Elsevier

Social Science & Medicine

Volume 60, Issue 4, February 2005, Pages 763-775
Social Science & Medicine

Neighbourhood inequalities in physical inactivity: the role of neighbourhood attractiveness, proximity to local facilities and safety in the Netherlands

https://doi.org/10.1016/j.socscimed.2004.06.013Get rights and content

Abstract

We investigated the association between the neighbourhood socioeconomic environment and physical inactivity, and explored the contribution of neighbourhood characteristics to this association. Data were analysed of 20–69 years old participants of the Dutch GLOBE study who lived in 78 neighbourhoods of Eindhoven (n=8.767). The neighbourhood socioeconomic environment was assessed from aggregated self-reported information of participants’ education and occupation level, and employment status. Aspects of physical inactivity investigated were based on the time spent on (a) walking and cycling to shops or work, (b) walking, cycling and gardening in leisure time, and (c) participation in sports activities. Characteristics of neighbourhoods included the proximity to food shops, general physical design of neighbourhoods, quality of green facilities, noise pollution from traffic and required police attention as evaluated by municipal services (professionals) responsible for these characteristics.

Compared to those living in the most advantaged neighbourhoods, residents living in the quartile of socio-economically most disadvantaged neighbourhoods were more likely to walk or cycle to shops or work, but less likely to walk, cycle or garden in leisure time and less likely to participate in sports activities (adjusted for age, sex and individual educational level). Neighbourhood inequalities in walking or cycling to shops or work were not mediated by specific neighbourhood characteristics included in our analyses. The increased probability of almost never walking, cycling and gardening in leisure time in the most disadvantaged neighbourhoods was partly mediated by a poorer general physical design in these neighbourhoods. Similarly, the increased probability of almost never participating in sports activities in the most disadvantaged neighbourhoods was partly mediated by larger amounts of required police attention.

The direction of neighbourhood inequalities differs for aspects of physical inactivity. Neighbourhood characteristics are related to physical inactivity and contribute to neighbourhood socioeconomic inequalities in physical inactivity.

Introduction

Physical inactivity is a major burden of public health, due to its established association with coronary heart diseases (CHD) and other chronic diseases (Berlin & Colditz, 1990; Helmrich, Ragland, Leung, & Paffenbarger, 1991; Powell, Thompson, Caspersen, & Kendrick, 1987; Wannamethee & Shaper, 1999; Wannamethee & Shaper, 2001) and high prevalence rates. For example, it has been estimated that 40–55% of Dutch adults do not meet the Dutch recommendations of being moderately active for at least five days per week (Hildebrandt, Urlings, Proper, Ooijendijk, & Stiggelbout, 1999; Kemper, Ooijendijk, & Stiggelbout, 2000; Schuit, Feskens, & Seidell, 1999). Thus, developing effective interventions aimed at reducing these rates are important for public health.

Knowledge about appropriate target populations and determinants of physical inactivity within these populations is essential for the development of such interventions. It becomes increasingly clear that different determinant levels can be distinguished, such as the family level, school or work level and neighbourhood level. Therefore, understanding the neighbourhood context in which physical activity takes place is an important contribution in achieving this (Owen, Leslie, Salmon, & Fotheringham, 2000).

In recent years, several studies have shown that living in socio-economically disadvantaged neighbourhoods is associated with increased incidence (Diez Roux et al., 2001) and mortality of CHD (Hart, Ecob, & Davey-Smith, 1997) and its major risk factors (Diez-Roux et al., 1999; Kleinschmidt, Hills, & Elliott, 1995; van Lenthe & Mackenbach, 2002). Yet, descriptive evidence of an association between the neighbourhood socio-economic environment and physical inactivity is limited, and explanations for such associations are still unanswered. Thus, the present study aims to investigate the existence of neighbourhood inequalities in physical inactivity and to explore the potential mediating roles of neighbourhood characteristics in the association between the neighbourhood socioeconomic environment and physical inactivity.

Studies investigating the association between the neighbourhood socio-economic environment and physical activity generally show a decreased likelihood of being physically active for those living in the socio-economically most disadvantaged compared to the most advantaged neighbourhoods. Ellaway and Macintyre (1996), as well as Ecob and Macintyre (2000), found more unfavourable exercise patterns in socio-economically disadvantaged neighbourhoods compared to those that are more advantaged. Yen and Kaplan (1998), using data from the Alameda County Study, found that subjects living in a poverty area showed a greater decline in physical activity as compared to those living in a non-poverty area (Yen & Kaplan, 1998). In an Australian study those living in the most disadvantaged areas showed a higher probability of not being sufficiently active compared to those living in the most advantaged areas (Giles-Corti & Donovan, 2002b). However, using specific aspects of physical activity as the outcomes, the results are less unequivocal. While the probability of walking for recreation was higher in those living in the most advantaged compared to the most disadvantaged areas, the latter study also reported a higher probability of walking for transport for those living in the most disadvantaged areas compared by those living in the most advantaged areas (Giles-Corti & Donovan, 2002b). Further, Ross (2000) reported that residents living in areas with a high percentage of households living in poverty were more likely to walk than those in the more wealthy areas (Ross, 2000), while associations were in the opposite direction if neighbourhoods were characterised by the percentage of college educated subjects. Thus, the results justify further research, preferably in large, population-based studies using specific outcomes of physical activity.

Moreover, an intriguing, but yet unanswered, question concerns the mechanisms responsible for the association between the neighbourhood socioeconomic environment and physical inactivity. The above-described patterns of physical inactivity by neighbourhood socioeconomic environment suggest that neighbourhood characteristics may be involved in the mechanisms. An important problem for carrying out empirical research however, is the paucity of conceptual models, linking the neighbourhood socioeconomic environment to neighbourhood characteristics and subsequently to physical activity. In order to develop a model from which hypotheses could be derived, we first identified neighbourhood characteristics that are potentially related to physical activity. Subsequently, we explored if and how these characteristics could be related to the neighbourhood socioeconomic environment.

The idea that neighbourhood characteristics may be related to physical activity seems to be recognised already by the Ottawa Charter in 1986 (Ottawa Charter for Health Promotion, 1986). More recently, it has been suggested that the neighbourhood environment may either facilitate or discourage the range of physical activity behaviour (Sallis & Owen, 1996). For such reasons, several theoretical models now distinguish ‘environmental factors’ as a separate category of behavioural determinants of physical activity (Cohen, Scribner, & Farley, 2000; Emmons, 2000; Flay & Petraitis, 1994; McLeroy, Bibeau, Steckler, & Glanz, 1988). However, empirical evidence of associations between area characteristics and physical activity remains scarce. Most available evidence for a relation between neighbourhood characteristics and physical activity suggests a role for the proximity of facilities. An inverse association is reported between the objectively measured distance to pay facilities for exercise and self-reported exercise (Sallis et al., 1990) and between the distance to a main rail-trail and its use (Troped et al., 2001). Further, some studies, though not all (Wilcox, Castro, King, Houseman, & Brownson, 2000) provide evidence for a positive association between the perceived proximity to physical activity related facilities (such as walking trails, and indoors gyms) and other neighbourhood facilities (such as shops and parks) and the likelihood of being physically active (Ball, Bauman, Leslie, & Owen, 2001; Booth, Owen, Bauman, Clavisi, & Leslie, 2000; Brownson, Baker, Housemann, Brennan, & Bacak, 2001; Parks, Housemann, & Brownson (2002a), Giles-Corti & Donovan (2002b)). Recently, a dose response relationship has been suggested between the number of places to exercise and the likelihood to meet recommendations for physical activity (Parks, Housemann, & Brownson, 2003).

Some studies have reported a positive association between the aesthetics of the neighbourhood and physical activity (Ball et al., 2001; Brownson et al., 2001; Corti, Donovan, & Holman, 1997; Takano, Nakamura, & Watanabe, 2002). There is also evidence that increased perceived safety is related to a decreased likelihood of being physically active (Booth et al., 2000; Craig, Brownson, Cragg, & Dunn, 2002; Eyler et al., 1998; King, Castro, Eyler, Sallis, & Brownson, 2000; Weinstein, Feigley, Pullen, Mann, & Redman, 1999).

Interestingly, the above-mentioned factors are now accommodated in a recently developed framework for assessing environmental determinants of recreational and transport related walking and cycling (Pikora, Giles-Corti, Bull, Jamrozik, & Donovan, 2003).

Support for an association between the neighbourhood socioeconomic environment and the above-mentioned neighbourhood characteristics can be derived from a theory in which the association between the neighbourhood socioeconomic environment and related neighbourhood characteristics is described in terms of a spiral of decay (Skogan, 1990). In summary, this theory describes that a decrease in the neighbourhood socioeconomic environment (in itself determined by a number of factors) triggers several processes. It is suggested that such a deterioration starts a process of selective migration, in which more wealthy persons move out, and those with more limited financial resources move in. As most of the individuals entering the disadvantaged neighbourhoods did not actively choose these neighbourhoods to live in, they may not intend to stay in the neighbourhood for a long period. Thus, they may be less inclined to take care of the maintenance of their houses and they may feel less responsible for other physical characteristics of the neighbourhood, resulting in structural degradation. In these neighbourhoods, it may be less inviting to investigate commercially and existing neighbourhood facilities may disappear. Additionally, crime may become more prevalent in such neighbourhoods. The above-described relation between the neighbourhood socioeconomic environment and structural degradation (and therefore the attractiveness), the accessibility and proximity to neighbourhood facilities and safety has been demonstrated in some studies (Macintyre & Ellaway, 1998; Macintyre, Maciver, & Sooman, 1993; Skogan, 1990; Sooman & Macintyre, 1995). On the contrary however, one study reported that objectively measured spatial access to recreational facilities was significantly higher for those living in socio-economically disadvantaged compared to advantaged areas (Giles-Corti & Donovan, 2002b).

Clearly, a number of other neighbourhood characteristics may be related to the neighbourhood socioeconomic environment. However, taking into account that mediating variables in our study should be characteristics that are related to these outcomes and to the socioeconomic environment, three groups of relevant characteristics were identified: (a) the general level of attractiveness, (b) the proximity to neighbourhood facilities and (c) the safety in the neighbourhoods.

In epidemiological studies, physical activity is often assessed from information about the frequency and duration of activities of different intensities, ranging from (recreational) walking and cycling to sports participation. We argue that for investigating neighbourhood inequalities in physical activity, aspects of the total amount of physical activity should be used as separate outcomes, as the direction of the associations of these outcomes with the neighbourhood socioeconomic environment seems to differ (Giles-Corti & Donovan, 2002b; Ross, 2000). Moreover, not all neighbourhood characteristics may be equally relevant for various physical activities. For example, where the proximity of sports facilities may be related to sports participation, it is less clear why it should influence walking and cycling to shops. For the latter aspect of physical activity, the proximity of shops may be much more relevant. In our study, information was gathered of three major aspects of physical activity: the amount of time spent on (a) walking and cycling to shops or work, (b) walking, cycling and gardening in leisure-time and (c) participation in sports activities.

In this study we explored the role of neighbourhood characteristics for neighbourhood inequalities in physical inactivity. Specifically, we tested the following hypotheses:

  • 1.

    (a) Residents of socio-economically disadvantaged neighbourhoods are less likely to spend time on walking and cycling to shops or work compared to residents in more socio-economically advantaged neighbourhoods.

    (b) These inequalities are mediated by (a) a lower general level of attractiveness, (b) a poorer proximity to shops, and (c) lower levels of safety in more disadvantaged neighbourhoods.

  • 2.

    (a) Residents of socio-economically disadvantaged neighbourhoods are less likely to spend time on walking and cycling and gardening in leisure time compared to residents in more socio-economically advantaged neighbourhoods.

    (b) These inequalities are mediated by (a) a lower general level of attractiveness, and (b) lower levels of safety in more disadvantaged neighbourhoods.

  • 3.

    (a) Residents of socio-economically disadvantaged neighbourhoods are less likely to participate in sports activities compared to residents in more socio-economically advantaged neighbourhoods.

    (b) These inequalities are mediated by (a) a poorer proximity to sports facilities, and (b) lower levels of safety in the more disadvantaged neighbourhoods.

Section snippets

Study population

The GLOBE-study is a prospective cohort study aimed at explaining socioeconomic inequalities in health. The objectives and design of the study are presented in detail elsewhere (Mackenbach, Mheen, & Stronks, 1994). For the baseline measurement, a random sample of 27,070 non-institutionalised subjects between 15 and 75 years of age and living in the south-eastern Netherlands were invited to complete a postal questionnaire, including questions on socioeconomic position (SEP) and physical

Results

Mean ages of participants across quartiles of the neighbourhood socioeconomic environment ranged from 45.2 years (SD 13.5) to 47.9 years (SD 14.6). Each quartile consisted of slightly more women compared to men, percentages ranging 50.1–51.9% (not tabulated). More then 38% of our study population reported almost never walking or cycling to shops or work. Compared to individuals living in the most advantaged neighbourhoods, those in the most disadvantaged neighbourhoods showed a significantly

