Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Parental perceived neighborhood attributes: associations with active transport and physical activity among 10–12 year old children and the mediating role of independent mobility

Parental perceived neighborhood attributes: associations with active transport and physical... Background: During the last decades, the use of active travel modes declined in all age groups. Childhood is a critical time to establish lifelong healthy patterns. To develop effective interventions in this age group, insight in the correlates of health behaviors and the possible mediating factors is necessary. Among children, the role of parents may not be overlooked. Therefore, this study aimed to examine the associations of parental perceptions of neighborhood environmental attributes with active transport and total physical activity in 10–12 year old Belgian boys and girls. Furthermore, this study examined the potential mediating effect of independent mobility on these associations. Methods: In the present study, 736 10–12 year old children and their parents from 44 elementary schools in Flanders, Belgium, participated. The children were asked to wear an activity monitor and to fill in a survey questioning demographic factors and the Flemish Physical Activity Questionnaire. The parents filled in a survey concerning demographic factors, the child’s level of independent mobility and environmental perceptions (Neighborhood Environmental Walkability Scale). Results: Overall, boys reported more active transport when parents perceived more land use mix diversity, shorter distances to school, good land use mix access, higher residential density and less pleasing neighborhood aesthetics. Higher total physical activity levels were reported when parents perceived shorter distances to school and availability of walking/cycling infrastructure. None of the associations was mediated by independent mobility in boys. Girls reported more active transport when parents perceived higher residential density, more land use mix diversity, shorter distances to school, good land use mix access, available walking/cycling infrastructure and convenient recreational facilities. Girls reported higher total physical activity levels when parents perceived high residential density, good land use mix access, well-maintained and high quality walking/cycling infrastructures and more traffic safety. Independent mobility was found to be an important mediator of these associations in girls. Conclusions: Neighborhood environmental interventions to increase children’s active transport and physical activity can be effective when combined with awareness raising programs for parents. Furthermore, among girls encouraging independent mobility may contribute to behavior change. * Correspondence: [email protected] Department of Movement and Sport Sciences, Faculty of Medicine and Health Sciences, Ghent University, Watersportlaan 2, B-9000 Ghent, Belgium Full list of author information is available at the end of the article © 2014 De Meester et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. De Meester et al. BMC Public Health 2014, 14:631 Page 2 of 14 http://www.biomedcentral.com/1471-2458/14/631 Background more attention has been given to active transport to school Regular physical activity during childhood is associated than to other forms of active transport [27,28]. A recent re- with different well-known health benefits, including the view of D’Haeseetal. (unpublished data)showed thatina maintenance of a healthy weight, protective effects on number of studies high neighborhood walkability, absence the development of diabetes mellitus type 2, promotion of freeway crossings, high land use mix accessibility and of bone health as well as prevention of cardiovascular high neighborhood safety were associated with higher levels disease risk factors [1-6]. Despite these health benefits, of active transport to school in children. Nevertheless, no many children do not achieve the public health recom- consistent results were found concerning the association mendations of 60 minutes moderate-to-vigorous phys- between active transport to school and other neighborhood ical activity per day [7]. environmental attributes like: land use mix diversity, street Walking and cycling for transport, also called ‘active connectivity, the availability of walking and cycling in- transport’ have been acknowledged as affordable and frastructure, the presence of sidewalks and cycle tracks, convenient sources of physical activity with a significant altitude differences, aesthetics, crime and traffic safety, rec- contribution to the total physical activity levels [8,9]. reational facilities and the degree of urbanization. Further- The results of the reviews of Davison et al. [9] and more, no definite conclusions could be drawn with regard Faulkner et al. [10] both showed that children who walk to the contribution of neighborhood built environmental or cycle to school have higher overall physical activity attributes to explain active transport during leisure time levels than those who use motorized transport. However, (e.g. to visit friends). during the past decades, there has been a consistent de- Thus, a number of reviews summarized the findings cline in the use of active travel modes [11-15]. As child- from studies on the association of neighborhood envir- hood is a critical to establish lifelong healthy patterns, onmental attributes with physical activity or with active the development of effective interventions to promote transport among children [25,26]. But, the findings de- physical activity and in particular active transport among scribed in these reviews are inconclusive. The inconclu- children has become a public health priority in most de- siveness of these findings can possibly be attributed to veloped countries [16]. the manner in which neighborhood environmental attri- During the last decade, ecological models have been butes were assessed, the nature of the participants and used more frequently than the more individually-oriented the underestimation of the accumulation of influences models to investigate the correlates of children’sphysical that affect physical activity. activity behavior [17-19]. Ecological models emphasize the As stated by Ding et al. [25] and Giles-Corti et al. [29], importance of environmental characteristics, after the in- the inconclusiveness of reviews summarizing the find- dividual and social factors, in explaining physical activity ings from studies on the association of neighborhood behavior [20]. Interventions designed to create environ- environmental attributes with physical activity or with ments that stimulate the targeted behavior change are active transport among children is potentially influenced appealing because they have the potential for having a by the different methods to measure environmental at- sustained impact on population groups rather than tributes. According to Ball et al. (2008), objectively de- short-term impacts on individuals [17,20]. termined built environmental attributes may potentially As a consequence of the shift from individually-oriented be indirectly associated with physical activity, whereas correlates of physical activity towards environmental- perceptions of environmental attributes have a more dir- oriented correlates as opportunity to shape physical activ- ect influence on physical activity [30]. The presence of ity, empirical evidence is accumulating documenting on environmental attributes might therefore not automatic- the role of neighborhood environmental attributes to ex- ally influence the behavior in the absence of awareness plain children’s physical activity behavior. However, in of those attributes [30]. Further, previous studies based on self-reported per- contrast to the consistent findings in adults [21-24], the association between neighborhood environmental attri- ceptions of neighborhood built environmental character- butes and children’s levels of physical activity is rather istics used parental perceptions as well as children’s perceptions. Parents are important gatekeepers and deci- blurred [25,26]. A recent review of Ding et al. (2011) con- cluded that in only 34 percent of the studies among chil- sion makers for their children’s physical activity and dren, investigating the association between perceived built travel behavior. It may therefore be the case that paren- tal perceptions of neighborhood environmental attri- environmental attributes and self-reported physical activity, a significant positive association was found. In the butes are of greater importance than the perceptions of other 66 percent of the studies, no association could be the children themselves. established [25]. Furthermore, as stated by Giles-Corti et al. [29] and Within the existing evidence on the importance of Panter et al. [27] it is likely that the impact of the neigh- neighborhood environmental attributes for active transport, borhood built environmental attributes on children’s De Meester et al. BMC Public Health 2014, 14:631 Page 3 of 14 http://www.biomedcentral.com/1471-2458/14/631 activity level is moderated or mediated by a range of fac- the last year of elementary school. To recruit the children, tors, such as certain characteristics of the children and their 148 schools were randomly selected from all elementary parents. An important overlooked factor for the association schools in East- and West-Flanders and the principals between neighborhood environmental attributes and chil- were contacted by phone. After this first contact, 44 prin- dren’s physical activity behavior might be the independent cipals agreed to let the final year of their school (1 class mobility granted by the parents. Children’s independent group per school) participate (response rate = 29.7%) and mobility can be defined as the geographical distance from gave written consent. The main reasons for refusal were children’s home to places where they are allowed to wander “no time” and “participated already in many other studies when playing and socializing [31]. It is possible that the or projects”. impact of the neighborhood built environmental attributes In total, 976 children and their parents could be reached. on children’s active transport and activity level may vary Both the children and their parents received an informative according to the level of independent mobility of the chil- letter about the study with an invitation to participate. Fi- dren. Different studies have shown that the parental per- nally, the parents of 749 children from 44 class groups ceptions of certain neighborhood environmental attributes agreed to let their child participate in this study and con- are related to the level of independent mobility of chil- sented to be involved in the study (response rate = 76.7%). dren. Neighborhoods characterized by a shorter distance to Data collection took place between September 2010 and school, less traffic danger, proximity to friends, no freeways June 2011. During school hours and under supervision of a and presence of facilities to assist active travel are found to research assistant, the children with written consent to par- be positively associated with the level of independent mo- ticipate completed a questionnaire on socio-demographic bility of children [32-35]. This, in turn, may increase their variables and their own PA behavior. Furthermore, the re- level of active transport as different studies have shown that search assistant explained the protocol of the activity moni- children who encounter less mobility restrictions have tor and its diary as the children were asked to wear an higher levels of active transport. The results of a study of activity monitor for seven consecutive days. Every child was Page et al., conducted in the UK, showed that 10–11 year also given a questionnaire to be completed by one of the old boys and girls with higher levels of independent mobil- parents. The parental questionnaire contained questions ity had higher levels of active transport [36]. Similar results concerning socio-demographic characteristics, neighborhood were foundinanAustralianstudy of Carver et al.inthe environmental perceptions and the level of independent mo- same age group [37]. Furthermore, children’s level of inde- bility of their child. Parents completed this questionnaire at pendent mobility was also found to be positively associated home. One week later, the research assistant visited the with children’s overall outdoor play [38] and accelerometer schools for a second time to collect the parental question- determined weekday physical activity [39]. naires, activity monitors and diaries. The parental question- To our knowledge no previous studies investigated the naires, activity monitors or diaries that were forgotten the mediating role of the level of independent mobility on the second visit, were redirected by the teacher or collected by associations between parental perceptions of neighbor- the research assistant during a third visit. hood environmental attributes and physical activity behav- The study protocol received approval from the Ethics iors such as active transport. When investigating these Committee of Ghent University Hospital. indirect pathways, it is important to take into account the gender of the children as the level of independent mobility Measures is found to be higher in boys compared to girls [39-42]. Demographic characteristics The first aim of the present study was to examine the Children’s age and gender was questioned in the child association of the parental perceptions of neighborhood questionnaire. In the parental questionnaire, the parents environmental attributes with active transport and total were asked to fill in their own and their partner’s level of physical activity in 10–12 year old Belgian boys and girls. education. Educational attainment of the children’s par- Secondly, this study examined the potential mediating ents was used as a proxy measure of children’sSES.The effect of independent mobility on the associations of the educational level of the child’s mother and father was parental perceptions of neighborhood environmental at- determined based on four options: less than high school, tributes with the level of active transport and total phys- completed high school, completed college or completed ical activity of 10–12 year old boys and girls. university. The educational level of mother and father was coded into ‘reached a college or a university level’ or Methods ‘did not reach a college or a university education level’. Participants and procedure In the present study, a large sample of 10–12 year old Neighborhood built environmental attributes children from 44 elementary schools in East- and West- To measure the parental perceptions of neighborhood Flanders, Belgium, participated. All children attended built environmental attributes, the parent version of the De Meester et al. BMC Public Health 2014, 14:631 Page 4 of 14 http://www.biomedcentral.com/1471-2458/14/631 Neighborhood Environmental Walkability Scale for Independent mobility Youth (NEWS-Y) was used [43]. The NEWS-Y has accept- To determine children’s level of independent mobility, able to good test-retest reliability with an intra-class correl- two items were included in the parent questionnaire ation coefficient between 0.56 and 0.87 [44] for parents of “How far is your child allowed to cycle from home with- children between 5 and 10 years. The neighborhood built out adult accompaniment?” and “How far is your child environmental attributes questioned were residential allowed to walk from home without adult accompani- density, land use mix diversity, distance to school, land ment?”