Journal of Public Health Advance Access published online on May 8, 2008
Journal of Public Health, doi:10.1093/pubmed/fdn037
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Overweight and obesity among adolescents in Norway: cultural and socio-economic differences
Else-Karin Grøholt, Researcher1
Hein Stigum, Senior Researcher2
Rannveig Nordhagen, Senior Researcher3
1 Department of Health Statistics, Division of Epidemiology, Norwegian Institute of Public Health, PO Box 4404, Nydalen, 0403 Oslo, Norway
2 Department of Chronic Diseases, Division of Epidemiology, Norwegian Institute of Public Health, PO Box 4404, Nydalen, 0403 Oslo, Norway
3 Department of Genes and Environment, Division of Epidemiology, Norwegian Institute of Public Health, PO Box 4404, Nydalen, 0403 Oslo, Norway
Address correspondence to Else-Karin Grøholt, E-mail: else-karin.groholt{at}fhi.no
Background The aim of this study was to investigate overweight and obesity among a representative population of 15 966 Norwegian 15–16 year olds and the associations with different socio-economic and cultural risk factors.
Methods Self-reported data were obtained from school-based surveys in six counties during 2000–04. Overweight and obesity were calculated using Cole's index.
Results The prevalence of overweight and obesity were 11.8% and 2.4%, respectively, higher among boys. Logistic regression analyses revealed that adolescents in Nordland, Troms and Finnmark (the northernmost counties) were 70–90% more likely to be overweight and obese compared with adolescents in Oslo (the capital and southernmost county) (OR for overweight in Finnmark = 1.7, CI = 1.3, 2.3). Lower educational plans and poor family economy were both significantly associated with overweight and obesity. So was physical inactivity (OR = 1.2, CI = 1.1, 1.3 and OR = 1.6, CI = 1.2, 2.1, respectively). Eating breakfast was positively associated with not being overweight/obese.
Conclusion Overweight and obesity is associated with socio-economic factors and with factors related to food habits and nutrition, suggesting important areas for prevention.
Keywords: epidemiology, obesity, public health
 |
Introduction |
|---|
 TOP Introduction MethodsResultsDiscussionFundingReferences |
|---|
Childhood overweight is rapidly emerging as a global epidemic with profound public health consequences. It has been estimated that about 10% of the world's school-aged children carry excess body fat,1 and several studies have found an increasing prevalence of overweight among children and adolescents.2,3 This has been prominent in USA, but also in European countries.
The factors accounting for the increase in overweight and obesity over the past 30–40 years is not fully understood. Evidence from previous studies suggests that the increase might be a result of an interaction between hereditary and environmental factors in the way that several genetic variants interact with an ‘at-risk’ environment.4 International research have identified several socio-economic and cultural factors that might be associated with childhood overweight,1 but few corresponding studies have been performed in Norway.
Health surveys among 40 year olds have shown that body mass index follows a normal distribution, but that the curve has been displaced to the right during the last decade.5 There is reason to believe that this might also be the case for children. It has been found that Norwegian conscripts born in 1980 have higher body mass index (BMI) compared with conscripts born in 1967.6 A substantial increase in the upper BMI-percentiles over time has also been found among adolescents.7 In addition, a significant increase in weight for height has been found among 4–15-year-old children over the three last decades, the weight gain being most pronounced in the pre-adolescent group.8 Andersen et al.3 found that the prevalence of overweight increased from 7.8% to 11.5% (girls) and from 7.3% to 11.5% (boys) during 1993–2000 and that social class, breakfast frequency and physical inactivity were associated with overweight.
Overweight and obesity are associated with increased risk of coronary heart disease, hypertension, hyperlipidaemia and type 2 diabetes mellitus.1 Overweight children have also an increased risk of becoming overweight as adults.9 From a preventive perspective, factors that may influence the development of overweight among children and adolescents are therefore important to study.
The aim of this study is to analyse the prevalence of overweight and obesity among adolescents in Norway. We will look for regional differences, and analyse the associations with socio-economic factors, nutrition and physical activity.
|
Methods |
|---|
|
TOPIntroductionMethodsResultsDiscussionFundingReferences |
|---|
The study protocol was approved by the Norwegian Data Inspectorate and cleared by the Regional Committee for Medical Research Ethics. The study has been conducted in full accordance with ethical principles as per the World Medical Association Declaration of Helsinki.
