1. Introduction
Worldwide, the majority of children spend more than two hours per day watching TV [
1,
2], and laptop and phone use in children is increasing substantially [
3]. Besides screen activities children engage in many other sedentary behaviors, e.g., sitting in the classroom and doing homework [
4,
5]. Studies in Europe and the United States, show that children spend approximately 8 h per day sedentary [
6,
7].
Recently, interest in the health consequences of excessive sedentary behavior is increasing. Systematic reviews only including longitudinal studies concluded no convincing evidence for a prospective relationship between childhood sedentary behavior and various health outcomes [
8,
9]. However, sedentary behavior tracks from early childhood to later in life [
10,
11], and for adults evidence for an adverse relationship between sedentary behavior and morbidity and mortality is emerging [
12,
13]. Therefore, excessive sedentary behavior among children remains a public health concern and developing interventions to reduce sedentary behavior seems appropriate. Unfortunately, existing interventions targeting reductions in children’s sedentary time have disappointing effects, possibly explained by a lack of knowledge on the most important determinants for engaging in sedentary behavior [
14].
Previous reviews found that the most commonly examined determinants can be grouped in the following domains: (1) demographic and biological; (2) psychological, cognitive and emotional; (3) behavioral; (4) social and cultural; and (5) physical environmental [
15,
16,
17,
18]. Recent reviews found insufficient evidence for determinants of sedentary behavior in children in these five domains, as few prospective studies were available, and the majority of the determinants were examined only once. Furthermore, most studies focused on screen time such as watching television or playing videogames as an indicator of sedentary behavior [
17,
19,
20]. Yet screen time takes up only a small part of the total time spent sedentary [
5,
21,
22,
23]. Finally, Brug and Chinapaw [
24] state that many studies exploring determinants were not specifically designed for this question, and the determinants identified in the different reviews mainly concern characteristics of sedentary children instead of motivational and contextual reasons for engaging in sedentary behavior, which is essential for the design of effective interventions.
Another major gap in the current evidence on determinants is the perspective of children on why they engage in sedentary behavior. As children are the experts of their own behavior [
25], their perspectives could bring new insights into relevant determinants of sedentary behavior. Furthermore, as parents play a significant role in facilitating and regulating their children’s behavior [
26], their view on potential determinants of their children’s sedentary behavior is of additional relevance. Previous studies found, for example, that parents’ screen time, and parents’ rules concerning screen time and playing outside, were associated with children’s sedentary time [
27,
28,
29]. Therefore, the aim of this study was to explore child- and parent-perceived determinants of children’s sedentary behavior, using concept mapping with 11 to 13 year old children and their parents. Concept mapping is a helpful participatory method to acquire knowledge about children’s and parents’ perspectives, by allowing them to provide their unique contribution to research concerning their own behavior [
30,
31].
2. Materials and Methods
2.1. Participants
Children and one group of parents were recruited via different primary schools in The Netherlands. Between April and October 2015 primary schools were selected through purposive sampling, contacted through information letters and follow-up telephone calls, aiming to include both rural and urban primary schools comprising children with different socio-economic backgrounds. After approaching 26 primary schools, the first school was included (response rate 4%). Thereafter, through purposive sampling based on socio-economic background and rural/urban area three other schools were contacted and included (response rate 100%). Thirty-eight children (response rate 40%) and seven parents were willing to participate (response rate 12%). Due to late inclusion of the fourth primary school, parents of the children at this school were not able to participate in the online concept mapping sessions. Therefore, a second group of parents (n = 14) were recruited through convenience sampling in January 2017.
When a school agreed to cooperate, information letters were sent to the children in the 8th grade, aged between 11–13 years old, and their parents. Both parents from non-participating and participating children were allowed to take part, as we were interested in the ideas of parents in general and not specifically the parents of participating children. Parents from the convenience sample received an information letter via email. As a reward for participation children received a small present, parents were offered an update on the results of the study, and the schools were offered a presentation about the results of the study.
The VU University Medical Ethical Committee concluded that the protocol does not fall within the scope of the Medical Research Involving Human Subjects Act. No identifying participant information was collected for the purpose of this study, and written informed consent was signed by one parent and the participating child, or the participating parent.
2.2. Procedures
Concept mapping is a research method in which group data are collected qualitatively in a structured and inductive way, and are analyzed quantitatively. The concept mapping procedure consists of several steps: generation of participants’ statements towards a seeding statement, grouping all statements and rating of their importance, summarizing these data into computer generated concept maps, and interpretation of these concept maps [
30,
31]. In this way, structure and objectivity to qualitative data are provided [
30,
31]. Generally, concept mapping studies include an average of 10–20 participants per study [
31].
