Gender Difference in Health Condition Among Emirati Adolescent: Role of Environment and Parent Perceptions

Background: The United Arab Emirates (UAE) is one of the developed nations in the middle-east with the obesity rate among the youth two to three times greater than the international standards. Therefore, this research aims to study the variation of perception of the built environment among Emirati male and female adolescents and their parents. Also, the role of determinants of the health condition of Emirati adolescents is analyzed. Method: A total of 335 students (aged 14-20) from six schools in the Al Ain region of Abu Dhabi Emirate and 250 parent responses were used to conduct the mean-test of perception of hindrances and multinomial logit of the health condition using the perception, behavior and built environment measures. Results: Emirati males perceive the built environment factors as barriers more than female adolescents. Parents perceive street-crossing (p<0.016) and sidewalk characteristics (p<0.020) to be more of a hindrance to walking to school than students. Trac exposure and walkability at home and schools report a signicant effect on the health condition of the Emirati adolescents along with the self-reported physical activity. Marginal effect reports sedentary and low-levels of activity predicts the probability of an Emirati adolescent being not just overweight or obese (p<0.001), but also underweight (p<0.05) Conclusion: Recommendations for parents and school authorities, Abu Dhabi Municipality and Urban Planning department, and to enhance the built environment are discussed. Also recommendation to address both, levels of obesity and being underweight, are discussed.

[18]. Other built-environment variables such as the land use mix, residential density, and street connectivity are the critical factors for assessing the walkability of a location and can provide insights for design and policy recommendations [19].

Study Area
This study was conducted in the Emirate of Abu Dhabi, the capital of UAE and also the largest of the seven emirates as it occupies 87% of the total area of the UAE. The Emirates of Abu Dhabi consists of three regions: Abu Dhabi-the capital central region, the Al Ain region, and the Western region, as shown in Fig. 1.

Study population
School-going children within the age group of 14-20 years (from 10th to 12th grade) from six schools from Al Ain were selected for this study. ADEK provided the approval to survey school students and their parents in public and private schools for this study, and IRB approval was obtained for human subject research from the institution. Informed consent was obtained from the parents and school children for the study. As public schools in the UAE are not co-ed schools, a pair of boys and girls school in the same district were selected. One private school from one of these districts was included in the study.

Survey Content and Administration
Parents were sent consent forms to request approval to volunteer and let their kids volunteer in the study. Student completed the survey in school while the parents who agreed to participate in the study completed the survey and sent it to the school. The school representatives were educated about the purpose and goals of this study to answer any questions. The survey inquired about the socio-economic, demographic, behavioral, and perception questions. Both the student and parent surveys were translated into Arabic (the local language in UAE) and were administered to the respondents.

Measures Dependent Variable: Health Condition
The measure of the Body Mass Index (BMI) based on age and gender speci cation from the International Obesity Task Force (IOTF) [20] were categorized as underweight, healthy weight, overweight, and obese health condition for students. These categorical measures of health condition were used as the dependent variable for the adolescents. The BMI for the participants was calculated by dividing the weight of each student by square of their height. Physical Activity Score The physical activity score was calculated based on student's responses about their active and sedentary behavior. The alpha value of the score was over 0.65 and hence was used further for the analysis. The outcome categories ranged from: Sedentary (1: -5-0); Low Activity (2: 1-6); Moderate Activity (3: 7-13); High Activity (4: 14-20); Vigorous (5: 21-27).

Built-Environment Perceptions
Perception of parent and the adolescent identi ed their concerns related to the constructs of sidewalk, street, tra c, safety, and appeal of the built environment. The concerns for each of the above constructs were summed to calculate the perception of concerns related to the ve built-environment constructs. Walkability Index The walkability index for schools and homes was calculated as the sum of intersection density, residential density, and land use mix [21]. The z-score for each variable -residential density, street connectivity, and land use mix was calculated and aggregated for 400m from the six schools and residences of individual student using: Walkability Index (WI) = z-score (residential density) + 2*z-score (intersection density) + z-score (Land use mix) Exposure to Tra c: The regulation by the ADEK restricts students from crossing major roads with designated speeds of 40kmph and above. Hence to evaluate the tra c exposure as a measure of speed, we used the the ratio of [length of (Highways + Majorroads)/ length of secondary roads] for the same catchments area as for the walkability index at both origin and destination [18].

Results And Discussion
Based on the International Obesity Task Force (IOTF) categories, male and female students were grouped into four health categories. About 30% of females and 40% of males are either overweight or obese (Table 1). While close to two-third of emirate females were in the normal weight category, only half of their counterparts were in the normal weight category. This suggests male adolescents are at higher risk of various health concerns related to physical inactivity than females.

Perception of Male and Female Adolescents
The perception of the built environment was compared for Male and Female adolescents to assess if their perceptions varied from each other. If so, who among both the sex perceived the environment to be less suitable to walk and also if the difference was statistically signi cant. Of the total 250 respondents, 158 were female respondents, and 92 were male ( Table 2). The respondents marked all concerns under each of these built-environment characteristics that they perceived to be a hindrance. Responses indicate, on average, male adolescents perceived higher average concerns than the females, and the differences were statistically signi cant for all the built-environment constructs. The street crossing concern for the males almost twice as much as the average concerns of female adolescents. We checked if the difference was due to exposure (walking to school) to the environment. The reported average walking for male and females were not signi cantly different, which indicates that male adolescent reported higher concerns about the built environment. Contrary to several studies, male adolescents report higher concerns or negative perception of the built-environment than female adolescents. This negative perception may be mediated because of two reasons. Firstly, male are associated with more sedentary behavior and engage with indoor recreational activities than being involved in outdoor sports. Secondly, males are known to receive their licenses earlier and drive more than their female counterparts. Further studies that can assess the activities of both male and female adolescents using objective measures such as accelerometer can provide more evidence to this posit.

Parent and Adolescent Perceptions
Parent perception of the built-environment was compared with the perception of adolescents (both male and female). On average, the parent perception varied from the children for the sidewalks and the street-crossing ( Table 3). The tra c, safety, and appeal of the environment were not signi cantly different for the parents and adolescents. While parent's perception of the street-crossing reported lower average than the children, sidewalk characteristics were of higher concern for parents in comparison to the children. Parent perceptions are generally more negative to let their children walk than their children's perception. Given that, it was surprising to observe that parents reported lesser concerns on average about the street-crossing in comparison to their children. While about 40-50% of adolescents were concerned about the width of the street with marked crosswalks, only about 30% of the parents were concerned about the same. The parents preferred the presence of crossing guard while the kids were keen about having a marked pedestrian crossing for them to cross the street.

The Determinant of Adolescent Health Condition
The determinants of the health condition of the adolescents were assessed using four multinomial models ( Table 4) While the r-square for the multinomial regression output is changed in log-likelihood from the intercept-only model to the four models, and unlike the linear regression, it cannot be interpreted the same way, higher the r-square value the better is generally accepted for comparison of models [22]. The r-square value improved the predictability of health condition from the base model with the perception of children and their characteristics from seven percent to 40% in the full model with parent's perception, physical activity levels, and objective measures of the built environment.
While the results of all four models are narrated, for discussion purposes, the full model will be elaborated for its implication on the health conditions of adolescents in UAE. The results from all four models indicate that the odds that adolescents who have higher concerns with street-crossing (p < 0.05) would be underweight was 2.2 times higher than being normal weight. However, the odds that they would be underweight decreased by three times if they were physically active while being overweight or obese were not statistically signi cant. Also, the odds that the male adolescent would be either underweight or obese increased by 2.5 times and about 20 times, respectively, than their female counterparts (p < 0.05).
The levels of physical activity reported similar results across the three models. Results indicate higher the levels of physical activity lesser are the odds that the adolescent would be underweight in comparison to being normal weight. This suggests that physical activity can not only mitigate the chances of overweight or obesity but can help reduce the chances of an adolescent being underweight.
The objective assessment of the built environment reported an interesting relation with the health condition of the adolescents. The exposure to tra c at destination (schools) reduces the odds that adolescents would be overweight by eight times in comparison to being normal weight. However, exposure to tra c at origin (home location) increases the odds by about four times (full model) that adolescents would be overweight or obese in comparison to being normal weight. This indicates, higher exposure to tra c near the adolescent's household tends to negatively impact the health condition of the adolescents, which complements other studies [23]. However, with an increase in exposure to tra c near schools, the odds that adolescents would be negatively impacted reduces. This could be as the exposure is calculated only for the six study schools. Also, opportunities to engage in PA at school may negate the effect of tra c exposure. However, further studies can con rm the postulate to inform transportation policies around the schools in UAE.
Better walkability at origin (homes) decreases the odds of adolescents being obese, while better walkability at destination (school) increases the odds of adolescents being overweight (by twice) or obese (by 13 times) in comparison to being normal weight. Does this indicate, a walkable environment around the schools does not help the health outcomes of the school-going children? While, again, this is contrary to most studies, this is indicative of several factors such as siting of school and policies related to walking around the school (policy restricting students walking on the road with speed higher than 40kms). Studies report siting of the school is an important consideration to encourage PA among school-going children [24].
*p < 0.05, **p < 0.01, ***p < 0.001 Also, higher concerns with sidewalk and tra c increased the odds of adolescents being underweight and obese, respectively. However, higher concerns with safety reduces the odds of the adolescents being underweight in comparison to being normal weight. Surprisingly, parent with higher concerns about streetcrossings was associated with lower odds of their children being overweight (p < 0.05) in comparison to being normal weight which is contrary to assumption.

