BOSTON—Becoming obese isn’t like catching a cold, but a handful of research groups are now trying to model obesity in a population by treating it like a “social contagion” that spreads among people through their interactions. Statistician David Allison and information scientist Keisuke Ejima of the University of Alabama at Birmingham and mathematician Diana Thomas of the United States Military Academy in West Point, New York, have been working on refining such obesity models. They explained their progress in a session today at the annual meeting of AAAS, which publishes Science. We caught up with them to learn more. This interview has been edited for clarity and length.
Q: How did this idea of modeling obesity as a contagion come about?
K. E.: The first paper was published in 2007 by [Nicholas] Christakis and [James] Fowler. … They found that the obesity can be transmissible through social networks. During my Ph.D. course, I was working on mathematical modeling of infectious disease. When I read their paper, I thought, “I can apply my mathematical techniques to describe the obesity epidemic model.”
Q: In what way is obesity a “social contagion”?
D. T.: If you’re somebody who loves to go to the gym and loves to eat healthy, it’s unlikely that you’re going to draw in a circle of friends that love to smoke cigarettes and love to eat at fast food restaurants. You’re going to kind of surround yourself and emulate the behavior of a cluster around you. There is some experimental evidence for that.
D. A.: I want to just be clear that actually there is a hypothesis out there, supported by some data, that certain microbes—particularly adenoviruses—do contribute to obesity, and in that sense, it would be contagious in the more literal sense of spreading viruses from one person to another. So we don’t want to dismiss that.
Q: What factors go into your model, and what does it predict?
D. A.: What we’re saying is a population that had these characteristics would start to have these obesity rates. [The model is] a series of interrelated equations, and they include several factors. One is a genetic factor. … We include another term [in the equation] for nongenetic transmission of obesity from mother to offspring, because there’s some evidence that the more obese the mother is, the more obese the offspring will be, irrespective of what she transmits genetically. Then another term deals with the overall population prevalence of obesity. … If the prevalence is higher, you run into more people who are obese, and you are more likely to pick it up from them.
Q: What kinds of things can these models tell us?
D. T.: All the models are indicating that over time, obesity rates will level off … around 35% to 40% [prevalence]. It is believed that obesity rates are leveling off because of the great work we’ve been doing in intervention. But that may not be the case. All of these models are indicating that a plateau is just a natural evolution of the system.
Q: Can the models tell us how to reduce obesity?
D. A.: Most of these models are, so far, fairly primitive. Our model only includes a single gene—we know that that’s not true. [And] we know that people have different numbers of offspring as a function of their BMI [body mass index, which measures body fat based on height and weight]. Overweight people tend to have more children than nonoverweight people. Only when a model has built in these many factors can we make good predictions and talk about manipulation. … We’re not ready to make public health recommendations.
Q: What questions would you like the models to answer?
D. A.: Should we intervene with people who are already at the high end of the BMI continuum, because maybe they’ll benefit the most or need it the most? Or should we intervene more broadly because we’ll have a bigger bang for our buck with a bigger population? These kinds of models can help inform those decisions.
Q: Are you concerned that describing obesity as a contagion might increase the stigma around it?
D. T.: I have a lost a tremendous amount of weight … and when I was classified with obesity, I know people looked at me and said, “Well she’s just fat because she’s lazy and doesn’t want to take care of herself.” What these types of models tell us is there’s more than one reason why someone might end up in that state, and not all of those reasons are in their personal control.
D. A.: Obesity is a highly stigmatized condition already. There’s ample data showing that obese people are discriminated against in the job market, in the housing market, [and] with respect to wages. We don’t want to stigmatize anybody further. … That’s where careful messaging has to get out, that this is something that we’re all in together.
Check out our full coverage of AAAS 2017.