Concept mapping sessions and analyzes were performed between May 2015 and February 2017. At the start of the concept mapping sessions, gender and age were inquired from children, and from parents their age, gender, family structure, amount of siblings, and highest educational level, i.e., pre-vocational-, vocational-, or higher education. The socio-economic status (SES) of the primary schools was assessed based on zip codes of the schools and the status-scores document from the Dutch Social and Cultural Planning Agency. This SES score was divided in tertiles, i.e., low SES, medium SES, or high SES. The whole concept mapping procedure is described in detail below.
2.2.1. Concept Mapping Sessions—Children
The concept mapping sessions with the children took place at their school during school hours, lasting approximately 2 h each. Four groups of 8–10 children, attending different primary schools, participated in these sessions that were facilitated by two trained researchers. Because of children’s limited attention span the whole procedure was divided over two sessions.
The first session comprised a warming up exercise, an individual brainstorm, and a group brainstorm. The purpose of the warming-up exercise was to broaden children’s understanding about sedentary behaviors, i.e., children wrote down their daily activities and divided them in active and inactive activities. Thereafter, the seeding statement was presented in two ways:
Can you think of different reasons why you choose to do activities in which you have a seated position?
“For some things I do, I do them in a seated position, I do these things because”
Subsequently, the individual brainstorm started, in which the participants were stimulated to finish the sentence with as many ideas as possible, and to write down all ideas. In doing this, they were stimulated to think about their own sedentary behavior, as well as the sedentary behaviors of 11–13-year old children in general. Thereafter, participants shared one by one all of their statements in the group brainstorm. This resulted in a list of unique statements.
In the second session, each child individually grouped the statements, printed on separate cards, into piles of related statements. A minimum of three and a maximum of ten piles were required due to settings of the software program, with at least two statements per pile, and a miscellaneous pile was prohibited. Subsequently, children named the different piles. Additionally, children individually rated the importance of the statements in influencing their sedentary time using a 5-point Likert-scale, ranging from very unimportant to very important.
2.2.2. Concept Mapping Sessions—Parents
Concept mapping sessions with the parents were performed online, making it easier for the parents to participate. A comparable procedure as with the children was followed, i.e., a comparable warming-up exercise and the individual brainstorm towards the seeding statement were performed by e-mail, thereby omitting the group brainstorm. The sorting, naming and rating tasks were performed by means of the online program Ariadne [
32]. The seeding statement was presented in two ways:
Can you think of different reasons why your child chooses to do activities in which he/she has a seated position?
“For some things my child does, he/she does them in a seated position, he/she does these things because…”
2.3. Statistical Analysis
Separate analyses were carried out for each school and two groups of parents. The data were analyzed by multidimensional scaling and hierarchical cluster analyses using the software program “Ariadne” [
32], specifically designed for concept mapping. These analyses showed every statement as a point on a figure, with statements sorted more often together appearing close to each other, and statements never/rarely sorted together widely separated. By default eight clusters were made, combining statements close to each other. Subsequently, three independent researchers (Lisan M. Hidding, Teatske M. Altenburg, and Evi van Ekris) explored whether more or less clusters would represent participant’s ideas more adequate, i.e., each cluster comprising statements related to a similar concept. Based on the importance rating of the individual statements, average ratings for each statement and each cluster were calculated.
2.4. Interpretation of the Maps
Two researchers (LH and TA) interpreted the maps, by discussing similarities in the meaning of statements in each cluster, and subsequently, by determining the concept represented by each cluster. Next, researchers (LH and TA) named the clusters, far as possible derived from the names given by the participants in the individual sorting task. As statistical significance not always results in the best representation of the qualitative data, the researchers critically reflected on the computer-generated clusters. According to the researchers, some statements were not optimally represented in the computer-generated cluster maps. To improve the logic of the concept map, these statements were moved to nearby clusters or newly created clusters, the latter where possible based on clusters created by the participants in the individual sorting task. Original and final cluster compositions are shown in the
supplemental material (
Appendix A1 and
Appendix A2, respectively). Finally, the maps were discussed, listing similarities and differences between the maps of the different primary schools and between the maps of the children and the parents, and an explanation for the findings was searched. A third researcher (Mai. J. M Chinapaw or Evi van Ekris) was consulted in case of disagreement. In the results section the final clusters are presented. The final clusters were interpreted as potential determinants as perceived by children and parents. Whether these potential determinants are actual determinants cannot be concluded based on this concept mapping study.
