Our study was the first to examine associations between neighborhood walkability and different types of PA in European older adults. Previous research on these relationships was predominantly conducted in North America, where the built environment can differ substantially from other continents . Consequently, this raised questions on the applicability of North American results to other geographical contexts and, therefore, the present study sought to reveal relationships in a Western European context.
A first aim of our study was to identify main effects of walkability in relation to older adults’ objectively-measured and self-reported PA. Regarding objective levels of PA, the present study investigated associations between walkability and LPA in addition to MVPA, given that the vast majority of older adults’ daily total PA consists of low-intensity activities ,. Accelerometer low-light PA was the most prevalent type of all objectively-measured PA and there was a negative association with neighborhood walkability. Presumably, residents of low-walkable neighborhoods spend more time doing activities indoors, such as cleaning up or doing other small household chores (e.g., ironing). House-based activities may be less physically challenging and of shorter duration than outdoor activities (e.g., walking to a shop). In contrast to the significant results found for low-light PA, walkability was unassociated with older adults’ high-light PA. A univocal explanation for this finding is difficult to formulate, as no other studies have investigated relationships between walkability and older adults’ accelerometer LPA yet and results cannot be compared. Nevertheless, these contrasting findings for low-light versus high-light PA indicate that splitting total LPA into low-light and high-light PA, according to the Copeland cut point of 1,041 counts.min−1 may be useful to get a better insight into how the built environment is related to older adults’ different types of LPA.
Next to objectively-measured PA, we examined self-reports of older adults’ domain-specific PA. Walkability was positively associated with older adults’ walking for transportation, and this finding supports results from Belgian Environmental Physical Activity Study in adults (BEPAS Adults) , as well as other non-European research among adults ,, and older adults ,. As walking from place to place is an accessible and low-cost activity, which older adults can easily integrate into their daily routine, this positive relationship is very promising from a health promotion perspective. Given that 65% of the participants reported to have at least two physical limitations that restricted them from doing daily-life activities, short trips may be most desirable. Access to destinations (e.g., shops, public services, places for social interaction) in the close vicinity of the home residence may thus be a key environmental correlate behind this significant relationship. Previous qualitative , and quantitative – research on the PA-environment relationship in older adults has already provided some evidence for the importance of destinations for walking in the elderly. However, having destinations nearby only relates to one of the three components of walkability (i.e., land use mix diversity), whereas the walkability index also includes street connectivity and residential density. Both of these factors have been positively associated with older adults’ transport-related walking in some prior studies ,. With regard to our findings, a higher street connectivity may facilitate walking because street segment length and block size are smaller and destinations are more directly accessible ,. Walkability was not associated with any measure of self-reported recreational PA in our study. This contrasts with the BEPAS Adults, where participants walked more for recreation in a high-walkable neighborhood , but is consistent with some previous work in older adults ,,. Perhaps, other aspects of the physical environment could be more important in explaining recreational PA among older adults, such as the presence of parks or traffic levels . Another possibility is that socio-ecological factors other than the physical environment, for instance social support or strength of one’s social network ,, may be more closely related to recreational PA in this age group.
The present study was the first to examine associations between walkability and cycling in the elderly, but no significant relationships were found. This might be due to a low prevalence of cycling in older adults. As cycling requires a certain basic fitness and balance/agility level, participants with physical limitations may have perceived themselves as more vulnerable to get injured. Another reason for the lack of association may be that older adults who did cycle, cycled longer distances, beyond the boundaries of their residential neighborhood where walkability was measured.
A second aim of our study was to investigate possible moderating effects of neighborhood income (a proxy for SES) on the walkability-PA relationships. In the present study, older adults living in low-income neighborhoods accumulated about 11 additional weekly minutes of objectively-measured MVPA when they resided in a high-walkable, compared to a low-walkable neighborhood. In contrast with prior studies among Iranian and US older adults, which could not identify any income x walkability interactions ,, this was the first study to show that the observed relationship between walkability and PA was only present among older adults living in low-income neighborhoods. Our observed interaction indicates that in Belgium, the built environment may be of greater importance for certain population subgroups and public health promotion initiatives should be adapted accordingly. Economic factors might partly explain this finding; low-income neighborhood residents may have less access to the private car (i.e., own a car themselves or have significant others owning a car) than residents of high-income neighborhoods and may therefore spend more time in their neighborhoods, making them more dependent of walkability of their residential environment. In contrast, high-income neighborhood residents might be more mobile and could spend more time outside the neighborhood boundaries, which in turn could explain why walkability of their residential neighborhood was unassociated with MVPA levels in this group of older adults.
Strengths and limitations
A first strength of BEPAS Seniors is that it is the first European study examining specific walkability-PA relationships in older adults and to our knowledge, the only study investigating associations with cycling and light-intensity PA. Hence, our results are useful in updating the current knowledge on the PA-environment relationship in older adults across the broad spectrum of geographical settings. Secondly, the study design of BEPAS Seniors is identical to the design of BEPAS Adults , making it possible to get a better insight in environment-PA relationships across different age groups and identify age-specific correlates of Belgians’ PA levels. Besides, the study design is similar to the US Neighborhood Quality of Life Study for Seniors ,,, providing the opportunity to compare study findings cross-nationally and even cross-continentally. A third strength of the present study is that we assessed domain-specific measures of self-reported PA. Our different outcomes regarding transport-related versus recreational PA confirm that the PA-environment relationship can be domain-specific ,. Moreover, self-reported PA levels were collected through face-to-face interviews. As older adults may experience more cognitive difficulties when responding to a questionnaire , the guidance by the interviewer was experienced to be beneficial for this population of elders because more accurate responses can be obtained. In addition, as a previous study suggested that older adults tend to over-report levels of total MVPA (Van Holle et al., submitted), accelerometers were used to objectively measure levels of total MVPA in the present study, which is another strength of the study. However, some limitations need to be acknowledged as well. First, our study only presents cross-sectional data and relationships cannot be assumed to be causal. Longitudinal studies in older adults are recommended to determine whether the identified relationships change or last over time. Additionally, using longitudinal designs will enable researchers to examine whether a more activity-friendly built environment (e.g., higher walkability) may help older adults to maintain their PA levels over time. Moreover, natural experiments are encouraged, to investigate if over-time structural changes in the built environment will cause changes in older adults’ PA. A second limitation is that no information on the specific context of PA was available, so we cannot make sure that the objectively-measured and self-reported PA occurred in the residential neighborhood. Future studies incorporating GPS measurements may provide such additional information.