HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS REGISTER		Author Keyword(s) Vol Page [Advanced]

This Article Full Text Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this link to a friend Similar articles in this journal Similar articles in PubMed Add article to my folders Download to citation manager Request Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Joshu, C E Articles by Ewing, R Search for Related Content PubMed PubMed Citation Articles by Joshu, C E Articles by Ewing, R Topic Collections Related Article

Personal, neighbourhood and urban factors associated with obesity in the United States

¹ Department of Community Health and Prevention Research Center, Saint Louis University School of Public Health, St Louis, Missouri, USA
² National Center for Smart Growth Education and Research, University of Maryland, College Park, Maryland, USA

Correspondence to:
C E Joshu, Department of Community Health, Prevention Research Center, Saint Louis University School of Public Health, Salus Center, 3545 Lafayette Avenue, St Louis, MO 63104, USA; joshuce{at}slu.edu

Introduction: Growing evidence suggests the built environment impacts obesity within urban areas; however, little research has investigated these relationships across levels of urbanisation in diverse and representative populations. This study aimed to determine whether personal and neighbourhood barriers differ by the level of urbanisation and the relative importance of personal barriers, neighbourhood barriers and land-use development patterns measured by a county-level sprawl index.

Methods: Population-based, cross-sectional telephone survey data were collected on 1818 United States adults of diverse ethnicity and income level. Primary analyses were stratified by the level of urbanisation at the county level (large metropolitan, small metropolitan, non-metro, rural). Associations between obesity and neighbourhood and personal barriers were estimated with logistic regression, controlling for demographic variables. Within metropolitan areas, the association between body mass index (BMI) and county-level sprawl was estimated using hierarchical linear modelling, controlling for individual-level neighbourhood and personal barriers and demographic variables and then assessing cross-level interaction.

Results: The prevalence of neighbourhood, but not personal, barriers differed widely across levels of urbanisation. Specific neighbourhood (eg traffic, unattended dogs) and personal (eg time, injury) barriers differentially correlated with obesity across strata. The impact of sprawl on BMI (B = –0.005) was consistent with previous findings; standardised coefficients indicate that personal (β = 0.10) and neighbourhood (β = 0.05) barriers had a stronger association than sprawl (β = –0.02). Furthermore, the effect of sprawl on BMI increased by –0.006 with each additional personal barrier.

Conclusions: Future intervention planning and policy development should consider that personal barriers and built environment characteristics may interact with each other and influence obesity differently across urbanisation levels.