The Worst Kind of Science Clickbait
Last year, I F*%king Love Science, blared a headline sure to pique the interest of any alcohol imbiber: "Beer Hops May Protect Against Liver Disease." IFLS readers were quick to offer their two cents. Many wondered whether specific beer styles or even a beer's International Bittering Units (IBU) rating could be linked to liver protection, but others, like JD Miller, rightly expressed skepticism.
"Jumping from the results of a single study of female mice (what's the sample size? My university doesn't have sub to this journal) to the broad conclusion that "drinking hops protects against (human) liver disease" is not only bad science, it's reckless and potentially dangerous clickbait which ifls should know better than to indulge in."
IFLS isn't exactly a bastion of solid science reporting, so it's understandable (although maddening) that they do partake in this sort of clickbait. It's less understandable and far more maddening that more noted news venues also publish it.
"Diet Soda May Alter Our Gut Microbes And Raise The Risk Of Diabetes," NPR reported in 2014. The impetus for that headline was a study in which researchers fed zero-calorie sweeteners to mice over eleven weeks and observed the effects. They then fed seven human subjects an amount of saccharin (an artificial sweetener) equal to drinking 40 cans worth of diet soda per day for a week and observed similar effects. The study drew widespread criticism, but that skepticism was mostly left out of NPR's story.
Both of these stories, along with their hyped headlines, were inspired by studies on rodents, and they're not alone. Dozens, perhaps hundreds, of stories relating studies on rodents to humans appear each year. The media loves them, and readers greedily click them by the tens of thousands. The problem is that these stories are not newsworthy.
Rodent studies are by definition preliminary, offering only the most tenuous hint that their respective findings will bear out in humans. Just a third of animal studies published in the most prestigious journal replicate to humans. It's likely that the rate is much lower for less reputable journals. Moreover, when we look at how studies on animals replicate to humans in stroke and cancer research, the rate is much worse. Less than one in ten results carry over to humans.
Science is often counterintuitive, but in this case, our intuitive senses serve us well: Mice are not humans. Sure, we may share a significant number of genes, but as I reported last year, the way mice are "wired" and how they live are drastically different to humans. These real physiological differences contribute to the dismal rates of replication.
I encourage other media outlets to take this reality into account when choosing whether or not to report on rodent studies. I also encourage IFLS to change the headline of the story I referenced earlier, from "Beer Hops May Protect Against Liver Disease" to "Beer Hops May Protect Against Liver Disease in Mice, But There's Less Than a One in Ten Chance That It Will Protect Against Liver Disease in Humans."
Admittedly, that's not quite as share-worthy as the original.