• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Am J Health Promot. Author manuscript; available in PMC Jun 1, 2011.
Published in final edited form as:
PMCID: PMC3105357
NIHMSID: NIHMS294509

Using Built Environmental Observation Tools: Comparing Two Methods of Creating a Measure of the Built Environment

Abstract

Purpose

Identify an efficient method of creating a comprehensive and concise measure of the built environment integrating data from geographic information systems (GIS) and the Senior Walking Environmental Assessment Tool (SWEAT).

Design

Cross-sectional study using a population sample.

Setting

Eight municipally defined neighborhoods in Portland, Oregon.

Subjects

Adult residents (N = 120) of audited segments (N = 363).

Measures

We described built environmental features using SWEAT audits and GIS data. We obtained information on walking behaviors and potential confounders through in-person interviews.

Analysis

We created two sets of environviental measures, one based on the conceptual framework used to develop SWEAT and another using principal component analysis (PCA). Each measure’s association with walking for transportation and exercise was then assessed and compared using logistic regression.

Results

A priori measures (destinations, safety, aesthetics, and functionality) and PCA measures (accessibility, comfort/safety, maintenance, and pleasantness) were analogous in conceptual meaning and had similar associations with walking. Walking for transportation was associated with destination accessibility and functional elements, whereas walking for exercise was associated with maintenance of the walking area and protection from traffic. However, only PCA measures consistently reached statistical significance.

Conclusion

The measures created with PCA were more parsimonious than those created a priori. Performing PCA is an efficient method of combining and scoring SWEAT and GIS data.

Keywords: Environment, Residence Characteristics, Walking, Principal Component Analysis, Research Design, Prevention Research

PURPOSE

Ensuring that built environments are conducive to walking is a central component of encouraging physical activity at the community level. The term built environment encompasses urban design, land use, and transportation systems; it defines and describes the communities in which individuals live and therefore influences lifestyles and choices made by inhabitants of those communities.1-3

Research into associations between characteristics of the built environment and health has relied primarily on self-reported descriptions of environmental elements,4-6 or more recently, on data collected from geographic information systems (GIS) or other existing transportation and urban planning data.7-10 Although GIS measures are based on objective data lacking in subjective self-reports, they describe only general environmental characteristics and are typically available only on broad scales and from large geographic areas defined by administrative boundaries.11

To improve the specificity of neighborhood characterization, observational audit instruments12-14 have been developed to quantify objective measures of specific environmental features within a precisely defined area. Microscale data generated by these tools can provide insightful and specific information for designing targeted interventions to encourage physical activity in a community. However, there has been little research on how to efficiently analyze the large amount of data generated by these audit instruments, and as a result, their utility has been limited thus far.

This study will address this gap in usability by identifying a scoring method that balances describing precise built environment features and reducing the number of measures for analysis using available data collected with the Senior Walking Environmental Assessment Tool (SWEAT), a microscale audit instrument developed from a clear conceptual framework. Specifically, this study will (1) develop two sets of environmental measures integrating GIS and audit data, one based on the conceptual framework used to design SWEAT and a second based on principal component analysis (PCA) and (2) evaluate the efficiency of each method by comparing the two sets of measures in their association with walking behavior. In determining which method is more efficient at creating comprehensive and concise composite measures, we will provide researchers investigating numerous aspects of the built environment with a method of managing observational data generated by audit tools such as SWEAT.

METHODS

Design

This cross-sectional study was conducted in 2007 using data collected in 2002 to 2003 as part of a study of characteristics of the built environment and health among urban seniors. SWEAT audits and resident interviews were conducted on a random sample of street segments (N = 363) in eight municipally defined neighborhoods recognized by the Portland Office of Neighborhood Involvement. These neighborhoods were selected from control neighborhoods in the Senior Health and Physical Exercise study, a walking-program randomized trial in 56 Portland neighborhoods.12 Characteristics of selected neighborhoods are provided in Table 1. The Oregon Health & Science University’s Institutional Review Board approved this study.

