Purpose: Few comprehensive assessments of the availability and prominence of healthy and less healthy (discretionary) food in supemarkets have been undertaken using the protocols developed by the INFORMAS network. This study aimed to evaluate Australian supermarket environments and determine if the marketing of discretionary food varied by supermarket chain and/or socioeconomic disadvantage.

Methods: The study was a cross-sectional supermarket audit conducted in metropolitan areas of Melbourne and Geelong, Victoria, Australia during 2019. A random sample of 104 stores was audited from four supermarket chains (Coles, Woolworths, ALDI, Independent), stratified by area-level socioeconomic position.

Measures included: i. Proportion of shelf-space in square metres allocated to selected discretionary foods (chips, chocolate, confectionery, sweet biscuits, soft/energy drinks) and healthy foods (fresh/frozen fruit and vegetables); ii. Percentage of end-of-aisle, checkout and island bin displays containing a discretionary food item; and iii. Proportion of space within end-of-aisle, checkout and island bin displays devoted to discretionary food. Linear regression adjusting for the clustered sample design was used to test for differences between supermarket chain and interactions with socioeconomic position.  

Results: Shelf space devoted to discretionary food was greater for Independent stores (75.3%) compared with Woolworths (65.7%), Coles (64.8%) and ALDI (63.2%) (all p<.0001). The proportion of shelf space devoted to discretionary food was 7.0% higher in the most compared to least disadvantaged areas (p<0.01). Across all stores, 87% of staffed checkouts and 55% of end-of-aisle displays facing the store front included displays of less healthy food. Less healthy discretionary food accounted for >70% of display space at i. Island bin displays near end-of-aisles, ii. self checkout displays, and iii. island bins near checkouts. ALDI was less likely to display discretionary food in both fixed displays and island bins at prominent end-of-aisles and checkout locations.

Conclusions: Extensive marketing of discretionary food in all Australian supermarket chains is likely to strongly influence purchasing patterns and diets at a population scale. Private and public sector policies to reduce consumer exposure to discretionary food in-store are recommended.

Purpose

A retailer-academic collaboration aimed to co-design and pilot commercially sustainable strategies to increase sales of healthier foods relative to less healthy foods within a product category in a major supermarket chain.

Methods

Two co-design workshops were held, involving supermarket corporate strategy staff and public health nutrition academics, to identify potential interventions. These were mapped against choice architecture frameworks and retailer strategic priorities before one intervention, more prominent shelf placement of healthier products within one category (breakfast cereals), was selected for piloting. A 36-week pilot study (12-weeks each, baseline, intervention and follow-up) was undertaken in six supermarkets (three intervention and three control stores) in Auckland, New Zealand. Products were ranked by nutrient levels and nutrient profile, and healthier products were placed at eye level and less healthy products were placed on lower shelves. No changes were made in matched control stores. The primary outcome was change in sales of healthier products relative to total category sales. Secondary outcomes included nutrient profile of category sales, in-store promotions, customer perceptions, and retailer feedback.

Results

Overall, there was no difference in the sales of more prominently positioned products between intervention (56%) and control stores (56%) during the intervention. There were also no significant differences in sales across the study time periods nor the nutritional composition of product sales. Intervention stores were found to have a greater number of in-store displays compared to control stores (685 vs 583) with a higher proportion of less healthy less prominent products (57% vs 43%) displayed. A significant interaction was evident between in-store promotions and sales. Most customers (265, 88%) supported shelf placement as a strategy to improve purchases but noted that brand preferences and price could override product choice.

Conclusions

Shelf placement alone was not an effective strategy to increase purchases of healthier products. Other important influences within the retail environment, including space management logistics, in-store promotions, and habitual shopping habits in the breakfast cereal category, may have diluted the effect of the shelf placement intervention.

Purpose: This research aims to improve healthfulness of choices in a supermarket in two stages: (1) co-design with consumers and staff, and (2) an ecological trial that modified real supermarket environment. Ecological models suggest that health is determined by both, individual factors and surrounding influences, including retail settings. Retail environments have been shown to (usually, negatively) influence healthfulness of food choices. In developed countries, over 70% of food comes from supermarkets – a major opportunity to improve healthfulness of food choices.

Methods: Stage 1 Co-design: five qualitative 90 minute co-design sessions were conducted with 24 consumers and eight retail staff. The tasks included: discussion on past experiences of making healthy and not choices; feedback on 28 previous campaigns; creating own campaign ideas in the form of colourful ‘mud maps’. All workshops were voice recorded and photos were taken of all ‘mud maps’. The thematic analysis was undertaken by two independent researchers, through the process of triangulation using the Seven Ps social marketing framework.

Stage 2 Ecological trial: Based on co-design findings, researchers and supermarket staff developed an intervention campaign “A healthy choice”. The three-month campaign consisted of: displaying shelf talkers on supermarket shelfs against 200 products identified as ‘healthy’ based on the levels of sugar, total/saturated fat, salt and fibre; cooking demonstrations, label reading supermarket tours, and a community-wide information campaign. 

Results: Product and Price strategies were the most frequently mentioned by consumers and staff. Interestingly, consumers wanted competitive offer, rather than just ‘cheap’ for healthy products. In Process, consumers desired enhanced in-store experiences (cooking demonstration, label reading workshops). The People aspect, suggested to bring other experts to retail environment (ie dieticians, chefs). Comparison of total sales (in units) for the 200 products labelled with “A Healthy Choice” campaign over the 14 weeks period, with the same period in the previous year (2018), shows a 3.7% increase in sales (or 606 units). This increase is substantive for the supermarket industry, considering there was no discounts offered, and the overall sales in that supermarket declined slightly compared to the previous year.

Purpose: SuperShelf is an innovative values-based intervention that uses behavioral economics and systems change to transform food pantries into welcoming environments for communities to access healthy, appealing foods. SuperShelf values include good food, respect for all, collaborative partnership, evidence-based practices, and systemic thinking. The effect of SuperShelf on pantry and client outcomes is being evaluated in a group-randomized study. Food pantries are challenging sites for intervention, and the SuperShelf intervention requires a high degree of organizational buy-in; thus, monitoring the successes and challenges of implementation of this values-based intervention is essential.

Methods: This study used qualitative interviews with food pantry managers and volunteers to explore their experience of transforming into a SuperShelf. Semi-structured interviews were conducted pre (n= 16) and post (n=16) transformation at 4 intervention sites and 4 control sites. Interviewers were audio recorded and transcribed verbatim. NVivo12 was used to facilitate analyses. A constructivist approach to grounded theory was used to conduct the analysis, allowing for themes to emerge and to be considered alongside the intervention values. 

Findings: SuperShelf values of ‘good food’ and ‘respect for all’ were evident in emerging themes pre and post transformation, in control and intervention sites, illustrating the commitment at all sites to provide a dignified client experience and food that was healthy, appealing, and culturally appropriate.  However expression of these values was greatly enhanced post-transformation in intervention sites. The value of ‘systemic thinking’, ‘evidence based practices’ and ‘collaboration’ were not evident in the thematic analysis pre-transformation, but did emerge post-transformation in intervention sites. Additionally some indication of increased ‘systemic thinking’ also emerged in control sites, indicating a shift in the broader hunger relief system. Finally, the physical transformation of pantries was described as particularly meaningful for participants, and formed a conduit around which a broader cultural transformation of values were experienced.

Conclusion: The SuperShelf transformation process was experienced positively by intervention sites; the physical transformation was particularly impactful, and served to facilitate adoption of all the SuperShelf values. Systems-wide changes may be experienced in non-intervention sites. Food pantries can successfully transform to integrate a values-driven, systems-change approach to improving the client experience.

Introduction: Conflicting evidence on the association between community food environment (CFE) and weight status has been reported in developed countries and evidence from developing countries is scarce.

