Purpose: Car driving contributes to physical inactivity, sedentary behaviour, congestion, air- and noise pollution. Environment-level factors can be important targets for shifting population behaviors away from car use and towards active forms of transport. A comprehensive index combining multiple environmental factors that drive car use in Toronto, Canada, was strongly associated with car use. However, no such index is available for a European setting. This study aimed to develop a neighborhood drivability index for the Netherlands, and to investigate associations with car use for different purposes and trip lengths.

Methods: In this cross-sectional study, we considered candidate variables for inclusion in our index, related to car that capture density, diversity, design, destination accessibility, distance to transit, and demand management. Geographical data were obtained for all Dutch 4-digit postal code areas (n=4,066). The index was tested against a  nationwide travel survey conducted in 2017 (n=21,376). We used a training set (random 2/3 of survey dataset) to create a composite drivability index by identifying environmental variables that predicted car use, based on a logistic prediction model with car use (yes/no) as the primary outcome. In a testing set (1/3 of the dataset), logistic regression was used to assess the association between neighborhood drivability as a standardized score (0-100) and car use, overall and according to trip characteristics (long/short distances, discretionary/non-discretionary purposes), adjusted for sociodemographic and other relevant confounders. We investigated effect modification by neighborhood-level socio-economic status and urbanicity.

Results: The drivability index consisted of land-use mix, population density, paid parking, public transit stops, and train stations; and predicted car use fairly well (Area Under the Curve: 0.63). The median drivability was 54.36 (IQR:14.25). A 1% higher neighborhood drivability score was associated with a higher odds of car use (OR:1.04; 95%CI:1.03-1.04). This association was stronger for discretionary trips (OR:1.04; 95%CI:1.03-1.05) than non-discretionary trips (OR:1.00; 95%CI:0.99-1.00). There was no difference between short (OR:1.03; 95%CI:1.02-1.03) and long trips (OR:1.04; 95%CI:1.03-1.04). No effect modification was observed.

Conclusions: This first nationwide drivability index for the Netherlands was associated with car use in general, and especially for discretionary trips. This could inform targeted transport and urban design policies.

Purpose:

To increase the enactment of evidence-based health policy, it is essential to build trusting relationships between public health researchers and policymakers. Researchers and policymakers have different timelines, priorities, and values, which necessitates a process that includes training and development of a non-partisan network. The Texas Research-to-Policy Collaboration (TX RPC) Project aims to develop and support partnerships between health researchers and policymakers, using learnings from a similar federal model. This presentation describes the adaptation of the federal RPC model to the Texas legislative process, and presents data on recruitment, training, and evaluation. This project uses a framework rooted in Agenda Setting Theory, with a focus on translational methods.

Methods:

The RPC model was modified to align with the Texas legislative cycle and prioritize child health legislation and state-level research related to nutrition, physical activity, obesity, and tobacco/e-cigarette use. Adaptation of project tools was informed by an advisory committee, community partners, and prior research with legislators to reflect state policymaking values and protocols. Researchers and legislators were recruited through various channels. Researchers completed baseline and post-training surveys; legislators completed a baseline interview and policy identification needs assessment. Pre- and post-survey data were analyzed to determine changes in knowledge and self-efficacy. Data from the interviews and needs assessment were compiled and categorized according to policy themes.

Results/findings:

61 researchers enrolled as network members, and 59 researchers were trained; 21 legislators participated in the interviews. Researchers’ self-efficacy in communicating with policymakers and policy knowledge significantly improved after training. Baseline interviews indicate legislators desire expert knowledge and data on several health issues. During the 87^th Texas Legislative Session in January-May 2021, TX RPC researchers/staff responded to 91 legislative requests, provided testimony on 3 bills, and held 31 collaboration meetings with legislators. TX RPC legislators filed 19 bills impacting child health; 8 bills passed at least one chamber of the Texas Legislature and 3 passed into law. ​

Conclusions:

Initial recruitment and project activities indicate a need for training researchers and providing data and related support to legislators. Developing public health researcher-policymaker partnerships at the state level can be an effective model to implement evidence-based policy.

