Skip to main content

International study of perceived neighbourhood environmental attributes and Body Mass Index: IPEN Adult study in 12 countries

Abstract

Background

Ecological models of health behaviour are an important conceptual framework to address the multiple correlates of obesity. Several single-country studies previously examined the relationship between the built environment and obesity in adults, but results are very diverse. An important reason for these mixed results is the limited variability in built environments in these single-country studies. Therefore, the aim of this study was to examine associations between perceived neighbourhood built environmental attributes and BMI/weight status in a multi-country study including 12 environmentally and culturally diverse countries.

Methods

A multi-site cross-sectional study was conducted in 17 cities (study sites) across 12 countries (Australia, Belgium, Brazil, China, Colombia, Czech Republic, Denmark, Mexico, New Zealand, Spain, the UK and USA). Participants (n = 14222, 18–66 years) self-reported perceived neighbourhood environmental attributes. Height and weight were self-reported in eight countries, and measured in person in four countries.

Results

Three environmental attributes were associated with BMI or weight status in pooled data from 12 countries. Safety from traffic was the most robust correlate, suggesting that creating safe routes for walking/cycling by reducing the speed and volume of traffic might have a positive impact upon weight status/BMI across various geographical locations. Close proximity to several local destinations was associated with BMI across all countries, suggesting compact neighbourhoods with more places to walk related to lower BMI. Safety from crime showed a curvilinear relationship with BMI, with especially poor crime safety being related to higher BMI.

Conclusions

Environmental interventions involving these three attributes appear to have international relevance and focusing on these might have implications for tackling overweight/obesity.

Background

Overweight and obesity are important health problems in developed and developing countries [1]. The ecological model has been increasingly used as a conceptual framework to address obesity at multiple levels (individual, inter-personal/social environment, built environment, and policies) [2]. The premise is that by reducing the obesogenicity of the environment, the obesity problem can be tackled on a larger scale and reach a wider population.

In the past decade, built environmental factors were mainly studied in relation to physical activity (PA). Results showed a consistent relationship between PA (active transport and recreational walking) and neighbourhood walkability in adults [3–6].

Investigating the relationship between characteristics of the built environment and adiposity (e.g., BMI, overweight/obesity) has often been secondary to investigating their relationships with PA or diet, and findings are less clear. A recent review of reviews [3] identified 8 review articles that examined the relationship between the built environment and overweight/obesity in adults [7–14]. The findings of these reviews are very diverse. A review focusing on smart growth factors (e.g., higher density, diverse land use) showed that few studies reported significant associations with adiposity measures [8]. In contrast, another review showed lower average BMI in neighbourhoods with higher perceived mixed land-use, improved walkability, and better access to facilities [10]. Also, significant associations between some aspects of the objectively measured built environment (e.g. residential density, street connectivity, greenery, access to destinations) and obesity were observed in 84 % of the studies [7].

Although some of the apparent discrepancies could be due to differences in methods, the main issue that may explain this diversity in results is the limited variability. Almost all studies of relationships between the built environment and overweight/obesity have been conducted within single countries and usually within a single city. Limited variability in built environments in these studies might be a reason for non-significant associations. However, a few single-country studies were explicitly designed to maximize built environment variability, and most of those studies are included in present analyses of pooled international data (NQLS, BEPAS, PLACE). Combining data from environmentally and culturally different contexts may help to better understand how neighbourhood built environments are related to residents’ adiposity.

The present study's purpose was to examine the strength, direction, and shape of the associations between perceived neighbourhood built environmental attributes and adiposity measures using pooled data from the 12 countries participating in the International Physical Activity and the Environment Network (IPEN) Adult study. Variation in the associations by study sites was also explored.

Methods

Study design

The IPEN Adult study is an observational, epidemiologic, multi-country, cross-sectional study, including 17 city-regions (hereafter, sites) within 12 countries: Australia (Adelaide), Belgium (Ghent), Brazil (Curitiba), China (Hong Kong), Colombia (Bogota), Czech Republic (Olomouc, Hradec Kralove), Denmark (Aarhus), Mexico (Cuernavaca), New Zealand (North Shore, Waitakere, Wellington, Christchurch), Spain (Pamplona), the UK (Stoke-on-Trent) and USA (Seattle, Baltimore).

