Walking and cycling to school are the two most common modes of active transport to and from school (ATS). They are convenient and environmentally friendly ways to foster daily physical activity among adolescents (Aubert et al., 2018; Guthold et al.; 2020; Khan et al., 2020) while also providing a range of health, personal, and social benefits (Kek et al., 2019; Mandic, Hopkins et al., 2017; Carver et al., 2005). In New Zealand, fewer than 50% of adolescents use ATS (Smith et al., 2018), with walking being more popular than cycling to school (Mandic, Hopkins et al., 2017). Walking to school is also more popular in other countries, including Canada (59% vs 3%) (Larsen et al., 2009), China (40% vs 16%) (Sun, Liu and Tao, 2015), Great Britain (40% vs 16%) (Department for Transport, 2017), Ireland (18% vs 2%) (Murtagh, Dempster and Murphy, 2016), Spain (46% vs 2%) (Chillón et al., 2013), and Switzerland (34% vs 10%) (Bringolf-Isler et al., 2008). However, in some countries, such as Denmark (Cooper et al., 2006), Germany (Reimers et al., 2013), and the Netherlands (Bere et al., 2008), cycling to school is more popular than walking among adolescents at least in part due to a complex mix of socio-historical and cultural factors, ongoing investment in cycling-friendly infrastructure, and supportive policies (Pucher and Buehler, 2008).

In general, most studies have examined walking and cycling to school combined under the term “active transport” (Martin, Lee and Lowry, 2007; Babey et al., 2009; Evenson et al., 2006; Robertson-Wilson, Leatherdale and Wong, 2008) or focused on walking only (Mitra and Buliung, 2015) or cycling only (de Bruijn et al., 2005). However, walking and cycling are distinct modes of transport with distinct characteristics. For example, mode-specific enablers and barriers have been identified by both New Zealand adolescents (Mandic, Hopkins et al., 2017; Mandic et al., 2022) and their parents (Mandic et al., 2020), suggesting the importance of examining these two modes of ATS separately. In these New Zealand studies, both parents and adolescents perceived cycling as less safe with less social support and more environmental barriers than walking to school.

Further, most studies examining adolescents’ school travel have been conducted in main urban centres (Martin, Lee and Lowry, 2007; Mitra, Buliung and Roorda, 2010; Jones and Sliwa, 2016; Easton and Ferrari, 2015; Babey et al., 2009; Chillón et al., 2013), with few studies focusing on rural areas (Bungum et al., 2008; Sjolie and Thuen, 2002). As these geographic settings have distinct social, cultural, economic, and built environment contexts, the rates of adolescents’ walking and cycling to school as well as enablers of and barriers to their use are likely to differ in rural and urban areas. Typically, a higher prevalence of ATS is observed among adolescents in urban versus rural areas (Pabayo and Gauvin, 2008; Babey et al., 2009; Yang, Diez Roux and Bingham, 2011; Potoglou and Arslangulova, 2017; Sjolie and Thuen, 2002; Robertson-Wilson, Leatherdale and Wong, 2008; Fulton et al., 2005). This observation may at least in part be attributed to longer distances to school travelled by rural adolescents. In contrast, a New Zealand study reported a higher prevalence of ATS among adolescents in rural versus urban areas (Mandic et al., 2015).

Distance from home to school is the strongest predictor of ATS in adolescents (Ikeda et al., 2018; Mandic et al., 2015; De Meester et al., 2013). Threshold distances for adolescents walking to school range from 1.4 km to 3.0 km (Bere et al., 2008; Pocock et al., 2018; Nelson et al., 2008, Chillón et al., 2015), whereas reasonable cycling distances range from 3.0 km to 8.0 km (D’Haese et al., 2011; Nelson et al., 2008; Bere et al., 2008; Van Dyck et al., 2010), with large variability between countries. In addition, parental (Mandic et al., 2020) and adolescents’ (Mandic et al., 2022) perceptions and rates of walking and cycling to school change with increasing home-to-school distance. Thus, efforts to promote ATS will most likely only be effective among adolescents living within a reasonable walking and/or cycling distance from their school. Therefore, studies examining adolescents’ perceptions of walking and/or cycling to school should be conducted in adolescents for whom using ATS is potentially feasible.

In New Zealand, adolescents living within 4.8 km of their closest school are ineligible for school transport financial assistance (Ministry of Education, no date); consequently, they rely on walking, cycling, private vehicle transport, or public transport for their school travel. Although fare-based public buses are available in the main urban centres and some medium-sized urban areas, limited public transport provision is characteristic of rural areas internationally, including in New Zealand (Nutley, 2003; Nutley, 1996).

