Physical activity, sedentary screen time and bullying behaviors: exploring differences between adolescents with and without disabilities

ABSTRACT The study’s purpose was to examine the prevalence of sedentary screen time behaviour and physical activity (PA) participation, and their association with environmental (socioeconomic status, bullying behaviours) and personal factors (age, sex). We analysed data from the 2013–14 Health Behaviour in School-aged Children survey among Israeli adolescents aged 11–17 [N = 4,241; 56% female; 9.3% = adolescents with disability (AWD)]. A minority (<10%) of all participants complied with the 2010 World Health Organization global PA recommendations. AWD were significantly more involved in bullying behaviours. PA participation was significantly decreased in AWD in grades 6 and 8 compared to adolescents without disability. Furthermore, > 30% of AWD and adolescents without disability engaged in sedentary screen time >3 hours/day. In conclusion, AWD in grades 6 and 8 should be given specific attention, to reduce the likelihood of bullying at school and to increase the inclusion of AWD in PA lessons.

attention of educators and health care providers for managing participation opportunities (Klitsie et al., 2013;Olds et al., 2010). The International Classification of Functioning, Disability and Health (ICF), officially endorsed by all World Health Organization (WHO) member states (World Health Organization, 2001), is a framework and common language used to describe and organize information on outcomes and sources of functioning and disability. Within the ICF, contextual factors from environmental variables such as attitudes of others and personal variables such as age and sex, interact with the individual's functional limitations and health condition, as well as the activity limitations and participation restrictions, in a holistic model of disability (World Health Organization, 2013). The principles of ICF have been well established within rehabilitation research (Reinhardt, 2011), as well as within adapted physical activity (PA) research (Hutzler & Sherrill, 2007). According to this framework, personal and environmental contextual factors are associated with activity limitation and participation restriction.

Sedentary screen time behaviour, physical activity, and adolescents with disabilities
Much of adolescents' voluntary sedentary time is related to screen utilization, including using smartphones, television viewing, playing computer games, and using the computer for other activities (Klitsie et al., 2013;Olds et al., 2010). Evidence of associations between sedentary behaviour and poor health outcomes among AWD, such as obesity, low fitness, and even low self-esteem (Haegele et al., 2020), compliments to recommended limits of less than two hours per day spent on screen time for all adolescents (On behalf of SBRN Terminology Consensus Project Participants, Tremblay, M. S., Aubert, S., Barnes Domoff et al., 2020). Despite these recommendations, there is evidence to suggest that a high proportion of AWD do not keep to these limits, and this has been increasing over time (Haegele et al., 2020;Ng et al., 2018Ng et al., , 2019. School-aged adolescents have the opportunity to spend more time partaking in sedentary screen-time behaviour (SSB) during the weekends, and therefore it is important to examine the behaviours during the week and weekends separately (Throuvala et al., 2020). Moreover, boys tend to report more time partaking in SSB than girls, especially regarding computer games (Humbert et al., 2006). Socioeconomic status (SES) is an important environmental factor in SSB and PA participation (Humbert et al., 2006). Adults and children from higher SES groups are more likely to participate in PA compared to lower SES groups (Crespo et al., 1999;Woodfield et al., 2002), and are less likely to partake in SSB (Kristjansdottir & Vilhjalmsson, 2001;Lowry et al., 1996).
AWD are at risk of developing further health conditions due to their sedentary behaviour (Li et al., 2019). They are often excluded from different types of exercise programs, both in school and afterschool activities, due to inaccessibility -making them more likely to engage in sedentary activities (Ganz et al., 2020). In the school setting, AWD are excluded more often from a variety of activities, and may be placed in environments with more sedentary activities. Furthermore, the experience of physical education for children with disabilities is the participation is in passive activities (Fitzgerald, 2012). Due to the lack of physical activity opportunities, AWD become more likely to engage in SSB such as watching television, and may have more school absenteeism as well as other exclusions from school activities, such as physical education class and field trips (Walker et al., 2019).
Due to the significantly higher incidence of obesity and the lower level of PA in AWD compared to adolescents without disabilities, which will likely result in greater health concerns in adulthood, PA levels should be increased in AWD (McPherson et al., 2016;Rimmer & Rowland, 2008). These authors suggested that environmental barriers imposed on AWD may be considered as the primary reason for their lack of participation in PA.

Adolescents bullying and adolescents with disabilities
Bullying is characterized by repeated and unwanted aggression where there is a power differential that results in physical, emotional, social, or educational harm (Gladden et al., 2014). According to the Indicators of School Crime and Safety (Harel-Fisch et al., 2016), approximately 20% of adolescents aged 12-18 years reported being bullied. In Israel about 10.5% of the adolescents in grades 6-12 have reported being bullied within the last month. In the general adolescent population, sex was found to be a consistent factor influencing the likelihood of bullying, with significantly more males involved in bullying than females (Carbone-Lopez et al., 2010;Von Marées & Petermann, 2010;Wang et al., 2009). Additional factors associated with bullying were older age and lower psycho-social status of the bullying participants (Von Marées & Petermann, 2010). AWD are commonly affected by bullying behaviour, oftentimes more than adolescents without disabilities. Based on meta analyses, children and adolescents with disabilities or chronic physical conditions were 1.65 times more likely to be victims of bullying behaviours than their peers without disabilities (Pinquart, 2016). Some authors also reported the troubling occurrence of bullying behaviours within PA participation in physical education and club sports (e.g., Dane-Staples et al., 2013;Haegele & Kirk, 2018).

Health Behaviour in School-aged Children Survey
The Health Behaviour in School-aged Children (HBSC) survey has been conducted every four years since the early 1980s, as a collaborative cross-national study, coordinated by the WHO Regional Office for Europe in Denmark (Currie et al., 2009). In the recent surveys, more than 40 countries were involved, mostly in Europe. PA participation has been documented in the HBSC study, and was found to be linked to overweight and obesity rates (Janssen et al., 2005) and sedentary behaviour (Guthold et al., 2010). School bullying is also reported in the HBSC study, and results from previous years have shown that boys have been more involved in physical and verbal bullying than girls (Wang et al., 2009).
PA participation of school AWD was reported in a few of the HBSC studies. For example, Steele et al. (2004) compared health behaviours among school adolescents with physical disabilities with the Canadian national sample. Significant differences were found, suggesting that AWD participated less in PA (Steele et al., 2004). More specifically, fewer boys with disabilities were reported to partake in a sufficient amount of PA compared to males without disabilities, when data from 15 countries were pooled together from the HBSC study (Ng et al., 2017). Moreover, more males and females with disabilities reported a greater amount of screen-time than their peers without disabilities (Ng et al., 2018). Regarding bullying behaviours, Sentenac et al. (2013) reported that the percentage of victims was significantly higher among those reporting disability than among their peers without disabilities in 11 countries studied within the HBSC study 2005/2006. Therefore, based on an Israeli 2014-15 HBSC data subset, the first purpose of the current study was to describe the SSB and PA participation of AWD compared to adolescents without disabilities in Israeli schools. The second purpose was to outline the associations between SSB and PA with environmental factors such as bullying behaviour and SES, as well as personal factors including sex and age (classified by grade level). The current study examined the role of environmental and personal contextual factors among AWD compared to adolescents without disability as correlates and predictors of SSB related to screen time behaviour and physical inactivity.

Participants
This study uses Israeli data from the 2014-15 Health Behaviour in School-aged Children: WHO Collaborative Cross-National survey/study (HBSC) cross-national survey conducted among adolescents aged 11-17. The HBSC is a school-based survey of adolescent health behaviours and psychosocial determinants, carried out among representative samples of school-aged adolescents every four years, using an international standardized methodological protocol involving standardized procedures for sampling and translation of items . In order to ensure a representative sample, according to the international HBSC protocol  the Israeli Ministry of Education's list of schools was used. Classrooms were randomly sampled (90% classroom response) and for each sampled school an additional class was also randomly sampled. All school pupils present in the sampled classrooms were included (>95% pupil response). The research protocol received approval from ethics committees of the Israeli Ministry of Education and Bar-Ilan University.
The research population included students from Hebrew-speaking state-secular and statereligious schools. Participation in the study was fully voluntary and anonymous, with no explicit incentives provided for participation. Questionnaires were administrated by trained research assistants in the absence of a teacher during regular class time. Parents were informed about the study via the school administration and could bar their children from participating if they disagreed with it. Pupils were divided into three similarly clustered and representative sub-samples, each of them answering all core and one segment of elective questions on disabilities.