Participants were mostly recruited from neighbourhoods chosen to maximize variance in neighbourhood walkability and socio-economic status (SES) [15]. For neighbourhood selection, all countries, except Spain, used an objective GIS-based neighbourhood walkability index [16, 17], with each country’s selection strategy reported elsewhere [15]. Administrative units were ranked by walkability index and median household income (neighbourhood-level SES), and approximately equal numbers of neighbourhoods were selected to represent four categories: high-walkable/high-SES, high-walkable/low-SES, low-walkable/high-SES, and low-walkable/low-SES.

Participant recruitment

Adults living in the selected neighbourhoods were systematically contacted over 2002–2011. Four countries used phone/mail/online surveys; seven countries delivered study materials directly to participants; whilst Hong Kong used intercept interviews. Although ages ranged from 16–94, we limited the present analyses to 18–66 years.

Quality control

All investigators completed the San Diego State University IRB training, and met the NIH Fogarty International Center and their local ethics requirements. All participants provided informed consent. Confidentiality for pooled data maintained using only identification codes before transmitting data to the IPEN Coordinating Center, where it was processed to maximize completeness, comparability, and consistent-coding.

Measures

Body mass index (BMI) and weight status

To calculate BMI, participants reported their height and weight in eight countries or were measured objectively in Brazil, Mexico, New Zealand, and UK. Self-reported and objectively measured BMI are highly correlated, and BMI is a proxy measure for adiposity in large-scale studies [18]. Participants were then categorised into normal weight and overweight/obese (BMI ≥ 25 kg/m2).

Perceptions of neighbourhood built environmental attributes

The Neighbourhood Environment Walkability Scale (NEWS) assesses perceived neighbourhood attributes believed to be related to physical activity and reflects key exposure measures in the IPEN study [19, 20]. As IPEN countries used adapted versions, extensive item comparisons (made by at least 2 independent raters) and confirmatory factor analyses were completed. Cerin et al. [21] reported the following 10 NEWS subscales that can be used for the IPEN multi-country pooled analyses: (1) Residential density; (2) Land use mix–access; (3) Land use mix–diversity; (4) Street connectivity; (5) Infrastructure and safety for walking; (6) Aesthetics; (7) Safety from traffic; (8) Safety from crime; (9) Streets having few cul-de-sacs; and (10) No major physical barriers to walking.

The Residential density subscale is a weighted sum of items reflecting perceived density of housing, ranging from predominantly single-family dwellings to high-rise buildings with more than 20 stories. Land use mix–diversity reflects average perceived walking proximity (i.e., average of five-point ratings ranging from ≤5 minute walk to 30+ minute walk) from home to 9 types of destinations (e.g., supermarket, school, transit stop, and other stores and services). The remaining eight scales are average ratings of items answered on a four-point Likert scale (1 = strongly disagree to 4 = strongly agree). Scoring details are described elsewhere (21). Additionally, for the current study, a composite perceived walkability index was computed by summing the z-scores of all perceived neighbourhood walkability attributes.

Socio-demographic characteristics

Age, gender, educational level, and marital status were assessed and included as covariates in all statistical models. Education level was categorized into having ‘university degree’, ‘high school diploma’ and ‘less than high school diploma’. Marital status was dichotomized into married/de facto or not-married.

Data analytic plan

Descriptive statistics were computed for the whole sample and by study site for all variables. There were 8.2 % cases with missing data on at least one variable. Consequently, ten imputed datasets were created for the main regression analyses (see below) as recommended by Rubin [22] and van Buuren [23]. Multiple imputations were performed using chained equations (MICE) [23] accounting for the two-stage stratified sampling strategy employed in each study site (see Methods section).

