Journal of Traffic and Logistics Engineering Vol. 3, No. 1, June 2015
Built Environment Towards Active Transport and
Health Improvements
Tayebeh A. Saghapour
Civil Engineering, RMIT University, Melbourne, Australia
Email: [email protected]
Sara B. Moridpour and Russell C. Thompson
RMIT University, Civil Engineering/University of Melbourne, Civil Engineering, Melbourne, Australia
Email: [email protected]; [email protected]
Abstract—Arrangement and distribution of land use
activities in an urban district is known as one of the
potential ways influencing the patterns of urban mobilities.
Providing facilities and services for residents in the vicinity
of their living areas can minimize the needs for long
distance travels and increase the chance of having active
travels. Hence, transport policy makers, urban planners
and health practitioners have been recently turning towards
promoting physical activities by environmental solutions.
Melbourne as a metropolitan area has encountered the risk
of increasing car dependency and subsequent health
problems among residents. This paper aims to investigate
the association between the built environment features and
active travel behaviour. Socioeconomic factors have been
also considered for their important influence on
transportation mode choice. Data used in this study has
been provided from Victorian Integrated Survey of Travel
and Activity (VISTA). Statistical analysis including MultiNominal Logistic (MNL) regression used to explore the
relationship between variables. Promoting healthy
behaviours and active commuting can be achieved by
understanding current levels of physical activities as well as
transport network and accessibility to various uses.
hand, many diseases such as diabetes, depression,
hypertension, overweight and obesity have been reported
as the negative health impacts of physical inactivity [1],
[2]. Furthermore, physical inactivity has also found to be
one of the major contributors to chronic disease,
disability, and premature mortality [2]. The link between
the built environment and travel behavior has received
considerable research attention during recent decades [3],
[4]. In following sections we review some studies
conducted on both positive and negative aspects of health
outcomes of active travels and sedentary trips. This study
aimed to investigate the impacts of built environmental
factors along with socioeconomic characteristics on
active travels (walking and cycling). Fig. 1 demonstrates
different mode choice (a) and weight status (b) among
age categories. As shown in Fig. 1a by increasing age,
use of motorized vehicle have risen. The peak of using
motorized vehicle in the figure (a) was shown between
35 to 64 groups, while people in these age categories
were considered as overweight or obese (see Fig. 1b).
Health issue and its hidden expenditures which burden
government annually make it a considerable subject for
researchers and policy makers. In this study we focus on
some factors may affect individuals’ mode choice.
In the following, we first present a literature review in
Section 2; Section 3 describes the dataset; Section 4
presents methods and analysis results; section 5 includes
discussion and conclusions.

Index Terms—built environment, active transport, health,
physical activities, multinomial logit model
I.
INTRODUCTION
Over the past decades, the link between the built
environment and travel behaviour has received
considerable research attention. After World War II,
Australian cities have become dependent on motorized
transport by huge increasing in car ownership. This
happening along with sprawling land use planning and
automobile-oriented development have encouraged
people to have less physical activities and spend more
times traveling by automobiles. The negative
consequences of this kind of passive behaviours have
been highlighted in metropolitan areas such as
Melbourne.
Numerous studies have been conducted regarding the
increasing the likelihood of obesity and cardio metabolic
disease as the cause of physical inactivity. On the other

II.
Active commuting which is mainly defined as nonmotorized trips such as walking and cycling has been
categorized as potential forms of physical activities.
There are few studies investigating the specific impacts
of active travels on health. Strong et al., [5] reviewed the
health consequences of physical activities and developed
recommendations based on evidence for physical
activities among youth. Their findings included
affirmative changes in cardiovascular disease and
metabolic syndrome and significant increase in bone and
muscle strength along with less adiposity. In a health
study by Samimi et al., [6], key results indicated that
obesity will be decreased by 0.4% if auto use decrease by
1%. According to Bassett et al., [7], active transportation
Manuscript received December 12, 2014; revised March 30, 2015.
©2015 Journal of Traffic and Logistics Engineering
doi: 10.12720/jtle.3.1.72-76
BACKGROUND
72
Journal of Traffic and Logistics Engineering Vol. 3, No. 1, June 2015
Most studies do not consider different type of physical
activities. In a meta-analysis review by Hamer and Chida
[8], a negative association has been found between active
commuting and cardiovascular risk. The results showed
that by increasing walking and cycling cardiovascular
risk decreased by 11% especially among women. Similar
outcomes have also observed from another meta-analysis
review by Oguma et al., [9]. They concluded that women
would be less in danger of cardiovascular disease if they
walked 1 hour per week or possibly less. Zheng et al.