Table 1
Characteristics From the 2000 Census of Audited Municipally Defined Neighborhoods (N = 8)*

Sample

The study population was defined as residents of audited street segments. Households whose tax lots were on or adjacent to the audited street segments were canvassed by trained interviewers in the summer of 2003. Interviewer training occurred over 2 days and included reviewing the extensive training manual and a practical competency exam. In addition to retaining the training manual for reference, interviewers and study staff met bi-weekly to review issues or questions that arose. Each face-to-face interview took approximately 15 minutes to complete. Canvassing occurred on weekdays between 3:00 and 7:00 pm; households at which no one answered the door were canvassed a second time. The first resident 18 years or older who was able to provide informed consent was interviewed. A total of 128 adults were interviewed. The survey response rate was 32% among households with someone answering the door. Participants with three or more audited street segments within one-quarter of a mile (402 m) of their household were included in this analysis.

Information on potential confounders (gender, race/ethnicity, years of residence in the area, frequency of physical activity, and overall health status) were self-reported during the structured interviews. Because of a low frequency of nonwhite respondents, race/ethnicity was collapsed to “white, non-Hispanic” and “other.” Years of residence in the area was dichotomized as “less than 10 years” and “10 years or longer.” Physical activity was assessed by asking, “On average, how often in a typical week do you exercise for 20 minutes or more at a level that increases your breathing rate enough to raise a sweat?” (responses: “never” to “a great deal”) and collapsed to “never” and “at least once per week.” General health was measured on a five-point Likert scale (“poor” to “excellent”) and collapsed to “poor/fair” and “good or better.” Initially, data on gender were not recorded to protect interviewees’ identities; however, a change in protocol during the latter half of the study period allowed gender data to be collected from participants interviewed during the second half of the data collection process.

Measures

Built Environment Features

Built environment features were measured through GIS and SWEAT audits by five teams of trained observers between October 2002 and August 2003. The 35-question SWEAT instrument measures a variety of street, sidewalk, and building characteristics hypothesized to influence walking behaviors, particularly among older adults.12 SWEAT was developed using a conceptual framework based on themes appearing in an extensive literature review and two published frameworks.6,14 Reliability of the SWEAT instrument was assessed and described previously.12 Briefly, inter-rater reliability of SWEAT was assessed from audits of 36 randomly sampled street segments in Portland, Oregon, using κ scores and intraclass correlation coefficients. Raters had acceptable agreement (κ > .6 or r > .6) on 67% of items; the aesthetics and destinations domains were less reliable than functionality and safety domains. GIS can augment information obtained from street-level SWEAT audits. For instance, a park could be missed if it is not adjacent to a sampled segment. Similarly, a random sample of segments may not accurately measure the availability of destinations because businesses tend to cluster on one street. To most completely describe neighborhood built environments with available resources, SWEAT audits and GIS data were combined.

In this study, SWEAT data for all segments within a participant’s local neighborhood were aggregated to create neighborhood-level built environment variables. One-quarter of a mile was selected as the definition of a “local neighborhood” because this distance is frequently used in the literature15-18 and is common for walking trips. Audited segments were included in a participant’s local neighborhood if any part of the segment fell within one-quarter of a mile of any point on the household’s segment. Alleyways (n = 7) were excluded from analysis. Details of neighborhood-level variables are provided in the appendix. Items with little or no variance across participants or with a high percentage of missing values were eliminated from further analysis.

GIS data on amenities within one-quarter of a mile of each audited street segment were obtained using Areview 3.3 (ESRI, Redlands, California) linked to the Regional Land Information System, a GIS database maintained by Metro, the regional planning agency for the Portland metropolitan area. Counts of amenities were summed into four mutually exclusive categories: (1) shopping (neighborhood grocery, convenience stores, supermarkets, malls), (2) medical (hospitals, medical offices), (3) services (libraries, pharmacies, banks, beauty shops, schools, senior services), and (4) activities (community centers, senior centers, pools, parks/trails, community gardens). Counts of bus stops and restaurants, as well as street segment length, were also obtained.

