Are our suburbs walkable for the elderly?

(1)

- Focus on walking for transport

David Pemberton Msc Thesis

Department of Environmental and Biological Sciences Faculty of Science and Forestry

University of Eastern Finland UEF December 2021

(2)

Research has found that walkability supports physical and social health by increasing physical activity and social interaction, thus reducing obesity, high blood pressure, type two diabetes and social isolation among other things. Additionally, connective pedestrian network, rich proximate services and sufficient density support walkability including the vital social interaction. Safe and accessible pedestrian environment with open, interactive building frontages, greenery balancing the density and interesting public open spaces complement the structural neighborhood level walkability features, supported by frequent public transport. On the other hand, suburban sprawl has been shown to decrease walkability by increasing walking distances, decreasing local diversity and destination availability, and lowering public transport availability. And lack of walkability decreases the independent mobility and social interaction of suburban elderly over 65 years old, who generally share the same features affecting walkability and the benefits of it as any of us but in the course of decreasing mobility require shorter distances and thus require closer proximity of essential amenities than the younger ones. Walking for transport is in focus here, as it is an especially relevant form of walking for those not having active mobility habits or preferences of active transport, thus helping them to stay closer to the health sustaining minimum level of physical activity to avoid increasing dependency and premature hospitalisation.

Based on the literature study, the case study aims at assessing the walkability in three differing suburban neighborhoods of Helsinki Capital Region and its effects over the elderly population. Walkability of selected neighborhoods was assessed by analysing the connectivity of the street network, functional diversity and destinations availability, streetscape, open green space and public transport accessibility, comparing them also with Walk Score^TM results, considering the needs and abilities of the elderly inhabitants.

The findings of this study can be used to consider, how would improving walkabilty support the health and independent living of suburban elderly by supporting walking for transport. Aside of being an increasing group of inhabitants the elderly represent one of the many vulnerable groups that require special attention in making the city supportive for all. Walkability improvements generally make cities more accessible, equally open and livable for everybody.

The results of the thesis show that some of the analysed suburban neighborhoods had poor walkability and local accessibility as a part of it, which may also reflect lack of social interaction.

As there are twice as many elderly inhabitants in the suburban neighborhoods of Helsinki than in the central city, it can be considered a health equality issue, taking also the simultaneous home care crisis into account, and especially so in relation to the elderly with passive mobility habits.

The current work proposes more walking-oriented planning and interventions to enhance physical activity and social interaction of the suburban elderly by supporting walking for transport, more detailed pre-post-studies of individual suburbs, combined with the retrofitting of the car era suburbs to improve local accessibility with better diversity and service availability supporting physical activity and social interaction. Creating and maintaining proper parks and public open spaces within the neighborhood is also relevant within a walkable neighborhood, linking walking for pleasure with walking for transport.

(4)

Preface

The walkability research literature is expanding rapidly, tempting to track the new developments endlessly. Five years ago I chose to study a narrow section of walkability, considering the elderly inhabitants, after following and supporting the past everyday life of my mother during her last decades in Helsinki till her final 92’th year, and as a n´th generation non-driver-walker.

Although biking is one groundbreaker in the re-vitalisation of our cities after the long, sclerotic era of the private car and splintered urban structures, there are vague signs that also walkability

elements could be integrated in the urban planning palette - possibly also in the suburbs, where the need is even more acute than in the more obvious and trendy central city areas.

If I would start this study now, I would put much more emphasis in the social interaction elements of walkability, as many people are left alone and without sufficient care, relying increasingly on externalized virtual interaction without real physical and humane contacts.

This work has taken too much time, and I am obliged to the University of Eastern Finland (UEF) and my tutors PhD Pertti Pasanen and PhD Xavier Albacete, for patience in supporting it, as also my family and some close friends have done. Also the inspiring contents, tutors and course-mates of the Aalto University course on Urban and community planning 2013-2014

(Yhdyskuntasuunnittelun pitkä kurssi) were significant influencers for me.

With the environmental health studies background in UEF it was interesting to combine urban planning with environmental health perspectives, keeping also in mind the unscientific wisdom of Jane Jacobs, a woman of the streets and neighborhoods.

I hope that this study is useful, as we will all be elderly inhabitants sooner or later, and walkability improvements are also selfish aside of equality and fairness aspects of urban and community planning.

Helsinki, December 2021 David Pemberton

(5)

1. INTRODUCTION

The built urban environment has been shown to influence our physical activity (PA) through walking behavior and active transport (AT) significantly (Frank et al., 2005; Lee & Moudon, 2008;

Ross et al., 2015; Sallis et al., 2016). Simultaneously, AT has been shown to improve physical and social health (Oja et al., 1998; Hagströmer et al, 2007; Buehler et al, 2011; Laverty et al., 2013).

AT has also been found to be especially important for passive walkers, as they do not walk much for leisure or exercise, and as it lowers the treshold of physical activity and is connected with everyday life (Oja et al., 1998; Kruger et al., 2008; Buehler et al., 2011; Millward et al., 2013).

The World Health Organisation (WHO) suggests that the lack of mobility, often combined with low quality and excessive diet, is an important source of cardiovascular diseases, obesity and type 2 diabetes and that increasing peoples mobility is a relevant way to relieve this problem (Hamer &

Chida, 2008; WHO, 2010). The prevalence of Active Transport (AT) such as walking against passive transport such as driving a car has shown a direct correlation with obesity levels in Europe and the U.S.A. (Bassett et al., 2008). As leisure-time physical activity is more prevalent among the more educated and well off socio-economic groups in Europe, and as occupational physical activity does not any more balance it properly, there is a health inequality problem developing between different socio-economic groups (Beenackers et al., 2012). Simultaneously, many popular forms of physical activities such as fitness clubs, have become restricted by membership or a significant fee, thus further elevating the treshold of physical activity among the less advantaged socio-economic groups (Beenackers et al., 2012), present also among the elderly inhabitants after retirement, as acquired mobility habits have their roots deep.

Walkability can be defined as those qualities of the urban environment that either support or discourage walking for transport or leisure (Moudon et al., 2006; Leslie et al., 2007; Pivo & Fisher, 2011; U.S. Department of Health and Human Services/The Surgeon General, 2015). Therefore, walkability is one of the most important determinants of the environmental health effects of our built environment (Saelens et al., 2003; Sundqvist et al, 2011). Walkability also increases the social capital, social interaction and social cohesion within the neighbourhood (Leyden, 2003; Torres and others, 2013; Neutens et al., 2013; Zhu and others, 2014) and helps in creating a sense of community (Wood et al., 2010). This may be especially important for the rapidly increasing elderly population suffering from social isolation and loneliness (Giuliano et al., 2003). Correspondingly, urban sprawl has been found to encourage passive transport by private car and therefore e.g. obesity (McCann & Ewing, 2003; Lopez, 2004), together with social isolation (Rosenbloom, 2004).

Walkability also enhances the economic and socio-economic viability of the neighborhood and

(6)

local community, improving the customer base of local businesses, making them economically feasible (Song & Knaap, 2013; Gilderbloom et al., 2015). Walkability also provides better accessibility, consumer cost savings, public cost savings by reduced external costs, more efficient land use, community livability, economic development, and support for equity objectives (Boarnet et al, 2008; Litman, 2003). The increasing socio-economic viability of the neighborhoods may also increase property values and rental profits (Pivo & Fisher, 2011), inducing gentrification, thus transforming formerly affordable walkable areas out of reach for many inhabitants, such as the elderly inhabitants living in subsidised suburban rentals or cheaper apartments. Higher prices together with minimal availability of affordable rentals in walkable areas prohibits many working age inhabitants to move into more walkable areas while they still can and would be willing to, as their decreasing mobility would at some stage require. Also ”social housing” targeted for low- income inhabitants often tends to be located in distant car-dependent neighborhoods with discouraging transport walking conditions (Kim & Woo, 2016). Walking also relieves traffic congestion, improves urban air quality, decreases car noise and decreases greenhouse gas emissions as well as fossil fuel consumption (Mindali et al., 2004; Frank et al., 2010; Mindell et al., 2011).

Walkability or the lack of it may therefore improve or deteriorate physical or social health of the inhabitants through various pathways. Especially the connectivity of the street network, availability of diverse walking destinations and services within a walking distance, a high quality public transport service along with a sustainable base of inhabitants to support these functions have been shown to contribute in establishing and maintaining walkable communities, along with various streetscape quality issues that are related with attractiveness, interactivity of the streets & street amenities, history, public open spaces, high quality parks and other greenery.

Although most trips in the capital City of Helsinki are already made by walking (Helsinki, 2016), developing walkability is often restricted to the city centers, e.g. by creating specific walking areas in the core city, to attract tourists and shoppers. See related policy statements in e.g.

https://www.hel.fi/helsinki/en/maps-and-transport/cycling/walking/ Planning related with walking is missing relevant mobility data of walking and is also predominantly route-related, not considering the other essential elements and levels of walkability, although walking route planners like this are useful for active walkers: https://pk.reittiopas.fi/en/

The PhD dissertation of Jenni Kuoppa (2016) emphasized the idea of not limiting the walkability discussion to simple metrics, or to city centers, but to take also suburbs and the importance of human experience in account when discussing or developing walkability and urban planning. I have tried to take her words into account here.

The closest academic work overlapping this work untill now is the PhD dissertartion of Tiina E.

(7)

Laatikainen (2019), who studied environments for healthy and active ageing. Her work was primarily focused on walking for leisure, as this work concentrates on walking for transport, which have differing environmental correlates, as stated later. As her work studied the situation of people aged 55 to 75 years, and this work focuses on people over 65 years, this work deals with people whose mobility is already more limited than the people over 55, and whose need of proximity is much more evident. Her method of using online participatory mapping in collecting locational mobility data from the active elderly leisure walkers may also emphasize those respondents who are faring relatively well and can actively take care of themselves, while this work has been concentrated on passive risk groups in need of enhanced activity and mobility by daily utilitary activities. Her work is methodologically thorough. Several parts of it are referred to and commented in this work.

Most elderly inhabitants in Helsinki live in suburbs that were designed during the automobile era, compared with the central city, where the proportion of the elderly people is less than half of the suburban neighborhoods. Considering the benefits of walkable neighborhoods, it could be hypothesized that they support the physical and social health and independent living of the elderly inhabitants, also allowing them to age in place, i.e. in their home neighborhoods, as many of them hope (Rosenbloom, 2004). It is therefore sensible to investigate, how walkable are the current suburbs, to consider how could suburban walkability be improved to support elderly inhabitants, which is the intention of this work.

This thesis is organised as follows. The theoretical concepts and literature background, measuring and benefits of walkability are presented in chapter 2, referring also to the scientific criticism such as the self-selection hypothesis, revealing relevant information concerning e.g. passive walkers. The walkability of three selected neighborhoods is analysed in chapter 3 as a case study, in relation to the needs and abilities of the elderly inhabitants as an example of various vulnerable groups that benefit from walkability or suffer from the lack of it. After this, aside of the results, also further study needs and the possibilities to improve walkability of urban neighborhoods are discussed and proposed.

(8)

2. DEFINING, MEASURING AND EVALUATING WALKABILITY

2.1 What is walkability?

Although there is no universally accepted definition of walkability, it can be understood by comparing some current definitions of it. In U.S.A. where walkability has been studied most extensively, Leslie et al. (2007) defined walkability of a community as the extent to which characteristics of the built environment and land use may or may not be conducive to walking for leisure, exercise or recreation, to access services, or to travel to work. The highest federal authority of health issues in the U.S.A., The Surgeon General described (2015) a walkable community as one where it is safe and easy to walk in and pedestrian activity is encouraged. Pivo & Fisher (2011) defined walkability as the degree to which an area within walking distance of a property encourages walking for recreational or functional purposes. These definitions relate walkability to the

neighboring community with its various physical and social features. This is also reflected in the Surgeon General´s statement (2015) that communities can benefit when they implement strategies that make them more walkable, by making streets pedestrian friendly; building houses, shops, and other destinations close together (mixed land use); and increasing access to public transit. The differing definitions of walkability also indicate that walkability is a multi-level function in relation to urban form, starting from an individual street or a block with its microfeatures to the whole neighborhood, and extending to the city level with its transit and mobility system as a whole.

2.1.1 Purposes and types of walking – the importance of walking for transport

The Surgeon General (2015) states that people walk for many purposes, such as for transportation to get to school, work, a store, or the library or for leisure to have fun, socialize with friends or family, walk their dog, or to improve their health. And as walking is multipurpose, it provides many opportunities for people to incorporate physical activity into their lives. In most research, such as Moudon et al. (2006), general walking has been divided into two main categories: walking for transport (WFT) and walking for leisure (WFL). Walking for transport is in focus here, as it seems to have the strongest effect on health and immobility inequalities often related to passive walkers and the elderly.

2.1.2 The prevalence of walking

The differences in the prevalence of walking are huge both geographically and demographically.

(9)

Bassett et al. (2008) found that although walking is the most common leisure time physical activity both in the United States and Europe, walking and bicycling are far more common in European countries than in the US, Australia, and Canada. Europeans walked more than United States residents (382 versus 140 km per person per year) in 2000 (Figure 1).

Figure 1: Percentage of walking, biking and public transit trips in Europe, North America and Australia as a percentage of the total number of trips for 17 countries. (Bassett et al., 2008)

In the United States only 8% of trips were taken by walking, cycling and public transit, whereas in Latvia 67% of trips were made by walking, cycling and public transit. Also the total accumulated length of walking trips was significantly longer in the European countries.

Basset et al. stated that Europeans are more likely to walk for utilitarian purposes (WFT/AT) such as shopping, commuting to work, and school trips. Short trips in Europe are often taken by walking, while in the United States the automobile is used for 55% of about 0.5 km trips, 85% of 1.0 km trips, and more than 90% of longer trips. Buehler et al. (2011) found that between 2001/2002 and 2008/2009, the proportion of “any walking” was stable in the U.S. (18.5%) but increased in Germany from 36.5% to 42.3%. In 2008/2009, the proportion of “30 minutes of walking and cycling” in Germany was 21.2% and 7.8%, respectively, compared to 7.7% and 1.0% in the U.S.

Oja et al. (1998) found in a Finnish mid-size town Tampere that the prevalence of regular walking to work was 18% in the fall, 21% in the winter and 14% in summer. Most walked on a nearly daily

(10)

basis, and the modal shares were quite stable, as 87% of those walking or cycling in winter did so also in summer. Oja et al. say that walking and cycling were reported to be brisk (causing slight sweating and shortness of breath) by most physically active commuters. The median one-way distance of walking was 1 km in summer and fall and 2 km in winter, and the respective activity time was 12-18 minutes, meaning about two thirds of the general weekly PA recommendation.

The modal share of walking for transport in Helsinki, the capital of Finland, has increased within the last decade, while women have increased their share of walking and transit most, as the share of passive private car transit has remained male dominated (Figure 2).

Figure 2: The principal mode of transport 2012 -2016 among all respondents within one day.

Walking = “kävely”; as a driver of a private car = “henkilöautolla kuljettajana”; bus = “bussi”; bike

= “polkupyörä”; metro = “metro”; tram “ “raitiovaunu”; passenger in a private car = “henkilöautolla matkustajana”; train = “juna”; taxi = “taksilla; other mode = “muu kulkutapa”.

The amount and mode share of walking in Helsinki increased more than any other transport mode (see ”kävely” yearly columns). Both central city (”kantakaupunki”) and suburban (”esikaupunki”) inhabitants have increased their amount of walking trips per day equally much, but the significant difference between central and suburban walking frequencies has remained the same.

Who are walking and who are not?

There are relevant differences in the prevalence of walking among different demographic groups,

(11)

related to e.g. education, socioeconomic position, sex, age, abilities and personal preferences, differing in between different types of walking. The U.S. Surgeon General (2015) stated that walking in general is common among people who are physically active. Among people who walked, the average time spent walking was 13 minutes a day or about 90 minutes a week, providing slightly more than half of the time needed to meet the guideline of at least 150 minutes of aerobic physical activity each week. Surgeon General therefore says that about 60% of people who walked met the guideline by walking alone or in combination with other forms of physical activity (e.g. running or biking), compared with 30% of those who did not walk. Walking decreased with increasing age.

Oja et al. (1998) found in Finland that a considerably higher proportion of women than men commuted by these modes, which may also reflect mobility habits after retirement. The proportion of walkers increased with age, and in the 50-64 years age group the proportion was 23-31% in different seasons, greatest in winter.

Ball et al. (2007) found that those in the lowest educational group are less likely to participate in leisure-time and transport-related walking. Kruger et al. (2008) found that in the United States that leisure walking was most prevalent among respondents with higher incomes and education levels, whereas transportation walking increased in prevalence with education level but decreased with income level. They concluded that WFT and WFL have different demographic correlates.

Beenackers et al. (2012) found that people in higher socioeconomic positions (SEP) were more vigorously physically active (VLTPA) during leisure-time than people with lower SEP, thus showing a clear socioeconomic inequality pattern. Nearly all studies (96%) conducted in the Western European region reported that VLTPA was more prevalent among people with a higher SEP. Also in Scandinavia and in the Anglo-Saxon countries, the same associations dominated (both 88% positive). They found income to be positively associated with VLTPA more frequently among men (83%) than among women (67%). Buehler et al. (2011) noticed that there is much less

variation in active travel among socioeconomic groups in Germany than in the U.S. as German women, children, and seniors walk and cycle much more than their counterparts in the U.S. They concluded that the high prevalence of active travel in Germany shows that daily walking and cycling can help a large proportion of the population to meet recommended physical activity levels.

Lifelong mobility habits are challenging to transform while ageing, and the challenges of immobility and excessive nutrition may further increase the burden of inequality among the elderly.

Beenackers et al. (2012) say that transport-related PA and walking can be relevant in leveling inequalities in PA, as it seems to be less inhibited by SEP-factors, as it is already popular, and as interventions to change people´s free time mobility habits have been found to be challenging.

(12)

Also Turrell et al. (2013) found that as the odds of their study sample spending between 60 and 150 minutes walking for transport in the week prior to the survey was 80% higher in residents of the most disadvantaged neighborhoods (quintile 1) than residents of the least disadvantaged

neighborhoods (Q4), this makes them close to meeting the physical activity recommendations exclusively by active transport. The association between the least and most disadvantaged

neighborhoods and walking for transport strengthened linearly across all levels of walking. Turrell et al. also say that despite residents of disadvantaged neighborhoods are more likely to walk for transport, they walk less for recreation or exercise than residents of advantaged neighborhoods, and are more likely to smoke, have poorer diets and consume alcohol to harmful levels.

Ecological models explaining the effects on walking

The various influencing factors of walking behavior have been described in the so-called ecological models, which also contain moderate intensity excercise like walking for transport and leisure in the context of health. These moderate intensity activities may be crucial for those people who are not vigorously active such as elderly people.

The pioneers of ecological models Saelens, Sallis & Frank (2003) found that psychological and social factors explain much less variance in moderate intensity activity such as walking than in vigorous physical activity. They say that as most vigorous physical activities are directly and intentionally health-related, walking and cycling are often done for multiple purposes, making them more susceptible to environmental influence. Walking and cycling are often done for basic

transportation such as shopping, leisure, recreation, exercise or for occupational purposes. To tackle the multi-purpose, moderate intensity activities of walking and biking, Saelens et al. developed an ecological model to describe the environmental interactions influencing walking. Ecological hypothesis suggests that the combination of psychosocial and environmental variables will best explain physical activity (Figure 3).