4. Discussion
To the best of our knowledge this is the first study exploring child- and parent-perceived determinants of children’s sedentary behavior, by performing concept mapping. Children and parents indicated several matching potential determinants, of which “I sit because I can work/play better that way” as one of the most important perceived determinants according to both children and parents. In addition, children indicated “I sit because it is the norm/I sit because I have to” and “I sit because there is nobody to play with” as considerable important determinants of their sedentary behavior.
The potential determinants found in our study fit within the frequently used domains to group determinants [
15,
16,
17,
18]. The majority of child- and parent-identified determinants belong to the psychological, cognitive and emotional domain, e.g., “I sit because I’m not in the mood to do anything”, and “I sit because seated activities are fun”. However, the perceived determinants that children and parents rated as most important, i.e., “I sit because of the norm/I sit because I have to” (children, one group of parents), “I sit because I can work/play better that way” (children, all parents), and “I sit because there is nobody to play with (children), in particular belong to the social and cultural domain and the physical environmental domain”. Interestingly, determinants belonging to the demographic and biological domain are extensively discussed in previous studies [
17,
19], yet they received far less attention in the present study, only one group of parents indicated such a determinant, i.e., “My child sits because it is in his/her nature”. In addition, determinants belonging to the behavioral domain received no attention.
The potential determinants “Sitting because it is the norm/Sitting because I have to” and “Sitting because I can work/play better that way” mentioned by both children and parents both include motives related to the current societal norm of a sedentary lifestyle. Interestingly, one group of parents rated “Sitting because it is the norm” as unimportant, while children and the second group of parents rated it as one of the most important motives for engaging in sedentary behavior. By establishing rules, and commanding children to sit during activities, some parents might unconsciously stimulate their children’s sedentary behavior [
27,
33]. Owen et al. [
34] already found the norm to be a potential determinant in adults, as nowadays it is the norm to sit during many activities, including meetings, classes, and while relaxing at home. Similarly, many of the statements in the “Sitting because I can work/play better that way” clusters relate to the school or home environment, e.g., it is easier to work/play in a sedentary posture as their school environment is decorated for sedentary activities. Lanningham-Foster et al. [
35] confirm that the school environment influences children’s sedentary behavior, as children were significantly more physically active when a classroom was made more suitable for active behaviors, e.g., availability of mobile white boards for active learning lessons.
One group of children and parents identified “Sitting because it is a habit” as a potential determinant. This finding is in line with studies in adolescents: de Bruijn et al. [
36] found that habit strength was the strongest correlate of television viewing and Chinapaw et al. [
37] found that changes in habit strength were significantly related to changes in screen time. In young adults Conroy et al. [
38] found that habit strength had a strong bivariate correlation with objectively assessed sedentary behavior. Thus far, no studies examined the relation of habit strength with total sedentary time in children. The mechanism behind habit as a behavioral determinant is previously described in relation to exercise behavior. When behavior is a habit it is automatically activated by environmental stimulus or cues, instead of consciously deciding to engage in the behavior [
39]. Due to the current societal norm of a sedentary lifestyle, children learn to be sedentary from an early age. Consequently, sedentary behavior has become a habit, e.g., ‘My child sits because others do so, and it is a habit’ indicated by one group of parents. This cluster additionally indicates that children also learn to sit by imitating others.
Children rated “I sit because there is nobody to play (actively) with” as an important potential determinant of their sedentary behavior, related statements were found in the cluster “Siting because there is nothing (active) to do” which was also indicated by their parents. This might imply that children naturally prefer physically active play but tend to sit when they have nothing to do or nobody to play with. This is in line with the participatory study of Caro et al. [
40], where children preferred to be physically active in the school playground and were sedentary when they were bored. The children in this study perceived both social and physical environmental factors of the playground to influence fun of physically active play.
The children also indicated weather conditions as potential determinants, e.g., coldness, rain, or too hot temperatures within the clusters “I sit because of the weather” and “I sit because the physical environment is not suitable”. Additionally, they indicated statements related to safety and distance. Yilderim et al. [
41], found a significant positive association between rainfall and sedentary time, confirming that the weather may indeed be a determinant of sedentary behavior. Additionally, a review found that parents may restrict children in playing outside when they have concerns about the safety of the neighborhood [
42]. In our study only the children mentioned statements regarding safety of the neighborhood, indicating that it is not just a concern of their parents, however, children might have adopted their parent’s ideas.
