To walk or not to walk? Examining non-linear effects of streetscape greenery on walking propensity of older adults

https://doi.org/10.1016/j.jtrangeo.2021.103099Get rights and content

Abstract

Population aging is a conspicuous demographic trend shaping the world profoundly. Walking is a critical travel mode and physical activity for older adults. As such, there is a need to determine the factors influencing the walking behavior of older people in the era of population aging. Streetscape greenery is an easily perceived built-environment attribute and can promote walking behavior, but it has received insufficient attention. More importantly, the non-linear effects of streetscape greenery on the walking behavior of older adults have not been examined. We therefore use readily available Google Street View imagery and a fully convolutional neural network to evaluate human-scale, eye-level streetscape greenery. Using data from the Hong Kong Travel Characteristic Survey, we adopt a machine learning technique, namely random forest modeling, to scrutinize the non-linear effects of streetscape greenery on the walking propensity of older adults. The results show that streetscape greenery has a positive effect on walking propensity within a certain range, but outside the range, the positive association no longer holds. The non-linear associations of other built-environment attributes are also examined.

Introduction

The aging of the population, as a result of rising life expectancy, a decreasing fertility rate, and the aging of baby boomers (born between 1946 and 1964), is profoundly shaping a wide variety of regions around the world (e.g., Japan and Hong Kong). The absolute number and percentage of older people worldwide are ballooning. In 2019, there were 700 million older adults (defined as people aged ≥65) worldwide. By 2050, this number will nearly double to 1.3 billion. Additionally, the percentage of older people was 9.1% in 2019 and is predicted to climb to 15.9% in 2050 (United Nations, 2019). Similarly, Hong Kong is facing the issue of population aging and dramatic and unprecedented changes in its demographic profile. Second only to Japan (28.4% in 2020, topping the global list) in the Asian context, Hong Kong has a high percentage of older people (18.2% in 2020). The percentage of its older people is predicted to grow to 34.7% in 2050 (United Nations, 2019).

Older people's mobility has been widely documented to be closely related to their participation in activities, social integration, physical and psychological health, and life satisfaction (Alsnih and Hensher, 2003; Chen et al., 2021, Metz, 2000). Their mobility is a determinant of quality of life and subjective wellbeing (Banister and Bowling, 2004). For older adults, an insufficient level of mobility leads to difficulty in participating/engaging in social activities and interacting with the wide community and, consequently, low morale, depression, and loneliness (Wong et al., 2018).

Walking is a form of active mobility (low-intensity physical activity) for residents and a travel mode with considerable economic, environmental, social, safety, and health advantages (Frank et al., 2007; Heath et al., 2012; Sælensminde, 2004; Wong et al., 2021). Walking has thus received much attention from governments, non-government organizations, and researchers. It has been advocated and promoted universally. As Cervero and Kockelman (1997) state, a paramount transportation objective is diverting travel demand from motorized to active modes such as walking. Moreover, walking is a critical travel mode for older adults, who generally have limited access to cars (Hu et al., 2013; Yang, 2018), in many locations (e.g., Chinese cities) (Cheng et al., 2019b, Liu et al., 2021b; van Wee and Handy, 2016). This is particularly true for Hong Kong. Owing to its mixed land use, pedestrian-friendly urban design, and high walkability, Hong Kongers have a predilection for and habit of walking, as evidenced by Hong Kongers taking the highest number of walking steps all over the world (Althoff et al., 2017).

Urban greenery may facilitate physical activity (due in part to humankind's biophilia nature) and alter people's walking behavior (Lu et al., 2018). Traditionally, its evaluation heavily relies on in-person assessments, field observations, aerial photography, and remote sensing imagery. However, these methods have limitations, such as high labor intensity, a restriction to small areas, a restriction to a bird's eye (overhead) view, and the inability to represent the human scale (Kang et al., 2020). Fortunately, owing to the availability of street-view imagery data and the rapid development of urban analytics, streetscape greenery has garnered increasing attention from researchers (especially those in the public health field) in recent years. We can now efficiently and accurately estimate streetscape greenery from street-view imagery (Kang et al., 2020).