Saelens et al. say that as psychosocial, demographic and environmental features influence walking, walking can, respectively, be modified by influencing these features, e.g. by interventions targeted at individuals or urban planning. They say that also environmental correlates of walking for transport (WFT) may differ from those correlating with walking for leisure/recreation (WFL). Lee (2004) stated that aside of the purpose of walking, environmental correlates of walking likely differ also by demographic features of the inhabitants. Such as age.

(13)

Figure 3: A proposed ecological model of neighborhood environment influence on walking and cycling. Double lines denote stronger relations; single lines denote weaker relations; dashed lines denote mediated relations. *Some examples of demographic variables are provided, but should not be considered comprehensive. ** Psychosocial correlates of physical activity would include, but are not limited to, such variables as self-efficacy, perceived benefits, perceived barriers, social support, and enjoyment of physical activity.

In line with Saelens et al., Ball et al. (2007) found that a combination of personal, social and environmental factors contribute to explaining lower levels of walking among women with low education, those in the lowest educational group being less likely to participate in both leisure-time and transport-related walking. They found that educational inequalities in leisure-time walking were influenced by e.g. neighborhood walking tracks, coastal proximity, friends’ social support, dog ownership, self-efficacy, enjoyment and intentions. Walking for transport was mostly explained by qualities of the neighborhood, coastal proximity, street connectivity and social support from family.

They say that also access to sidewalks or to a variety of destinations to walk to, urbanisation or land use mix and perceiving neighbors as being active may influence walking for transport among women with lower education. They say that walking may provide a key avenue for reducing

socioeconomic inequalities in physical activity within this group, and that urban planning strategies to build more walkable neighborhoods in disadvantaged areas, improving access to walking tracks and coastal localities, and considering greater connectivity among streets.

(14)

Also Benackers et al. (2012) say that inequalities in leisure-time PA and vigorous activity are often thought to be caused by either an educational effect on knowledge about the positive health

consequences of PA, or financial possibilities to engage in leisure-time PA, for example to afford memberships or admission rates for sports and PA facilities. As the patterns in inequalities in PA were similar for education and income, Beenackers et al. think that both may be important, although education was stronger associated with PA than income. Such factors as SEP differences in social or cultural capital or differences in physical environmental opportunities for PA, may also be

relevant determinants of SEP inequalities in PA. They say that external factors such as connectivity, density and the availability of public transport might be especially important for active transport PA.

2.1.3 Correlation of urban form with walking

While differences in personal walking habits and preferences are influenced by social and

socioeconomic features, also our physical enviroment effects the amount, type and other qualities of walking.

Environmental correlates of walking for transport

Bassett et al. (2008) claim that travel-related walking, bicycling and the use of public transit (i.e.

active transportation) is more common in Europe because of

1. Compact, dense cities with mixed land uses that generate short trips;

2. Restrictions on car use such as car-free zones, low speed limits, and prohibitions of through traffic;

3. Extensive, safe, and convenient facilities for walking and cycling;

4. Traffic calming of residential neighborhoods;

5. Coordination of public transit with walking and cycling to transit stations and stops, including bike parking, as well as safe sidewalks and bikeways;

6. Traffic regulations and policies that favor pedestrians and cyclists over motorists; and

7. High cost of owning and operating a car resulting from expensive driver licensing, high gasoline prices, and high taxes on car purchases.

They also say that active transportation tends to be more prevalent in older cities with mixed land use having residential, commercial, and civic buildings interspersed, as well as having sidewalks and well-developed public transit systems.

(15)

McCann & Ewing (2003) found that residents of sprawling places are likely to walk less, weigh more, and have high blood pressure than residents of compact areas. They say that the way our communities are built – the urban form – appears to have an impact on physical activity and health.

Ross & Dunning (1997) found that residents of densely populated areas reported the fewest vehicle trips, vehicle miles traveled, and vehicle miles per trip decreased dependency on single occupancy vehicles and increased use of transit. Less densely populated areas had more drivers per adult and more vehicles per adult. Also Distance to work and travel time to work decreased as the

percentage of retail trade in an area increased. At the home block group, increasing housing density was associated with greater transit availability and closer proximity to transit. Bicycle and walk trips increased along with increasing residential density, which was also associated with increasing employment density.

Sugiyama et al. (2012) found that better access to relevant neighborhood destinations (e.g., local stores, services, transit stops) can be conductive to adults' utilitarian walking. Adults' utilitarian walking was consistently associated with presence and proximity of retail and service destinations (in 80% of the studies) and functional aspects of routes (sidewalks and street connectivity) in 50%

of studies.

Craig et al. (2002) say that the positive association between the environment score and walking to work supports the development of integrated communities for housing, shops, workplaces, schools, and public spaces. Salon (2006) concluded that the built environment accounted for from one half to two thirds of the difference in walking levels associated with changes in population density in New York City. Zhou & Kockelman (2008) found that the built environment accounted for 58% to 90% of the total influence of residential location on VMT. Cao (2010) found that neighborhood type accounted for about 61% of the observed effect of the built environment on utilitarian walking frequency and 86% of the total effect on recreational walking frequency.

Urban correlates of walking for leisure

Sugiyama et al. (2002) found that 28.8% of respondents reported using public open spaces (POS) for physical activity and the likelihood of using POS increased with increasing levels of access, but those with very good access to large, attractive POS were 50% more likely to achieve high levels of walking. In 2012 they found that recreational walking was associated with presence, proximity, and

(16)

quality of recreational destinations (35% of studies) and route aesthetic (35% of studies), but unrelated to route safety and traffic in most studies. Sugiyama et al. (2010) also found that shorter distance to attractive open spaces was associated with doing recreational walking, but adults with larger attractive POS within 1.6 km of their home were more likely to walk 150 minutes or more in a week. They say that, provided there are no physical barriers affecting access such as major roads, distance is a major determinant of park use, with most users being drawn from within a 500-m radius of the park. As POS use is sensitive to distance, it is important to have them in proximity.

They say that well-designed public open spaces are an important component of the recreational mix providing opportunities for physical activity and social interaction. However, having good access to larger POS supports higher levels of walking. Larger parks tend to have more attributes that

provide satisfying experiences for the user. Respondents also described trees, water features, bird life, and size to provide more opportunities to “lose oneself”. The say, referring to Kaplan (1995), that natural environments are restorative when they give users a sense of being away from their usual setting, and a sense of fascination resulting from exposure to (for example) birdlife or natural beauty, balancing the often overburdened human information processing fatigue. Sugiyama et al.

said that those running or walking through urban parks perceive their experience as more

restorative, happier and relieving aggression, anxiety and depression, compared with those running in the streets. They also found that although larger POS generally have more attributes that make them attractive, even smaller POS with more attributes may attract more users. It may be possible to attract more users to POS by creating walking trails that link smaller local parks through better guided/directed routes, by developing shaded walking paths landscaped with trees and shrubs, and by creating interest by developing undulating areas around the flat POS, along with better

maintenance and care. Encouraging more use will lure more users, thus making the POS also safer.

They say that to increase walking, thoughtful design and redesign of POS is required to create attractive POS with facilities that encourage active use by multiple users (e.g., walkers, sports participants, picnickers). They say that despite the popularity of walking, a disproportionate amount of community POS is zoned for organised sports rather than for informal activities such as walking or jogging. For example playing fields are usually empty outside of the time used for organized sports, and typically well irrigated, flat lawns that are dull for walkers. They found that playing fields with public access dcan be redesigned for multiple users—organized sports participants, walkers, and passive recreational users — making better use of this important community resource.

They concluded that good access to attractive and large POS is associated with higher levels of walking. But simply providing proximate POS is not enough to increase walking, as more emphasis should be given to its size and attributes that make it attractive.

(17)

Although parks and greenery improve WFL, and sprawled neighborhoods may be green with vegetation, this in itself does not increase WFL if other elements of walkable neighborhoods are missing. McCann & Ewing (2003) found that people in more sprawling places spent less time walking in their leisure time than people living in compact locations. New Yorkers walked for exercise 79 minutes more each month than the people in the sprawled Geauga County in Ohio. Bassett et al. (2008) say that many of the same elements of the built environment that impact walking for utilitarian purposes can also explain the choice to walk for leisure, exercise or recreation, such as pedestrian infrastructure, street network connecting places of residence with parks and open space.

Young et al. (2017) say that the social dimension of environments have a significant influence on the health and well-being of the elderly. They studied public open space designs in three urban renewal districts with surveys, finding that social connection was considered the most important criteria among the three selected parks, which all performed unsatisfactorily when it came to enhancing social participation and inclusion. They say that to satisfy the specific needs of the elderly, future park designs could provide better space and opportunities for incorporating

participation in the general planning and design in their local parks and neighborhoods to enhance healthy aging.

2.1.4 Accessibility – function of proximity, connectivity and destination availability

Humpel et al. (2002) found that accessibility, opportunities, and aesthetic attributes had significant associations with physical activity, and that weather and safety showed weaker relationships.

Proximate places to go and flexible routes to get there are essential for walkability.

Handy (1993) found that the willingness of an individual to travel a certain distance depends on both the type and the amount of activity at the destination, finding also differing attributes for local and regional accessibility. Activity centers within a shorter distance nearby the community contribute to local accessibility and are generally small, such as grocery stores, supermarkets, drugstores and dry cleaners. Activities beyond that distance contribute to regional accessibility and are often larger – such as department stores, suburban shopping malls or downtown commercial areas, which offer a wide range of "comparison" goods, attracting customers from a wider area.

Local accessibility is associated with short and frequent "local" trips, and the choice of a particular destinations depends on the distance to that destination. Regional accessibility is associated with longer and less frequent trips, where distance is less important. A high level of accessibility implies

(18)

that more opportunities are located close by.

Handy found that communities with high levels of both local and regional accessibility had 40 percent less shopping travel in kilometers than communities with low levels of both. However, better local or regional accessibility was not associated with smaller trip frequency as residents in areas with poor accessibility do not compensate by taking fewer trips. Also, residents in areas with good accessibility do not take advantage of this fact by taking more trips, suggesting that there is an average or standard number of trips that residents make, regardless of the distance they must travel.

Correspondingly, the average shopping distances increased as levels of both types of accessibility decreased. When regional accessibility is low, better local accessibility is associated with less PkmT, as also high local accessibility (Table 1).

Table 1: Travel characteristics by superdistrict type (Handy et al., 1993).

Handy says that good local accessibility may reduce the frequency of regional trips, and good regional accessibility may reduce the frequency of local trips. Also, high local accessibility was most important when regional accessibility was low and vice versa. They therefore say that

providing high levels of local accessibility in new developments or improving local accessibility in existing developments may result in less automobile travel, but as better regional accessibility decreases the distance but not the number of trips, people will travel a certain amount indifferent of the distance, as some needs cannot be met locally. Enhancing both local and regional accessibility is therefore wise.

For the elderly inhabitants, local accessibility is more essential than regional, as their dependence on other people increases along with the decreasing mobility, making it increasingly challenging to rely on regional accessibility by using public transport or by asking for a ride from their friends or relatives (See Rosenbloom, 2004 and chapter 3.2). While Ross and Dunning (1997) found that

(19)

approximately 83% of all “trips” from a origin to a destination are short, for nonwork purposes, and occur relatively close to home, for the elderly people the proportion of these trips is even bigger, as work related trips are mostly excluded. Saelens et al. (2003) therefore say that the majority of nonwork trips are within walking or cycling distance and of interest to the physical activity, air quality, and transportation planning.

Ross and Dunning (1997) also say that factors that influence the choice to use motorized or non- motorised transport are based primarily on: (a) proximity (distance) and (b) connectivity (directness of travel), while also travel cost, environmental quality, including car noise, and aspects of

convenience and access such as parking availability are likely influential. Proximity relates to the distance between trip origins and destinations. They say that as proximity considers straight-line distances between land uses, connectivity characterizes the ease of moving between origins (e.g.

homes or workplaces) and destinations (e.g. stores and employment) within the existing street and sidewalk–pathway structure. Connectivity is therefore high if streets are laid out in a grid pattern and if there are no barriers such as freeways, walls or other physical obstacles. Choosing direct routes and several route options is therefore possible. Many interconnected streets in a regular grid pattern facilitate walking for transport (Saelens et al., 2003; Frank et al., 2003).

Saelens et al (2003) say that low connectivity is often found in modern suburbs, characterized by low density of intersections (e.g., long block size), barriers to direct travel (e.g., cul de sacs, private properties, fences), and few route choices. Also different land uses are purposefully separated, based on single use, resulting in a lack of land use mix. In older cities there are many residences above street-level shops, making it more convenient to walk to shops or work. They illustrate proximity and connectivity by comparing two contrasting types of neighborhood designs (Figure 4)

(20)

Figure 4: Two distinct community designs. From “The Traditional Neighborhood Development:

How Will Traffic Engineers Respond?” by F. Spielberg, 1989, ITE Journal, 59, 18. (Saelens et al., 2003)

The upper part of figure 4 represents a conventional recent suburban layout, forcing one to travel large distances to achieve only short straight-line distances (low connectivity). The lower part represents a traditional network layout providing better interconnected streets with more direct and shorter pathways between origins and destinations of the trip (high connectivity). The upper community also has lower density and poor land use mix, while the community below integrates different types of land uses within smaller areas. Saelens et al. say that these differences in proximity and connectivity are essential for a walker, as walkable distances are shorter than for passive motorised transport.