Personal factors
Individual variables included sex and age. Individuals were asked to report their sex (male or female). Age was determined by their school grade (6 th , 8 th , 10 th ,12 th ). Disability type was established via the participant's self-report. Participants answered the question 'Do you suffer from a disability?' A list of specific impairment types followed: visual impairment, hearing impairment, mobility impairmentnot using a wheelchair, mobility impairment -using a wheelchair, and other. Answers were yes or no.

Environmental factors
SES -SES was estimated with the Family Affluence Scale II (FAS). The scale is composed of four items (number of cars, holidays taken, computers owned, and bedrooms) and has a total score that ranges from 4 (lowest) to 15 (highest), which determines the FAS level. The FAS has been found to be a better proxy of SES than measures that rely on adolescent reports of parental occupation or income (Boyce et al., 2006).
School bullying -We measured school bullying from nine questions pertaining to bullying others and nine questions pertaining to being bullied. The questions focused on the following bullying behaviours: fighting, harassment, using derogatory names, hitting, spreading lies, ridiculing someone because of his/her colour or ethnicity, boycotting others, sexual comments, and texting related harassment. In each question, the participant was asked to report the amount of times he or she was engaged in or was the subject of such behaviours in school. The participant could respond with: none, only once or twice, two or three times a month, about once a week, or several times a week. Scores ranged from 9 to 45, with higher scores representing a greater extent of school bullying (bullying others or being bullied). The items were based on the Olweus Bully/Victim Questionnaire (Olweus, 1993).

Sedentary screen-time behaviour
SSB was assessed using six questions from the HBSC survey pertaining to the number of hours the participant engaged in screen watching behaviours (i.e., television, computer, video, DVD, and other screen-related entertainment) during the weekdays (Sunday through Thursday) and weekend (Friday through Saturday). For each question there was a 9-point response scale ranging from 'none at all' to 'seven or more hours a day' (Iannotti & Wang, 2013). Weighted means for weekday and weekend use were used to calculate the total hours per week (average weekday screen time*5 + average weekend screen time*2). A previous validation study examined a brief questionnaire used to measure television time, similar to that used in the HBSC study (Bucksch et al., 2014). In addition, these HBSC items have acceptable test-retest reliability (intraclass correlation coefficients ranging from 0.76-to-0.81) and validity (correlations ranging from 0.36-to-0.54) (Vereecken et al., 2006). As adolescents who watch television for three hours or more per day have a 65% higher chance of being obese than adolescents who watch less than one hour per day (Boone et al., 2007), we further grouped the sample based on their SSB to adolescents with SSB of up to three hours a day and adolescents with SSB of three hours or more a day.

Physical activity level
PA was defined by 'any activity that usually increases your heart rate and makes you get out of breath some of the time'. An example of a question was, 'How often over the past seven days have you been physically active for a total of at least 60 minutes per day?' Answers were given on an 8-point scale (0 = none to 7 = daily). The measure has reasonable validity (r = 0.37) with 5-day accelerometer data (Walker et al., 2019) and acceptable test-retest reliability when used as a dichotomous variable (Ng et al., 2019). According to the WHO (World Health Organization, 2010), adolescents and young people aged 5-17 should engage in at least 60 minutes of moderate-to-vigorous intensity PA (MVPA) daily. Accordingly, participants' responses were converted into the following two categories: not meeting and meeting the WHO recommendations.

Study participants' characteristics
Descriptive statistics for the study participants (AWD and adolescents without disability) were calculated and compared using chi-squared tests for categorical variables and independent t-tests for continuous variables.

Physical activity level, sedentary screen time behaviour, and bullying behaviours (being bullied and bullying others)
Within-group differences in PA level, SSB, and bullying behaviours in the different age groups, and sex, were compared using independent t-tests or one-way analysis of variance with Tukey's honest significant difference post-hoc test. Between-group differences (differences between adolescents with and without disability) were compared using independent t-tests. In addition, SSB was examined via calculation of the percentage of adolescents in each study group and in each grade who engaged in the sum SSB of up to three hours a day and adolescents with SSB of three hours or more a day. For example, a participant who engaged in two hours of watching TV and one hour of working on the computer has a total of three hours of SSB. Accordingly, this participant will be part of the SSB of three hours or more a day group. Between-group differences in SSB were compared using chi-squared test of independence. Based on the question 'How often over the past seven days have you been physically active for a total of at least 60 minutes per day?,' participants were grouped into two groups: participants meeting and not meeting the WHO 2010 PA recommendations (i.e., engaging vs. not engaging in at least 60 minutes of MVPA daily). The prevalence of adolescents in each study group was calculated and compared via chi-square test of independence.

Variables related to and predicting sedentary screen time behaviour and physical activity level
To evaluate factors associated with SSB and PA levels, Pearson correlations were conducted. In addition, differences between categorical variables in SSB and PA level were examined separately with independent t-tests. Next, four separate forward multiple stepwise regression procedures were conducted to analyse factors predicting SSB and PA level among adolescents with and without disability. Only variables that had significant correlations or presented significant differences within the different sub-categories (categorical variables) with the dependent variables were included. All independent variables were checked for multicollinearity using the variance of inflation factor (variance of inflation factor > 10; Myers, 1990). The criterion for inclusion in the model was an alpha level of 0.05, and the exclusion criterion was an alpha level of 0.10. The aforementioned analyses were conducted separately for adolescents with and without disability. In all statistical analysis procedures, the level of significance was set to P < 0.05 (2-tailed).

Study participants' characteristics
This study included 4,612 adolescents, 429 of whom AWD (9.3%), who completed the items on disabilities. Additional information on group characteristics and between-group differences is presented in Table 1.

Sedentary screen time behaviour and physical activity level
During the week, 37% (12 th grade) to 49% (10 th and 11 th grades) of adolescents without disability engaged in SSB for ≥ three hours/day. Among the AWD, this ranged from 43% (11 th grade) to 52% (8 th grade). During the weekend, 33% (12 th grade) to 55% (10 th grade) of adolescents without disability were engaged in SSB for ≥ three hours/day. For the AWD, the range was from 40% (8 th grade) to 58% (8 th grade). No statistically significant between-group differences were observed in SSB in either the weekdays or weekends (chi-square ranged from 0.01 to 1.41; p > 0.05). Moreover, among AWD, no statistically significant differences were observed in SSB between males and females, between adolescents in different grades (Table 2), or between adolescents with different aetiologies (Table 3). However, among adolescents without disability, adolescents in 12 th grade engaged in less SSB than adolescents in 6 th through 10 th grades. No statistically significant betweengroup differences were observed in SSB of adolescents with and without disability (Table 2).
Overall, 422 (9.15%) of AWD and adolescents without disability who participated in this study reported that they carry out a sufficient amount of PA to be considered as physically active. More specifically, 27 AWD (6.29% of the sample with disability) and 395 adolescents without disability (9.44% of the sample without disability) met the 2010 PA recommendations of at least 60 minutes of MVPA per day. Moreover, among adolescents without disability, this compliance level ranged from 3% (12 th grade) to 16% (6 th grade). The prevalence of males who met the PA guidelines was greater than females (12.97% and 6.53%, respectively). Among AWD, the compliance level ranged from 2% (12 th grade) to 12% (6 th grade). A higher frequency of adolescents without disability reached the recommended PA level in the 6 th and 8 th grades (p < 0.05). No statistically significant differences were observed between males and females in the prevalence of participants achieving the recommended PA level (7.40% and 3.26%, respectively). Additional information is presented in Figure 1.
PA levels of adolescents without disability in the 6 th and 8 th grades were significantly higher than those observed among AWD in the same age groups (p = 0.01). No statistically significant betweengroup differences were observed in the 10 th , 11 th or 12 th grades. Moreover, in adolescents without   Notes: 'Being bullied' and 'Bullying others' scores range from 9 to 45, with higher scores representing a greater extent of school bullying (bullying others or being bullied); * within-group differences -significantly different than females; a) within-group differences -significantly different than 6 th grade; b) within-group differences -significantly different than 7 th grade; c) withingroup differences -significantly different than 8 th grade; d) within-group differences -significantly different than 9 th grade; e) within-group differences -significantly different than 10 th grade; SD, standard deviation. disability, in comparison to females, males presented a higher level of PA (2.21 days of the week vs. 3.18 days of the week, respectively). Similarly, in comparison to males with disability, males without disability presented a greater level of PA (3.18 days of the week vs. 2.15 days of the week, respectively) (Table 2). Finally, no statistically significant differences in PA level were observed in different aetiologies (Table 3).

