Background: Meta-analyses (MA) of behavioral interventions represent the highest level of evidence for public health recommendations. Conclusions from MA could be misleading if they include pilot and small sample size trials because effects from such trials are associated with inflated estimates of intervention effectiveness. This study examines the impact of including pilot and small trials within MA on topics related to childhood obesity.

Methods: Searches were conducted across four databases for MA published since January 2016 of behavioral interventions on topics related to the prevention/treatment of childhood obesity (e.g., physical activity, sleep, weight loss/maintenance). The computed meta-analytic effect sizes (ES) were extracted from each MA and constituent articles. Articles were classified as a self-identified pilot (feasibility, preliminary efficacy, exploratory) trials, or small (N≤100), mid (N=101-191), and large trials (N>191, median sample size). Effects were compared across trial classifications using multi-level meta-regressions.

Results: A total of 1,602 ES were extracted from 48 MA containing 604 unique articles (avg. 22 per MA, range 2 to 108) including 227,217 participants. On average, MA were comprised of 22% (0-58%) pilot and 21% (0-83%) small trials. Across all MA, ES from large trials (0.128, 95CI 0.053-0.203) were significantly smaller compared to ES from pilot (0.245, 95CI 0.165-0.324) and small trials (0.382, 95CI 0.259-0.505). At the MA level, the agreement between statistically significant (p<.05) ESs from the original MA and re-estimated ESs restricted to large trials was low-to-moderate (kappa 0.52), with 33% of significant ES reported in the original MA rendered non-significant, whereas 9% of non-significant ES were rendered significant using only large trials. Every 10% increase in the proportion of pilot or small trials included within a MA was associated with a 5-11% increase in the ES absolute difference from the original MA estimated ES.

Conclusions: Summary ES should be interpreted with caution when MA of behavioral interventions include a substantial proportion of small or pilot trials. 

Purpose: The relationship between the fast food environment and diet has gained growing attention, but it remains poorly understood how the food environment relates to eating behaviours. There are initial indications that the physical food environment contributes to social consumption norms, yet, this has not been studied on the neighbourhood level. This study aimed to investigate residents’ neighbourhood fast food consumption norms as a potential mediating pathway between residential fast food outlet exposure and consumption.
Methods: A cross-sectional study was conducted. A nationwide sample of 1038 respondents living across the Netherlands recruited through a panel bureau completed a survey. Respondents reported their fast food consumption (amount/week) as well as descriptive and injunctive norm perceptions regarding fast food consumption in their neighbourhood. The Locatus database provided addresses of all fast food outlets in the Netherlands. Fast food exposure was measured as the count of fast food outlets within a 400m street-network buffer around the respondents’ zip-codes. Regression models were used to assess associations between fast food exposure, fast food consumption, and norm perceptions. A bootstrapping procedure was used to test indirect effects.
Results: There was no direct association between fast food exposure and fast food consumption (OR=1.02, p=.22). Residential exposure to fast food outlets was positively associated with descriptive norms (B=.05, p<.001), and injunctive norms (B=.03, p<.001), which in turn were associated with increased odds of consuming fast food (OR=1.16, p=.03, and OR=1.44, p<.001, respectively). There was evidence of indirect effects of fast food exposure on fast food consumption, via descriptive norms (B=.01, 95% CI =.0003 to .013) and injunctive norms (B=.01, 95% CI =.004 to .017).
Conclusions:  Individuals who have a higher residential availability of fast food outlets, perceive fast food consumption in the neighbourhood as more common and appropriate. Future research may explore how the changing food environment may shift consumption norms. Insights may support policymakers in urban planning to develop healthier neighbourhoods and ultimately stimulate healthier consumption.

Purpose: Robust evidence on the most effective delivery of health education to the Chinese remains limited, both in Australia and worldwide. The current practice of Chinese diabetes education relies on language-translating Western evidence-based material. Such translated health education however, was shown to conflict with Chinese expectations, creating additional stress, frustration and anger. Type 2 diabetes disproportionately affects the Chinese population yet there are no structured Chinese diabetes education programs in Melbourne. This project aimed to develop and evaluate a pilot diabetes group education program specifically for Chinese Australians.

Methods: A culturally-tailored group education program (Not Scared of Sugar™) was developed and piloted with Melbourne-based Cantonese-speaking people with type 2 diabetes. Program teaching styles were aligned with the Confucian process of learning and incorporated cultural strategies to promote healthy behaviour change. Thirty-four individuals (35% male) attended five education sessions over ten weeks, delivered by a Cantonese-speaking facilitator and multidisciplinary clinicians. Data was collected from participants at baseline, on program completion and at 6 months follow-up. Differences between baseline and follow-up data were determined using paired sample t-tests or Mann-Whitney U tests.

