Here's a great example of how weak findings in small, underpowered studies -- findings which should be at most viewed as hypothesis-generating -- become transmuted into Serious Studies With Important Implications when the lay press give them too much credit. In this case, the Los Angeles Times, Huffington Post, and CNN took the bait after the American Heart Association gave the story a boost.

The facts: Jessica Kepplinger of the University of Technology in Dresden, Germany, got up at the American Stroke Association's International Stroke Conference to report a small observational series of 56 people, all of whom had already had strokes (selection bias alert). Upon testing with polysomnography and MRIs, 91% had obstructive sleep apnea. Comparing those who did vs. those who didn't have sleep apnea (all of whom were stroke patients, remember), those with OSA were more likely to also have additional white matter lesions. Worse sleep apnea was correlated with more white matter lesions.

The LA Times blog's conclusion (borrowed from the study's press release):

Having severe sleep apnea may make people more at risk for silent strokes and small brain lesions, researchers found.

Huffington Post nodded sagely, and agreed:

A new study suggests that the sleep disorder is also linked with small brain lesions and a symptomless form of stroke, known as silent stroke.

CNN got it almost right (an MD, Lisa Shives wrote this one):

It's a small study that adds further evidence to what most sleep experts already know - that obstructive sleep apnea (OSA) is linked with a high risk of having silent strokes.

Of course, controlling for obesity, high blood pressure, diabetes, activity levels, smoking, etc. was not mentioned -- and certainly couldn't have been; the sample size is too small (there were only 5 patients without sleep apnea in the cohort). Sleep apnea and strokes share many comorbidities; their relationship is complex and unknown, and can't be elucidated by a small, highly selected series of patients. Sleep apnea and its severity track closely with increasing severity of other cardiovascular risk factors, as does stroke. And incidentally discovered white matter disease (what they call "silent strokes" here) is of course exceedingly common in people with cardiovascular risk factors or overt disease.

This is an uncontrolled case series whose natural lifecycle is to buzz around its conference, then disappear, with its findings maybe stimulating a more-believable longitudinal cohort later. But because it could be spun to sound scary, the press elevated it to science. That's not news either, but it's a nice case study in how health "news" gets made.

Gary Schwitzer's blog, HealthNewsReview.org, is a great watchdog for reining in hype among science writers, educating them on how not to get taken in by poor-quality studies, and calling them out (including the major news media) when they don't -- which is often. Check it out for a perspective-changing view of health reporting.

"Sleep apnea may make people more prone to silent strokes," reporting on the American Stroke Assn.'s International Stroke Conference, in the Los Angeles Times' Booster Shots blog.

American Heart Association's original press release

Lay press eats helping of hype on sleep apnea…

Comments on this post are for paid subscribers