Bullying victimization and bullying behaviours
Within the group AWD, no significant differences were found across the different age groups in either being bullied or bullying others (p > 0.05). However, in comparison to females, males engaged more in both types of bullying activities. Moreover, in comparison to adolescents with vision and hearing problems, adolescents with mobility problems using a wheelchair were significantly more bullied and there were significantly higher rates of bullying others (Table 3).
Within the group of adolescents without disability, older adolescents (grades 10 through 12) reported less frequent occasions of being bullied. Differences across age in bullying others were not statistically significant (Table 2).
In addition, across the different age groups and in both males and females, in comparison to adolescents without disability, AWD were bullied more and also bullied others more (P < 0.001). Additional information is presented in Table 2.

Sedentary screen-time behaviour
Within the AWD group, all variables associated with SSB were not statistically significant. Accordingly, a regression analysis to predict SSB of AWD was not conducted. Within the group of adolescents without disability, grade and bullying others exhibited statistically significant associations with SSB (r = −0.08 and r = 0.03, respectively; Table 4). Both variables also significantly predicted SSB of adolescents without disability (adjusted R 2 = 0.04, F ratio = 8.48, p < 0.0.001; Table 5).

Physical activity level
Among AWD, PA level was associated with grade and SES (r = −0.17 and 0.12, respectively; Table 4). Grade and SES were also statistically significant predictors of PA level of AWD (adjusted R 2 = 0.04, F ratio = 16.13, p < 0.001; Table 5). Within the group of adolescents without disability, grade (r = −0.23), SES (r = 0.22), being bullied (r = 0.07) and bullying others (r = 0.15) exhibited all statistically significant associations with PA participation level (Table 4). However, a regression analysis showed that only sex (being a male) grade (lower grade) and SES (higher) predicted higher PA participation level (adjusted R 2 = 0.10, F ratio = 183.03; p < 0.001). Additional information is presented in Table 5.

Discussion
Based on the 2014-15 HBSC Israeli data subset, the present study's main purpose was to compare between adolescents with and without disability regarding the outcomes of and associations between (a) health behaviour (SSB and PA) and (b) personal factors (sex and age), and the environmental factors outlined in the ICF (SES and bullying). The secondary purpose was to outline potential predictors of PA and SSB. Given the data provided by Munikedem-Givon (2019) (2018), the proportion of AWD within the sample (9.3%) appears representing the population of students with disability in the general schools in Israel.

Sedentary screen-time behaviour
Based on data from 429 AWD and 4,183 adolescents without disability, only a minority of the participants met the health behaviour recommendations of the WHO. The mean time spent on SSB was more than four daily hours on weekdays (Table 2). Regarding age, nearly half of participants exceeded the recommended screen time threshold of up to three daily hours, with younger students exhibiting longer SSB. Based on our findings, the SSB time did not significantly differ across adolescents with and without disability. Our findings conform with most of the pooled HBSC scores representing 15 European countries (Ng et al., 2018). However, in the latter study the authors reported that after adjusting for age and SES, boys with disabilities spent significantly longer times in SSB compared to those without disabilities, while in girls the difference was not significant. Furthermore, when individual country data were addressed, the differences between those with and without disabilities were barely observable. These inconsistencies highlight the need for future studies to use a different method of examining the associations by sex.
While no personal or environmental factors addressed in this study predicted SSB in AWD, a finding that is of concern was that bullying others predicted SSB in adolescents without disabilities (Table 5). This finding may suggest that in addition to the health-oriented importance of reducing SSB, it also has social implication. Reducing SSB may be associated with less socially disruptive behaviour. In this study we used a pooled bullying score with no splitting by type of bullying behaviour, such as physical, verbal, or cyberbullying. In future studies, it would be important to explore which specific disruptive behaviours are associated with excessive SSB.