. Response categories were: not, 0 m-200 m, use mix access, street network connectivity, availability 200 m-500 m, 500 m-1 km, 1 km-3 km, 3 km-5 km, of walking and cycling infrastructure, maintenance and 5 km-10 km, +10 km. The question “How far is your quality of walking and cycling infrastructures, aesthetics of child allowed to cycle from home without adult accom- the neighborhood, convenience of recreational facilities paniment?” was found to have good reliability with an and crime and traffic safety. The neighborhood was de- ICC of 0.79 [46]. The second question was informed by fined as the immediate environment around the house this question. The highest score on these two items was within a distance of 1 kilometer (10 – 15 minutes walk- taken as an indication of how far they were allowed to ing distance). The NEWS-Y scoring guidelines were walk or cycle from home without adult accompaniment. used to calculate the subscales. Following the NEWS-Y scoring guidelines [45], the subscale residential density Physical activity was computed by the following formula: score on The Flemish Physical Activity Questionnaire (FPAQ) question 1a (single family residences) + 12*score on was used to determine the duration (hours and minutes question 1b (row houses) + 25*score on question 1c per day) of school related active transportation (walking (apartments). The other subscales were calculated by and cycling to and from school), walking and cycling for taking the mean of the different item scores. The con- transport during leisure time and total physical activity tent, response options and descriptive statistics are level. The total physical activity level [47] was calculated giveninTable 1. using the minutes of all activities questioned in the FPAQ Table 1 Content, response options and descriptive statistics of the parental perceived neighborhood built environmental attributes Content of the item Response Mean Mean category score score boys girls Residential density (3 items) Presence of different types of residences (e.g. detached single 5-point 78.8 (26.9) 79.5 (26.3) family residences, row houses, apartments) scale Land use mix diversity (9 items) Distance to local facilities (e.g. supermarket, post office, library) 5-point 3.4 (0.9) 3.4 (0.9) scale Land use mix access (5 items) Access to neighborhood services for their child (e.g. ease to walk to 5-point 3.6 (1.1) 3.6 (1.0) public transport, ease to walk to school) scale Distance to school (1item) Distance to the school of the adolescent 5-point 2.9 (1.4) 3.0 (1.4) scale Connectivity (3 items) Connectedness of street network (e.g. presence of intersections, 5-point 3.3 (0.7) 3.3 (0.7) dead-end streets, alternate routes) scale Availability of walking and cycling Availability of walking and cycling infrastructure (e.g. footpaths and cycling 5-point 2.7 (0.9) 2.8 (0.9) infrastructure (4 items) lanes in most streets, footpaths and cycling lanes separated from streets ) scale Quality and maintenance of walking Quality and maintenance of walking and cycling infrastructure (e.g. 5-point 3.1 (1.0) 3.1 (0.9) and cycling infrastructure (5 items) maintenance of cycling lanes and footpaths, presence of lighting) scale Aesthetics (4 items) Presence of aesthetic features (e.g. green spaces, attractive buildings, 5-point 3.6 (0.8) 3.5 (0.7) streets free from litter and graffiti) scale Safety for traffic (6 items) Perceived safety from traffic problems (e.g. speed of traffic in neighborhood, 5-point 2.9 (0.7) 2.8 (0.7) availability of pedestrian crossings and traffic signals) scale Safety for crime (4 items) Perceived safety from crime (e.g. crime prevalence in the neighborhood, 5-point 3.5 (0.8) 3.4 (0.7) perceived safety from strangers) scale Convenience of recreation facilities Distance to PA facilities (e.g. sports field, sports hall, swimming pool, park) 5-point 3.5 (0.9) 3.4 (0.9) (5 items) scale PA: physical activity. none, a few, about half, a lot, all. > 30 min, 21–30 min, 11–20 min, 6–10 min, 1–5 min. strongly disagree, somewhat disagree, neither agree or disagree, somewhat agree, strongly agree. Note: all perceived built environmental attributes were positively scored: higher score = more walkable. De Meester et al. BMC Public Health 2014, 14:631 Page 5 of 14 http://www.biomedcentral.com/1471-2458/14/631 (walking and cycling to and from school, walking and cyc- 4.1.0). For every minute of reported moderate or vigor- ling for transport during leisure time, school-related sport- ous physical activity for which the activity monitor was ing activities and leisure time sporting activities). The removed, 150 steps were added to the daily number of FPAQ was found to be a reliable and reasonably valid reported step counts [53]. Among children, the import- questionnaire for the assessment of these different dimen- ance of including non-wear activities registered in diar- sions of physical activity in youth [47]. ies when using activity monitors has been demonstrated To measure free-living step counts, the children were previously [54]. asked to wear an activity monitor for 7 consecutive days, including two weekend days. The Yamax Digiwalker Data analyses SW-200 (Yamax cooperation, Tokio, Japan) and the Preliminary analyses were conducted to obtain descrip- Actigraph accelerometer, model GT1M (Actigraph MTI, tive information about the demographic characteristics Manufacturing Technology Inc., Pensacola, FL, USA) of the study sample and the study variables (children’s were used. The Yamax Digiwalker has been acknowl- level of independent mobility, self-reported minutes per edged as a valid, accurate and reliable instrument to day of active transport, total physical activity level and measure free-living step-counts [48]. The GT1M acceler- objectively (pedometer or accelerometer) determined ometer has demonstrated good reliability for measuring daily number of steps) using SPSS 17.0. steps [49]. Evidence exists that neither accelerometers Tests for normal distribution revealed some skewed nor pedometers are affected by reactivity among adoles- physical activity variables (self-reported school related cents [50,51]. Although the step counts measured by the active transportation, walking and cycling for transport Yamax Digi-walker CW-701 (the update of the Yamax during leisure time, total physical activity level). To ob- Digiwalker SW-200) have been shown to be highly cor- tain distributions that more closely approximated sym- related with the step counts of the GT1M accelerometer metry, logarithmic transformations of these variables (r = 0.78), the overall agreement between the step counts were conducted, and the transformed variables were of both monitors is rather low [52]. In the study of Kinnu- used in the analyses [55]. For ease of interpretation, nen et al. (2011), the 95% limits of agreement ranged be- summary data of untransformed physical activity vari- tween −2690 to 2656 steps/day for the mean value (mean ables are reported in minutes/day (Table 2). of accelerometer and pedometer steps/day = 6026). Further, Furthermore, all analyses were done separately for the limits of agreement varied substantially over the range boys and girls given the existing evidence that boys’ level of values. At the lowest recorded step count (mean of acce- of independent mobility is higher than girls’ level of in- lerometer and pedometer steps/day = 906) the accelero- dependent mobility [36,39-42]. meter was on average recording more steps/day than the To examine the associations between the parental per- pedometer. In contrast, at the highest step count value ceptions of neighborhood environmental attributes and (mean of accelerometer and pedometer steps/day = 12,018) self-reported minutes/day of active transport to and the accelerometer recorded less steps/day than the pedom- from school, walking for transport during leisure time, eter on average [52]. To overcome this problem, all analyses cycling for transport during leisure time, total physical were controlled for the type of monitor used. activity level and objectively determined number of steps/ The children were asked to wear the activity monitor day, multiple linear regression analyses were conducted (pedometer or accelerometer) during waking hours but using MLwin version 2.22. The dependent variables were to remove the monitor for aquatic activities and for ac- self-reported minutes/day of active transport to and from tivities that prohibit activity monitors. Together with school, walking for transport during leisure time, cycling their activity monitor, all children received a diary. Chil- for transport during leisure time, total physical activity level dren who wore a pedometer were asked to record the and objectively determined number of steps/day. These lin- dates, steps taken at the end of the day and a description ear regression analyses provided τ-coefficients (Figure 1). and duration of the activities for which the pedometer To investigate if children’s level of independent mobil- was removed in the diary. The children who wore an ac- ity mediated the association of neighborhood environ- celerometer were asked to record the duration and a de- mental perceptions with their level of physical activity, scription of the activities for which the accelerometer the product-of-coefficient test of McKinnon et al. was was removed. The children who wore an accelerometer used (Figure 1) [56]. This test included three steps that were not asked to record the steps taken at the end of the were conducted in MLwin version 2.22. In a first step, the day because accelerometers have a memory and are able to effects of the neighborhood environmental perceptions on save the steps taken during the entire recording period. At the potential mediator (children’s independent mobility) the end of the measurement period, the activity monitors were examined by regressing children’s independent mo- were collected and the accelerometer determined step bility onto the neighborhood environmental perceptions. count data were downloaded (Actilife software version Step one was conducted only for the parental perceptions De Meester et al. BMC Public Health 2014, 14:631 Page 6 of 14 http://www.biomedcentral.com/1471-2458/14/631 Table 2 Demographic characteristics and descriptive statistics for adolescents’ level of independent mobility and physical activity behavior Total sample Girls Boys Age: mean (SD) 11.2 (0.5) 11.1 (0.5) 11.2 (0.6) Weight status: % Overweight 12.0 14.2 10.0 Obese 5.0 5.1 5.0 Gender: % Male 51.9 Female 48.1 Educational level: % Mother: No college/university degree 45.6 48.2 43.2 College or university degree 54.4 51.8 56.8 Father No college/university degree 54.2 55.1 53.4 College or university degree 45.8 44.9 46.6 Self-reported physical activity (min/day (SD)) Active transport to and from school 10.2 (12.7) 9.3 (11.8) 11.2 (13.5) Walking for transport during leisure time 9.0 (12.4) 7.8 (11.2) 10.2 (13.3) Cycling for transport during leisure time 10.5 (13.1) 8.1 (11.2) 12.8 (14.3) Total physical activity 81.3 (43.1) 68.7 (35.4) 93.2 (46.3) Daily step counts (steps/day (SD)) 10 766 (3503) 9739 (2979) 11 740 (3684) Level of independent mobility (%) Not 3.5 5.7 1.4 0m – 200 m 1.3 1.8 0.8 200 m – 500 m 8.1 7.5 8.7 500 m – 1 km 17.7 20.9 14.8 1km – 3 km 32.6 31.6 33.5 3km – 5 km 22.4 20.9 23.7 5km – 10 km 10.0 7.8 12.0 +10 km 4.5 3.9 5.0 that showed a significant direct association with the chil- calculated by dividing the product-of-coefficient (αß) by the dren’s level of active transport, total physical activity level total main effect of the neighborhood environmental per- and objectively determined daily number of step counts. ceptions on the dependent physical activity variables (τ). This step provided estimates of the α-coefficients. For all analyses, 95% confidence intervals (CI) were re- In the second step, the independent effect of the poten- ported. All analyses were controlled for two proxy mea- tial mediator on the dependent physical activity variables sures of individual SES (educational attainment of mother was investigated by regressing the dependent physical ac- and father). Clustering of individuals in schools was taken tivity variables onto the neighborhood environmental per- into account by using multi-level modeling with children ceptions and the potential mediator (children’s level of at the first level and schools at the second level. All ana- mobility) (ß-coefficients). Step two was only conducted lyses including the daily step counts as dependent variable for the parental perceptions that showed a significant as- were controlled for the type of monitor (pedometer or sociation with the potential mediator. accelerometer). In the third step, the mediated effect was calculated by multiplying the two coefficients (αß). Results The statistical significance of the mediated effect was cal- Sample characteristics culated by dividing the product-of-coefficients (αß) by its Table 2 provides an overview of the descriptive statistics standard error. Moreover, the proportion mediated was for demographic characteristics, physical activity behavior De Meester et al. BMC Public Health 2014, 14:631 Page 7 of 14 http://www.biomedcentral.com/1471-2458/14/631 Figure 1 Path diagram for the mediational effect of children’s independent mobility on the association of neighborhood environmental perceptions with the children’s level of physical activity. and independent mobility for the total sample and by gen- Main effects of neighborhood environmental perceptions der. In total, 736 (98.3%) children completed the question- on active transport, the total physical activity level and naire and handed it back to the research assistant. From the daily amount of step counts (τ-coefficients) 93.5% of the parents (n = 701), a complete questionnaire Table 3 shows the main effects of neighborhood environ- returned to school. In 92.1% the children and the parent, mental perceptions on active transport, the total physical who filled in the questionnaire, had the same address. In activity level and the daily amount of step counts among 6.3% the children lived partly with the parent who filled in boys and girls. the questionnaire. In 1.6% the child did not live with the The results of the regression analyses revealed that person who filled in the questionnaire. boys reported more minutes of active transport to and In total, 649 children (86.6%) had complete step count from school when their parents perceived more land use data, 370 children wore a pedometer and 279 an acceler- mix diversity, a shorter distance to the school of their ometer. The sample consisted of 354 girls (48.1%) and child and good land use mix access. 382 boys (51.9%). Mean age was 11.2 ± 0.5 years. The When parents perceived a higher residential density, children reported on average 10.2 (±12.7) minutes/day more land use mix diversity, a shorter distance to school of walking and cycling to and from school, 9.0 (±12.4) and less pleasing neighborhood aesthetics, their male minutes/day of walking during leisure time, 10.5 (±13.1) children reported more walking for transport during leis- minutes/day of cycling during leisure time and 81.3 ure time. Furthermore, the parental perception of a (±43.1) minutes/day of physical activity. On average, the shorter distance to school was associated with more cyc- children took 10 766 (±3503) steps/day. ling for transport during leisure time and the parental Most of the boys and girls were allowed to go between perception of a shorter distance to school and availabil- 500 m and 5 km from home unsupervised. A significant ity of walking and cycling infrastructure were associated difference in independent mobility was found between with more self-reported total physical activity. No asso- boys and girls (t = 3.438; p < 0.01). Boys were allowed to ciations were found between the parental neighborhood go further from home unsupervised using active trans- environmental perceptions and the objectively determined port compared to girls. daily number of step counts. De Meester et al. BMC Public Health 2014, 14:631 Page 8 of 14 http://www.biomedcentral.com/1471-2458/14/631 Table 3 Main effects of neighborhood environmental perceptions on physical activity variables among boys and girls (τ-coefficients) Active transport to Walking for transport Cycling for transport Overall level of Daily number of step counts and from school during leisure time leisure during leisure time leisure physical activity MAIN EFFECTS AMONG BOYS τ (SE) 95% CI τ (SE) 95% CI τ (SE) 95% CI τ (SE) 95% CI τ (SE) 95% CI Residential density 0.001 (0.001) −0.001-0.003 0.004 (0.001) 0.002-0.006 0.002 (0.001) 0.000-0.004 0.000 (0.001) −0.002-0.002 7.322 (8.766) −9.859-24.503 Land use mix diversity 0.108 (0.036) 0.037-0.179 0.076 (0.039) 0.000-0.152 0.050 (0.040) −0.028-0.128 0.005 (0.015) −0.024-0.034 −4.615 (226.984) −449.504-440.274 Distance to school 0.129 (0.023) 0.084-0.174 0.065 (0.025) 0.016-0.114 0.062 (0.026) 0.011-0.113 0.026 (0.010) 0.006-0.046 −23.777 (149.181) −316.172-268.618 Land use mix access 0.185 (0.030) 0.126-0.244 0.061 (0.033) −0.004-0.126 0.036 (0.034) −0.031-0.103 0.022 (0.013) −0.003-0.047 147.627 (195.926) −236.388-531.642 Connectivity 0.091 (0.049) −0.005-0.187 0.015 (0.052) −0.087-0.117 0.010 (0.053) −0..094-0.114 0.012 (0.020) −0.027-0.051 423.100 (312.254) −206.558-1052.758 Availability of walking and cycling 0.070 (0.037) −0.003-0.143 −0.020 (0.039) −0.096-0.056 0.047 (0.040) −0.031-0.125 0.033 (0.015) 0.004-0.062 72.449 (233.694) −385.591-530.489 infrastructure Maintenance and quality of 0.035 (0.033) −0.030-0.100 −0.021 (0.034) −0.088-0.046 0.032 (0.035) −0.037-0.101 0.015 (0.013) −0.010-0.040 36.869 (202.103) −359.253-432.991 walking and cycling infrastructure Aesthetics −0.064 (0.047) −0.156-0.028 −0.116 (0.048) −0.210-(−0.022) −0.049 (0.050) −0.147-0.049 −0.012 (0.019) −0.049-0.025 −214.570 (292.276) −787.431-358.291 Safety for traffic 0.055 (0.051) −0.045-0.155 0.029 (0.054) −0.077-0.135 0.051 (0.056) −0.059-0.161 −0.010 (0.021) −0.051-0.031 −11.255 (316.694) −631.975-609.465 Safety for crime −0.010 (0.046) −0.100-0.080 −0.076 (0.047) −0.168-0.016 −0.028 (0.049) −0.124-0.068 −0.012 (0.019) −0.049-0.025 −205.080 (287.894) 769.352-359.192 Convenience of recreational 0.050 (0.040) −0.028-0.128 −0.011 (0.042) −0.093-0.071 −0.022 (0.043) −0.106-0.062 −0.006 (0.017) −0.039-0.027 150.952 (250.034) −339.115-641.019 facilities MAIN EFFECTS AMONG GIRLS τ (SE) 95% CI τ (SE) 95% CI τ (SE) 95% CI τ (SE) 95% CI τ (SE) 95% CI Residential density 0.004 (0.001) 0.002-0.006 0.005 (0.001) 0.003-0.007 0.001 (0.001) −0.001-0.003 0.001 (0.000) 0.001-0.001 6.298 (7.407) −8.220-20.816 Land use mix diversity 0.089 (0.037) 0.016-0.162 0.102 (0.039) 0.026-0.178 −0.001 (0.037) −0.074-0.072 0.026 (0.015) −0.003-0.055 21.472 (181.128) −333.539-376.483 Distance to school 0.126 (0.024) 0.079-0.173 0.046 (0.026) −0.005-0.097 −0.007 (0.025) −0.056-0.042 0.015 (0.010) −0.005-0.035 −84.890 (122.967) −325.905-156.125 Land use mix access 0.143 (0.032) 0.080-0.206 0.129 (0.035) 0.060-0.198 −0.015 (0.034) −0.082-0.052 0.035 (0.014) 0.008-0.062 286.596 (167.068) −40.857-614.049 Connectivity 0.008 (0.050) −0.090-0.106 0.019 (0.054) −0.087-0.125 0.013 (0.051) −0.087-0.113 −0.024 (0.021) −0.065-0.017 182.826 (248.738) −304.700-670.352 Availability of walking and 0.101 (0.039) 0.025-0.177 