UWA Research Publication

Knuiman, M., Christian, H., Divitini, M., Foster, S., Bull, F., Badland, H. M., & Giles-Corti, B. (2014). A Longitudinal Analysis of the Influence of the Neighborhood Built Environment on Walking for Transportation. American Journal of Epidemiology, 180(5), 453-461. © The Author 2014. Published by Oxford University Press on behalf of the Johns Hopkins Bloomberg School of Public Health. All rights reserved.

This is a pre-copyedited, author-produced PDF of an article accepted for publication in American Journal of Epidemiology, following peer review. The version of record Knuiman, M., Christian, H., Divitini, M., Foster, S., Bull, F., Badland, H. M., & Giles-Corti, B. (2014). A Longitudinal Analysis of the Influence of the Neighborhood Built Environment on Walking for Transportation. American Journal of Epidemiology, 180(5), 453-461, is available online at: http://dx.doi.org/10.1093/aje/kwu171

This version was made available in the UWA Research Repository on the 1st of September 2015, in compliance with the publisher’s policies on archiving in institutional repositories.

Use of the article is subject to copyright law.

1

A longitudinal analysis of the influence of the neighborhood built

environment on transport walking: The RESIDE Study

Matthew W. Knuiman*, Hayley E. Christian, Mark L. Divitini, Sarah A. Foster, Fiona C. Bull, Hannah

M. Badland and Billie Giles-Corti

* Correspondence to Professor Matthew Knuiman, School of Population Health (M431), The University

of Western Australia, 35 Stirling Highway, Crawley WA 6009 Australia

Email: matthew.knuiman@uwa.edu.au

Telephone: 61-8-6488-1250

Fax: 61-8-6488-1188

Running head

Built environment and transport walking

Author affiliations

Centre for the Built Environment and Health, School of Population Health, The University of Western

Australia, Perth, Australia (Matthew Knuiman, Hayley Christian, Mark Divitini, Sarah Foster, Fiona

Bull); McCaughey VicHealth Centre for Community Wellbeing, Melbourne School of Population and

Global Health, University of Melbourne, Melbourne, Australia (Hannah Badland, Billie Giles-Corti).

Abbreviations

RESIDE, RESIDential Enviroments; SD, standard deviation; OR odds ratio; CI Confidence Interval

2

ABSTRACT

The purpose of this analysis was to use longitudinal data over seven years (four surveys) from the

Residential Environments (RESIDE) Study (Perth, Australia 2003-2012) to more carefully examine the

relationship of neighborhood walkability and destination accessibility to neighborhood transport walking

seen in many cross-sectional studies. We compared effect estimates from three types of logistic

regression models; two that utilise all available data (a population marginal model and a subject-level

mixed model), and a third subject-level conditional model that exclusively uses within-person

longitudinal evidence. The results support that neighborhood walkability (especially land use mix and

street connectivity) and local access to public transit stops and the variety of destinations are important

determinants of transport walking. The similarity of (subject-level) effect estimates from logistic mixed

models and conditional logistic models indicates there is little or no bias from uncontrolled time-constant

residential preference (self-selection) factors, but confounding by uncontrolled time-varying factors such

as health status remains a possibility. These findings provide policy-makers and urban planners with

further evidence of features of the built environment that may be important in the design of

neighborhoods to increase transport walking and meet the health needs of residents.

MESH-heading keywords:

Built environment, Logistic Regression, Longitudinal Study, Neighborhood, Transport, Urban Planning,

Walking

3

Over the last decade there has been increased research examining built environment features that

enable or facilitate or are barriers to improved health (1-5). A strong focus of this research has been

understanding the relationship between the neighborhood built environment and physical activity (2, 5-7)

and an effective strategy has been to estimate neighborhood built environment effects on neighbourhood

walking (8). In particular, measures of neighborhood walkability and the number and diversity of

destinations (land use mix) that are accessible by walking from home have received considerable

attention (9-13). In a recent review, 80% of studies reported a positive association between walking and

the presence and proximity of retail and service destinations (14). Overall, more built environment studies

have reported associations with transport-related walking compared with other types of physical activity

(11). Since the 1996 US Surgeon General’s report (15), promoting walking has been a global priority with

public health agencies actively promoting walking for transport (16).

However, most studies to date have relied upon cross-sectional analyses with few studies having a

longitudinal design (17-20). Cross-sectional studies are subject to confounding by unmeasured and thus

uncontrolled factors, including factors related to why people choose to live in certain neighborhoods

(often called self-selection factors) (21). Longitudinal or repeated measures data on a cohort of

individuals allow temporal order to be considered and the examination of within-person changes in both

built environment factors and physical activity behaviours. These within-person relationships, if analysed

appropriately, are not subject to confounding by time-constant (selection and other) factors as each

individual serves as his/her own control (22, 23). Thus longitudinal data, if analysed with appropriate

regression models, and if there is sufficient variation over time within as well as between individuals,

provides better evidence for likely causal relationships between the built environment and physical

activity behavior.

The RESIDential Environments (RESIDE) study provides a unique opportunity to examine

longitudinal data from a cohort of individuals and provide stronger evidence of a causal relationship

between the built environment and transport-related walking. Importantly, the study has been designed to

answer the question of self-selection (21, 24, 25). The primary aim of this analysis was to use longitudinal

4

data over seven years (four surveys) from the RESIDE study to examine neighborhood walkability and

destination accessibility in relation to neighborhood transport walking. A secondary aim was to examine

these relationships using three types of logistic regression models, two that utilise all available data (a

population marginal model and a subject-level mixed model) and a third subject-level conditional model

that exclusively uses within-person longitudinal evidence, and to compare effect estimates from these

different models. The comparison of effect estimates from the subject-level mixed and conditional models

enables assessment of bias arising from unmeasured time-constant confounders such as self-selection

factors.

MATERIALS AND METHODS

Sample and data collection

The RESIDE study commenced in 2003, and is a longitudinal natural experiment of 1813 people

building homes in 73 new housing developments across metropolitan Perth, Western Australia. Details of

participant recruitment procedures have been reported elsewhere (26). Briefly, participants moving to

each development were invited to participate by the state water authority following the land transfer

transaction. The following eligibility criteria were applied: English proficiency; 18 years; intention to

relocate by December 2005; and willingness to complete surveys four times over seven years. Participants

were recruited by telephone and one person from each household randomly selected. Participants were

surveyed four times: baseline (T1: n=1813); then approximately one (T2: n=1467), three (T3: n=1230),

and seven years (T4: n=565) later. The numbers included in analyses are slightly less than these figures

due to missing data. Almost all participants (99%) moved to their new home between T1 and T2 and 10%

moved home after T2. The University of Western Australia’s Human Research Ethics Committee

(#RA/4/1/479) provided ethics approval.

Measures

Outcome variable Participants reported the frequency of transport-related walking within their

neighborhood (defined as a 15 minute walk from their home) over a usual week using the Neighborhood

Physical Activity Questionnaire (27). This questionnaire has been shown to have acceptable reliability

(27). The frequency of neighbourhood transport walking was reduced to a binary yes/no variable for the

5

main analyses as the majority of participants did no neighbourhood transport walking and the distribution

was heavily skewed (68% did none, mean frequency was 1.2 times/week and median was zero) and

dominated by the none/some dichotomy. Thus the focus was on factors that influenced whether or not

participants did any neighbourhood walking for transport rather than the amount (for those that did some)

and the results therefore directly facilitate the study of built environment interventions aimed at

increasing neighborhood transport walking specifically through encouraging the uptake or maintenance of

the behavior.

Adjustment variables All models adjusted for age, gender, marital status, education level,

occupation (including whether or not in the workforce), hours of work per week, household income,

number of adults in household, children at home and motor vehicle access (see Table 1). These were

included as time-varying factors in all models.

Objective measures of the neighborhood environment At each time point objective built

environment measures were generated using Geographic Information Systems, and included

(standardised) neighbourhood walkability measures street connectivity, residential density and land use

mix which were calculated for the area accessible along the street network within 1600m from the

participant’s home(28, 29). The measure of land use mix used was “Model 2” in Christian et al(30). The

count of different types of services (dry cleaner, post office, pharmacy, video store; range 0-4),

convenience stores (deli, general store, supermarket, green grocer, seafood shop, petrol station shop, other

food shop, shopping center; range 0-8), public open space destinations (park, sports field, beach; range 0-

3), and bus stops, and railway station (yes/no) within a 1600m service area of participant’s homes were

also calculated. A commercial electronic database of services and stores (Sensis Pty Ltd, Sydney,

Australia) was used to generate these counts based on the geocoded locations of destinations and

participants’ homes. There is moderate to good agreement between these measures and what is actually

on the ground (31).

Perceptions of the neighborhood environment The count of different types of services (dry cleaner,

post office, pharmacy, video store; range 0-4), convenience stores (local shop, supermarket, green grocer,

petrol station shop; range 0-4), and public open space destinations (park, sports field, beach; range 0-3),

6

and the presence of a bus stop (yes/no) and railway station (yes/no) participant’s perceived they had

access to within the neighborhood (≤15min walk from home) were calculated from responses to the

Neighborhood Environment and Walking Scale questionnaire and which have been demonstrated as

reliable(32).

Statistical analysis

We used logistic regression models as they are the most commonly used regression method for

binary (dichotomous) outcome data and because there are well established and reliable approaches for

fitting the three different types of logistic regression models. All models were fitted using SAS software

(Statistical Analysis System, SAS Institute, Inc., Cary, North Carolina). The first type of logistic model

was a marginal model that was fitted using all available data from the four time points for each person. It

provided population-average estimates of the effects of factors on neighborhood transport walking

(yes/no) using Generalised Estimating Equations with robust standard errors. This allowed for repeated

measures on the same person over time to be correlated. SAS Proc Genmod with a Repeated statement

was used to fit these marginal models. The second model that also used all available data was a

conditional model that included a random effect for each subject (which allowed a correlation between

repeated measures on the same subject over time). This type of model provided subject-level estimates of

the effects of factors. SAS Proc Glimmix with a Random statement was used to fit these conditional

models. This second type of model is often termed a logistic mixed model due to measuring both fixed

and random effects. If the factor is time-varying, the estimated effect of the factor from these first two

types of model is a combined cross-sectional and within-person change effect estimate(22, 33). The third

was a conditional model that included a fixed effect for each subject (which absorbed the effects of all

measured and unmeasured time-constant factors) and provided subject-level estimates of the effects of

time-varying factors. SAS Proc Logistic with a Strata statement was used to fit this model (commonly

termed conditional logistic regression). The effects of time-constant factors cannot be estimated in this

third type of model and, because the estimated effect of a time-varying factor is based entirely on within-

person changes, there is no potential for bias from measured or unmeasured time-constant

confounders(22, 33).

