C E Joshu1, T K Boehmer, R C Brownson, R Ewing. 1. 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@slu.edu
Abstract
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 (beta = 0.10) and neighbourhood (beta = 0.05) barriers had a stronger association than sprawl (beta = -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.
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 (beta = 0.10) and neighbourhood (beta = 0.05) barriers had a stronger association than sprawl (beta = -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.
Authors: Fuzhong Li; Peter Harmer; Bradley J Cardinal; Mark Bosworth; Deborah Johnson-Shelton; Jane M Moore; Alan Acock; Naruepon Vongjaturapat Journal: Am J Epidemiol Date: 2009-01-19 Impact factor: 4.897
Authors: Lindsay M Braun; Daniel A Rodríguez; Kelly R Evenson; Jana A Hirsch; Kari A Moore; Ana V Diez Roux Journal: Health Place Date: 2016-03-01 Impact factor: 4.078
Authors: Philip J Troped; Heather A Starnes; Robin C Puett; Kosuke Tamura; Ellen K Cromley; Peter James; Eran Ben-Joseph; Steven J Melly; Francine Laden Journal: J Aging Phys Act Date: 2013-03-26 Impact factor: 1.961