Wednesday, October 23, 2013

On Getting Fat

A few months ago, FLG watched the video below.  It shocked him because it was 1) contrary to what he thought/been told and 2) seemed so compelling.  (FLG is posting a condensed version, but longer versions are all over YouTube.)

FLG had a couple of questions.  First, what about Asians?  Don't Asian diets have a high rice component?  Second, how could the medical establishment have been so far off?

Turns out Asian diets are lower in sugar, which may be a catalyst for all carbohydrates becoming problematic.  Also, Asians have lower BMIs, but higher body fat percentages.  So, their level of leanness might be overstated in the data.

As far as the medical establishment being so far off, the best explanation is a donut.   Basically, if you see that people who eat more donuts have heart attacks and you are looking for a casual dietary factor, it could either be the fat or the carbs.  Back in the 50s, the fat hypothesis caught on and the carb hypothesis was ignored.

But there is a larger issue.  FLG remembers sitting in his econometrics class when the professor, in a bit of a digression, said that a lot of medical research has problems with the statistical analysis, particularly omitted variable bias.  But there is even a larger issue with correlation not equaling causation.  One example that comes to mind is the claim that flossing prevents heart disease.  FLG has seen several attempts to explain a causal link, but the most likely explanation, at least in FLG's opinion, is that people who floss are fundamentally different than non-flossers.  To use a Smithian term, they are more prudent.

As similar thing applies to donuts.  Maybe it's not the fat or the carbs, per se, but that people who eat lots of donuts are less concerned about their health than people who don't eat as many.  This creates a whole host of potential omitted variable problems

As FLG started looking at some of these studies himself, it became clear that Taubes was correct.  A lot of public health recommendations are based on very constrained or inconclusive findings extrapolated to far larger recommendations based upon assumptions and an overly precautionary stance by health experts, who seemed rather oblivious to potential unintended consequences.   It also turns out that recent clinical trials seem to support the claim that carb restriction appears to be the most effective.

This led FLG to reduce his sugar, flour, rice, and potato intake about a month or so ago, and sure enough he's been losing weight pretty steadily.  He still eats pizza and has beer on the weekends though.


The Maximum Leader said...

You should read his (Taubes') book. I have it. I recommended it to a few people and gifted it to my buddy Kevin. (Who wrote extensively about it on his blog. I also would recommend the Skeptic Magazine podcast with Taubes:

All very interesting.

Anonymous said...

My periodontist claims that it's oozings and secretions from the plaque bacteria which inflame the arteries. Why not? dave.s.

FLG said...


I've heard that before. My take is whether, common sense-wise, does it seem more likely that bacteria in your mouth causes, through a long chain of causation, artery inflammation OR that people who floss just so happen to be more prudent with their health overall, and this leads to better heart health?

I'm not saying that the bacteria theory isn't possible, but it's better to go with Occam's Razor.

Creative Commons License
This work is licensed under a Creative Commons Attribution-No Derivative Works 3.0 United States License.