The long-term goal is to reduce the prevalence of child obesity, which has been declared one of the major public health challenges of the 21st century by the World Health Organization. The immediate purpose of the proposed study is to evaluate the impact of obesity-related interventions and policies on early childhood obesity risk in communities throughout Los Angeles County, a socio-demographically diverse region with a population of 10 million. An interdisciplinary team from UCLA and UC Berkeley will form partnerships with the Los Angeles County Department of Public Health (LACDPH), Public Health Foundation Enterprises WIC (PHFE WIC) [the largest local agency Supplemental Nutrition Assistance Program for Women, Infants and Children (WIC) in the country], and California Food Policy Advocates (CFPA) to develop new evaluation research methodology using systems science approaches, for this purpose. The proposed study will challenge and shift current paradigms in evaluation research methodology by (1) using outcome data (obesity status in preschool-aged children) from a large administrative database;(2) developing a methodology for retrospectively quantifying the impact of community efforts to address obesity or obesity-related behaviors (diet and physical activity), and (3) pioneering the combined use of causal inference methods and systems science approaches for the analysis. The administrative database providing the outcome data has been maintained by PHFE WIC since 2003;it contains height and weight data gathered from about ~ 200,000 two to five year old WIC participants each year. These data have recently been evaluated for validity and shown to correlate highly with measurements taken by a research team from UCLA. The methodology for quantifying the impact of obesity-related community interventions and policies on early childhood obesity will use existing data gathered by LACDPH as well as information from interviews with key informants identified with the help of LACDPH and CFPA (two organizations highly knowledgeable about community efforts to address child obesity in LA County), to develop an index;development of this index will borrow from the experiences of research groups that have conducted similar work in the field of tobacco control. The application of causal inference methods, which perhaps represent the fastest growing methodological advances in epidemiology, in combination with systems science approaches to analyze the data will shift current paradigms in evaluation research methodology, allowing for the assessment of the effects of unplanned mixtures of possibly time-varying interventions.
The high rates of obesity experienced in the 2000s led to community efforts to improve eating patterns and increase physical activity. The proposed study will apply systems science methodologies to study the impact of such efforts on obesity risk in low income preschool-aged children so as to better understand what it takes to reduce climbing rates of obesity in this vulnerable population. We are focusing on early childhood obesity because research has shown that early childhood is a critical time to develop healthy eating and physical activity habits.
|Nianogo, Roch A; Wang, May C; Wang, Aolin et al. (2016) Projecting the impact of hypothetical early life interventions on adiposity in children living in low-income households. Pediatr Obes :|
|Wang, Aolin; Arah, Onyebuchi A (2015) Body Mass Index and Poor Self-Rated Health in 49 Low-Income and Middle-Income Countries, By Sex, 2002-2004. Prev Chronic Dis 12:E133|
|Nianogo, Roch A; Arah, Onyebuchi A (2015) Agent-based modeling of noncommunicable diseases: a systematic review. Am J Public Health 105:e20-31|
|Wang, Aolin; Arah, Onyebuchi A (2015) G-computation demonstration in causal mediation analysis. Eur J Epidemiol 30:1119-27|
|Liew, Zeyan; Olsen, JÃ¸rn; Cui, Xin et al. (2015) Bias from conditioning on live birth in pregnancy cohorts: an illustration based on neurodevelopment in children after prenatal exposure to organic pollutants. Int J Epidemiol 44:345-54|
|Thompson, Caroline A; Arah, Onyebuchi A (2014) Selection bias modeling using observed data augmented with imputed record-level probabilities. Ann Epidemiol 24:747-53|