0.086 (0.042) 0.004-0.168 −0.000 (0.040) −0.079-0.078 0.004 (0.017) −0.029-0.037 −143.455 (196.440) −529.477-241.567 cycling infrastructure Maintenance and quality of 0.025 (0.037) −0.048-0.098 0.050 (0.039) −0.026-0.126 0.020 (0.037) −0.053-0.093 0.042 (0.015) 0.013-0.071 110.504 (185.024) −252.143-473.151 walking and cycling infrastructure Aesthetics −0.089 (0.050) −0.187-0.009 −0.077 (0.053) −0.181-0.027 −0.043(0.050) −0.141-0.055 0.012 (0.021) −0.029-0.053 −67.354 (250.140) −557.628-422.920 Safety for traffic −0.002 (0.050) −0.100-0.096 0.078 (0.054) −0.028-0.184 0.060 (0.051) −0.040-0.160 0.053 (0.021) 0.012-0.094 139.887 (253.962) −357.879-637.653 Safety for crime 0.039 (0.052) −0.063-0.141 −0.094 (0.055) −0.202-0.014 0.041 (0.052) −0.061-0.143 0.023 (0.022) −0.020-0.066 −44.984 (257.382) −549.453-459.485 Convenience of recreational 0.059 (0.036) −0.012-0.130 0.108 (0.039) 0.032-0.184 −0.028 (0.037) −0.101-0.045 0.022 (0.015) −0.007-0.051 25.069 (184.481) −336.514-386.652 facilities Note: All regression models were adjusted for educational attainment of mother and father and clustering by schools. All regression models including the daily step counts as dependent variable were controlled for type of monitor (pedometer or accelerometer). The significant effects were shown in bold. De Meester et al. BMC Public Health 2014, 14:631 Page 9 of 14 http://www.biomedcentral.com/1471-2458/14/631 Table 4 Regression analyses for the possible mediated effects of independent mobility among boys and girls BOYS: potential mediator: independent mobility α (SE) 95% CI B (SE) 95% CI αß (SE) 95% CI Proportion Mediated (%) Residential density −0.002 (0.003) −0.008 – 0.004 Land use mix diversity −0.128 (0.081) −0.287 – 0.031 Distance to school −0.045 (0.053) −0.149 – 0.059 Land use mix access 0.076 (0.071) −0.063 – 0.215 Availability of walking and cycling infrastructure −0.083 (0.083) −0.246 – 0.080 Aesthetics 0.078 (0.104) −0.126 – 0.282 GIRLS: potential mediator: independent mobility α (SE) 95% CI B (SE) 95% CI αß (SE) 95% CI Proportion Mediated (%) Residential density 0.008(0.003) 0.002 – 0.014 Active transport to and from school 0.059(0.024) 0.012 – 0.106 0.000 (0.000) 0.000 – 0.001 11.8% Walking for transport during leisure time leisure 0.038 (0.026) −0.013 – 0.089 Total physical activity level 0.054 (0.010) 0.034 – 0.074 0.000 (0.000) 0.000 – 0.001 43.2% Land use mix diversity 0.050(0.092) −0.130 – 0.230 Distance to school −0.011(0.061) −0.131 -0.109 Land use mix access 0.250(0.080) 0.093 – 0.407 Active transport to and from school 0.047(0.022) 0.004 – 0.090 0.012 (0.007) −0.001 – 0.0025 Walking for transport during leisure time leisure 0.028 (0.025) −0.021 – 0.077 Total physical activity level 0.050 (0.009) 0.032 – 0.068 0.013 (0.005) 0.004 – 0.023 37.7% Availability of walking and cycling infrastructure 0.200(0.099) 0.006 – 0.394 Active transport to and from school 0.057(0.023) 0.012 – 0.102 0.011 (0.007) −0.003 – 0.026 Walking for transport during leisure time leisure 0.035 (0.025) −0.014 – 0.084 Maintenance and quality of walking and cycling infrastructure 0.108(0.091) −0.070 – 0.286 Safety for traffic 0.411(0.123) 0.170 – 0.652 Total physical activity level 0.051 (0.009) 0.033 – 0.069 0.021 (0.007) 0.007 – 0.035 39.5% Convenience of recreational facilities 0.205(0.091) 0.027 – 0.383 Walking for transport during leisure time leisure 0.037 (0.025) −0.012 – 0.085 SE, standard error; CI confidence interval. Notes: α-coefficients were estimated by regressing children’s level of independent mobility onto the neighbourhood environmental perceptions. ß-coefficients were estimated by regressing the dependent active transport and physical activity variables onto the neighbourhood environmental perceptions and children’s level of independent mobility. αß-coefficients represent the mediated effect. All regression models were adjusted for educational attainment of mother and father and clustering by schools. All regression models including the daily step counts as dependent variable were controlled for type of monitor (pedometer or accelerometer). The significant effects were shown in bold. De Meester et al. BMC Public Health 2014, 14:631 Page 10 of 14 http://www.biomedcentral.com/1471-2458/14/631 Among girls (Table 3), the regression analyses revealed minutes of walking for transport during leisure time that when parents perceived a higher neighborhood resi- after taking into account the parental perceptions of the dential density, more land use mix diversity, a shorter neighborhood residential density, land use mix access, distance to the school of their child, good land use mix the availability of walking and cycling infrastructure and access and availability of walking and cycling infrastruc- the convenience of recreational facilities (Table 4). ture, their female children reported more minutes of ac- tive transport to and from school. Mediated effects of independent mobility on the More walking for transport during leisure time was as- associations between neighborhood environmental sociated with a higher parental perception of residential perceptions and PA (αß-coefficients) density, more land use mix diversity, good land use mix Among girls, the level of independent mobility signifi- access, availability of walking and cycling infrastructure cantly mediated the association of the parental percep- and convenient recreational facilities. When parents per- tion of residential density with active transport to and ceived high neighborhood residential density, good land from school (11.8%) and the total level of PA (43.2%). use mix access, well-maintained and high-quality walk- Furthermore, the level of independent mobility medi- ing and cycling infrastructures and more safety for traf- ated 37.7% of the association between the parental per- fic, their children reached a higher level of total physical ception of land use mix access and total PA. activity. For the number of minutes of cycling for trans- Finally, 39.5% of the association between the parental port during leisure time and the objectively determined perception of safety for traffic and the self-reported total daily number of step counts no associations with the par- level PA was mediated by the level of independent mo- ental perceptions of neighborhood environmental attri- bility (Table 4). butes were found. The level of independent mobility was no significant mediator of the association between the parental percep- Step 1: effects of the neighborhood environmental tion of land use mix access and active transport to and perceptions on the potential mediator (α coefficients) from school. The level of independent mobility was also Girls had higher levels of independent mobility when no significant mediator of the association between the their parents perceived a higher neighborhood residen- parental perception of the availability of walking and tial density, good land use mix access, high availability of cycling infrastructure and active transport to and from walking and cycling infrastructure, more safety for traffic school. and convenient recreational facilities. Among boys, no significant association was found of the parental percep- Discussion tions of residential density, land use mix diversity, dis- The first aim of this study was to examine the associ- tance to school, land use mix access, availability of ation of the parental perception of neighborhood envir- walking and cycling infrastructures and aesthetics with onmental attributes with active transport and the total their level of independent mobility (Table 4). physical activity level among 10–12 year old Belgian boys and girls. The results presented in this paper clearly Step 2: effect of the potential mediator on the dependent indicate that among 10–12 year old Belgian boys and physical activity variables (ß coefficients) girls, the way in which their parents perceive their Among girls, the level of independent mobility showed a neighborhood is related to their level of active transport positive association with the self-reported minutes of ac- and to a lesser extent to their total physical activity level. tive transport to and from school and total physical ac- Specifically, a high degree of residential density, a short tivity after controlling for the parental perception of distance to school, a high degree of land use mix diversity, residential density. Similar results were found after ad- accessible neighborhood services and available walking and justment for the parental perception of land use mix cycling infrastructures were identified as important parental access. perceptions to explain physical activity in boys and girls After adjustment for the parental perception of the of this age group. The existing reviews summarizing the availability of walking and cycling infrastructure, the results of studies examining the association between level of independent mobility was positively associated neighborhood environmental attributes and physical ac- with the self-reported number of active transport to and tivity behavior among children repeatedly concluded from school. Furthermore, after adjustment for the par- that for a lot of neighborhood built environmental attri- ental perception of safety for traffic in the neighborhood, butes no definite conclusions could be drawn with regard the level of independent mobility was positively associ- to their contribution to explain active transport and over- ated with the total physical activity level. all physical activity level [25,26]. However, in these reviews No association was found between the level of inde- no distinction was made between child and parental pendent mobility and the self-reported number of perceptions of neighborhood environmental attributes De Meester et al. BMC Public Health 2014, 14:631 Page 11 of 14 http://www.biomedcentral.com/1471-2458/14/631 to draw conclusions. Probably, parental perceptions of arts). We used non-wearing diaries to register the activ- environmental attributes are more important to explain ities performed without an activity monitor and adjusted active transportation and total physical activity than per- the daily amount of step counts based on the reported ceptions of the children themselves. This is also confirmed number of minutes of non-wear activities. However, the in a previous Belgian study, investigating the association children were not asked to register cycling activities in between children’s perception of neighborhood environ- the non-wearing diaries. Consequently, the registration mental attributes and active transport among 13–15 year of cycling activities for which the activity monitor was olds [57]. The results of this previous study showed that not removed may be inaccurate. As cycling is a common some perceived environmental attributes were found to be form of active transport among Belgian children [58] this important for active transport to school, but the import- can be a possible explanation for the lack of associations ance of the perceived environmental attributes was rather with the objectively determined daily step counts. negligible for active transport during leisure time. Although The second aim of this study was to investigate the the age group of that study was not comparable with the possible mediating role of the level of independent mo- present study age group, it appears that parental percep- bility on the association of the parental perception of tions of environmental attributes might be of greater im- neighborhood environmental attributes with the level of portance than the perceptions of the children themselves, active transport and the total physical activity level. The certainly to explain active transport during leisure time. data of this study showed that about 50% of the children Additional research is certainly needed to confirm this con- in this age group were allowed to walk or cycle between clusion. However, if confirmed, this finding emphasizes the 1 km and 5 km from home without parental accompani- need to incorporate environmental changes but also aware- ment. The studies found in the literature, investigating ness raising programs directed to parents to substantially the level of independent mobility of children, operation- increase active transport and physical activity in children. alized the term independent mobility in different ways. Thus, the results of this study showed that in 10–12 To our knowledge no recent other study operationa- year old boys and girls the parental perceptions of differ- lized the level of independent mobility in terms of the ent neighborhood environmental attributes are associ- geographical distance from youth’shome to places where ated with different forms of active transport and in some they are allowed to wander when playing and socializing cases with their self-reported total physical activity level. [31]. This makes comparison difficult. Consistent with The significant associations found with the self-reported other studies [37,39-41], the results of this study showed a total physical activity level can mostly be explained by marked difference in the level of independent mobility the associations found with the active transport variables between boys and girls. Boys were allowed to walk or as the total physical activity level is a sum of different cycle further from home without supervision than girls types of physical activity including the active transport of the same age. Parents are clearly more protective of variables. However, for some neighborhood environmen- daughters [37]. tal attributes a significant association was found with the The association between the parental perceptions of total physical activity level while no significant associ- neighborhood environmental attributes and the level of ation was found with the different active transport be- independent mobility was different between girls and haviors (i.e. the parental perceptions of the availability of boys. The level of independent mobility of female chil- walking and cycling infrastructure among boys and the dren was higher in neighborhoods that were perceived parental perceptions of safety for traffic and the main- as more activity-friendly or walkable by the parents of tenance/quality of walking and cycling infrastructure the children. A higher degree of residential density in among girls). However, it is possible that some small, the neighborhood, higher land use mix accessibility, high non-significant correlations between the environmental availability of walking and cycling infrastructure, more safety for traffic and convenient recreational facilities characteristics and the different active transport behav- iors exist and accumulate which may give a significant were associated with higher levels of independent mobil- association with the total physical activity level. ity among girls. Among boys, none of the parental per- ceptions of neighborhood environmental attributes that Furthermore, no associations were found between the parental perceptions of neighborhood environmental at- were found to be associated with active transport or the tributes and the objectively determined daily step total physical activity level, showed a significant associ- ation with the level of independent mobility. Thus, out counts. The activity monitors used to determine the daily step counts are able to measure ambulatory move- of these results we can conclude that among boys the ment; however, they sometimes do not detect or under- level of independent mobility is not dependent of how estimate some types of physical activity (e.g. bicycling). parents perceive their direct neighborhood environment Furthermore, activity monitors are often removed to while among girls, the more activity friendly or ‘walk- undertake certain activities (e.g. swimming, marterial able’ the parents perceive their neighborhood, the less De Meester et al. BMC Public Health 2014, 14:631 Page 12 of 14 http://www.biomedcentral.com/1471-2458/14/631 restricted girls are. Comparable studies relating neighbor- activity in girls, through improved neighborhood environ- hood environmental attributes to the level of independent mental attributes, may be more effective if the mobility re- mobility also showed the significance of the neighborhood strictions for girls are targeted as well. However, to our environment for the willingness of parents to allow their knowledge, this is the first time that the possible mediat- children to be active independently [33,35,39,40,59]. These ing role of the level of independent mobility in the associ- studies also found differences between boys and girls. How- ation of the neighborhood environmental attributes with ever, in none of these studies the distinctness was so obvious. children’s level physical activity has been investigated. So, The environmental attributes questioned, are the en- results cannot be compared and these results should first vironmental attributes of the direct neighborhood envir- be confirmed in other studies before drawing definite onment. The parents reported their perception about conclusions. the neighborhood environment within walking distance Strengths of this study included the use of both object- (within a 10–15 minute walk from their homes). Since ive and self-reported measures to assess physical activity. the level of independent mobility for a lot of adolescents Furthermore, different physical activity behaviors related (and especially for boys) is higher, it is possible that the to distinct physical activity contexts were included. This