How To Read Through The Science Hype In Newspapers
Many newspapers either move over general interest reporters (Hello, LA Times!) to science reporting or have sloppy reporters (Gina Kolata and Jane Brody at The New York Times) who err in their thinking and reporting with great regularity. Unfortunately, that's hard for the average reader to see -- and many tend to give these reporters more respect than they deserve, simply because they're writing for a major media source.
Well, there's an excellent piece in Forbes by a medical reporter I hadn't read (or maybe noticed that much) before but now have respect for, Larry Husten. An exerpt from his piece, "Is Red Meat A Fish Story? Why You Should Never Believe Health Headlines":
Don't believe the the hype! That's the cardinal rule to obey when reading health news. "Breakthroughs" and "cures" are rare, and should always be viewed with caution and skepticism.This week was a great example. Last Sunday, the New York Times, the major networks, and a host of other media outlets (including this one) reported on a paper in Nature Medicine about the discovery of a novel and potentially significant pathway linking red meat to heart disease. Briefly, the research suggested that carnitine, which is found naturally in high concentrations in red meat, can lead to atherosclerosis when it is converted by gut bacteria to a chemical called TMAO. Almost immediately I received a lot of comment from experts who raised serious questions about the research. Then today, a separate study was published with an entirely different perspective on carnitine. Although the two studies don't directly contradict each other, they suggest that the real truth about carnitine is likely to be quite complex and will never be adequately summarized in a headline.
...First let me address the problems with the original paper. As several critics were quick to point out (and as I should have reported originally), a key assumption of the original paper- that eating red meat can in fact cause atherosclerosis- has never been demonstrated with scientific rigor and is actually the subject of intense controversy. I asked Stanley Hazen, the senior author of the Nature Medicine paper, about this. He admitted that "in truth, we did not examine this," though he pointed to several lines of evidence linking carnitine to heart disease.
Second, the Nature Medicine paper did not mention a key fact that could dramatically change one's interpretation of the research. The Nature Medicine paper claims that red meat leads to atherosclerosis through the conversion of carnitine to TMAO. But the paper didn't mention that fish- which almost everyone thinks of as being heart healthy- often have very high levels of TMAO. I asked Hazen about this and whether he had measured TMAO levels in fish. He didn't say whether he had measured TMAO in fish, but agreed that "some forms of fish that have very high levels of TMAO," such as deep sea fish in Arctic waters, which use TMAO as anti-freeze and "and have extraordinarily high TMAO" levels.
Do read the whole thing at the link.
via @ChrisKresser







Headlines are almost always written by someone other than the author, and are just a tease to get you to read the article.
In truth, you shouldn't rely on getting accurate science news from mass media of any sort. They're usually...inaccurate.
I R A Darth Aggie at April 15, 2013 10:57 AM
I've done pretty well to extending that caveat to "anything written by a journalist " is suspect.
Most of them are little magpies repeating what some so called expert told them with no understanding of either the science or the issues behind the story.
(This is why they are a journalist, because they have neither the math skills or the background to actually analyze data.)
Trust no one, and if you are not a subject matter expert in the field being reported on, find at least three people who can explain possible flaws in both the study, and the underlying data, before coming to any conclusions at all.
Isab at April 15, 2013 4:32 PM
One that I heard about today was Microsoft Research: The probabilistic terrestrial carbon model: towards actionable predictions.
So supposedly MS is involved in building a scientifically neutral research effort into climate change. A good idea, yes?
But then you have MS involved in Common Core which is most decidedly not neutral.
So which side do you trust?
Jim P. at April 15, 2013 7:49 PM
The short answer is neither. No one should be trusted either because of their credentials or who is paying them, or not paying them.
There is a two part or three part test which most lay people cannot do themselves, and that is why real science is not only peer reviewed, but peer reviewed on every level of understanding such as not only the validity of the conclusions, but the validity of the assumptions, and why a particular data set was selected for a study, and was the sample large enough to begin with?
For example, the basic assumption behind global warming is that increased atmospheric CO2 produced by mans industrial activities is causing the earth to warm, and there is a tipping point that will occur when the concentration gets high enough.
However if Co2 is a lagging indicator, as many suspect, increased CO2 may be caused by warming.
When people are desperate to push a particular theory because it has political rewards, all consensus is suspect.
Isab at April 15, 2013 8:12 PM
Leave a comment