Predicting Probability of Health Condition
The marginal effect of levels of physical activity on health conditions were estimated (Fig. 2). Table 5 reports the probability of an adolescent being underweight is 13.4% and 6.4% given that they report sedentary and low levels. It is hence important to propose speci c policies to encourage high levels of physical activity not only to deal with overweight ot obesity in adolescents, but also to mitigate negative health impact of being underweight. Also the probability of an adolescent being overweight or obese is signi cantly related to sedentary, low and moderate levels of physical activity. However, the probability of adolescent being overweight and obeses when engaged is vigorous physical activity is about 13% and 20% respectively, while the rest of the predictors are set to their mean values. This could indicate the in uence of improper diet, psychological factor or genetic [25]. Although outside the scope of this study, future studies to assess these factors could provide more insight.

Implications for Built-Environment
In general, the built environment is perceived to be non-conforming for walking by both females and males. They had major concerns with the constructs of street-crossing and safety of the built environment. Within the construct of street-crossing, the issue of the width of the street, and availability of marked crosswalks was of major concern to the adolescents. Within the construct of safety, the adolescents were concerned about the amount of tra c and unsafe driving behavior on the roads adjoining schools and homes. Street design approaches such as the introduction of "road diets" with curb-cuts at the intersections with textured crosswalks, crossing island, and lane narrowing could improve the perception of safety and concerns related to street-crossing.
Both tra c exposure and walkability around the homes of the adolescents have an impact on the health conditions of the adolescents. A better walkable environment with a higher land-use mix and density can encourage more walking [28,29], and reduce the chances of both being underweight and obese in Adolescents. The Department of Urban Planning and Municipalities of Abu Dhabi (ADPM) have suggested implementing the Urban Street Design Manual projects in Al Ain with special consideration to the needs of pedestrians [28]. Implementation of this approach in core residential areas (trip origins) can reduce vehicle speeds, cut-through tra c, and general tra c noise. Additionally, reduction of speed limits on major roads, especially on designated routes for children to walk to and from school could encourage walking within the neighborhood.
In Abu Dhabi, the main concern is about the policies surrounding built environment that make it hard to walk to school [30]. ADEK and ADPM can nd transformative strategic plans to promote active commuting to school as the primary mode of transportation. They could follow the example of projects in California that enhanced tra c safety along routes to school (for example, widening of crosswalks and sidewalk improvements), which led to signi cantly greater walking or biking for children [31].