Purpose: To examine the association between availability of franchised food outlets (FFO) and overweight, including obesity (OW) in 6-17-year-old Malaysians in states of Perak (high OW prevalence) and Terengganu (low OW prevalence).

Methods: Information on residential location and body mass index (BMI) based on measured weight and height for individuals residing in Perak and Terengganu was extracted from the National Health and Morbidity Survey 2015. The International Obesity Task Force BMI criteria was used to define OW. CFE data was obtained from the official websites of FFO, including supermarkets (SM), fast food outlets (FF) and convenience stores (CS). Spatial analysis was performed using ArcGIS 10.5 to assess the CFE. Availability of the FFO was defined as the presence of ≥1 FFO within 1000m radius from individual residence. Log-binomial regression was performed to quantify the associations between CFE and OW, adjusted for sociodemographic factors (age, sex, race, residential area and household income).

Results: In total, a complete data was available for 880 of 932 individuals and 916 FFO were identified (SM=321, FF=314, CS=281). The age-standardised OW prevalence was 30.1% in Perak and 13.6% in Terengganu (p<0.001). A significantly higher number of FFO within 1000m radius of individual residence was observed in Perak than Terengganu (p<0.001). In Perak, availability of CS (PR 2.27;95%CI 1.35, 3.80) and SM (PR 1.63;95%CI 1.00, 2.64), but not FF (PR 1.05; 95%CI 0.66, 1.67) were associated with higher OW prevalence. The respective associations were insignificant in Terengganu.

Conclusions: Disparities in the availability of FFO between states in Malaysia exist. FFO within walkable distance from individual residence were more common in the state with high prevalence of OW, and the availability of convenience stores and supermarkets was associated with higher prevalence of OW in that state. Future research should incorporate other built environment features including physical activity environment, and individual behaviours.

Purpose: Childhood overweight and obesity is a major public health issue in Australia with 25.8% of children aged 2-17 years overweight/obese. The $45M South Australian OPAL intervention applied a whole-of-community approach in a multi-site childhood obesity prevention initiative across metropolitan and regional areas. This evaluation assessed the main effect of OPAL and the contextual factors modifying it.  Methods: OPAL involved 5 years of implementation for 20 intervention communities matched with 20 comparison communities. Effectiveness was assessed using the non-equivalent control group quasi-experimental evaluation design with 5 years pre-intervention and 2 years post-intervention measures. Change over time was assessed for the annual community prevalence of overweight/obesity based on body mass index derived from measured height and weight for 4-5-year-old pre-school children (n=68,763 records) with addresses geocoded and assigned to suburbs nested in communities (n=216 intervention, 225 comparison suburbs). Analyses accounted for baseline differences in overweight/obesity prevalence and the social (median suburb income and education) and built (suburb fast-food outlet density, walkability, dwelling density) environments.  Results: Exposure to OPAL yielded reductions in overweight/obesity for 4-5-year-old children in the intervention relative to comparison condition. This continued in the two years after funding ceased. The preventive fraction was 12.2%. OPAL advantaged higher SES areas: overweight/obesity decreased for suburbs in the highest education and income tertiles for the intervention but not comparison condition. For suburbs in the intervention but not comparison condition, overweight/obesity decreased in the highest walk score tertile. Overweight/obesity increased over time for suburbs in the highest fast-food tertile. High dwelling density was associated with reduced overweight/obesity. Conclusions: In Australia, from 2007/8 to 2014/15 the prevalence of overweight/obesity in 5-12-year-old children rose from 21.8% to 26.3%. This evaluation shows that for the same time frame, in South Australia, OPAL was associated with a 12.2% reduction in overweight/obesity prevalence for 4-5-year-old children. Both social and built environments modified OPAL's impact. That high walk scores amplified the impact of OPAL, but high concentrations of fast-food outlets attenuated and reversed the otherwise positive intervention effect suggests future interventions need to account for built environmental factors that can block the success of an otherwise positive program.

