Associations of built environment and proximity of food outlets with weight status

Analysis from 14 cities in 10 countries

Thomas Cochrane, Yan Yu, Rachel Davey, Ester Cerin, Kelli L Cain, Terry L Conway, Jacqueline Kerr, Lawrence D Frank, James E Chapman, Marc A Adams, Duncan Macfarlane, Delfien Van Dyck, Poh-Chin Lai, Olga L Sarmiento, Jens Troelsen, Deborah Salvo, Rodrigo Reis, Josef Mitáš, Grant Schofield, Neville Owen & 1 others James F Sallis

Research output: Contribution to journalJournal articleResearchpeer-review

29 Downloads (Pure)

Abstract

The study aimed to examine associations of neighborhood built environments and proximity of food outlets (BE measures) with body weight status using pooled data from an international study (IPEN Adult). Objective BE measures were calculated using geographic information systems for 10,008 participants (4463 male, 45%) aged 16-66 years in 14 cities. Participants self-reported proximity to three types of food outlets. Outcomes were body mass index (BMI) and overweight/obesity status. Male and female weight status associations with BE measures were estimated by generalized additive mixed models. Proportion (95% CI) of overweight (BMI 25 to <30) ranged from 16.6% (13.1, 19.8) to 41.1% (37.3, 44.7), and obesity (BMI ≥ 30) from 2.9% (1.3, 4.4) to 31.3% (27.7, 34.7), with Hong Kong being the lowest and Cuernavaca, Mexico highest for both proportions. Results differed by sex. Greater street intersection density, public transport density and perceived proximity to restaurants (males) were associated with lower odds of overweight/obesity (BMI ≥ 25). Perceived proximity to public transport stops (females) was associated with higher odds of overweight/obesity. Composite BE measures were more strongly related to BMI and overweight/obesity status than single variables among men but not women. One standard deviation improvement in the composite measures of BE was associated with small reductions of 0.1-0.5% in BMI but meaningful reductions of 2.5-5.3% in the odds of overweight/obesity. Effects were linear and generalizable across cities. Neighborhoods designed to support public transport, with food outlets within walking distance, may contribute to global obesity control.

Original languageEnglish
Article number105874
JournalPreventive Medicine
Volume129
Number of pages13
ISSN0091-7435
DOIs
Publication statusPublished - 22. Oct 2019

Fingerprint

Weights and Measures
Food
Body Mass Index
Restaurants
Geographic Information Systems
Hong Kong
Mexico

Cite this

Cochrane, Thomas ; Yu, Yan ; Davey, Rachel ; Cerin, Ester ; Cain, Kelli L ; Conway, Terry L ; Kerr, Jacqueline ; Frank, Lawrence D ; Chapman, James E ; Adams, Marc A ; Macfarlane, Duncan ; Van Dyck, Delfien ; Lai, Poh-Chin ; Sarmiento, Olga L ; Troelsen, Jens ; Salvo, Deborah ; Reis, Rodrigo ; Mitáš, Josef ; Schofield, Grant ; Owen, Neville ; Sallis, James F. / Associations of built environment and proximity of food outlets with weight status : Analysis from 14 cities in 10 countries. In: Preventive Medicine. 2019 ; Vol. 129.
@article{d869e5ea4da24491b45e0de318163041,
title = "Associations of built environment and proximity of food outlets with weight status: Analysis from 14 cities in 10 countries",
abstract = "The study aimed to examine associations of neighborhood built environments and proximity of food outlets (BE measures) with body weight status using pooled data from an international study (IPEN Adult). Objective BE measures were calculated using geographic information systems for 10,008 participants (4463 male, 45{\%}) aged 16-66 years in 14 cities. Participants self-reported proximity to three types of food outlets. Outcomes were body mass index (BMI) and overweight/obesity status. Male and female weight status associations with BE measures were estimated by generalized additive mixed models. Proportion (95{\%} CI) of overweight (BMI 25 to <30) ranged from 16.6{\%} (13.1, 19.8) to 41.1{\%} (37.3, 44.7), and obesity (BMI ≥ 30) from 2.9{\%} (1.3, 4.4) to 31.3{\%} (27.7, 34.7), with Hong Kong being the lowest and Cuernavaca, Mexico highest for both proportions. Results differed by sex. Greater street intersection density, public transport density and perceived proximity to restaurants (males) were associated with lower odds of overweight/obesity (BMI ≥ 25). Perceived proximity to public transport stops (females) was associated with higher odds of overweight/obesity. Composite BE measures were more strongly related to BMI and overweight/obesity status than single variables among men but not women. One standard deviation improvement in the composite measures of BE was associated with small reductions of 0.1-0.5{\%} in BMI but meaningful reductions of 2.5-5.3{\%} in the odds of overweight/obesity. Effects were linear and generalizable across cities. Neighborhoods designed to support public transport, with food outlets within walking distance, may contribute to global obesity control.",
author = "Thomas Cochrane and Yan Yu and Rachel Davey and Ester Cerin and Cain, {Kelli L} and Conway, {Terry L} and Jacqueline Kerr and Frank, {Lawrence D} and Chapman, {James E} and Adams, {Marc A} and Duncan Macfarlane and {Van Dyck}, Delfien and Poh-Chin Lai and Sarmiento, {Olga L} and Jens Troelsen and Deborah Salvo and Rodrigo Reis and Josef Mit{\'a}š and Grant Schofield and Neville Owen and Sallis, {James F}",
note = "Copyright {\circledC} 2019. Published by Elsevier Inc.",
year = "2019",
month = "10",
day = "22",
doi = "10.1016/j.ypmed.2019.105874",
language = "English",
volume = "129",
journal = "Preventive Medicine",
issn = "0091-7435",
publisher = "Elsevier",