7

As models with the separate counts of destination types for services, convenience stores and public

open spaces did not provide a better fit than the models with the total number of types of these

destinations, only fitted models with the total counts of types of destinations are shown (for both

objective and perceived destinations).

RESULTS

Baseline socio-demographic characteristics of study cohort

At baseline, the mean age of participants was 40 years, 60% were female, 82% were either married

or in a de facto relationship (i.e. living with partner but not married), 23% had a bachelor degree or

higher, 43% had professional or managerial-administrator occupations and 25% lived in households

earning ≥$90,000 per year. Many participants worked 39-59 hours per week (43%), had access to a motor

vehicle (98%), and lived in a household which had two adults (59%) and children (49%) living at home

(Table 1).

Transport walking and characteristics of the neighborhood environment over time

At baseline (T1) 37% of participants did some neighborhood transport walking, with 28% at one

(T2), 29% at three (T3), and 36% at seven (T4) years follow-up. The mean frequency of walking/week

for transport in the neighborhood showed a similar pattern of change over time; 1.4 trips at baseline with

a decrease at one (1.1) and three years (1.1) follow-up but then returning to baseline levels at seven years

follow-up (1.4).

Many features of the neighborhood environment of the participants decreased substantially upon re-

location to a new housing development (between T1 and T2), and then gradually increased but even by

seven years follow-up (T4) many had not yet returned to their baseline levels (number of bus stops,

railway station, types of services and convenience stores) (Table 2). Access to types of public open spaces

was similar across all time points. Street connectivity increased slightly over time, land use mix decreased

between T1 and T2 and then remained stable, and residential density was relatively stable over all time

points.

Effect of the built environment on neighborhood transport walking over time

8

Table 3 shows the multivariate model estimates obtained from the three different modelling

approaches with objective measures of neighborhood walkability and objective measures of

neighbourhood public transit and destinations. Table 4 shows estimates from the equivalent model with

objective measures of neighborhood walkability and perceived measures of neighbourhood public transit

and destinations.

Both the connectivity and land use mix walkability components (but not residential density), and

neighborhood access to public transit were significantly related to walking for transport in the

neighborhood and this was consistent across all three analysis methods. Connectivity had an estimated

population-average odds ratio of 1.09 per unit change (Method I) and estimated subject-level odds ratios

of 1.12 and 1.13 for Methods II and III respectively. Land use mix had an estimated population-average

odds ratio per unit change of 1.21 (Method I) and estimated subject-level odds ratios of 1.29 and 1.33 for

Methods II and II respectively. Participants with 30+ bus stops within 1600m had approximately doubled

odds of walking for transport compared to participants with 0-14 bus stops and the presence of a train

station within 1600m increased the odds of walking for transport by about 50%. The objectively

measured count of types of destinations was significantly related to walking for transport in a dose-

response manner (trend P-values < 0.05) but when categorised into levels the odds ratio comparing

participants’ neighborhoods with 8-15 to 0-3 destinations types within 1600m was only about 1.3 (P

value= 0.04).

When the objective public transit and destinations measures (Table 3) were replaced by the

measures of perceived access to public transit and destinations (Table 4), the estimated odds ratios for the

walkability measures were slightly attenuated, the odds ratio for railway station access was essentially the

same, and although the perceived bus stop access measure was binary (yes/no) it remained significant in

Methods I and II. There was however a substantial increase in the strength and significance of the

relationship with the total count of types of destinations when the objective measure was replaced with

the perceived measure. The odds of walking for transport more than doubled when comparing participants

who perceived access to 7-11 compared with 0-2 destinations types within a 15 minute walk from home.

DISCUSSION

9

In our fitted models of population-average and subject-level (within and between) effects on

transport walking we found that neighborhood connectivity and land use mix, public transit stops, and the

diversity of destination types that were accessible from home by walking, were significantly related to

neighborhood transport walking. These longitudinal results were evident in three types of regression

models and provide stronger evidence than is available from previous findings from cross-sectional

studies (11, 14, 34).

