This morning NPR had a story about this study, which followed high-school grads to age 40. It uses “growth mixture models” (I’m not sure what these are) to identify two latent classes: one of “normative,” gradual growth in weight from normal weight at high school graduation to higher weight at 35 or 40, the other of “persistent overweight,” i.e., being overweight at high school graduation and staying overweight. There are important differences in health at age 40, which I don’t think is that surprising (though worth demonstrating). But there are also differences in social outcomes, including having “ever had a partner” (romantic, I assume); welfare receipt; and not having pursued education after high school, all of them more likely among the persistently overweight group.
The paper also, though, demonstrates that low childhood SES is a significant predictor of overweight at high school graduation, as is (independently) high school GPA. Read in this way, it strikes me that we ought not understand persistent obesity as a biological cause of social outcomes, as the NPR story (and particularly Kelly Brownell’s commentary therein) suggests, but rather as a mediator between childhood SES and adult SES.