Existing literature has extensively emphasized the effects of socio-demographic characteristics and built-environment attributes (e.g., population density and street connectivity) on the travel behavior of older adults. However, few studies have evaluated the role of streetscape greenery (Yang et al., 2020; Yang et al., 2019; Zang et al., 2020). More importantly, most, though not all, of these studies have assumed a pre-determined (often linear) relationship between travel behavior and its contributory factors. Nevertheless, the connection between the built environment and the travel behavior of older people may be non-linear (Ding et al., 2019, Liu et al., 2021a; van Wee and Handy, 2016). This non-linearity can be explained by the peer effect (or collective socialization) and travel utility (Cheng et al., 2020a; Galster, 2018; Mokhtarian and Salomon, 2001). In recent years, pioneering research has focused on the issue of non-linearity (Cheng et al., 2020a; Ding et al., 2018b; Ding et al., 2018c; Zhang et al., 2020). However, to our best knowledge, only one study (Cheng et al., 2020a) concentrates on the non-linearity of the connection between the built environment and the travel behavior of older people. No studies on the effect of streetscape greenery on travel behavior have considered the issue of non-linearity. As ignoring a genuinely non-linear effect may result in misunderstandings and erroneous practical implications (Cheng et al., 2020a), delving deeper into the non-linearity issue is crucial.

To address the above issues, we use data from the Hong Kong Travel Characteristic Survey (TCS) 2011 and Google Street View (GSV) imagery (detailed in Section 3) to assess the walking propensity (or propensity of walk trip-making, propensity to walk) of older people and streetscape greenery, respectively. A random forest model is adopted to evaluate the non-linear effects of streetscape greenery on walking propensity. A binary logistic regression model is generated to make a comparison with the random forest model. Notably, socio-demographic and built-environment attributes (measured by TCS 2011 data or geo-data) are controlled. The contributions of this study are (1) the examination of the connection between streetscape greenery and older people's walking propensity; (2) the scrutinization of the non-linear and threshold effects of streetscape greenery on travel behavior for the first time; and (3) the assessment of the non-linear and threshold effects of built-environment attributes on older people's walking behavior.

The remainder of this paper is organized as follows. The ensuing section reviews the literature on the contributory factors of the walking behavior of older adults. Sections 3 introduces the data and methodology of random forest modeling. Section 4 presents the results of random forest modeling and compares them with the outcomes of logit modeling. Section 5 winds up the paper, discusses theoretical and practical implications, and summarizes research limitations.

Section snippets

Literature review

A voluminous body of the existing literature has confirmed the effects of the built environment on the walking behavior of older people. We conducted electronic searches in the database of Web of Science and retrieved pertinent articles published between 2006 and 2020. We first screened the papers by title and then by abstracts. We considered their relevance to this study and manually chose papers for detailed reviews.

Table 1 summarizes selected studies on the link between the built environment

TCS 2011 data

TCS 2011 is a wide-ranging and comprehensive travel survey that was conducted by the Transport Department of the Hong Kong government during September 2011 and January 2012. The TCS is conducted roughly every ten years. It includes three main surveys, namely household interview survey (for collecting the trip information of Hong Kong residents), stated preference survey (for identifying the contributory factors of transport mode choices for users), and hotel/guesthouse tourists survey (for

Relative importance of predictor variables

Table 3 presents the random forest results and shows the relative importance of predictor variables. Fig. 3 graphically describes the relative importance. The variable of primary interest, namely Streetscape greenery, has the second-highest relative importance (12.82%), exceeded only by Age (16.65%). This outcome supports the critical role of streetscape greenery in determining the walking propensity of older adults. Furthermore, the other “5Ds” built-environment variables closely follow

Conclusions and discussions

The aging of the population has become a popular phenomenon in many cities worldwide (Jing et al., 2021), such as Hong Kong. Given that walking is a prevalent travel mode for older adults and has multiple health and wellbeing benefits, determining the correlates of the walking behavior of older adults is of paramount importance. Furthermore, as an easily perceived built-environment attribute, streetscape greenery has seldom been investigated in the travel behavior literature of older adults but

Acknowledgments

This research was supported by the Fundamental Research Funds for the Central Universities of China (No. 2682021CX097), Education and Scientific Research Grant of Sichuan Province (No. SCJG20A110), and Talent Cultivation Quality and Teaching Reform Project of Ideological and Political Theory Course of Chengdu University of Technology in 2020 (20800-2020SZ009). The authors are grateful to the reviewers for their constructive comments.