Our longitudinal results on measures of walkability support several cross-sectional studies (35-37)

as well as studies of self-selection, neighborhood design and walking (11, 38-40). However, whereas we

found both connectivity and land use mix to be positively related to transport walking, a small

longitudinal study of neighborhood design and women’s pedometer-measured walking found that

increases in land use mix were associated with less walking (ie negative relationship) but fewer cul de

sacs (higher connectivity) was positively associated with more walking (41). Our results suggest that the

three traditional components of walkability (street connectivity, residential density and land use mix) are

not always equally important determinants of transport walking, and perhaps should not always be

weighted equally when constructing overall walkability indices (42). In our study, land use mix had a

greater and more significant effect than either street connectivity or residential density. Of the few

longitudinal studies conducted to date there is some evidence to suggest that increased mix of businesses

(25, 43) and more public and pay recreational facilities are determinants of walking and physical activity

(23). The relative importance of the three components of walkability however may differ across studies

and geographic location as the magnitude of effects per standard deviation change (i.e. per unit of Z-

score) are influenced by the degree of heterogeneity in these component measures within the population

under study. Perth is a relatively low density city with small variation in these measures - in the

multivariate models the estimated odds ratio for land use mix was modest at around 1.25 for a one

standard deviation change. Interactions between these variables could not be adequately investigated in

our study due to the relatively small variation in these measures and the moderate sample size but remain

a possibility. For example, as pointed out in a review report(6), high residential density is likely to be a

pre-requisite for the presence of mixed land uses, as without increased density there are insufficient

10

people to support local shops services, and public transit. Newman and Kenworthy (44) have observed

that both housing and employment densities of at least 35 per hectare are required to support public

transit. Thus, future research should examine the relative importance of, and interactions between, the

three components of neighborhood walkability in cities with greater variation in these measures.

Our longitudinal results, as expected, confirm that access to public transit (bus, rail) is a key

determinant of walking for transport. As the RESIDE study involves a cohort of people who re-located

(between T1 and T2) to new greenfield developments mostly on the outer regions of the Perth

metropolitan areas, there was a decrease in the number of neighborhood public transit stops from T1 to

T2 and this has previously been shown to be related to a reduction in their local transport-related walking

(18). The current results utilising T1 through T4 data and a variety of regression modelling approaches

re-confirm this relationship. Results from two transit rail quasi-experiments (45-47) show that the

implementation of rail infrastructure may increase walking. Our results support this since, although only a

small percentage of participants had local access to a train station it was (independently) related to

walking for transport with an odds ratio of about 1.5.

Our longitudinal results also confirm that local access to a variety of destinations is an important

determinant of walking for transport. However, the relationship was much stronger for the measure of

perceived access to types of destinations than for the objective Geographic Information Systems -derived

measure. Further, when both the perceived and objective measures were included together in the same

multivariate models (results not shown), the perceived measure remained highly significant and the

objective measure no longer significant. This may be because perceptions rather than objective measures

are more proximal to the individual’s intention, self-efficacy and decision-making in relation to walking

for transport. Alternatively, the weaker relationship with objective measures could reflect measurement

error and/or under-counting of the number of types of destinations using the destination databases

available for this analysis(31). Nevertheless without actual destinations being present, it would not seem

possible for participants to hold positive perceptions about their presence and so the presence of

accessible destinations remains important.

11

Longitudinal studies, with repeated measures on a cohort of individuals, allow both cross-sectional

(i.e., between people) comparisons and within-person comparisons. Residential preferences (i.e., self-

selection factors) are an important issue in ascertaining the causal effect of the neighborhood environment

on transport walking. Thus direct comparison of (subject-level) effect estimates from logistic mixed

models (method II) and conditional logistic models (method III) enables an assessment of the influence of

uncontrolled (time-constant) self-selection and other confounders. If the effect estimates are similar then

there can be no bias from uncontrolled confounding by time-constant factors in the logistic mixed model

(method II), therefore, its effect estimate is preferred as it uses all available cross-sectional and

longitudinal evidence. Furthermore, the effect estimate from the corresponding logistic Generalised

Estimating Equations Generalised Estimating Equations model (method I) is thus also preferred but of

course it provides population-average estimates. If the logistic mixed model (method II) and conditional

logistic model (method III) produce different effect estimates then the presumption is that there is

confounding by time-constant factors and so estimates from the conditional logistic model (method III)

which are not affected by this confounding are preferred. Note however, that as the conditional logistic

model only uses within-person (ie longitudinal) comparisons, its estimate will be less precise (i.e., greater

standard error).

In our multivariate models, we found similar effect sizes from methods II and III for all

neighborhood environment factors suggesting little confounding by uncontrolled time-constant factors,

including time-constant selection factors. As expected, we also observed more precise (confidence

intervals not as wide) effect estimates from method II as compared with method III. Further, as is

generally the case, the effect sizes were slightly larger in the subject-level logistic mixed model (method

II) than in the equivalent marginal model (method I). That is, the effect of these environmental factors is

greater when considered at the individual-level than when considered at a population average-level. The

estimated subject-level odds ratio represents the estimated increase in the odds of an individual doing

transport walking if his/her environmental factor level changes. It is important to note that logistic models

applied to cross-sectional data produce population-average effect estimates and hence can only be validly

compared with effect estimates from method I for longitudinal data.