[10], in a meta-analysis review found a statistically
significant association between coronary heart disease
risk and walking. Their results reported that 8 METh/week increasing in walking resulted in 19% decline in
the coronary heart disease risk. Other studies have also
found significant relationship between physical activities
and different kinds of cancers. For instance, in a
systematic review (29 case–control studies) by
Manninkhof et al., [10], an inverse association was found
between physical activities and breast cancer risk.
Furthermore, increasing the level of physical activities
causes modest reductions in colon cancer among men and
women [11].
On the other hand, walking has been considered as
“therapeutic mobilities” and reported as a treatment
which makes significant improvements in mood
especially level of anger, depression, fatigue and
confusion [12]-[14]. In general, investment on promoting
active travels can be considered as potential resources to
fund additional health care or decrease government
overall expenditure on health. In a research by Jarrett et
al., [15] in England and Wales, it was estimated that
costs of National Health Service (NHS) could be reduced
up to UK£17 billion (in 2010 prices), if the amount of
active travels (walking and cycling) increased. Generally,
both positive and negative impacts of active transport and
passive travel, respectively investigated widely in recent
years. The majority of studies concluded travel behavior
can affect health directly and indirectly. Although there is
not enough studies concentrating on different built
environment factors along with socio economic
characteristics on the duration of time people spend on
motorized vehicles.
can be used as the explanation for international
differences in obesity rates. They conducted a study to
examine the relationship between active transportations
and obesity rates in different countries including Europe,
North America, and Australia. The key findings of the
study showed that European countries with more active
travels (walking and cycling) were less likely to be
overweight or obese. In contrast in North America or
Australia people were more dependent automobile and
had the highest rate of obesity.
Private Motorized Vehicle
Walk/Bike
Public Transport
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Age Categories
a. Percentage of using different modes, VISTA (2009)
underweight/Normal
Overweight
Obese
70%
60%
50%
40%
30%
20%
10%
0%
Age Categories
b. Weight status according to body mas index
Figure 1. Weight status and mode use in different age categories
TABLE I.
NUMBER AND PERCENTAGE OF TRIPS MADE IN DIFFERENT MODES AND TIME DURATIONS
Transport Mode
Private Motorized
Vehicle
Walking
Cycling
Public Transport
Total
<15 min
15-30
min
30-45
min
45-60
min
60-120
min
120-180
min
>180
min
Total
Count
40837
15725
4845
2195
1386
132
34
65154
% of Total
43.5%
16.8%
5.2%
2.3%
1.5%
.1%
.0%
69.4%
Count
17221
2128
268
64
29
1
0
19711
% of Total
18.4%
2.3%
.3%
.1%
.0%
.0%
0.0%
21.0%
Count
726
397
173
84
54
3
0
1437
% of Total
.8%
.4%
.2%
.1%
.1%
.0%
0.0%
1.5%
Count
3364
2433
1145
453
131
5
5
7536
% of Total
3.6%
2.6%
1.2%
0.5%
0.1%
0.0%
0.0%
8.0%
Count
% of Total
62148
66.2%
20683
22.0%
6431
6.9%
2796
3.0%
1600
1.7%
141
.2%
39
.0%
93838
100.0%
©2015 Journal of Traffic and Logistics Engineering
73
Journal of Traffic and Logistics Engineering Vol. 3, No. 1, June 2015
III.
The methods used in this study included ANOVA as
descriptive analysis to compare the means between
different modes and Multinomial Regression Model has
been also applied to examine the probability of using
different modes in four main groups of private motorized
vehicles, walking, cycling and public transport
considering both environmental and socioeconomic
factors. Table II shows the descriptive analysis for trips
(considering trip stages) made by different modes. As
shown average travel time spent as vehicle driver was
19.24 minutes. School bus and train and has the highest
mean by respectively 28.08 and 27.03 minutes.
To analyze data Multi-nominal Regression (MNL) has
been applied to examine the likelihood of using each
mode
considering
socioeconomics
and
built
environments factors. Table III shows the results of the
analysis.
Odds Ratio in Table IV shows the likelihood that
individuals may go for specific modes in comparison to
the reference category (private motorized vehicle). OR>1
and OR<1 show the more and less likelihood for a mode
to be chosen by people.
DATASET
Data used has been obtained from Victorian Integrated
Survey of Travel and Activity (VISTA) to analyse
whether different built environment features can affect
time duration individuals spent in motorized vehicle in
their daily trips (trips are considered as trip segments for
example one journey with a specific purpose may contain
several trips). Data contained the details information of
93838 trips of 80332 numbers of journeys made by
22184 individuals in Melbourne region. Table I shows
the result of cross-tabulation results containing number
and percentage of trips made by different modes in
different time intervals. As shown in Table I among short
trips (less than 15 minutes) 43.5% was made by
motorized vehicle and 19.2% was walking/cycling trips.