Findings: Mean participant age was 69 ±9 years, with a mean time of 25.7 ±10.8 years in Australia and a median duration of diabetes of 10(IQR=2.8, 20.5) years. On completion of the program, participant waist circumference was reduced from 90.5 ±9.9 cm to 88.9 ±9.9 cm (P<0.001) and was maintained at 6 months. There was a significant increase in the number of diabetes self-care behaviours undertaken (American Association of Diabetes Educators Questionnaire Score: 30(22, 32.3) versus 34(30.3, 35.8), P<0.001) at 6-month follow-up. HbA1c was unchanged after 6 months (6.8 ±0.7% versus 6.7 ±0.7%, P=0.318). Participants rated the program very highly. 

Conclusions: Culturally tailoring health education is more than language-translation of the materials. To optimise clinical effectiveness and participant experience, health education should align with cultural learning orientation and incorporate specific behaviour change motivators.

Objectives: Accelerated gains in weight, particularly age and sex-specific body mass index (zBMI), during summer break from school are well established in elementary age children (6-11yrs.). It is unclear if young adolescents (12-14 yrs.) experience increases in zBMI or percentage body fat (%BF) during summer break. This study examined the summer trajectories of zBMI and %BF via bioelectrical impedance (BIA) in a sample of young adolescents.

Methods: Two middle schools in the southeastern United States participated in this observational study. Measures of height, weight, and %BF measured via the Tanita DC-430U (Tanita Corporation of American, Arlington Heights, IL) were conducted in the spring (T₀), just before summer break from school, and again in the fall (T₁), as students arrived for new school year approximately 12 weeks apart. All students who provided body composition measures were included in this analysis. Separate multilevel mixed effects linear regression with measures nested within participants estimated change in zBMI and %BF. Age, sex, and race/ethnicity were included in the models as covariates. A sex-by-time interaction was included to examine differences in zBMI and %BF change over summer by sex. 

Results: A total of 73 adolescents (45% female, 52% non-white) were measured at both T₀ and T₁. At T₀ (baseline). Unadjusted regression models showed zBMI and %BF increased over the summer by 0.09 (95%CI 0.02, 0.15) and 1.37% (95%CI, 0.69, 2.04), respectively. Adjusting for age, sex, and race/ethnicity, %BF increased over the summer by 2.13% (95%CI=0.28, 3.99), but zBMI showed no statistically significant increase (0.03, 95%CI= -0.16, 0.23). Sex-by-time interactions indicated no statistically significant difference in %BF or zBMI change by sex.

Conclusions: Summer appears to be associated with increases %BF in young adolescents and possibly increases in zBMI. It is unclear if this represents an acceleration in %BF and zBMI gains when compared to the school year. To understand and compare summer vs. school body composition change, a third measurement is planned in the spring of 2020 following school.

Purpose: To develop a food literacy curriculum utilizing theoretical frameworks and to pilot test the curriculum with adolescents in underserved communities. 

Methods: Development of the curriculum, informed by Social Cognitive Theory and constructivism, followed a Backward Design approach. Curriculum concepts were guided by the definition of food literacy and its components described by Vidgen and Gallegos (2014). Knowledge and skills in three fundamental areas were determined as essential for achieving food literacy: agriculture/food systems, nutrition, and cooking. After internal pilot testing, the draft curriculum was reviewed by an expert committee and revisions were incorporated into the final draft for pilot testing within target populations. Two low-income communities participated in the pilot. Modules were implemented by the same trained facilitator and observer within both communities. Any necessary revisions to the curriculum were made within a 4-month period between pilots. Structured observations were completed by the facilitator and observer to assess whether each lesson met the defined learning objectives and authentic assessments.

Results: The resulting curriculum, Teens CAN: Comprehensive Food Literacy in Cooking, Agriculture, and Nutrition, includes twelve modules featuring experiential lessons. The first pilot demonstrated that adolescents were able to achieve authentic assessments for almost all lessons. Results from the first pilot led to minor modifications throughout to improve clarity. Additionally, three procedures were rewritten to better align with lesson objectives and aid in achievement of authentic assessments. Lastly, two lessons with similar learning objectives were combined into one for brevity. The second pilot confirmed that adolescents were able to achieve authentic assessments for all lessons and that revisions made following the first pilot were successful. Only slight revisions were needed following the second pilot to further improve lessons.     

