The SNAP-Ed Evaluation Framework: demonstrating the impact of a national framework for obesity prevention in low-income populations.

The United States Department of Agriculture's Supplemental Nutrition Assistance Program Education, known as SNAP-Ed, is the country's largest and most diverse community nutrition program. In 2017, nearly 140 SNAP-Ed implementing agencies (SIAs) and hundreds of contractors delivered nutrition education to almost 5 million people in nearly 60,000 low-resource sites. Millions more were impacted with social marketing campaigns and policy, systems, and environmental changes. This article introduces and describes the benefits of the newly developed SNAP-Ed Evaluation Framework (Framework) and companion Interpretive Guide to consistently measure SNAP-Ed outcomes across different settings. The Framework uses the social ecological model as its underlying theory and features 51 indicators across four levels: Individual, Environmental Supports, Sectors of Influence, and Population Results. Topline findings from the first-year Census to track Framework adoption found that most SIAs intended to impact indicators closer to the inner levels of influence: Individual (mean = 59% of SIAs; SD = 22%) and Environmental Settings (mean = 48%; SD = 23%). As yet, few SIAs targeted outcomes for long-term indicators (mean = 26%; SD = 15%), Sectors of Influence (mean = 20%; SD = 12%), or Population Results (mean = 30%; SD = 11%). An in-depth example of how one state is using the Framework is described. The SNAP-Ed Evaluation Framework offers a new suite of evaluation measures toward eliminating disparities that contribute to poor diet, physical inactivity, food insecurity and obesity. Practitioners will need technical assistance to implement the Framework, especially to measure longer-term, multi-sector and population results, and to maximize effectiveness in SNAP-Ed.

Impact of California Children's Power Play! Campaign on Fruit and Vegetable Intake and Physical Activity among Fourth- and Fifth-Grade Students.

PURPOSE: Examine the impact of the Children's Power Play! Campaign on fruit and vegetable (FV) intake and physical activity (PA). DESIGN: Study design was a cluster randomized, controlled trial. SETTING: Forty-four low-resource public schools in San Diego County, California, were included in the study. SUBJECTS: Study subjects comprised a total of 3463 fourth/fifth-graders (1571 intervention, 1892 control), with an 86.9% completion rate. INTERVENTION: Throughout 10 weeks, activities were conducted during/after school, including weekly FV/PA lessons and PA breaks; biweekly classroom promotions/taste tests; posters displayed in/around schools; and weekly nutrition materials for parents. MEASURES: Self-reported FV intake (cups/d) and PA (min/d) were collected at baseline and follow-up using a diary-assisted, 24-hour dietary recall and Self-Administered Physical Activity Checklist. ANALYSIS: Multivariate regression models adjusted for demographics and cluster design effects were used, with change as the dependent variable. RESULTS: Intervention children, compared with controls, showed gains in daily FV intake (.26 cups, p < .001) and PA time at recess/lunch (5.1 minutes, p = .003), but not total daily PA minutes. CONCLUSION: Power Play! can help schools and community organizations improve low-income children's FV intake and PA during recess/lunch.

Fruit and Vegetable, Fat, and Sugar-Sweetened Beverage Intake Among Low-Income Mothers Living in Neighborhoods With Supplemental Nutrition Assistance Program-Education.