We examined associations of built environment with body mass index (kg/m2) and weight status (normal vs. overweight/obese) using generalized additive mixed models (GAMMs) [24]. A set of GAMMs used binomial variance and logit link functions, appropriate for dichotomous measures of weight status. These models yielded odds ratios of being vs. not being overweight/obese. BMI was modelled using GAMMs with the residual variance proportional to the outcome mean (variance corresponding to the cube of the mean) and with a logarithmic link function [24]. The reported antilogarithms of the regression coefficient estimates of these two GAMMs represent the proportional increase in BMI associated with a unit increase in the predictor.

A main-effect set of GAMMs estimated the dose–response relationships of all perceived environmental attributes and, in separate models, of the composite walkability index with the continuous and categorical BMI outcome variables, adjusting for study site, socio-demographic covariates, and the design variable administrative-unit-level socio-economic status (low versus high), as well as accounting for dependency in error terms due to clustering of participants sampled from pre-selected administrative units. Curvilinear relationships of environmental attributes with outcomes were estimated using non-parametric smooth terms, which were modelled using thin-plate splines [24]. Separate GAMMs were run to estimate environmental attributes by study site interaction effects. The significance of interaction effects was evaluated by comparing QAIC values of models with and without a specific interaction term (the model with the smaller QAIC was preferred). Significant interaction effects were probed by computing site-specific associations using linear functions. Finally, the proportion of city-level variance of probability of being overweight/obese and BMI explained by perceived environmental attributes was computed. All analyses were conducted in R [25] using the packages ‘car’ [26], ‘mgcv’ [24], ‘gmodels’ [27], and ‘mice’ [28].

Results

Table 1 shows descriptive statistics for each study site including socio-demographic characteristics, weight status, and BMI. The total sample consisted of 14,222 participants; 57 % women, 60 % lived with a partner, 44 % with a college/university degree, and 74 % worked. The mean age was 42 years (SD = 12.8). After adjusting for environmental predictors and socio-demographic covariates, Cuernavaca (Mexico) had the highest and Hong Kong (China) the lowest prevalence of overweight/obesity (Fig. 1). In general, the average BMI and prevalence of overweight/obesity were lower in European sites (except for Stoke-on-Trent, UK) and higher in North and South American (except for Bogota, Colombia), Australian and New Zealand sites (Fig. 1).

Table 1 Overall and site-specific sample characteristics: socio-demographics, body mass index (BMI), and weight status
Fig. 1
figure 1

Between-site differences in probability of being overweight/obese and average body mass index (kg/m2) adjusted for socio-demographic characteristics and perceived environmental attributes

Average perceived residential density was the lowest in three out of four study sites located in New Zealand, and the highest in Hong Kong, followed by Pamplona (Spain) and Curitiba (Brazil) (Table 2). All study sites had relatively high average scores on land use mix. Pamplona (Spain) had the highest scores on seven out of ten perceived environmental characteristics, while Cuernavaca (Mexico) and Waitakere (New Zealand) had one of the lowest average scores on five characteristics. The composite walkability indices of the study sites in New Zealand, Cuernavaca (Mexico), Bogota (Colombia) and Ghent (Belgium) were relatively low, while those of Hong Kong (China), Aarhus (Denmark) and Pamplona (Spain) were high.

Table 2 Overall and site-specific perceived environmental scores [mean (SD)], all scores are in the positive direction, higher scores meaning more agreement with the attribute

Female, younger, single, highly educated participants and those living in higher SES areas were less likely to be overweight/obese (Table 3). When accounting for other perceived environmental attributes, traffic safety was the only attribute negatively related to the odds of being overweight/obese (Table 3). A significant interaction effect of study site by perceived pedestrian infrastructure and safety on the odds of being overweight/obese was observed. Namely, a positive association between pedestrian infrastructure and safety and the odds of being overweight/obese was observed in Hong Kong (China) (OR = 1.45; 95 % CI: 1.07, 1.96; p < .05), a negative association in Adelaide (Australia) (OR = 0.80; 95 % CI: 0.70, 0.94; p < .01), while the other sites showed no significant association.

Table 3 Linear and curvilinear associations of socio-demographic and perceived environmental attributes with weight status (normal vs. overweight/obese) and body mass index (kg/m2) (N = 14222)

The composite walkability index was significantly negatively related to the odds of being overweight/obese (OR = 0.98; 95 % CI: 0.97, 0.99; p = .002). The difference in minimum and maximum scores across study sites in the walkability index was nearly 30 units. Thus, the odds of being overweight/obese for a person living in the most walkable neighbourhood were 45 % lower than those of a person living in the least walkable neighbourhood. These effects did not significantly differ across study sites. Perceived environmental attributes explained 11.8 % of site-level variance in the probability of being overweight/obese, corresponding to a standard deviation of 4.7 % in prevalence of overweight/obesity and 28.2 % in prevalence between the lowest and highest ranked study site on perceived neighbourhood attributes.

BMI was lower in female, higher educated, younger participants, and those living in higher socio-economic status neighbourhoods (Table 3). Higher levels of land use mix–diversity and traffic safety were predictive of lower BMI. A unit increase on each of these two environmental measures was associated with a 1 % decrease in average BMI (Table 3). The association of perceived safety from crime with BMI was negative and curvilinear; BMI decreases as safety increases, but is relatively constant at the higher end of the safety from crime scale (Fig. 2). Study site moderated the associations of BMI with perceived residential density and no major barriers to walking. Specifically, only two study sites showed significant, negative associations between perceived residential density and BMI. These were Adelaide, Australia (OR = 0.9996; 95 % CI: 0.9994, 0.9998; p < 0.001) and Waitakere, New Zealand (OR = 0.9992; 95 % CI: 0.9985, 0.9999; p < 0.05). Negative associations of BMI and no perceived major barriers to walking were observed in Adelaide, Australia (OR = 0.987; 95 % CI: 0.977, 0.998; p < 0.05) and Waitakere, New Zealand only (OR = 0.967; 95 % CI: 0.943, 0.993; p < 0.05).

Fig. 2
figure 2

Non-linear relationship between perceived safety from crime and body mass index (kg/m2)

The composite walkability index was significantly negatively related to BMI (OR = 0.992, 95 % CI = 0.88, 0.995; p < 0.001). Study site did not significantly moderate this association. Perceived environmental attributes explained 5.9 % of city-level variance in BMI, corresponding to a standard deviation of 0.37 kg/m2 and 2.24 kg/m2 difference between the lowest and highest ranked study site on perceived neighbourhood attributes.

Discussion

In the pooled analyses including all 12 countries, safety from traffic was the only environmental attribute that was associated in the expected direction both with lower odds of being overweight/obese, and lower BMI. Participants’ BMI was linearly related to land use mix-diversity and curvilinearly associated with safety from crime. It should be noted that these three environmental attributes (safety from traffic, land use mix-diversity, and safety from crime) show a consistent relationship with the outcome measures in 12 environmentally and culturally diverse countries. The linear relationships found for land use mix-diversity and traffic safety in these countries suggest that the closer the perceived walking proximity to destinations (supermarket, grocery, post office, transit stop, restaurant, park, gym, school), and the safer it is to reach these destinations (no heavy traffic, slow traffic speeds), the lower the BMI of the inhabitants of the regions is. Previous research showed a consistent relationship between land use mix-diversity and physical activity [3], while associations between traffic safety and physical activity are less consistent [29, 30]. Because land use mix and traffic safety have been related to more walking for transportation purposes, transport-walking is a potential mediator of the observed association between these two environmental attributes and body mass. However, clear and consistent evidence on the relationships of these environmental attributes with body mass is lacking. Black et al. [10] concluded from their review of 37 studies that a lower average BMI seemed to be related to higher mixed land use, improved walkability, better access to facilities, and low perceived hazards. However, their conclusion was derived by examining mainly studies conducted in the US. The present study supports the importance of diverse land use (the presence of destinations nearby) and neighbourhood safety in residents’ weight status through examining the large, international data with a wider variability in environmental factors and obesity levels.

The present study makes an important contribution by investigating the shape of the relationship between built environment variables and BMI. The curvilinear relationship with perceived safety from crime can be understood as a threshold effect. The curvilinear association suggests that in neighbourhoods with low perceived crime safety, a small increase in perception of crime safety could be associated with a decrease in BMI. However, in relatively safe neighbourhoods, few additional effects on BMI could be expected from further increasing the perceived safety from crime. Thanks to the large between-country variance in built environmental characteristics and weight status, the present findings can give further insights into the often conflicting results found in the past for the relationship between perceived safety from crime and physical activity/body mass [30, 31]. Especially in countries that have relatively high crime rates, lower perceived safety from crime at night seems to be an important feature related to higher BMI [32].

Very few interactions by study site were revealed in the analyses, suggesting that associations were similar across countries. A site-specific result was found for perceived pedestrian infrastructure and safety on the odds of being overweight/obese. A positive association was found in Hong Kong, a negative association in Australia, and nonsignificant associations in all other countries. Another site-specific result was observed for residential density and no major barriers on BMI. Negative associations between BMI and both environmental attributes were only found in Australia and New Zealand. These results suggest that the relationship between perceived built environmental attributes and BMI/overweight might be somewhat stronger in Australia compared to other countries. Previous studies on the relationship between the built environment and body mass in Australian adults showed mixed results. A recent study by Christian et al. [33] showed almost no relationship of perceived environmental factors with BMI, while other studies showed significant associations [34, 35]. Future research is needed to further clarify this finding, in order to be able to decide whether similar intervention strategies could possibly have generalizable effects across countries.

Despite the consistent association of BMI with land-use mix diversity, safety from traffic and safety from crime in the present study, the other perceived environmental attributes that were examined, such as residential density, land use mix access (to shops and services), street connectivity, pedestrian/bicycling infrastructure, aesthetics, absence of cul-de-sacs, and other major barriers were not associated with BMI or weight status across the countries. As discussed in the previously cited review [3], neighbourhood environmental attributes that were found to be consistently related to PA do not necessarily have direct associations with BMI. However, the composite walkability index including the sum of all perceived environmental attributes was significantly negatively related to both the odds of being overweight/obese and to BMI. As these effects did not significantly differ across study sites, they show that a combination of more favourable environmental attributes is related to less overweight across the world. It is notable that the composite walkability score had a relatively large effect size. In the case of the odds of overweight/obesity, the most walkable study site had 45 % lower odds than the least walkable site. In contrast, a single item showed a smaller effect size. For instance, the odds of overweight/obesity in the safest site (from crime) was 11 % lower than that of the least safe site. This can be interpreted as showing that multiple environmental attributes may have an accumulated impact on residents’ weight status. Such accumulated environmental influence has been also suggested for physical activity [36]. Future research can investigate the impact of multiple environmental attributes on adults’ adiposity.

The major strength of the present study was the unique pooling of data across 12 countries, to increase the variability in built environment characteristics. Another strength is the use of advanced statistical analyses investigating linear as well as curvilinear relationships. Furthermore, the study used a large sample, including more than 14,000 adults. Limitations included the mix of self-report and objective measures for BMI across countries. This variation in measurement methods could have under- or over-estimated observed associations. A second limitation is the use of standard BMI cut-points in all countries. As no ethnicity information was available, more specific BMI cut-points (eg. Asian specific) could not be used. The use of perceived environmental attributes is sometimes considered to be a limitation. Within the IPEN study, objective GIS-based environmental measures are also available in almost all countries. Previous research has shown that perceived as well as objective measures of the built environment can explain physical activity or BMI [37]. In addition, it is well documented that people’s environment perceptions often differ from objectively-identified walkability measures [38, 39]. The present study focused on the perceived environment related to weight status/BMI, and GIS based measures will be included in future manuscripts dependent on the availability of the measures in subsamples of countries. It was beyond the scope of the present manuscript to study the potential mechanisms through which neighbourhood environmental attributes relate to BMI. Future manuscripts will examine these mechanisms by conducting analyses to examine how multiple measures of physical activity and sedentary behaviours may mediate the relation between built environments and BMI. Diet intake should also be included in future studies in the pathway between environmental attributes and BMI. However, diet was not measured in the IPEN Adult study.