Walking Behavior

Walking behaviors were assessed through structured interviews. Walking for transportation was measured by asking participants, “What form of transportation do you use to get where you need to go?” Participants could report up to three modes of transportation (car, motorcycle, or moped; public transport; bicycle; walking; other; never goes out). Interviewers allowed for spontaneous responses but read options if the participant did not offer a response. For this analysis, participants were dichotomized into those who reported walking and those who did not. Walking for exercise was measured by the question, “How often in a typical week do you walk for exercise in your neighborhood?” Participants selected from the responses: never (0 times per week), a little bit (1–2 times per week), a moderate amount (3–4 times per week), quite a bit (5–6 times per week), or a great deal (everyday). The response categories were collapsed to “never” and “at least once per week” to be comparable with walking for transportation.

Analysis

Two sets of built environment summary variables were created. The first set of summary variables was based on SWEAT’S conceptual model, which classifies built environment features into four categories hypothesized a priori to be associated with walking: functionality, safety, aesthetics, and destinations.12 The second set of data-driven summary variables was created using PCA.

An exploratory PCA with varimax rotation was used to group built environment variables into theoretical indices. All variables were included unless they were visually highly skewed and unable to be transformed (to comply with assumptions of PCA). The appropriate number of factors to retain was determined by reviewing the resulting rotated factor pattern matrix and scree plot. Variables were considered to load on a given factor if the factor loading was at least .40 and did not load on more than one retained factor.19

Summary variables based on the PCA and a priori categorizations were constructed by summing built environment variables standardized to a normal distribution (mean = 0, SD = 1). Standardized variables were assigned negative values in the a priori measures if the variable was hypothesized to be inversely associated with walking, whereas variables were assigned negative values in the data-driven measures if the original (unstandardized) variable had a negative correlation greater than .15 with walking for transportation or exercise. PCA loading coefficients were not used in summation because they are often data driven and the aim of this study was to develop a robust composite measure.

Logistic regression was used to investigate the association between built environment summary variables and walking outcomes. Potential confounders that changed an environment measure’s odds ratio (OR) by more than 10% or were significantly related to environment measures and walking outcomes were retained in multivariable models. The significance of between-group differences in potential confounders was assessed using χ2 or Fisher’s exact test when cell values were less than five. Possible interaction effects were examined by stratifying models by covariates. Analyses were conducted using SAS, version 9.1 (SAS Institute lnc, Cary, North Carolina).

RESULTS

After interviewed participants were matched with audited segments, data from 120 adults who had three to 18 audited segments within their local neighborhoods were available for analysis. Men and minorities were undersampled, similar to other surveyed samples.20 Most participants were white (83.4%); of the 59 participants whose gender was recorded, 64.4% were female. On average, participants had lived in their neighborhoods for 13.7 years, and half reported engaging in some physical activity three or more times per week (50.8%). Table 2 provides the distribution of participant characteristics stratified by walking outcomes. Walking was a primary mode of transportation for 29.2% of participants, and 80.8% of participants walked for exercise at least once per week.

Table 2
Participant Characteristics by Use of Transportation and Exercise Walking

Twenty-six environmental variables were included in the PCA (Table 3) after exclusion of 11 visually highly skewed variables. Four components, accessibility, maintenance, comfort/safety, and pleasantness, were retained from the PCA, accounting for 57.2% of the total variance. Accessibility described the presence of transportation pathways and destinations. Presence of unbarred windows loaded in this component, likely because it was correlated with characteristics of commercial areas. The second component, maintenance, described the physical condition of environmental features. The third component, comfort/safety, described a pathway that may feel more comfortable and safe to a pedestrian because of a level sidewalk, separation from traffic, and low traffic volume. The fourth component, pleasantness, included features that may make a walking trip more serene or attractive, such as porches, buffer zones, and traffic-calming devices.

