My friend Angus is the editor of a story on how B vitamins can cut brain tissue loss in older adults.  Very interesting, and potentially very useful.

But badly edited.

I’m not giving the blame to the reporter – I’m guessing she rewrote the press release.  But it’s the editor’s job to do fact-verification, the “does this make sense?” question when dealing with numbers.  And Angus didn’t.

The headline reads

B vitamins found to halve brain shrinkage in old

and the story repeats the theme

People who had the highest levels of homocysteine at the start of the trial benefited the most from the treatment, with their brains shrinking at half the rate of those on the placebo.

And then there’s the actual numbers:

on average the brains of those taking the vitamin treatment shrank at a rate of 0.76 percent a year, while those taking the dummy pill had an average brain shrinkage of 1.08 percent.

Now, from my perspective, half of 1.08 is 0.59.  Going from 1.08 to 0.76 is more like thirty percent, not fifty percent.  Maybe they are only looking at the pull quote, but we aren’t given those percentages, and the headline doesn’t specify the condition for the best results.

It would have been better to say something like

B vitamins cut brain shrinkage 30% in old

and explain the better results in the story, along with reporting on how much of the population has these high levels of the long-named amino acid.

Angus, m’boy, you tried to pull a fast one, and you got caught.  No matter how much you wish it, or try to use it, we are not yet a post-numerate society.