1. Takemi Sugiyama PhD MArch Spatial Epidemiology & Evaluation Research Group University of South Australia, Adelaide, Australia Perceived Neighbourhood Environmental Attributes Associated with Adults’ Recreational Walking IPEN Adult Study in 12 Countries Sugiyama T, Cerin E, Owen N, Oyeyemi AL, Conway TL, Van Dyck D, Schipperijn J, Macfarlane DJ, Salvo D, Reis RS, Mitáš J, Sarmiento OL, Davey R, Schofield G, Orzanco-Garralda R, Sallis JF. Perceived neighbourhood environmental attributes associated with adults’ recreational walking: IPEN Adult study in 12 countries. Health & Place, 2014;28:22–30.

2. Co-Authors Ester Cerin (Hong Kong) Neville Owen (Australia) Adewale L Oyeyemi (Nigeria) Terry L Conway (USA) Delfien Van Dyck (Belgium) Jasper Schipperijn (Denmark) Duncan J Macfarlane (Hong Kong) Deborah Salvo (Mexico) Rodrigo S Reis (Brazil) Josef Mitáš (Czech Republic) Olga L Sarmiento (Colombia) Rachel Davey (UK) Grant Schofield (New Zealand) Rosario Orzanco-Garralda (Spain) James F Sallis (USA) Acknowledgements IPEN Coordinating Centre Nicole Bracy Kelli Cain Carrie Geremia Lisa Husak Marc Adams IPEN Executive Committee James F Sallis Jacqueline Kerr Ilse De Bourdeaudhuij Neville Owen Lawrence Frank Terry Conway Takemi Sugiyama And all IPEN investigators

3. An increasing number of studies have examined associations of neighbourhood environmental attributes with physical activity. To date, most studies on this topic have examined data collected in a single country. Non-significant or weak associations reported in single- country studies may be due partly to limited variation in environmental attributes. Multi-country studies can fill this methodological gap by provide larger variance in environmental attributes. Background

4. Q1. To examine the strength and shape of associations of perceived neighbourhood environmental attributes with adults’ recreational walking using data obtained from 12 countries Q2. To examine whether these associations differ across countries Aims

5. The International Physical Activity and Environment Network (IPEN) Adult Study was an observational, cross- sectional, multi-country study, involving 13,745 adults aged 18–66 years from 12 countries (17 study sites). Countries included: Australia (AUS; Adelaide), Belgium (BEL; Ghent), Brazil (BRA; Curitiba), Colombia (COL; Bogota), the Czech Republic (CZ; Olomouc, Hradec Kralove), Denmark (DEN; Aarhus), Hong Kong (HK), Mexico (MEX; Cuernavaca), New Zealand (NZ; North Shore, Waitakere, Wellington, Christchurch), Spain (ESP; Pamplona), the United Kingdom (UK; Stoke-on-Trent), and the United States of America (USA; Baltimore, Seattle). IPEN Adult Study

6. Outcome: Recreational Walking Self-reported frequency (days/week) and duration (minutes/week) of walking for recreation in the last week, obtained from the International Physical Activity Questionnaire (IPAQ; long version), were used.

7. Exposure: Neighbourhood Environment Perceived neighbourhood environmental attributes were determined using the Neighborhood Environment Walkability Scale (NEWS). The following 10 items were used: residential density score land use mix–access street connectivity infrastructure and safety aesthetics safety from traffic safety from crime few cul-de-sacs no major barriers proximity to parks

8. Analysis Generalized Additive Mixed Models (GAMMs) were used to estimate the strength and shape of associations of perceived environmental attributes with recreational walking. All analyses were conducted in R. Main effects of environmental attributes on the outcome (linear and curvilinear components) were examined for the whole sample, adjusting for study site, socio- demographic covariates, and area-level SES. Interaction effects were probed by computing the site- specific association of a perceived environmental attributed with the outcome via linear functions. Analyses stratified by sites were conducted when significant.

9. Models The following three models were examined: Model 1: likelihood of any walking for recreation versus no walking for recreation; Model 2: frequency (days/week) of walking for recreation among those reporting walking during the past week; Model 3: duration (minutes/week) of walking for recreation among those reporting walking during the past week

10. Sample Characteristics (N=13,745) Mean age 42 years; 57% women; 44% with a college/university degree; 75% working; 60% married or living with a partner Descriptive Results WalkingMean (SD)Lowest 2 Highest 2 Frequency (days/wk)1.9 (2.3)1.0 [BEL]3.6 [ESP] Non-zero frequency (days/wk) 1 3.4 (2.2)2.1 [BEL]4.6 [ESP] Duration (min/wk)115 (226)54 [BRA]234 [ESP] Non-zero duration (min/wk) 1 204 (249)146 [BRA]316 [CZ] %reporting no walking43%21% [DEN]66% [MEX] 1 among those who reported walking, 2 site-level lowest/highest [country] BEL: Belgium; BRA: Brazil; CZ: Czech Republic (Olomouc); DEN: Denmark; ESP: Spain; MEX: Mexico

11. Descriptive Results Environmental attributesMean (SD)Lowest 4 Highest 4 Residential density score 1 77 (114)18 [NZ]440 [HK] Land use mix–access 2 3.4 (0.66) 3.0 [US1]3.7 [ESP] Connectivity 2 3.0 (0.73)2.7 [NZ]3.3 [BRA] Infrastructure and safety 2 3.0 (0.57)2.6 [MEX]3.3 [ESP] Aesthetics 2 2.8 (0.70)2.2 [UK]3.1 [US2] Safety from traffic 2 2.6 (0.67)2.4 [BRA]3.1 [CZ] Safety from crime 2 3.0 (0.80)2.1 [COL]3.5 [ESP] Few cul-de-sacs 2 2.8 (1.00)2.3 [NZ]3.5 [ESP] No major barriers 2 3.3 (0.84)2.2 [HK]3.7 [US1] Proximity to parks 3 4.1 (1.21)3.1 [BEL]4.8 [ESP] 1 range: 0–1024, 2 range: 1–4, 3 range: 1–5, 4 site-level lowest/highest [country] BEL: Belgium; BRA: Brazil; COL: Colombia; CZ: Czech Republic (Hradec Kralove); ESP: Spain; HK: Hong Kong; NZ: New Zealand (North Shore); US1: USA (Baltimore); US2: USA (Seattle)

12. Linear Association (Model 1) Odds of walking for recreation 1 OR95%CIp Residential density score1.080.88, 1.320.482 Land use mix–access1.020.90, 1.170.726 Connectivity 1.010.96, 1.070.745 Infrastructure and safety 1.010.94, 1.090.826 Aesthetics 1.261.18, 1.35<0.001 Safety from traffic 1.050.99, 1.110.128 Safety from crime 1.071.00, 1.140.036 Few cul-de-sacs 0.940.90, 0.980.024 No major barriers 1.000.95, 1.050.886 Proximity to parks 1.071.03, 1.110.031 1 Analysis for the whole sample (n=13,745), GAMMs with binomial variance and logit link functions

13. Non-zero walking frequency 1 (days/wk) exp(b)95%CIp Residential density score1.0010.999, 1.0030.138 Land use mix–access1.020.99, 1.050.062 Connectivity 1.021.00, 1.050.025 Infrastructure and safety 0.970.94, 1.000.053 Aesthetics 1.051.03, 1.08<0.001 Safety from traffic 0.990.96, 1.010.313 Safety from crime 1.000.97, 1.020.737 Few cul-de-sacs 0.990.98, 1.010.274 No major barriers 1.000.98, 1.020.898 Proximity to parks 1.010.99, 1.020.213 1 Analysis for those who reported walking (n=7,838), GAMMs with negative binomial and logarithmic link functions Linear Association (Model 2)

14. Non-zero walking duration 1 (minutes/wk) exp(b)95%CIp Residential density score1.0011.000, 1.0010.005 Land use mix–access1.071.02, 1.110.003 Connectivity 1.020.98, 1.050.298 Infrastructure and safety 0.980.92, 1.040.519 Aesthetics 1.020.95, 1.100.569 Safety from traffic 0.970.93, 1.010.101 Safety from crime 0.980.94, 1.020.334 Few cul-de-sacs 0.980.95, 1.000.059 No major barriers 1.000.97, 1.040.795 Proximity to parks 1.010.99, 1.040.344 1 Analysis for those who reported walking (n=7,838), GAMMs with negative binomial and logarithmic link functions Linear Association (Model 3)

15. The solid line represents point estimates (and black dashed lines their 95% confidence intervals) of the odds of walking at various levels of perceived environmental attribute for the imputed dataset producing median values of odds of walking (relative to other imputed models). The black dots and grey dashed lines represent the same estimates produced by the other nine imputed datasets. Curvilinear Association (Model 1) Residential density score Odds of walking for recreation