Table 3
Principal Component Factor Loadings and Percentage of Variance Explained*

Variables included in each built environment summary measure are shown in Table 4. Curb height was hypothesized to have an inverse association with walking and was assigned a negative value in the a priori measure calculation. No variables loading in PCA factors had negative correlations greater than .15 with walking for transportation or exercise; thus, none were assigned negative values.

Table 4
Composition of Built Environmental Measures*

The adjusted associations between environmental measures and walking outcomes are provided in Table 5. Walking for transportation was significantly associated with the PCA measures accessibility (95% OR = 1.15; CI, 1.04–1.26) and pleasantness (OR = 1.33; CI, 1.09–1.62) and with the a priori functionality measure (OR = 1.19; CI, 1.08–1.32). Walking for exercise was significantly associated with PCA comfort/safety (OR = 1.51; CI, 1.05–2.18) and marginally associated with PCA maintenance (OR = 1.15; CI, 1.00–1.33) and a priori safety (OR = 1.19; CI, 1.00–1.42).

Table 5
Odds Ratios for Environmental Measures and Likelihood of Walking*

To understand the effect of excluding non-normal variables from the PCA, a sensitivity analysis was performed using a third set of measures created by manually adding the exeluded variables into individual PCA summary measures, based on their interpreted meanings. Results of all subsequent analyses using these PCA-expanded measures were similar but slightly less significant than results of the original PCA measures, suggesting that the additional variables added noise to the measures rather than improving them.

DISCUSSION

Two methods of creating neighborhood-level measures from data obtained from GIS and SWEAT audits were compared in their associations with walking for transportation and exercise tising a sample of 120 community-dwelling adults in Portland, Oregon. Summary measures were created for the functionality, safety, aesthetics, and destinations indices on which SWEAT was based. A PCA created four stimmary measures interpreted as accessibility, maintenance, comfort/safety, and pleasantness.

A priori and PCA measures were analogous in conceptual meaning. A measure in each set described destinations (a priori destinations and PCA accessibility), functional elements (a priori, functionality and PCA pleasantness), appearance (a priori aesthetics and PCA maintenance), and safety (a priori safety and PCA comfort/safety). In addition, both sets of environmental measures had similar associations with walking for transportation and exercise. Participants who walked for transportation were more likely than those who did not to live in neighborhoods with high scores for destination accessibility and functional elements, such as wide sidewalks and buffer zones. On the other hand, participants who walked for exercise were more likely to live in well-maintained neighborhoods with protection from traffic. Additionally, no built environmental measure was associated with both transportation and exercise walking. This difference illustrates that environmental features that enable or discourage pedestrians depend on the purpose of the trip, confirming observations in previous studies.5,18,21

However, whereas built environmental measures created using the conceptual model and PCA are similar in meaning and association with walking for transportation and exercise, a priori measures consistently failed to meet statistical significance. Our findings suggest that the SWEAT conceptual framework should be refined. The a priori measures safety and destinations did not have a statistically significant association with walking, despite similarity in conceptual meaning and built environmental components to significant PCA measures. The failure of the a priori measures to meet statistical significance may be due to the inclusion of built environmental items unrelated to the other items in the measures, thereby reducing their efficiency. These findings stiggest that PCA may be a more efficient method of scoring data generated by SWEAT.

Built environmental features associated with walking in this study are consistent with existing research. Destinations, accessibility, and functional elements—appearing in accessibility, pleasantness, destinations, and functionality measures—were significantly associated with walking for transportation. This supports previous research finding walking for transportation to be more common in neighborhoods with greater ease of access18,21,22 and increased density and variety of destinations.18,21,23,24 Condition of facilities and pathways, pathway comfort, and traffic protection—forming maintenance, safety/comfort, aesthetics, and safety measures—were associated with walking for exercise, although the associations were not consistently significant. Previotis research has also identified aesthetics21 and safety25 as potential enablers of walking for exercise or leisure, but as in our study, the findings are not consistent. Safety has been found to be inversely related to walking.23 Additionally, other studies failed to identify any significant built environmental predictors of walking for recreation.22,23,26 In contrast to our findings, previous studies have suggested destinations, particularly shopping24 and recreational facilities,20 to be conducive to walking for recreation. This may not have been observed in this study becatise recreational destinations were combined with other destinations.

