Modeling Social Contagion of Obesity

Obesity is considered to be a result of several non-communicable factors, such as inactive lifestyle, environment, health condition, age, etc. However, recent studies found that the obesity epidemic can be considered a social contagion process, or “spread” from person to person, like influenza. For example, if your friends are obese, your chance of becoming obese becomes higher. Prior studies also show that obesity risk is greater among individuals whose mother is obese during the pregnancy. For example, if your mother is obese, then your chance of becoming obese is greater.

The objective of this work is to investigate the impact of both contagious and non-contagious factors in order to predict the future prevalence of obesity. We developed a simple mathematical model assuming obesity risk is dependent on the (1) non-communicable risk, (2) mothers’ condition of obesity and (3) obesity prevalence (social or peer effect). Furthermore, we assume the individual susceptibility to obesity risk is dependent on genotype.

Fitting our model to the empirical data set, approximately 27.6% of our population will be considered obese in the future. Model simulations from our mathematical model enabled us to predict changes in the prevalence of obesity when any intervention to the obesity epidemic is implemented. Results suggest that if we can reduce non-communicable risk, mother to child transmission risk, and peer effect by 50% from the current values, then the saturated prevalence reduces to 26.77%, 24.69%, and 22.58%, respectively.

In conclusion, our study incorporated both non-contagious and contagious risk factors into a mathematical model for predicting the prevalence of obesity at the population-level. Moreover, our results demonstrate how mathematical models can be a useful tool for assessing the effect of intervention programs before they are carried out.