UCLA

For decades, obesity has been a major public health problem in the US and has been one of the most predominant players in the increase of type 2 diabetes (T2DM) incidence. Obesity is thought to be the result of the interplay between individual and environmental factors which can occur early and throughout an individual’s life course. Despite major ongoing prevention efforts, obesity is still on the rise and this has warranted its description as a complex health problem calling for the use of systems science methods to disentangle such complexity. The overarching goal of this dissertation was to apply systems science and causal analytical approaches to study the life-course development of obesity and its effects on T2DM. Specifically, we developed an agent-based model of a cohort of children born in Los Angeles county—ViLA (i.e. Virtual Los Angeles Cohort) and followed from birth into adulthood in order (i) to forecast the incidence and trends of obesity and T2DM, (ii) to investigate the mechanisms through which childhood obesity affects T2DM and (iii) to evaluate the effectiveness of key health interventions on obesity and T2DM in ViLA. We used simulated data from 98,230 individuals in ViLA and observational data from 1054 children enrolled in the Special Supplemental Nutrition Program for Women, Infants, and Children and applied the g-computation algorithm to estimate causal quantities. Our results suggest that the incidence and prevalence of obesity and T2DM are generally high with notable racial disparities and will continue rising over time and with age at an alarming rate. Furthermore, much of the effect attributable to childhood obesity in the development of incident T2DM was due to pathways other than through adult obesity. Additionally, engaging in moderate-to-vigorous physical activity and eliminating fast-food consumption were the most effective interventions for preventing obesity and T2DM. For maximum effectiveness, interventions have to be implemented in combination with one another and virtually at every critical life stages throughout the life span. Agent-based simulation models could be used as virtual laboratories for integrating best existing evidence, gaining new insights, exploring new mechanisms and evaluating intervention effectiveness in obesity and diabetes research.