This study supports prior research that identified built environment correlates of walking for exercise or transportation and adds to the body of research by proposing a method of creating conceptually meaningful measures of neighborhood walkability from objective, microscale data. Despite the development of reliable audit tools such as SWEAT, which is capable of providing researchers a measure of specific built environment features, such tools are underutilized in research of the association between the built environment and physical activity. Although many recently published articles measured the built environment using GIS and existing databases,8,9,17,20,22,24,26-29 relatively few used audit instruments.15,16,18,23,30 Gathering microdata with audit tools remains more resource intensive than using existing GIS data, but the methodology outlined here can encourage the tise of audit tools to more precisely describe environments by providing researchers with summary measures that can be more easily incorporated into models than individual items. Furthermore, previotis stvidies that reduced audit tool data into summary measures focused on a priori hypotheses,18,23 whereas this study demonstrates that a data driven approach may yield more efficient measures.

Limitations of this study include the inability to control for potential confounders not incltided in the interview. We could not adjust for some potential confounders controlled for in other studies,8,9,15-17,21,31 such as age, socioeconomic status, educational attainment, and body mass index. In addition, participants may not be representative of the study population because canvassing typically occurred during the day on weekdays, when many working adults are not home. The relatively small number of participants who walked for transportation or did not walk for exercise limits the ability to find significant associations, perhaps explaining the marginally significant associations observed. Neighborhood built environment was represented by a sample of segments, sometimes as few as three, which may not be adequate for an acctirate description.

Although outcome measures were not independently validated, they do have face validity. The walking for transportation question was derived from the social capital module of the 2000 to 2001 General Household Survey of Great Britain,32 which underwent several rounds of testing prior to implementation. The walking for exercise question is similar to an item in the Neighborhood Physical Activity Questionnaire,33 which demonstrated excellent reliability (κ > .75).

The response rate in this study was low, although comparable to similar studies8,9,11,21,23 and surveys in the same population.28 Selection bias may have been introduced if the response rate was associated with both neighborhood built environment and walking behaviors. For instance, effect size may have been underestimated if nonresponders were more likely to be nonwalkers from low-walkability neighborhoods. However, this is unlikely because the average response rate in the most walkable neighborhoods (33.5%) was similar to the rate in the least walkable neighborhoods (30.6%). The low response rate did result in reduced power owing to a smaller number of participating residents.

This study was relatively small in size; however, the study had a priori 80% power to detect a 13 percentage point difference in walking prevalence with a one-SD change in the accessibility measure of interest, assuming a low-walkability prevalence of 40% to 60% suggested by previous studies.23,24,26,29,34,35 In addition, an appiopriate sample size for PCA may be estimated by five times the number of variables being analyzed.19 As the sample at the lower boundary suggested, the fact that the PCA constructs mapped well into a priori constructs and the direction of the results is similar between the two approaches provides confidence in the results of this current study. Validity of the methodologic approach proposed in this study can be further validated by using it in additional larger studies in more diverse geographic areas.

This study may not be generalizable outside of Portland, Oregon. Observation was done within established, traditional neighborhoods in a city in the Northwest United States. Studies in other communities with variations in environment and behavior are needed to assess whether the constructs identified here are reproduced and therefore valid outside of this study population.

This study provides researchers with an efficient method of creating a comprehensive and concise measure of the built environment, integrating two powerful and complementary objective measurement tools: audit instruments and GIS.

Acknowledgments

This project was supported in part by funding to Dr. Yvonne L. Michael from the Oregon Alzheimer Disease Center, Borchard Foundation Center on Law and Aging, National Institutes of Aging (G022240), and National Cancer Institute (GA109920). We thank Dr. Irina V. Sharkova for her contribution to the GIS analysis.