References (66)

  • C. Ding et al.

    Applying gradient boosting decision trees to examine non-linear effects of the built environment on driving distance in Oslo

    Transp. Res. A Policy Pract.

    (2018)
  • C. Ding et al.

    Non-linear effects of the built environment on automobile-involved pedestrian crash frequency: a machine learning approach

    Accid. Anal. Prev.

    (2018)
  • C. Ding et al.

    How does the station-area built environment influence Metrorail ridership? Using gradient boosting decision trees to identify non-linear thresholds

    J. Transp. Geogr.

    (2019)
  • L.D. Frank et al.

    Stepping towards causation: do built environments or neighborhood and travel preferences explain physical activity, driving, and obesity?

    Soc. Sci. Med.

    (2007)
  • F. Ghani et al.

    Do differences in built environments explain age differences in transport walking across neighbourhoods?

    J. Transp. Health

    (2018)
  • G.W. Heath et al.

    Evidence-based intervention in physical activity: lessons from around the world

    Lancet

    (2012)
  • Y. Hou

    Polycentric urban form and non-work travel in Singapore: a focus on seniors

    Transp. Res. Part D: Transp. Environ.

    (2019)
  • Y. Hou et al.

    The associations between older adults’ daily travel pattern and objective and perceived built environment: a study of three neighbourhoods in Singapore

    Transp. Policy

    (2020)
  • X. Hu et al.

    Understanding the travel behavior of elderly people in the developing country: a case study of Changchun, China

    Procedia Soc. Behav. Sci.

    (2013)
  • P. Koh et al.

    Mobility of the elderly in densely populated neighbourhoods in Singapore

    Sustain. Cities Soc.

    (2015)
  • K.M. Leung et al.

    Impact of physical and social environments on the walking behaviour of Hong Kong’s older adults

    J. Transp. Health

    (2018)
  • J. Liu et al.

    Non-linear associations between built environment and active travel for working and shopping: An extreme gradient boosting approach

    J. Transp. Geogr.

    (2021)
  • Z. Liu et al.

    Heterogeneity in physical activity participation of older adults: a latent class analysis

    J. Transp. Geogr.

    (2021)
  • Y. Lu et al.

    The effect of street-level greenery on walking behavior: evidence from Hong Kong

    Soc. Sci. Med.

    (2018)
  • Y. Lu et al.

    A dose–response effect between built environment characteristics and transport walking for youths

    J. Transp. Health

    (2019)
  • D.H. Metz

    Mobility of older people and their quality of life

    Transp. Policy

    (2000)
  • P.L. Mokhtarian et al.

    How derived is the demand for travel? Some conceptual and measurement considerations

    Transp. Res. A Policy Pract.

    (2001)
  • M. Moniruzzaman et al.

    An investigation of the attributes of walkable environments from the perspective of seniors in Montreal

    J. Transp. Geogr.

    (2016)
  • M. Moniruzzaman et al.

    Mode use and trip length of seniors in Montreal

    J. Transp. Geogr.

    (2013)
  • K. Sælensminde

    Cost–benefit analyses of walking and cycling track networks taking into account insecurity, health effects and external costs of motorized traffic

    Transp. Res. A Policy Pract.

    (2004)
  • T. Tao et al.

    Exploring the non-linear associations between spatial attributes and walking distance to transit

    J. Transp. Geogr.

    (2020)
  • J. Van Cauwenberg et al.

    Relationship between the physical environment and physical activity in older adults: a systematic review

    Health Place

    (2011)
  • B. van Wee et al.

    Key research themes on urban space, scale, and sustainable urban mobility

    Int. J. Sustain. Transp.

    (2016)
  • Cited by (220)

    View all citing articles on Scopus
    View full text