I have an old friend from college who I’m in touch with via Facebook. He looks great. In fact, he looks like how I’d like to look. After a year of running and lifting I’m noticeably more muscular. I’m 150-155 pounds at 5’8, and according to my Aria I’ve lost about ~3% body fat (I’m not reporting the absolute value because it’s not that accurate, though the readings seem consistent day to day). But, since I’m South Asian and tend toward a “baby face” mien I’m not where I want to be in terms of definition. In contrast my friend is very toned. I asked him how he had changed his physique so much over the years, since in college he was rather “soft” looking. His response? Gay peer pressure.
This is why articles like this in Salon drive me crazy, You should never diet again: The science and genetics of weight loss. It’s excerpted from the book Secrets From the Eating Lab: The Science of Weight Loss, the Myth of Willpower, and Why You Should Never Diet Again. The author, Traci Mann, is a professor at the University of Minnesota, as she proudly notes:
To tease apart the effects of genes from the effects of the shared environment, researchers located identical twins that were raised in separate homes without knowing each other. It may seem surprising that there are enough sets of twins that meet this criteria, but there are. This type of twin research was partly pioneered in the very psychology department in which I work, at the University of Minnesota (coincidentally located in the twin cities of Minneapolis and St. Paul). If you go up to the fifth floor, the walls are covered with photographs of identical twins that were separated at the age of five months (on average) and had been apart for about thirty years before being reunited as adults. The visible similarities are remarkable, as are the many documented behavioral similarities.
The crucial twin study of body weight (which comes from the Swedish Adoption/Twin Study of Aging) included 93 pairs of identical twins raised apart (and 154 pairs of identical twins raised together). Sure enough, the weights of identical twins, whether they were raised together or apart, were highly correlated. That study, along with several others, led scientists to conclude that genes account for 70 percent of the variation in people’s weight. Seventy percent! What is truly remarkable is that this is only slightly lower than the role genes play in height (about 80 percent of the variation). Don’t get me wrong. I’m not saying you can’t influence your weight at all, just that the amount of influence you have is fairly limited, and you’ll generally end up within your genetically determined set weight range.
Many readers will be familiar with twin studies. If you aren’t, read Born that Way by William Wright. The strange thing about this reference to this scientific project is that it’s incongruous to see it in a Left-wing publication like Salon. The more moderate Slate trashed twin studies about four years ago, even though cutting edge genomic methods are now validating their results. Why all the hostility? Mostly it has to do with non-trivial heritabilities for intelligence and personality which come out of this type of research, which are not congenial to a particular sort of Left-wing mentality (see Steven Pinker’s The Blank Slate). I suspect the issue here is that the political valence is not alarming to the readers of Salon, who are very open arguments about “thin privilege”, and the idea that the heavy might be considered another protected class. The analogy then in terms of heritability and a biological basis for this trait is with homosexuality, where the Left favors genetic determinism. The same studies also come with high heritabilities for traits such as weight in developed world populations. The quoted results above are correct, but the implications that the public takes from them are misleading.
I’m pretty sure that Traci Mann knows the technical definition of heritability by the way she writes. That is, the proportion of phenotypic variation that can be explained by genetic variation within the population. The problem is that the general public is going to see “70 percent heritable” and think “70 percent genetic,” when that’s not even wrong. The way the piece is written also misleads in this fashion. One thing to note is that even though height is 80 to 90 percent heritable, the correlation between full siblings for this trait is only 0.50. I suspect this would surprise people since it is such a heritable trait, but that goes to show that a population wide heritability statistic has only modest utility on the individual scale. More importantly by analogy to height the norm of reaction matters a great deal. We know empirically that genetically similar populations vary in mean and variance of weight and height over time and in different environments.
Which goes back to social context. My friend now works out a lot and watches what he eats. It’s something that he does every day, and something that is enabled by the social environment in which he is embedded (see this music video about “Mean Gays”). Weight varies in the United States by class and region, and from what I recall this remains after you control for demographic variables. What this probably means is that it takes a village to sustain weight loss. So in a way those who argue that dieting is useless are correct. But they’re being misleading when they imply that your weight is ultimately “genetic.” It’s social.
Update: A friend emailed me and pointed out figure 2 in a recent Nature paper. It gets at what’s producing the heritability statistic: