Concurrent assessment of urban environment and cardiometabolic risk over 10 years in a middle‐aged population‐based cohort. (8th November 2018)

• Record Type: Journal Article
• Authors: Daniel, Mark; Carroll, Suzanne J.; Niyonsenga, Theophile; Piggott, Ellie J.; Taylor, Anne; Coffee, Neil T.
• Is Part Of: Geographical research. Volume 57:Number 1(2019)
• Page Start: 98
• Page End: 110
• Publication Date: 2018-11-08
• Subjects: built environment -- walkability -- public open space -- food environment -- area‐level wealth -- cardiometabolic risk; Geography -- Research -- Periodicals; Geography -- Australasia -- Periodicals; Human geography -- Periodicals; Physical geography -- Periodicals; 304.2072
• DOI: 10.1111/1745-5871.12318
• Languages: English
• ISSNs: 1745-5863

Abstract

Abstract: Inference regarding the impact of urban areas on health is limited by cross‐sectional studies assessing few dimensions and ignoring area‐level socio‐economic status. This study simultaneously assessed several dimensions of the built environment against incident cardiometabolic risk (CMR) arising over 10 years. It tested the hypothesis that, accounting for local area relative wealth, features of the built environment would not predict incident CMR. Initially, disease‐free adults in a biomedical cohort in Adelaide, Australia, provided address and clinical data over three waves of follow‐up. CMR was defined as the count of five clinical CMR factors. Built environment measures were derived for urban form, and natural, and food environments. Local area wealth was expressed using the relative location factor index. Poisson growth models accounting for within‐suburb clustering, age, sex, and education were used to estimate associations between built environment measures and increasing CMR. Fitted linear trajectories had statistically significant mean values of intercepts and slopes. CMR trajectories were associated with age, male sex, and low education. In models including measures of the food, natural, and urban form environments, per standard deviation increase, only POS count predicted incident CMR, which was more strongly predicted by relative location factor. Not accounting for local area socio‐economic status may overestimate the strength of relationships between health and the built environment. Inequity in accessible POS is robustly related to incident CMR.