OBJECTIVE: To examine among low-income mothers the consumption of fruits and vegetables (FV), high-fat foods, and sugar-sweetened beverages (SSBs) and overall diet quality in relation to levels of reach of Supplemental Nutrition Assistance Program-Education (SNAP-Ed) interventions across 2,907 California census tracts. DESIGN: Cross-sectional telephone survey conducted from April through October, 2014 using the Automated Self-administered 24-Hour Recall dietary assessment. PARTICIPANTS: Mothers or primary caregivers (n = 6,355) from randomly selected SNAP households. The sample was 42.6% Latina, 25.5% white, and 17.6% African American. The response rate was 60.5%. MAIN OUTCOME MEASURES: Cups of FV; calories from high-fat foods; and cups of SSBs, overall and from items purchased from fast-food restaurants. Overall diet quality was assessed by the Healthy Eating Index-2010. ANALYSIS: Linear regression controlling for race/ethnicity and education, with significance at P ≤ .05. RESULTS: Mothers from high SNAP-Ed reach census tracts ate more cups of FV, consumed fewer calories from high-fat foods, and drank fewer cups of SSBs. Healthy Eating Index-2010 scores did not vary by levels of SNAP-Ed reach. CONCLUSIONS AND IMPLICATIONS: Supplemental Nutrition Assistance Program-Education interventions are related to increased intake of FV and decreased consumption of high-fat foods and SSBs, but not overall diet quality. Future studies should include assessment of physical activity to investigate caloric balance in association with levels of SNAP-Ed interventions.

Support for Policies to Improve the Nutritional Impact of the Supplemental Nutrition Assistance Program in California.

The Supplemental Nutrition Assistance Program (SNAP) provides a vital buffer against hunger and poverty for 47.6 million Americans. Using 2013 California Dietary Practices Survey data, we assessed support for policies to strengthen the nutritional influence of SNAP. Among SNAP participants, support ranged from 74% to 93% for providing monetary incentives for fruits and vegetables, restricting purchases of sugary beverages, and providing more total benefits. Nonparticipants expressed similar levels of support. These approaches may alleviate the burden of diet-related disease in low-income populations.

Reach of Supplemental Nutrition Assistance Program-Education (SNAP-Ed) interventions and nutrition and physical activity-related outcomes, California, 2011-2012.

INTRODUCTION: This study combined information on the interventions of the US Department of Agriculture's Supplemental Nutrition Assistance Program-Education with 5,927 interview responses from the California Health Interview Survey to investigate associations between levels of intervention reach in low-income census tracts in California and self-reported physical activity and consumption of fruits and vegetables, fast food, and sugar-sweetened beverages. METHODS: We determined 4 levels of intervention reach (low reach, moderate reach, high reach, and no intervention) across 1,273 program-eligible census tracts from data on actual and eligible number of intervention participants. The locations of California Health Interview Survey respondents were geocoded and linked with program data. Regression analyses included measures for sex, age, race/ethnicity, and education. RESULTS: Adults and children from high-reach census tracts reported eating more fruits and vegetables than adults and children from no-intervention census tracts. Adults from census tracts with low, moderate, or high levels of reach reported eating fast food less often than adults from no-intervention census tracts. Teenagers from low-reach census tracts reported more physical activity than teenagers in no-intervention census tracts. CONCLUSION: The greatest concentration of Supplemental Nutrition Assistance Program-Education interventions was associated with adults and children eating more fruits and vegetables and adults eating fast food less frequently. These findings demonstrate the potential impact of such interventions as implemented by numerous organizations with diverse populations; these interventions can play an important role in addressing the obesity epidemic in the United States. Limitations of this study include the absence of measures of exposure to the intervention at the individual level and low statistical power for the teenager sample.

Reliability of a retail food store survey and development of an accompanying retail scoring system to communicate survey findings and identify vendors for healthful food and marketing initiatives.

OBJECTIVE: To develop a retail grocery instrument with weighted scoring to be used as an indicator of the food environment. PARTICIPANTS/SETTING: Twenty six retail food stores in low-income areas in California. INTERVENTION: Observational. MAIN OUTCOME MEASURE(S): Inter-rater reliability for grocery store survey instrument. Description of store scoring methodology weighted to emphasize availability of healthful food. ANALYSIS: Type A intra-class correlation coefficients (ICC) with absolute agreement definition or a κ test for measures using ranges as categories. RESULTS: Measures of availability and price of fruits and vegetables performed well in reliability testing (κ = 0.681-0.800). Items for vegetable quality were better than for fruit (ICC 0.708 vs 0.528). Kappa scores indicated low to moderate agreement (0.372-0.674) on external store marketing measures and higher scores for internal store marketing. "Next to" the checkout counter was more reliable than "within 6 feet." Health departments using the store scoring system reported it as the most useful communication of neighborhood findings. CONCLUSIONS AND IMPLICATIONS: There was good reliability of the measures among the research pairs. The local store scores can show the need to bring in resources and to provide access to fruits and vegetables and other healthful food.

