NY Times ‘Well’ section continues to mislead readers — this time on facial exercises that ‘may make you look 3 years younger’

The New York Times “Well” section is once again … unwell.  

Before exercises (left) • At 8 weeks (center) • After 20 weeks (right) | Northwestern University

This time, it’s an article titled “Facial Exercises May Make You Look 3 Years Younger,” which employs a clickbait headline and the veneer of scientific legitimacy to hype a small study of highly questionable significance. It quotes two sources –both co-authors of the study–and buries the significant limitations toward the end of the story; limitations which render the provocative headline unsubstantiated. 

This is a recurring issue that we’ve documented many times before:

• NYT blog post on sugar’s effect on heart disease markers was a little too short and sweet
• Walnuts for weight loss? Times makes huge leap based on brain scans of nine people
• Conflicts of interest abound in NYT post on Transcendental Meditation
• … and many more examples dating back to 2013.

Again and again we find the Times’ wellness stories falling down on crucial responsibilities: quoting biased sources rather than independent ones, not scrutinizing the evidence or citing alternative studies, erroneously implying cause-and-effect from observational studies, and not raising conflict of interest issues.

Again and again we find that readers of their wellness stories send in comments that are often more informative and germane regarding the topic at hand than the article itself. Here are some examples from the facial exercises article:

What the study found and how

The Northwestern University study is entitled, “Association of Facial Exercises With the Appearance of Aging,” and was recently published in JAMA Dermatology. Here are the basics:

• 16 middle-aged women (11 dropped out), selected by the authors, were trained in 32 facial exercises designed by one of the co-authors
• Participants were photographed before, 8 weeks in, and at the end of 20 weeks of doing the exercises for 30 minutes every day
• Two dermatologists used a scale (Merz-Carruthers Facial Aging Photoscale) which uses roughly 20 facial features (mostly wrinkles and cheek features) to estimate the ages of the women. They guessed the “before” ages of the women to be 50.8 and the after 20 weeks” ages to be 48.1
• A participant satisfaction questionnaire found the women were “highly satisfied” with changes in 18 out of 20 facial features

“But what does that ‘three years younger’ really mean in real-world terms for real patients?” asked UNC-Chapel Hill dermatologist Ade Adamson, MD.

Ade Adamson MD, MPP

“Both this pilot study, and the media coverage of it, have problems. The study has multiple limitations including: non-objective measures of the outcomes, the very small size, the drop-out rate, no control group, and that they didn’t really account for the multiple comparisons — about 20 I think — which means their results could easily be attributed to chance alone. As for the media coverage it was quite variable but, for the most part, was way too optimistic given the limitations of the study.” (Learn more about the problems with multiple comparisons in research studies here.)

This is a study in which the two primary outcomes — “Did the patients appear younger after the exercises?” and “Were the patients pleased?” — rely heavily on subjective interpretations. The women who invested substantial time in the program may well want to believe that it’s working for them. Could this be reflected in their responses to the survey questions? The only way to assess this effect would be through a control group of women who underwent a comparable sham intervention.

Also, having 11 subjects drop out can introduce bias: Does this leave only the women who perceived that the exercises were working for them (skewing the results in favor of the intervention)? Does it leave us with a highly motivated (and highly select) subgroup that is even more unlikely to be representative of the general public?

Let’s be clear: Some of these limitations were mentioned in The Times ‘Well’ article. But they were buried near the end of an article that didn’t seek critical opinions from people unaffiliated with the study.

I agree with the comment of Ellen from Williamsburg (above) who felt the piece read like an advertisement and I would add this: If it reads like promotional fluff, even if you do include limitations, but don’t make it clear how those limitations render your headline worthless, haven’t you already done your readers more harm than good?

Perhaps it’s not surprising that stories directed at ‘improving’ our faces are quite common. And also quite suspect. Here are a few more we’ve written about: “resting bitch face” and “are your lips attractive?“