Conclusion

In summary, three environmental attributes were associated with BMI or weight status in pooled data from 12 countries, over and above the effect of socio-demographic characteristics and neighbourhood-level SES. Safety from traffic was the most robust correlate, which suggests that creating safe routes for walking or cycling by reducing the speed and volume of traffic might contribute to improved weight status and BMI across various geographical locations. Close proximity to several destinations was associated with lower BMI across all countries, suggesting that compact neighbourhoods with more places to walk to could be an important policy goal for controlling obesity internationally. Safety from crime showed a curvilinear relationship with BMI, with the relation of perceived crime safety to lower BMI being more pronounced in areas with poorer perceived safety. This suggests that enhanced efforts to improve neighbourhood safety (through reducing crime or incivilities, or enhancing surveillance) should be targeted mainly in areas with low perceived crime safety, and if effective, could play a role in obesity prevention. Environmental interventions involving these three attributes appear to have international relevance and may help reduce the prevalence of overweight and obesity across the world. To achieve such changes, it will likely be necessary for public health professionals to work with policy makers and practitioners in transportation (to improve traffic safety), urban planning (to ensure mixed use development), and law enforcement (to reduce crime) [40]. Future longitudinal or quasi-experimental studies documenting weight changes in neighbourhoods are needed to build more robust evidence in support of policy changes implied by the findings of this study.

References

  1. Finucane MM, Stevens GA, Cowan MJ, Goodarz D, Lin JK, Paciorek CJ, et al. National, regional, and global trends in body-mass index since 1980: systematic analysis of health examination surveys and epidemiological studies with 960 country-years and 9.1 million participants. Lancet. 2011;377:557–67.

    Article  Google Scholar 

  2. Sallis JF, Owen N, Fisher EB, Glanz K, Rimer BK, Viswanath K. Ecological models of health behavior. In: Health behavior and health education: Theory, research, and practice. San Francisco: Jossey-Bass; 2008. p. 465–86.

    Google Scholar 

  3. Ding D, Gebel K. Built environment, physical activity, and obesity: what have we learned from reviewing the literature? Health Place. 2012;18:100–5.

    Article  Google Scholar 

  4. Grasser G, Van Dyck D, Titze S, Stronegger W. Objectively measured walkability and active transport and weight-related outcomes in adults: a systematic review. Int J Public Health. 2013;58:615–25.

    Article  Google Scholar 

  5. Van Holle V, Deforche B, Van Cauwenberg J, Goubert L, Maes L, Van de Weghe N, et al. Relationship between the physical environment and different domains of physical activity in European adults: a systematic review. BMC Public Health. 2012;12:807.

    Article  Google Scholar 

  6. Reis RS, Hino AAF, Rech CR, Kerr J, Hallal PC. Walkability and Physical Activity: Findings from Curitiba, Brazil. Am J Prev Med. 2013;45:269–75.

    Article  Google Scholar 

  7. Papas MA, Alberg AJ, Ewing R, Helzlsouer KJ, Gary TL, Klassen AC. The built environment and obesity. Epidemiol Rev. 2007;29:129–43.

    Article  Google Scholar 

  8. Durand CP, Andalib M, Dunton GF, Wolch J, Pentz MA. A systematic review of built environment factors related to physical activity and obesity risk: implications for smart growth urban planning. Obes Rev. 2011;12:e173–82.

    Article  CAS  Google Scholar 

  9. Feng J, Glass TA, Curriero FC, Stewart WF, Schwartz BS. The built environment and obesity: a systematic review of the epidemiologic evidence. Health Place. 2010;16:175–90.

    Article  Google Scholar 

  10. Black JL, Macinko J. Neighborhoods and obesity. Nutrition Reviews. 2008;66:2–20.

    Article  Google Scholar 

  11. Booth KM, Pinkston MM, Poston WS. Obesity and the built environment. JADA. 2005;105:S110–7.

    Article  Google Scholar 

  12. Lachowycz K, Jones AP. Greenspace and obesity: a systematic review of the evidence. Obes Rev. 2011;12:e183–9.

    Article  CAS  Google Scholar 

  13. Lovasi GS, Hutson MA, Guerra M, Neckerman KM. Built environments and obesity in disadvantaged populations. Epidemiol Rev. 2009;31:7–20.

