SEARCH

SEARCH BY CITATION

Keywords:

  • education;
  • discrimination;
  • stigmatization;
  • school marks;
  • university

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Objective: To study the relationship between BMI at age 18 years and later attained education, with control for intelligence and parental social position.

Research Methods and Procedures: A cohort of 752,283 Swedish men born from 1952 to 1973 were followed in registers with respect to attainment of high education (≥15 years of education) until December 31, 2001. Intelligence and BMI (kilograms per meter squared) were measured at compulsory military conscription at age 18 years. Ninth grade school marks were available for a subgroup born from 1972 to 1973 (N = 93,374). The hazard ratio for attaining high education was estimated with proportional hazard regression analysis controlling for intelligence, height, parental socioeconomic position, country of birth, conscription center, and municipality.

Results: Young men who were obese (BMI ≥ 30) at age 18 years (N = 10,782) had a much lower chance of attaining a high education than normal-weight subjects [(18.5 ≤ BMI < 25); adjusted hazard ratio 0.48 (95% confidence interval, 0.45, 0.52)]. Young men who were obese at age 18 had lower mean ninth grade school marks than young men with normal weight at any given intelligence level.

Discussion: Obese men in Sweden are doing much worse in the educational system than their normal-weight counterparts even after adjustments for intelligence and parental socioeconomic position. Discrimination in the educational system and other sectors of the society may explain these strong associations.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

The prevalence of obesity has increased dramatically all over the world. In the footsteps of the obesity epidemic, comorbidities such as diabetes and cardiovascular diseases prevail, but obesity does not result only in medical consequences. Strong inverse associations between social position and obesity have been reported from many affluent societies (1, 2, 3, 4, 5). Sobal (1) suggested that obese people are exposed to stigmatization and discrimination in societies where thinness is the norm. A study from the 1980s found that obese men did not attain as high a social position as their normal-weight counterparts when intelligence, education, and parental social position were taken into account (2). However, a recent study found no association between obesity and social disadvantage among men (4); thus, knowledge is inconsistent in this field of research, and the causality question is unsolved.

Education is often used as a measure of social position. Attained education influences which occupation might be attainable and, consequently, income. If overweight and obese subjects attain lower education than their normal-weight counterparts, it may have profound consequences for their future possibilities in life. We have been able to conduct a longitudinal population-based cohort study, much larger than any previous study, of the relationship between BMI of military conscripts and their later attained education, controlling for intelligence and parental social position.

Research Methods and Procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Population and Record Linkage

In 2003, all Swedish men born from 1952 to 1973 (N = 1547,008) were identified in Statistic Sweden's Register of the Total Population. Men who had died before conscription (N = 19,823) or emigrated before conscription (N = 52,179) and immigrated men who were not Swedish citizens at the time of conscription (N = 244,893) were excluded. There were also men who were not called to conscription for unknown reasons (N = 131,980). The study population consisted of 1,119,219 men eligible for conscription examination.

Information on the study subjects was obtained by a linkage of the Swedish Military Service Conscription Register, the Population and Housing Censuses of 1960, 1970, 1990, and the Longitudinal Database of Education, Income, and Occupation of 2001. Statistics Sweden's register on school marks and school leaving certificates was also used. Mortality data were obtained from the Cause of Death Register, and migration data were obtained from the Register of the Total Population. Parents of study cohort members were identified in the Multi-Generation Register using the unique personal identification number ascribed to all individuals with permanent residence in Sweden. We followed the index cohort and their parents through the registers mentioned above from birth until the end of 2001.

Among these 1,119,219 men eligible for conscription examinations, 56,566 (5.1%) did not go through conscription for unknown reasons, 136,755 (12.2%) took part in the conscription examinations but had missing information on some of the variables from conscription, 14,448 (1.3%) lacked information on education in the registers used, 73,682 (6.6%) lacked information on time for attaining a high educational level, and 102,740 (9.2%) lacked parental information. Of the 1,119,219 men, 752,283 (67.2%) had information on all variables used, and 925,898 men (82.7%) had information on all variables except time for educational attainment. We identified 93,374 men born from 1972 to 1973 with additional register information on school marks from ninth grade and type of educational program in Grades 10 to 12. The youngest cohort, born in 1973, was followed from birth to age 28 years and the oldest cohort, born 1952, from birth to age 49 years.

Ethical Considerations

The study was approved by the Ethics committee at the Karolinska Institute, Stockholm, Sweden (reference no. 03-354).

Outcome Variables

Sweden has 9 years of compulsory schooling. Thereafter, adolescents may continue with a practically oriented program or a theoretically oriented program (Years 10 to 12). Marks are set from 1 to 5, where 1 is the lowest. Universities are free of charge, and admissibility criteria are based mainly on marks. High education, defined as at least 15 years of education (corresponding to the minimal requirements for a university degree in Sweden), was the main outcome variable. An additional definition of high education defined as at least 13 years of education (corresponding to at least 1 year of university education) was used for reasons of comparison.

Explanatory Variables and Confounders

Height and weight were measured and intellectual performance tested at the conscription examinations. BMI was calculated as weight (kilograms) divided by squared height (meters squared) and categorized into deciles or categorized as underweight (BMI < 18.5), normal weight (18.5 ≤ BMI < 25), overweight (25 ≤ BMI < 30), and obese (BMI ≥ 30) according to the World Health Organization criteria (6). The intelligence test used is included in the Swedish Enlistment Battery and matches the concept of general ability. It has been described in detail elsewhere (7, 8, 9, 10). There are four basic tests: a logic/general intelligence test, a verbal test of synonym detection, a test of visual-spatial/geometric perception, and a technical/mechanical skills test with mathematical/physics problems. A combined (global) intelligence score is derived from the conscripts’ performance on all four tests, and a Gaussian distributed score between 1 and 9 is generated. There were six conscription areas in Sweden, and the analyses were additionally adjusted for these areas as a means of eliminating potential confounding.

