of data I've ever seen .. there are just layers and layers of problems". Several groups have requested that the journal postpone publication for further investigation. Among the "grave concerns" of these scientists are:
  • Omitted studies (that would have significantly changed the results and conclusions);
  • Incomplete picture (particularly as regards what other foods the participants were eating);
  • Inappropriate analysis (especially the absence, even the impossibility, of randomised control trials, and the use of observational data);
  • Contradictory data (using different assessment methods on the same data results in very different conclusions);
  • Confusing message (drawing conclusions nd making recommendations diametrically opposed to every other study can only result in confusions for consumers, especially given that there IS a consensus in the scientific community);
  • Unusual inclusions (such as a review of attitudes towards eating meat, which obviously resulted in a slant towards eating meat among meat-eaters);
  • Ignoring the environmental impact and animal welfare considerations (although, frankly, one would not expext these to be included in a nutritional study).
Anyway, the genie is out of the proverbial bottle. However poor the actual science, the comvenient conclusions will probably lodge in people's minds, and confirmation bias will probably take hold, much like the discredited but still often-quoted study on the link between vaccinations and autism. Hell, Donald Trump will probably get in on the act at some point.
The other conclusion we can draw from all this is that nutrition research is hard and emotive, and some people are always going to disagree.  Also, pulling together several studies with different methodologies and parameters and objectives is even harder, and more subject to misinterpretation. Plus, even if people can agree on the science carried out, some people will always disagree with the conclusions drawn (for example, if reducing red meat consumption results in 6 fewer heart attacks or seven fewer cases on cancer, should we conclude, as this study does, that the effect is minimal, or that it is substantial?) Kudos to the Globe and Mail (and specifically André Picard) for making this clear to their readers.
But, make no mistake, whatever this study suggests, you should probably still avoid it reduce your consumption of red and processed meats. As the BBC concludes, "The weight of scientific opinion falls on the side of reducing red and processed meat consumption". And if we can't trust the BBC, then what can we trust in this world?

No comments: