Parents’ and adolescents’ perception of traffic- and crime-related safety as correlates of independent mobility among Belgian adolescents

The independent mobility (IM), defined as the freedom of young people to travel without adult supervision, has been related to the physical activity time, the acquisition of personal autonomy, to less intense fear of crime, and to a stronger feeling of being part of their community and other health and social benefits. The aims of this study were to compare parents’ and adolescents’ traffic- and crime-related safety perceptions of their neighborhood and to analyze the associations of these perceptions with adolescents’ IM. A total of 291 adolescents and their parents completed the Neighborhood Environment Walkability Scale (NEWS) questionnaire. Multilevel (two-level models: individual level—neighborhood level) regression analyses were conducted to examine whether the environmental perceptions differed between parents and adolescents and the association between the parental and adolescents’ perception to the IM and the active independent mobility (AIM). Parents reported a more negative perception of traffic (except for amount and speed) and crime-related safety. Adolescents’ environmental perceptions were not associated with their IM but parental perceptions of traffic- and crime-related safety were associated with IM and with active IM, although not all associations were in the expected direction. Future urban policy efforts should address environments where parents perceive sufficient levels of safety to increase the levels of IM in adolescents.


Background
The World Health Organization (WHO) recommends at least 60 minutes of moderate-to-vigorous physical activity (PA) per day in order to achieve health benefits in youth aged 5 to 17 years old [1]. These benefits include a better mental health status, better self-esteem and the neighborhood [39] and Schoeppe et al. related higher rates of IM when parents had a higher educational level and reported higher social connection [40]. Systematic reviews about IM assessed the methodological approach [41] or its relation to physical activity and weight status [15]. To the best of our knowledge, the relation between parents' and adolescents' perception of traffic-and crime-related safety and their association with IM has not been studied before. Addressing whose perceptions (adolescents or parents) are more relevant to determine IM is important because it would allow researchers and practitioners to prepare effective interventions to increase IM and, indirectly, to increase the rates of active transportation [42]. The purpose of this study was two-fold: the first aim was to compare parental and adolescents' traffic-and crime-related safety perceptions of the same neighborhood environment. The second aim was to analyze the association of parental and adolescents' neighborhoodrelated traffic and crime safety perceptions with adolescents' IM.

Protocol and procedures
The data were collected between September 2014 and March 2016 as part of the follow-up measurements of the Belgian Environmental Physical Activity Study in children (BEPAS-child) [43] that took place between December 2011 and May 2013. The study was conducted in Ghent, Belgium. The participants were nested within neighborhoods (n = 98). The number of participants per neighborhood ranged from 1 to 13 and the included neighborhoods were spread across the city and suburbs (i.e. low and high walkable areas) in Ghent, and across low and high SES areas. The participants first received an information letter, informing that a researcher would visit them at home a few days later and introducing the purpose of the follow-up study. Of the 606 children who participated in BEPAS-child, 375 adolescents agreed to participate in this followup study and completed a survey about their PA and perceptions of their neighborhood environment. The study was limited to 12-15 years old because it corresponded to the follow-up of a study conducted three years later with children. Therefore the sample accessible to this work was young adolescents. After selecting those who provided complete data, the final sample was composed of 291 adolescents and their parents. Ethical approval (i.e. sample, procedure, consents (active and opt-out), instruments) was obtained from the Ethics Committee of Ghent University, and participants' informed consent was obtained prior to the data collection. The adolescents provided active consent (adolescents signed the informed consent if they want to participate) and the parents provided passive consent a letter was delivered to all the parents and they could retract permission returning the document signed).

Measures
Socio-demographic characteristics. The adolescents self-reported their age, gender, height, and weight, and their parents self-reported their age, gender, height, weight, and highest educational level. The educational level of the parents was categorized into "low education" (no education, primary, lower secondary, or higher secondary) or "high education" (bachelor or master degree).
Perceived neighborhood environmental factors. The Neighborhood Environment Walkability Scale for Youth (NEWS-Y) and its version for adults [44] were used to assess parents' and adolescents' perception of the neighborhood environment. For this study, the 8 items of traffic safety and the 6 items of the safety from crime subscales were used, since these are common items asked exactly in the same way to both parents and adolescents (Table 1). All items were answered on a four-point Likert scale (from 1 = strongly disagree to 4 = strongly agree). The reliability of the questionnaire is acceptable for parents (ICC between 0.61-0.78) and for adolescents (ICC between 0.56-0.87). The internal consistency of the two subscales was acceptable for all participant groups (α between 0.72 and 0.90) [44].
Independent mobility (IM). Adolescents' IM is commonly assessed by means of a questionnaire [41]. In the current study, we asked adolescents about the average time they spent travelling without accompaniment per trip in a usual week, separately for (i) walking, (ii) cycling, and (iii) using public transport in order to assess adolescents' IM. Only 15% of the participants reported 0 minutes of IM in all modes of commuting. Therefore, the mean of the time walking, cycling, and using public transport without adult accompaniment was used to define the variable "IM". The variable "active IM (AIM)" was determined by the mean of the time using active modes of commuting (i.e. walking and cycling) without adult accompaniment.

Statistical analyses
Descriptive statistics were reported for all relevant variables. Multilevel (two-level models: individual level-neighborhood level) regression analyses were conducted to examine whether 1. There is so much traffic in nearby streets that it's hard or unpleasant to walk in my neighborhood (alone or with someone else).
2. In the nearby streets traffic usually drives slowly (50km/hour or less).
2. In the nearby streets traffic usually drives slowly (50km/hour or less).
3. Most drivers drive faster than the maximum speed limit in our neighborhood.
3. Most drivers drive faster than the maximum speed limit in my neighborhood.
4. There are a lot of exhaust fumes when I walk in my neighborhood.
4. There are a lot of exhaust fumes when I walk in my neighborhood.
5. At night the streets are well-lit in our neighborhood. 5. At night the streets are well-lit in my neighborhood.
6. In the streets in our neighborhood pedestrians and cyclists are very visible to people from their houses.
6. In the streets in my neighborhood pedestrians and cyclists are very visible to people from their houses.
7. There are zebra crossings and traffic lights in our neighborhood to help pedestrians at busy places.
7. There are zebra crossings and traffic lights in my neighborhood to help pedestrians at busy places.
8. I think it is safe to allow my child to cross the road in our neighborhood.
8. I feel safe crossing the roads in my neighborhood.

Crime-related safety Crime-related safety
9. There is a lot of crime in our neighborhood. 9. There is a lot of crime in my neighborhood the environmental perceptions differed between parents and adolescents. The age group (parents versus adolescents) was included as an independent variable and the perceptions -the 14 items of traffic safety and crime safety-were alternately entered as dependent variables in the regression models. All analyses were controlled for adolescents' gender and parents' highest educational level. Multilevel regression analyses were conducted to examine the associations of the parental and adolescents' perceptions of traffic safety and crime safety with IM. The perceptions of traffic safety and crime safety were included as independent variables, and both IM and AIM were alternately entered as dependent variables. Again, these analyses were controlled for adolescents' gender and parents' highest educational level. The analyses were performed using the Statistical Package for Social Sciences (SPSS Version 20.0 for Windows, IBM Corp., Armonk, NY, USA), and the level of significance was set at p<0.05.

Descriptive statistics
The descriptive data of the participants are presented in Table 2. The parents' mean age was 41.4 ± 8.9 years old, mainly mothers (76.5%) completed the parental questionnaire, and 40% of them had a college or university degree. The adolescents were mostly girls (54.7%) and their mean age was 13.2 ± 1.0 years old. No differences in socio-demographic characteristics were found between boys and girls. We also found no gender differences for overall IM and AIM, but boys were allowed to travel further by bike than girls (p = 0.03). The means of the parents' and adolescents' perceptions of traffic safety and crime-related safety are presented in Table 3.

