Gender and age differences in recreational and transport walking: The contribution of the neighbourhood social and built environments

Populations are growing and ageing globally, and they concentrate in urban centres, placing greater pressure on city infrastructure and resources. The burden of non-communicable diseases partly reflects the increasing inactivity trends in populations, possibly from living in environments primarily designed for private motor vehicle transportation rather than active transportation.

Walking as regular physical activity (PA) is an important behaviour to facilitate active living in ageing communities. Compared to men and younger adults, women and older adults are less physically active, and favour walking rather than more vigorous PA. Research underpinned by social-ecological frameworks indicates that certain social and built neighbourhood features influence the walking patterns of residents. Therefore, the identification of specific environmental features that facilitate walking in populations, particularly in those demographic groups predisposed to inactivity, can inform the social and physical planning or retrofitting of urban forms that might potentially reduce the gender and age disparities in overall PA participation.

Multilevel neighbourhood-based studies to date mostly reported the average neighbourhood effects of gender and age differences in recreational and transport walking, implicitly assuming that neighbourhood environments influence the walking patterns of men and women, and younger and older persons, similarly. However, this might not be a true reflection of what is actually happening. Through three multilevel cross-sectional studies underpinned by a social-ecological framework, this thesis explored the contribution of the neighbourhood built and social environments to explaining gender and age differences in recreational and transport walking. The thesis contains seven chapters outlined below.

Chapter 1 provides an overview of the context, purpose and structure of this thesis, including a conceptual framework underpinning three studies within this thesis. This chapter highlights the important role that neighbourhood designs have in facilitating walking, and notes the benefits of using a social-ecological framework to inform multilevel interventions (targeted both at individuals and communities) which have the maximum potential to increase population walking levels.

Chapter 2 presents the historical perspective of the influence of neighbourhoods on health behaviours such as walking, as well as an overview of the relevant policy and research frameworks supporting the investigation of the environmental correlates and determinants of recreational and transport walking. This chapter also presents an overview on the health benefits, recommendations and measurement of PA, with a focus on walking patterns, noting that it usually occurs within neighbourhoods, and that recreational walking has different environmental correlates than transport walking. Furthermore, the literature indicates that walking patterns vary by gender and age, and these differences are discussed within a social-ecological framework. A critical appraisal of the social and physical neighbourhood features influencing walking is also presented. Finally, this review provides a summary of the literature gaps which informed the research questions addressed within this thesis.

Chapter 3 describes the methodology in two parts: the first part describes the sampling design, selection methods and data collection instruments of the How Areas in Brisbane Influence healTh And acTivity (HABITAT) survey, the multilevel study underpinned by a social-ecological framework used in this thesis; the second part provides more specific information on relevant measures as well as the analytic and statistical modelling strategies undertaken to address the research questions.

Chapter 4 presents Study 1, which examined whether gender and age differences in walking for recreation (WfR) and walking for transport (WfT) were similar or different across neighbourhoods. This study used Wave 2 of HABITAT (collected in 2009), involving a sample of 7,866 residents aged 42-68 years living within 200 Brisbane neighbourhoods. On average, women were significantly more likely to engage in WfR at moderate and high levels and no gender differences in WfT were observed. Older adults were significantly less likely to walk for transport and more likely to walk for recreation at high levels. More interestingly, the relationships between gender and walking, and age and walking, were not the same in all neighbourhoods (i.e. the Brisbane average concealed important information), suggesting that neighbourhood-level factors influenced the walking patterns of men and women, and younger and older adults, differently. The subsequent two studies focused on identifying these neighbourhood-level factors.

Chapter 5 presents Study 2, which investigated the contribution of the neighbourhood social environment (assessed through neighbourhood-level perceptions of social cohesion, incivilities, and safety from crime) to explaining the gender differences in WfR observed across neighbourhoods in Study 1. Study 2 used Wave 2 of HABITAT (collected in 2009), involving a sample of 7,866 residents aged 42-67 years living within 200 Brisbane neighbourhoods. On average, women were more likely than men to walk for recreation prior to adjustment for covariates. Gender differences in WfR varied significantly across neighbourhoods (as previously established in Study 1), and the magnitude of the between-neighbourhood variation for women was twice that of men, suggesting that women are more sensitive to their neighbourhood environments in regards to WfR. However, the social environment did not explain neighbourhood differences in the gender-WfR relationship, nor did it explain the observed between-neighbourhood variation in WfR for men or women. This is most likely due to the noted limited variation in social environments across Brisbane neighbourhoods, an urban setting where structural differences between neighbourhoods might not be as extreme as in other cities.

Chapter 6 presents Study 3, which investigated the contribution of the neighbourhood built environment (objectively assessed through neighbourhood-level measures of residential density, street connectivity and land-use mix) to explaining the age differences in WfT observed across neighbourhoods in Study 1. Study 3 used Wave 1 of HABITAT, (collected in 2007) involving a sample of 11,035 residents aged 40-65 years living within 200 neighbourhoods. On average, older adults were less likely to walk for transport. Age differences in WfT varied significantly across neighbourhoods (as previously established in Study 1), and the magnitude of the between-neighbourhood variation for older groups was twice that of the youngest group, suggesting that older adults are more sensitive to their neighbourhood environment than their younger counterparts. The built environment played a limited role in explaining neighbourhood differences in the age-WfT relationship. Residential density and street connectivity (but not land use mix) partially explained the observed between-neighbourhood variation in WfT for across age groups.

Finally, Chapter 7 provides the discussion and conclusions of this research program. Collectively, the three studies comprising this thesis confirmed that the walking patterns of men and women, and younger and older persons are differently shaped and circumscribed by different neighbourhood environments. In particular, women and older adults seemed more sensitive to their environments than their counterparts, suggesting that they might require more supportive environments to walk. While Brisbane’s social environment did not contribute to explaining gender differences in WfR across neighbourhoods, the age differences in WfT across neighbourhoods were partly attributed to the contextual effects of residential density and street connectivity. Thus, in designing neighbourhoods that facilitate active living and ageing communities, governments should consider denser and more connected urban forms which would produce more equitable increases in WfT across age groups.

This body of evidence contributes to the literature investigating the important role that the neighbourhood design has in facilitating the healthy lifestyle of residents who are regularly exposed to it. More specifically, the findings from this thesis favour the ongoing multilevel analyses of demographic heterogeneity around the neighbourhood averages –rather than mean centric approach– as they more realistically reflect the impact of neighbourhood exposures on the walking patterns of different demographic groups. As cities vary widely in their social and built environments, such research –especially when undertaken in urban settings characterised by larger variation in their environments– is relevant for informing ecological interventions which facilitate walking opportunities everywhere for all demographic groups, particularly those predisposed to inactivity, resulting in sustainable public health, socioeconomic and environmental gains for the overall population.