environmental attributes at a further distance than the allowed us to investigate the importance of the parental attributes within the direct neighborhood will contribute perceptions of neighborhood environmental attributes to explaining the level of independent mobility. Future for specific physical activity behaviors in specific con- research should look further into this. texts. This is important for the design of future physical Higher levels of independent mobility were related to activity interventions. Furthermore, this study included a higher levels of active transport to school and higher total relatively large study sample. physical activity levels, but no association was found with Limitations of the present study included the cross- the daily amount of step counts. These findings concur sectional study design, which does not permit to infer a with the conclusions drawn in the existing international causal relationship. Secondly, adolescents were not asked literature. Page et al. (2009) and also Carver et al. (2010) to register cycling activities for which the activity monitor investigated the relation between the level of independent was not removed. As activity monitors are not able to mobility and the level of physical activity among 10–11 register cycling behavior accurately, this has a reflection year old children. In both studies, independent mobility on the daily amount of step counts. Third, step-counts appeared to be an important correlate of the level of active were determined using the Yamax Digi-Walker CW701 transport and physical activity for this age group [37]. and the GT1M accelerometer. Although the step counts Since active commuting has been acknowledged by a num- measured by the Yamax Digi-walker CW-701 have been ber of studies as a physical activity behavior with a signifi- shown to be highly correlated with the step counts of the cant contribution to the total physical activity level, this has GT1M accelerometer, the overall agreement between the important implications for health promotion. Parents may step counts of both monitors is rather low [52]. To over- limit their child’s level of independent mobility and conse- come this problem, all analyses were controlled for the quently restrict their child’sopportunitiestobephysically type of monitor used. Finally, the level of independent active. Encouraging independent mobility may therefore mobility was determined using self-reported data. The use contribute to increasing levels of active transport and as a of objective measures of GIS- and GPS-technologies in consequence the total physical activity level. combination with self-reported data would ascertain the The mediating analyses showed that in boys none of places visited without adult accompaniment. This may be the associations between the parental perceptions of a possible solution for the lack of consensus about the neighborhood environmental attributes and active trans- operationalization of the level of independent mobility port or total physical activity was mediated by independ- which limits the comparability between the results of this study and the results of other studies. ent mobility. Among girls on the other hand, the level of independent mobility mediated the association of the parental perception of residential density with active Conclusion transport to and from school (11.8%) and with the total A first conclusion of this study is that among 10–12 self-reported level of physical activity (43.2%). Further- year old boys and girls, the parental perceptions of more, the associations between the parental perception neighborhood environmental characteristics are asso- of land use mix access and the parental perception of ciated with their level of active transport and total safety for traffic with the total physical activity level were physical activity. This may be important for Belgian mediated respectively for 37.7% and 39.5% by the level of policy makers and urban planners to make well-considered independent mobility. Based on these results, we might decisions concerning built environmental redevelopments assume that independent mobility is of higher importance of existing Belgian neighborhoods and planning of new in girls than in boys. Promoting the level of physical neighborhoods. De Meester et al. BMC Public Health 2014, 14:631 Page 13 of 14 http://www.biomedcentral.com/1471-2458/14/631 Secondly, in girls, the level of independent mobility 12. Salmon J, Timperio A, Cleland V, Venn A: Trends in children's physical activity and weight status in high and low socio-economic status areas of plays a role in the association of the parental perceptions Melbourne, Victoria, 1985–2001. Aust N Z J Public Health 2005, 29:337–342. of neighborhood environmental characteristics with their 13. Mc Donald NC: Active transportation to school: trends among US level of active transport and their total physical activity schoolchildren, 1969–2001. Am J Prev Med 2007, 32:509–516. 14. Van Der Ploeg HP, Merom D, Corpuz G, Bauman AE: Trends in Australian level. When designing programs, focusing on neighbor- children traveling to school 1971–2003: burning petrol or hood environmental characteristics, to promote active carbohydrates? Psychol Aging 2008, 46:60–62. transport or physical activity, it should be kept in mind 15. Dollman J, Norton K, Norton L: Evidence for secular trends in children's physical activity behaviour. Br J Sports Med 2005, 39:892–897. that the level of independence given to girls also needs 16. World Health Organization: Global Stragey on diet, Physical activity and to be considered to obtain behavior change. However, Health. Geneva: Geneva, Zwitzerland: World Health Organization; 2004. before the results of this study can be generalized, these 17. Giles-Corti B, Timperio A, Bull F, Pikora T: Understanding physical activity environmental correlates: increased specificity for ecological models. results should be first confirmed in other studies. Exerc Sport Sci Rev 2005, 33:175–181. 18. Davison KK, Lawson CT: Do attributes in the physical environment Competing interest influence children's physical activity? A review of the literature. Int J The authors declare that they have no competing interests. Behav Nutr Phys Activ 2006, 3:19. 19. Ferreira I, Horst K, Wendel-Vos W, Kremers S, van Lenthe FJ, Brug J: Authors’ contribution Environmental correlates of physical activity in youth - a review and FDM coordinated the data collection, assisted in the recruitment of the update. Obes Rev 2007, 8:129–154. participants, conducted the statistical analyses and drafted the manuscript. 20. Sallis JF, Owen N, Fisher EB: Ecological Models of Health Behaviour. In GC, DVD and IDB participated in the interpretation of the data, revised the Health Behavioour and Health Education: Theory, Research and Practice. 4th draft versions of the manuscript and provided critical comments during the edition. Edited by Glanz K, Rimer BB, Viswanath K. San Francisco: process. All authors read and approved the final version of the manuscript. Jossey-Bass; 2008:462–484. 21. Gebel K, Bauman A, Petticrew M: The Physical Environment and Physical Acknowledgments Activity: A Critical Appraisal of Review Articles. Am J Prev Med 2007, The authors wish to acknowledge the contribution of Coppin Evelyne, Bieke 32:361–369. Moerman, Elien Moerman, De Backer Ann-Sofie, Fauve Vandendorpe, Delrue 22. Heath GW, Brownson RC, Kruger J, Miles R, Powell KE, Ramsey LT, The Task Marlies, Everaert Marloes and Vanden Bergh Tine in the data collection. This Force on Community Preventive Services: The effectiveness of Urban work was supported by a Ghent University Special Research Fund (BOF) Design and Land Use and Transport Policies and Practices to Increase [grant number BOF 08/24J/134]. Physical activity: A Systematic Review. J Phys Act Health 2006, 3(Suppl 1):S55–S76. Author details 23. Owen N, Humpel N, Leslie E, Bauman A, Sallis JF: Understanding Department of Movement and Sport Sciences, Faculty of Medicine and environmental influences on walking; Review and research agenda. Am J Health Sciences, Ghent University, Watersportlaan 2, B-9000 Ghent, Belgium. Prev Med 2004, 27:67–76. Research Foundation Flanders (FWO), Brussels, Belgium. 24. Wendel-Vos W, Droomers M, Kremers S, Brug J, van Lenthe F: Potential environmental determinants of physical activity in adults: a systematic Received: 16 May 2013 Accepted: 12 June 2014 review. Obes Rev 2007, 8:425–440. Published: 20 June 2014 25. Ding D, Sallis JF, Kerr J, Lee S, Rosenberg DE: Neighborhood environment and physical activity among youth: a review. Am J Prev Med 2011, 41:442–455. References 26. Wong BYM, Faulkner G, Buliung R: GIS measured environmental correlates 1. Warburton DE, Nicol CW, Bredin SS: Health benefits of physical activity: of active school transport: A systematic review of 14 studies. Int J Behav the evidence. Can Med Assoc J 2006, 174:801–809. Nutr Phys Activ 2011, 8:39. 2. Strong WB, Malins RM, Blimkie CRJ, Daniels SR, Dishman RK, Gutin B, 27. Panter JR, Jones AP, van Sluijs EM: Environmental determinants of active Hergenroeder AC, Must A, Nixon PA, Pivarnik JM: Evidence based physical travel in youth: A review and framework for future research. Int J Behav activity for school-age youth. J Pediatr 2005, 146:732–737. Nutr Phys Act 2008, 5:34. 3. Steele RM, Brage S, Corder K, Wareham NJ, Ekelund U: Physical activity, 28. Pont K, Ziviani J, Wadley D, Bennett S, Abbott R: Environmental correlates cardiorespiratory fitness, and the metabolic syndrome in youth. J Appl of children's active transportation: a systematic literature review. Health Physiol 2008, 105:342–351. Place 2009, 15:827–840. 4. Jimenez-Pavon D, Kelly J, Reilly JJ: Associations between objectively 29. Giles-Corti B: Encouraging Walking for Transport and Physical Activity in measured habitual physical activity and adiposity in children and Children and Adolescents How Important is the Built Environment? adolescents: Systematic review. Int J Pediatr Obes 2010, 5:3–18. Sports Med 2009, 39:995–1009. 5. Kokkinos P: Physical Activity and Cardiovascular Disease Prevention: 30. Ball K, Jeffery RW, Crawford DA, Roberts RJ, Salmon J, Timperio AF: Current Recommendations. Angiology 2008, 59:26S–29S. Mismatch between perceived and objective measures of physical 6. Vuori I: Physical activity and health: Metabolic and cardiovascular issues. activity environments. Prev Med 2008, 47:294–298. Adv Physiother 2007, 9:50–64. 31. van Vliet W: Exploring the fourth environment. An examination of the home 7. Verloigne M, Van Lippevelde W, Maes L, Yildirim M, Chinapaw M, Manios Y, range of city and suburban teenagers. Environ Behav 1983, 15:567–587. Androutsos O, Kovacs E, Bringolf-Isler B, Brug J, De Bourdeadhuij I: Levels of 32. Fyhri A, Hjorthol R: Children's independent mobility to school, friends and physical activity and sedentary time among 10- to 12-year-old boys and leisure activities. J Transport Geogr 2009, 17:377–384. girls across 5 European countries using accelerometers: an observational 33. Fyhri A, Hjorthol R, Mackett RL, Fotel TN, Kytta M: Children's active travel study within the ENERGY-project. Int J Behav Nutr Phys Act 2012, 9:34. and independent mobility in four countries: Development, social 8. Lee MC, Orenstein MR, Richardson MJ: Systematic review of active contributing trends and measures. Transport Pol 2011, 18:703–710. commuting to school and childrens physical activity and weight. J Phys 34. Hillman M, Adams JGU: Children's Freedom and Safety. Child Youth Act Health 2008, 5:930–949. Environ 1992, 9:12–33. 9. Davison KK, Werder JL, Lawson CT: Children's active commuting to school: 35. Jago R, Thompson JL, Page AS, Brockman R, Cartwright K, Fox KR: current knowledge and future directions. Prev Chronic Dis 2008, 5:A100. Licence to be active: parental concerns and 10-11-year-old children's 10. Faulkner G, Buliung R, Flora PK, Fusco C: Active school transport, physical ability to be independently physically active. J Public Health (Oxf ) activity levels and body weight of children and youth: A systematic 2009, 31:472–477. review. Prev Med 2008, 48:3–8. 11. Sturm R: Childhood Obesity - What we can learn from existing data on 36. Page AS, Cooper AR, Griew P, Jago R: Independent mobility, perceptions secular trends. Prev Chronic Dis 2015, 2:A20. of the built environment and children's participation in play, active De Meester et al. BMC Public Health 2014, 14:631 Page 14 of 14 http://www.biomedcentral.com/1471-2458/14/631 travel and structured exercise and sport: the PEACH Project. Int J Behav Nutr Phys Activ 2010, 7:17. 37. Carver A, Timperio A, Hesketh K, Crawford D: Are children and adolescents less active if parents restrict their physical activity and active transport due to perceived risk? Soc Sci Med 2010, 70:1799–1805. 38. Wen LM, Kite J, Merom D, Rissel C: Time spent playing outdoors after school and its relationship with independent mobility: a cross-sectional survey of children aged 10–12 years in Sydney, Australia. Int J Behav Nutr Phys Activ 2009, 6:15. 39. Page AS, Cooper AR, Griew P, Davis L, Hillsdon M: Independent mobility in relation to weekday and weekend physical activity in children aged 10–11 years: The PEACH Project. Int J Behav Nutr Phys Activ 2009, 6:2. 40. Carver A, Timperio AF, Crawford DA: Young and free? A study of independent mobility among urban and rural dwelling Australian children. J Sci Med Sport 2012, 15:505–510. 41. Hillman M, Adams J, Whitelegg J: One false move: a study of children's independent mobility. London: PSI Publishing; 1990. 42. O'Brien M, Jones D, Sloan D, Rustin M: Children's Independent Spatial Mobility in the Urban Public Realm. Childhood 2000, 7:257–277. 43. De Bourdeaudhuij I, Sallis JF, Saelens BE: Environmental correlates of physical activity in a sample of Belgian adults. Am J Health Promot 2003, 18:83–92. 44. Rosenberg D, Ding D, Sallis JF, Kerr J, Norman GJ, Durant N, Harris SK, Saelens BE: Neighborhood Environment Walkability Scale for Youth (NEWS-Y): Reliability and relationship with physical activity. Prev Med 2009, 49:213–218. 45. Scoring for the Neighborhood Environment Walkability Scale - Youth (NEWS-Y). In [http://www.drjamessallis.sdsu.edu/Documents/NEWS-Yscoring.pdf] 46. Ducheyne F, De Bourdeaudhuij I, Lenoir M, Cardon G: Test-retest Reliability and Validity of a Child and Parental Questionnaire on Specific Determinants of Cycling to School. Pediatr Exerc Sci 2012, 24:289–311. 47. Philippaerts RM, Matton L, Wijndaele K, Balduck AL, De Bourdeaudhuij I, Lefevre J: Validity of a physical activity computer questionnaire in 12- to 18-year-old boys and girls. Int J Sports Med 2006, 27:131–136. 48. Crouter SE, Schneider PL, Karabulut M, Bassett DR Jr: Validity of 10 electronic pedometers for measuring steps, distance, and energy cost. Med Sci Sports Exerc 2003, 35:1455–1460. 49. Silva P, Mota P, Esliger D, Welk G: Technical Reliability Assessment of the Actigraph GT1M Accelerometer. Meas Phys Educ Exerc Sci 2010, 14:79–91. 50. McNamara E, Hudson Z, Taylor SJ: Measuring activity levels of young people: the validity of pedometers. Br Med Bull 2010, 95:121–137. 51. Behrens TK, Dinger MK: Motion sensor reactivity in physically active young adults. Res Q Exerc Sport 2007, 78:1–8. 52. Kinnunen TI, Tennant PWG, McParlin C, Poston L, Robson SC, Bell R: Agreement between pedometer and accelerometer in measuring physical activity in overweight and obese pregnant women. BMC Public Health 2011, 11:501. 53. Miller R, Brown W, Tudor-Locke C: But what about swimming and cycling? How to ‘count’ non-ambulatory activity when using pedometers to assess physical activity. J Phys Act Health 2006, 3:257–266. 54. De Meester F, De Bourdeaudhuij I, Deforche B, Ottevaere C, Cardon G: Measuring physical activity using accelerometry in 13 to 15 year old adolescents: the importance of including non-wear activities. Public Health Nutr 2011, 14:2124–2133. 55. Keene ON: The log transformation is special. Stat Med 1995, 14:811–819. 56. MacKinnon: An introduction to Mediation Analysis. New York, USA: Erlbaum Psych Press; 2008. 57. De Meester F, Van Dyck D, De Bourdeaudhuij I, Deforche B, Cardon G: Does the perception of neighborhood built environmental attributes influence active transport in adolescents? Int J Behav Nutr Phys Activ 2013, 10:38. 58. Cardon G, Maes L, De Bourdeaudhuij I: Bicycling to school during the Submit your next manuscript to BioMed Central transition from childhood into adolescence. A 6 year longitudinal study. and take full advantage of: Pediatr Exerc Sci 2012, 24:369–383. 59. Veitch J, Bagley S, Ball K, Salmon J: Where do children usually play? A qualitative study of parents’ perceptions of influences on children's • Convenient online submission active free-play. Health Place 2006, 12:383–393. • Thorough peer review • No space constraints or color figure charges doi:10.1186/1471-2458-14-631 Cite this article as: De Meester et al.: Parental perceived neighborhood • Immediate publication on acceptance attributes: associations with active transport and physical activity • Inclusion in PubMed, CAS, Scopus and Google Scholar among 10–12 year old children and the mediating role of independent mobility. BMC Public Health 2014 14:631. • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png BMC Public Health Springer Journals

Parental perceived neighborhood attributes: associations with active transport and physical activity among 10–12 year old children and the mediating role of independent mobility

Download PDF
14 pages

Loading next page...
