Last night I wrote about a new CDC study showing a 43 percent drop in obesity rates among 2-5 year-olds. It seemed inexplicably large to me, especially because no other age group showed any decline at all. Today, Zachary Goldfarb helpfully 
publishes a bit more of the data, and I’ve extracted two lines from his chart. This only deepens the mystery.
As you can see, there’s a fair amount of noise in the chart, and it’s possible that this explains the whole thing. But if we take the data seriously, you can see something even more dramatic than a 43 percent drop over a decade. Between 2003-04 and 2005-06, there’s a 25 percent drop. That’s a gigantic decline over the space of two years.
But there’s more. If there’s anything real going on here, you’d expect to see some kind of correlation between 2-5 year-olds and 6-11 year-olds with a time lag of a few years. But I don’t see anything. The 2005-06 cohort of 2-5 year-olds is noticeably less obese, but the 2007-12 cohort of 6-11 year-olds shows barely any change at all.
So this whole thing is very strange. As I said, it’s possible that noise is responsible for a lot of this. But even if there really is something going on, it doesn’t seem to be having any impact at all once children get a few years older. That’s both strange and disappointing. I wouldn’t expect miracles, but the whole point of obesity interventions in small children is that it prevents a lifetime of bad habits. As the New York Times put it, “New evidence has shown that obesity takes hold young: Children who are overweight or obese at 3 to 5 years old are five times as likely to be overweight or obese as adults.” But if that’s true, it sure isn’t showing up in the data. As near as I can tell, reducing obesity among 2-5 year-olds has precisely zero effect on obesity later in childhood.