12

Note that whilst comparison of estimates from the Method II and Method III models does allow

assessment of potential confounding by unmeasured time-constant (eg self-selection) factors it does not

address potential confounding by unmeasured time-varying factors. Uncontrolled time-varying factors

that are associated with change in walking behaviour and also associated with changes in the built

environment may lead to bias in the estimated effect of the built environment. Changing health status and

demographic transitions such as marriage, divorce, having children, and leaving workforce may be

associated with change in walking behaviour and also likely to be associated with change in built

environment if the participant re-locates due to the change in health or demographic status. In our models

we have adjusted for demographic transitions but not for changing health status so bias from changing

health status remains a potential source of bias.

A limitation and further potential source of bias in effect estimates from this study is the drop-out of

participants (especially between T3 and T4). An analysis of the factors related to participant drop-out

revealed that drop-out was associated with some demographic variables (age, sex, children at home) but

was not related to prior values of walking behaviour (the outcome variable). When drop-out is related to

covariates only and not to prior or missing values of the outcome variable, the drop-out pattern is called

(conditionally on the covariates) missing completely at random. Estimates from method I are unbiased

under this pattern of drop-out (provided the covariates related to drop-out are included in the models and

that there are no further unmeasured covariates related to drop-out) and estimates from the likelihood-

based Methods II and III are unbiased under this drop-out pattern and also under the less restrictive

pattern of missing at random where drop-out can also be related to the prior value of the outcome variable

(22). If drop-out is related to the missing values of the outcome variable (this cannot be checked from the

data) the drop-out pattern is called non-ignorable and all methods may produce biases estimates.

In a recent Australian study physical inactivity ranked fourth in terms of measured population

health loss(48) and since this is largely amenable to change there is a need to identify effective built

environment supports for increased physical activity. This longitudinal study provides stronger evidence

than cross-sectional studies of a possible supportive effect of the neighborhood built environment on local

transport-related walking. The results suggest that people living in neighborhoods with mixed land uses,

13

connected streets, access to public transit and a diverse number of destinations (services, convenience

stores, public open space) benefit from higher levels of local transport walking. These findings provide

policy-makers and urban planners with further evidence of features of the built environment that may be

important in the design of neighborhoods to increase transport walking and meet the health needs of

residents. Future research should examine the relative importance of the components of neighborhood

walkability in cities with greater variation in these measures. They should also consider using the

analytic methods described here when examining the relationships between the built environment and

walking.

ACKNOWLEDGEMENTS

The RESIDE Study was funded by grants from the Western Australian Health Promotion Foundation

(11828), the Australian Research Council (LP0455453) and an Australian National Health and Medical

Research Council Capacity Building Grant (458688). BG-C was supported by an National Health and

Medical Research Council Principal Research Fellow Award (1004900). HEC is supported by an

National Health and Medical Research Council /National Heart Foundation Early Career Fellowship

(1036350), and SAF is supported by a Healthway Health Promotion Research Fellowship (21363). The

authors gratefully acknowledge the Geographic Information Systems team (Nick Middleton, Sharyn

Hickey, Bridget Beasley and Dr Bryan Boruff) for their assistance in the development of the Geographic

Information Systems measures used in this study.

Conflict of interest: None declared

14

Table 1. Baseline Socio-demographic Characteristics of RESIDE Study Cohort, Perth, Western Australia

2003-2005 (n=1,703).

Characteristic %

Female 59.8

Age in yearsa 39.9 (11.8)

Marital status

Married/de facto 81.6

Separated/divorced/widowed 7.6

Single 10.8

Education level

Secondary or less 39.4

Trade/apprenticeship/certificate 37.4

Bachelor degree or higher 23.2

Occupation

Manager/administrator 15.0

Professional 28.0

Blue collar 16.5

Clerical/sales/service/other 23.0

Not in workforce 17.4

Hours of work per week

≤ 19 10.8

20 – 38 24.4

39 – 59 42.6

≥ 60 4.7

Not in workforce 17.4

15

Household income

≤ $50,000 25.7

$50 – 69,000 25.0

$70 – 89,000 23.2

≥ $90,000 26.0

Number of adults in household

One 16.4

Two 59.4

Three 12.7

≥ Four 11.5

Children at home 49.0

Access to a motor vehicle 98.1

a Value is mean (SD)

16

Table 2: Objective and Perceived Neighborhood Environment of Participants at Baseline (T1) and in New Neighborhood at one (T2), three (T3) and seven

(T4) Years Follow up: RESIDE Study, Perth, Western Australia, 2003 -2012.