 
/lp/springer-journals/parental-perceived-neighborhood-attributes-associations-with-active-6Tac1zckiC

References (63)

Publisher
Springer Journals
Copyright
2014 De Meester et al.; licensee BioMed Central Ltd.
eISSN
1471-2458
DOI
10.1186/1471-2458-14-631
Publisher site
See Article on Publisher Site

Abstract

Background: During the last decades, the use of active travel modes declined in all age groups. Childhood is a critical time to establish lifelong healthy patterns. To develop effective interventions in this age group, insight in the correlates of health behaviors and the possible mediating factors is necessary. Among children, the role of parents may not be overlooked. Therefore, this study aimed to examine the associations of parental perceptions of neighborhood environmental attributes with active transport and total physical activity in 10–12 year old Belgian boys and girls. Furthermore, this study examined the potential mediating effect of independent mobility on these associations. Methods: In the present study, 736 10–12 year old children and their parents from 44 elementary schools in Flanders, Belgium, participated. The children were asked to wear an activity monitor and to fill in a survey questioning demographic factors and the Flemish Physical Activity Questionnaire. The parents filled in a survey concerning demographic factors, the child’s level of independent mobility and environmental perceptions (Neighborhood Environmental Walkability Scale). Results: Overall, boys reported more active transport when parents perceived more land use mix diversity, shorter distances to school, good land use mix access, higher residential density and less pleasing neighborhood aesthetics. Higher total physical activity levels were reported when parents perceived shorter distances to school and availability of walking/cycling infrastructure. None of the associations was mediated by independent mobility in boys. Girls reported more active transport when parents perceived higher residential density, more land use mix diversity, shorter distances to school, good land use mix access, available walking/cycling infrastructure and convenient recreational facilities. Girls reported higher total physical activity levels when parents perceived high residential density, good land use mix access, well-maintained and high quality walking/cycling infrastructures and more traffic safety. Independent mobility was found to be an important mediator of these associations in girls. Conclusions: Neighborhood environmental interventions to increase children’s active transport and physical activity can be effective when combined with awareness raising programs for parents. Furthermore, among girls encouraging independent mobility may contribute to behavior change. * Correspondence: [email protected] Department of Movement and Sport Sciences, Faculty of Medicine and Health Sciences, Ghent University, Watersportlaan 2, B-9000 Ghent, Belgium Full list of author information is available at the end of the article © 2014 De Meester et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. De Meester et al. BMC Public Health 2014, 14:631 Page 2 of 14 http://www.biomedcentral.com/1471-2458/14/631 Background more attention has been given to active transport to school Regular physical activity during childhood is associated than to other forms of active transport [27,28]. A recent re- with different well-known health benefits, including the view of D’Haeseetal. (unpublished data)showed thatina maintenance of a healthy weight, protective effects on number of studies high neighborhood walkability, absence the development of diabetes mellitus type 2, promotion of freeway crossings, high land use mix accessibility and of bone health as well as prevention of cardiovascular high neighborhood safety were associated with higher levels disease risk factors [1-6]. Despite these health benefits, of active transport to school in children. Nevertheless, no many children do not achieve the public health recom- consistent results were found concerning the association mendations of 60 minutes moderate-to-vigorous phys- between active transport to school and other neighborhood ical activity per day [7]. environmental attributes like: land use mix diversity, street Walking and cycling for transport, also called ‘active connectivity, the availability of walking and cycling in- transport’ have been acknowledged as affordable and frastructure, the presence of sidewalks and cycle tracks, convenient sources of physical activity with a significant altitude differences, aesthetics, crime and traffic safety, rec- contribution to the total physical activity levels [8,9]. reational facilities and the degree of urbanization. Further- The results of the reviews of Davison et al. [9] and more, no definite conclusions could be drawn with regard Faulkner et al. [10] both showed that children who walk to the contribution of neighborhood built environmental or cycle to school have higher overall physical activity attributes to explain active transport during leisure time levels than those who use motorized transport. However, (e.g. to visit friends). during the past decades, there has been a consistent de- Thus, a number of reviews summarized the findings cline in the use of active travel modes [11-15]. As child- from studies on the association of neighborhood envir- hood is a critical to establish lifelong healthy patterns, onmental attributes with physical activity or with active the development of effective interventions to promote transport among children [25,26]. But, the findings de- physical activity and in particular active transport among scribed in these reviews are inconclusive. The inconclu- children has become a public health priority in most de- siveness of these findings can possibly be attributed to veloped countries [16]. the manner in which neighborhood environmental attri- During the last decade, ecological models have been butes were assessed, the nature of the participants and used more frequently than the more individually-oriented the underestimation of the accumulation of influences models to investigate the correlates of children’sphysical that affect physical activity. activity behavior [17-19]. Ecological models emphasize the As stated by Ding et al. [25] and Giles-Corti et al. [29], importance of environmental characteristics, after the in- the inconclusiveness of reviews summarizing the find- dividual and social factors, in explaining physical activity ings from studies on the association of neighborhood behavior [20]. Interventions designed to create environ- environmental attributes with physical activity or with ments that stimulate the targeted behavior change are active transport among children is potentially influenced appealing because they have the potential for having a by the different methods to measure environmental at- sustained impact on population groups rather than tributes. According to Ball et al. (2008), objectively de- short-term impacts on individuals [17,20]. termined built environmental attributes may potentially As a consequence of the shift from individually-oriented be indirectly associated with physical activity, whereas correlates of physical activity towards environmental- perceptions of environmental attributes have a more dir- oriented correlates as opportunity to shape physical activ- ect influence on physical activity [30]. The presence of ity, empirical evidence is accumulating documenting on environmental attributes might therefore not automatic- the role of neighborhood environmental attributes to ex- ally influence the behavior in the absence of awareness plain children’s physical activity behavior. However, in of those attributes [30]. Further, previous studies based on self-reported per- contrast to the consistent findings in adults [21-24], the association between neighborhood environmental attri- ceptions of neighborhood built environmental character- butes and children’s levels of physical activity is rather istics used parental perceptions as well as children’s perceptions. Parents are important gatekeepers and deci- blurred [25,26]. A recent review of Ding et al. (2011) con- cluded that in only 34 percent of the studies among chil- sion makers for their children’s physical activity and dren, investigating the association between perceived built travel behavior. It may therefore be the case that paren- tal perceptions of neighborhood environmental attri- environmental attributes and self-reported physical activity, a significant positive association was found. In the butes are of greater importance than the perceptions of other 66 percent of the studies, no association could be the children themselves. established [25]. Furthermore, as stated by Giles-Corti et al. [29] and Within the existing evidence on the importance of Panter et al. [27] it is likely that the impact of the neigh- neighborhood environmental attributes for active transport, borhood built environmental attributes on children’s De Meester et al. BMC Public Health 2014, 14:631 Page 3 of 14 http://www.biomedcentral.com/1471-2458/14/631 activity level is moderated or mediated by a range of fac- the last year of elementary school. To recruit the children, tors, such as certain characteristics of the children and their 148 schools were randomly selected from all elementary parents. An important overlooked factor for the association schools in East- and West-Flanders and the principals between neighborhood environmental attributes and chil- were contacted by phone. After this first contact, 44 prin- dren’s physical activity behavior might be the independent cipals agreed to let the final year of their school (1 class mobility granted by the parents. Children’s independent group per school) participate (response rate = 29.7%) and mobility can be defined as the geographical distance from gave written consent. The main reasons for refusal were children’s home to places where they are allowed to wander “no time” and “participated already in many other studies when playing and socializing [31]. It is possible that the or projects”. impact of the neighborhood built environmental attributes In total, 976 children and their parents could be reached. on children’s active transport and activity level may vary Both the children and their parents received an informative according to the level of independent mobility of the chil- letter about the study with an invitation to participate. Fi- dren. Different studies have shown that the parental per- nally, the parents of 749 children from 44 class groups ceptions of certain neighborhood environmental attributes agreed to let their child participate in this study and con- are related to the level of independent mobility of chil- sented to be involved in the study (response rate = 76.7%). dren. Neighborhoods characterized by a shorter distance to Data collection took place between September 2010 and school, less traffic danger, proximity to friends, no freeways June 2011. During school hours and under supervision of a and presence of facilities to assist active travel are found to research assistant, the children with written consent to par- be positively associated with the level of independent mo- ticipate completed a questionnaire on socio-demographic bility of children [32-35]. This, in turn, may increase their variables and their own PA behavior. Furthermore, the re- level of active transport as different studies have shown that search assistant explained the protocol of the activity moni- children who encounter less mobility restrictions have tor and its diary as the children were asked to wear an higher levels of active transport. The results of a study of activity monitor for seven consecutive days. Every child was Page et al., conducted in the UK, showed that 10–11 year also given a questionnaire to be completed by one of the old boys and girls with higher levels of independent mobil- parents. The parental questionnaire contained questions ity had higher levels of active transport [36]. Similar results concerning socio-demographic characteristics, neighborhood were foundinanAustralianstudy of Carver et al.inthe environmental perceptions and the level of independent mo- same age group [37]. Furthermore, children’s level of inde- bility of their child. Parents completed this questionnaire at pendent mobility was also found to be positively associated home. One week later, the research assistant visited the with children’s overall outdoor play [38] and accelerometer schools for a second time to collect the parental question- determined weekday physical activity [39]. naires, activity monitors and diaries. The parental question- To our knowledge no previous studies investigated the naires, activity monitors or diaries that were forgotten the mediating role of the level of independent mobility on the second visit, were redirected by the teacher or collected by associations between parental perceptions of neighbor- the research assistant during a third visit. hood environmental attributes and physical activity behav- The study protocol received approval from the Ethics iors such as active transport. When investigating these Committee of Ghent University Hospital. indirect pathways, it is important to take into account the gender of the children as the level of independent mobility Measures is found to be higher in boys compared to girls [39-42]. Demographic characteristics The first aim of the present study was to examine the Children’s age and gender was questioned in the child association of the parental perceptions of neighborhood questionnaire. In the parental questionnaire, the parents environmental attributes with active transport and total were asked to fill in their own and their partner’s level of physical activity in 10–12 year old Belgian boys and girls. education. Educational attainment of the children’s par- Secondly, this study examined the potential mediating ents was used as a proxy measure of children’sSES.The effect of independent mobility on the associations of the educational level of the child’s mother and father was parental perceptions of neighborhood environmental at- determined based on four options: less than high school, tributes with the level of active transport and total phys- completed high school, completed college or completed ical activity of 10–12 year old boys and girls. university. The educational level of mother and father was coded into ‘reached a college or a university level’ or Methods ‘did not reach a college or a university education level’. Participants and procedure In the present study, a large sample of 10–12 year old Neighborhood built environmental attributes children from 44 elementary schools in East- and West- To measure the parental perceptions of neighborhood Flanders, Belgium, participated. All children attended built environmental attributes, the parent version of the De Meester et al. BMC Public Health 2014, 14:631 Page 4 of 14 http://www.biomedcentral.com/1471-2458/14/631 Neighborhood Environmental Walkability Scale for Independent mobility Youth (NEWS-Y) was used [43]. The NEWS-Y has accept- To determine children’s level of independent mobility, able to good test-retest reliability with an intra-class correl- two items were included in the parent questionnaire ation coefficient between 0.56 and 0.87 [44] for parents of “How far is your child allowed to cycle from home with- children between 5 and 10 years. The neighborhood built out adult accompaniment?” and “How far is your child environmental attributes questioned were residential allowed to walk from home without adult accompani- density, land use mix diversity, distance to school, land ment?”. Response categories were: not, 0 m-200 m, use mix access, street network connectivity, availability 200 m-500 m, 500 m-1 km, 1 km-3 km, 3 km-5 km, of walking and cycling infrastructure, maintenance and 5 km-10 km, +10 km. The question “How far is your quality of walking and cycling infrastructures, aesthetics of child allowed to cycle from home without adult accom- the neighborhood, convenience of recreational facilities paniment?” was found to have good reliability with an and crime and traffic safety. The neighborhood was de- ICC of 0.79 [46]. The second question was informed by fined as the immediate environment around the house this question. The highest score on these two items was within a distance of 1 kilometer (10 – 15 minutes walk- taken as an indication of how far they were allowed to ing distance). The NEWS-Y scoring guidelines were walk or cycle from home without adult accompaniment. used to calculate the subscales. Following the NEWS-Y scoring guidelines [45], the subscale residential density Physical activity was computed by the following formula: score on The Flemish Physical Activity Questionnaire (FPAQ) question 1a (single family residences) + 12*score on was used to determine the duration (hours and minutes question 1b (row houses) + 25*score on question 1c per day) of school related active transportation (walking (apartments). The other subscales were calculated by and cycling to and from school), walking and cycling for taking the mean of the different item scores. The con- transport during leisure time and total physical activity tent, response options and descriptive statistics are level. The total physical activity level [47] was calculated giveninTable 1. using the minutes of all activities questioned in the FPAQ Table 1 Content, response options and descriptive statistics of the parental perceived neighborhood built environmental attributes Content of the item Response Mean Mean category score score boys girls Residential density (3 items) Presence of different types of residences (e.g. detached single 5-point 78.8 (26.9) 79.5 (26.3) family residences, row houses, apartments) scale Land use mix diversity (9 items) Distance to local facilities (e.g. supermarket, post office, library) 5-point 3.4 (0.9) 3.4 (0.9) scale Land use mix access (5 items) Access to neighborhood services for their child (e.g. ease to walk to 5-point 3.6 (1.1) 3.6 (1.0) public transport, ease to walk to school) scale Distance to school (1item) Distance to the school of the adolescent 5-point 2.9 (1.4) 3.0 (1.4) scale Connectivity (3 items) Connectedness of street network (e.g. presence of intersections, 5-point 3.3 (0.7) 3.3 (0.7) dead-end streets, alternate routes) scale Availability of walking and cycling Availability of walking and cycling infrastructure (e.g. footpaths and cycling 5-point 2.7 (0.9) 2.8 (0.9) infrastructure (4 items) lanes in most streets, footpaths and cycling lanes separated from streets ) scale Quality and maintenance of walking Quality and maintenance of walking and cycling infrastructure (e.g. 5-point 3.1 (1.0) 3.1 (0.9) and cycling infrastructure (5 items) maintenance of cycling lanes and footpaths, presence of lighting) scale Aesthetics (4 items) Presence of aesthetic features (e.g. green spaces, attractive buildings, 5-point 3.6 (0.8) 3.5 (0.7) streets free from litter and graffiti) scale Safety for traffic (6 items) Perceived safety from traffic problems (e.g. speed of traffic in neighborhood, 5-point 2.9 (0.7) 2.8 (0.7) availability of pedestrian crossings and traffic signals) scale Safety for crime (4 items) Perceived safety from crime (e.g. crime prevalence in the neighborhood, 5-point 3.5 (0.8) 3.4 (0.7) perceived safety from strangers) scale Convenience of recreation facilities Distance to PA facilities (e.g. sports field, sports hall, swimming pool, park) 5-point 3.5 (0.9) 3.4 (0.9) (5 items) scale PA: physical activity. none, a few, about half, a lot, all. > 30 min, 21–30 min, 11–20 min, 6–10 min, 1–5 min. strongly disagree, somewhat disagree, neither agree or disagree, somewhat agree, strongly agree. Note: all perceived built environmental attributes were positively scored: higher score = more walkable. De Meester et al. BMC Public Health 2014, 14:631 Page 5 of 14 http://www.biomedcentral.com/1471-2458/14/631 (walking and cycling to and from school, walking and cyc- 4.1.0). For every minute of reported moderate or vigor- ling for transport during leisure time, school-related sport- ous physical activity for which the activity monitor was ing activities and leisure time sporting activities). The removed, 150 steps were added to the daily number of FPAQ was found to be a reliable and reasonably valid reported step counts [53]. Among children, the import- questionnaire for the assessment of these different dimen- ance of including non-wear activities registered in diar- sions of physical activity in youth [47]. ies when using activity monitors has been demonstrated To measure free-living step counts, the children were previously [54]. asked to wear an activity monitor for 7 consecutive days, including two weekend days. The Yamax Digiwalker Data analyses SW-200 (Yamax cooperation, Tokio, Japan) and the Preliminary analyses were conducted to obtain descrip- Actigraph accelerometer, model GT1M (Actigraph MTI, tive information about the demographic characteristics Manufacturing Technology Inc., Pensacola, FL, USA) of the study sample and the study variables (children’s were used. The Yamax Digiwalker has been acknowl- level of independent mobility, self-reported minutes per edged as a valid, accurate and reliable instrument to day of active transport, total physical activity level and measure free-living step-counts [48]. The GT1M acceler- objectively (pedometer or accelerometer) determined ometer has demonstrated good reliability for measuring daily number of steps) using SPSS 17.0. steps [49]. Evidence exists that neither accelerometers Tests for normal distribution revealed some skewed nor pedometers are affected by reactivity among adoles- physical activity variables (self-reported school related cents [50,51]. Although the step counts measured by the active transportation, walking and cycling for transport Yamax Digi-walker CW-701 (the update of the Yamax during leisure time, total physical activity level). To ob- Digiwalker SW-200) have been shown to be highly cor- tain distributions that more closely approximated sym- related with the step counts of the GT1M accelerometer metry, logarithmic transformations of these variables (r = 0.78), the overall agreement between the step counts were conducted, and the transformed variables were of both monitors is rather low [52]. In the study of Kinnu- used in the analyses [55]. For ease of interpretation, nen et al. (2011), the 95% limits of agreement ranged be- summary data of untransformed physical activity vari- tween −2690 to 2656 steps/day for the mean value (mean ables are reported in minutes/day (Table 2). of accelerometer and pedometer steps/day = 6026). Further, Furthermore, all analyses were done separately for the limits of agreement varied substantially over the range boys and girls given the existing evidence that boys’ level of values. At the lowest recorded step count (mean of acce- of independent mobility is higher than girls’ level of in- lerometer and pedometer steps/day = 906) the accelero- dependent mobility [36,39-42]. meter was on average recording more steps/day than the To examine the associations between the parental per- pedometer. In contrast, at the highest step count value ceptions of neighborhood environmental attributes and (mean of accelerometer and pedometer steps/day = 12,018) self-reported minutes/day of active transport to and the accelerometer recorded less steps/day than the pedom- from school, walking for transport during leisure time, eter on average [52]. To overcome this problem, all analyses cycling for transport during leisure time, total physical were controlled for the type of monitor used. activity level and objectively determined number of steps/ The children were asked to wear the activity monitor day, multiple linear regression analyses were conducted (pedometer or accelerometer) during waking hours but using MLwin version 2.22. The dependent variables were to remove the monitor for aquatic activities and for ac- self-reported minutes/day of active transport to and from tivities that prohibit activity monitors. Together with school, walking for transport during leisure time, cycling their activity monitor, all children received a diary. Chil- for transport during leisure time, total physical activity level dren who wore a pedometer were asked to record the and objectively determined number of steps/day. These lin- dates, steps taken at the end of the day and a description ear regression analyses provided τ-coefficients (Figure 1). and duration of the activities for which the pedometer To investigate if children’s level of independent mobil- was removed in the diary. The children who wore an ac- ity mediated the association of neighborhood environ- celerometer were asked to record the duration and a de- mental perceptions with their level of physical activity, scription of the activities for which the accelerometer the product-of-coefficient test of McKinnon et al. was was removed. The children who wore an accelerometer used (Figure 1) [56]. This test included three steps that were not asked to record the steps taken at the end of the were conducted in MLwin version 2.22. In a first step, the day because accelerometers have a memory and are able to effects of the neighborhood environmental perceptions on save the steps taken during the entire recording period. At the potential mediator (children’s independent mobility) the end of the measurement period, the activity monitors were examined by regressing children’s independent mo- were collected and the accelerometer determined step bility onto the neighborhood environmental perceptions. count data were downloaded (Actilife software version Step one was conducted only for the parental perceptions De Meester et al. BMC Public Health 2014, 14:631 Page 6 of 14 http://www.biomedcentral.com/1471-2458/14/631 Table 2 Demographic characteristics and descriptive statistics for adolescents’ level of independent mobility and physical activity behavior Total sample Girls Boys Age: mean (SD) 11.2 (0.5) 11.1 (0.5) 11.2 (0.6) Weight status: % Overweight 12.0 14.2 10.0 Obese 5.0 5.1 5.0 Gender: % Male 51.9 Female 48.1 Educational level: % Mother: No college/university degree 45.6 48.2 43.2 College or university degree 54.4 51.8 56.8 Father No college/university degree 54.2 55.1 53.4 College or university degree 45.8 44.9 46.6 Self-reported physical activity (min/day (SD)) Active transport to and from school 10.2 (12.7) 9.3 (11.8) 11.2 (13.5) Walking for transport during leisure time 9.0 (12.4) 7.8 (11.2) 10.2 (13.3) Cycling for transport during leisure time 10.5 (13.1) 8.1 (11.2) 12.8 (14.3) Total physical activity 81.3 (43.1) 68.7 (35.4) 93.2 (46.3) Daily step counts (steps/day (SD)) 10 766 (3503) 9739 (2979) 11 740 (3684) Level of independent mobility (%) Not 3.5 5.7 1.4 0m – 200 m 1.3 1.8 0.8 200 m – 500 m 8.1 7.5 8.7 500 m – 1 km 17.7 20.9 14.8 1km – 3 km 32.6 31.6 33.5 3km – 5 km 22.4 20.9 23.7 5km – 10 km 10.0 7.8 12.0 +10 km 4.5 3.9 5.0 that showed a significant direct association with the chil- calculated by dividing the product-of-coefficient (αß) by the dren’s level of active transport, total physical activity level total main effect of the neighborhood environmental per- and objectively determined daily number of step counts. ceptions on the dependent physical activity variables (τ). This step provided estimates of the α-coefficients. For all analyses, 95% confidence intervals (CI) were re- In the second step, the independent effect of the poten- ported. All analyses were controlled for two proxy mea- tial mediator on the dependent physical activity variables sures of individual SES (educational attainment of mother was investigated by regressing the dependent physical ac- and father). Clustering of individuals in schools was taken tivity variables onto the neighborhood environmental per- into account by using multi-level modeling with children ceptions and the potential mediator (children’s level of at the first level and schools at the second level. All ana- mobility) (ß-coefficients). Step two was only conducted lyses including the daily step counts as dependent variable for the parental perceptions that showed a significant as- were controlled for the type of monitor (pedometer or sociation with the potential mediator. accelerometer). In the third step, the mediated effect was calculated by multiplying the two coefficients (αß). Results The statistical significance of the mediated effect was cal- Sample characteristics culated by dividing the product-of-coefficients (αß) by its Table 2 provides an overview of the descriptive statistics standard error. Moreover, the proportion mediated was for demographic characteristics, physical activity behavior De Meester et al. BMC Public Health 2014, 14:631 Page 7 of 14 http://www.biomedcentral.com/1471-2458/14/631 Figure 1 Path diagram for the mediational effect of children’s independent mobility on the association of neighborhood environmental perceptions with the children’s level of physical activity. and independent mobility for the total sample and by gen- Main effects of neighborhood environmental perceptions der. In total, 736 (98.3%) children completed the question- on active transport, the total physical activity level and naire and handed it back to the research assistant. From the daily amount of step counts (τ-coefficients) 93.5% of the parents (n = 701), a complete questionnaire Table 3 shows the main effects of neighborhood environ- returned to school. In 92.1% the children and the parent, mental perceptions on active transport, the total physical who filled in the questionnaire, had the same address. In activity level and the daily amount of step counts among 6.3% the children lived partly with the parent who filled in boys and girls. the questionnaire. In 1.6% the child did not live with the The results of the regression analyses revealed that person who filled in the questionnaire. boys reported more minutes of active transport to and In total, 649 children (86.6%) had complete step count from school when their parents perceived more land use data, 370 children wore a pedometer and 279 an acceler- mix diversity, a shorter distance to the school of their ometer. The sample consisted of 354 girls (48.1%) and child and good land use mix access. 382 boys (51.9%). Mean age was 11.2 ± 0.5 years. The When parents perceived a higher residential density, children reported on average 10.2 (±12.7) minutes/day more land use mix diversity, a shorter distance to school of walking and cycling to and from school, 9.0 (±12.4) and less pleasing neighborhood aesthetics, their male minutes/day of walking during leisure time, 10.5 (±13.1) children reported more walking for transport during leis- minutes/day of cycling during leisure time and 81.3 ure time. Furthermore, the parental perception of a (±43.1) minutes/day of physical activity. On average, the shorter distance to school was associated with more cyc- children took 10 766 (±3503) steps/day. ling for transport during leisure time and the parental Most of the boys and girls were allowed to go between perception of a shorter distance to school and availabil- 500 m and 5 km from home unsupervised. A significant ity of walking and cycling infrastructure were associated difference in independent mobility was found between with more self-reported total physical activity. No asso- boys and girls (t = 3.438; p < 0.01). Boys were allowed to ciations were found between the parental neighborhood go further from home unsupervised using active trans- environmental perceptions and the objectively determined port compared to girls. daily number of step counts. De Meester et al. BMC Public Health 2014, 14:631 Page 8 of 14 http://www.biomedcentral.com/1471-2458/14/631 Table 3 Main effects of neighborhood environmental perceptions on physical activity variables among boys and girls (τ-coefficients) Active transport to Walking for transport Cycling for transport Overall level of Daily number of step counts and from school during leisure time leisure during leisure time leisure physical activity MAIN EFFECTS AMONG BOYS τ (SE) 95% CI τ (SE) 95% CI τ (SE) 95% CI τ (SE) 95% CI τ (SE) 95% CI Residential density 0.001 (0.001) −0.001-0.003 0.004 (0.001) 0.002-0.006 0.002 (0.001) 0.000-0.004 0.000 (0.001) −0.002-0.002 7.322 (8.766) −9.859-24.503 Land use mix diversity 0.108 (0.036) 0.037-0.179 0.076 (0.039) 0.000-0.152 0.050 (0.040) −0.028-0.128 0.005 (0.015) −0.024-0.034 −4.615 (226.984) −449.504-440.274 Distance to school 0.129 (0.023) 0.084-0.174 0.065 (0.025) 0.016-0.114 0.062 (0.026) 0.011-0.113 0.026 (0.010) 0.006-0.046 −23.777 (149.181) −316.172-268.618 Land use mix access 0.185 (0.030) 0.126-0.244 0.061 (0.033) −0.004-0.126 0.036 (0.034) −0.031-0.103 0.022 (0.013) −0.003-0.047 147.627 (195.926) −236.388-531.642 Connectivity 0.091 (0.049) −0.005-0.187 0.015 (0.052) −0.087-0.117 0.010 (0.053) −0..094-0.114 0.012 (0.020) −0.027-0.051 423.100 (312.254) −206.558-1052.758 Availability of walking and cycling 0.070 (0.037) −0.003-0.143 −0.020 (0.039) −0.096-0.056 0.047 (0.040) −0.031-0.125 0.033 (0.015) 0.004-0.062 72.449 (233.694) −385.591-530.489 infrastructure Maintenance and quality of 0.035 (0.033) −0.030-0.100 −0.021 (0.034) −0.088-0.046 0.032 (0.035) −0.037-0.101 0.015 (0.013) −0.010-0.040 36.869 (202.103) −359.253-432.991 walking and cycling infrastructure Aesthetics −0.064 (0.047) −0.156-0.028 −0.116 (0.048) −0.210-(−0.022) −0.049 (0.050) −0.147-0.049 −0.012 (0.019) −0.049-0.025 −214.570 (292.276) −787.431-358.291 Safety for traffic 0.055 (0.051) −0.045-0.155 0.029 (0.054) −0.077-0.135 0.051 (0.056) −0.059-0.161 −0.010 (0.021) −0.051-0.031 −11.255 (316.694) −631.975-609.465 Safety for crime −0.010 (0.046) −0.100-0.080 −0.076 (0.047) −0.168-0.016 −0.028 (0.049) −0.124-0.068 −0.012 (0.019) −0.049-0.025 −205.080 (287.894) 769.352-359.192 Convenience of recreational 0.050 (0.040) −0.028-0.128 −0.011 (0.042) −0.093-0.071 −0.022 (0.043) −0.106-0.062 −0.006 (0.017) −0.039-0.027 150.952 (250.034) −339.115-641.019 facilities MAIN EFFECTS AMONG GIRLS τ (SE) 95% CI τ (SE) 95% CI τ (SE) 95% CI τ (SE) 95% CI τ (SE) 95% CI Residential density 0.004 (0.001) 0.002-0.006 0.005 (0.001) 0.003-0.007 0.001 (0.001) −0.001-0.003 0.001 (0.000) 0.001-0.001 6.298 (7.407) −8.220-20.816 Land use mix diversity 0.089 (0.037) 0.016-0.162 0.102 (0.039) 0.026-0.178 −0.001 (0.037) −0.074-0.072 0.026 (0.015) −0.003-0.055 21.472 (181.128) −333.539-376.483 Distance to school 0.126 (0.024) 0.079-0.173 0.046 (0.026) −0.005-0.097 −0.007 (0.025) −0.056-0.042 0.015 (0.010) −0.005-0.035 −84.890 (122.967) −325.905-156.125 Land use mix access 0.143 (0.032) 0.080-0.206 0.129 (0.035) 0.060-0.198 −0.015 (0.034) −0.082-0.052 0.035 (0.014) 0.008-0.062 286.596 (167.068) −40.857-614.049 Connectivity 0.008 (0.050) −0.090-0.106 0.019 (0.054) −0.087-0.125 0.013 (0.051) −0.087-0.113 −0.024 (0.021) −0.065-0.017 182.826 (248.738) −304.700-670.352 Availability of walking and 0.101 (0.039) 