To address the paucity of rural transport knowledge, rural-specific transport habits and perceptions need to be investigated to develop appropriate place-based interventions and policies to promote ATS and ultimately increase adolescents’ physical activity. Research questions for this study were guided by socio-ecological frameworks of travel behaviour (Panter, 2008; Sallis, 2006) and the theory of planned behaviour (Ajzen, 2002). Socio-ecological frameworks of travel behaviour suggest that a wide range of individual, social, environmental, and policy factors influence how adolescents travel to school (Panter et al., 2010; Davison and Lawson, 2006; Pont et al., 2009). In addition, the theory of planned behaviour (Ajzen, 2002) illustrates the role of individual beliefs and perceptions. Briefly, this theory posits that the intention to use ATS, as a direct antecedent of the behaviour itself, can be predicted based on attitudes towards walking and cycling, subjective norms, and perceived behavioural control. Therefore, this study compared the perceptions of walking and cycling to school among adolescents attending a secondary school within 4.8 km (3.0 mi) of their home in rural settings in the Otago region of New Zealand.

The key findings are that the adolescents in rural settlements living within 4.8 km of school in Otago, New Zealand, (1) more frequently walked to school than cycled; (2) reported greater peer, parental, and school support for walking versus cycling to school; (3) perceived better physical infrastructure for walking compared to cycling to school; and (4) expressed greater desire and intention to walk than cycle to school. In contrast, their perceptions of safety, distance to school, and trip duration were not significantly different for walking versus cycling to school. This is one of the first studies to compare perceptions of walking versus cycling to school exclusively among rural adolescents. The findings illustrate differences in prevalence of and preferences for walking versus cycling to school as well as mode-specific barriers among adolescents living in rural New Zealand.

These findings extend the existing evidence from urban centres of New Zealand and other countries demonstrating that, compared to cycling, walking is more common, more favourable, and a better supported mode of transport to school among rural New Zealand adolescents who are ineligible for the subsidised school bus. In this sample of adolescents living within 4.8 km from their school, approximately three-quarters used ATS regularly (74%), with the rates of walking to school being significantly higher than the rates of cycling (73% vs 11%). These data are not directly comparable to findings from previous studies conducted in similar populations due to the lack of an internationally accepted definition of “rural” areas (Statistics New Zealand, no date), and in this study, only those living within 4.8 km from their school were included. It is also possible that the very short distance to school in our study sample (median distance: 700 m) may also have had an influence on mode choice for school travel for rural adolescents. For example, for such a short journey, getting the bicycle out, cycling to school, and locking it up at school may take more time than the time required to walk to school. However, although our inclusion criteria resulted in a relatively small sample size and very short median distance to school, the higher prevalence of walking versus cycling to school is similar to findings from other studies conducted in rural adolescents (Bungum et al., 2008; Nelson et al., 2008) as well as urban adolescents in many countries (Timperio et al., 2006; Sun, Liu and Tao, 2015; Department for Transport, 2017; Murtagh, Dempster and Murphy, 2016; Larsen et al., 2009; Buliung, Mitra and Faulkner, 2009; Chillón et al., 2013; Chillón et al., 2010; Bringolf-Isler et al., 2008), including New Zealand (Mandic, Hopkins et al., 2017; Mandic et al., 2015).

In the present study, 61% of adolescents stated that they like to cycle for recreation, and only one in five adolescents perceived that cycling to school is not cool. This suggests that most rural adolescents seem interested in cycling, despite the low percentage that actually cycled to school. Adolescents’ enjoyment of cycling should be considered in future actions aimed to encourage cycling to school in this age group. However, recreational cycling is most likely happening in locations different from the roads with traffic that rural adolescents encounter on their school route.

In this study, adolescents reported greater peer and parental support and peer and parental role modelling for walking compared to cycling to school. Only one-fifth of adolescents often cycled with friends or parents. Similar findings were previously reported among urban adolescents (Mandic, Hopkins et al., 2017) and their parents (Mandic et al., 2020) from Dunedin city, New Zealand. Extensive evidence from the literature points to the importance of peer support (Mandic, Hopkins et al., 2017; Hinckson, 2016; Deforche et al., 2010; Carver et al., 2005; Ducheyne et al., 2012) and parental support (Carver et al., 2005; Deforche et al., 2010; Mandic, Hopkins et al., 2017; Ducheyne et al., 2012) for both walking and cycling to school among adolescents. Change over time in social capital, social trust, and parental safety concerns limited adolescents’ active transport and independent mobility among Otago adolescents (Porskamp et al., 2019), further emphasising the importance of interventions to increase social support for adolescents’ ATS in rural areas.