Physical activity participation
While according to 2010 WHO PA recommendations, adolescents should participate in MVPA at least one hour every day (World Health Organization, 2010), the mean participation rates in this sample were below 10%, with substantial reduced MVPA participation rates among AWD (6.29%), and significant differences presented across the 6 th and 8 th grade levels. The lack of significant differences in MVPA participation in older adolescents may have been a result of a 'floor effect' (i.e. having a large concentration of participants' scores at or near the lower limit of the scale (Hessling et al., Table 3. Adolescents with disability: One-way analysis of variance for differences in physical activity level, sedentary behaviour and school bullying based on disability type. Notes: a different than vision; b different than hearing; c different than mobility -no wheelchair; d different than mobilitywheelchair; e other; p < 0.05 (two-tailed) SD, standard deviation.

Sedentary behaviourweekdays and weekend, hours: mean (SD)
2004), as adolescents from the 11 th grade only exhibited a 5% compliance with the 2010 WHO PA recommendations, and adolescents from the 12 th grade only a 3% and 2% compliance for adolescents without and with disabilities, respectively. Based on the regression analyses, age represented by grade was the main predictor of participation in PA in adolescents with and without disabilities. In adolescents without disabilities, sex was also a significant predictor, suggesting that males were more compliant with the PA recommendations than females. This predictor did not hold true in AWD, probably also due to the 'floor effect'. Compliance with 2010 WHO PA recommendations in our sample (6.53% for females and 12.97% for males) was far below the overall mean from 2010, which was 18.6%, based on data from 32 countries in the HBSC study (Kalman et al., 2015). Furthermore, this rate of compliance was far below the Israeli rates reported in 2010 (23.6% for males and 10% for females). With regard to AWD, the rates presented in our sample ( Figure 1) were also far below the rates described in Finland, where an increasing trend was reported in compliance with MVPA recommendations between 2002 and 2014 (8.7% to 15.6% for females and 13.0% to 26.6% for males). It should, however, be acknowledged that the definition of AWD was different in Finland and also included participants with long-term illnesses (Ng et al., 2016). The profoundly low participation rate of MVPA particularly in AWD exhibited in our findings should be considered a major threat to the overall health of this population, which is increasingly being included and integrated within the general school system framework. According to a recent Canadian study (Brown et al., 2019), participation in MVPA confers protective benefits for internalizing disorders in AWD.
Our finding that higher SES significantly predicted MVPA participation in adolescents with or without disabilities is supported in several studies on adults (Crespo et al., 1999), adolescents (Humbert et al., 2006) or children (Woodfield et al., 2002). However, in contrast to our findings, Voorhees et al. (2009) reported that none of the indices of SES within a sample of adolescents from 36 middle schools across six geographic regions in the United States were related to device-based measures of PA. Regarding the potential mechanism explaining the SES effect, Humbert et al. (2006) reported that differences existed between adolescents from high compared to low SES regarding factors associated with their PA participation. Those coming from families with higher SES reported less participation barriers. However, in a comprehensive study analysing the relationship between adolescents' PA and family SES over 28 years, Telama et al. (2009) reported that while increased participation in PA organized by sport clubs was strongly associated with higher SES across sexes, no effect was reported in PA participation at the school. The impact of higher SES on PA participation should be considered by policymakers, and specific programs should be developed that encourage adolescents from families with lower SES to increase their participation in PA.
Based on our data, neither being bullied by others nor actively bullying others significantly influenced PA participation of AWD. This finding may be reassuring, as Haegele and Sutherland (2015) warned that bullying was the most persistently described negative peer interaction during PA in AWD. Interestingly, being bullying victims and bullying others were strongly associated with adolescents with and without disabilities. This finding can be supported by previous reports informing that some children developed a retaliation response with the same bullying behaviour to their peers (Davis & Watson, 2001;Fitzgerald, 2005). Such a response may, however, illustrate the beginning of a vicious cycle of being teased and teasing, which may ultimately result in negative social outcomes. Teasing and bullying experiences may have detrimental outcomes for school achievement (Cornell et al., 2013), and be associated with higher rates of anxiety disorders, such as social phobia, obsessive compulsive behaviour or panic disorders in adults (McCabe et al., 2003).
Given our findings, parents, educators and other stakeholders should be encouraged to support the participation of AWD within inclusive PA frameworks. Doing so may increase the underscored participation rate of AWD in PA. However, based on our data (Table 2), bullying others and being bullied were significantly more common in AWD compared to adolescents without disability. Adolescents who use a wheelchair for their mobility were most likely to be bullied and to engage in bullying others (Table 3). These data are important because they indicate that AWD, and particularly those with severe physical disability, still present a vulnerable minority in the school landscape. Educators and stakeholders should consider precautions during the inclusion process, minimizing the occurrence of bullying behaviours confronting AWD. Specific exposure to disability programs, such as Paralympic School Day (McKay et al., 2015), appears to be one way to cope with this concern.
Limitations There are several methodological limitations to consider when generalizing these conclusions. First, this study utilized a cross-sectional design. Longitudinal studies are warranted to support the magnitude of the associations found in this study. Second, PA and SSB variables were determined by self-report rather than direct measurement. Therefore, under -or over-estimation may have occurred. Third, variables were measured with questions taken from a larger survey, including additional potential personal and environmental factors. The confounding effect of such variables should be considered.