T1

(n=1703)

T2

(n=1273)

T3

(n=1150)

T4

(n=504)

Neighborhood environment measure Mean (SD) % Mean (SD) % Mean (SD) % Mean (SD) %

Connectivity (z-score) -0.0 (1.0) 0.7 (1.5) 1.0 (1.4) 1.2 (1.5)

Residential density (z-score) -0.0 (1.0) -0.3 (0.7) -0.1 (0.5) -0.1 (0.5)

Land use mix (z-score) -0.0 (1.0) -0.5 (0.9) -0.7 (0.7) -0.5 (1.0)

Number of bus stopsa

0-14

15-29

30+

37 (21)

10

32

58

21 (14)

37

43

20

23 (15)

35

43

22

24 (13)

21

53

27

Railway station presentb 10 1 2 4

Count of type of servicesc 1.3 (1.2) 0.7 (1.1) 0.8 (1.2) 0.9 (1.1)

Count of type of convenience storesd 2.2 (1.9) 0.7 (1.3) 0.8 (1.4) 1.1 (1.5)

Count of type of public open spacese 1.2 (0.5) 1.1 (0.4) 1.1 (0.4) 1.1 (0.5)

Total count of types of destinationsf

0-3

4-7

8-15

4.7 (2.9)

44

35

21

2.4 (2.4)

79

14

7

2.6 (2.5)

79

13

8

3.1 (2.6)

68

23

9

Perceived access to bus stopg 85 84 88 91

Perceived access to railway stationh 10 3 6 6

17

Count of perceived access to types of servicesi 1.6 (1.6) 0.7 (1.2) 0.9 (1.4) 1.1 (1.3)

Count of perceived access to types of

convenience storesj

2.0 (1.5) 0.9 (1.3) 1.2 (1.5) 1.7 (1.5)

Count of perceived access to types of public

open spacek

1.6 (0.7) 1.5 (0.7) 1.5 (0.7) 1.6 (0.7)

Total count of perceived access to types of

destinationsl

0-2

3-6

7-11

5.2 (3.1)

28

34

38

3.1 (2.7)

62

24

15

3.7 (3.0)

53

25

22

4.4 (3.0)

39

34

27

T1= time point 1 (baseline); T2= time point 2 (1 year); T3=time point 3 (3 years); T4=time point 4 (7 years); SD = standard deviation a Number of bus stops within 1600m. b Railway station present within 1600m. c Count of types of services: dry cleaner, post office, pharmacy, video store (range 0-4) within 1600m. d Count of types of convenience stores: deli, general store, supermarket, green grocer, seafood shop, petrol station shop, other food shop, shopping centre (range 0-8) within

1600m. e Count of types of public open spaces: parks ≥2 ha, sports fields, beach (range 0-3) within 1600m. f Total count of types of services, convenience shops and public open spaces (range 0-15) within 1600m. g Perceived access to bus stop within 15minute walk from home. h Perceived access to railway station within 15minute walk from home. i Count of types of services: dry cleaner, post office, pharmacy, video store (range 0-4) participant perceives has access to within 15minute walk from home. j Count of types of convenience stores: local shop, greengrocer, petrol station shop, supermarket (range 0-4) participant perceives has access to within 15minute walk from

home. k Count of type of public open spaces: parks, sports fields, beach (range 0-3) participant perceives has access to within 15minute walk from home. l Total count of types of services, convenience shops and public open spaces (range 0-11) participant perceives has access to within 15minute walk from home.

18

Table 3. Multivariate Models of the Effect of Neighborhood Walkability and Objective Public Transit and Destination Measures of the Built Environment

on Neighborhood Transport Walking Over Time Using Three Different Analysis Methods. RESIDE Study, Perth, Western Australia, 2003 -2012.

Analysis method I:

Marginal-Population average

(N=1703 participants with 4630

observations)

Analysis method II:

Conditional-Subject level-Mixed

(N=1703 participants with 4630

observations)

Analysis method III:

Conditional-Subject level-Fixed effect

(N=1703 participants with 2858

within-person comparisons)

OR 95% CI OR 95% CI OR 95% CI

Connectivity (z-score) 1.09 1.03 , 1.15 1.12 1.04 , 1.21 1.13 1.01 , 1.26

Residential density (Z-score) 1.02 0.92 , 1.14 1.03 0.90 , 1.18 0.96 0.80 , 1.15

Land use mix (z-score) 1.21 1.12 , 1.30 1.29 1.17 , 1.43 1.33 1.16 , 1.52

Number of bus stopsa

0 – 14 1.00 1.00 1.00

15 – 29 1.63 1.34 , 1.98 1.88 1.49 , 2.39 1.99 1.46 , 2.71

≥ 30 1.75 1.39 , 2.19 2.07 1.56 , 2.74 2.33 1.57 , 3.45

Railway station presentb 1.34 1.00 , 1.81 1.52 1.03 , 2.26 1.79 1.02 , 3.16

Total count of types of destinationsc

0 – 3 1.00 1.00 1.00

4 – 7 1.03 0.87 , 1.22 1.03 0.82 , 1.29 1.08 0.80 , 1.45

19

8 - 15 1.29 1.02 , 1.64 1.38 1.02 , 1.87 1.40 0.93 , 2.10

CI=confidence interval; OR=odds ratio; All models adjusted for age, gender, marital status, education level, occupation (including whether or not in the workforce), hours

of work per week, household income, number of adults in household, children at home, motor vehicle access and time. a Number of bus stops within 1600m. b Railway station present within 1600m. c Total count of types of services, convenience shops and public open spaces (range 0-15) within 1600m.

20

Table 4. Multivariate Models of the Effect of Neighborhood Walkability and Perceived Public Transit and Destinations Measures of the Built Environment

on Neighborhood Transport Walking Over Time Using Three Different Analysis Methods. RESIDE Study, Perth, Western Australia, 2003 -2012.

