October 22, 2018
Kevin Lomangino is the managing editor of HealthNewsReview.org. He tweets as @KLomangino.
Here’s another example in our occasional series documenting how the New York Times’ ‘Well’ section is often unwell.
This time it’s a brief story claiming that “Coffee may tame the redness of rosacea.”
Let’s start with the fact that this study was observational and had many limitations, which we’ll explain below. (Read our primer on observational studies to learn more.)
It is journalistic malpractice not to mention at least some of those issues in a news story about this type of research.
But you will find no such discussion in the skimpy 220 words that the Times devoted to the topic. Not a single word.
That’s a stark contrast with the journal article upon which the news story is based, which considered those limitations important enough to warrant 230 words — more than the entirety of the Times’ coverage.
That’s why context is so important for a study like this, said dermatologist Ade Adamson, MD, an assistant professor in the Department of Internal Medicine at Dell Medical School at the University of Texas at Austin, who was not involved with the study.
“We don’t know why certain people get rosacea, but family history and skin type are important factors which were not controlled for (the authors admit this),” he wrote in an email. He added that “health-seeking behavior” — meaning how often someone has contact with doctors and the health care system — is another factor that probably plays into this. The more often you see doctors, he explained, the more likely you are to come away from those encounters with a diagnosis of rosacea. “This was also not controlled for,” he said.
Adamson noted that dermatologists generally tell patients with rosacea that coffee makes the condition worse. But the story didn’t mention this, and Times readers will likely draw the opposite conclusion — even though the study itself looked only at coffee’s possible role in the prevention of rosacea, and not at whether coffee has any role in the treatment of people who already have the condition. Those are two separate issues that are conflated in the Times coverage.
“To say that coffee ‘may tame the redness,’ as in the NY Times, is a huge leap,” said Adamson. “You need prospective randomized trials to show that. We still don’t know why people get rosacea, but it’s probably not coffee.”
We’ve said it many times before and we’ll say it again: Stories like this one are peddling a certainty that simply doesn’t exist. It hooks readers with the claim that coffee “may be good for the complexion,” but ignores that this type of study isn’t designed to show such a cause-and-effect relationship.
Times readers deserve better and have asked for better on many occasions. A commenter who identifies herself as Sharon Kahn wrote the following in response to this article:
This was a correlational study, which cannot prove causation, only that, in the authors’ words, an “inverse relationship” was found between the two variables. It is possible that a third, unstudied variable is the real factor. Secondly, please note the limitations which the authors’ acknowledge in the article.
Will the editors ever listen to readers like these?
Disclosure: Ade Adamson, MD, is on the editorial board of JAMA Dermatology where the study that was the basis for the Times story was published. He had no role in the decision to publish the journal article and did not participate in its peer review.