Conclusions
In conclusion, this study targeted personal and environmental factors associated with SSB and PA participation in adolescents with and without disabilities in a nationwide sample. Based on our findings, it may be suggested that only a minority of the adolescents with and without disabilities in this sample reported complying with WHO recommendations regarding PA and SSB. As to personal factors, the data indicate that adolescents with and without disabilities did not differ across sexes in SSB. However, males without disability were more physically active than females without disability. Furthermore, younger adolescents in Grades 6 and 8 presented increased SSB compared to those in Grades 9 through 12. Furthermore, PA participation was significantly lower among AWD in grades 6 and 8 compared to adolescents without disability. Regarding environmental factors, adolescents with and without disabilities were not influenced by SES. However, bullying behaviour was associated with SSB in adolescents without disabilities. While adolescents with mobility restrictions were significantly more involved in bullying behaviour, this did not seem to impact their PA participation. This study has outlined, for the first time, the risk of increased bullying behaviour and lack of physical activity in Israeli AWD. This said, an association of the bullying behaviour with physical activity was not demonstrated. Therefore, based on our findings, several recommendations can be set forward to educators and policymakers: (a) PA participation of AWD should be encouraged at all ages; (b) While the processes enabling PA participation of AWD are beyond the scope of this article, reducing the likelihood of being bullied seems to be a particularly important goal; and (c) reduction of SSB should be encouraged for all adolescents, thereby increasing desirable social behaviours.

Disclosure statement
No potential conflict of interest was reported by the authors.

Notes on contributors
Prof. Yeshayahu Hutzler is Full professor and chair of the Innovation and Entrepreneurship Centre at the Academic College at Wingate, Israel. Between 2007-2011 he assumed Presidency of the International Federation of Adapted Physical Activity and in the year 2017 was awarded for outstanding contributions to this field of study. Between 2013 and 2016 he served as Editor in Chief of the "Adapted Physical Activity Quarterly. By 2020 he published about 150 scholarly contributions. Since 1987, he has designed and developed several innovative exclusive as well as inclusive physical activity models and programmes for persons with disabilities.

Dr. Riki Tesleris
Senior Lecturer at the Department for Health systems management. She is also a member of the Health Behaviors in School Age Children research program and co-leader for Active Healthy Kids Global Matrix. Dr. Tesler focuses on research in the fields of health promotion, risk and health behaviors, In particular, her research is engaged in evaluation of health promotion strategies and intervention programs targeting the well-being and salutogenic attitudes among populations.