Yes, another new member for the coffee club. Nary a week goes by without a new observational study about the benefits or risks of coffee consumption. And almost always, many news organizations fail to explain the limitations of the observational studies.
Today’s example, though, gives us one of the better cross-media comparisons of people who clearly tried to get it right – with varying degrees of success.
What: A lot of news coverage apparently generated by a news release from the American Academy of Neurology about a paper entitled, “Greater Consumption of Coffee is Associated with Reduced Odds of Multiple Sclerosis.”
When: The news release was about a paper that won’t even be presented until two months from now at the Academy’s annual meeting.
Note: The researchers use the term “associated with.” And in the abstract, you’ll see the term association (in various forms) a total of eight times.
Note further: The Academy’s news release immediately introduces language of causation – not association – in its headline, “Can Coffee Reduce Your Risk of MS?” As Jon Stewart of The Daily Show has pointed out, some writers/journalists feel you can get away with saying almost anything as long as you follow it with a question mark. Question mark journalism, fostered by question mark public relations. Examples of headlines I found, practicing question mark journalism:
Here’s a further sampling:
The Los Angeles Times headlined it, “Researchers find new reason to drink coffee: It may reduce risk of MS.” And the story’s first sentence:
“Drink up, coffee lovers: Neurologists say a healthy appetite for coffee may reduce your risk of developing multiple sclerosis.”
Hmm. Was the observational study convincing enough to be a call to action to “drink up”? If you keep reading, it’s not until 80% deep down into the story that we’re finally told:
“The results don’t prove that coffee was responsible for reducing the risk of MS. Nor did the study explain why coffee might offer some protection against the disease.”
HealthDay, in comparison, put the caveats way up high – in the 3rd sentence of the story:
“Researchers stressed that the findings do not prove that coffee fights MS.”
Better yet, the story continued….
“This doesn’t mean we should be recommending rampant coffee drinking,” said lead researcher Dr. Ellen Mowry, an assistant professor of neurology at Johns Hopkins University in Baltimore. … She is scheduled to present the findings — which are considered preliminary until published in a peer-reviewed medical journal — in April, at the annual meeting of the American Academy of Neurology in Washington, D.C. The academy released the results Thursday, ahead of the conference.
That certainly counters the LA Times’ lead line of “Drink up…”
Medscape included this independent perspective:
Marie B. D’hooghe, MD, from the National Multiple Sclerosis Center, Melsbroek, Belgium, called the data “intriguing” and said she is “looking forward to reading the full paper.” “We always need to be cautious,” she added. “Bias and errors cannot be excluded.
Ah, yes, let’s wait for the chance to read the full paper, rather than jumping to conclusions about an abstract and a news release published two months prior to the abstract even being publicly presented and open for discussion in a scientific conference.
Leaving journalism per se, but with a glimpse of how a big name hospital website might deal with advice to its patients…a Cleveland Clinic website piece ends:
“So if you love a cup of coffee in the morning and a few over the course of the day, don’t feel guilty. Drink up!”
First, did the Cleveland Clinic web writers miss what the latest study’s lead researcher told HealthDay? Here it is again: “This doesn’t mean we should be recommending rampant coffee drinking.” There was no call to action in this study.
Second, I can’t tell you how many times I’ve seen that tired old “don’t feel guilty” line, but I can share with you some of what I found in just a quick web search of media messages about coffee this morning:
So, MS now joins the coffee club, in which association ≠ causation, along with: