Social Support and Modelling in Relation to Physical Activity Participation and Outdoor Play in Preschool Children

Physical activity during early childhood is a prerequisite for healthy development in many cases. The aim of this study was to assess the relationships of social modelling and support from parents, peers, and siblings and domain-specific physical activity participation in a nationwide sample of preschool boys and girls from Germany. 519 preschool children aged 4–6 and one of their parents participated in the ‘MoMo’ Wave 1 Study between 2009 and 2012. Participants and their parents provided self-reported data on social support modelling, and domain-specific physical activity participation (physical activity in sports clubs, physical activity outside of sports clubs, and outdoor play). Parental, peer, and sibling support and modelling were related to domain-specific physical activity: Parental support was particularly relevant for physical activity in sports clubs, and peer support for outdoor play. Parental modelling was only related to physical activity outside of sports clubs: Maternal modelling was a positive correlate in girls and paternal modelling in boys, respectively. Sibling and peer modelling were especially relevant for physical activity in sports clubs. The results were heterogeneous regarding types and providers of support and modelling. Thus, different providers and types of support should be targeted in physical activity promotion programs for preschool children.


Introduction
Physical activity and outdoor play are essential in early childhood and contribute to a healthy development, as they enable cognitive, physical, emotional, social, and motor learning, as well as well-being [1,2]. Especially in early childhood, extensive physical activity and bodily movement are necessary to develop a sound base for lifelong health and active living throughout one's lifespan [3]. Thus, the German National Guidelines on Physical Activity recommend at least 180 minutes of physical activity per day in children aged 4-6 years [4].
However, 2 out of 3 preschool children in Germany [5,6] are not as physically active as recommended by the guidelines, with a higher risk of inactivity in girls compared to boys [5,7,8].
A large accelerometer study with over 700 participants conducted in a federal state in Germany,

Study Design
The data for this study was obtained from a subsample of the German nationwide MoMo-Study on physical fitness and physical activity in children and adolescents from Germany, and its umbrella study, the German Health Interview and Examination Survey for Children and Adolescents, KiGGS [27,28]. For both studies, a nationwide diverse sample of children and adolescents from Germany was planned. Thus, a stratified, multi-stage sample with three evaluation levels was drawn [29]. First, a systematic sample of 167 primary sampling units was selected from an inventory of German communities stratified according to the BIK classification system that measures the level of urbanization and geographic distribution [28]. Second, an age-stratified sample of randomly selected children and adolescents was drawn from the official registers of local residents. A total of 12,368 children and adolescents participated in the KiGGS Wave 1 [30], of which 6,076 were randomly assigned to MoMo Wave 1. Thus, the MoMo Study is based on a subsample of the KiGGS Study. The final number of preschool children aged 4-6 years in MoMo Wave 1 was 519, building the final sample of this study. Response rates of the MoMo Study are reported elsewhere [31].

Data Collection
MoMo Wave 1 data was collected between 2009 and 2012. The data collection took place over three years, because a large representative nationwide sample was drawn for KiGGS and the MoMo Study [27,28]. The data collection teams had to go to the 167 study sides, which were distributed over the whole of Germany. In order to avoid systematic bias in the study results by regional or seasonal trends, the sequence of sample points visited for data collection was laid down beforehand in a random route planning.
Parents and their children were invited to examination rooms at central locations in the 167 cities and municipalities that were within close proximity of their homes. A parent gave written informed consent and then completed questionnaires on physical activity behaviors, social support, social modelling, and socio-demographic aspects together with the child in the presence of a qualified interviewer on site. The KiGGS and the MoMo Studies were approved by the Charité/Universitätsmedizin Berlin ethics committee and the Federal Office for the Protection of Data and were conducted according to the Declaration of Helsinki [32].

Measures
The measures on physical activity participation and the social environment were captured within the MoMo Study, while confounding factors (socio-demographic variables) were captured within the KiGGS Study.

Physical Activity Participation
We used the MoMo Physical Activity Questionnaire (MoMo-PAQ) preschool version to assess habitual physical activity in different domains (physical activity in sports clubs, leisure-time physical activity outside of sports clubs, and outdoor play) in preschool children [33]. The MoMo-PAQ consists of 28 items and measures different domains of physical activities in a habitual week. Data obtained with the MoMo-PAQ are sufficiently valid and reliable (test-retest reliability: ICC = 0.68). Further information on the MoMo-PAQ is reported elsewhere [33,34].
Participants were asked if they regularly participate in physical activity in sports clubs. They could list up to four different physical activities in sports clubs they regularly engage in. A dichotomous variable "physical activity in sports clubs" was built according to 1-"regular participation in physical activities in sport clubs" or 0-"no physical activities in sport clubs".
Additionally, participants were asked to if they regularly participate in physical activities outside of sports clubs (e.g., swimming in a public pool with parents, Inline skating, or cycling). They could state up to four unorganized, leisure-time physical activities taking part outside of sports clubs. A dichotomous variable "physical activity outside of sports clubs" was built according to 1-"regular participation in physical activities outside of sports clubs" or 0-"no physical activities outside of sports clubs".
We used an 8-scaled item about days per week in which the child plays outside ("How often do you normally play outside during a week (for example: playing tag, skipping rope, or going to the swimming pool)?") and an item about minutes spent on average during one of those days to assess unorganized outdoor play. According to the WHO guidelines on physical activity [35], a dichotomous variable "daily outdoor play" was built according to 1-"seven days per week with outdoor play" or 0-"no to six days per week with outdoor play"

Parental, Peer and Sibling Modelling
Paternal, maternal and sibling modelling were all measured using a single item (e.g., "Does your father regularly do sports?"), which had a dichotomous answer format of 1-"yes" and 0-"no". We also measured peer modelling with a single item ("How many of your friends regularly do sports?") with a four-point rating scale ranging from 1-"none" to 4-"most of my friends".

Parental and Peer Support
Parental support scales followed the theoretical basis from Uchino [20]. Each scale included two items which were based on a four-point rating scale (e.g., for emotional support: "How important is sport in your family?" 1-"not at all" to 4-"very important"). We assessed peer support by a scale containing three items which were also based on a four-point rating scale (e.g., "How often do your friends ask you if you want to play outside or do sports with them (e.g., playing soccer, riding a bicycle, inline skating)?" 1-"never" to 4-"always"). The scales on social support had good or moderate test-retest reliability over a period of one week. Further information on these measures have been published elsewhere [36].

Confounding Factors
A migration background was assumed if the participant themselves had immigrated to Germany, at least one parent was not born in Germany, or if both parents immigrated to Germany or had no German nationality [37]. Individual-level socioeconomic status (SES) was derived separately for both parents and included items on educational and professional status and the total household income [38]. Children with separated parents were assigned the socioeconomic status of the parent they lived with. Income, educational, and professional status were scored on a scale from 1 to 7, and a sum score was created (range: 3-21) and categorized into low (3)(4)(5)(6)(7)(8), medium (9)(10)(11)(12)(13)(14) and high (15)(16)(17)(18)(19)(20)(21) socioeconomic status [39]. The type of residential area was defined according to the number of residents living in the participants hometown, differentiated into rural area (<5000 residents), small town (5000-19,999 residents), medium-sized town (20,000-99,999 residents), and city (>99,999 residents). Additionally, we captured the "region in Germany", meaning former east and former west Germany. Furthermore, we also took siblings (yes/no) and the family situation (living at bodily parents/mother/other) into account.

Statistical Analysis
We conducted all analyses with SPSS Version 25. Socio-demographic characteristics were analyzed using descriptive statistics (mean and standard deviation for continuous variables and frequency in percentage for categorical variables). We used chi2-tests and t-tests to determine gender differences in social support, social modelling, and physical activity outcome variables.
To analyze the different effects of parental and peer modelling and support on domain-specific physical activity participation stratified by gender, multiple separate logistic regressions were run with the different dichotomous physical activity variables as outcomes and one of the modelling and support variables as a potential correlate. In addition, age, socioeconomic status, region in Germany, residential area, migration background, siblings, and family situation were included as possible confounders in every model. From these logistic regressions, we obtained odds ratios which express the influence of different modelling and support scales on whether a participant is active or not. Higher values express a higher chance of being active in the specific domain with higher amounts of social support or positive modelling. Finally, we ran an additional overall model with every peer and parental modelling and support variable and confounders to determine the overall explained variance for each physical activity outcome variable (R 2 in Tables 3 and 4). For all analysis, the significance level was set to p < 0.05.

Sample Description
Data from 519 preschool children (244 girls and 275 boys) from the MoMo Study were eligible for analysis in the current study. Socio-demographic characteristics of the participating children are presented in Table 1. The mean age of the preschool children participating in the study was 5.40 (SD = 0.75) years, ranging from 4-6 years. The study population was comprised of 490 (94.4%) native German children and 29 children with a migration background. About 6.2% of the children were assigned to a low, 61.9% to a middle, and 31.9% a high socio-economic position. Most children lived with their biological parents (87.7%).

Social Support and Social Modelling in Relation to Gender
Physical activity, social support, and social modelling variables in boys and girls are presented in Table 2. In the current study, more than half of the preschool children regularly participated in physical activity in sports clubs (54.1%) and in outdoor play activities every day (60.7%). Additionally, 38.0% of the children regularly engaged in physical activity outside of sports clubs. Girls were more likely to participate in physical activity in sports clubs than boys (χ 2 = 4.406, p = 0.036), whereas no significant gender differences were found in physical activity outside of sports clubs and outdoor play. Regarding gender differences in social environmental variables, girls were more likely to get emotional support from their parents than boys. The highest support scores were found in parental informational support. 67.8% of the children had siblings who were regularly active, half of the children had regularly active mothers, and 42.6% had regularly active fathers.

Social Support and Social Modelling in Relation to Domain-Specific Physical Activity
The results of the logistic regression analyses are presented separately for boys and girls in Tables 3  and 4, respectively. Results were similar for boys and girls concerning physical activity in sports clubs: parental emotional, informational, and instrumental support and peer and sibling modelling were related to the outcome in both genders. Additional parental companionship support was only related to the outcome in boys (β = 2.06, p < 0.01).

Discussion
Using data from the MoMo Study, our present study aimed to identify social environmental correlates (social support and social modelling) of participation in physical activity in and outside of sports clubs and outdoor play. Overall, our results revealed that parental and peer support and modelling, as well as sibling modelling are related to preschool children's physical activity participation in different domains. Furthermore, we examined differences between boys and girls. However, the results are heterogeneous regarding types and providers of support and modelling behavior. It seems like parental support and modelling of peers and siblings were most important for supporting physical activity in sports clubs in girls and boys. As physical activity in sports clubs was the only domain of structured physical activity, this may indicate that preschool children are particularly in need of support to enable the participation in structured physical activity. Particularly parental instrumental support (e.g., driving children to sports facilities), which has been a strong predictor of physical activity in sports clubs in boys and girls, could be a necessary prerequisite. In contrast, social instrumental support might be less important to make unstructured activities possible, but could encourage them.
Regarding parental support, our results are in line with results published earlier: Loprinzi and Trost [40] found a positive relationship between parental support and preschool children's home physical activity, but not for MVPA in child care. Schoeppe and Trost [41] differentiated between maternal and paternal support for preschool children's physical activity and found that both maternal and paternal support were correlates of preschool children's physical activity, whereby stronger associations were found in girls in comparison to boys. However, they did not differentiate between different domains.
In the present study, parental modelling was only related to physical activity outside of sports clubs. Precisely, girls' physical activity outside of sports clubs was associated with maternal modelling and boys' with paternal modelling, which corroborates the same-sex hypotheses [18]. Other studies identified positive associations between parental and preschool children's physical activity [42,43]. However, Spurrier and colleagues [43] found correlations for parental modelling, particularly maternal participation in physical activity with preschool children's outdoor playtime. They concluded that during preschool, maternal modelling might have a stronger influence than paternal modelling, which has been found to be more relevant in older children [44]. Nevertheless, our study suggests that the identification with a same-sex model is relevant in preschool children.
Aside from parental correlates, we observed domain-specific relationships between peer and sibling social behaviors and preschool children's physical activity participation. Concretely, peer support was a relevant predictor of outdoor play, but not for physical activity in or outside of sports clubs, whereas peer modelling was relevant for physical activity in sports clubs and, in boys, also outside of sports clubs. Thus, peers already seem to be important role models and supporters for physical activity in different domains for preschool-aged children, as also shown in previous studies [45,46]. Preschool-aged children should have the chance to mingle with other children in the same or different age groups and to engage in shared physical activities and play.
Moreover, our results demonstrate that sibling modelling was associated with boys' and girls' physical activity in sports clubs and with outdoor play in boys only. Preschool children might be motivated to participate in physical activity in sports clubs if they have siblings that do the same. Additionally, younger children may be inspired by older siblings [47], as siblings are potential playmates and thus, facilitate the participation in physical activities together [48,49].
Regarding outdoor play, our study indicates different impacts of parents and siblings for boys and girls, respectively. In early childhood, parents are more protective of girls than boys [50], as fears of strangers or traffic danger are greater for girls than boys [16]. Therefore, parental companionship could be of greater relevance for girls, whereas sibling modelling could be more important for boys [16].
The strength of the current study is its examination of the relationships between the social environment and different domains of physical activity and outdoor play in a nationwide sample of preschool children. We showed that these relationships were different regarding different physical activity domains, and the mechanisms of social influences on children's physical activity participation seem to differ between physical activities in different domains. Thus, our study contributes to a better understanding of social environmental correlates of physical activity by taking domain-specific physical activity into account instead of focusing on overall physical activity or on MVPA. Furthermore, since our data were drawn from the nationwide MoMo Study conducted in 167 communities in Germany, the results of this study are highly representative.
Nevertheless, some limitations of the study have to be mentioned. First, the data are cross-sectional and do not allow for the analysis of causal relationships. Therefore, we do not know the direction of the relationships found, and it is also possible that the physical activity behavior of the children influenced the social environment instead of the other way around [13,51]. Second, all data was captured from proxy reports and is prone to bias. For example, it is possible that parents misjudged the data. In addition, we did not capture how many fathers and mothers participated in the study, but fathers and mothers may have answered the questionnaire items differently. Furthermore, it is possible that the children did not perceive the same level of support and modelling as gauged by their parents. Thus, we do not know how the questionnaire was answered if there was a disagreement between the child and its parent. Third, only unspecified social support and modelling from parents, peers, and siblings has been considered, and no domain-specific support and modelling data was captured (e.g., parental companionship support for outdoor play or peer support for physical activity in sports clubs). Thus, odds ratios may have been underestimated in our study. Analyzing the relationships of more specific measures of social support and social modelling with the domain-specific physical activity outcomes might lead to even greater odds ratio estimates as found in the current study.

Conclusions
The present study provides nationwide data from Germany and showed differences of social environmental relationships with physical activity in different domains and in preschool boys and girls. Subsequently, further research should take different domains of physical activity into account to enable the development and successful implementation of programs to promote physical activity. In interventions to promote physical activity in preschool children beside the domain of physical activity, the providers of support and modelling should be targeted. Showing that peer modelling and support were related to a number of domain-specific physical activity measures and interventions including peer groups could be promising in the promotion of physical activity. For the participation in organized sport activities in sport clubs, parents and peers are important providers of support and modelling for both genders and, thus, can be objectives in intervention programs.

Acknowledgments:
The authors wish to thank the children and parents who participated in the MoMo Study, and the team of dedicated study staff for their assistance in data collection. Additionally, we would like to thank Lisa Christine Peterson for her language editing service and Katrin Käppner for her assistance with formatting of the document.

Conflicts of Interest:
The authors declare no conflict of interest.