Shut up, Grandma. You too, AP.
Here's a modest step we can take, like, today toward improving the public reporting of science: Ban all ledes that use fictional grannies to put an anecdotal face on statistics. As in this:
A few extra pounds not risky, study says
Obesity, however, still tied to greater threat of cancer, heart disease
Being 25 pounds overweight doesn't appear to raise your risk of dying from cancer or heart disease, says a new government study that seems to vindicate Grandma's claim that a few extra pounds won't kill you.
Let's touch on the desk-induced offenses first. The study doesn't say "a few extra pounds [are] not risky"; it says that for "overweight" people, meaning those between normal and obese in body-mass index, there's no increased risk of death from some diseases. That's not the same thing as "not risky," in that death isn't the only risk one faces in the great health spectrum, and it overlooks the finding that overweight is associated with a significantly higher risk of death from other diseases. (If you're not up for paying the AMA 12 bucks or whatever it costs to download the article, this observation is actually in the AP story.) So the hed's both off base and false.
Now, from the text on hand, we're not going to settle the question of how regularly AP writers read the original article before writing (unfailingly, often, sometimes, whenever). But it's worth noting -- heck, the abstract is free -- that the study isn't about "being 25 pounds overweight," and it doesn't measure the impact or whatever folk wisdom your grannies did or didn't dispense. It's about whether falling in a particular BMI range is associated with higher likelihood of death from a couple of sets of causes. Twenty-five pounds, conveniently, will often take you from the top of the "normal" band to the top of the "overweight" band," but that doesn't make it a magical safety figure.
Doesn't mean it's a bad study. Does mean we need to not make something of it that it isn't. And if grandma complains, kick her in the knee.
A few extra pounds not risky, study says
Obesity, however, still tied to greater threat of cancer, heart disease
Being 25 pounds overweight doesn't appear to raise your risk of dying from cancer or heart disease, says a new government study that seems to vindicate Grandma's claim that a few extra pounds won't kill you.
Let's touch on the desk-induced offenses first. The study doesn't say "a few extra pounds [are] not risky"; it says that for "overweight" people, meaning those between normal and obese in body-mass index, there's no increased risk of death from some diseases. That's not the same thing as "not risky," in that death isn't the only risk one faces in the great health spectrum, and it overlooks the finding that overweight is associated with a significantly higher risk of death from other diseases. (If you're not up for paying the AMA 12 bucks or whatever it costs to download the article, this observation is actually in the AP story.) So the hed's both off base and false.
Now, from the text on hand, we're not going to settle the question of how regularly AP writers read the original article before writing (unfailingly, often, sometimes, whenever). But it's worth noting -- heck, the abstract is free -- that the study isn't about "being 25 pounds overweight," and it doesn't measure the impact or whatever folk wisdom your grannies did or didn't dispense. It's about whether falling in a particular BMI range is associated with higher likelihood of death from a couple of sets of causes. Twenty-five pounds, conveniently, will often take you from the top of the "normal" band to the top of the "overweight" band," but that doesn't make it a magical safety figure.
Doesn't mean it's a bad study. Does mean we need to not make something of it that it isn't. And if grandma complains, kick her in the knee.
posted by fev at 10:51 PM
0 Comments:
Post a Comment
<< Home