Leslie, Coffee, Frank & others (2007) argue that as the choices to use motorised or non-motorised active transport are based on proximity (distance) and connectivity (directions of travel), as distance is a highly relevant for walking in competing with other modes of travel. Increasing distances decrease the utility of walking compared to other modes of transport quickly (Frank, 2004). This effect can be supposed to be even stronger among the elderly inhabitants. Also O’Sullivan and Morrall (1996) say that distances of less than 0,8 km between residences, shops, employment, and to regional transit service are desirable, to keep walking competitive

Tenkanen et al. (2016) found time to be an essential determinant in access to healthy food and other health related services. They also found that the current spatial concept of “food desert”, referring to a geographical area where disadvantaged (e.g., low-income) residents lack access to affordable and healthy food, should be revised to include also different modes of transport. For example, while low-income population – like a significant part of the elderly population - tend to use public transport (PT) as their main transport mode for daily mobility, it is relevant to measure potential food deserts based on accessibility by PT and not only by private car. For elderly people with limited mobility, accessibility via PT has only limited benefits, as also for e.g. single parents, who cannot leave their small children unattended. Or those elderly who take care of their spouses in need of immediate presence. They depend on local accessibility.

2.1.5 Levels of walkability

Some key features of walkability are connected to the street and neighborhood levels, while e.g.

(21)

transit availability is also a vital feature of the city system level walkability, connecting streets and neighborhoods to the whole city and regional accessibility level. These interactive levels can complement or inhibit each others, defining the total walkability of a city.

2.1.5.1 The street

The most obvious perception of walkability is attached to the street level. Microscale attributes of streetscapes may alter pedestrian safety and comfort, and they are cheaper to modify than macro scale neighborhood features. Boarnet et al. (2011) found that measures of physical infrastructure, such as sidewalks, street characteristics, pedestrian crossings, and traffic signals, can be linked to physical activity and walking.

Sallis et al. (2015) studied micro level built environmental factors with a Brief Streetscape Audit Tool and found that streetlights, benches, curb cuts, the presence of a sidewalk, and buffers between streets and sidewalks were related to active transport in 3 of 4 age groups. They say that these attributes may improve the experience of pedestrians and bicyclists, or they could be indicators of a broader pattern of activity-supportive design features. For example, as sidewalks may be the most basic attribute for supporting pedestrian activity, also streets with curb cuts benefit pedestrians by e.g. improving the safety of intersections and slowing traffic speeds, improving access for older adults, people with disabilities, and parents with baby strollers. Aesthetic elements such as colorful buildings may enrich the walking experience. Benches may be an unexpected correlate of physical activity, but their presence signals consideration for pedestrians, and they are important for older adults who need to rest during walks. Streetlights support nighttime activity and increase feelings of security. Separating pedestrians from the higher velocity car and bike traffic with e.g. a planting strip improves pedestrian comfort and safety. They say that items with the most consistent associations with active transport appear to serve many functions, and that crossings and

intersections were particularly important for older adults, as all 3 crossings items were related to active transport. The total score was significantly related to walking for transport in all age groups, and seems to be the best indicator of activity supportiveness, as no single attribute dominated in encouraging active transport (Figure 5). MAPS-Mini total streetscape quality scores were linearly related to active transport in all age groups, suggesting that multiple environmental attributes

supportive of activity need to be provided or improved to have a large effect on walking. Sallis et al.

therefore say that making one improvement to streetscape environments would probably have a small effect, but making several improvements could have cumulatively large effects on walking and bicycling for transportation. They noticed that modest differences produced large percentage

(22)

changes for e.g. older adults, and larger improvements in MAPS-Mini scores might lead to an almost 250% increase in walking for transport. They say that streetscape environments appear to be less important for leisure physical activity than for active transportation.

Figure 5: Association of active transport with MAPS-Mini scores (percentage of total possible) ranked in quintiles from the poorest (lowest quintile) to the best (highest quintile) activity supportive microscale attributes of the built environment in the 2 older age groups. Quintiles for younger adults ranged from 14.6% to 59.2%; for older adults, 14.4% to 64.0%.(Sallis et al., 2015)

Also Cain et al. (2014) showed that microscale features were related to physical activity in 4 age groups. Destinations and land use, streetscape, street segment, and intersection variables were more related to transport walking/biking, while aesthetic variables were related to leisure PA. They found that the overall score was related to objective PA in children and older adults and concluded that microscale environment attributes are related to PA across the lifespan.

Ewing et al. (2016) said that street network design is widely thought to include street network characteristics of a neighborhood or district (e.g. Saelens et al., 20003). However, urban design also incorporates the streetscape features that affect the pedestrian experience. Ewing & Clemente (2013) studied 588 block faces in New York City to identify variables that explain pedestrian traffic volumes. They found that the proportion of windows on the street, the proportion of active street frontage, and the number of street furniture contribute to pedestrian counts, suggesting that urban design and streetscapes have a significant influence on pedestrian activity. Regarding “windows as a percentage of ground floor façade,” which relates to

(23)

transparency, they say that street vitality depends on its interaction with adjacent buildings, and a high transparency at the ground level can facilitate interaction. They found that having equal proportions of residential, retail, and office on a block face is less conductive to pedestrian activity than having a disproportionally high share of retail frontage. But they also found that context is important, particularly FAR and population density within about 400 meters of commercial streets, recommending higher values of these variables. They say that street furniture such as signs,

benches, parking meters, trash cans, newspaper boxes, bollards, streetlights, and such human scale items increase the complexity of the street, making it less dull. They also say that installing urban furniture does not guarantee improvements in pedestrian activity, if all other important factors, such as land use and public safety are forgotten. Also the percentage of active uses such as shops,

restaurants, public parks, and other generate pedestrian traffic. Inactive uses like blank walls, driveways, parking lots, vacant lots, abandoned buildings, and offices with no apparent activity degenerate pedestrian traffic.

According to Borst et al. (2008), elderly inhabitants found slopes and/or stairs, zebra crossings, trees along the route, front gardens, bus and tram stops, shops, business buildings, catering establishments, passing through parks or the city center, and traffic volume positively related to perceived attractiveness of a street. Litter on the street, high-rise buildings, and neighborhood density of dwellings were negatively related to perceived attractiveness. They say that especially perceived tidiness of the street, its scenic value and the presence of activity or other people along the street were most important for making the street attractive for walking.