The built environment also influences adolescents’ and their parents’ decisions about walking and/or cycling to school. In this study, rural adolescents reported greater availability and quality of infrastructure for walking (footpaths) compared to cycling (cycle lanes) to school. Similar findings have previously been reported by adolescents (Mandic, Hopkins et al., 2017; Hopkins and Mandic, 2017; Mandic et al., 2022) and their parents (Mandic et al., 2020) from Dunedin city, highlighting less physical infrastructure support for cycling in general, compared to walking, in both urban and rural areas of New Zealand. Previous studies conducted in urban centres reported that the presence of footpaths (Boarnet, Anderson et al., 2005a; Ewing, Schroeer and Greene, 2004; Ewing and Cervero, 2010; Fulton et al., 2005; Kerr et al., 2006) and/or cycle lanes (Kerr et al., 2006) and higher quality and greater availability of walking and cycling infrastructure (Aarts et al., 2012; Boarnet, Day et al., 2005b; De Meester et al., 2013; Fulton et al., 2005; McDonald et al., 2013; Stewart, 2011) are positively associated with higher rates of walking and/or cycling to school among adolescents. A systematic review by Smith et al. (2017) suggested that improvements in footpaths and the availability of cycle lanes could increase walking and cycling among adolescents. However, rural settings are fundamentally different from urban areas in terms of the provision, use, need, and likelihood of development of ATS-supportive physical infrastructure. For example, in rural areas, adolescents can walk on verges/edges and shoulders on roadsides. In contrast, cycling requires more specific infrastructure. Many towns and villages in the Otago region of New Zealand are located along main cycle trails built for tourism (New Zealand Cycle Trails, no date), but those cycle trails do not necessarily link residential areas to adolescents’ schools.

Few adolescents in rural New Zealand expressed concern about the safety of walking and cycling to school, which differs from findings in urban adolescents in New Zealand (Mandic, Hopkins et al., 2017; Mandic et al., 2022) and elsewhere (De Meester et al., 2013; Leslie et al., 2010; Aarts et al., 2012), where adolescents’ and their parents’ concerns about both personal safety (Evenson et al., 2006; Leslie et al., 2010; Aarts et al., 2012) and traffic safety (Mitra and Buliung, 2014; Carver et al., 2005, Hume et al., 2009; Giles-Corti et al., 2011) are negatively associated with adolescents’ ATS. The urban New Zealand study revealed a complex range of factors that contributed to perceptions of cycling safety, including features and perceptions of the built environment and traffic safety as well as previous cycling experiences (including crashes) and adolescents’ cycling skills and on-road experiences (Hopkins and Mandic, 2017). Since parental (Mandic et al., 2020) and adolescents’ (Mandic et al., 2022) perceptions of barriers to walking and cycling (e.g., perceived safety) vary by home-to-school distance, it is possible that the lack of safety concerns reported by rural adolescents in the present study is at least in part attributed to the study sample selection which resulted in half of rural adolescents living 0.7 km (0.4 mi) from their school.

Approximately one-quarter to one-third of rural adolescents in this study identified that cycle-friendly uniforms, having a school locker, being able to cycle without a helmet, safer bicycle storage at school, and slower traffic on the roads would encourage them to cycle to school more frequently. Similar findings were previously reported in urban New Zealand adolescents (Mandic, Hopkins et al., 2017). These findings provide valuable information for future interventions aimed at encouraging cycling to school among adolescents living in both rural and urban areas.

Although 41% of adolescents in the present study believed that cycling skills training would make them safer in traffic, approximately one-quarter stated that they would participate in such training at their schools, which is similar to findings previously reported in urban adolescents in New Zealand (Mandic, Flaherty et al., 2016; Mandic, Flaherty et al., 2017).

Previous studies have shown that cycle skills training can improve cycling-related knowledge of road rules and confidence to cycle among children (Mandic et al., 2018) and adolescent girls (Mandic et al., 2018), although such training did not increase the rates of cycling to school among urban adolescent girls following the 6- to 10-week training programme (Mandic et al., 2018). However, in countries like New Zealand, where traffic safety is a major barrier to cycling for transport, several weeks of cycle skills training is unlikely to prepare adolescents for safely sharing the roads with motorised vehicles. Instead, a cultural shift around cycling (i.e., perceived acceptability and thus appropriate safety measures in place) may be necessary to increase the rates of cycling for transportation.

Rural-specific research is necessary for advancing knowledge and understanding and for appropriately designing interventions that encourage walking and cycling within this context. Among adolescents living in rural New Zealand and residing within 4.8 km (3 mi) of their school, ATS participation was high, with the prevalence of walking to school being seven times higher than cycling. Compared to cycling to school, rural adolescents had more favourable perceptions of walking and considered walking as a more desired mode of transport to school, with better social and built environment support. Therefore, future initiatives to encourage ATS among rural adolescents need to consider adolescents’ perceptions of walking versus cycling, use mode-specific approaches, and address mode-specific barriers. Interventions in rural areas should also consider context-specific opportunities for, and barriers to, encouraging walking and cycling to and from school among adolescents. The current findings present initial evidence of differences in perceptions of walking versus cycling in a sample of rural adolescents that needs to be replicated in future studies using larger samples and, ideally, longitudinal designs.

Data used in data analysis for this project will not be shared due to sensitivity of the collected data as well as participants having received assurances that the collected data will not be shared.