As the time increases active travels decrease while there
was still 1.5% motorized trips.
A. Variables and Indicators
Variables were mainly considered in three categories
including socio-economic variables (age, gender,
employment status and household income), built
environment factors (land use mix entropy index and
population density) and travel variables (travel time and
number of vehicles). Land use mix entropy index was
calculated using Eq. (1) formula. the values vary from 0
to 1, while 1 indicates a perfect balance among different
type of land uses and 0 shows the homogeneity (Eq. 1)
[16].
𝐸𝐼𝑖 = −(∑𝐽𝑗=1(𝑃𝑗 . 𝑙𝑛𝑃𝑗 ))/𝑙𝑛𝐽
V.
In this paper we examined different socioeconomic
and environmental factor influencing transport mode
choice. Socioeconomic variables included age, gender,
and employment status, number of vehicles in household
and household income; while built environment variables
were population density and land use mix entropy. Both
groups of variables had significant, generally, affect
mode choice. As shown in Table IV, travel time, number
of vehicles, income, age, population density, and land use
mix entropy significantly influence walking and cycling
trips. However, for trips made by public transport gender
and income level have not had any significant impacts.
The key findings of the analysis can be summarized as
follows:
 Areas with more mixed development of uses
individuals are more likely to go for walking
(OR=2.274, p<.001), cycling (OR=4.084, p<.001)
and public transport (OR=1.637, p<.001)
 As time increases the likelihood for going for
walking decreases (OR=.931, p<.001)
 As the age increases people are less likely to go
for walking, cycling and public transport
(respectively OR=.980, p<.001, OR=.940, p<.001
and OR=.942, p<.001)
 As income level increases people are more likely
to go for cycling (OR=1.211, p<.001)
 Males are more likely to go for cycling
(OR=2.189, p<.001)
 Individuals with full time careers are more likely
to go for public transport (OR=1.168, p<.001),
while people with part time or casual jobs are less
likely to go for walking (respectively, OR=.709,
p<.001 and OR=.835, p<.001).
Hence, it can be concluded that in neighborhoods with
higher diversity in uses and more population density, the
(1)
where EIi indicates the entropy index within a buffer i
(SA2)1. Pj represents the proportion of a type of land use
j, and J is the number of land use categories. Six different
Land use categories including residential, commercial,
Industrial, transport and infrastructure, community
services and sport and recreation centers, have been
chosen to calculate LU mix index entropy. These
categories are defined from ten main uses categories
defined by Australian Valuation Property Classification
Codes [17].
Population Density calculated as the number of people
per square kilometer (large scale non-residential area
excluded). Age, gender household income level and
employment status have been considered as most
influential factors affecting individuals transport mode
choice. Travel indicators included time (minutes) spent
for each trip and motorized vehicle ownership. Contents
of Table II show the descriptive analysis of variables.
Average age was 38 years and household income was
mostly between third and fourth categories varied from
1100 to 2499 ($/week). Average travel time spent in trip
stages was 22.3 minutes.
IV.
METHOD AND RESULTS
1 According to Australia Bureau of Statistics (ABS), the ABS
structure of Melbourne region contains, 53074 Mesh Blocks, 9514
Statistical Area Level 1 (SA1), 277 Statistical Area Level 2 (SA2), 42
Statistical Area Level 3 (SA3) and 12 Statistical Area Level 4 (SA4).
©2015 Journal of Traffic and Logistics Engineering
DISCUSSION AND CONCLUSIONS
74
Journal of Traffic and Logistics Engineering Vol. 3, No. 1, June 2015
level of physical activities can be higher. However,
cultural and social factors as Cairns et al., [18] argued
may affect travel behaviors, as well. Future studies may
consider those factors as well as neighborhood design
factors to have a better understanding of travel behaviors.
TABLE II. DESCRIPTIVE STATISTICS OF SAMPLE
Age
Gender
Mean
Std. Deviation
Skewness
Kurtosis
38.41
1.88
0.01
-1.07
1.53
0.50
-0.12
-1.98
*
2.50
1.39
0.05
-1.83
**
3.63
1.33
-0.57
-0.90
0.33
0.20
0.52
-0.61
3958.64
2935.63
1.85
7.70
Travel Time
22.29
16.68
3.24
20.11
Number of Vehicles
1.94
1.02
1.03
2.84
Employment Status
Household Income
Land use mix Entropy
Population Density (p/km2)
* Employment status categories: full time, part time, casual, not-working and other.
