School- and intervention-related factors associated with institutionalization of health promotion interventions in elementary schools. / Facteurs liés aux écoles et facteurs liés aux interventions dans le contexte de l'institutionnalisation des interventions de promotion de la santé dans les écoles primaires.

INTRODUCTION: Long-term availability of health-promoting interventions (HPIs) in school settings can translate into health benefits for children. However, little is known about factors associated with HPI institutionalization in schools. In this study, we identified correlates of the institutionalization of HPIs offered in elementary schools in Quebec, Canada. METHODS: In two-part, structured telephone interviews over three academic years (2016-2019), elementary school principals (or their designees) throughout Quebec identified an index HPI offered at least once in their school during the previous three years, and were asked whether it was institutionalized (i.e. explicitly written in the school's educational project, e.g. in the form of educational objectives and means of achieving them). We examined associations between institutionalization and 10 school-related and 16 HPI-related characteristics in univariable and multivariable logistic regression analyses. RESULTS: School key informants (n = 163) reported on 147 different HPIs that had been available in their schools in the past three years, 56% of which were institutionalized. Three aspects of school culture-parent/community engagement with the school, school/teacher commitment to student health and school physical environment-were positively associated with HPI institutionalization. HPI-related characteristics positively associated with HPI institutionalization included number of competencies addressed by the HPI, number of teaching strategies employed, modifications made to the HPI prior to or during implementation and perceived success of the HPI. Inviting families or community groups to participate in the HPI was inversely associated with institutionalization. CONCLUSION: Better understanding of factors associated with HPI institutionalization may inform the development of school-based HPIs that have the potential for sustainability.

Lack of consistency in measurement methods and semantics used for network measures in adolescent health behaviour studies using social network analysis: a systematic review.

BACKGROUND: Social network analysis (SNA) is often used to examine how social relationships influence adolescent health behaviours, but no study has documented the range of network measures used to do so. We aimed to identify network measures used in studies on adolescent health behaviours. METHODS: We conducted a systematic review to identify network measures in studies investigating adolescent health behaviours with SNA. Measures were grouped into eight categories based on network concepts commonly described in the literature: popularity, position within the network, network density, similarity, nature of relationships, peer behaviours, social norms, and selection and influence mechanisms. Different subcategories were further identified. We detailed all distinct measures and the labels used to name them in included articles. RESULTS: Out of 6686 articles screened, 201 were included. The categories most frequently investigated were peer behaviours (n=201, 100%), position within the network (n=144, 71.6%) and popularity (n=110, 54.7%). The number of measurement methods varied from 1 for 'similarity on popularity' (within the 'similarity' category) to 28 for the 'characterisation of the relationship between the respondent and nominated peers' (within the 'nature of the relationships' category). Using the examples of 'social isolation', 'group membership', 'individuals in a central position' (within the 'position within the network' category) and 'nominations of influential peers' (sub within the 'popularity' category), we illustrated the inconsistent reporting and heterogeneity in measurement methods and semantics. CONCLUSION: Robust methodological recommendations are needed to harmonise network measures in order to facilitate comparison across studies and optimise public health intervention based on SNA.

Prospective associations between neighborhood features and body mass index in Montreal adolescents.

PURPOSE: To investigate the association between the neighborhood built environment and trajectories of body mass index (BMI) in youth. METHODS: Data were collected in a prospective study of 1293 adolescents in Montreal. Built environment variables were obtained from public databases for road networks, land use, and the Canadian Census. Anthropometric data were collected when participants were ages 12.5, 15 and 17 years. We undertook hierarchical cluster analysis to identify contrasting neighborhood types based on features of the built environment (e.g., vegetation, population density, walkability). Associations between neighborhood type and trajectories of BMI z-score (BMIz) were estimated using multivariable linear mixed regression analyses, stratified by sex. RESULTS: We identified three neighborhood types: Urban, Suburban, and Village. In contrast to the Urban type, the Suburban type was characterized by more vegetation, few services and low population density. Village and Suburban types were similar, but the former had greater land use diversity, population density with more parks and a denser food environment. Among girls, living in Urban types was associated with decreasing BMIz trajectories. Living in Village types was associated with increasing BMIz trajectories. No associations were observed among boys. CONCLUSIONS: Neighborhoods characterized by greater opportunities for active living appear to be less obesogenic, particularly among girls.