The aim of this study was to examine factors associated with ATS. including, for the first time, NHSS status. This was also the first study to consider ATS and weather patterns as daily repeated measures, improving sensitivity and modelling robustness. Results showed a significant association between NHSS status and ATS, whereby children who lived in a low-walkable neighbourhood, but whose parents preferred a highly walkable neighbourhood (NHSSPHLL) were three times less likely to use ATS than their counterparts (NHSSOTH). In other words, children residing in a highly walkable neighbourhood (irrespective of parental neighbourhood preference) or those residing in a low-walkable neighbourhood whose parents preferred a low-walkable neighbourhood, were significantly more likely to use ATS. The latter may be indicative of an issue of socio-economic status, whereby a “match” in neighbourhood walkability and preference was indicative of a family’s ability to afford to live in a neighbourhood of their choosing. The former supports adult research that shows the positive influence of neighbourhood walkability on active transport behaviours. NHSS has explained approximately 42% of differences in latent modelling of adult vehicle miles travelled between similar households living in urban/more walkable versus rural/less walkable neighbourhoods . Likewise, preferring and residing in a more walkable neighbourhood was associated with active transport in a large sample of New Zealand adults . No other comparable examinations for children exist.
As observed in previous research , no relationship was found between daily weather patterns and ATS. The dichotomisation of rainfall as none versus some may have hindered our ability to detect any association between substantial rainfall and ATS. Due to the nature of the rainfall data however, this approach was necessary to ensure that modelling remained robust. Moreover, previous research has shown that even when comparing days with no versus some rain, significant differences in physical activity levels can be found in children [37, 38]. ATS may be less amenable to temporal factors such as weather and determined predominantly by pre-existing built environment and social variables such as time and convenience . While improving on earlier research that has considered seasonality or weekly weather patterns in relation to activity, the use of a daily measure of weather data may still have been insufficiently sensitive to identify relationships between ATS and weather patterns. Moreover, weather data were captured from one primary weather station for each respective city. As such, differences existed in distance to these weather stations across and within cities. Therefore, it is possible that differential weather patterns were observed for neighbourhoods and individuals within each city and so the association between weather factors and ATS may have been diluted accordingly. Future research should consider better spatio-temporal matching of weather exposures for individuals. For example, this might involve extracting weather data for periods of the day where ATS might be expected to occur, and undertaking measures of weather at finer spatial resolutions (e.g., at the school or neighbourhood, rather than city, level). It is possible, however, that decision making regarding travel mode is not limited to exact temporal or spatial exposure. For example, predicted weather patterns or heavy rainfall in the early morning may influence travel plans for later periods of the day, irrespective of actual weather at the time of the journey.
Significant differences were observed in ATS between cities, with children residing in North Shore City approximately twice as likely to use ATS than their counterparts living in other cities. These findings conflict with national prevalence data for New Zealand that suggests a greater proportion of trips are made by walking or cycling in Christchurch and Wellington Cities than in Auckland City (4%, 3%, and 2% of kilometres travelled per person (children and adults), per year, respectively) . Reasons for this finding are unclear; it is possible that local initiatives such as the Travelwise school travel plan programme, initiated in North Shore City prior to data collection for the current study, may have influenced children’s travel behaviours in this region . It is also possible that variables such as home ownership and length of residency may reflect a greater ability to ‘self select’ a neighbourhood, and that these variables differed between cities, however we were unable to assess these differences in the current investigation.
Household income was not significantly related to ATS after accounting for other factors in the multivariate modelling. Earlier New Zealand research has shown that children residing in high deprivation areas are more likely to use ATS than those living in the least deprived areas . However, internationally, research investigating associations between socio-economic status and ATS has been equivocal, with positive, negative, and insignificant relationships found [26, 61]. Similarly, after accounting for other significant factors from the bivariate analyses, we found no relationship between car access (or lack thereof) and ATS in the current study. Nearly all (90%) respondents had frequent or unlimited car access, thus homogeneity in this factor may have hindered our ability to detect a relationship with ATS .
In keeping with previous research, our findings showed increasing distance to school was significantly related to a reduced likelihood of ATS [30, 63]. A substantial drop in prevalence of ATS was seen even for those children who lived further than 700 m from school. It is worth noting that almost all of the ATS observed in the current study was via walking. A study with parents of Belgian youth aged 11–12 years suggested that criterion distances of 1.5 km and 3.0 km are optimal for encouraging ATS via walking and cycling respectively , however whether these findings hold true for other populations remains to be determined. School catchment zones vary widely in New Zealand (up to 90 km using a Euclidean diameter). US data suggest that only 20% of children live within 1.6 km (1 mile) from school . Even so, for children who do live within this distance, surveys have shown that a high proportion of children do not actively commute to school [32, 64]. A number of Australian studies exemplify the discrepancies between residing close enough to school for children to use active transport despite little uptake of active travel modes. Parents of children aged 5–6 and 10–12 years identified a walking distance of 800 m in one direction as being appropriate for their children, roughly equivalent to a 15-minute walk . Despite this observation, a later study of 4–13 year old children found that while over half of the participants lived within a 15 minute walk to school, parents still reported their child’s school was too far away to reach by walking . Yet another study showed that of children living within 400 metres of school, 21% were still driven by car, even though trip durations by car or walking were strikingly similar (mean duration of 8 and 7 minutes, respectively) .
Cumulatively, these results suggest that localised schools nested within communities may facilitate increased uptake of ATS. Irrespective of actual school zoning, recent trends in school siting and consequent effects on upsizing have created a significant barrier to children actively travelling to and from school. For example, student numbers in the US have grown, yet the number of small local schools has dropped and there has been a consequent increase in ‘supersized’ schools that service a wider geographic spread . When considering school siting, it may also be important to take into account other factors such traffic volume, which may mediate or moderate the positive effect of street connectivity on ATS .
Age was significantly related to ATS, whereby children aged 11–14 years (reflecting intermediate and secondary school ages) were approximately three times more likely to use ATS than their younger counterparts (aged 5–10 years). Although contradictory findings have been reported regarding school travel and age , our findings are in keeping with those from the national New Zealand Travel Survey, which show a greater prevalence of walking and cycling for transport in youth aged 13–17 years (31%), compared with children aged 5–12 years (29%) . All other factors being equal, an increase in ATS with age/school level is unsurprising, and may be indicative of increasing parental licence, whereby older children have more freedom to travel independent of adult supervision .
Aside from NHSS, parental factors such as safety concerns (e.g., about crime, traffic, sidewalks and cycle lanes/bikeways), supports for ATS, and factors influencing these were not examined. As such, we cannot determine the relative contribution that distance to school has above and beyond these parental factors, which may also be independently associated with ATS, or moderate or mediate the relationships found [22, 30, 33]. We also focused on the trip to school only, a pragmatic choice based on the expectation of less trip-chaining on the trip to school , as recently evidenced in a study of independent mobility in New Zealand children .