** Household income ($/week): less than 650, 650-1099, 1100-1649, 1650-2499 and more than 2500)
TABLE III. ONE WAY ANOVA FOR MODES AND TRAVEL TIME
95% Confidence Interval for Mean
Lower Bound
Upper Bound
N
Mean
Std. Deviation
Std. Error
Vehicle Driver
43339
19.24
17.57
.084
19.07
19.40
21645
170
17.52
25.32*
18.28
22.56
.12
1.73
17.28
21.91
17.76
28.74
Walking
Vehicle Passenger
Motorcycle
Walking
19711
9.92
8.36
.06
9.80
10.03
cycling
cycling
1437
23.19
19.52
.52
22.18
24.20
Public Transport
Taxi
307
23.81
15.84
.90
22.03
25.59
Train
3462
27.03
16.33
.28
26.49
27.58
Tram
School Bus
Public Bus
1861
322
1489
15.16
28.08
19.20
10.63
18.54
16.91
.25
1.03
.44
14.67
26.05
18.34
15.64
30.11
20.06
93838
17.21
16.68
.054
17.10
17.31
Private Motorized
Vehicles
Total
*Not reliable because of several observation having trips out of Melbourne area
TABLE IV. MNL MODEL FOR TRIPS
Transport Modes a
Walking
Intercept
Travel Time
Number of Vehicles
Income
Age
Population Density
Land use Mix Entropy
Male b
Employment Status c
Full Time
Part Time
Casual
Cycling
Intercept
Travel Time
Number of Vehicles
Income
Age
Population Density
Land use Mix Entropy
©2015 Journal of Traffic and Logistics Engineering
B
-.122
-.072
-.467
.024
-.020
.000
.822
-.006
Wald
7.852
4197.814
1649.773
9.933
71.587
1888.365
315.807
.113
Sig.
.005
.000
.000
.002
.000
.000
.000
.737
-.035
-.344
-.181
-5.070
.013
-.505
.191
-.062
.000
1.407
2.634
131.576
11.461
1320.430
120.012
202.866
61.143
55.989
478.821
102.852
.105
.000
.001
.000
.000
.000
.000
.000
.000
.000
75
Odds
Ratio
95% Confidence Interval for OR
Lower Bound
Upper Bound
.931
.627
1.024
.980
1.000
2.274
.994
.929
.613
1.009
.976
1.000
2.077
.959
.933
.641
1.040
.985
1.000
2.490
1.030
.966
.709
.835
.926
.668
.752
1.007
.752
.927
1.013
.604
1.211
.940
1.000
4.084
1.011
.563
1.154
.925
1.000
3.112
1.015
.647
1.270
.956
1.000
5.360
Journal of Traffic and Logistics Engineering Vol. 3, No. 1, June 2015
Male b
.783
Employment Status c
Full Time
.131
Part Time
.030
Casual
-.100
Public Transport Intercept
-1.519
Travel Time
.012
Number of Vehicles
-.599
Income
.013
Age
-.060
Population Density
.000
Land use Mix Entropy
.493
Male b
-.069
Employment Status c
Full Time
.155
Part Time
-.417
Casual
-.149
a. The reference category is: Private Motorized Vehicle.
b. The reference category is: Female.
c. The reference category is: Unemployed.
176.512
.000
2.189
1.950
2.457
3.734
.082
.286
639.469
426.680
1235.487
1.345
284.636
738.621
55.630
7.101
.053
.774
.593
.000
.000
.000
.246
.000
.000
.000
.008
1.141
1.030
.905
.998
.841
.628
1.303
1.262
1.305
1.013
.550
1.013
.942
1.000
1.637
.933
1.011
.532
.991
.936
1.000
1.438
.887
1.014
.568
1.034
.949
1.000
1.863
.982
26.052
75.258
3.528
.000
.000
.060
1.168
.659
.862
1.100
.600
.738
1.239
.724
1.006
[13] J. Pretty, M. Griffin, J. Peacock, R. Hine, M. Sellens, and N.
South N, “A countryside for health and wellbeing: The physical
and mental health benefits of green exercise–executive summary,”
Countryside Recreation Network, 2005.
[14] R. Robertson, A. Robertson, R. Jepson, and M. Maxwell,
“Walking for depression or depressive symptoms: A systematic
review and meta-analysis,” Mental Health and Physical Activity,
vol. 5, no. 1, pp. 66-75 2012.
[15] J. Jarrett, J. Woodcock, U. K. Griffiths, Z. Chalabi, P. Edwards, I.
