Monday, June 4, 2012

To understand obesity, genetics is essential/irrelevant

Both or the following statements are true:

1. Genetics is a major factor in understanding obesity

2. Genetics has no significant role in obesity

The trick is the framing of the question. If we are talking about the factors that determine why person A is obese while person B isn't, genetics should certainly be considered. If we're talking about the rise in obesity nationally and internationally, genetics has no significant role. This is one of those cases where the level of aggregation can completely change the validity of conclusions but you'll seldom hear the distinction made explicitly and you sometimes won't be able to tell even from context.

Making subtle changes in the way we frame a question can, and often does, make huge differences in both the way we approach the question and the inferences we can draw from it. Unfortunately, these fine points are often overlooked even in technical discussions and almost never make it to the popular press. The result is that intelligent people of good faith are often sucked into pointless pseudo-arguments. Worse yet, people of bad faith can produce dishonest but deceptively persuasive arguments by reframing questions mid-discussion.


  1. I agree with the actual reasoning above, but for the sake of argument: populations have been increasing rapidly, this increase could in principle be genetically non-uniform and therefore related to the rise in obesity nationally and internationally.

    Also epigenetic factors would throw the distinction between the two statements into the toilet.

  2. @Paul: I agree that there are nuances with Mark's argument re: genetics. But it seems unlikely that selection for obesity could work so fast given the much slower rate seen in animals (in terms of generations).

    Epigenetics is a whole other can of worms but still depends on some secular shift in exposure.