Obesity Prevention Among American Indian Adults: Impact Evaluation of a Multi-level, Multi-component Intervention
Johns Hopkins University
Prevalence of adult overweight and obesity is alarmingly high, disproportionately so in American Indian and Alaska Native (AIAN) populations. Factors contributing to this health disparity in this population are multifaceted, and include several environmental factors, extreme poverty, and the tendency for AIAN to live in remote and economically-deprived areas where access to healthy foods and physical activity opportunities is low but access to inexpensive and unhealthy foods is high. Many public health experts and researchers have explored the prospect of school, food store, or other institutional obesity interventions as a means of increasing healthy food availability and improving food and physical activity (PA) related behaviors. However, none of these interventions have taken a multi-level, multi-component (MLMC) approach, which is essential for exposure and reinforcement of the intervention messages to all segments of the population. The Obesity Prevention Research and Evaluation of InterVention Effectiveness in NaTive North Americans (OPREVENT) was a multi-level, community-based adult obesity intervention pilot-study (2012-2015) funded by the U.S. Department of Agriculture (USDA). The program took place within food stores, schools, worksites, and media outlets and was designed to intervene at multiple environmental levels to promote healthier food and PA behaviors such as healthy food purchasing, improved dietary intake, and increased PA. Five AI communities in Michigan and New Mexico were randomized to either immediate intervention (n=3) or delayed intervention (n=2). Food stores, worksites, and schools were recruited from each community and received intervention materials and support from the OPREVENT study team throughout six intervention phases over the course of one year. Materials and support included posters, educational displays, flyers, booklets, giveaways, and interactive sessions such as taste tests and cooking demonstrations. Respondents were randomly selected from each community, and baseline and follow-up interviews were conducted to assess changes in food and PA related behaviors, dietary intake, PA level, and psychosocial variables (PSV) as well as to evaluate level of exposure to the intervention. Data for this analysis were taken from the qualitative food frequency questionnaire, the International Physical Activity Questionnaire-Short Form, select sections of the Adult Impact Questionnaire, and the Intervention Exposure Evaluation. Baseline interviews were completed with 424 respondents, and follow-up interviews were completed with 299 respondents. Results suggest that MLMC interventions such as OPREVENT can be effective in increasing volume of moderate PA (MET-min. per week) and decreasing daily servings of regular soda. No changes were observed in psychosocial variables. Process data (not included in this dissertation) and anecdotal evidence suggest that the dietary and PA materials and messages may have been unclear, and that the intervention was not implemented with high exposure, thereby potentially decreasing the probability of observing significant changes. These results provide important information related to designing and implementing large MLMC interventions within AIAN communities. First, AIAN adults appear to be more amenable to decreasing unhealthy beverage intake than decreasing unhealthy food intake. There also seems to be potential for increased volume of PA. There was no change observed in any PSV, however there were intermediate to high levels of some PSV (knowledge and self-efficacy) at baseline. Exposure to the intervention was low. Future research should work to identify ways to improve messaging to see changes in additional foods and beverages, and also look into expanding the current MLMC framework to emphasize intervention messages across all environmental levels and community institutions. To increase effectiveness in the future, researchers should ensure that MLMC interventions are implemented with high exposure. Intervention materials should be pilot-tested to ensure that the messages are culturally acceptable, clearly presented, and related to the evaluation instruments.
obesity, American Indian