Elsevier

Social Science & Medicine

Volume 72, Issue 8, April 2011, Pages 1266-1273
Social Science & Medicine

Neighborhood walkability, physical activity, and walking behavior: The Swedish Neighborhood and Physical Activity (SNAP) study

https://doi.org/10.1016/j.socscimed.2011.03.004Get rights and content

Abstract

More knowledge concerning the association between physical activity and objectively measured attributes of the built environment is needed. Previous studies on the association between objectively measured neighborhood walkability, physical activity, and walking have been conducted in the U.S. or Australia and research findings are available from only one country in Europe – Belgium. The first aim of this Swedish study of 2269 adults was to examine the associations between neighborhood walkability and walking for active transportation or leisure, and moderate-to-vigorous physical activity (MVPA) and whether these hypothesized associations are moderated by age, gender, income, marital status and neighborhood-level socioeconomic status. The second aim was to determine how much of the total variance of the walking and physical activity outcomes can be attributed to neighborhood-level differences. Neighborhood walkability was objectively measured by GIS methods. An index consisting of residential density, street connectivity, and land use mix was constructed to define 32 highly and less walkable neighborhoods in Stockholm City. MVPA was measured objectively during 7 days with an accelerometer and walking was assessed using the validated International Physical Activity Questionnaire (IPAQ). Multilevel linear as well as logistic models (mixed-effects, mixed-distribution models) were used in the analysis. The statistically significant and “adjusted” results for individuals living in highly walkable neighborhoods, as compared to those living in less walkable neighborhoods, were: (1) 77% and 28% higher odds for walking for active transportation and walking for leisure, respectively, (2) 50 min more walking for active transportation/week, and (3) 3.1 min more MVPA/day. The proportion of the total variance at the neighborhood level was low and ranged between 0.0% and 2.1% in the adjusted models. The findings of the present study stress that future policies concerning the built environment must be based on context-specific evidence, particularly in the light of the fact that neighborhood redevelopments are time-consuming and expensive.

Introduction

Physical activity is associated with a number of positive health outcomes, such as increased longevity (Sundquist, Qvist, Sundquist, & Johansson, 2004) and decreased risks of cardiovascular disease (Sesso et al., 2000, Sundquist, Johansson et al., 2005, Sundquist, Qvist et al., 2005), diabetes mellitus type 2 (Burchfiel et al., 1995, Manson et al., 1992), and some types of cancer (Ratnasinghe, Modali, Seddon, & Lehman, 2010). The World Health Organization (WHO) has stated that it is important to increase people’s levels of physical activity in order to decrease the global burden of these widespread diseases (WHO, 2010).

However, it is difficult to change people’s behavior towards a healthier lifestyle, such as including more physical activity in their daily lives. Physical activity is influenced by a complex array of personal, behavior-specific, socioenvironmental and physical environmental factors (Giles-Corti, Timperio, Bull, & Pikora, 2005). Recently, there has been an increasing focus on studies of the association between physical activity and attributes of the built environment, such as neighborhood walkability. However, few studies have been based on objectively measured neighborhood walkability. Previous studies on the association between objective neighborhood walkability, physical activity, and walking were conducted in the U.S. (Sallis, Saelens, Frank, Conway, Slymen, Cain et al., 2009) or Australia (Owen et al., 2007) and research findings are available from only one country in Europe – Belgium (Van Dyck et al., 2010). The concept of neighborhood walkability includes such items as residential density (number of residential units per residential square kilometer), street connectivity (number of intersections per square kilometer), land use mix (the evenness of distribution of residential, commercial, and office developments), and the retail floor area ratio (ratio of retail building floor area to land area) (Frank et al., 2005, Leslie et al., 2007).

Previous studies from the U.S. and Australia have found positive associations between neighborhood walkability and physical activity in adults. The Neighborhood Quality of Life Study (NQLS), conducted in the U.S., found positive associations between neighborhood walkability and walking for active transportation, walking for leisure, and accelerometer-measured moderate-to-vigorous physical activity (MVPA). Individuals living in highly walkable neighborhoods had 31.5 more minutes per week of walking for active transportation, 4.3 more minutes per week of walking for leisure and 5.8 more minutes per day of MVPA compared to individuals living in less walkable neighborhoods (Sallis, Saelens, Frank, Conway, Slymen, Cain et al., 2009). The Physical Activity in Localities and Community Environments (PLACE) study, conducted in Australia, found positive associations between neighborhood walkability and walking for active transportation. In addition, the relationship between neighborhood walkability and walking for active transportation was stronger for weekly frequency of walking than it was for weekly minutes spent walking (Owen et al., 2007).

There is, however, a need to examine whether the associations found in the U.S. and Australia also hold up in a European context. This is important because there are large differences in the built environment between Europe and the U.S. or Australia. In addition, Europe is characterized by a high degree of heterogeneity in the approximately 750 million people living in around 50 countries. This heterogeneity can be seen in the very different economic and political systems, the cultural mix across the European continent, and the many languages spoken.