The street level features are connected with the neighborhood level walkability, which contain them. As Ewing and Clemente (2013) noted, streetscape features are closely connected with the network design variable of the neighborhood level walkability, but also with the diversity variable, by e.g. active street frontages, and the density variable, through supporting the diversity and

survival of local services. A well designed major street is functionally approaching the concept of a neighborhood, as it contains both residential and other functions like shops, other everyday

amenities, public transport connections, social contacts within the community, local economy, security in both physical and social sense, greenery, places to rest and to mix with people, etc. This may be one of the reasons why the pioneering urban activist Jane Jacobs wrote so much about the street, e.g. in her cornerstone book The Death and Life of Great American Cities (1961).

2.1.5.2 The neighborhood

Moudon et al. (2006) say that characterizing neighborhood is especially important in relation to

(24)

health and physical activity, as neighborhoods are perceived by both residents and policy makers as meaningful congregations of people with common interests, making them essential for e.g.

interventions and planning. They say that defining the walkable neighborhood extends beyond pedestrian concerns, as the ability to walk in a neighborhood indicates, aside of a type of mobility and means of travel, also a type of sociability between neighbors, which, together, likely affect the physical, mental, and spiritual health of people in the community. Neighborhood evokes, according to Moudon et al., socio-physical homogeneity, a shared sense of place, connection, and access. It therefore has multiple cognitive, economic, geographic, behavioral, cultural, and temporal dimensions. When considering elderly inhabitants, both the social and physical dimensions are essential for their health and wellbeing.

Clarence Perry pioneered in 1929 by proposing the concept of “Neighborhood Unit,” based on children and families being able to walk safely from their homes to elementary schools and

community centers. Perry said that while the neighborhood community has no political structure, it frequently has greater unity and coherence than a village or a city. His neighborhood unit was bounded by arterial streets where apartment buildings, retail and services were located. Ideal

neighborhood size was 5000 to 6000 people, determined by the population necessary to yield 800 to 1200 elementary school age children, which was considered to be the most advantageous school size. Galster (2001) says that the theory of Perry derived from social reforms aimed at growing urban populations within congested and polluted cities, and from Ebenezer Howard’s Garden City theory (Howard, 1898), which modeled “new” British cities of 6000 to 30,000 people living within walkable distance to services and linked by rail transit. Perry, unlike Howard, wanted to improve normal people´s lives within the cities themselves, and not by moving them outside the

industrialised and crowded cities, as also the sub-urbanists of the late 19´th century did, creating the city of a private car.

The concept of neighborhood is widely used today, e.g. among the New Urbanists, advocating walkable neighborhoods. They use concepts like ”Traditional Neighborhood Design” (TND),

”Transit Oriented Development” (TOD), (Cervero, 2007); Pedestrian Pockets, Transit Villages, Urban Villages, etc to promote a planning and development based on the principles of how cities and towns had been built for several centuries: walkable blocks and streets, housing and shopping in close proximity, and accessible public spaces. They focus on human-scaled urban design,

articulating their principles in e.g. the Charter of the New Urbanism , to offer alternatives to the sprawling, single-use, low-density patterns typical of post-WWII development. They create tools to reform zoning and street design and develop underutilized building types like shopfront houses and courtyard units—that contribute to diverse neighborhoods. They advocate for villages, towns, and

(25)

cities consisting of neighborhoods designed around a five-minute walk from center to edge.

Diez Roux (2002) says that disadvantaged people suffer the consequences of bad quality physical environments most severely, as their financial constraints limit their mobility and access to daily activities and services, not being equally free to choose their neighborhood. These disadvantaged people also include numerous elderly inhabitants, often living in the cheaper suburban areas.

Threshold values for perceived walkable neighborhoods

Moudon et al. (2006) compared the perceptions and experience of those who walk sufficiently to meet the health recommendations and those who do not generally walk. They found that the

perception of neighborhood social environment (based on knowledge of neighbors and the presence of people who walked and biked in the neighborhood) was significant for the experience of

walkability. Also other physical neighborhood level walkability attributes were strongly and positively associated with walking sufficiently to meet health recommendations, such as denser neighborhoods with activities closer together, more sidewalks and smaller blocks.

Within the 1-km buffer, grocery stores, parks, and schools performed well in predicting their perceived presence. They found that threshold distances for eating/drinking establishments and grocery stores were 262 and 440 meters. Only grocery stores were strongly

associated with sufficient walking, while parks and schools were more insignificant. Distances between sufficient walkers’ homes and attractor land uses (restaurants,

grocery stores and agglomerations of grocery stores, restaurants, and retail stores) were less than one-third mile. Office complexes located within about 0,4 km from the respondents’ homes deterred walking, while more direct routes to schools supported walking. The threshold of transportation

infrastructure needed to support walking sufficiently for health was slightly more than 16mkilometers of sidewalks along major streets (excluding local streets) in the area within 1 km of home. Moudon et al. found that the distances between the respondent’s home and selected destinations were consistently different with the two extremes of walking behavior: walking sufficiently vs not walking in the neighborhood (Figure 6). They noticed that sufficient walkers were significantly more likely than non-walkers and moderate walkers to perceive the presence of grocery stores and parks in their

neighborhood. Residents were more likely to report the presence, rather than the absence of particular land uses when those land uses were closer to and

(26)

more abundant around their homes.

Figure 6: Mean airline distance from respondent’s home to the closest

destination and group of destinations. Distances presented in feet (0,34 m).

Moudon et al., (2006).

Moudon et al. found that respondents living in parcels with a net density higher than 21.7 residential units per 0,45 Ha were more likely to walk than those whose home were in a parcel with lower density. However, net residential densities measured within the 1-km buffer of the respondents’

homes were negatively associated with more walking when higher than 15.5 residential units per acre, converting to about 4800 residential units/km2 and circa 15 000 inhabitants/km2. Moudon et al. say that if the average unit size would be about 233 m²(including parking) with two stories high units, less than 50% of the net residential land would be covered by buildings or structures, not being difficult to achieve and leaving room for also parks, trees and other greenery. Also, sufficient walkers lived on street-blocks of less than 4 to 5 acre, meaning 54 apartments per hectar or 5360 apartments/km2. The street-block size threshold of less than 2 hectars is about the size of a medium-sized city blocks developed before World War II, found also in contemporary single- family housing areas. The mean values of block size within the 1-km buffer of the respondent were

(27)

slightly larger, at 2,75 Ha for sufficient walkers. This means that the block size would be less than 152,5 x 152,5 m (excluding streets). Large differences in block size measures between sufficient walkers and the total population indicated that reducing block size could enhance neighborhood walkability. They say that objectively derived threshold measures of walkable neighborhood attributes were fairly close to the typical conditions in their study area, indicating that the creation of supportive environments for walking could be achieved within current development practices in the urbanized areas of the country. They consider that the greatest challenge is locating food and daily retail uses within very short distances of residences 450 meters (1500 feet ) or less, as

threshold distances for eating/drinking establishments and grocery stores were 207 meters and 440 meters.