Subjects
The data were obtained from school-based surveys in six (of totally 19) counties in Norway during 2000–04: Oslo 2000–01, Oppland 2001, Hedmark 2002, Nordland 2004, Troms 2003 (city of Tromsø 2002) and Finnmark 2003. These counties represent the capital, two inland southern and three costal northern counties, respectively. All adolescents in the 10th grade (last year in compulsory school) in the respective counties were invited to fill in a questionnaire about health, lifestyle, welfare and living conditions. The majority was 15–16 years old and had reached puberty. Less than 1% of students were younger than 15 years or older than 17 years.
Mean response rate was 86.4% and varied from 71% to 90%, highest among students in Oppland and lowest in Finnmark. The low response rate in Troms and Finnmark might be caused by the fact that 9.5% (7 of 74) and 23% (12 of 52) of the schools, respectively, did not participate in the school-based survey. The pupils at these schools instead received a postal questionnaire.
A total of 15 966 adolescents answered the questionnaire, 7342 in Oslo, 1939 in Hedmark, 1877 in Oppland, 2657 in Nordland, 1514 in Troms and 637 in Finnmark. The number of respondents reflected the population size in the six counties.
Outcome variables
The outcome variables were either overweight or obesity or BMI.
Weight and height were reported according to the questions: ‘What was your weight the last time you weighed yourself?’ and ‘What was your height the last time you measured your height?’ The weight was given in kilograms and the height in centimetres. BMI was calculated by dividing weight in kilograms with squared height in metre (weight/(height*height)). Overweight and obesity were calculated according to Cole's index.10
Explanatory variables
The explanatory variables in the models were county, socio-economic factors, physical activity, TV/PC use, dietary habits and nutrition. Other independent variables that were included in the models were place of residence, sex, self-reported health and smoking.
Two socio-economic factors were included: the adolescent's educational plans11 and the family's economic situation. Educational plans were divided into four different categories: (i) university or college, (ii) upper secondary (general), (iii) upper secondary (vocational) and (iv) other educational plans/not decided yet. The family's economic situation reported by the adolescents was coded as: (i) good/excellent, (ii) average and (iii) poor. Physical activity was measured according to two variables: transport to school and physical activity outside school time. Transport to school was categorized into (i) bus, car or moped and (ii) bicycling or walking. Physical activity outside school time was measured according to how many hours per week the adolescents were active in athletics/sports: (i) 
8 h per week, (ii) 3–7 h per week and (iii) 0–2 h per week. TV/PC use was measured according to how many hours per day (outside school time) the adolescents were watching television, video or playing computer games, and was categorized into: (i) 0–2 h per day and (ii) 3 h per day. Food habits were measured according to how often the adolescents used to eat breakfast and lunch/packed lunch. These two variables were divided into three categories: (i) 5–6 times/week, (ii) 1–4 times/week and (iii) seldom/never. The nutrition variables were derived from factor analysis based on 16 questions about how often the participants used to eat different kinds of food: fruit/soft fruit, cheese, potato, cooked vegetables, uncooked vegetables, fish, whole milk, semiskilled milk, skimmed milk, water, juice, chocolate/sweets, chips, mineral water with sugar, mineral water without sugar and squash. Factor analysis revealed two distinct factors: (i) one factor made out of chocolate/sweets, chips, mineral water with and without sugar, labelled ‘unhealthy diet’ and (ii) one factor made out of the other variables labelled ‘healthy diet’. In the logistic regression model, these factors were categorized into quartiles, in the linear model as continuous variables. Place of residence (coded rural or urban) was based on information about the geographic area of the schools in the northern counties and the geographic area of living in the southern counties. Self-reported health was divided into two categories: (i) good/excellent and (ii) not so good/bad. Smoking habits was categorized into: (i) never smoked/stopped smoking and (ii) daily/sometimes smoking.
Statistical analysis
Unadjusted analyses were carried out using contingency tables with percentages and chi-square tests. The adjusted associations between overweight and obesity and the explanatory variables are expressed as odds ratios (OR) with 95% confidence intervals (CI), using logistic regression. Multiple linear regression analyses were also performed to estimate the association between BMI and the same explanatory variables. We performed regression diagnostics by plotting delta-beta (for each beta-estimate) versus observation number to look for data points with strong influence. Variables not significant at the 25% level were excluded from the models. The analyses were performed using SPSS v 12.0.
|
Results |
|---|
|
TOPIntroductionMethodsResultsDiscussionFundingReferences |
|---|
Tables 1, 2 and 3 show the estimated associations between overweight and obesity and the explanatory variables based on one logistic regression model with all covariates included. Table 4 shows the estimated associations between BMI and the same independent variables in a linear regression model.
|
|
|
|
Overweight and obesity in different regions in Norway
The prevalence of overweight and obesity was 11.8% and 2.4%, respectively (Table 1). Adolescents in the two inland and the three northernmost counties, respectively, were about 20–30 and about 70–90% more likely to be overweight and obese compared with adolescents in Oslo.
The prevalence of overweight and obesity was 14.8% and 2.9% among boys and 8.8% and 1.9% among girls, respectively. The association between overweight and obesity and the explanatory variables did not differ between boys and girls in the study (data not shown).
The association between overweight and obesity and socio-economic factors
Adolescents planning a vocational education had about 1.3 times the odds of being overweight and about two times the odds of being obese compared with adolescents planning a university or college education.
Compared with adolescents living in families with good or excellent economy, the odds of being overweight and obese was about 1.3–1.5 times higher among adolescents living in families with poorer economy.
The association between overweight and obesity and factors related to physical activity
Table 2 shows that adolescents using bus or car transport to school were about 30% more likely to be obese compared with adolescents who cycled or walked (CI = 1.0, 1.6). Furthermore, whereas overweight increased moderately with declining physical activity outside school, the association with obesity was much stronger, with a doubling from the high to the low activity group.
There was also an independent association between overweight and obesity and TV/PC use. Adolescents watching television/video 3 h or more per day were about 60% more likely to be obese compared with adolescents watching television/video 0–2 h per day (Table 2).
The association between overweight and obesity and factors related to food habits and nutrition
Breakfast and lunch habits were associated with both overweight and obesity (Table 3). Results from the adjusted analyses showed that overweight and obesity were associated with reported consumption of healthy food. In addition, overweight and obese adolescents were less likely to report eating unhealthy food (Table 3).
The association between BMI as a continuous variable and the independent variables in the model
Linear regression analyses supported the findings from the logistic regression. Table 4 shows that when all covariates are at their reference category, the expected BMI is 20.7 (the constant term). Respondents from the northern county Finnmark were about 1 unit higher in BMI, and girls were 0.7 units lower. The BMI for any covariate combination can be found simply by adding the corresponding terms together. The delta-beta plots showed no data points/subjects with undue high influence.
|
Discussion |
|---|
|
TOPIntroductionMethodsResultsDiscussionFundingReferences |
|---|
Main finding of this study
This study presents data about overweight and obesity among adolescents in Norway. The overall prevalence of overweight and obesity was 11.8% and 2.4%, respectively, and higher among adolescents living in northern compared with southern regions. Overweight and obesity were in addition associated with low socio-economic status and with factors related to physical activity, food habits and nutrition. These associations were also found to include the whole spectrum of BMI (when measured as a continuous variable), and not merely the upper end (defined as overweight and obesity).
Limitations of this study
Even if the response rates were relatively high, the data might be biased by different health-profiles among responders and non-responders in the study. It might not be fortuitous who was present and who was not present at school the current day of the survey. The lower response rate in Finnmark (71%) might partly be due to the fact that a quarter of the schools did not participate in the school-based survey. The effect of response bias, however, cannot be large. If the average BMI was 21 among responders, and we assume that it was 25 among non-responders, the population average BMI would be 86.4%*21 + (100% – 86.4%)*25 = 21.5.
Calculation of BMI in the study was based on self-reported height and weight the last time of measurement. ‘Last time’ was not clearly defined, and the self-reported data might therefore date from different points of time. Earlier studies have shown that self-reported height and weight correlate highly with measured data in adolescence, but the validity of self-reports has also been questioned.12,13 However, in a Norwegian study, sensitivity and specificity according to identified overweight based on self-reports were found to be 83% and 100%, respectively.3 The gender differences in our study might partly be a result of unequal self-reports of height and weight among girls and boys.
What is already known on this topic
The overall prevalence of overweight found in the study is in accordance with the International Obesity Task Force (IOTF) and with what is found in other Nordic studies.1,3 We may assume that a small country like Norway, though geographical divergent, have a relatively genetically homogenous population. It has been shown that adults in the northern counties often have unfavourable outcomes on health and lifestyle factors,14 which probably also is the case for adolescents. The regional differences found are thus most probably caused by local factors acting in different ‘obesogenic’ environments.15
The results in our study confirm previous findings showing that breakfast was dropped more often among overweight and obese adolescents.16 Breakfast skipping is a habit that seems to increase with age and be associated with other lifestyle factors detrimental to health.17,18 However, eating breakfast per se has not necessarily been related to lower BMI.18 Particularly, the cereal content has been shown to be important,19,20 and it has been found that eating cereal, whether cooked, ready to eat, or quick breads for breakfast is associated with significantly lower BMI when compared with either skipping breakfast or consuming other types of breakfast.19 The cereal content in the traditional Norwegian breakfast is usually high, and may underline the importance of healthy breakfast habits among Norwegian adolescents. Our contradictory finding of an inverse relationship between overweight and obesity and reported intake of food might be explained by confounding effects of underreporting or changed dietary pattern as a consequence of dieting.
Our study confirms the findings from previous studies showing that physical inactivity promotes whereas physical activity protects against overweight and obesity.1,3,21,22 However, in our study hours in front of the TV/PC per day was used as a proxy for physical inactivity. Physical activity and TV/PC use had independent effects on overweight and obesity, suggesting different causal pathways. TV/PC use may, for instance, have a direct effect through reduced energy expenditure. In addition, it may be associated with increased intake of sweets and snacks, both due to changed eating behaviour and to greater exposure to advertisements for foods high in sugar and fat content.
Overweight and obesity are, like other health measures, related to socio-economy as the highest prevalence is found in the lowest socio-economic groups. This is well known from previous studies.1,23 Measuring socio-economic status in adolescence is, however, difficult, 24,25 and health in youth might be characterized more by the absence than the presence of class variation.26 One important objection to this theory is the relative role of parental social class versus current socio-economic circumstances of young people themselves. In our analyses, perceived family economy was used as a proxy for parental socio-economic position and educational plans as a proxy for current socio-economic position in adolescence. Both were subjectively evaluated, and might be considered as the adolescents interpretation of own and parental socio-economic status. Plans of an academic education have, however, been found to be stable over time, and track into studying rather than working at the age of 18.11 The prevalence of overweight and obesity has also been found to be higher among adolescents in vocational secondary education schools compared with other schools.27
In our industrialized part of the world, availability and consumption of energy dense food and drink is increasing, and the demands of physical activity are less important than ever before.28 In many ways, we might be approaching an ‘obesogenic’ environment.15 However, children and adolescents are also to a certain degree influenced by the family environment, for example, the family's habits according to nutrition, physical activity and television viewing. Such family habits are, however, in turn influenced by socio-economy, and many of the factors related to an ‘obesogenic’ environment seem to be adversely associated with groups in the lower social strata. Previous research (including this) has in many ways concentrated on individual lifestyle factors linked to energy intake and energy expenditure and less on environmental factors such as food legislation and transport systems. The fact that the ‘obesogenic lifestyle’ seems more appealing to disadvantaged groups is a great challenge for public health, and must be widely addressed in future research.
What this study adds
This is a comprehensive report from a sample of Norwegian adolescents on the association between overweight and obesity and factors related to lifestyle and socio-economy. Whereas several socio-economic and cultural factors that might be associated to childhood overweight have been identified in international research, few corresponding studies have been performed in Norway. In order to prevent the growing epidemic of overweight and obesity, factors that may influence this development are important to identify.
|
Funding |
|---|
|
TOPIntroductionMethodsResultsDiscussionFundingReferences |
|---|
The data collection was carried out and funded by the Norwegian Institute of Public Health in collaboration with the University of Oslo and the Centre for Sami Health Research. The Municipality of Oslo contributed to the funding of the Oslo part of the study.
|
References |
|---|
|
TOPIntroductionMethodsResultsDiscussionFundingReferences |
|---|
- Lobstein T, Baur L, Uauy R. Obesity in children and young people: a crisis in public health. Obes Rev (2004) 5(Suppl. 1):4–104.[CrossRef][Medline]
- Ekblom O, Oddsson K, Ekblom B. Prevalence and regional differences in overweight in 2001 and trends in BMI distribution in Swedish children from 1987 to 2001. Scand J Public Health (2004) 32:257–63.
[Abstract/Free Full Text] - Andersen LF, Lillegaard IT, Overby N, et al. Overweight and obesity among Norwegian schoolchildren: changes from 1993 to 2000. Scand J Public Health (2005) 33:99–106.
[Abstract/Free Full Text] - Clement K. Genetics of human obesity. C R Biol (2006) 329:608–22.[CrossRef][Web of Science][Medline]
- Meyer HE, Tverdal A. Development of body weight in the Norwegian population. Prostaglandins Leukot Essent Fatty Acids (2005) 73:3–7.[CrossRef][Web of Science][Medline]
- Bjerkedal T, Brevik J. Geographic variation in body mass index among Norwegian conscripts born 1967–8. Tidsskr Nor Lægeforen (2001) 121:674–7.
- Bjørnelv S, Lydersen S, Mykletun A, et al. Changes in BMI-distribution from 1966–69 to 1995–97 in adolescents. The Young-HUNT study, Norway. BMC Public Health (2007) 7:279.[CrossRef][Medline]
- Júlíusson P, Roelants M, Eide GE, et al. Overweight and obesity in Norwegian children: Secular trends in weight for height and skinfolds. Acta Paediatr (2007) 96:1333–7.[CrossRef][Web of Science][Medline]
- Pietilainen KH, Kaprio J, Rasanen M, et al. Tracking of body size from birth to late adolescence: contributions of birth length, birth weight, duration of gestation, parents' body size, and twinship. Am J Epidemiol (2001) 154:21–9.
[Abstract/Free Full Text] - Cole TJ, Bellizzi MC, Flegal KM, et al. Establishing a standard definition of child overweight and obesity worldwide: international survey. Br Med J (2000) 320:1240–3.
[Abstract/Free Full Text] - Friestad C, Lien N, Klepp K-I. Educational plans—when are they established? Implications for the measurement of socioeconomic status in youth. Young (2001) 9:18–32.
[Free Full Text] - Gorber SC, Tremblay M, Moher D, et al. A comparison of direct vs. self-report measures for assessing height, weight and body mass index: a systematic review. Obes Rev (2007) 8:307–26.[CrossRef][Medline]
- Rasmussen F, Eriksson M, Nordquist T. Bias in height and weight reported by Swedish adolescents and relations to body dissatisfaction: the COMPASS study. Eur J Clin Nutr (2007) 61:870–6.[CrossRef][Web of Science][Medline]
- Ulseth E, Undheim R, Malterud K. Has the obesity epidemic reached Norway? Tidsskr Nor Laegeforen (2007) 127:34–7.[Medline]
- Washington R. One way to decrease an obesogenic environment. J Pediatr (2005) 147:417–8.[CrossRef][Web of Science][Medline]
- Keski-Rahkonen A, Kaprio J, Rissanen A, et al. Breakfast skipping and health-compromising behaviors in adolescents and adults. Eur J Clin Nutr (2003) 57:842–53.[CrossRef][Web of Science][Medline]
- Sjoberg A, Hallberg L, Hoglund D, et al. Meal pattern, food choice, nutrient intake and lifestyle factors in The Goteborg Adolescence Study. Eur J Clin Nutr (2003) 57:1569–78.[CrossRef][Web of Science][Medline]
- Barton BA, Eldridge AL, Thompson D, et al. The relationship of breakfast and cereal consumption to nutrient intake and body mass index: the National Heart, Lung, and Blood Institute Growth and Health Study. J Am Diet Assoc (2005) 105:1383–9.[CrossRef][Medline]
- Cho S, Dietrich M, Brown CJ, et al. The effect of breakfast type on total daily energy intake and body mass index: results from the Third National Health and Nutrition Examination Survey (NHANES III). J Am Coll Nutr (2003) 22:296–302.
[Abstract/Free Full Text] - Rampersaud GC, Pereira MA, Girard BL, et al. Breakfast habits, nutritional status, body weight, and academic performance in children and adolescents. J Am Diet Assoc (2005) 105:743–60.[CrossRef][Web of Science][Medline]
- Gortmaker SL, Must A, Sobol AM, et al. Television viewing as a cause of increasing obesity among children in the United States, 1986–1990. Arch Pediatr Adolesc Med (1996) 150:356–62.
[Abstract/Free Full Text] - Berkey CS, Rockett HR, Field AE, et al. Activity, dietary intake, and weight changes in a longitudinal study of preadolescent and adolescent boys and girls. Pediatrics (2000) 105:56.[CrossRef]
- Lien N, Kumar BN, Holmboe-Ottesen G, et al. Assessing social differences in overweight among 15- to 16-year old ethnic Norwegians from Oslo by register data and adolescent self-reported measures of socio-economic status. Int J Obes (2007) 31:30–8.[CrossRef][Web of Science][Medline]
- Vuille J-C, Schenkel M. Social equalization in the health of youth. The role of the school. Eur J Publ Health (2001) 11:287–93.
[Abstract/Free Full Text] - Spencer NJ. Social equalization in youth: evidence from a cross-sectional British survey. Eur J Public Health (2006) 16:368–75.
[Abstract/Free Full Text] - West P. Health inequalities in the early years: is there equalisation in youth? Soc Sci Med (1997) 44:833–58.[CrossRef][Web of Science][Medline]
- Vissers D, Devoogdt N, Gebruers N, et al. Overweight in adolescents: differences per type of education. Does one size fit all? J Nutr Educ Behav (2008) 40:65–71.[CrossRef][Web of Science][Medline]
- Meyer K. Prevention of obesity and physical inactivity—a socio-cultural challenge. J Publ Health (2008) 16:1–2.[CrossRef]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||