0.025-0.177 0.086 (0.042) 0.004-0.168 −0.000 (0.040) −0.079-0.078 0.004 (0.017) −0.029-0.037 −143.455 (196.440) −529.477-241.567 cycling infrastructure Maintenance and quality of 0.025 (0.037) −0.048-0.098 0.050 (0.039) −0.026-0.126 0.020 (0.037) −0.053-0.093 0.042 (0.015) 0.013-0.071 110.504 (185.024) −252.143-473.151 walking and cycling infrastructure Aesthetics −0.089 (0.050) −0.187-0.009 −0.077 (0.053) −0.181-0.027 −0.043(0.050) −0.141-0.055 0.012 (0.021) −0.029-0.053 −67.354 (250.140) −557.628-422.920 Safety for traffic −0.002 (0.050) −0.100-0.096 0.078 (0.054) −0.028-0.184 0.060 (0.051) −0.040-0.160 0.053 (0.021) 0.012-0.094 139.887 (253.962) −357.879-637.653 Safety for crime 0.039 (0.052) −0.063-0.141 −0.094 (0.055) −0.202-0.014 0.041 (0.052) −0.061-0.143 0.023 (0.022) −0.020-0.066 −44.984 (257.382) −549.453-459.485 Convenience of recreational 0.059 (0.036) −0.012-0.130 0.108 (0.039) 0.032-0.184 −0.028 (0.037) −0.101-0.045 0.022 (0.015) −0.007-0.051 25.069 (184.481) −336.514-386.652 facilities Note: All regression models were adjusted for educational attainment of mother and father and clustering by schools. All regression models including the daily step counts as dependent variable were controlled for type of monitor (pedometer or accelerometer). The significant effects were shown in bold. De Meester et al. BMC Public Health 2014, 14:631 Page 9 of 14 http://www.biomedcentral.com/1471-2458/14/631 Table 4 Regression analyses for the possible mediated effects of independent mobility among boys and girls BOYS: potential mediator: independent mobility α (SE) 95% CI B (SE) 95% CI αß (SE) 95% CI Proportion Mediated (%) Residential density −0.002 (0.003) −0.008 – 0.004 Land use mix diversity −0.128 (0.081) −0.287 – 0.031 Distance to school −0.045 (0.053) −0.149 – 0.059 Land use mix access 0.076 (0.071) −0.063 – 0.215 Availability of walking and cycling infrastructure −0.083 (0.083) −0.246 – 0.080 Aesthetics 0.078 (0.104) −0.126 – 0.282 GIRLS: potential mediator: independent mobility α (SE) 95% CI B (SE) 95% CI αß (SE) 95% CI Proportion Mediated (%) Residential density 0.008(0.003) 0.002 – 0.014 Active transport to and from school 0.059(0.024) 0.012 – 0.106 0.000 (0.000) 0.000 – 0.001 11.8% Walking for transport during leisure time leisure 0.038 (0.026) −0.013 – 0.089 Total physical activity level 0.054 (0.010) 0.034 – 0.074 0.000 (0.000) 0.000 – 0.001 43.2% Land use mix diversity 0.050(0.092) −0.130 – 0.230 Distance to school −0.011(0.061) −0.131 -0.109 Land use mix access 0.250(0.080) 0.093 – 0.407 Active transport to and from school 0.047(0.022) 0.004 – 0.090 0.012 (0.007) −0.001 – 0.0025 Walking for transport during leisure time leisure 0.028 (0.025) −0.021 – 0.077 Total physical activity level 0.050 (0.009) 0.032 – 0.068 0.013 (0.005) 0.004 – 0.023 37.7% Availability of walking and cycling infrastructure 0.200(0.099) 0.006 – 0.394 Active transport to and from school 0.057(0.023) 0.012 – 0.102 0.011 (0.007) −0.003 – 0.026 Walking for transport during leisure time leisure 0.035 (0.025) −0.014 – 0.084 Maintenance and quality of walking and cycling infrastructure 0.108(0.091) −0.070 – 0.286 Safety for traffic 0.411(0.123) 0.170 – 0.652 Total physical activity level 0.051 (0.009) 0.033 – 0.069 0.021 (0.007) 0.007 – 0.035 39.5% Convenience of recreational facilities 0.205(0.091) 0.027 – 0.383 Walking for transport during leisure time leisure 0.037 (0.025) −0.012 – 0.085 SE, standard error; CI confidence interval. Notes: α-coefficients were estimated by regressing children’s level of independent mobility onto the neighbourhood environmental perceptions. ß-coefficients were estimated by regressing the dependent active transport and physical activity variables onto the neighbourhood environmental perceptions and children’s level of independent mobility. αß-coefficients represent the mediated effect. All regression models were adjusted for educational attainment of mother and father and clustering by schools. All regression models including the daily step counts as dependent variable were controlled for type of monitor (pedometer or accelerometer). The significant effects were shown in bold. De Meester et al. BMC Public Health 2014, 14:631 Page 10 of 14 http://www.biomedcentral.com/1471-2458/14/631 Among girls (Table 3), the regression analyses revealed minutes of walking for transport during leisure time that when parents perceived a higher neighborhood resi- after taking into account the parental perceptions of the dential density, more land use mix diversity, a shorter neighborhood residential density, land use mix access, distance to the school of their child, good land use mix the availability of walking and cycling infrastructure and access and availability of walking and cycling infrastruc- the convenience of recreational facilities (Table 4). ture, their female children reported more minutes of ac- tive transport to and from school. Mediated effects of independent mobility on the More walking for transport during leisure time was as- associations between neighborhood environmental sociated with a higher parental perception of residential perceptions and PA (αß-coefficients) density, more land use mix diversity, good land use mix Among girls, the level of independent mobility signifi- access, availability of walking and cycling infrastructure cantly mediated the association of the parental percep- and convenient recreational facilities. When parents per- tion of residential density with active transport to and ceived high neighborhood residential density, good land from school (11.8%) and the total level of PA (43.2%). use mix access, well-maintained and high-quality walk- Furthermore, the level of independent mobility medi- ing and cycling infrastructures and more safety for traf- ated 37.7% of the association between the parental per- fic, their children reached a higher level of total physical ception of land use mix access and total PA. activity. For the number of minutes of cycling for trans- Finally, 39.5% of the association between the parental port during leisure time and the objectively determined perception of safety for traffic and the self-reported total daily number of step counts no associations with the par- level PA was mediated by the level of independent mo- ental perceptions of neighborhood environmental attri- bility (Table 4). butes were found. The level of independent mobility was no significant mediator of the association between the parental percep- Step 1: effects of the neighborhood environmental tion of land use mix access and active transport to and perceptions on the potential mediator (α coefficients) from school. The level of independent mobility was also Girls had higher levels of independent mobility when no significant mediator of the association between the their parents perceived a higher neighborhood residen- parental perception of the availability of walking and tial density, good land use mix access, high availability of cycling infrastructure and active transport to and from walking and cycling infrastructure, more safety for traffic school. and convenient recreational facilities. Among boys, no significant association was found of the parental percep- Discussion tions of residential density, land use mix diversity, dis- The first aim of this study was to examine the associ- tance to school, land use mix access, availability of ation of the parental perception of neighborhood envir- walking and cycling infrastructures and aesthetics with onmental attributes with active transport and the total their level of independent mobility (Table 4). physical activity level among 10–12 year old Belgian boys and girls. The results presented in this paper clearly Step 2: effect of the potential mediator on the dependent indicate that among 10–12 year old Belgian boys and physical activity variables (ß coefficients) girls, the way in which their parents perceive their Among girls, the level of independent mobility showed a neighborhood is related to their level of active transport positive association with the self-reported minutes of ac- and to a lesser extent to their total physical activity level. tive transport to and from school and total physical ac- Specifically, a high degree of residential density, a short tivity after controlling for the parental perception of distance to school, a high degree of land use mix diversity, residential density. Similar results were found after ad- accessible neighborhood services and available walking and justment for the parental perception of land use mix cycling infrastructures were identified as important parental access. perceptions to explain physical activity in boys and girls After adjustment for the parental perception of the of this age group. The existing reviews summarizing the availability of walking and cycling infrastructure, the results of studies examining the association between level of independent mobility was positively associated neighborhood environmental attributes and physical ac- with the self-reported number of active transport to and tivity behavior among children repeatedly concluded from school. Furthermore, after adjustment for the par- that for a lot of neighborhood built environmental attri- ental perception of safety for traffic in the neighborhood, butes no definite conclusions could be drawn with regard the level of independent mobility was positively associ- to their contribution to explain active transport and over- ated with the total physical activity level. all physical activity level [25,26]. However, in these reviews No association was found between the level of inde- no distinction was made between child and parental pendent mobility and the self-reported number of perceptions of neighborhood environmental attributes De Meester et al. BMC Public Health 2014, 14:631 Page 11 of 14 http://www.biomedcentral.com/1471-2458/14/631 to draw conclusions. Probably, parental perceptions of arts). We used non-wearing diaries to register the activ- environmental attributes are more important to explain ities performed without an activity monitor and adjusted active transportation and total physical activity than per- the daily amount of step counts based on the reported ceptions of the children themselves. This is also confirmed number of minutes of non-wear activities. However, the in a previous Belgian study, investigating the association children were not asked to register cycling activities in between children’s perception of neighborhood environ- the non-wearing diaries. Consequently, the registration mental attributes and active transport among 13–15 year of cycling activities for which the activity monitor was olds [57]. The results of this previous study showed that not removed may be inaccurate. As cycling is a common some perceived environmental attributes were found to be form of active transport among Belgian children [58] this important for active transport to school, but the import- can be a possible explanation for the lack of associations ance of the perceived environmental attributes was rather with the objectively determined daily step counts. negligible for active transport during leisure time. Although The second aim of this study was to investigate the the age group of that study was not comparable with the possible mediating role of the level of independent mo- present study age group, it appears that parental percep- bility on the association of the parental perception of tions of environmental attributes might be of greater im- neighborhood environmental attributes with the level of portance than the perceptions of the children themselves, active transport and the total physical activity level. The certainly to explain active transport during leisure time. data of this study showed that about 50% of the children Additional research is certainly needed to confirm this con- in this age group were allowed to walk or cycle between clusion. However, if confirmed, this finding emphasizes the 1 km and 5 km from home without parental accompani- need to incorporate environmental changes but also aware- ment. The studies found in the literature, investigating ness raising programs directed to parents to substantially the level of independent mobility of children, operation- increase active transport and physical activity in children. alized the term independent mobility in different ways. Thus, the results of this study showed that in 10–12 To our knowledge no recent other study operationa- year old boys and girls the parental perceptions of differ- lized the level of independent mobility in terms of the ent neighborhood environmental attributes are associ- geographical distance from youth’shome to places where ated with different forms of active transport and in some they are allowed to wander when playing and socializing cases with their self-reported total physical activity level. [31]. This makes comparison difficult. Consistent with The significant associations found with the self-reported other studies [37,39-41], the results of this study showed a total physical activity level can mostly be explained by marked difference in the level of independent mobility the associations found with the active transport variables between boys and girls. Boys were allowed to walk or as the total physical activity level is a sum of different cycle further from home without supervision than girls types of physical activity including the active transport of the same age. Parents are clearly more protective of variables. However, for some neighborhood environmen- daughters [37]. tal attributes a significant association was found with the The association between the parental perceptions of total physical activity level while no significant associ- neighborhood environmental attributes and the level of ation was found with the different active transport be- independent mobility was different between girls and haviors (i.e. the parental perceptions of the availability of boys. The level of independent mobility of female chil- walking and cycling infrastructure among boys and the dren was higher in neighborhoods that were perceived parental perceptions of safety for traffic and the main- as more activity-friendly or walkable by the parents of tenance/quality of walking and cycling infrastructure the children. A higher degree of residential density in among girls). However, it is possible that some small, the neighborhood, higher land use mix accessibility, high non-significant correlations between the environmental availability of walking and cycling infrastructure, more safety for traffic and convenient recreational facilities characteristics and the different active transport behav- iors exist and accumulate which may give a significant were associated with higher levels of independent mobil- association with the total physical activity level. ity among girls. Among boys, none of the parental per- ceptions of neighborhood environmental attributes that Furthermore, no associations were found between the parental perceptions of neighborhood environmental at- were found to be associated with active transport or the tributes and the objectively determined daily step total physical activity level, showed a significant associ- ation with the level of independent mobility. Thus, out counts. The activity monitors used to determine the daily step counts are able to measure ambulatory move- of these results we can conclude that among boys the ment; however, they sometimes do not detect or under- level of independent mobility is not dependent of how estimate some types of physical activity (e.g. bicycling). parents perceive their direct neighborhood environment Furthermore, activity monitors are often removed to while among girls, the more activity friendly or ‘walk- undertake certain activities (e.g. swimming, marterial able’ the parents perceive their neighborhood, the less De Meester et al. BMC Public Health 2014, 14:631 Page 12 of 14 http://www.biomedcentral.com/1471-2458/14/631 restricted girls are. Comparable studies relating neighbor- activity in girls, through improved neighborhood environ- hood environmental attributes to the level of independent mental attributes, may be more effective if the mobility re- mobility also showed the significance of the neighborhood strictions for girls are targeted as well. However, to our environment for the willingness of parents to allow their knowledge, this is the first time that the possible mediat- children to be active independently [33,35,39,40,59]. These ing role of the level of independent mobility in the associ- studies also found differences between boys and girls. How- ation of the neighborhood environmental attributes with ever, in none of these studies the distinctness was so obvious. children’s level physical activity has been investigated. So, The environmental attributes questioned, are the en- results cannot be compared and these results should first vironmental attributes of the direct neighborhood envir- be confirmed in other studies before drawing definite onment. The parents reported their perception about conclusions. the neighborhood environment within walking distance Strengths of this study included the use of both object- (within a 10–15 minute walk from their homes). Since ive and self-reported measures to assess physical activity. the level of independent mobility for a lot of adolescents Furthermore, different physical activity behaviors related (and especially for boys) is higher, it is possible that the to distinct physical activity contexts