Implications for Health of Adolescents
A study by Al-Haddad et al [32] reported that both male and female in UAE tend to be about twice overweight compared to international standards. In this study, obesity and overweight in male adolescents were about 10% more than female adolescents. Several socio-cultural and behavior factors speci c to the region could be attributed to this difference. Culturally, males are the majority of the workforce in the UAE and are engaged in driving more than women. About 75 percent of all trips in the emirate of Abu Dhabi is made by the male [33]. Studies indicate a high correlation of driving behavior and obesity [34], which could be one of the driving factors for higher rates of overweight and obesity among men. A recent study con rmed the high rates of obesity in males in the UAE. It established sedentary behavior as one of the major causes of the obesity epidemic in males [35]. Creating opportunities to engage them in outdoor activities such as soccer and basketball games, a preferred sports activity in the region, within their neighborhood with a directive to engage in outdoor activity should be encouraged among the adolescents and youth in the emirate.
Another less addressed issue with health outcomes is the assessment and impact of underweight [36]. Being a male adolescent is positively associated with being underweight. While signi cant factors relate to malnutrition and dietary habits, fatigue, emotional, genetic, and psychological factors, and smoking are other factors that can lead to adolescence' underweight health condition. Research suggests using "effective public health programs that can simultaneously promote a healthy lifestyle, improve diet quality, and address both undernutrition and chronic disease" [25]. Health policies that regulate dietary intake both at school and at home need to be promoted. Recently schools in Abu Dhabi restricted the availability of high-calorie food at school. Extending that by providing good dietary plans for students in school can regulate this further and provide the necessary support to curb the issue of underweight in adolescents.

Implications for Enhancing Physical Activity: Planning and Policy
The school environment can provide "quality physical education and supportive environment for a long-lasting healthy and active lifestyle" [37]. Health and physical literacy at schools can provide needed knowledge, motivation, and engagement in physical activity. Control competence allows self-directed structuring and pacing of physical activity, which has shown to be more effective [38]. ADEK policies to actively engage schools to develop control competence approaches for children and adolescents in school can instill PA behaviors that could last until adulthood.
Outside the school environment, the physical environment around the homes of adolescents can provide opportunities for engagement in an active lifestyle.
Emirati housing communities lack sizeable open play areas and mixed use development to provide opportunities for after-school activities [39]. Also, alternative approaches such as pedestrian bridges for students to circumvent major roads with high speed to reduce exposure to high tra c should be incorporated in the design manual. In existing neighborhoods, retro tting and improving alleys' physical conditions and appearance could also enhance their use and thus support walkability. Recommended physical improvement plans include paving, landscaping, and upgrading lighting to promote walking. Such improvements may be particularly helpful in suburban neighborhoods, where alleys are often unpaved, not utilized, improperly lit, and obstructed by debris and vegetation [41]. Although it's important to improve the environment to increase the physical activity levels, it is also important to increase the awareness of the bene ts of physical activity among parents. A recent study stated that to enhance the ability to engage in physical activity, it is important to improve the built environment and raise awareness by educating the parents [42].
The outcomes of this study provide implications to improve the overall health condition for adolescents in Abu Dhabi. While these implications are speci c to Abu Dhabi and UAE, major cities in the middle-east have similar socio-cultural, topographic, and demographic characteristics. They can implement similar approaches to enhance the health condition of their local population. However, the policy and planning recommendations are adaptable for all municipalities and schools proactive in improving the built environment and health conditions of school-aged children.

BMI -Body Mass Index
Declarations Ethics approval and consent to participate The study was approved by the Abu Dhabi Department of Education and Knowledge (ADEK). The survey instruments for the school, parent and child were approved by the Khalifa University Institutional Review Board. Consent forms were signed by the representative of each study school. Signed approval was obtained from parents for their own participation and participation of their children in the study. Only students who received approvals from their parents were administered the survey for this study.

Consent for publication
The consent for publication was obtained from all school representative, parents for themselves and their children.

Availability of data and materials
The ADEK approvals do not permit us to share the data with any third party and restricts us from use and distribution of the same without their permission, given the privacy rules in the United Arab Emirates.

Competing interests
The authors declare that they have no competing interests

Funding
This study was conducted without any nancial support from any agencies or institutions.
Authors' contributions PM was responsible for conception, analysis and monitoring of the study. AW was responsible for data collections and approvals. PM, KA and AW analysed and interpreted the data and its implications. PM and KA were responsible for drafting the manuscript. All the authors have read and approved the content of the manuscript.  The city of Al Ain is located in the eastern side of the Capital Center and shares its borders with the Sultanate of Oman. Source: Author. Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.