Analysis method I:

Marginal-Population average

(N=1703 participants with 4630

observations)

Analysis method II:

Conditional-Subject level-Mixed

(N=1703 participants with 4630

observations)

Analysis method III:

Conditional-Subject level-Fixed effect

(N=1703 participants with 2858

within-person comparisons)

OR 95% CI OR 95% CI OR 95% CI

Connectivity (z-score) 1.05 0.99 , 1.11 1.06 0.99 , 1.15 1.07 0.95 , 1.19

Residential density (z-score) 1.04 0.94 , 1.15 1.05 0.92 , 1.19 0.97 0.81 , 1.15

Land use mix (z-score) 1.16 1.08 , 1.25 1.22 1.10 , 1.34 1.27 1.11 , 1.45

Perceived access to bus stopa 1.35 1.10 , 1.66 1.49 1.14 , 1.96 1.31 0.92 , 1.87

Perceived access to railway stationb 1.44 1.13 , 1.85 1.64 1.17 , 2.28 1.80 1.13 , 2.85

Total count of perceived access to

types of destinationsc

0 – 2 1.00 1.00 1.00

3 – 6 2.07 1.76 , 2.43 2.58 2.09 , 3.20 2.35 1.81 , 3.05

7 - 11 2.32 1.95 , 2.77 3.01 2.38 , 3.80 3.11 2.28 , 4.25

CI=confidence interval; OR= odds ratio; All models adjusted for age, gender, marital status, education level, occupation (including whether or not in the workforce), hours

of work per week, household income, number of adults in household, children at home, motor vehicle access and time. a Perceived access to bus stop within 15minute walk from home. b Perceived access to railway station within 15minute walk from home.

21

c Total count of types of services, convenience shops and public open spaces (range 0-11) participant perceives has access to within 15minute walk from home.

22

REFERENCES

1. Corburn J. Special section on urban planning and public health. J Am Plann Assoc 2006;72(1):119-121. 2. Humphreys DK, Ogilvie D. Social inequalities in physical activity: Do environmental and policy

interventions help to reduce the gap? A pilot systematic review. Lancet 2012;380:S50. 3. Renalds A, Smith TH, Hale PJ. A systematic review of built environment and health. Fam Community

Health 2010;33(1):68-78. 4. Barton H. Land use planning and health and well-being. Land Use Policy 2009;26(Sup1):S115-S23. 5. Kopelman P, Jebb SA, Butland B. Executive summary: Foresight tackling Obesities: Future choices' project.

Obes Rev 2007;8 Suppl 1:vi-ix. 6. National Research Council, Committee on Physical Activity Land Use Institute of Medicine. Does the built

environment influence physical activity? Examining the evidence. Washington, DC, USA: Transportation Research Board National Research; 2005.

7. Badland H, Schofield G. Transport, urban design, and physical activity: An evidence-based update. Transportation Res D-Tr E 2005;10(3):177-196.

8. Giles-Corti B, Timperio A, Bull F, et al. Understanding physical activity environmental correlates: Increased specificity for ecological models. Exercise and Sport Sciences Reviews 2005;33(4):175-181.

9. Durand CP, Andalib M, Dunton GF, et al. A systematic review of built environment factors related to physical activity and obesity risk: Implications for smart growth urban planning. Obes Rev 2011;12(5):e173-e182.

10. Lee C, Moudon AV. Physical activity and environment research in the health field: Implications for urban and transportation planning practice and research. J Plan Lit 2004;19(2):147-181.

11. McCormack G, Shiell A. In search of causality: A systematic review of the relationship between the built environment and physical activity among adults. Int J Behav Nutr Phy 2011;8(1):125-136.

12. Saelens B, Sallis JF, Frank L. Environmental correlates of walking and cycling: Findings from the transportation, urban design, and planning literatures. Ann Behav Med 2003;25:80-91.

13. Ferdinand AO, Sen B, Rahurkar S, et al. The Relationship Between Built Environments and Physical Activity: A Systematic Review. Am J Public Health 2012;102(10):e7-e13.

14. Sugiyama T, Neuhaus M, Cole R, et al. Destination and route attributes associated with adults' walking: A review. Med Sci Sports Exer 2012;44(7):1275-1286.

15. Services. UDoHaH. Physical activity and health: a report of the Surgeon General. Atlanta GA: U.S. Department of Health and Human Services, Centres for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, 1996.

16. World Health Organization. Interventions on diet and physical activity: what works: summary report. Geneva: WHO 2009.

17. Giles-Corti B, Bull F, Knuiman M, et al. The influence of urban design on neighbourhood walking following residential relocation: Longitudinal results from the RESIDE study. Soc Sci Med 2013;77:20-30.

18. Christian H, Knuiman M, Bull F, et al. A new urban planning code’s impact on walking: The Residential Environments Project. Am J Public Health 2013;103(7):1219-1228.

19. Ding D, Gebel K. Built environment, physical activity, and obesity: What have we learned from reviewing the literature? Health Place 2012;18(1):100-105.