Both children and parents identified enjoying sedentary activities as a potential determinant of sedentary behavior. Previous studies emphasized the importance of fun to children’s behavior as well. Norman et al. [
43] found a positive association between the enjoyment of sedentary activities and high amounts of sedentary time. Moreover, a review found evidence for a negative association between preference of sedentary activities and time spent on physical activity [
17].
Children mentioned statements related to feeling not so well and feeling bored, indicating children’s mood is another potential determinant of their sedentary behavior. This finding is in line with previous research indicating that lower psychological wellbeing was associated with higher TV/Video viewing and computer use during adolescence [
44] and higher TV viewing later in life [
45]. Rideout et al. [
3] found that children who spent more time using media, e.g., mobile/online media and television, were more likely to feel unhappy, sad or bored.
The influence of new media on children’s behavior is highlighted in the cluster “I sit because I want to make contact with my friends”. Underlying statements show the influence of smartphones, social media, and online games. Children use these platforms and devices to keep in touch with their friends and family.
4.1. Strengths and Limitations
An important strength of this study is the focus on both child- and parent perspectives, giving not only new insights in potential determinants but also their relevance/importance in influencing sedentary time. Yet the use of the concept mapping method further strengthens this study by structuring qualitative data in a quantitative way and providing the opportunity to visualize the composite ideas of a group in one map, including participants from different backgrounds. In addition, the method consists of several educational tasks, e.g., sharing ideas in a group context, and sorting related motives into piles. Furthermore, concept mapping gives all participants the opportunity to share their individual ideas, as children share their ideas one by one during the group brainstorm, and parents provided their individual ideas by e-mail. Despite that a few children needed more guidance during the sorting task concept mapping proved a feasible and valuable method for this age group. The online concept mapping sessions with the parents may have reduced social desirable answers.
A limitation of the study is the limited age range of the children (11–13 years old). A strength is that a diverse group of children participated, i.e., with low and medium SES, living in urban and rural areas. In addition, similar clusters were identified across schools, and at the last school no new topics emerged in the underlying statements (i.e., indicating saturation), thereby strengthening the evidence for the representativeness of our findings for 11–13 years old in The Netherlands. Similarly, in the second group of parents no new topics emerged. The participating parents, who were not necessarily the parents of the participating children, were mainly highly educated, which may have limited the representativeness of parent-identified determinants. The low recruitment rate among parents is an additional limitation of our study, which may have resulted in selection bias. Lastly, a few children had difficulties with the sorting task, resulting in more statements being rearranged by researchers in the final concept maps. As the rearranged statements were included in the clusters emerging from the individual sorting task of the other children, the final concept maps are still representative for children’s perspectives.
4.2. Future Recommendations
Although the evidence for the adverse health effects of excessive sedentary behavior in youth is unconvincing, there is little potential harm of implementing sedentary time limiting interventions [
8]. Current interventions aiming to reduce sedentary behavior mainly focused on determinants in the psychological, cognitive, emotional domain and behavioral domain. Strategies include for example enhancing children’s knowledge and awareness of their sedentary time, goal setting, and limiting TV time e.g., by using TV budgets [
14]. Our study suggests that these may not be the most important determinants for children. The children and most parents in our study rated potential determinants belonging to the social/cultural and the physical environmental domain as most important, namely “I sit because of the norm/I sit because I have to”, and “I sit because I can work/play better that way”. These child- and parent-perceived determinants call for very different intervention strategies targeting the social or physical environment, giving them more freedom being non-sedentary e.g., by allowing children to be active during lessons, homework or screen time or providing gymnastic balls, standing and exercise desks. Additionally, future interventions might benefit from focusing on both the school and home setting, as children indicate being limited by physical activity restricting rules and environments in both settings. Moreover, involving both children and parents in the development of future interventions seems important, as their perceptions were not identical. Lastly, interventions targeting sedentary behavior reductions may be incorporated in interventions promoting moderate-to-vigorous physical activity, for which ample evidence exists of its beneficial health effects in youth. Although, whether the identified factors are indeed determinants of sedentary behavior should first be confirmed in prospective observational and intervention studies. In addition to perceived determinants of overall sedentary behavior, future studies should focus on child- and parent-perceived determinants of specific sedentary behaviors such as screen time and sitting at school, as it is likely that determinants of these behaviors may deviate from that of overall sedentary behavior.