Friday, 14 June 2013

Close encounters of the third variable

From yesterday's Metro...

Chemical in plastics 'causes obesity in many young girls'

Young girls exposed to a chemical found in a variety of plastics could be up to five times greater risk of obesity, a study suggests.

Sceptical minds will have noticed that this epidemiological finding relates only to girls, not boys—always an indicator of a data dredge. Notwithstanding the unlikelihood of a single chemical 'causing' obesity, what kind of bizarre substance would cause obesity in females but not males?

Those aged nine to 12 with 'extremely high' levels of bisphenol-A in their urine were most likely to be grossly overweight, the Plos One journal reports.

What could possibly explain this phenomenon?

The substance can be found in drinks bottles, food cans and till receipts.

So people who drink lots of drinks and eat lots of food are more likely to be obese? Who'd a thunk it?*

(* "Surely the researchers adjusted their findings to take into account diet?!", I hear you cry. Not really, your honour. They adjusted for the amount of 'junk food' and vegetables eaten because these are seen as good/bad by public health boneheads, but there was no accounting for the number of calories consumed. You can read the study here.)

No comments: