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



To determine to what extent self-reported and objective data on socio-economic status (SES) are associated with overweight/obesity among 15 to 16-year-old ethnic Norwegians.


A cross-sectional questionnaire study on health and health-related behaviors.


All school children aged 15–16 years old in 2000 and 2001 in Oslo, Norway. Response rate 88% (n=7343). This article is based on the data from the 5498 ethnic Norwegians.


Self-reported height and weight were used to measure overweight (including obesity) as defined by the International Obesity Task Force cutoffs at the nearest half-year intervals. SES was determined by register data from Statistics Norway on residential area, parental education and income and by adolescent self-reported measures on parental occupation and adolescents' educational plans.


The prevalence of overweight/obesity was low, but higher among boys (11%) than among girls (6%). Parental education (four levels) showed the clearest inverse gradients with overweight/obesity (boys: 18, 13, 10 and 7%; girls: 11, 6, 6 and 4%). Parental education remained significantly associated with overweight/obesity when adding occupation and income to the model for the boys, whereas there were no significant associations in the final model for the girls. Overweight/obesity was associated with a lower odds ratio of planning for higher education (college/university) among boys only.


For the boys, parental education was most strongly associated with overweight/obesity, and the association between overweight/obesity and educational plans appears to imply downward social mobility. The relationships between the various SES measures and overweight/obesity appeared more interrelated for the girls.


According to WHO, obesity has become a worldwide epidemic with latent effects on health in adulthood that will become fully evident as today's children grow up.1 A recent systematic review of health consequences of obesity concludes that obese children suffer from both psychological consequences and increased levels of cardiovascular risk factors in the short term, and from an increased risk of morbidity and mortality in adulthood.2 The need to prevent obesity in childhood is recognized, but the success of interventions has been limited so far.3, 4 Social inequalities in health among adults are well established, but are more debated for the adolescent age period.5 It has been shown that body mass index (BMI) track from childhood to adulthood.6 Thus, overweight could be a latent cause of health inequalities among adults, as the consequences of overweight may not manifest into health symptoms until adulthood.

In their review on socio-economic status (SES) and obesity, Sobal and Stunkard7 conclude that whereas there was a clear inverse relationship between SES and prevalence of obesity in adult women, the relationship was less clear for men and children/adolescents. A more recent review of childhood predictors of adult obesity found 12 longitudinal studies spanning from childhood to adulthood, and the majority showed an inverse relationship between childhood SES (based on parents' SES) and obesity in adulthood,8 but still the cross-sectional associations were less clear. The measures of SES used were mostly parental occupation, some studies used parental education and a few used parental income,8 but neither Sobal and Stunkard7 nor Parsons et al.8 elaborate on whether there were differences in the relationship based on different SES measures or whether the measures were obtained from the parents or by proxy report from the adolescents. The most commonly used measures of social status are occupation, education and income. Occupation is especially related to working conditions (both physical and psychological), education is related to lifestyle, problem-solving abilities and values, and income is related to better housing, medical care and diet.9 Although these measures are usually correlated, independent effects on dietary habits have been shown.10, 11

Parson et al.8 present two models that may explain the association between low SES and greater fatness in studies of adults: low SES may promote development of fatness or greater fatness may lead to downward social mobility. Sobal and Stunkard7 argue that weight control and pressure for thinness is more common among women of high SES than women of low SES, but state that little is known about why this is so. However, they suggest that during early socialization, SES could transmit pressure for thinness,7 which we postulate would be reflected in differences by SES in ‘desired’ BMI for girls. They also point to physical activity, social mobility and inheritance of SES as well as obesity as mediating factors explaining the negative association between SES and fatness.

Measuring the SES of adolescents is much discussed owing to the difficulty of obtaining valid data on parental SES,12, 13, 14, 15 and because adolescents have not obtained their own SES. Friestad et al.16 studied the stability of educational plans in a longitudinal study, and found that planning for academic studies showed high stability between ages 13 and 15, and did track into studying rather than working at age 18. Others have found that adolescents' health behaviors (including physical activity and dietary measures) predict adult educational attainment independent of the SES of the parents.17 Together, these findings make educational plans a potential measure of SES for studying health-related inequalities among adolescents. In addition, there is a growing interest in health differences by geographical areas,18 and it has previously been shown that morbidity differs by areas in Oslo.19

Norwegian data support that increasing BMI in adolescence increases risk of death in adulthood,20 and that obesity track from adolescence to adulthood.21 In a study of 18–19 year olds from Hordaland county in Norway, overweight did not differ by SES (parents' income and education), whereas it did for both indicators in an American sample described in the same article.22 However, a recent study of a representative sample of Norwegian 9 and 13 year olds did find increased odd ratio of being overweight among those whose parents had the shortest education.23

The purpose of this paper was to investigate whether any of five measures of SES were more strongly associated with overweight (including obesity) among 15- to 16-year-old ethnic Norwegians from Oslo, and whether the relationships could be described as gradients or thresholds. Furthermore, social differences in desired BMI were assessed, and the association between educational plans and overweight/obesity were studied to address some of the suggested mechanisms behind social differences in overweight/obesity.

Subjects and methods

The Oslo Health Study (HUBRO), a population-based cross-sectional study encompassed both adolescents and adults living in Oslo, Norway. The Norwegian Institute of Public Health (NIPH) in collaboration with the Municipality of Oslo and the University of Oslo conducted the study in the springs of 2000 and 2001. The study was submitted to the Regional Committee for Ethics in Medical Research and approved by the Data Inspectorate of Norway. Additional concession has been granted for the linkage of data between the youth part of the HUBRO study and Statistics Norway concerning parents' education and income. This study was conducted in full accordance with the ethical principles as per the World Medical Association Declaration of Helsinki.


The study population, n=8435, comprised all tenth graders (15 to 16-year-olds) in all the lower secondary schools in Oslo in the springs of 2000 and 2001. Of these, 31 special needs students were unable to fill in the questionnaires and 88 students moved out of Oslo before the survey started, leaving 8316 eligible for participation. Of those eligible, 39 did not consent and 88 withdrew before participation. Of those remaining 7343 (response rate 88.3%), both gave their written consent and completed at least one question, but owing to a mistake during data entry data on gender was missing for 36 of these. In this article, only the ethnic Norwegian adolescents with reported gender (n=5498) attending public or private schools (n=5372) were included. BMI of ethnic minorities in this school-based population has previously been thoroughly presented and discussed.24

Data collection and data entry methods

The Oslo School Authorities approved the study and informed all the schools that had to consent to participate. Parents/guardians received via post an information brochure describing the youth study, and could withdraw their child through a written refusal. The student's written consent was obtained before participation. Two groups of four field workers collected the data by means of two questionnaires filled in by the participants during a double school period in 62 schools in the spring of 2000 and in 60 schools in the spring of 2001. The questionnaires were based on similar studies previously conducted in Norway, existing scientific knowledge and current needs and priorities of both the users and producers of data. No physical or clinical examination was included. The data were entered manually by the staff from NIPH. A random sample, n=200, was entered twice and showed high correspondence (99.9%). All personal identifiers were removed from the research file.

Weight for height-related measures

Those who would turn 16 years old during the calendar year of the data collection were included, whereas the others (n=128) were excluded. BMI (kg/m2) was calculated based on self-reported weight and height. Those with missing data on BMI (n=306) were excluded. In addition, BMI greater than 35 (n=29) or BMI less than 15 (n=22) were excluded based on the BMI reference percentiles for children/adolescents.25 The distribution was calculated based on these reference percentiles. Age at the time of the data collection and the International Obesity Task Force gender-specific cutoffs at ages 15, 15.5, 16 and 16.5 were used to determine overweight (including obesity).26 A ‘desired BMI’ was calculated based on their reported height and what they considered their ideal weight.

Socio-demographic variables

Age, gender and residential address of the participants, and education and income of their parents (n=4420) were obtained from Statistics Norway. The categorization of the city into west (high SES) and east is based on previous studies of social inequalities.19 Each of the 25 administrative areas within the city was ranked on an area deprivation index consisting of prevalence of low education, single parenthood and unemployment. Significant differences in mortality and disability by area deprivation were found.19 Parents' education was grouped into four levels: lower secondary education/uncompleted upper secondary education (primary), secondary education/uncompleted college education (secondary), less than 4 years of college/university (bachelor) and 4 years or more of college/university (master). The parent with the longest education was used or else the one available. The mother's and father's income after tax were added and grouped into quintiles.

Occupational status was based on the adolescents' response on open-ended questions. The final variable consisted of five occupational categories, based on the model used by Skogen:27 upper managerial, technical/economic intermediate strata, humanistic/social intermediate strata, non-manual workers and manual workers. The two middle classes (intermediate strata) are not hierarchically related to each other, but differ by culture. The parent with the highest occupational status was used or else the one available.

The seven answer categories on adolescents' educational plans were collapsed into five: not decided/other/1 year at upper secondary, vocational secondary education, academic secondary education, less than 4 years of college/university (bachelor) and 4 years or more of college/university (master).

Data analysis

Data were analyzed using the SPSS package 11.0. Cross-tabulations with χ2 test, differences in means with analysis of variance and stepwise, multivariate logistic regressions were run to determine the associations of the SES measures with the BMI and overweight/obesity measures. Four of the six correlations between the SES measures were r0.2, area and education was correlated at r=0.3 and education and income at r=0.5. Including them in the same model was thus not considered a problem. The models were controlled for family structure (living with both parents versus not) for both gender, as this could influence the potential effect of the parental SES factors, and for age at menarche for the girls (11 years, 12–13 years, 13 years), as this is known to be associated with BMI. No measure of pubertal onset for the boys was available.


Table 1 shows the distribution between the different categories of the SES measures among boys and girls. According to the American Center for Disease Control and Prevention (CDC)-reference distribution of BMI,25 50% of the sample was expected to be in the group between the 25th and the 74th percentile, but about 60% of our sample was between these percentiles (Figure 1). The proportion above the 85 percentile was lower than 15%, and here there were clear gender differences (boys: 9%; girls: 5%).

Table 1 Description of socio-economic measures in a sample of 15- to 16-year-old ethnic Norwegians from Oslo, participating in a health survey in 2000/2001
Figure 1

Distribution of BMI in a sample of 15- to 16-year-old ethnic Norwegians from Oslo, 2000–2001, according to the CDC-reference distribution.25

The differences in mean BMI by socio-economic indicators were small, but consistent, indicating that those with the highest SES had a lower BMI than those with the lowest SES (Table 2 and 3). Parental education showed the clearest gradient, whereas the occupational measures differentiated between workers and middle/high classes and the income measures differentiated between the two lowest and the three highest quintiles. For educational plans, planning a vocational training was associated with a higher mean BMI. When overweight/obesity was the dependent variable, residential area, parental education and educational plans remained highly significant (P<0.001), but income was no longer significant for the boys (Table 2a and b). In the multivariate analyses, the association with residential area was cancelled out by parental education for both genders (Table 3a and b). For boys, parental education remained the only SES measure significantly associated with overweight/obesity when adding occupation and income to the model (Table 3a), whereas none of the SES measures were significantly associated with overweight/obesity in the full model for the girls (Table 3b).

Table 2 Mean BMI, prevalence of overweight/obesitya and mean desired BMIb by socio-economic status in a sample of 15- to 16-year-old ethnic Norwegian boysc from Oslo, 2000–2001
Table 3 Mean BMI, prevalence of overweight/obesitya and mean desired BMIb by socio-economic status in a sample of 15- to 16-year-old ethnic Norwegian girlsc from Oslo, 2000–2001
Table 4 Association of overweight/obesitya with socio-economic status among 15- to 16-year-old ethnic Norwegian boys in Oslo, 2000–2001. Stepwise, multivariate logistic regression modelsb, (N=2315)
Table 5 The association of overweight/obesitya with socio-economic status among 15 to 16-year-old ethnic Norwegian girls in Oslo, 2000–2001. Stepwise, multivariate logistic regression modelsb, (N=2274)

The desired BMI of boys was the lowest among those with the highest educated parents and those living on the west side of the city was the lowest (Table 2a). For girls, those with the lowest desired BMI were from the west side of the city, had parents of middle/high class (based on occupation), parental income in the two upper quintiles or was planning for a master degree (Table 2b).

Table 4 shows that a lower odds ratio of planning for a higher education was found among overweight/obese boys even when residential area and parental education were controlled for.

Table 6 The association of planning higher educationa with overweight/obesityb among 15 to 16-year-old ethnic Norwegians in Oslo, 2000–2001. Stepwise, multivariate logistic regression modelsc

The 611 students excluded owing to attending special schools, being outside the age range, lacking data on BMI or unrealistic BMI values, did not differ from those included in the analyses with respect to the distribution of gender, parental occupation, family structure, perception of own health or preoccupation with weight (data not shown). In the group of girls with missing data on desired BMI, there were fewer dieters (8%) than in the group with data (23%) and the mean BMI was lower than among those with data on desired BMI, 19.6 and 20.3, respectively (data not shown).


In this study, five measures of SES were applied to investigate social inequalities in BMI and overweight/obesity among ethnic Norwegian adolescents. All the SES measures showed a higher BMI among those with the lowest SES compared to the highest SES, but only parental education showed a clear gradient. Educational plans distinguished between those planning a vocational training (higher BMI) and those planning for a master's degree (lower BMI). Among boys, parental education was the only SES measure associated with overweight/obesity in the multivariate model, and overweight/obesity was negatively associated with planning for a higher education implying downward social mobility. None of these associations were found for the girls.

It should be noted that the prevalence of overweight/obesity was low in our study, both applying the International Obesity Task Force's definition and the CDC-reference population and compared with data from other European countries.28 Furthermore, the rate was lower among the girls than among the boys, which may be part of the explanation of the gender differences found in the multivariate analyses owing to the lack of power in the analyses. However, both comparison across Europe28 and findings from a national sample of 9 and 13 year olds in Norway23 indicate that the prevalence is higher among younger children than among older ones, implying that overweight may become an increasing problem also in Norway.

Consistent with findings from previous reviews of SES and obesity in developed countries,7, 8 an inverse relationship was found: the children of workers, those with the lowest income and those with the less educated parents had the highest prevalence of overweight/obesity. However, only parental education showed a clear gradient, which may indicate that social inequality in overweight is caused by differences in knowledge levels and health-related behaviors, and less by material circumstances. Although the measure of parental occupation is not designed to be hierarchical, the lack of variation between the different groups was surprising. It could be owing to validity issues of the reporting and coding of this self-reported measure, but could also be caused by the small differences by social status in a socio-democratic country such as Norway where the middle class is dominating. The attempt to sift out differences within the large middle class group was made by dividing the middle class into two according to the type of profession. Those in social/humanistic jobs (i.e. teachers, health workers) would be expected to be more concerned about health and weight than the group with technical/economic jobs, but this was not found. However, the results did support the findings of others29, 30 that overweight is more common among children of the working/lower managerial class than the middle/upper class. This was also reflected in the relation with parental income where the two lowest quintiles had the highest prevalence of overweight, although this was not significant for boys.

There were clear gender differences in the multivariate models investigating whether any of the SES measures were independently associated with overweight/obesity. For boys, parental education was the most important measure, but among girls this association was no longer significant when occupation and income was added to the model. This parallels the gender differences found for the associations between desired BMI and the SES measures. In their review, Sobal and Stunkard7 discuss attitude toward obesity in developed societies and point out that pressure for thinness shows a strong relationship with SES among women. We used desired BMI as an indicator of internalization of social norms. Although the results did not strongly support the idea that desired BMI was associated with SES already in adolescence, some gender differences were observed. This could imply that general social norms on weight for girls cut across the effect of all the different indicators of SES, whereas for boys awareness of social norms for weight is mostly linked to parental education, perhaps through parental knowledge and role modelling within the family.

Furthermore, we found that for boys, being overweight/obese was associated with lower educational aims. This is contrary to Sobal and Stunkard,7 who found that the influence of obesity on social mobility is especially reported for women, but fits with the findings discussed above. Sobal and Stunkard7 also proposed that inheritance of obesity as well as SES influences the relationship between obesity and SES.7 Unfortunately, we did not have the possibility to control for the effect of parental BMI while assessing the effect of SES. Power and Parsons31 conclude in their review that ‘few studies have allowed for parental BMI when examining the SES of origin–adult obesity relationship’. Later, two studies have been published, which support that the relationship remains after adjusting for parental BMI,32, 33 whereas a longitudinal study from Oslo21 found that neither the father's education nor own education was significant predictors of BMI at age 33, when adolescent BMI and the father's BMI were included. More data on this is clearly needed.

A major strength of this study was the use of five different measures of SES and especially the register data on parental education and income as the adolescents' ability to provide valid data on these measures has been questioned.12 Methodologically, it is also a strength that all the schools in the city of Oslo participated. Furthermore, the high response rate in combination with the lack of bias owing to attrition, whether owing to missing data on weight/height or SES, imply that the results may be seen as representative of urban, ethnic Norwegian adolescents. The discussion of whether BMI can be used as a measure of nutritional status in adolescence has led to two different approaches: the CDC-reference curves,25 and the International Obesity Task Force cutoffs for overweight and obesity.26 We presented data on overweight/obesity with regard to both methods, and found similar prevalences of overweight/obesity.

There are, however, some potential weaknesses that should be acknowledged. BMI was based on self-reported weight and height, which raises questions about the reliability and validity of this measure. A methodological study among 17–19 year olds in Oslo (n=156) concluded that both test–retest reliability (Pearson's r=0.98 and 0.99) and validity (Pearson's r=0.94 and 0.96) of self-reported compared with measured height and weight was very good.34 Also, in American studies self-reported height and weight has been shown to be reliable and valid, and it is being used to assess overweight in the Youth Risk Behavior Surveillance System as it is easy to obtain, although it is acknowledged that adolescents who are overweight tend to underreport their weight.35 The occupational measure was based on the adolescents' reports, but although the proportion of missing could be substantial on this type of question, the adolescents are usually able to provide valid answers.14, 15 There were, however, some difficulties in assigning the occupations into different social groups, especially the upper managerial group. Social desirability in reporting could bias the association between BMI and SES as both were self-reported. However, as similar results were found for the register-based data on SES and BMI, this is not seen as a large problem. Although we did control for family structure, it is difficult to know how the SES of parents influence the children when they are not living together. Furthermore, the measure of income did not include other measures of wealth. We acknowledge that the exclusion of the ethnic minority adolescents limits the generalizability of the findings to ethnic Norwegian adolescents. However, we24 have previously shown that for the ethnic minority adolescents in our study, there were clear differences in BMI between six broad ethnic groups, but no association between BMI and social class based on occupation or mother's education. This could be owing to heterogeneity in SES within these six ethnic groups, lower validity of the data on education, occupation or income for immigrants even when based on register data or that the majority of the non-western immigrants were classified as lower SES.24

In conclusion, SES differences in overweight/obesity were present in the cohort, but an independent significant effect of the different SES measures was found only for parents' educational level and only among the boys. Indication of downward social mobility was found for boys, and this needs further investigation. The causes of differences in overweight/obesity by education should be investigated in qualitative studies with special emphasis on gender differences in awareness of social norms for body weight. For those working in obesity prevention, the clear difference by parental educational level may indicate that the methods should be differentiated by educational level.


  1. 1

    WHO. Obesity. Preventing and Managing the Global Epidemic Report of a WHO Consultation on Obesity, 3–5 June 1997, WHO: Geneva 1998.

  2. 2

    Reilly JJ, Methven E, McDowell ZC, Hacking B, Alexander D, Stewart L et al. Health consequences of obesity. Arch Dis Child 2003; 88: 748–752.

    CAS  Article  Google Scholar 

  3. 3

    Campbell K, Waters E, O'Meara S, Summerbell C . Interventions for preventing obesity in childhood. A systematic review. Obes Rev 2001; 2: 149–157.

    CAS  Article  Google Scholar 

  4. 4

    Summerbell CD, Ashton V, Campbell KJ, Edmunds L, Kelly S, Waters E . Interventions for treating obesity in children. Cochrane Database of Systematic Reviews 2003. CD001872.

  5. 5

    West P, Sweeting H . Evidence on equalization in health in youth form the West of Scotland. Soc Sci Med 2004; 59: 13–27.

    Article  Google Scholar 

  6. 6

    Whitaker RC, Wright JA, Pepe MS, Seidel KD, Dietz WH . Predicting obesity in young adulthood from childhood and parental obesity. N Engl J Med 1997; 337: 869–873.

    CAS  Article  Google Scholar 

  7. 7

    Sobal J, Stunkard AJ . Socioeconomic status and obesity: a review of the literature. Psychol Bull 1989; 105: 260–275.

    CAS  Article  Google Scholar 

  8. 8

    Parsons TJ, Power C, Logan S, Summerbell CD . Childhood predictors of adult obesity: a systematic review. Int J Obes Relat Metab Disord 1999; 23: S1–107.

    PubMed  PubMed Central  Google Scholar 

  9. 9

    Liberatos P, Link BG, Kelsey JL . The measurement of social class in epidemiology. Epidemiol Rev 1988; 10: 87–121.

    CAS  Article  Google Scholar 

  10. 10

    Galobardes B, Morabia A, Bernstein MS . Diet and socioeconomic position: does the use of different indicators matter? Int J Epidemiol 2001; 30: 334–340.

    CAS  Article  Google Scholar 

  11. 11

    Turrell G, Hewitt B, Patterson C, Oldenburg B . Measuring socio-economic position in dietary research: is choice of socio-economic indicator important? Pub Health Nutr 2003; 6: 191–200.

    Google Scholar 

  12. 12

    Looker ED . Accuracy of proxy reports of parental characteristics. Sociol Educ 1989; 62: 257–276.

    Article  Google Scholar 

  13. 13

    Currie CE, Elton RA, Todd J, Platt S . Indicators of socioeconomic status for adolescents: the WHO health behaviour in school-aged children survey. Health Educ Res 1997; 12: 385–397.

    CAS  Article  Google Scholar 

  14. 14

    Lien N, Friestad C, Klepp K-I . Adolescents' proxy reports of parents' socioeconomic status – how valid are they? J Epidemiol Community Health 2001; 55: 731–737.

    CAS  Article  Google Scholar 

  15. 15

    West P, Sweeting H, Speed E . We really do know what you do: a comparison of reports from 11 year olds and their parents in respect of parental economic activity and occupation. Sociology 2001; 35: 339–359.

    Article  Google Scholar 

  16. 16

    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.

    Article  Google Scholar 

  17. 17

    Koivusilta L, Rimpela A, Vikat A . Health behaviours and health in adolescence as predictors of educational level in adulthood: a follow-up study from Finland. Soc Sci Med 2003; 57: 577–593.

    Article  Google Scholar 

  18. 18

    Pickett KE, Pearl M . Multilevel analyses of neighbourhood socioeconomic context and health outcomes: a critical review. J Epidemiol Community Health 2001; 55: 111–122.

    CAS  Article  Google Scholar 

  19. 19

    Rognerud MA, Kruger O, Gjertsen F, Thelle DS . Strong regional links between socio-economic background factors and disability and mortality in Oslo, Norway. Eur J Epidemiol 1998; 14: 457–463.

    CAS  Article  Google Scholar 

  20. 20

    Engeland A, Bjorge T, Sogaard AJ, Tverdal A . Body mass index in adolescence in relation to total mortality: 32-year follow-up of 227,000 Norwegian boys and girls. Am J Epidemiol 2003; 157: 517–523.

    Article  Google Scholar 

  21. 21

    Kvaavik E, Tell GS, Klepp K-I . Predictors and tracking of body mass index from adolescence into adulthood: follow-up of 18 to 20 years in the Oslo Youth Study. Arch Pediatr Adolesc Med 2003; 157: 1212–1218.

    Article  Google Scholar 

  22. 22

    Friestad C, Pirkis J, Biehl M, Irwin Jr CE . Socioeconomic patterning of smoking, sedentary lifestyle, and overweight status among adolescents in Norway and the United States. J Adolesc Health 2003; 33: 275–278.

    Article  Google Scholar 

  23. 23

    Frost-Andersen L, Lillegaard ITL, Øverby N, Lytle L, Klepp K-I, Johansson L . Overweight and obesity among Norwegian school children: changes from 1993 to 2000. Scand J Pub Health 2005; 33: 99–106.

    Article  Google Scholar 

  24. 24

    Kumar BN, Holmboe-Ottesen G, Lien N, Wandel M . Ethnic differences in body mass index and associated factors of adolescents from Minorities in Oslo, Norway: a cross sectional study. Public Health Nutr 2004; 7: 999–1008.

    CAS  Article  Google Scholar 

  25. 25

    Kuczmarski RJ, Ogden CL, Grummer-Strawn LM, Flegal KM, Guo SS, Wei R et al. CDC growth charts: United States. Adv Data 2000; 314: 1–27.

    Google Scholar 

  26. 26

    Cole TJ, Bellizzi MC, Flegal KM, Dietz WH . Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ 2000; 320: 1240–1243.

    CAS  Article  Google Scholar 

  27. 27

    Skogen K . Cultures and Natures. Cultural Patterns, Environmental Orientations and Outdoor Recreation Practices among Norwegian Youth. Norwegian Social Research NOVA Report 16/99: Oslo, 1999.

    Google Scholar 

  28. 28

    Lobstein T, Frelut ML . Prevalence of overweight among children in Europe. Obes Rev 2003; 4: 195–200.

    CAS  Article  Google Scholar 

  29. 29

    Mikkila V, Lahti-Koski M, Pietinen P, Virtanen SM, Rimpela M . Associates of obesity and weight dissatisfaction among Finnish adolescents. Public Health Nutr 2003; 6: 49–56.

    Article  Google Scholar 

  30. 30

    O'Dea JA, Caputi P . Association between socioeconomic status, weight, age and gender, and the body image and weight control practices of 6- to 19-year-old children and adolescents. Health Educ Res 2001; 16: 521–532.

    CAS  Article  Google Scholar 

  31. 31

    Power C, Parsons T . Nutritional and other influences in childhood as predictors of adult obesity. Proc Nutr Soc 2000; 59: 267–272.

    CAS  Article  Google Scholar 

  32. 32

    Laitinen J, Power C, Jarvelin MR . Family social class, maternal body mass index, childhood body mass index, and age at menarche as predictors of adult obesity. Am J Clin Nutr 2001; 74: 287–294.

    CAS  Article  Google Scholar 

  33. 33

    Power C, Manor O, Matthews S . Child to adult socioeconomic conditions and obesity in a national cohort. Int J Obesity 2003; 27: 1081–1086.

    CAS  Article  Google Scholar 

  34. 34

    Skaarer C . En metodevurdering av spørsmål stilt ved 16-, 18-, og 19-års alder i den longitudinelle spørreundersøkelsen; ‘Voksen i år 2000’. (A methodological evaluation of questions asked at the age of 16, 18 and 19 in the longitudinal survey; ‘The Norwegian Longitudinal Health Behaviour Study’) Cand Scient Thesis, Institute for Nutrition Research, Oslo, 2001.

    Google Scholar 

  35. 35

    Goodman E, Strauss RS . Self-reported height and weight and the definition of obesity in epidemiological studies. J Adolesc Health 2003; 33: 140–142.

    Article  Google Scholar 

Download references


The Oslo Health Study was funded by the NIPH, University of Oslo and the Municipality of Oslo. The National Health Screening Services (SHUS) (now part of the NIPH) conducted the data collection. The post-doctoral fellowship for Dr Lien is funded by a grant from the Norwegian Research Council.

Author information



Corresponding author

Correspondence to N Lien.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Lien, N., Kumar, B., 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 31, 30–38 (2007).

Download citation


  • BMI
  • overweight
  • socio-economic status
  • adolescents

Further reading