Purpose: Numerous schools have implemented programs to address childhood obesity; however, few have evaluated obesity-related health behaviors of classroom teachers implementing these programs. This cross-sectional study examined daily physical activity (PA) including time spent engaged in light PA and moderate-vigorous (MVPA) and PA-related classroom practices (e.g. punishment by withholding recess; role modeling healthy PA habits) among teachers in 20 urban, suburban and rural U.S. schools participating in a childhood obesity prevention intervention. 

Methods: Light and MVPA were assessed using 24-hour accelerometry (Actical) worn on the ankle for seven consecutive days. PA-related classroom practices were assessed via survey (7 items; sum score range: 0-35; alpha=0.80; higher score=more health promoting practices). Demographics included gender, race, self-reported height/weight (calculated BMI; >30kg/m²=obese), grade taught, years teaching experience, and education level. Analyses included linear mixed regression models, accounting for within-school effects and adjusting for demographic covariates.

Results: Classroom/special area teachers and teaching assistants with complete survey data and one or more 24-hour periods of accelerometry data were included (n=290, 90% female; 63% elementary school; 60% white; aged 43.2+11.3 years; 41% obese). The majority of teachers had <20 years teaching experience (73%) and 62% had a graduate degree. Means and standard deviations were calculated for average minutes/day spent in light (399.2+85.2) and MVPA (24.1+14.4) and PA-related classroom practices sum score (22.4+4.9). From linear regression models, for every one-minute increase in MVPA, there was a 0.06 increase in PA-related classroom practices sum score (SE=0.02; p<0.01). Fewer years teaching (β:-2.03; SE:0.74, p=0.01) and female gender (β:-2.25; SE:1.06, p=0.03) were significantly associated with lower sum scores; whereas, elementary school teachers had higher scores (β: 3.56; SE:0.67, p<0.01). Light PA was not associated with PA-related classroom practices sum score.

Conclusions: Teachers with higher MVPA were more likely to engage in classroom practices that could positively impact their students’ PA patterns. Future studies in school settings should incorporate evaluations of teacher health behaviors and explore whether improvement of teacher health behaviors subsequently impacts student health behaviors.

Purpose

Well-connected cycling infrastructure creates a transportation environment that promotes physical activity. To choose active transport underserved users need access to affordable bicycles and bike maintenance, elements often overlooked by urban planners. Nonprofit, community bike shops provide refurbished bicycles and mechanical assistance at low cost or through sweat equity programs. This study documents the role of community bike shops in providing equitable access to bikes and bike maintenance and makes the case for their inclusion in public policy.

Methods

A thirty-six question survey was sent to the 205 US community bike shops. Eighty-eight completed surveys were returned (42.9%). Question categories consisted of Likert scales and qualitative data regarding bike shop goals, barriers, programming, and communities served. Neighborhood demographics (race, sex, education, median income, poverty rate) and transportation mode (car, public transportation, walk, bike) were compared to averages using 2017 US Census Data and analyzed using a one-way t-test.

Results

Findings supported the hypothesis that community bike shops provide needed services to help underserved populations mode shift to active transport. Primary goals reported in the survey were to provide bicycles for active transport (88.1%) and to empower people in the community (88.1%).  Sweat equity opportunities for bikes and maintenance are provided at 62.5% of the shops. They serve diverse populations: low income/underemployed (98.9%), homeless (89.8%), unemployed (89.8%), refugees (82%), veterans (84.1%), LGBTQ (88.6%), underserved teens/youth (86.4%), and individuals with disabilities (73.9%). They provide services to minority communities – significantly below the national average for whites (p=0.001) and significantly above the average for blacks (p=0.002), and are located in areas significantly below the median income (p<0.001) and significantly higher in poverty (p<0.001). People served are significantly less likely to use a car to commute (p=0.005) and significantly more likely to use public transportation (p=0.004), walking (p<0.001), and biking (p<0.001). 