Appendix

Appendix

*

Local Neighborhood-Level
Variable
DescriptionMean (SD)Source
Residential densityPercent of buildings that are multiple-family dwellings0.14 (0.13)SWEAT
Nonresidential usePercent of buildings that are nonresidential (institutional, retail, commercial, public, religious)0.11 (0.14)SWEAT
Mixed usePercent of buildings that are mixed use0.01 (0.02)SWEAT
Short buildingsPercent of buildings less than 3 stories tall0.92 (0.11)SWEAT
Unbarred windowsPercent of buildings without bars on windows0.84 (0.10)SWEAT
PorchesPercent of residential buildings with porches0.54 (0.15)SWEAT
Well-maintained yardsPercent of segments with well-maintained yards0.82 (0.12)SWEAT
Well-maintained buildingsPercent of segments with well/fair-maintained buildings0.89 (0.12)SWEAT
Tree densityAverage number of trees ≥15 feet tall per 100 segment feet0.01 (0.00)SWEAT
Resting place densityAverage number of resting places per 100 segment feet0.00 (0.00)SWEAT
Litter scoreAverage litter rating across segments (1 = little litter, 0 = heavy litter)0.69 (0.21)SWEAT
RestroomsTotal number of restrooms1.00 (1.51)SWEAT
Streetlight densityAverage number of streetlights per 100 segment feet0.01 (0.00)SWEAT
Curbside parkingPercent of segment sides with curbside parking (for retail)0.10 (0.11)SWEAT
Lot parkingPercent of segment sides with parking lots (for retail) in front of buildings0.08 (0.09)SWEAT
Continuous sidewalksPercent of segment sides with continuous sidewalks0.96 (0.05)SWEAT
Gentle slopePercent of segment sides with flat/gentle sidewalk slope0.89 (0.14)SWEAT
Hard-surface sidewalksPercent of segment sides with concrete and/or asphalt sidewalks1.00 (0.00)SWEAT
Well-maintained sidewalksPercent of segment sides with sidewalks in good/fair condition0.59 (0.20)SWEAT
Wide sidewalksAverage sidewalk width (inches)63.89 (4.04)SWEAT
Unobstructed sidewalksPercent of segment sides without sidewalk obstructions0.45 (0.18)SWEAT
Buffer zonesPercent of segment sides with buffer zones0.91 (0.10)SWEAT
Buffer zone widthAverage buffer zone width (inches)61.65 (15.66)SWEAT
Through streetsPercent of segments that are through roads (vs. dead-ends)0.97 (0.05)SWEAT
Narrow roadsPercent of segments with 1 or 2 lanes of traffic0.93 (0.10)SWEAT
Bike lanesPercent of segments with bike lanes0.07 (0.11)SWEAT
Traffic-calming devicesPercent of segments with traffic-calming devices0.31 (0.18)SWEAT
Posted speed limitPercent of segments with speed limit ≤25 miles per hour0.97 (0.05)SWEAT
Crossing signalsPercent of signaled intersections with pedestrian signals0.60 (0.44)SWEAT
Crossing signal timeAverage pace (feet per second) to cross street during pedestrian signal1.15 (0.66)SWEAT
Curb cutsPercent of segments with curb cuts at crossings0.50 (0.19)SWEAT
Curb heightAverage curb height (inches)4.97 (0.43)SWEAT
Traffic loadAverage number of cars per minute3.16 (2.26)SWEAT
Block lengthAverage segment length (feet)318.24 (53.69)GIS/RLIS
Bus stopsNumber of bus stops within one-quarter of a mile13.08 (6.04)GIS/RLIS
RestaurantsNumber of restaurants within one-quarter of a mile3.69 (4.07)GIS/RLIS
ShopsNumber of retail/shopping businesses (neighborhood grocery, convenience stores,
 supermarkets, malls) within one-quarter of a mile
1.91 (1.64)GIS/RLIS
Medical facilitiesNumber of medical facilities (hospitals, medical offices) within one-quarter of a mile0.70 (0.91)GIS/RLIS
ServicesNumber of services (libraries, pharmacies, banks, beauty shops, schools, senior services)
 within one-quarter of a mile
3.12 (2.54)GIS/RLIS
ActivitiesNumber of activities (community centers, senior centers, pools, parks/trails, community
 gardens) within one-quarter of a mile
0.44 (0.79)GIS/RLIS
*SWEAT indicates Senior Walking Environmental Assessment Tool; GIS, geographic information systems; and RLIS, Regional Land Information System.
Excluded from analysis because of no variance across participants.
Excluded from analysis because of high percentage of missing values.