    Article  Google Scholar 

  14. Casagrande SS, Whitt-Glover MC, Lancaster KJ, Odoms-Young AM, Gary TL. Built environment and health behaviors among African Americans: a systematic review. Am J Prev Med. 2009;36:174–81.

    Article  Google Scholar 

  15. Kerr J, Sallis JF, Owen N, De Bourdeaudhuij I, Cerin E, Sugiyama T, et al. Advancing science and policy through a coordinated international study of physical activity and built environments: IPEN adults methods. J Phys Act Health. 2013;10:581–601.

    Google Scholar 

  16. Frank LD, Sallis JF, Saelens BE, Leary L, Cain K, Conway TL, et al. The development of a walkability index: application to the Neighborhood Quality of Life Study. Br J Sports Med. 2010;44:924–33.

    Article  CAS  Google Scholar 

  17. Cerin E, Leslie E, Owen N, Bauman A. An Australian version of the neighborhood environment walkability scale: validity evidence. Meas Phys Educ Exercise Sci. 2008;12:31–51.

    Article  Google Scholar 

  18. McAdams MA, Van Dam RM, Hu FB. Comparison of self-reported and measured BMI as correlates of disease markers in US adults. Obesity. 2007;15:188–96.

    Article  Google Scholar 

  19. Saelens BE, Sallis JF, Black JB, Chen D. Neigborhood-based differences in physical activity: an environment scale evaluation. Am J Public Health. 2003;93:1552–8.

    Article  Google Scholar 

  20. Cerin E, Saelens BE, Sallis JF, Frank LD. Neighborhood environment walkability scale: validity and development of a short form. Med Sci Sports Exerc. 2006;38:1682–91.

    Article  Google Scholar 

  21. Cerin E, Conway TL, Cain KL, Kerr J, De Bourdeaudhuij I, Owen N, et al. Sharing good NEWS across the world: developing comparable scores across 12 countries for the Neighborhood Environment Walkability Scale (NEWS). BMC Public Health. 2013;13:309.

    Article  Google Scholar 

  22. Rubin DB. Multiple imputation for non-response in surveys. New York: Wiley J & Sons; 1987.

    Book  Google Scholar 

  23. van Buuren S. Flexible imputation of missing data. Boca Raton, Florida: Chapman & Hall; 2012.

    Book  Google Scholar 

  24. Wood SN. Generalized additive models: an introduction with R. Boca Raton, Florida: Chapman & Hall; 2006.

    Google Scholar 

  25. R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing. Vienna, Austria. 2013. ISBN 3-900051-07-0, URL http://www.R-project.org/.

  26. Fox J, Weisberg S. An R companion to applied regression. 2nd ed. Thousand Oaks, CA: Sage; 2011.

    Google Scholar 

  27. Warnes GR. gmodels: Various R programming tools for model fitting. R package version 2.15.3. URL http://CRAN.R-project.org/package=gmodels.

  28. van Buuren S, Groothuis-Oudshoorn K. Mice: Multivariate imputation by chained equations in R. J Stat Software. 2011;45:1–67.

    Google Scholar 

  29. Van Dyck D, Cerin E, Conway TL, De Bourdeaudhuij I, Owen N, Kerr J, et al. Perceived neighborhood environmental attributes associated with adults’ transport-related walking and cycling: Findings from the USA, Australia and Belgium. Int J Behav Nutr Phys Act. 2012;9:70.

    Article  Google Scholar 

  30. Saelens BE, Handy SL. Built environment correlates of walking: a review. Med Sci Sports Exerc. 2008;40:S550–66.

    Article  Google Scholar 

  31. Foster S, Giles-Corti B. The built environment, neighborhood crime and constrained physical activity: An exploration of inconsistent findings. Am J Prev Med. 2008;47:241–51.

    Article  Google Scholar 

  32. Oyeyemi AL, Deforche B, Sallis JF, De Bourdeaudhuij I, Van Dyck D. Behavioral mediators of the association between neighborhood environment and weight status in Nigerian adults. Am J Health Promot. 2013;28:23–31.

