Behavioral dynamics of social propagation for weight loss: Evidence from a randomized controlled trial in a low-income Appalachian community

Leveraging social propagation of health interventions for disease prevention is a public health holy grail. However, the dynamics of social causal induction have not been well studied. We present evidence from a randomized trial, demonstrating social-behaviorally “infectious” risk factor dynamics for obesity and diabetes. Specifically, we present the Social R 0 of weight loss propagation, which is itself adapted from the widely used infectious disease R 0 . Via the trial design to infer causality, we calculated the Social R 0 across multiple time points for weight change propagation ( social-R 0 = 1.2, p < 0.01). We further show that health propagation of an intervention program can attain epidemic proportions; and that public health systems can intervene to modify the R 0 value thereby potentially managing, preventing, or reversing social infections at epidemic scale (yielding population-averaged comparative R 0 ratios, 7.77 < R c < 15.58, p < 0.01, for weight loss propagation). To facilitate adoption of the methodologies, macros and code for use with various statistical software packages are included. The results indicate that the social induction of health interventions are not only possible, but that propagation can both be isolated causally via trial design and be quantified over time. This bodes enormous promise for developing and quantifying future self-sustaining public health interventions.