Effect of fresh fruit availability at worksites on the fruit and vegetable consumption of low-wage employees.

OBJECTIVE: To examine the impact of fresh fruit availability at worksites on the fruit and vegetable consumption and related psychosocial determinants of low-wage employees. DESIGN: A prospective, randomized block experimental design. SETTING: Seven apparel manufacturing and 2 food processing worksites. PARTICIPANTS: A convenience sample of 391 low-wage employees in 6 intervention worksites and 137 low-wage employees in 3 control worksites in Los Angeles, CA. INTERVENTION: Fresh fruit deliveries with enough for 1 serving per employee, 3 days a week for 12 consecutive weeks. The control worksites did not receive the fruit deliveries. MAIN OUTCOME MEASURES: Participants' fruit and vegetable consumption, fruit and vegetable purchasing habits, self-efficacy, job satisfaction, and overall health were measured at baseline, weeks 4 and 8, and following the 12-week intervention. ANALYSIS: Descriptive statistics and growth curve analysis using hierarchical linear modeling were employed to analyze the data. RESULTS: Participants in the intervention worksites showed a significant increase in fruit, vegetable, and total fruit and vegetable consumption, purchasing of fruit, family purchasing of vegetables, and self-efficacy toward eating 2 servings of fruit each day compared to the control worksites. CONCLUSIONS AND IMPLICATIONS: Improving access to fruit during the workday can improve fruit and vegetable consumption, purchasing habits, and self-efficacy of low-income employees.

The Power Play! Campaign's School Idea & Resource Kits improve determinants of fruit and vegetable intake and physical activity among fourth- and fifth-grade children.

OBJECTIVE: Examine the effect of the California Children's Power Play! Campaign's School Idea & Resource Kits for fourth/fifth grades on the psychosocial determinants of fruit and vegetable (FV) intake and physical activity (PA). METHODS: Randomized, controlled trial (n = 31 low-resource public schools; 1,154 children). Ten grade-specific, 50-minute nutrition and PA lessons over an 8-week period. Pre/post surveys measuring knowledge, outcome expectations, and self-efficacy (SE) were analyzed using analysis of covariance, controlling for baseline values; chi-square comparing positive/negative changes, adjusted for cluster design effects. RESULTS: Intervention children made significant gains for FV knowledge (4 items, P < .05 to P < .001); positive outcome expectations (fifth grade only, P < .001); asking/shopping and eating SE (P = .04 and P < .001); PA knowledge (2 items, P < .01 to P < .001); outcome expectations (P < .05); and support seeking SE (P = .04); but not SE to overcome barriers. CONCLUSIONS AND IMPLICATIONS: The Power Play! Campaign's School Idea & Resource Kits improved the determinants of FV intake and PA.

Building evaluation capacity in local programs for multisite nutrition education interventions.

From 2004-2008, capacity to conduct program evaluation was built among the Network for a Healthy California's 48 largest local partners. Capacity building was done within a framework of Empowerment Evaluation and Utility-Focused evaluation. Tools included: a Scope of Work template, a handbook, a compendium of surveys, an evaluation plan and report template, data entry and analysis templates, teleconferences, workshops, and technical assistance. Over time, more programs completed an evaluation, used a more rigorous design, and reported statistically significant results. Acting as an evaluation guide, rather than a funder, Network staff built rapport with local programs that annually undertake new evaluations.

Using an opinion poll to build an obesity-prevention social marketing campaign for low-income Asian and Hispanic immigrants: report of findings.

