Background
Childhood overweight and obesity have been increasing dramatically worldwide, even in young children. Despite a possible stabilization, the high prevalence remains a great public health concern [
1]. Among several environmental factors, a sedentary behaviour (especially TV viewing) and a reduction in physical activity (PA) are implicated in this increase in body fatness [
2]. Some data indicate that children have become less physically active [
3‐
5]. One potential environmental influence on children's PA that may be implicated in this decrease is the amount of time spent outdoors [
6]. It has recently been reported that 3- to 5-year old children spend around 80% of their time in activities classified as sedentary or at most light PA [
7]. Furthermore, a trend towards a decline in motor performance has already been noticed in young children [
8].
Individual socio-cultural determinants such as selected parental characteristics (migrant background, low educational level and high workload) are known risk factors for childhood overweight/obesity and sedentary behaviours [
9‐
12]. In addition, there exist also regional variations of overweight and PA [
13,
14]. Even within Europe, prevalence in overweight differs among countries, ranging from 5 to 25% [
13,
14]. However, there is a lack of data on the respective roles of individual and environmental determinants in a well defined setting. With its linguistic and cultural diversity, Switzerland offers the opportunity to study diverse cultural environments within the same country. It can thus serve as a model to examine the impact of the regional environment on adiposity, PA and motor skills in children.
In the present study, we assessed differences in adiposity, objectively measured PA, sedentary behaviour and agility performance in preschool children according to different socio-cultural determinants (parental migrant status, education, and workload) and the regional environment.
Discussion
Our results strengthen the importance of the regional environment on lifestyle behaviours and obesity in very young children. The main findings of this study are: Within the same country with the same national health policy, preschool children from the French speaking, south-western part of Switzerland had an increase in adiposity, were more sedentary, less physically active and less agile compared to the German speaking, north-eastern part. Those differences were in the range of 10% or more and persisted after adjustment for parental characteristics. On the other side, parental characteristics like migrant status, low educational level or workload had less impact on adiposity and lifestyle behaviour and their impact was attenuated after adjustment for the regional environment.
Children of migrant parents had more body fat, were more sedentary, less active, and less agile than children of Swiss parents. These findings are in agreement with several studies, which reported that children with a migration background had higher BMI and were less active [
10,
11,
27,
28]. Maternal workload and low parental educational level resulted in differences in some PA measures and/or in agility performance, but, in contrast to previous studies [
10], not in adiposity. Ethnic and genetic factors can contribute to the increase in adiposity in some migrants, while economic reasons, lack of parental time and support and individual socio-cultural attitudes might also explain some of our data [
29]. We could hypothesize that the healthy behaviour found in children of part-time working mothers could be a combination of time, motivation and sufficient financial resources. Parental characteristics had a particularly strong influence on the time spent watching TV. Existing differences in parental characteristics between the two regions could theoretically explain the observed differences in the regional environment. However, the impact of the regional environment on adiposity and all lifestyle behaviours except TV viewing was much larger and persisted after adjustment for the above mentioned parental characteristics.
The regional environment represents the broader social, cultural, economic and built environment, within which individual behaviour occurs [
30]. Differences in the socio-cultural environment could explain our findings which are also in concordance with the observed North-South gradient in overweight and obesity within Europe [
13,
14]. In those studies, a particularly high prevalence of overweight/obesity was described among southern European countries [
13,
14]. This North-South gradient has also been reported within Italy [
31]. Moreover, lower levels of overweight and obesity were found among children in countries of Central and Eastern Europe compared to Western Europe [
13,
14]. Different regional patterns have been observed for reported PA in children across European countries [
13,
32,
33]. According to our results, the German speaking part of Switzerland reflects the trends in the north-eastern countries, whereas the French speaking region reflects those in the south-western (and Latin) European countries. Even within the same country, with almost identical climatic conditions, we could observe this North-South gradient in two urban areas. This could be explained by different prevailing cultural norms, priorities, attitudes and beliefs. Indeed, differences in health beliefs have been associated with differences in PA in young adults [
33]. Furthermore, similarly to many northern countries, general efforts in health promotion have a stronger tradition in the north-eastern part of Switzerland compared to the south-western counterparts. Indeed, differences in health between the German and French speaking part of Switzerland have been shown to reflect broader European patterns [
26]. Similar differences in PA within Switzerland were also observed in large representative samples of adults, where PA levels, reported by questionnaires, were found to be higher in citizens of the German compared to the French speaking part of Switzerland [
26,
34]. In addition, reported time spent playing outdoors has been shown to be higher in school-aged children and adolescents of the German compared to the French speaking part of Switzerland [
35]. Similarly, substantial differences in "active transportation to school" have been documented between German- and French speaking school children and adolescents that were independent of socio-demographic characteristics and environmental factors like distance to school or bike availability [
36]. On the other side, lifestyle behaviours in different regions within the French or within the German part of Switzerland, respectively, are reported to be very similar [
26,
34,
35]. In our sample, the observed regional differences were comparable, if the whole population or the subgroup of non-migrant children were analyzed. On an even larger perspective outside of the health sector, differences between the German and French speaking part exist regarding educational and employment levels [
37]. But also public transport and school system differ between the two parts and are possibly tightly linked to the key outcome measures of our study. For example, preschool in the French part of Switzerland resembles primary school more than it does in the German part of Switzerland (choice of academic objectives, classrooms).
A range of social and environmental factors have been described as potential influences on children's PA and sedentary behaviours [
38]. The social environment includes social factors in the home and neighbourhood environments, as well as social peer networks. Within the social environment, the social network plays a major role. In adults, it can even influence adiposity more strongly than first grade relatives [
39]. Since diverse phenomena can spread within social networks, it seems also possible that obesity or PA might spread from person to person [
39]. Adults as well as children are embedded in social networks and are influenced by the evident behaviours of those around them. This fact suggests that weight gain or activity levels in one person might influence weight gain or activity levels in others. Having obese or sedentary social contacts might influence the adoption of such specific behaviours. In adolescent girls, the role of peer social network factors is crucial for participating in PA [
40]. It is quite possible, that also in younger age not only home support and parental modelling, but also larger social peer networks play an important role [
38].
The built environment has also been associated with PA and adiposity and includes access to recreation facilities, parks, playgrounds and traffic [
30]. Our regional differences in the outcome parameters observed within the same country could also be influenced by differences in the built environment. Observed differences in time playing outdoors may reflect potential disparities in the access to playgrounds between the two regions, but might also be influenced by the social environment. In this age group, playing outdoors has been shown to be a good marker of PA [
6]. Perceived barriers for children to play outdoors like road traffic, lack of playgrounds or courtyards and danger of crime did not differ between both regions, but the parental perception of barriers might be influenced by socio-cultural factors. Furthermore, ways of commuting and school-based structured PA as important factors of PA did not have a large impact on our results, as the great majority of preschoolers in both areas (90%) still walked to school and physical education at school did not differ.
Strengths and novelties of the study include the joint comparison of regional and individual determinants within the same small country taking advantage of two distinct socio-cultural regions of similar size and with a similar climate. Further strengths are the comprehensive assessment of adiposity and objectively measured PA and the inclusion of motor skills in a very young population. In this study, we focused on socio-cultural characteristics and therefore we did not take into account other essential factors that influence lifestyle and adiposity, such as parenting practices and believes. On the other side, parenting practices and believes may also be influenced by individual and regional socio-cultural characteristics. Since our investigation is not based on a representative sample, we cannot be certain to draw conclusions for the whole population, although differences in reported PA between the French and German regions in large representative samples of adults and in a population of school-aged children and adolescents confirm our findings [
26,
34‐
36]. Another limitation is the cross sectional design of the study that limits the investigation of clear cause-effect relationships.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
JJP and SK and designed the study. JJP was the principal investigator and is guarantor. JJP, SK, FB, IN, VE, UM, UG and PM established the methods and questionnaires. FB, IN, VE and JP were the main coordinators of the study. FB, IN, VE, UM, PM and JJP conducted the study. PM gave statistical and epidemiological support. FB wrote the article under the assistance of JJP and got additional help from SK, UG and PM. JJP obtained the funding, with the assistance of SK. All authors provided comments on the drafts and have read and approved the final version.