Relation between parents' and adolescents' perception of traffic safety and crime-related safety
The results of the regression analyses are presented in Table 4. All the perceptions of traffic safety and crime-related safety differed between parents and adolescents, except for the amount of traffic and speed of traffic. Regarding traffic, parents had a higher perception of speeding and of exhaust fumes than adolescents (both p<0.001). Adolescents, however, reported more presence of good lighting at night, that pedestrians were visible from the houses nearby, a higher presence of crossings and signals for pedestrians, and more safety when crossing the streets than parents (all p<0.001). Concerning crime-related safety, parents reported more crime in the neighborhood and more fear of being hurt by a stranger (when commuting alone in the neighborhood, commuting with a friend, walking through the neighborhood and being in a nearby park), than adolescents (all p<0.001 except for crime p = 0.035). Furthermore, parents rather perceived that walking at night was not safe because of crime (p = 0.001) compared to adolescents.

Associations of parents' and adolescents' perceptions of traffic-and crimerelated safety with IM
The associations of both parents' and adolescents' perceptions of traffic safety and crimerelated safety with IM and AIM are shown in Table 5. Several parental perceptions were associated with IM and AIM, but none of the adolescent perceptions were associated with IM or AIM.
More parental concerns about being hurt by a stranger when the adolescent is with friends and having clear view of pedestrians from houses nearby were associated with lower IM.   When parents reported more concerns about being hurt by a stranger when the adolescent was alone, higher IM was reported. Concerning AIM, parental perceptions of more safety at crosswalks and less presence of exhaust fumes from the cars were associated with higher AIM, and when parents reported more concerns about the adolescent being hurt when being with a friend, AIM was lower. Again, more parental concern about being hurt by a stranger when the adolescent is alone was associated with higher AIM and a higher presence of crosswalks and signals to cross busy roads was associated with lower AIM.

Discussion
The aims of this study were (i) to analyze the differences in perceptions of traffic-and crimerelated safety in the same neighborhood environment between adolescents and parents, and (ii) to analyze the associations of adolescents' and parents' perceptions with IM and AIM. The main results showed that parents and adolescents have different traffic and crime-related safety perceptions in the same environment, and only some parental perceptions, but not adolescents', were associated with IM and AIM.
The perceptions of traffic safety and crime-related safety in the same environment were different between adolescents and parents, except for the amount and the speed of traffic. Parents showed more concerns about traffic safety and crime-related safety, while adolescents were more confident about both of them. This result is consistent with other studies comparing perceptions of the environment between parents and children in relation to active commuting to school. These studies indicated that parents reported almost twice more presence of traffic and stranger danger than both children and adolescents [45,46]. In fact, children reported that parents were more worried than them about traffic and crime safety [45]. Nevertheless, fear of traffic may not be unfounded, since 44% of the injury-related mortality in youth in developed countries is caused by transport accidents [47]. Belgium is the most congested country in Europe in terms of hours spent in traffic [48] and, during the past years, traffic jams have more than doubled. In this sense, parents are probably more aware than adolescents of the actual situation, and they probably have a deeper background resulting in more negative perceptions of traffic safety. A strategy to increase the parental perceptions of traffic safety may be to develop educational programs for both parents and their children, including activities where they commute together in order to increase parents' confidence in their child's commuting behavior [49] and improve their perceptions of the neighborhood. Our findings showed that the adolescents' perceptions of traffic and crime-related safety were not associated with IM or AIM. To date, no previous studies have examined the association of these factors with IM, but some studies have investigated the link with the active commuting behavior. A study in US children (10-12 years old) indicated that only children's perceived presence of parks for children was associated with active mobility [45], and a Swedish study showed that adolescent-reported fear of traffic and crime was related to active commuting (47).
In the present study, the parental perception of traffic, particularly crosswalks and exhaust fumes, was associated with IM and AIM, but our findings were contradictory. Perceiving more safety at crosswalks, less exhaust fumes, and less presence of crossing signals were associated with higher AIM, whereas a clearer view of pedestrians from nearby houses was associated with a lower IM. The perception of traffic is one of the most important barriers for parents in relation to active commuting [50], and it has been associated with lower rates of active commuting in USA adolescents [51]. In fact, in the current study more traffic safety items were related to AIM than to overall IM. In relation to parental perception of traffic safety, a study conducted in Belgian children (10-12 years old) found that parental-perceived traffic safety was positively associated with AIM (for cycling) [25]. The presence of cars and air pollution have become a problem in Belgium because of the congestion of traffic [48], and further policy actions to develop programs focusing on reducing car emission and the amount of cars will be important to solve this public health problem. In addition, when there was a clear view of the adolescent from the houses nearby, IM was lower. This result might be explained by the current perceived erosion of trust and social relations in many communities in the last years [52], which may cause less faith in the good intentions of the community nearby.
In relation to crime-related safety, the parental perception of fear of the adolescent being hurt by a stranger when they were with friends was associated with a lower IM, whereas the parental fear of the adolescent being hurt when they were alone was associated with higher IM. In this regard, Australian children (10-12 years old) of parents with more fear of strangers had limited IM, and this association differed between boys and girls [53]. Consequently, strategies that help parents to recognize fear and its consequences are important to increase IM. The mixed results about stranger danger might be related to the assumption that adolescents take more risks when they are accompanied by friends due to their peers' influences in comparison to when they are alone [54]. However, it is only a speculation and further qualitative research with focus groups could offer a deeper insight into the reasons of this counterintuitive result. Furthermore, it is important to improve the community and social relations in order to increase IM and AIM, since social networks are key to achieve higher levels of independence [55]. This is also crucial to rise active commuting rates, because the fear of crime is higher when there is less pedestrian movement [56].
In addition, active commuting is related to IM [57]. Thus, strategies to increase the active commuting to school and other destinations may be important to increase adolescents' IM. A high variety of interventions focusing on increasing active commuting rates in children and adolescents have been conducted worldwide [58,59], specifically though interventions on school promotion more than policy and environmental changes [60]. Unfortunately, the interventions conducted did not achieve a high increase of active travels [61]. Few interventions targeted secondary school students and they focussed mostly on educational changes [62]. Thus, it is important to continue implementing new global interventions (i.e. environmental, social, personal) to promote active commuting and IM in children and adolescents.
The advantages of this study are the use of a large sample of paired adolescents and parents from different neighborhoods and the fact that traffic safety and crime-related safety were assessed using a validated questionnaire. Nevertheless, some limitations of this study should be acknowledged. First, since the study was executed in only one city of Belgium, generalizability of the findings is limited. Second, a cross-sectional design was used, which precludes any assumptions about causality. Lastly, we measured IM using the average minutes that participants commuted independently. Since information about the total mobility (i.e. minutes/week or distance) is lacking, we do not know how much of the participants' total mobility is independent. In future research, it would be interesting to include a more detailed assessment of IM and to test the longitudinal changes in adolescents' and parents' perceptions across different geographical contexts. In addition, according to the ecological model explained before [23] it would be interesting to assess personal correlates as motivation or behavioral patterns to build a more complete model of the independent mobility.
In conclusion, parents and adolescents living in the same environment had different perceptions of traffic and crime-related safety, and only some parental perceptions, but not adolescents' perceptions, were associated with IM and AIM. Future urban policy efforts should address environments where parents perceive sufficient levels of safety to increase the levels of IM in adolescents. In addition, educational programs focusing on improving parental perceptions of traffic-and crime-related safety and on the importance of reducing car use will be required to increase adolescents' IM and AIM.