Roberts, et al, “Effect of increasing active travel in urban England
and Wales on costs to the National Health Service,” The Lancet,
vol. 379, no. 9832, pp. 2198-205, 2012.
[16] J. S. Lee, J. Nam, and S. S. Lee, “Built environment impacts on
individual mode choice: An empirical study of the HoustonGalveston metropolitan area,” International Journal of
Sustainable Transportation, vol. 8, no. 6, pp. 447-70, 2014.
[17] (ACLUMP) ACLUaMP, Guidelines for Land Use Mapping in
Australia, Agriculture Do, Editor. 2011.
[18] S. Cairns, C. Harmer, J. Hopkin, and S. Skippon, “Sociological
perspectives on travel and mobilities: A review,” Transportation
Research Part A: Policy and Practice, vol. 63, pp. 107-117, 2014.
REFERENCES
A. De Nazelle, et al, “Improving health through policies that
promote active travel: A review of evidence to support integrated
health impact assessment,” Environment International, vol. 37, no.
4, pp.766-77, 2011.
[2] W. L. Haskell, S. N. Blair, and J. O. Hill, “Physical activity:
Health outcomes and importance for public health policy,”
Preventive Medicine, vol. 49, no. 4, pp. 280-282, 2009.
[3] S. L. Handy, M. G. Boarnet, R. Ewing, and R. E. Killingsworth,
“How the built environment affects physical activity: Views from
urban planning,” American Journal of Preventive Medicine, vol.
23, no. 2, pp. 64-73, 2002.
[4] D. Wang, Y. Chai Y, and F. Li, “Built environment diversities and
activity–travel behaviour variations in Beijing, China,” Journal of
Transport Geography, vol. 19, no. 6, pp. 1173-86, 2011.
[5] W. B. Strong, R. M. Malina, C. J. Blimkie, S. R. Daniels, R. K.
Dishman, and B. Gutin B, et al. “Evidence based physical activity
for school-age youth,” The Journal of Pediatrics, vol. 146, no. 6,
pp. 732-737, 2005.
[6] A. Samimi, A. K. Mohammadian, and S. Madanizadeh, “Effects
of transportation and built environment on general health and
obesity,” Transportation Research Part D: Transport and
Environment, vol. 14, no. 1, pp. 67-71, 2009.
[7] D. R. Bassett Jr, J. Pucher, R. Buehler, D. L. Thompson, and S. E.
Crouter, “Walking, cycling, and obesity rates in Europe, North
America, and Australia,” Journal of Physical Activity and Health,
vol. 5, no. 6, pp. 795-814, 2008.
[8] M. Hamer and Y. Chida, “Active commuting and cardiovascular
risk: A meta-analytic review,” Preventive Medicine, vol. 46, no. 1,
pp. 9-13, 2008.
[9] Y. Oguma and T. Shinoda-Tagawa, “Physical activity decreases
cardiovascular disease risk in women: Review and meta-analysis,”
American Journal of Preventive Medicine, vol. 26, no. 5, pp. 40718, 2004.
[10] H. Zheng, N. Orsini, J. Amin, A. Wolk, and F. Ehrlich,
“Quantifying the dose-response of walking in reducing coronary
heart disease risk: Meta-analysis,” European Journal of
Epidemiology, vol. 24, no. 4, pp. 181-92, 2009.
[11] E. M. Monninkhof, et al, “Physical activity and breast cancer: A
systematic review,” Epidemiology, vol. 18, no. 1, pp. 137-57,
2007.
[12] A. C. Gatrell, “Therapeutic mobilities: Walking and ‘steps’ to
wellbeing and health,” Health & Place, vol. 22, pp. 98-106, 2013.
[1]
©2015 Journal of Traffic and Logistics Engineering
Tayebeh A. Saghapour is a PhD student in RMIT University the
school of Civil, Environment and Chemical Engineering. She is doing
her research on the impacts of transport mode choice and its impacts on
general health. She holds Bachelor of Economics and Master’s degree
in Urban and Regional Planning from Shiraz University, Iran.
Sara B. Moridpour holds a Bachelor of Civil Engineering and
Master’s degree in Transportation Planning and Engineering from
Sharif University of Technology, Iran. She also received her PhD
degree from Monash University. She has 9 years of work and research
experience in the field of traffic and transport. Her main research
interests include on driving behaviour modelling and analysis, micro
simulation, transport network modeling and optimization, transport
infrastructure asset management. She has been lecturer in the School of
Civil, Environmental and Chemical Engineering, RMIT University,
from 2010.
Russell C. Thompson is an associate professor at University of
Melbourne. His main research areas include Urban Transportation and
Logistics in terms of Health, Safety, and Security, infrastructure
engineering.
76