Background
Adolescents are becoming increasingly overweight and obese in rich as well as poor societies worldwide [
1]. For example, in the US between the years 1980–2000 rates of obesity among adolescents 12–19 years old more than tripled [
2]. Further, between the years 2003 and 2007 rates of obesity among adolescents 10–17 years old in the US increased by 6.1 % among boys and 17.6 % among girls [
3]. In Europe, the annual increase in prevalence of childhood overweight was 0.1 % during the 1970s, 0.4 % during the 1980s, 0.8 % during the 1990s and 2.0 % by the 2000s [
4]. These patterns are of clinical as well as public health concern, since obesity in childhood is associated with both short- and long-term health consequences [
4‐
6].
Childhood overweight and obesity are the consequences of excessive energy intake relative to energy output together with a genetic disposition [
6]. The mechanisms for such imbalance are complex though, involving both individual- and societal level factors. Individual factors include dietary intake characteristics e.g., frequency of eating, portion sizes, energy density, high fat intake and high intake of sodas and sugar. Dietary intake characteristics in childhood and adolescence are important to evaluate and act on if necessary, because they also predict dietary intake in adult life and thereby risk of chronic diseases [
4,
7].
In Sweden, unfortunately not many large and comprehensive population-based surveys of dietary intake among adolescents have been carried out during the last decade. The National Food Agency in 2005, 2006 and 2008 carried out population-based telephone surveys of indicators of a healthy diet among Swedes aged 16–80 years. In 2008, in total only 211 persons in the age group 16–29 years were included [
8]. Also, at three times (1989, 1997–1998 and 2010–2011), The National Food Agency has carried out more comprehensive dietary intake surveys among about 2000 Swedes; however the number of participants in the younger age groups are too small to be evaluated separately. Further, since 1985 the National Institute of Health in collaboration with the World Health Organization (WHO) carries out national lifestyle evaluations every four years. In 2009–2010, dietary intake and physical activity among school children 11–15 years old were evaluated as part of such a survey [
9]. In addition, since 2004 the National Institute of Health yearly carries out a survey among 20,000 Swedes aged 16–84 years on lifestyle characteristics. However, this includes questions on frequency of intake of fruit and vegetables as sole indicators of dietary intake.
On the local level, diet information may also exist. For example, in the Gothenburg school district, school nurses have since 2004 systematically collected data on indicators of diet in addition to regular anthropometric measurements. Youth aged 11, 15 and 17 years answer questions on whether they daily consume breakfast, lunch and/or dinner as sole indicators of a healthy diet [
10].
In short, there is a dearth of comprehensive information on a sufficiently large number of participants on dietary intake and food choices among adolescents in Sweden that go beyond a few indicators of a healthy diet. If the current obesity epidemic is to be reversed and future chronic diseases prevented, comprehensive dietary information is urgently needed. In the current study, the aim was to describe dietary intake and food choices as well as leisure time activities in relation to overweight and obesity in a total sample of all schoolchildren aged 15 years in Western Sweden.
Discussion
A comprehensive survey of dietary intake and lifestyle factors was carried out among the total population of adolescents in Västra Götaland Region, Sweden (n = 21 651 in 2008). Important findings included that more girls than boys reached the goal of consuming fruit and vegetables at least daily (49.5 % vs. 34.4 %), that only 15 % of both sexes reached the goal of consuming fish at least twice weekly and that about two-thirds of both sexes reached the goal of regular moderate or vigorous leisure time physical activity weekly. In total, 12.4 % of all adolescents were overweight and 2.4 % were obese; both being more prevalent among boys than girls. Girls reported a proportionally higher intake of carbohydrates, fibre and sucrose, whereas boys reported a proportionally higher intake of protein as well as total and saturated fat. Boys reported more frequent consumption of sodas and concentrated fruit juices, milk with 3 % fat, bread, potatoes and fast food. Significant risk factors for being overweight or obese included living in rural areas, living in apartments and reporting no frequent leisure time physical activity.
We have chosen to compare our results with indicators of nutrient quality (ie, intake of fruit and vegetables and fish) developed by the National Food Agency [
8] as well as with the Nordic Nutrition Recommendations on consumption of fat, saturated fat, sugar and fibre [
21]. In addition, we have highlighted the consumption of butter and margarine because of trends in fat consumption recently noted among Swedish adults [
22]. Based on population-based cross-sectional consumption data collected with the same methodology during 1986–2010, this study described a significant decrease in fat consumption between 1986 and 1992 and a significant increase from 2002 for women and from 2004 for men. During this period, among both sexes use of butter as bread spread remained constant, use of low fat margarine decreased and use of butter-rapeseed oil blend increased. Concurrently, serum cholesterol levels decreased during 1986–2004 and increased after 2007. Hence, it is of interest to investigate recent patterns of fat intake also among Swedish adolescents.
The study is unique in that it invited the total population of 21, 651 adolescents in Västra Götaland Region. The participation rate was 54.3 %, which may be compared with the Swedish national dietary survey Riksmaten, with a participation rate of 36 % [
23]. Thus, the participation rate in our survey was in the upper range of what could be expected. Still, due to the varying participation rates among subgroups of adolescents, our results are likely not representative for foreign-born adolescents or adolescents with parents of short education.
Research has shown that by the age of 8–10 years, children can report their food intake as reliably as their parents and by adolescence cognitive abilities are fully developed, although issues of motivation and body image may hinder adequate reporting [
24]. Thus, recall errors are to be expected and include both underreporting and overreporting. Previous research on adolescents in Göteborg, Sweden, found that over-reporting was more common among boys and underreporting was more common among girls [
25]. Also among adolescents, underreporting tends to increase with increasing overall eating frequency as well as with higher body mass index, and less common foods are more easily forgotten than are main courses [
24,
26]. Still, modified dietary history and food frequency questionnaire are likely able to rank adolescents according to most nutrients [
25,
27], and adolescents do have the capacity to use photographs to estimate portion sizes [
28].
In our survey, energy intake was underreported as indicated by differences between reported energy intake and calculated total energy expenditure; a difference similar to that reported by others for this age group [
26]. In that study, underreporting of energy intake among adolescents from seven days food records, in comparison with total energy expenditure measured by the doubly labeled water method, was 78.3 % for girls and 81.9 % for boys. In our survey, the magnitude of underreporting for other nutrients is probably similar to that of energy. In our evaluations of macronutrient intake, all values are expressed as percentage of energy. Hence, effects of overall underreporting are less likely to bias comparisons of macronutrient intakes, although possible bias due to selective underreporting remains unknown. Still, diet intake data based on food frequency questionnaires with a limited number of food items are not expected to cover the total daily intake. Hence, such data should not be used in comparisons with absolute recommendations but are suitable for comparisons between subgroups, which is how data are used in this paper.
The sex-specific consumption patterns observed in our study are supported by previous Swedish findings. A national survey in 2009–2010 among 11, 13, and 15 year old school children noted that girls reported a higher consumption of fruit and vegetables whereas boys reported a higher consumption of sodas and candies and also more time on physical activity and computers [
9]. These patterns were established already among the 11-year-olds and differences increased with age. Such patterns also have been reported from other Nordic countries [
29]. Also, the clustering of unhealthy consumption patterns in this age group has been reported also in other Nordic countries, e.g. Norway [
30,
31].
Results on reported consumption of healthy and unhealthy food items among Swedish adolescents vary. In the national survey in 2009–2010 [
9], among the 15-year-olds only 27 % of the girls and 22 % of the boys reported consuming fruits daily or more often and only 39 % of the girls and 31 % of the boys reported consuming vegetables daily or more often. Furthermore, a survey from Gothenburg, Sweden, among 18-year-olds [
32] reported a daily intake of fruit and berries among 40 % of the girls and 21 % of the boys, and a daily intake of vegetables among 59 % of the girls and 36 % of the boys. Our recent survey thus found higher proportions of both girls (61 %) and boys (39 %) reporting at least daily consumption of fruits, and higher proportions of both girls (69 %) and boys (54 %) reporting at least daily consumption of vegetables. In our survey, sodas and concentrated fruit juices consumed at least daily were reported among 10 % of the girls and 20 % of the boys, which is slightly higher than previously reported in a survey from Gothenburg, Sweden [
33] (girls 3 %, boys 8 %) and in the national survey in 2009–2010, (girls 6 %, boys 11 %) [
9]. International comparison reveals that the proportion reporting daily fruit consumption in Sweden was below the international average for both sexes for all 40 countries included in the WHO international comparison [
9].
Possible explanations for the more healthy consumption patterns reported in our survey include the lower representation of foreign-born adolescents and adolescents with parents of short education. Also, the proportion of adolescents living in apartments in our survey (19.3 %) was lower than the corresponding national average for 0–24 years living in apartments (33 %) [
34]. Finally, it is possible that our survey captures a trend over time of improved dietary intake in this age group. Previous data on secular trends of vegetables and fruit consumption among adolescents in Sweden is sparse. According to the annual surveys from Public Health Agency of Sweden the consumption of fruit and vegetables 1–5 times/day has increased the last ten years among boys 16–29 years of age, although the increase is negligible among girls [
35].
Among girls, we found lower prevalence of both overweight (9.9 % vs. 15.1 % among boys) and obesity (1.9 % vs. 3.0 % among boys). Similarly, the national 2009/10 survey found a prevalence of self-reported overweight and obesity of 20 % among boys and 8 % among girls [
9]. The survey was part of an international initiative and presented an average prevalence of overweight and obesity among 15-year olds in 40 countries of 21 % among boys and 12 % among girls [
36]. Further, prevalence of measured overweight and obesity among 16-year-olds in a small national school based sample from 2001 were almost identical with that of ours for boys but somewhat higher for girls [
37]. Likewise, more recent measured data on adolescents from southeastern Sweden reported similar prevalence of overweight and obesity for boys but again somewhat higher for girls than our survey [
38]. In contrast, in an environment of high education in Uppsala, Sweden, the prevalence of measured overweight and obesity was much lower for 16-year old boys but similar to our prevalence for girls [
39]. In sum, the higher prevalence of overweight and obesity among boys found in our study are consistent with that of many other reports and warrant attention.
In our survey, significant risk factors for being overweight and obese included living in rural areas, living in apartments and reporting no frequent leisure time physical activity. Surprisingly, infrequent intake of fruits, vegetables and fish were not significant risk factors in the multivariate models. Only these indicators of a healthy diet were included in these analyses; hence it is possible that other components of the diet may have relation with risk of being overweight and obese.
The Norwegian national survey identified social class, time spent watching TV or in front of a computer and breakfast frequency as risk factors for being overweight among 13 -year-olds [
27]. Further, more frequent TV viewing in adolescence and early adulthood was in the 1958 British birth cohort associated with greater BMI gains through to mid-adulthood and with central adiposity in mid-life [
40]. In Norway, among young teenagers a twenty-month change in screen time was positively associated with changes in the consumption of sodas and unhealthy snacks in the same period [
41]. Clustering of extended screen time and high consumption of sodas as well as infrequent breakfast habits has been reported among 11–12 year olds in Gothenburg, Sweden [
42]. In our survey, extended screen time was in bivariate analyses significantly associated with increased risk of being overweight or obese. However, the significance disappeared in multivariate analyses, perhaps due to collinearity with factors such as time spent outdoors and regular physical activity. We noted a sex difference with boys reporting more screen time than girls, especially for time on computer games. In Sweden, total screen time has increased markedly for 15-year-olds since 2001 even though time for TV/DVD takes a smaller part [
9]. From a health perspective, this is a reason for concern.
In summary, our results on reported intake of healthy and unhealthy food items among adolescents in western Sweden describe sex differences similar to those of previous Swedish studies, with girls reporting more healthy habits than boys. Overall, our results point to a somewhat more healthy consumption pattern than found in previous studies, although this may partly reflect misreporting and selection bias. Prevalence of overweight and obesity support previous findings, and risk factors for being overweight and obese include socio-demographic characteristics such as living in rural area and living in apartments, and having low reported frequency of leisure time physical activity. The latter likely also reflects more frequent screen time, a phenomenon that has increased in this age group in Sweden in the last decade. In a public health perspective, these findings are a reason for concern. Public health actions should be taken to increase physical activity and consumption of vegetables, fruit and fish, as well as decrease consumption of sodas and candies with special attention to sex differences.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
AW, BH, KT, JT and HBF designed the research; JB and HBF oversaw data collection; IJ was responsible for the dietary database; AW and HBF performed statistical analyses, wrote the manuscript, and had primary responsibility for the final content; and AW, BH, JLK, IJ, KT, JB and HBF were involved in revision of the manuscript and the final version of the manuscript. All authors read and approved the final manuscript.