20. Larouche R, Trudeau F. Active commuting: Its impact on physical activity and health, and its main determinants. Sci Sport 2010;25(5):227-237.

21. Mokhtarian PL, Cao X. Examining the impacts of residential self-selection on travel behavior: A focus on methodologies. Transportation Res B-Meth 2008;42(3):204-228.

22. Fitzmaurice G, Laird N, Ware J. Applied longitudinal analysis. 2nd ed.: Wiley; 2011: Hoboken, New Jersey. 23. Boone-Heinonen J, Guilkey D, Evenson K, et al. Residential self-selection bias in the estimation of built

environment effects on physical activity between adolescence and young adulthood. Int J Behav Nutr Phy 2010;7(70):1-11.

24. Cao X, Mokhtarian PL, Handy SL. Examining the impacts of residential self-selection on travel behaviour: A focus on empirical findings. Transport Rev 2009;29(3):359-395.

25. Handy S, Cao X, Mokhtarian PL. Self-selection in the relationship between the built environment and walking: Empirical evidence from Northern California. J Am Plann Assoc 2006;72(1):55-74.

23

26. Giles-Corti B, Knuiman M, Timperio A, et al. Evaluation of the implementation of a state government community design policy aimed at increasing local walking: Design issues and baseline results from RESIDE, Perth Western Australia. Prev Med 2008;46(1):46-54.

27. 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(6):455-459.

28. Frank L, Schmid T, Sallis J, et al. Linking objectively measured physical activity with objectively measured urban form: findings from SMARTRAQ. Am J Prev Med 2005;28(2S2):117-125.

29. Villanueva K, Pereira G, Knuiman M, et al. The impact of the built environment on health across the life course: design of a cross-sectional data linkage study. Bmj Open 2013;3(1).

30. Christian H, Bull F, Middleton N, et al. How important is the land use mix measure in understanding walking behaviour? Results from the RESIDE study. Int J Behav Nutr Phy 2011;8:55-67.

31. Hooper P, Middleton N, Knuiman M, et al. Measurement error in studies of the built environment: Validating commercial data as objective measures of neighbourhood destinations. J Phys Act Health 2013;10(6):792-804.

32. Cerin E, Saelens BE, Sallis JF, et al. Neighborhood Environment Walkability Scale: validity and development of a short form. Med Sci Sports Exerc 2006;38(9):1682-1691.

33. Allison PD. Fixed effects regression models for longitudinal data using SAS. Cary NC: SAS Institute Inc; 2005.

34. Saelens B, Handy S. Built environment correlates of walking: A review. Med Sci Sports Exerc 2008;40(7):S550-S566.

35. Frank LD, Saelens BE, Powell KE, et al. Stepping towards causation: Do built environments or neighborhood and travel preferences explain physical activity, driving, and obesity? Soc Sci Med 2007;65(9):1898-1914.

36. Owen N, Cerin E, Leslie E, et al. Neighborhood walkability and the walking behavior of Australian adults. Am J Prev Med 2007;33(5):387-395.

37. Sallis JF, Saelens BE, Frank LD, et al. Neighborhood built environment and income: Examining multiple health outcomes. Soc Sci Med 2009;68(7):1285-1293.

38. Khattak AJ, Rodriguez D. Travel behavior in neo-traditional neighborhood developments: A case study in USA. Transportation Res A-Pol 2005;39(6):481-500.

39. Cao XJ. Exploring causal effects of neighborhood type on walking behavior using stratification on the propensity score. Environ Plann A 2010;42(2):487-504.

40. Cao X, Mokhtarian PL, Handy SL. The relationship between the built environment and nonwork travel: A case study of Northern California. Transportation Res A-Pol 2009;43(5):548-559.

41. Wells NM, Yang Y. Neighborhood design and walking: A quasi-experimental longitudinal study. Am J Prev Med 2008;34(4):313-319.

42. Witten K, Blakely T, Bagheri N, et al. Neighborhood Built Environment and Transport and Leisure Physical Activity: Findings Using Objective Exposure and Outcome Measures in New Zealand. Environ Health Persp 2012;120(7):971-977.

43. Cao X, Mokhtarian P, Handy S. Do changes in neighborhood characteristics lead to changes in travel behavior? A structural equations modeling approach. Transportation 2007;34(5):535-556.

44. Newman P, Kenworthy J. Urban design to reduce automobile dependence. Opolis: An international journal of suburban and metropolitan studies 2006;2(1):35-52.

45. Brown BB, Werner CM. A new rail stop: Tracking moderate physical activity bouts and ridership. Am J Prev Med 2007;33(4):306-309.

46. Brown BB, Werner CM. Before and after a new light rail stop: Resident attitudes, travel behavior, and obesity. J Am Plan Assoc 2008;75(1):5-12.

47. MacDonald JM, Stokes RJ, Cohen DA, et al. The effect of light rail transit on body mass index and physical activity. Am J Prev Med 2010;39(2):105-112.

48. Begg SJ, Vos T, Barker B, et al. Burden of disease and injury in Australia in the new millennium: Measuring health loss from diseases, injuries and risk factors. Medical Journal of Australia 2008;188(1):36-40.