MIT Sloan professor develops model to track and predict U.S. obesity trends

CAMBRIDGE, Mass., December 15, 2014—Obesity is an alarming problem around the world, affecting people of all different ages, genders and ethnic groups. It’s a major driver of healthcare costs and increased mortality rates. In light of the need to better understand and manage this crisis, MIT Sloan School of Management Visiting Prof. Hazhir Rahmandad used systems dynamics to create a model to track and predict U.S. adult obesity trends. His findings were published in the American Journal of Public Health.

In the study, Rahmandad and his colleagues looked at the energy imbalance gap (EIG), which is an individual’s average daily excess energy intake minus the total daily energy expenditure. The EIG controls the speed of change in body mass and “is at the core of understanding obesity,” says Rahmandad, noting that the energy gap is largely caused by eating habits. “It’s similar to the gas pedal in a car. If you push the pedal, the gap is positive and obesity trends speed up. If you push the brake, then the gap becomes negative and we would have fewer obese people. A zero gap is like cruise control with a steady obesity prevalence.”

Developing one of the first applications of systems dynamics to study EIG, the researchers’ quantified the gap for three adult population groups in the U.S. over four decades: Non-Hispanic Whites, Non-Hispanic Blacks, and Mexican-Americans. The model showed significant differences among the three population groups as well as between genders. “This is an important step toward the design of obesity prevention and intervention programs targeted at specific population groups,” says Rahmandad.

For the largest population group of Non-Hispanic Whites, they found that the EIG has been positive over the last four decades. In other words, this group has consistently been gaining weight, a trend reflected in the current obesity epidemic. However, the model shows that the gap is shrinking. Once the gap reaches zero, the obesity rate will have stabilized. “This doesn’t mean that this group has solved the problem of obesity, but that the problem is no longer getting worse,” explains Rahmandad.

The model tells a different story for Non-Hispanic Blacks and Mexican-Americans. For Non-Hispanic Blacks, the rate of obesity is growing and the EIG is still not close to zero. “The gap is positive and very significant,” observes Rahmandad, noting that the model shows the obesity trend continuing for at least another decade before the gap begins to close. “This signals that in the future we will see more challenges with obesity in the African-American community.”

The fastest growth in the gap, according to the study, is among Mexican-Americans. Not only is obesity worsening, but the speed of change in obesity is worsening too. “This population group needs much more attention to turn the tide of the obesity epidemic,” he says.

As for gender, the most notable finding pertains to African-American females, who generally have had higher EIGs than African-American males. “This means that obesity has been getting worse for females faster than for males,” observes Rahmandad. He adds that Mexican-Americans show a similar gender gap in the recent years, while for the Non-Hispanic White population, energy gaps are bigger for males than females, thus obesity trends are growing faster for men than women.

“We know obesity is an epidemic so these findings aren’t that surprising, however what is remarkable is the differences among ethnicities,” he says. “Our research supports the need for policies targeting specific population groups rather than blanket approaches.”

He adds that the individual level model of growth and body weight dynamics can also be used to understand malnutrition in places like Africa. “We’re currently working with the Gates Foundation to develop programs to treat malnourished children. By modeling how small children’s bodies respond to different nutritional interventions, we can come up with approaches that are more impactful in the field.”

To read the American Journal of Public Health article, “Modeling US Adult Obesity Trends: A System Dynamics Model for Estimating Energy Imbalance Gap,” please visit:

To read more about the individual level model of growth dynamics, please visit Prof. Rahmandad’s paper, “Human growth and body weight dynamics: an integrative systems model.”

Hazhir Rahmandad is Visiting Associate Professor at the MIT Sloan School of Management