Hacker News Reader

Comment by awolf | Hacker News Reader

>In other words, according to this study, obesity doesn’t cause diabetes: sugar does.

Sigh. No, according to this study, there is a stronger correlation between sugar and diabetes than between obesity and diabetes.

>Each 150 kilocalories/person/day increase in total calorie availability related to a 0.1 percent rise in diabetes prevalence (not significant), whereas a 150 kilocalories/person/day rise in sugar availability (one 12-ounce can of soft drink) was associated with a 1.1 percent rise in diabetes prevalence.

Much better! Lead with that next time, medical research reporter.

All in all, this seems like a pretty good study that controls for a lot of confounding variables that a lot of diet related studies completely neglect. I just wish the reporting here, and in general in all medical research reporting, wasn't so sloppy. The goal should be to spread awareness of new findings, not to spread misconceptions.

View on HN

"The study controlled for poverty, urbanization, aging, obesity and physical activity. It controlled for other foods and total calories. In short, it controlled for everything controllable, and it satisfied the longstanding “Bradford Hill” criteria for what’s called medical inference of causation by linking dose (the more sugar that’s available, the more occurrences of diabetes); duration (if sugar is available longer, the prevalence of diabetes increases); directionality (not only does diabetes increase with more sugar, it decreases with less sugar); and precedence (diabetics don’t start consuming more sugar; people who consume more sugar are more likely to become diabetics)."

As far as observational studies, (of which I am a huge skeptic), this one is pretty well controlled for. It's not like they studied 7 nations, or 11 nations. They studied 175 nations. That allows for a whole lot of controllability for various factors.

Not on that list: family history/genetics. I believe that has already been identified as a significant risk factor for diabetes. So it's not just the sugar.

It controls for etnic variables, by using trends instead of comparing raw numbers and some more advanced techniques.

It can not control for family because it was population-wide, not person-wide (and yes, that is a flaw that the authors acknoledge).

This study doesn't do the impossible and prove causation, but it goes beyond mere correlation in several key ways; follow Uhhrrr's link to http://epianalysis.wordpress.com/2013/02/27/sugardiabetes/ . It's not something you can casually dismiss with a wave of the "correlation is not causation" mantra.

This is commentary, not reporting: "Opnionator - Exclusive online commentary from The Times". It's little more than a guy on a soap box or one of the talking heads on TV. Unlike the article, the study itself does appear to be legitimate however.

I would be interested to see if the arguments about sugar we're having now are similar to the arguments about cigarettes in the 60's.

Reply on news.ycombinator.com