Discussion

In this study, we found inequalities in physical inactivity by the neighbourhood socioeconomic environment, but the direction of the inequalities varied for the different categories of physical inactivity. Moreover, for the outcomes for which inequalities were found in the hypothesised direction (i.e. higher probability of inactivity in the more socioeconomic disadvantaged neighbourhoods) mediating neighbourhood characteristics attenuated these associations. Our study is amongst the first to

Acknowledgements

The GLOBE study is carried out by the Department of Public Health of the Erasmus Medical Centre Rotterdam, in collaboration with the Public Health Services of the city of Eindhoven and region South-East Brabant. The authors are indebted to Ilse Oonk and Roel Faber for maintaining the database, the participants for their willingness to participate in the study, and Katrina Giskes for comments on the pre-final draft of this paper. The present study is supported by grants of the Ministry of Public

References (53)

  • T. Pikora et al.

    Developing a framework for assessment of the environmental determinants of walking and cycling

    Social Science & Medicine

    (2003)
  • C.E. Ross

    Walking, exercising an smokingDoes neighbourhood matter?

    Social Science & Medicine

    (2000)
  • A. Sooman et al.

    Health and perception of the local environment in socially contrasting neighbourhoods in Glasgow

    Health & Place

    (1995)
  • P.J. Troped et al.

    Associations between self-reported and objective physical environmental factors and use of a community rail-trail

    Preventive Medicine

    (2001)
  • A.B.V. Berkel-van Schaik et al.

    Naar een standaardoperationalisatie van sociaal-economische status voor epidemiologisch en sociaal-medisch onderzoek [towards a standard operationalisation of socioeconomic status for epidemiological and sociomedical research]

    (1990)
  • J.A. Berlin et al.

    A meta-analysis of physical activity in the prevention of coronary heart disease

    American Journal of Epidemiology

    (1990)
  • R.C. Brownson et al.

    Environmental and policy determinants of physical activity in the United States

    American Journal of Public Health

    (2001)
  • B. Corti et al.

    Factors influencing the use of physical activity facilitiesResults from qualitative research

    Health Promotion Journal of Australia

    (1997)
  • G. Davey Smith et al.

    Area based measures of social and economic circumstancesCause specific mortality patterns depend on the choice of the index

    Journal of Epidemiology and Community Health

    (2001)
  • A.V. Diez Roux et al.

    Neighborhood of residence and incidence of coronary heart disease

    New England Journal of Medicine

    (2001)
  • A.V. Diez-Roux et al.

    Neighbourhood differences in dietThe atherosclerosis risk in communities (ARIC) study

    Journal of Epidemiology and Community Health

    (1999)
  • A. Ellaway et al.

    Does where you live predict health related behaviours?A case study in Glasgow

    Health Bulletin (Edinb)

    (1996)
  • K.M. Emmons

    Health behavior in a social context

  • R. Erikson et al.

    Intergenerational class mobility and the convergence thesisEngland, France and Sweden

    British Journal of Sociology

    (1983,[object Object])
  • A.A. Eyler et al.

    Physical activity and minority womenA qualitative study

    Health Education Behaviour

    (1998)
  • B.R. Flay et al.

    The theory of triadic influenceA new theory of health behaviour with implications for preventive interventions

    Advances in Medical Sociology

    (1994)
  • Cited by (274)

    • Investigating what makes people walk or cycle using a socio-ecological approach in seven European cities

      2021, Transportation Research Part F: Traffic Psychology and Behaviour
    • Changes in neighborhood-level socioeconomic disadvantage and older Americans’ cognitive functioning

      2021, Health and Place
      Citation Excerpt :

      Sixthly, in addition to fears concerning exiting one's home (Foster and Giles-Corti 2008; Fried and Barron 2005; Piro et al., 2006), various consequences of worsening neighborhood socioeconomic circumstances might prevent regular engagement in physical activity. Indeed, residents of more socioeconomically disadvantaged neighborhoods were less likely to engage in physical activity partly through lower neighborhood aesthetic appeal, more constrained community social networks (Kamphuis et al., 2009), poorer neighborhood structural design, and more required police surveillance, monitoring, and control (van Lenthe et al., 2005). All these concomitants of worsening neighborhood socioeconomic circumstances discourage physical activity within one's community.

    View all citing articles on Scopus
    View full text