Conclusions

Nonprofit, community bike shops are needed to ensure equitable access to active transport and to increase physical activity in underserved populations. 

Purpose: Many college students accumulate insufficient aerobic and/or muscle strengthening activities to benefit their health, despite the availability of physical activity resources. The purpose of this study was to examine disparities in physical activity, as well as campus recreation facility use and comfort at a large university located in Northeastern United States of America.

Methods: This cross-sectional mixed-methods study collected data from students using an online survey that assessed demographics, physical activity behaviors (aerobic and muscle-strengthening), campus recreation facility use, comfort using campus recreation facilities, as well as reasons for discomfort, and strategies used to feel comfortable.

Results. The majority of participants (n = 319) identified as women (68.9%), non-Hispanic Caucasian (81.6%) and heterosexual (93.3%). While aerobic physical activity did not differ between men and women, women reported significantly less muscle-strengthening activity (p < 0.001, η² = 0.06), lower frequency of weight use (p = 0.035, η² = 0.02) and informal sport participation (p < 0.001, η² = 0.07), and higher frequency of cardio equipment use (p < 0.001, η² = 0.08) and group exercise participation (p < 0.001, η² = 0.09) compared to men. Women also reported significantly lower comfort using facilities in general (p=0.002, η²=0.05), as well as machine weights (p < 0.001, η² = 0.11), free weights (p < 0.001, η² = 0.15), and indoor running tracks (p = 0.023, η² = 0.03) compared to men. The presence of men, a lack of skill, and self-consciousness/judgement emerged as dominant themes explaining women’s discomfort using campus recreational facilities, in particular weight training areas. Regarding strategies used to feel comfortable, the most common theme was using music to tune into working out and/or tune out the surrounding environment, followed by going with friends for social support, gaining knowledge pertaining to facility and equipment use, and selecting clothing for comfort or appearance reasons.

Conclusions: Meaningful differences in physical activity behaviors and comfort using campus recreational facilities emerged between men and women. Administrators should consider making changes to environments and/or policies to provide equitable physical activity opportunities.    

Purpose: Promoting healthy behaviors requires focusing on priority communities. In Colombia, where women are less physically active than men, the Bogotá government offers free physical activity (PA) classes in public parks through the Recreovía program. Despite broad uptake, especially among women, there is notable variability in participation across Recreovía sites. We aimed to 1) evaluate built and social environmental factors associated with Recreovía local park environments; 2) identify perceptions that influence Recreovía engagement; and 3) facilitate an advocacy process among community members, policymakers and academic researchers to improve uptake of the program.

Methods: Using an explanatory sequential mixed methods design, we first collected contextual quantitative data using the System for Observing Play and Recreation in Communities and the Physical Activity Resource Assessment. We assessed participants’ anthropometrics, sociodemographic characteristics, and PA levels (using accelerometers). The qualitative component involved the Our Voice citizen science model in which community members use a mobile application to identify park features that promote or hinder PA, then analyze and use data to advocate for changes to enhance PA. We selected two public parks, representing the lowest and highest 10% in Recreovía attendance. We used descriptive statistics to analyze park and users’ characteristics. Qualitative data were thematically analyzed.

Results: Although both parks presented similar levels of moderate to vigorous PA (P=0.290), one had better quality than the other (P= 0.009). Participants (n=48) were primarily women (65%) who were meeting PA recommendations (93%), and highly satisfied with Recreovía (97%). Factors influencing use of the program included perceptions that it enhanced social and individual well-being, improving the friendliness of the park environment, and fostering self-confidence, social capital, mental health and civic engagement. Reported barriers to usage were related to park cleanliness, and safety. Presentations to stakeholders sparked planning aimed at reducing these barriers.

Conclusions: This study provides an adaptable framework for using technology-driven participatory methods to evaluate park-based interventions and understand factors supporting community-based PA behaviors. The Our Voice process elicited community dialogue, empowerment, and advocacy, facilitating interaction among stakeholders to optimize the Recreovía program and facilitate active living.