Conclusions: Achieving food literacy has the potential to improve nutrition-related behaviors and health outcomes. To our knowledge, Teens CAN is the first comprehensive curriculum developed for high school-aged adolescents that incorporates the components of food literacy as defined by Vidgen and Gallegos (2014). Teens CAN is currently being implemented to test for effects on food literacy knowledge, skills self-efficacy, and dietary behaviors.

Purpose: SuperShelf transforms food pantries, creating welcoming environments for communities to access healthy, appealing foods. Using a 6-step process, SuperShelf goes beyond typical behavioral economics-based interventions and supports both nutrition and client-centered practices by considering how food-insecure individuals may make choices in food pantry settings. SuperShelf methods aim to make the healthy choice the easy choice, while respecting individual preferences, cultural appropriate options, and creating a dignified experience. This presentation will focus on the feasibility for pantries to implement practices that are both nutritionally-focused and client-centered, as measured by an environmental implementation tool.

Methods: The SuperShelf evaluation is an NIH-funded study conducted in two waves. Pantries (n=16) are randomized into “transformation” or “delayed-transformation” groups and evaluated over one year. Implementation is measured by a pre/post tool (total score range 0-100), comprised of four subscores that measure practices encouraging healthy choices while respecting individual preferences: stocking standards (29 points), healthy food prominence/appeal 21 points), unhealthy food placement/competition (22 points), and aesthetics/use of space (28 points). We measured all 16 food pantries with the implementation tool pre/post and calculated changes in total scores and subscores.

Results: Among wave 1 pantries (n=8), transformation site implementation scores increased 32.1 points on average, while delayed-transformation site decreased 1.2 points on average. All four subscores also increased among transformation sites and delayed-transformation sites did not change. For example, the healthy food prominence subscore increased an average of 6.5 points and the unhealthy food placement subscore increased an average of 6.6 points, while the delayed-transformation site subscores decreased less than 1 point each. Results from all 16 pantries will be presented.

Conclusion: Score changes in the transformation groups indicate adoption of the SuperShelf model. Although interventions in food pantry settings are challenging, implementation of an intensive model such as SuperShelf is possible and shows promise for improving health equity and the client experience among food insecure populations. Additional evidence evaluating implementation and the factors that translate to a more client-centered food pantry are forthcoming. 

Purpose: Young adulthood has become synonymous with the development of poor lifestyle behaviours associated with an increased risk of preventable chronic disease in later years. Interventions aiming to improve health behaviours may be more engaging and effective if they are targeted to males or females than gender-neutral. This review is the first to examine the outcome effectiveness of gender targeted (males or females only) versus gender-neutral (males and females collectively) interventions targeting nutrition, physical activity or obesity in young adults.

Methods: Six electronic databases were searched for RCTs published up to October 2018 that evaluated nutrition, physical activity or obesity interventions in young adults (17-35 years). An effective intervention was one where the change in one or more primary outcome was positive and statistically significantly different from baseline, compared with control, or if no control comparator, compared with another active intervention. Effectiveness of outcomes was compared between gender targeted and gender-neutral studies.

Results: A total of 18,779 manuscripts were identified, and 94 RCTs were included. Sixty-seven studies were gender-neutral (71%) and 27 were gender targeted (29%, 20 female and seven male targeted). Primary outcome/s were adiposity (n=31, 33%), nutrition (n=26, 28%), or physical activity outcomes (n=25, 27%), or a combination (n=12, 13%). A greater proportion of gender targeted than gender-neutral studies were effective in improving nutrition (n=5, 100% and n=15, 71% of studies respectively) and physical activity outcomes (n=6, 86% and n=10, 56% respectively). A greater proportion of gender-neutral studies were effective in improving adiposity outcomes (n=11, 61% and n=5, 38% respectively). No differences were statistically significant.

Conclusions: Although differences in outcome effectiveness were identified between gender targeted and gender-neutral studies, these were not significantly different. This is likely due to an insufficient number of studies to detect a difference. However, the findings indicate a potential difference. To truly determine the effectiveness of gender targeted interventions, well-designed RCTs comparing gender targeted interventions with gender-neutral and control are needed.

Purpose: Children involved in organized sport report higher levels of physical activity (PA) than those who are not in sport. However, children who are meeting PA guidelines also have the unhealthiest diets. This may be due in part to the pervasive use of unhealthy “treats” as rewards within the culture of youth sport. Therefore, the purpose of this study was to explore the use of food as a reward related to youth sport participation. 

