How many walking and cycling trips made by elderly are beyond commonly used buffer sizes: Results from a GPS study

Highlights

• Distances walked are mostly beyond 400 and 500 m buffers.

• Distances cycled were mostly beyond 400, 500, 1000, 1200 and 1600 m buffers.

• Adults ≥75 years and those with functional limitations walked and cycled shorter distances.

• Males cycled shorter distances than females.

• Studies may miss important environmental characteristics if very small buffers are used.

Abstract

In choosing appropriate buffer sizes to study environmental influences on physical activity, studies are hampered by insufficient insight into the distance elderly travel actively. This study aims at getting insight into the number of trips walked and cycled within various buffer sizes using GPS measures. Data were obtained from the Elderly And their Neighborhood study (Spijkenisse, the Netherlands (2011–2012)). Trip length and mode of transport were derived from the GPS data (N=120; total number of trips=337). Distance decay functions were fitted to estimate the percentage of trips to grocery stores within commonly used buffer sizes. Fifty percent of the trips walked had a distance of at least 729 m; for trips cycled this was 1665 m. Elderly aged under 75 years and those with functional limitations walked and cycled shorter distances than those over 75 years and those without functional limitations. Males cycled shorter distances than females. Distance decay functions may aid the selection of appropriate buffer sizes, which may be tailored to individual characteristics.

Introduction

In aging societies, health promotion among elderly increasingly focuses on aging in place (Farber et al., 2011, Zantinge et al., 2011). This focus covers both societal need to prevent further increases in costs for institutionalized care, as well as the desire of most elderly people to age in place (Cheek et al., 2005). Physical activity plays an important role in this respect. Being able to walk or cycle to grocery stores contributes substantially to independent living and to health and well-being of elderly (Matthews et al., 2007, Noda et al., 2005, Savela et al., 2010, Smith et al., 2007). It is recognized that physical activity depends on individual characteristics (e.g. age, physical conditions, cognitions towards physical activity) and environmental characteristics (Alfonzo, 2005, Kremers et al., 2006, Sallis et al., 2002) such as the presence of shops (Chaudhury et al., 2012, Van Cauwenberg et al., 2012). Various studies have explored the role of physical environmental factors for walking among elderly. A review of the literature reveals that evidence for such associations is mixed (Van Cauwenberg et al., 2011). These mixed results have been attributed to methodological limitations such as the way buffers to measure environmental characteristics were constructed (i.e. circular vs. road networks) (Oliver et al., 2007) and their size (i.e. “buffer sizes”) (Brownson et al., 2009). Circular buffers become less accurate with the occurrence of physical barriers (e.g. rivers, highways). Therefore, road network buffers are often the preferred methods for constructing a buffer to study environmental correlates of physical activity behavior (Oliver et al., 2007). When choosing an appropriate buffer size to capture environmental characteristics, it is important to know how far people travel actively, which gives an indication of the environmental factors to which people are likely exposed. Auchincloss et al. concluded that there is still little knowledge regarding the relevant buffer size to study the impact of the environment on physical activity (Auchincloss et al., 2012). Some studies have investigated the importance of the size of buffers by studying correlates of physical activity within various buffer sizes (Dowda et al., 2009, Nagel et al., 2008, Prins et al., 2011), but little is known about the number of trips, made by means of active travel, within and beyond these buffers. Information on the relative number of trips made within and beyond these buffers provides insight into the exposure of elderly to their neighborhood environment. We therefore aimed to get insight into the number of trips made beyond various network buffer sizes using GPS measures. This knowledge may aid researchers in constructing appropriately sized network buffers.

A factor that complicates the determination of distances that elderly are willing to travel actively for their grocery shopping is that these distances may differ according to personal factors, such as age and gender (Perchoux et al., 2013). This may have consequences for relevant buffer sizes in which environmental influences on active transport behavior are studied. It has been shown that self-reported trip length decreases with increasing age (Yang and Diez-Roux, 2012) and that men seem to cycle less than women (Krizek et al., 2005). In a similar way it is likely that people with functional limitations are more bound to their residential environment and therefore travel shorter distances actively. Insight into how personal characteristics affect the distance people travel actively is needed to come closer towards the individualized buffer sizes where has been called upon (Diez Roux et al., 2007, Perchoux et al., 2013). Therefore, we also examined differences in distances traveled actively according to gender, age, and level of functional limitations.

In summary, it is the aim of this study to a) get insight into the number of trips walked and cycled beyond various buffer sizes using GPS measures and b) examine whether these differ according to gender, age, and level of functional limitations. We hypothesized that, given the high connectivity and relatively good infrastructure for walking and cycling in The Netherlands, most elderly will walk or cycle to get their groceries for distances well beyond 1000 m. In addition, we hypothesized that trip length decreases with increasing age and functional limitations.