Open Access Highly Accessed Research

Patterns of neighborhood environment attributes related to physical activity across 11 countries: a latent class analysis

Marc A Adams1*, Ding Ding18, James F Sallis2, Heather R Bowles3, Barbara E Ainsworth1, Patrick Bergman5, Fiona C Bull6, Harriette Carr9, Cora L Craig7, Ilse De Bourdeaudhuij10, Luis Fernando Gomez11, Maria Hagströmer4, Lena Klasson-Heggebø8, Shigeru Inoue12, Johan Lefevre13, Duncan J Macfarlane14, Sandra Matsudo15, Victor Matsudo15, Grant McLean9, Norio Murase12, Michael Sjöström4, Heidi Tomten16, Vida Volbekiene17 and Adrian Bauman18

Author Affiliations

1 Exercise and Wellness, School of Nutrition and Health Promotion, Arizona State University, Phoenix, AZ, USA

2 Active Living Research, University of California, San Diego, CA, USA

3 Risk Factor Monitoring and Methods Branch, Applied Research Program, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA

4 Unit for Preventive Nutrition, Department of Biosciences and Nutrition, Karolinska Institutet, Stockholm, Sweden

5 School of Education, Psychology and Sports Science, Linnaeus University, Kalmar, Sweden

6 School of Population Health, The University of Western Australia, Crawley, WA, Australia

7 Canadian Fitness and Lifestyle Research Institute, School of Public Health, Ottawa, Canada

8 Valnesfjord Rehabilitation Centre, Valnesfjord, Norway

9 Sport New Zealand, Ministry of Health, Wellington, New Zealand

10 Department of Movement and Sport Sciences, Ghent University, Ghent, Belgium

11 Pontificia Universidad Javeriana, Bogota, Colombia

12 Department of Preventive Medicine and Public Health, Tokyo Medical University, Tokyo, Japan

13 Department of Kinesiology, Katholic University, Leuven, Belgium

14 Institute of Human Performance, The University of Hong Kong, Hong Kong, China

15 Center of Studies of the Physical Fitness Research Center from São Caetano do Sul, CELAFISCS, Sao Paulo, Brazil

16 Oppegård Municipality, Oppegård County, Norway

17 Department of Sport Science, Lithuanian Academy of Physical Education, Kaunas, Lithuania

18 Prevention Research Collaboration, University of Sydney, Sydney, Australia

For all author emails, please log on.

International Journal of Behavioral Nutrition and Physical Activity 2013, 10:34  doi:10.1186/1479-5868-10-34

Published: 14 March 2013

Abstract

Background

Neighborhood environment studies of physical activity (PA) have been mainly single-country focused. The International Prevalence Study (IPS) presented a rare opportunity to examine neighborhood features across countries. The purpose of this analysis was to: 1) detect international neighborhood typologies based on participants’ response patterns to an environment survey and 2) to estimate associations between neighborhood environment patterns and PA.

Methods

A Latent Class Analysis (LCA) was conducted on pooled IPS adults (N=11,541) aged 18 to 64 years old (mean=37.5 ±12.8 yrs; 55.6% women) from 11 countries including Belgium, Brazil, Canada, Colombia, Hong Kong, Japan, Lithuania, New Zealand, Norway, Sweden, and the U.S. This subset used the Physical Activity Neighborhood Environment Survey (PANES) that briefly assessed 7 attributes within 10–15 minutes walk of participants’ residences, including residential density, access to shops/services, recreational facilities, public transit facilities, presence of sidewalks and bike paths, and personal safety. LCA derived meaningful subgroups from participants’ response patterns to PANES items, and participants were assigned to neighborhood types. The validated short-form International Physical Activity Questionnaire (IPAQ) measured likelihood of meeting the 150 minutes/week PA guideline. To validate derived classes, meeting the guideline either by walking or total PA was regressed on neighborhood types using a weighted generalized linear regression model, adjusting for gender, age and country.

Results

A 5-subgroup solution fitted the dataset and was interpretable. Neighborhood types were labeled, “Overall Activity Supportive (52% of sample)”, “High Walkable and Unsafe with Few Recreation Facilities (16%)”, “Safe with Active Transport Facilities (12%)”, “Transit and Shops Dense with Few Amenities (15%)”, and “Safe but Activity Unsupportive (5%)”. Country representation differed by type (e.g., U.S. disproportionally represented “Safe but Activity Unsupportive”). Compared to the Safe but Activity Unsupportive, two types showed greater odds of meeting PA guideline for walking outcome (High Walkable and Unsafe with Few Recreation Facilities, OR= 2.26 (95% CI 1.18-4.31); Overall Activity Supportive, OR= 1.90 (95% CI 1.13-3.21). Significant but smaller odds ratios were also found for total PA.

Conclusions

Meaningful neighborhood patterns generalized across countries and explained practical differences in PA. These observational results support WHO/UN recommendations for programs and policies targeted to improve features of the neighborhood environment for PA.

Keywords:
Built environment; International; Recreation; Surveillance; Exercise