Methods: A constructivist paradigm with a relativist ontology guided this project. An instrumental case study explored food as reward in youth soccer. Criterion-based sampling was used to recruit children aged 4-12 (N = 24), parents (N = 7), and coaches (N = 6) within Ontario, Canada. Data generation included surveys about sport participation, focus groups with parents and children, and one-on-one interviews with coaches. Transcribed audio recordings underwent thematic analysis.

Results: Key themes included: the use of food (treats) as a reward for effort, time constraints dictating food choice, friends as motivators, and the role of sponsorship in youth sport. Parents, coaches, and athletes indicated food, typically from a fast-food restaurant chain because of limited time, is given as a treat or reward for motivating and praising effort. Children consistently reported their friendships with teammates as a primary motivator for participating in sport; parents and coaches explained that snacks foster those friendships and promote comradery. Monetary support of sponsorship was seen as critical to the success of youth sport, regardless of the sponsor. Parents and coaches indicated fast-food sponsorship did not influence their choice of rewards for their child. 

Conclusions: Although parents, coaches, and children agreed that participation and effort in soccer would continue without external motivators, food rewards are still offered to motivate and celebrate performance. Given the children claimed they are intrinsically motivated to play soccer, it begs the question of why treats are offered at all. Knowing that using food as a reward is not required to motivate children’s participation suggests a need to educate parents on what truly motivates their child, and how to celebrate success and effort in a healthier way. 

Purpose: Current physical activity guidelines encourage individuals to be active during any life domain. However, recent evidence shows that work-related physical activity does not have the same mental health benefits as leisure-time physical activity. This is particularly important given that many people participate in large amounts of work-related physical activity without engaging in any leisure-time physical activity. Work-related physical activity however is likely to include a variety of different behaviours for people with different occupations. As such, the purpose of this study was to determine if occupation type moderated the association between work-related physical activity and psychological distress.

Methods: A randomly selected sample of 1,080 women from Melbourne, Australia completed the International Physical Activity Questionnaire (IPAQ) and General Health Questionnaire (GHQ-30), and reported their current occupation.

Results: Linear regression analyses indicated that occupation significantly moderated the association between work-related walking and psychological distress (F [8, 55] = 2.26, p = .036). Given evidence of moderation, we fitted linear regression models to test the associations between work-related physical activity and psychological distress for three separate groups: professionals, sales and services workers, and tradespersons. Female tradespersons who engaged in a low (B = -3.81, p = .006) or high amount of work-related walking (B = -3.23, p = .029), had significantly lower psychological distress symptoms than those who engaged in no work-related walking. There were no significant associations between work-related physical activity of any intensity and psychological distress for professionals, or sales and service workers.

Conclusions: Physical activity at work is not consistently associated with psychological distress as the relationship varies between different occupations. As such, a number of mechanisms and potential moderators need to be investigated before we can truly understand how to best promote mental health through workplace physical activity. Nevertheless, walking at work for female tradespersons may reduce psychological distress, and therefore, work-related physical activity should not be completely discounted in terms of its potential for improving mental health.

Purpose: Health information is more available and accessible than ever. However, information from multiple sources is often conflicting, making it difficult for people to form judgments in order to make health-related decisions. Research in educational psychology has proposed the importance of epistemic beliefs as a way to explain how people process and comprehend information. In addition, research on the public understanding of science has proposed that the public’s understanding of the topic may depend on the way they explain the existence of the disagreement. Therefore, this research examines individuals’ experiences with conflicting information related to health and nutrition, their epistemic beliefs and the way they explain conflicts.

Methods: Cross-sectional data are collected through a survey administered to University students (n=188). Validated scales are used for the measurement of epistemic beliefs and explanations for scientific disagreement. We use correlation analyses to explore associations between the different measures and compare those with more versus less advanced epistemic beliefs.

Results/findings: We describe people’s experiences with conflicting information related to health and nutrition and explore associations between their epistemic beliefs and the attributes they use to explain scientific disagreement. Preliminary analysis indicates that those with more advanced beliefs use a wider variety of explanations for expert disagreement and are more likely to attribute such conflicts to the complexity and uncertainty of the scientific process. Associations with a set of covariates, including socio-demographic characteristics, health consciousness, and topic involvement are analysed.

Conclusions:  Implications for the potential role of a manipulation of epistemic beliefs to improve one’s ability to handle conflicting information are discussed.