16. The solid line represents point estimates (and black dashed lines their 95% confidence intervals) of the odds of walking at various levels of perceived environmental attribute for the imputed dataset producing median values of odds of walking (relative to other imputed models). The black dots and grey dashed lines represent the same estimates produced by the other nine imputed datasets. Curvilinear Association (Model 1) Land use mix – access Odds of walking for recreation

17. The solid line represents point estimates (and black dashed lines their 95% confidence intervals) of the odds of walking at various levels of perceived environmental attribute for the imputed dataset producing median values of odds of walking (relative to other imputed models). The black dots and grey dashed lines represent the same estimates produced by the other nine imputed datasets. Curvilinear Association (Model 3) Aesthetics Minutes of walking for recreation

18. Associations were not significant in Australia, Belgium, Brazil, Columbia, Czech Republic (Olomouc, Hradec Kralove), Denmark, Mexico, New Zealand (Christchurch), Spain, UK, USA (Seattle) Site-Specific Associations (Aesthetics) Country (study site) exp(b)exp(95%CI)p Hong Kong 1.261.01, 1.560.037 New Zealand (North Shore) 1.361.06, 1.730.014 New Zealand (Waitakere) 1.351.00, 1.820.048 New Zealand (Wellington) 1.220.97, 1.540.089 USA (Baltimore) 1.191.01, 1.410.039 Site-specific linear associations of aesthetics with duration of walking for recreation among walkers

19. Of the 10 perceived environmental attributes examined, seven were (either linearly or curvilinearly) positively associated with at least one walking outcome. Stronger evidence was obtained for Aesthetics, which was associated with all three walking measures, with a larger effect size (26% higher odds of walking for a unit increase). Summary Q1. To examine the strength and shape of associations of perceived neighbourhood environmental attributes with adults’ recreational walking

20. Results show some support for the importance of: Safety from crime (linear association) Proximity to parks (linear association) Residential density (curvilinear association) Land use mix (curvilinear association) Q1. To examine the strength and shape of associations of perceived neighbourhood environmental attributes with adults’ recreational walking Summary

21. Mixed findings were obtained for: Connectivity and Few cul-de-sacs (higher street connectivity and many cul-de-sacs were both positively associated with walking) No linear/curvilinear associations were found for: Infrastructure and safety Safety from traffic No major barriers Q1. To examine the strength and shape of associations of perceived neighbourhood environmental attributes with adults’ recreational walking Summary

22. Effect modification by study sites was found for aesthetics, which was associated linearly with walking in 4 sites. The associations observed were mostly consistent across countries, suggesting the generalisability of the findings. Summary Q2. To examine whether associations of environmental attributes with recreational walking differ across countries

23. Associations of perceived neighbourhood environmental attributes with adults’ walking for recreation may be similar in environmentally, culturally diverse countries. Aesthetics appears to be an important factor in whether and how often adults walk for recreation. Higher population density (but not excessively high) and higher land use diversity seem to contribute to longer duration of recreational walking. Safe neighbourhoods (from crime) with parks may facilitate recreational walking. Conclusions

24. Australian data collection was supported by National Health and Medical Research Council (NHMRC) of Australia Project Grant #213114. The contributions of Neville Owen were supported by NHMRC Grant #569940, NHMRC Senior Principal Research Fellowship#1003960, and by the Victorian Government׳s Operational Infrastructure Support Program. Data collection in Hong Kong was supported by the HK Research Grants Council GRF Grants (#HKU740907H and #747807H) and HKU URC Strategic Research Theme (Public Health). US data collection and Coordinating Center processing was supported by the NIH Grants R01 HL67350 (NHLBI) and R01 CA127296 (NCI). The Danish study was partly funded by the Municipality of Aarhus. Data collection in the Czech Republic was supported by the Ministry of Education Youth and Sports Grant #MSM6198959221. The study conducted in Colombia was funded by Colciencias Grant 519_2010, Fogarty and CeiBA. Data collection in New Zealand was supported by the Health Research Council of New Zealand Grant #07/356. Data collection in Mexico was supported by the CDC Foundation (project #550), which received an unrestricted training grant from the Coca-Cola Company. The UK study was funded by the Medical Research Council Grant number G0501287 under the National Preventive Research Initiative. Funding Sources