were included. This environmental attributes at a further distance than the allowed us to investigate the importance of the parental attributes within the direct neighborhood will contribute perceptions of neighborhood environmental attributes to explaining the level of independent mobility. Future for specific physical activity behaviors in specific con- research should look further into this. texts. This is important for the design of future physical Higher levels of independent mobility were related to activity interventions. Furthermore, this study included a higher levels of active transport to school and higher total relatively large study sample. physical activity levels, but no association was found with Limitations of the present study included the cross- the daily amount of step counts. These findings concur sectional study design, which does not permit to infer a with the conclusions drawn in the existing international causal relationship. Secondly, adolescents were not asked literature. Page et al. (2009) and also Carver et al. (2010) to register cycling activities for which the activity monitor investigated the relation between the level of independent was not removed. As activity monitors are not able to mobility and the level of physical activity among 10–11 register cycling behavior accurately, this has a reflection year old children. In both studies, independent mobility on the daily amount of step counts. Third, step-counts appeared to be an important correlate of the level of active were determined using the Yamax Digi-Walker CW701 transport and physical activity for this age group [37]. and the GT1M accelerometer. Although the step counts Since active commuting has been acknowledged by a num- measured by the Yamax Digi-walker CW-701 have been ber of studies as a physical activity behavior with a signifi- shown to be highly correlated with the step counts of the cant contribution to the total physical activity level, this has GT1M accelerometer, the overall agreement between the important implications for health promotion. Parents may step counts of both monitors is rather low [52]. To over- limit their child’s level of independent mobility and conse- come this problem, all analyses were controlled for the quently restrict their child’sopportunitiestobephysically type of monitor used. Finally, the level of independent active. Encouraging independent mobility may therefore mobility was determined using self-reported data. The use contribute to increasing levels of active transport and as a of objective measures of GIS- and GPS-technologies in consequence the total physical activity level. combination with self-reported data would ascertain the The mediating analyses showed that in boys none of places visited without adult accompaniment. This may be the associations between the parental perceptions of a possible solution for the lack of consensus about the neighborhood environmental attributes and active trans- operationalization of the level of independent mobility port or total physical activity was mediated by independ- which limits the comparability between the results of this study and the results of other studies. ent mobility. Among girls on the other hand, the level of independent mobility mediated the association of the parental perception of residential density with active Conclusion transport to and from school (11.8%) and with the total A first conclusion of this study is that among 10–12 self-reported level of physical activity (43.2%). Further- year old boys and girls, the parental perceptions of more, the associations between the parental perception neighborhood environmental characteristics are asso- of land use mix access and the parental perception of ciated with their level of active transport and total safety for traffic with the total physical activity level were physical activity. This may be important for Belgian mediated respectively for 37.7% and 39.5% by the level of policy makers and urban planners to make well-considered independent mobility. Based on these results, we might decisions concerning built environmental redevelopments assume that independent mobility is of higher importance of existing Belgian neighborhoods and planning of new in girls than in boys. Promoting the level of physical neighborhoods. De Meester et al. BMC Public Health 2014, 14:631 Page 13 of 14 http://www.biomedcentral.com/1471-2458/14/631 Secondly, in girls, the level of independent mobility 12. Salmon J, Timperio A, Cleland V, Venn A: Trends in children's physical activity and weight status in high and low socio-economic status areas of plays a role in the association of the parental perceptions Melbourne, Victoria, 1985–2001. Aust N Z J Public Health 2005, 29:337–342. of neighborhood environmental characteristics with their 13. Mc Donald NC: Active transportation to school: trends among US level of active transport and their total physical activity schoolchildren, 1969–2001. Am J Prev Med 2007, 32:509–516. 14. Van Der Ploeg HP, Merom D, Corpuz G, Bauman AE: Trends in Australian level. When designing programs, focusing on neighbor- children traveling to school 1971–2003: burning petrol or hood environmental characteristics, to promote active carbohydrates? Psychol Aging 2008, 46:60–62. transport or physical activity, it should be kept in mind 15. Dollman J, Norton K, Norton L: Evidence for secular trends in children's physical activity behaviour. Br J Sports Med 2005, 39:892–897. that the level of independence given to girls also needs 16. World Health Organization: Global Stragey on diet, Physical activity and to be considered to obtain behavior change. However, Health. Geneva: Geneva, Zwitzerland: World Health Organization; 2004. before the results of this study can be generalized, these 17. Giles-Corti B, Timperio A, Bull F, Pikora T: Understanding physical activity environmental correlates: increased specificity for ecological models. results should be first confirmed in other studies. Exerc Sport Sci Rev 2005, 33:175–181. 18. Davison KK, Lawson CT: Do attributes in the physical environment Competing interest influence children's physical activity? A review of the literature. Int J The authors declare that they have no competing interests. Behav Nutr Phys Activ 2006, 3:19. 19. Ferreira I, Horst K, Wendel-Vos W, Kremers S, van Lenthe FJ, Brug J: Authors’ contribution Environmental correlates of physical activity in youth - a review and FDM coordinated the data collection, assisted in the recruitment of the update. Obes Rev 2007, 8:129–154. participants, conducted the statistical analyses and drafted the manuscript. 20. Sallis JF, Owen N, Fisher EB: Ecological Models of Health Behaviour. In GC, DVD and IDB participated in the interpretation of the data, revised the Health Behavioour and Health Education: Theory, Research and Practice. 4th draft versions of the manuscript and provided critical comments during the edition. Edited by Glanz K, Rimer BB, Viswanath K. San Francisco: process. All authors read and approved the final version of the manuscript. Jossey-Bass; 2008:462–484. 21. Gebel K, Bauman A, Petticrew M: The Physical Environment and Physical Acknowledgments Activity: A Critical Appraisal of Review Articles. Am J Prev Med 2007, The authors wish to acknowledge the contribution of Coppin Evelyne, Bieke 32:361–369. Moerman, Elien Moerman, De Backer Ann-Sofie, Fauve Vandendorpe, Delrue 22. Heath GW, Brownson RC, Kruger J, Miles R, Powell KE, Ramsey LT, The Task Marlies, Everaert Marloes and Vanden Bergh Tine in the data collection. This Force on Community Preventive Services: The effectiveness of Urban work was supported by a Ghent University Special Research Fund (BOF) Design and Land Use and Transport Policies and Practices to Increase [grant number BOF 08/24J/134]. Physical activity: A Systematic Review. J Phys Act Health 2006, 3(Suppl 1):S55–S76. Author details 23. Owen N, Humpel N, Leslie E, Bauman A, Sallis JF: Understanding Department of Movement and Sport Sciences, Faculty of Medicine and environmental influences on walking; Review and research agenda. Am J Health Sciences, Ghent University, Watersportlaan 2, B-9000 Ghent, Belgium. Prev Med 2004, 27:67–76. Research Foundation Flanders (FWO), Brussels, Belgium. 24. Wendel-Vos W, Droomers M, Kremers S, Brug J, van Lenthe F: Potential environmental determinants of physical activity in adults: a systematic Received: 16 May 2013 Accepted: 12 June 2014 review. Obes Rev 2007, 8:425–440. Published: 20 June 2014 25. Ding D, Sallis JF, Kerr J, Lee S, Rosenberg DE: Neighborhood environment and physical activity among youth: a review. Am J Prev Med 2011, 41:442–455. References 26. Wong BYM, Faulkner G, Buliung R: GIS measured environmental correlates 1. Warburton DE, Nicol CW, Bredin SS: Health benefits of physical activity: of active school transport: A systematic review of 14 studies. Int J Behav the evidence. Can Med Assoc J 2006, 174:801–809. Nutr Phys Activ 2011, 8:39. 2. Strong WB, Malins RM, Blimkie CRJ, Daniels SR, Dishman RK, Gutin B, 27. Panter JR, Jones AP, van Sluijs EM: Environmental determinants of active Hergenroeder AC, Must A, Nixon PA, Pivarnik JM: Evidence based physical travel in youth: A review and framework for future research. Int J Behav activity for school-age youth. J Pediatr 2005, 146:732–737. Nutr Phys Act 2008, 5:34. 3. Steele RM, Brage S, Corder K, Wareham NJ, Ekelund U: Physical activity, 28. Pont K, Ziviani J, Wadley D, Bennett S, Abbott R: Environmental correlates cardiorespiratory fitness, and the metabolic syndrome in youth. J Appl of children's active transportation: a systematic literature review. Health Physiol 2008, 105:342–351. Place 2009, 15:827–840. 4. Jimenez-Pavon D, Kelly J, Reilly JJ: Associations between objectively 29. Giles-Corti B: Encouraging Walking for Transport and Physical Activity in measured habitual physical activity and adiposity in children and Children and Adolescents How Important is the Built Environment? adolescents: Systematic review. Int J Pediatr Obes 2010, 5:3–18. Sports Med 2009, 39:995–1009. 5. Kokkinos P: Physical Activity and Cardiovascular Disease Prevention: 30. Ball K, Jeffery RW, Crawford DA, Roberts RJ, Salmon J, Timperio AF: Current Recommendations. Angiology 2008, 59:26S–29S. Mismatch between perceived and objective measures of physical 6. Vuori I: Physical activity and health: Metabolic and cardiovascular issues. activity environments. Prev Med 2008, 47:294–298. Adv Physiother 2007, 9:50–64. 31. van Vliet W: Exploring the fourth environment. An examination of the home 7. Verloigne M, Van Lippevelde W, Maes L, Yildirim M, Chinapaw M, Manios Y, range of city and suburban teenagers. Environ Behav 1983, 15:567–587. Androutsos O, Kovacs E, Bringolf-Isler B, Brug J, De Bourdeadhuij I: Levels of 32. Fyhri A, Hjorthol R: Children's independent mobility to school, friends and physical activity and sedentary time among 10- to 12-year-old boys and leisure activities. J Transport Geogr 2009, 17:377–384. girls across 5 European countries using accelerometers: an observational 33. Fyhri A, Hjorthol R, Mackett RL, Fotel TN, Kytta M: Children's active travel study within the ENERGY-project. Int J Behav Nutr Phys Act 2012, 9:34. and independent mobility in four countries: Development, social 8. Lee MC, Orenstein MR, Richardson MJ: Systematic review of active contributing trends and measures. Transport Pol 2011, 18:703–710. commuting to school and childrens physical activity and weight. J Phys 34. Hillman M, Adams JGU: Children's Freedom and Safety. Child Youth Act Health 2008, 5:930–949. Environ 1992, 9:12–33. 9. Davison KK, Werder JL, Lawson CT: Children's active commuting to school: 35. Jago R, Thompson JL, Page AS, Brockman R, Cartwright K, Fox KR: current knowledge and future directions. Prev Chronic Dis 2008, 5:A100. Licence to be active: parental concerns and 10-11-year-old children's 10. Faulkner G, Buliung R, Flora PK, Fusco C: Active school transport, physical ability to be independently physically active. J Public Health (Oxf ) activity levels and body weight of children and youth: A systematic 2009, 31:472–477. review. Prev Med 2008, 48:3–8. 11. Sturm R: Childhood Obesity - What we can learn from existing data on 36. Page AS, Cooper AR, Griew P, Jago R: Independent mobility, perceptions secular trends. Prev Chronic Dis 2015, 2:A20. of the built environment and children's participation in play, active De Meester et al. BMC Public Health 2014, 14:631 Page 14 of 14 http://www.biomedcentral.com/1471-2458/14/631 travel and structured exercise and sport: the PEACH Project. Int J Behav Nutr Phys Activ 2010, 7:17. 37. Carver A, Timperio A, Hesketh K, Crawford D: Are children and adolescents less active if parents restrict their physical activity and active transport due to perceived risk? Soc Sci Med 2010, 70:1799–1805. 38. Wen LM, Kite J, Merom D, Rissel C: Time spent playing outdoors after school and its relationship with independent mobility: a cross-sectional survey of children aged 10–12 years in Sydney, Australia. Int J Behav Nutr Phys Activ 2009, 6:15. 39. Page AS, Cooper AR, Griew P, Davis L, Hillsdon M: Independent mobility in relation to weekday and weekend physical activity in children aged 10–11 years: The PEACH Project. Int J Behav Nutr Phys Activ 2009, 6:2. 40. Carver A, Timperio AF, Crawford DA: Young and free? A study of independent mobility among urban and rural dwelling Australian children. J Sci Med Sport 2012, 15:505–510. 41. Hillman M, Adams J, Whitelegg J: One false move: a study of children's independent mobility. London: PSI Publishing; 1990. 42. O'Brien M, Jones D, Sloan D, Rustin M: Children's Independent Spatial Mobility in the Urban Public Realm. Childhood 2000, 7:257–277. 43. De Bourdeaudhuij I, Sallis JF, Saelens BE: Environmental correlates of physical activity in a sample of Belgian adults. Am J Health Promot 2003, 18:83–92. 44. Rosenberg D, Ding D, Sallis JF, Kerr J, Norman GJ, Durant N, Harris SK, Saelens BE: Neighborhood Environment Walkability Scale for Youth (NEWS-Y): Reliability and relationship with physical activity. Prev Med 2009, 49:213–218. 45. Scoring for the Neighborhood Environment Walkability Scale - Youth (NEWS-Y). In [http://www.drjamessallis.sdsu.edu/Documents/NEWS-Yscoring.pdf] 46. Ducheyne F, De Bourdeaudhuij I, Lenoir M, Cardon G: Test-retest Reliability and Validity of a Child and Parental Questionnaire on Specific Determinants of Cycling to School. Pediatr Exerc Sci 2012, 24:289–311. 47. Philippaerts RM, Matton L, Wijndaele K, Balduck AL, De Bourdeaudhuij I, Lefevre J: Validity of a physical activity computer questionnaire in 12- to 18-year-old boys and girls. Int J Sports Med 2006, 27:131–136. 48. Crouter SE, Schneider PL, Karabulut M, Bassett DR Jr: Validity of 10 electronic pedometers for measuring steps, distance, and energy cost. Med Sci Sports Exerc 2003, 35:1455–1460. 49. Silva P, Mota P, Esliger D, Welk G: Technical Reliability Assessment of the Actigraph GT1M Accelerometer. Meas Phys Educ Exerc Sci 2010, 14:79–91. 50. McNamara E, Hudson Z, Taylor SJ: Measuring activity levels of young people: the validity of pedometers. Br Med Bull 2010, 95:121–137. 51. Behrens TK, Dinger MK: Motion sensor reactivity in physically active young adults. Res Q Exerc Sport 2007, 78:1–8. 52. Kinnunen TI, Tennant PWG, McParlin C, Poston L, Robson SC, Bell R: Agreement between pedometer and accelerometer in measuring physical activity in overweight and obese pregnant women. BMC Public Health 2011, 11:501. 53. Miller R, Brown W, Tudor-Locke C: But what about swimming and cycling? How to ‘count’ non-ambulatory activity when using pedometers to assess physical activity. J Phys Act Health 2006, 3:257–266. 54. De Meester F, De Bourdeaudhuij I, Deforche B, Ottevaere C, Cardon G: Measuring physical activity using accelerometry in 13 to 15 year old adolescents: the importance of including non-wear activities. Public Health Nutr 2011, 14:2124–2133. 55. Keene ON: The log transformation is special. Stat Med 1995, 14:811–819. 56. MacKinnon: An introduction to Mediation Analysis. New York, USA: Erlbaum Psych Press; 2008. 57. De Meester F, Van Dyck D, De Bourdeaudhuij I, Deforche B, Cardon G: Does the perception of neighborhood built environmental attributes influence active transport in adolescents? Int J Behav Nutr Phys Activ 2013, 10:38. 58. Cardon G, Maes L, De Bourdeaudhuij I: Bicycling to school during the Submit your next manuscript to BioMed Central transition from childhood into adolescence. A 6 year longitudinal study. and take full advantage of: Pediatr Exerc Sci 2012, 24:369–383. 59. Veitch J, Bagley S, Ball K, Salmon J: Where do children usually play? A qualitative study of parents’ perceptions of influences on children's • Convenient online submission active free-play. Health Place 2006, 12:383–393. • Thorough peer review • No space constraints or color figure charges doi:10.1186/1471-2458-14-631 Cite this article as: De Meester et al.: Parental perceived neighborhood • Immediate publication on acceptance attributes: associations with active transport and physical activity • Inclusion in PubMed, CAS, Scopus and Google Scholar among 10–12 year old children and the mediating role of independent mobility. BMC Public Health 2014 14:631. • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit

Journal

BMC Public Health – Springer Journals

Published: Dec 1, 2014

Keywords: public health; medicine/public health, general; epidemiology; environmental health; biostatistics; vaccine

There are no references for this article.