Contributor Information

Erin M. Keast, Oregon Health & Science University, Portland, Oregon.

Nichole E. Carlson, Oregon Health & Science University, Portland, Oregon.

Nancy J. Chapman, Portland State University, Portland, Oregon.

Yvonne L. Michael, Oregon Health & Science University and Kaiser Permanente Northwest, Portland, Oregon.

References

1. Frank LD, Engelke PO. The built environment and human activity patterns: exploring the impacts of urban form on public health. J Plann Lit. 2001;16:202–218.
2. Lee C, Vernez-Moudon A. Physical activity and environment research in the health field: implications for urban and transportation planning practice and research. J Plann Lit. 2004;19:147–181.
3. Patterson PK, Chapman NJ. Urban form and older residents’ service use, walking, driving, quality of life, and neighborhood satisfaction. Am J Health Promot. 2004;19:45–52. [PubMed]
4. Brownson RC, Chang JJ, Eyler AA, et al. Measuring the environment for friendliness toward physical activity: a comparison of the reliability of 3 questionnaires. Am J Public Health. 2004;94:473–483. [PMC free article] [PubMed]
5. Humpel N, Owen N, Iverson D, et al. Perceived environment attributes, residential location, and walking for particular purposes. Am J Prev Med. 2004;26:119–125. [PubMed]
6. Saelens B, Sallis JF, Black JB, Chen D. Neighborhood-based differences in physical activity: an environment scale evaluation. Am J Public Health. 2003;93:1552–1558. [PMC free article] [PubMed]
7. Drewnowski A, Rehm CD, Solet D. Disparities in obesity rates: analysis by ZIP code area. Soc Sci Med. 2007;65:2458–2463. [PMC free article] [PubMed]
8. Frank LD, Schmid TL, Sallis JF, Chapman J. Linking objectively measured physical activity with objectively measured urban form: findings from SMARTRAQ. Am J Prev Med. 2005;28:117–125. [PubMed]
9. King WC, Belle SH, Branch JS, et al. Objective measures of neighborhood environment and physical activity in older women. Am J Prev Med. 2005;28:461–469. [PubMed]
10. Nelson MC, Gordon-Larson P, Song Y, Popkin BM. Built and social environments: associations with adolescent overweight and activity. Am J Prev Med. 2006;31:109–117. [PubMed]
11. Coulton CJ, Korbin J, Chan T, Su M. Mapping residents’ perceptions of neighborhood boundaries: a methodological note. Am J Community Psychol. 2001;29:371–383. [PubMed]
12. Cunningham GO, Michael YL, Farquhar SA, Lapidus J. Developing a reliable senior walking environmental assessment tool. Am J Prev Med. 2005;29:215–217. [PubMed]
13. Day K, Boarnet MG, Alfonzo M, Forsyth A. The Irvine-Minnesota Inventory to measure built environments: development. Am J Prev Med. 2006;30:144–152. [PubMed]
14. Pikora T, Giles-Corti B, Bull F, et al. Developing a framework for assessment of the environmental determinants of walking and cycling. Soc Sci Med. 2003;56:1693–1703. [PubMed]
15. Hoehner C, Brennan Ramirez LK, Elliott MB, et al. Perceived and objective environmental measures and physical activity among urban adults. Am J Prev Med. 2005;28(2 Suppl 2):105–116. [PubMed]
16. Jago R, Baranowski T, Zakeri I, Harris M. Observed environmental features and the physical activity of adolescent males. Am J Prev Med. 2005;29:98–104. [PubMed]
17. Rutt CD, Coleman KJ. Examining the relationships among built environment, physical activity, and body mass index in El Paso, TX. Prev Med. 2005;40:831–841. [PubMed]