}

Cochrane, T, Yu, Y, Davey, R, Cerin, E, Cain, KL, Conway, TL, Kerr, J, Frank, LD, Chapman, JE, Adams, MA, Macfarlane, D, Van Dyck, D, Lai, P-C, Sarmiento, OL, Troelsen, J, Salvo, D, Reis, R, Mitáš, J, Schofield, G, Owen, N & Sallis, JF 2019, 'Associations of built environment and proximity of food outlets with weight status: Analysis from 14 cities in 10 countries', Preventive Medicine, vol. 129, 105874. https://doi.org/10.1016/j.ypmed.2019.105874

Associations of built environment and proximity of food outlets with weight status : Analysis from 14 cities in 10 countries. / Cochrane, Thomas; Yu, Yan; Davey, Rachel; Cerin, Ester; Cain, Kelli L; Conway, Terry L; Kerr, Jacqueline; Frank, Lawrence D; Chapman, James E; Adams, Marc A; Macfarlane, Duncan; Van Dyck, Delfien; Lai, Poh-Chin; Sarmiento, Olga L; Troelsen, Jens; Salvo, Deborah; Reis, Rodrigo; Mitáš, Josef; Schofield, Grant; Owen, Neville; Sallis, James F.

In: Preventive Medicine, Vol. 129, 105874, 22.10.2019.

Research output: Contribution to journalJournal articleResearchpeer-review

TY - JOUR

T1 - Associations of built environment and proximity of food outlets with weight status

T2 - Analysis from 14 cities in 10 countries

AU - Cochrane, Thomas

AU - Yu, Yan

AU - Davey, Rachel

AU - Cerin, Ester

AU - Cain, Kelli L

AU - Conway, Terry L

AU - Kerr, Jacqueline

AU - Frank, Lawrence D

AU - Chapman, James E

AU - Adams, Marc A

AU - Macfarlane, Duncan

AU - Van Dyck, Delfien

AU - Lai, Poh-Chin

AU - Sarmiento, Olga L

AU - Troelsen, Jens

AU - Salvo, Deborah

AU - Reis, Rodrigo

AU - Mitáš, Josef

AU - Schofield, Grant

AU - Owen, Neville

AU - Sallis, James F

N1 - Copyright © 2019. Published by Elsevier Inc.

PY - 2019/10/22

Y1 - 2019/10/22

N2 - The study aimed to examine associations of neighborhood built environments and proximity of food outlets (BE measures) with body weight status using pooled data from an international study (IPEN Adult). Objective BE measures were calculated using geographic information systems for 10,008 participants (4463 male, 45%) aged 16-66 years in 14 cities. Participants self-reported proximity to three types of food outlets. Outcomes were body mass index (BMI) and overweight/obesity status. Male and female weight status associations with BE measures were estimated by generalized additive mixed models. Proportion (95% CI) of overweight (BMI 25 to <30) ranged from 16.6% (13.1, 19.8) to 41.1% (37.3, 44.7), and obesity (BMI ≥ 30) from 2.9% (1.3, 4.4) to 31.3% (27.7, 34.7), with Hong Kong being the lowest and Cuernavaca, Mexico highest for both proportions. Results differed by sex. Greater street intersection density, public transport density and perceived proximity to restaurants (males) were associated with lower odds of overweight/obesity (BMI ≥ 25). Perceived proximity to public transport stops (females) was associated with higher odds of overweight/obesity. Composite BE measures were more strongly related to BMI and overweight/obesity status than single variables among men but not women. One standard deviation improvement in the composite measures of BE was associated with small reductions of 0.1-0.5% in BMI but meaningful reductions of 2.5-5.3% in the odds of overweight/obesity. Effects were linear and generalizable across cities. Neighborhoods designed to support public transport, with food outlets within walking distance, may contribute to global obesity control.

AB - The study aimed to examine associations of neighborhood built environments and proximity of food outlets (BE measures) with body weight status using pooled data from an international study (IPEN Adult). Objective BE measures were calculated using geographic information systems for 10,008 participants (4463 male, 45%) aged 16-66 years in 14 cities. Participants self-reported proximity to three types of food outlets. Outcomes were body mass index (BMI) and overweight/obesity status. Male and female weight status associations with BE measures were estimated by generalized additive mixed models. Proportion (95% CI) of overweight (BMI 25 to <30) ranged from 16.6% (13.1, 19.8) to 41.1% (37.3, 44.7), and obesity (BMI ≥ 30) from 2.9% (1.3, 4.4) to 31.3% (27.7, 34.7), with Hong Kong being the lowest and Cuernavaca, Mexico highest for both proportions. Results differed by sex. Greater street intersection density, public transport density and perceived proximity to restaurants (males) were associated with lower odds of overweight/obesity (BMI ≥ 25). Perceived proximity to public transport stops (females) was associated with higher odds of overweight/obesity. Composite BE measures were more strongly related to BMI and overweight/obesity status than single variables among men but not women. One standard deviation improvement in the composite measures of BE was associated with small reductions of 0.1-0.5% in BMI but meaningful reductions of 2.5-5.3% in the odds of overweight/obesity. Effects were linear and generalizable across cities. Neighborhoods designed to support public transport, with food outlets within walking distance, may contribute to global obesity control.

U2 - 10.1016/j.ypmed.2019.105874

DO - 10.1016/j.ypmed.2019.105874

M3 - Journal article

VL - 129

JO - Preventive Medicine

JF - Preventive Medicine

SN - 0091-7435

M1 - 105874

ER -