    Article  Google Scholar 

  33. Christian H, Giles-Corti B, Knuiman M, Timperio A, Foster S. The influence of the built environment, social environment and health behaviors on body mass index. Results from RESIDE. Prev Med. 2011;53:57–60.

    Article  Google Scholar 

  34. Garden FL, Jalaludin BB. Impact of urban sprawl on overweight, obesity, and physical activity in Sydney, Australia. J Urban Health. 2009;86:19–30.

    Article  Google Scholar 

  35. Giles-Corti B, Macintyre S, Clarkson JP, Pikora T, Donovan RJ. Environmental and lifestyle factors associated with overweight and obesity in Perth, Australia. Am J Health Promot. 2003;18:93–102.

    Article  Google Scholar 

  36. Sallis JF, Bowles HR, Bauman A, Ainsworth BE, Bull FC, Craig CL, et al. Neighborhood environments and physical activity among adults in 11 countries. Am J Prev Med. 2009;36:484–90.

    Article  Google Scholar 

  37. McGinn AP, Evenson KR, Herring AH, Huston SL, Rodriguez DA. Exploring associations between physical activity and perceived and objective measures of the built environment. J Urban Health. 2007;84:162–84.

    Article  Google Scholar 

  38. Arvidsson D, Kawakami N, Ohlsson H, Sundquist K. Physical activity and concordance between objective and perceived walkability. Med Sci Sports Exerc. 2012;44:280–7.

    Article  Google Scholar 

  39. Gebel K, Bauman AE, Sugiyama T, Owen N. Mismatch between perceived and objectively assessed neighborhood walkability attributes: prospective relationships with walking and weight gain. Health Place. 2011;17:519–24.

    Article  Google Scholar 

  40. Institute of Medicine (IOM) and National Research Council. Local Government Actions to Prevent Childhood Obesity. Washington, DC: The National Academies Press; 2009.

    Google Scholar 

Download references

Acknowledgements

US data collection and Coordinating Center processing was supported by the following NIH grants: R01 HL67350 (NHLBI) and R01 CA127296 (NCI). Data collection in Hong Kong was supported by the grants (#HKU740907H and #747807H) and the HKU URC Strategic Research Theme (Public Health). The study conducted in Bogota was funded by Colciencias grant 519 2010, Fogarty and CeiBA (Center in Complex-Systems, Basic and Applied Research at the Universidad de los Andes). The contributions of Neville Owen were supported by a NHMRC Program Grant (#569940), a NHMRC Senior Principal Research Fellowship (#1003960), and by the Victorian Government’s Operational Infrastructure Support Program. The Danish study was partly funded by the Municipality of Aarhus. Data collection in the Czech Republic was supported by the grant MEYS (# MSM 6198959221). 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 which received an unrestricted training grant from The Coca-Cola Company. The funding bodies had no input in study design and the collection, analysis and interpretation of data and the writing of the article and the decision to submit it for publication. All authors are independent from the funding bodies.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Delfien Van Dyck.

Additional information

Competing interests

The authors declare that they have no competing interests.

Authors’ contributions

IDB, DVD, EC, RD, DS and RR conceptualised the manuscript. IDB and DVD drafted the introduction, methods and discussion sections. EC performed the analyses and drafted the data analytic plan and results sections. Everyone else participated in the country-level coordination, recruitment, and study implementation within each participating country. All authors read, edited, or revised the manuscript for important intellectual content, and approved of the version submitted.

Rights and permissions

Open Access  This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.

The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

To view a copy of this licence, visit https://creativecommons.org/licenses/by/4.0/.

The Creative Commons Public Domain Dedication waiver (https://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

De Bourdeaudhuij, I., Van Dyck, D., Salvo, D. et al. International study of perceived neighbourhood environmental attributes and Body Mass Index: IPEN Adult study in 12 countries. Int J Behav Nutr Phys Act 12, 62 (2015). https://doi.org/10.1186/s12966-015-0228-y

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12966-015-0228-y

Keywords