Background: To promote the success of interventions it is essential to match them to the community’s level of readiness. Community Readiness (CR) is the degree to which a community is prepared to address a health issue and measured in the community’s knowledge of the issue and existing efforts, leadership, resources, and the community climate. Aim of this study is to analyse CR for the prevention of childhood obesity in municipalities using the Community Readiness Model (CRM). Based on this, participatory strategies to improve CR for obesity prevention in children will be developed.

Methods: A systematic literature review (SLR) was conducted searching the databases PubMed, LIVIVO, Cochrane and Google Scholar. The SLR included studies that used the CRM for the prevention of childhood obesity to analyse the application in practice. To assess CR, guided interviews with key informants from five Bavarian municipalities were carried out. Interviewees were identified in a modified stakeholder analysis. The transcribed interviews were analysed by two researchers following the CRM scoring system.

Results: A total of 285 records were identified in the SLR. After review of titles, abstracts, and full texts, 17 studies remained. The CRM has already been applied to childhood obesity prevention in the USA (n = 10), Australia (n = 4), the UK (n = 1), Iran (n = 1), and South Africa (n = 1). In Germany, no studies could be identified. From January to April 2021, semi-structured interviews (n = 27) with key informants from the participating municipalities were conducted. This sample comprised administration departments, medical, health and educational facilities. The municipalities average CR score reached 3.8 on a 9-point rating scale. According to the CRM, this corresponds to the "Vague Awareness" stage of readiness.

Conclusion: The CR assessment showed that childhood obesity is a concern in municipalities without an immediate motivation to actively address it. Municipalities have only vague knowledge about childhood obesity and there are limited resources that could be used for prevention efforts. These results will be reflected in a workshop with each participating municipality. Aim is to develop an action plan to increase CR for childhood obesity prevention.

Purpose: Ecological models posit that interventions should be most potent when environments support the target behavior. This study was powered a priori to examine interactions between walkability and mHealth intervention components in a large-scale trial to increase PA.

Methods: Healthy, inactive adults (n=512) were randomized to WalkIT Arizona, a 12-month, 2x2 factorial mHealth intervention evaluating goal setting (adaptive versus static) and financial reinforcement timing (immediate versus delayed) to increase accelerometer-measured MVPA. Participants were recruited from neighborhoods based on GIS-measured walkability index (high/low) and socioeconomic status (SES, high/low) strata and block randomized into one of four interventions. Participants wore an ActiGraph GT9X Link daily for a year. After recruitment, walkability (summed Z-scores for residential density, land use mix, intersection density, and transit access) was calculated within a 500m street network buffer for each participant’s home. Generalized linear mixed effects hurdle models examined 1) likelihood of any (versus no) MVPA and 2) daily MVPA minutes based on the joint effects of walkability, goal type, reinforcement timing, and time, adjusting for accelerometer wear time, SES, and calendar month. Walkability was probed at 5^th, 25^th, 50^th, 75^th, and 95^th percentiles to explore the full range of effects.

Results: Walkability moderated effects of goal type for any MVPA (OR = 1.08, 95% CI: 1.01-1.15) and MVPA duration (OR=0.97, 95% CI: 0.95-0.99). A larger between-group difference in likelihood of any PA for adaptive versus static goals was observed; this effect was largest for low walkability (5^th, 25^th percentile) and disappeared at the 95^th percentile. Between-group differences in daily MVPA minutes were greater in low relative to high walkability areas favoring adaptive goals, although the reverse pattern was observed for static goals. For reinforcement timing, walkability moderated intervention effects for likelihood of any MVPA (OR=0.91, 95% CI: 0.85-0.97): immediate reinforcement showed a stronger increase in high relative to low walkability neighborhoods. There was no significant walkability x reinforcement x time interaction for MVPA duration.

Conclusions: Results show interactions between walkability and PA interventions depended on the intervention parameters, suggesting ecological model hypotheses may need to be refined.