Moudon et al (2006) found that walkable neighborhoods seem to be centered on basic daily retail and food-related activities. Their findings did not mostly support planning theories of the early 20th century which placed schools, community centers, and open space at the heart of community. The new centers typically house activities related to discretionary rather than necessary spending, and exclude grocery or drug stores, which are believed to be associated with car travel, and located to serve not one, but several neighborhoods. The strength of association between walking and the objective presence of proximate grocery stores and non-fast food restaurants likely reflect the rise of smaller families, changing lifestyles, time constraints that lead to frequent eating out or purchase of take-out meals, and diets that favor deli over frozen foods.

Some destinations attract and some deter from walking

Significant neighborhood characteristics within 1 km of residents’ homes consisted of both

“attractor” and “deterrent” land uses. The former included grocery stores, markets, supermarkets and eating and drinking establishments. Thresholds for attractive walking environments included two or more agglomerations of grocery stores, non-fast food restaurants and retail stores, but no more than four individual grocery stores within 1 km of the respondents’ homes. The presence of large office complexes and too many educational facilities in a neighborhood deterred from walking.

Threshold parameters for environments discouraging walking included office complexes larger than 3,97 hectares (one and a half to two traditional blocks) within 3 km, and more than five schools within 1 km of respondents’ homes. Moudon et al. (2006) found that although offices as such did not seem to deter from walking, the size of office complexes was relevant, as office complexes grown bigger – causing trips that exceed the threshold of walkability. They therefore suggest that in walkable neighborhoods office and retail should be mixed to attenuate the impact of office uses at

(28)

the street level. The same applies to schools: they estimate that the threshold of fewer than five such uses within the 1-km neighborhood seems achievable. Food environments were surprisingly

relevant in defining walkable neighborhood in his study, unlike parks. They concluded that previous theoretical constructs of neighborhoods centered on recreation and educational uses should be revised.

Perceptions speak for multiple parallel neighborhoods

Moudon et al. (2006) say that while people generally perceived the presence of land uses in their neighborhood if there were more such uses and they were closer, only sufficient walkers seemed significantly more likely than non-walkers to perceive the presence of land uses in relation to their actual number and distance. Therefore, only relatively high levels of walking corresponded to enhanced perceptions of neighborhood. Walking may therefore help people to “know” their neighborhood or people who know their neighborhood walk more.

Moudon et al. also noticed that sufficient walkers who reported having no grocery store in their neighborhood, had an average of 2.46 such uses within 1-km airline distance of their home, but the figure dropped to a mean of 1.26 grocery stores in the 1-km network buffer, suggesting that network measures and/or shorter buffer distances need to be considered to define neighborhood.

Counts of parks and schools were even higher than counts of groceries for those reporting an absence of such uses in their neighborhood (Figure 7).

Figure 7: Perception of presence versus absence of destinations in the neighborhood, by non- walkers and sufficient walkers and by mean values of parcel vounts in 1-km airline buffer;

**significant at the 0,05 level (Moudon et al., 2006).

(29)

Moudon et al. (2006) say that the lack of awareness of the supposedly important neighborhood land uses may indicate that perception is related to the need for or the practice of an activity, as e.g.

childless people may not notice the schools in their neighborhood. They therefore confirmed that multiple neighborhoods exist, defined as those of sufficient walkers, non-walkers, and the general population, as well as e.g., weekend versus weekday, morning versus evening, summer versus winter neighborhoods.

Perceptions speak for smaller walkable neighborhoods

Moudon et al. (2006) say that routine activities in small geographic areas can yield health-

supportive levels of walking. They also say that the size of spatial units of analysis and intervention to capture neighborhood walkability should be adjusted to be much smaller than generally applied in research and planning, as the walkable neighborhood seems to be both geographically contained within a 1-km circle of the residents’ homes, considering both perception and objective measures.

They also say that e.g. census-based spatial units generally used in urban planning seem to be too coarse to capture walkable neighborhoods, except in dense urban areas, as e.g. the distances

produced by normal 1000 persons per census block groups are considerably longer than the 860 and 1445 meters threshold distances for sufficient walkers to the closest restaurant and grocery store, respectively, and even more exaggerated for passive walkers. They say that census geography is also based on streets and often “cuts through” main street intersections with retail activities, instead of using them as the focal points of the walkable areas.

Also Smith et al. (2010) say that although existing measures of perceptions of the environment associated with walking commonly rely on providing a definition of 'neighborhood', e.g. 1 mile area around the home, our understanding of adults' own geographical definitions of their

neighborhood area are vague. They also say that as much of our daily physical activity is

accumulated through the activities of daily living, such as walking for transport, work and domestic activity, rather than through active leisure pursuits, we need to understand how people's local 'neighborhood' environment impacts on their walking behavior to create 'walkable' environments.

They conducted 58 face-to-face interviews with participants randomly selected from 10 areas of Stoke-on-Trent, England. Participants were shown printed maps of their local area with road names and places of interest (e.g. shops, services, green space). When asked to draw their 'neighborhood' boundary on the map, their perceived neighborhood areas generally represented only 16 ± 20% of the commonly used 1 mile (1.6 km) Euclidean buffer area. The 1 km network buffer was at least twice the size of the neighborhood area in 64% of cases (Figure 8).

(30)

Figure 8: GIS-defined and perceived neighborhood boundaries and reported walking destinations (Smith et al., 2010). The home location point and the neighborhood area of a participant drawn in the same color.

The results of Smith et al. (2010) indicate that perceived neighborhoods tend to be significantly smaller than those often used in physical activity research, e.g. from 1 to 1,6 kilometers (1 mile) Euclidean buffer areas or even more. E.g. Colabianchi et al. (2007) reported that a 1,2 km (0.75 mile) network buffer was appropriate to define the walking neighborhood within "easy walking distance" for older female adolescents. The study sample of Smith were adults, and they consider that the average walkable neighborhood area in their sample would be smaller than1,2 km/0.75 miles and closer to estimated '5-minute walking' distances used elsewhere (e.g. 400 m/0.25 miles).

Operational definitions of neighborhood would therefore need to be smaller than those typically used. But as 42% of all walking destinations fell outside of areas that participants perceived as 'their neighborhood', the use of the larger (1,6 km/1 mile) Euclidean buffer, which captured 96% of destinations, could be acceptable. Visited green spaces (marked green spots in the Figure 8 map) were related to leisure (WFL) destinations, and were the outmost destinations here, as utilitarian WFT-destinations such as retail and other local services were more proximate. Also Laatikainen (2019) found the green area WFL destinations (parks etc) to be often further away from the actual neighbourhood area of the people from 55 to 75 years she studied.

Are our suburbs walkable for the elderly? - Focus on walking for transport