OBJECTIVE: To gain opinions from low-income, limited-English-speaking Hispanic and Asian immigrants for formative research in a social marketing campaign. DESIGN: Nineteen questions on obesity prevention-related topics were embedded into a larger random digit-dial survey investigating the effects of language and cultural barriers on health care access. Participants were selected by ethnic encoding from consumer databases. SETTING: California's northern, southern, and Central Valley regions. PARTICIPANTS: Nine hundred and five adult Hispanic, Chinese, Vietnamese, Hmong, and Korean Californians from households < 130% of the Federal Poverty Level interviewed in 2005. VARIABLES MEASURED: Media usage, food stamp participation, health insurance, health problems, access and availability of fruits and vegetables (FVs) and physical activity, beliefs about overweight, and related regulation and policy change. ANALYSIS: Descriptive statistics and percentages for all questions. RESULTS: Latinos reported receiving most information from television; Hmong from radio. Hispanics, Koreans, and Vietnamese thought diabetes was the greatest health issue in California. Among Hmong, 83% thought FVs were too expensive, and 49% of Vietnamese thought good quality, affordable fresh FVs were too hard to find. CONCLUSIONS AND IMPLICATIONS: Identifying characteristics and opinions that distinguish these ethnic immigrant populations better enables the Network for a Healthy California to develop culturally relevant social marketing campaigns and materials.

California adults increase fruit and vegetable consumption from 1997-2007.

OBJECTIVE: To determine whether fruit and vegetable consumption among California adults significantly increased from 1997-2007. DESIGN: Biennial telephone surveillance surveys of California adults' dietary practices. PARTICIPANTS: California adults (n = 9,105 total all 6 surveys). INTERVENTION: Surveillance data reporting. MAIN OUTCOME MEASURES: Changes in fruit and vegetable consumption over time between 1997-2007, measured by mean servings and percentage of adults eating ≥ 5 servings on any given day. ANALYSIS: Comparisons of subsets both within the same year and across years were made using t tests, chi-square, and Tukey Studentized Range tests at 5% procedure-wise error rate. RESULTS: California adults significantly increased mean daily servings of fruits and vegetables from 3.8 servings in 1997 to 5.2 servings in 2007. CONCLUSIONS AND IMPLICATIONS: Since 1998, notable improvements in fruit and vegetable consumption have occurred to California populations, including the target audience groups of the Network for a Healthy California.

Differences in fruit and vegetable intake among Hispanic subgroups in California: results from the 2005 California Health Interview Survey.

OBJECTIVE: To compare total fruit and vegetable intake in cup equivalents and its individual components among Hispanic subgroups in California. METHODS: Data are from the adult portion of the 2005 California Health Interview Survey. Hispanic/Latino subjects (n=7,954) were grouped into six subcategories (Mexican, Central American, Caribbean, Spanish American, South American, and >1 group). Total fruit and vegetable intake in cup equivalents was estimated from frequency responses about seven food categories. Both t test and chi(2) test were used to assess differences in sociodemographic characteristics across Hispanic subgroups. Multivariate linear regressions using SUDAAN software (Survey Data Analysis, version 9.0.1, 2005, Research Triangle Institute, Research Triangle Park, NC) were conducted to obtain means of total fruit and vegetable intake in cup equivalents and its components by Hispanic subgroups controlling for confounders. RESULTS: Hispanic subgroups did not differ in their intake of total fruit and vegetable intake in cup equivalents (mean 3.4 c and 2.9 c for men and women, respectively). Small but significant differences (P<0.01) were found across Hispanic subgroups in individual fruit and vegetable components (green salad [women only], cooked dried beans and nonfried white potatoes) after adjusting for potential sociodemographic and acculturation confounders. CONCLUSIONS: Fruit and vegetable intake by Hispanic respondents did not meet the national recommendation, although their reported intake is higher compared to other race/ethnicity groups. The public health message remains the same: Increase fruit and vegetable intake. Examination of intake for subgroups of Hispanics may enhance the utility of dietary information for surveillance, program and message design, and intervention and evaluation.