Associations between food environment typologies and body mass index: Evidence from Yorkshire, England. (October 2019)

• Record Type: Journal Article
• Authors: Hobbs, M.; Green, M., A.; Wilkins, E.; Lamb, K.E.; McKenna, J.; Griffiths, C.
• Is Part Of: Social science & medicine. Volume 239(2019)
• Publication Date: 2019-10
• Subjects: Food outlets -- Longitudinal -- Body mass index -- Food environment -- Obesity -- Obesogenic environment; Social medicine -- Periodicals; Medical anthropology -- Periodicals; Public health -- Periodicals; Psychology -- Periodicals; Medicine -- Periodicals; Medicine -- Periodicals; Médecine sociale -- Périodiques; Anthropologie médicale -- Périodiques; Santé publique -- Périodiques; Psychologie -- Périodiques; Médecine -- Périodiques; Electronic journals; 362.105
• DOI: 10.1016/j.socscimed.2019.112528
• Languages: English
• ISSNs: 0277-9536

Abstract

Abstract: International research linking food outlets and body mass index (BMI) is largely cross-sectional, yielding inconsistent findings. However, addressing the exposure of food outlets is increasingly considered as an important adult obesity prevention strategy. Our study investigates associations between baseline food environment types and change in BMI over time. Survey data were used from the Yorkshire Health Study (n=8, 864; wave one: 2010-2012, wave two: 2013-2015) for adults aged 18-86. BMI was calculated using self-reported height (cm) and weight (kg). Restaurants, cafés, fast-food, speciality, convenience and large supermarkets were identified from the Ordnance Survey Point of Interest database within 1600m radial buffer of home postcodes. K-means cluster analysis developed food environment typologies based on food outlets and population density. Large supermarkets, restaurants, cafés, fast-food, speciality and convenience food outlets all clustered together to some extent. Three neighbourhood typologies were identified. However, multilevel models revealed that relative to cluster one all were unrelated to change in BMI (cluster 2, b= -0.146 [-0.274, 0.566]; cluster 3, b= 0.065 [-0.224, 0.356]). There was also little evidence of gender-based differences in these associations when examined in a three-way interaction. Policymakers may need to begin to consider multiple types of food outlet clusters, while further research is needed to confirm how these relate to changed BMI. Highlights: Developed a typology of neighbourhood food environments by population density. Considers if food environments were associated with change in BMI by gender. Supermarkets, restaurants, cafés, fast-food, speciality and convenience outlets clustered. A focus on multidimensional food environments may be warranted in policy and research. Food environments were unrelated to change in BMI over time with no difference by gender.