18. Pikora TJ, Giles-Corti B, Knuiman MW, et al. Neighborhood environmental factors correlated with walking near home: using SPACES. Med Sci Sports Exerc. 2006;38:708–714. [PubMed]
19. Hatcher L. A Step-by-Step Approach to Using SAS for Factor Analysis and Structural Equation Modeling. SAS Institute Inc; Cary, NC: 1994.
20. Owen N, Cerin E, Leslie E, et al. Neighborhood walkability and the walking behavior of Australian adults. Am J Prev Med. 2007;33:387–395. [PubMed]
21. Suminski RR, Carlos Poston WS, Petosa RL, et al. Features of the neighborhood environment and walking by U.S. adults. Am J Prev Med. 2005;28:149–155. [PubMed]
22. Nagel CL, Carlson NE, Bosworth M, Michael YL. The relation between neighborhood built environment and walking activity among older adults. Am J Epidemiol. 2008;168:461–468. [PMC free article] [PubMed]
23. Gauvin L, Riva M, Barnett T, et al. Association between neighborhood active living potential and walking. Am J Epidemiol. 2008;167:944–953. [PubMed]
24. McCormack GR, Giles-Corti B, Bulsara M. The relationship between destination proximity, destination mix and physical activity behaviors. Prev Med. 2008;46:33–40. [PubMed]
25. Humpel N, Owen N, Iverson D, et al. Perceived environment attributes, residential location, and walking for particular purposes. Am J Prev Med. 2004;26:119–125. [PubMed]
26. Lovasi GS, Moudon AV, Pearson AL, et al. Using built environment characteristics to predict walking for exercise. Int J Health Geographics. 2008;7:10. [PMC free article] [PubMed]
27. Boer R, Zheng Y, Overton A, et al. Neighborhood design and walking trips in ten U.S. metropolitan areas. Am J Prev Med. 2007;32:298–304. [PMC free article] [PubMed]
28. Li F, Fisher KJ, Brownson RC, Bosworth M. Multilevel modelling of built environment characteristics related to neighbourhood walking activity in older adults. J Epidemiol Community Health. 2005;59:558–564. [PMC free article] [PubMed]
29. Oliver LN, Schuurman N, Hall AW. Comparing circular and network buffers to examine the influence of land use on walking for leisure and errands. Int J Health Geographics. 2007;6:41. [PMC free article] [PubMed]
30. Suminski RR, Heinrich KM, Poston WS, et al. Characteristics of urban sidewalks/streets and objectively measured physical activity. J Urban Health. 2008;85:178–190. [PMC free article] [PubMed]
31. Van Lenthe FJ, Brug J, Mackenbach JP. Neighbourhood inequalities in physical inactivity: the role of neighbourhood attractiveness, proximity to local facilities and safety in The Netherlands. Soc Sci Med. 2004;60:763–775. [PubMed]
32. Coulthard M, Walker A, Morgan A. People’s Perceptions of Their Neighborhood and Community Involvement: Results From the Social Capital Module of the General Household Survey 2000. The Stationery Office; London, United Kingdom: 2002.
33. Giles-Corti B, Timperio A, Cutt H, et al. Development of a reliable measure of walking within and outside the local neighborhood: RESIDE’s Neighborhood Physical Activity Questionnaire. Prev Med. 2006;42:455–459. [PubMed]
34. Fisher KJ, Li F, Michael Y, Cleveland M. Neighborhood-level influences on physical activity among older adults: a multilevel analysis. J Aging Phys Act. 2004;12:45–63. [PubMed]
35. Simpson ME, Serdula M, Galuska DA, et al. Walking trends among US adults: the Behavioral Risk Factor Surveillance system. Am J Prev Med. 2003;25:95–100. [PubMed]

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

  • PubMed
    PubMed
    PubMed citations for these articles

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...