The first European study on the association between neighborhood walkability and physical activity was conducted in 2010 in Ghent, Belgium. The Belgian Environmental Physical Activity Study (BEPAS) confirmed that the previously found associations between neighborhood walkability and physical activity in the U.S. and Australia also exist in Belgium (Van Dyck et al., 2010), although some discrepancies in the main results were found. For example, Belgian adults living in highly walkable neighborhoods had more accelerometer-measured minutes of MVPA, more walking for active transportation, and more walking for recreation than those living in less walkable neighborhoods. The authors of that study concluded, “…conducting European studies is important because walkability is likely to be a context-relative construct…” and “…other European investigators are encouraged to examine SES interactions with walkability…” Previous studies of the association between neighborhood walkability and physical activity have included measures of neighborhood-level socioeconomic status (SES), although the results of the association between neighborhood-level SES and physical activity are inconsistent (Owen et al., 2007, Sallis et al., 2009, Van Dyck et al., 2010).

The first aim of this large Swedish study of adults was to examine the associations between objective neighborhood walkability and walking for active transportation, walking for leisure and accelerometer-measured MVPA and whether these hypothesized associations are moderated by individual-level sociodemographic factors and neighborhood-level SES. The second aim was to examine random effects in a multilevel fashion, which quantifies how much of the total variance of the walking and physical activity outcomes could be due to differences at the neighborhood level. This second aim constitutes a novel contribution and has the potential to provide important information to decision-makers and city planners because a knowledge of the magnitude of the total variance at the neighborhood level could contribute to cost-effective decisions concerning developments of new and redevelopments of already existing urban neighborhoods. Another strength of the present study is that the study design is similar to the designs of the NQLS, the PLACE Study, and the BEPAS, which entails the potential to make results comparable across countries.

Section snippets

Neighborhood walkability

Data for the Swedish Neighborhood and Physical Activity (SNAP) Study were collected in the city of Stockholm in Sweden. The city of Stockholm is divided into 408 small administrative units with homogeneous types of buildings. They contain approximately 2000 individuals per unit. The geographic boundaries of the administrative units follow the road/street network and they are also well-known geographic units that could be used for future health interventions. They constituted a basis for the

Descriptive statistics on the 2269 individuals

Table 1 shows that the median objectively measured MVPA of SNAP participants amounted to 41 min/day (SD = 23 min). The participants reported a median of 125 min/week of walking for active transportation (SD = 275 min) and a median of 60 min/week of walking for leisure (SD = 222). The proportion of female participants was 55% and the proportion of married/cohabiting participants was 74% of the entire study sample. Forty percent were over 50 years old and 42% were found among those with middle

Discussion

The main findings of the present study of the association between objectively measured neighborhood walkability, physical activity, and walking behavior, conducted in a Swedish context, are mainly in agreement with previous research from the U.S., Australia, and Belgium. In the adjusted models, we found the following statistically significant results among individuals living in highly walkable neighborhoods, compared to those living in less walkable neighborhoods: (1) 77% and 28% higher odds

Conclusions

The findings of this study show a positive association between objective neighborhood walkability and physical activity outcomes in a Swedish context. Although these average effects were significant, the low values in the calculations of ICC indicate that the variance at the neighborhood level is low; the largest proportion of the total variance was at the individual level. Moreover, the objective assessment of the individuals’ level of physical activity showed a relatively slight difference of

Acknowledgments

This work was supported by grants to Dr Kristina Sundquist from the Swedish Council for Working Life and Social Research, FAS (2006-0386), and The Swedish Research Council for Environment, Agricultural Sciences, and Spatial Planning, Formas (2006-1196). Though listed as first and second authors, Sundquist and Eriksson contributed equally to this article.

References (33)

  • C.M. Burchfiel et al.

    Physical activity and incidence of diabetes: the Honolulu Heart Program

    The American Journal of Epidemiology

    (1995)
  • A. Forsyth

    Twin cities walking study, environment and physical activity: GIS protocols

    (2007)
  • L.D. Frank et al.

    Many pathways from land use to health: associations between neighborhood walkability and active transportation, body mass index, and air quality

    Journal of the American Planning Association

    (2006)
  • P.S. Freedson et al.

    Calibration of the Computer Science and Applications, Inc. accelerometer

    Medicine & Science in Sports & Exercise

    (1998)
  • K. Gebel et al.

    Correlates of non-concordance between perceived and objective measures of walkability

    Annals of Behavioral Medicine

    (2009)
  • B. Giles-Corti et al.

    Understanding physical activity environmental correlates: increased specificity for ecological models

    Exercise and Sport Sciences Reviews

    (2005)
  • Cited by (183)

    View all citing articles on Scopus
    1

    Contributed equally to this work.

    View full text