Data from Population and Housing Censuses and the Longitudinal Database of Education, Income, and Occupation included parental education and socioeconomic index (SEI)1 group, as well as country of birth and educational attainment of study subjects. Information on parental education and SEI group was used as a proxy of childhood socioeconomic position. We used the highest SEI group and the highest education of either parent, nearest to the time of birth of the study subjects, either 1960 or 1970. SEI groups were non-manual high, non-manual intermediate, non-manual low, skilled workers, unskilled workers, farmers, and others. Parental education was dichotomized into high (at least 15 years) and low (<15 years) education. Ethnicity was categorized as being born in Sweden or in a foreign country. The area of residence at the time of conscription was categorized as main cities and suburbs, large cities and industrial areas, and rural areas by using an abbreviated version of a classification of municipalities described elsewhere (11).

Statistical Methods

Associations between BMI categories and attained education were estimated by Cox proportional hazard regression analyses with PROC PHREG in SAS (12). We used age as the time axis. This means that all analyses were controlled for age. Individuals were censored at the time of emigration, death, or December 31, 2001, whichever came first. Because the data set includes families with multiple siblings who are correlated and, thus, violates the usual independence assumption, we adjusted standard errors with a robust sandwich estimator (covsandwich option in phreg). The proportional hazard assumption was checked graphically, and we found no evidence that this assumption was violated. The hazard ratios (HRs) for high education among men who were underweight, overweight, or obese were estimated using men of normal weight as reference and with adjustment for performance on the intelligence test, parental education and SEI group, country of birth, and conscription center. To assess secular trends over calendar time, we analyzed the data with models stratified by birth year. To adjust for time trends in height and BMI, we created birth year-standardized z scores of height and BMI that were categorized into deciles. Cox regression models estimated the HR for high education with the 3rd decile as reference group and with adjustment for performance on the intelligence test, parental education and SEI group, country of birth, and conscription center. Supplementary analyses were conducted by logistic regression in which odds ratios (ORs) were adjusted for the prevalence of high education (13). In these analyses, 173,615 additional men with information on educational level but missing information on time for high educational attainment could be included. The ORs for high educational attainment in 2001 were estimated for underweight, overweight, and obese men with normal-weight men as reference. To further understand the determinants of attained education, we analyzed the impact of ninth grade school marks with multivariate Cox regression models. To analyze the statistical interaction between BMI and intellectual performance in relation to attained education, a fully adjusted Cox regression model stratified by intelligence test performance was conducted, and the interaction was tested with the likelihood ratio test.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

The main analyses included 752,283 men. The mean age of attaining a high education was 27.3 years (range, 20.0 to 48.8 years). When normal-weight individuals were set as reference in the univariate Cox regression model, underweight [HR, 0.93; 95% confidence interval (CI), 0.91, 0.95], overweight (HR, 0.54; 95% CI, 0.53, 0.56), and obese (HR, 0.30; 95% CI, 0.28, 0.32) individuals had a decreased likelihood of attaining high education (Table 1). We stratified the analyses by birth year, but no significant change in the relation between BMI and attained education was observed over time (data not shown).

Table 1.  HRs for attaining high education among 752,283 Swedish men
VariableN = 752,283Prevalence of high education (%)HR (95% CI)*HR (95% CI)
  • HR, hazard ratio; CI, confidence index; SEI, socio-economic index; (ref), reference level for analysis.

  • *

    Crude HRs were estimated by Cox regression models controlled for age.

  • HRs were estimated with Cox regression models controlled for age and adjusted for all variables in the table and additionally adjusted for country of birth, municipality, and conscription center.

BMI category    
 Underweight63,90619.30.93 (0.91, 0.95)0.94 (0.92, 0.95)
 Normal weight(ref)617,45920.21.001.00
 Overweight60,32611.30.54 (0.53, 0.56)0.70 (0.68, 0.72)
 Obese10,5926.30.30 (0.28, 0.32)0.48 (0.44, 0.52)
Intelligence test score (IQ)    
 113,8011.10.09 (0.08, 0.10)0.11 (0.10, 0.13)
 238,3131.40.11 (0.10, 0.12)0.14 (0.13, 0.15)
 365,4843.10.25 (0.24, 0.27)0.30 (0.28, 0.31)
 4103,7555.60.47 (0.46, 0.48)0.52 (0.50, 0.53)
 5(ref)171,54111.51.001.00
 6140,02420.41.84 (1.81, 1.88)1.69 (1.66, 1.72)
 7110,36331.33.04 (2.98, 3.09)2.52 (2.48, 2.56)
 870,10743.94.70 (4.61, 4.78)3.50 (3.44, 3.57)
 938,89558.07.07 (6.94, 7.21)4.72 (4.63, 4.82)
Height decile    
 171,95712.90.46 (0.45, 0.47)0.75 (0.73, 0.77)
 272,53315.50.56 (0.55, 0.58)0.80 (0.78, 0.82)
 373,85716.90.62 (0.61, 0.64)0.83 (0.81, 0.85)
 474,27217.70.68 (0.66, 0.69)0.87 (0.85, 0.89)
 575,52418.40.71 (0.70, 0.73)0.89 (0.87, 0.90)
 674,77120.20.75 (0.73, 0.77)0.90 (0.88, 0.92)
 775,36020.40.80 (0.78, 0.81)0.93 (0.91, 0.95)
 877,94821.60.83 (0.81, 0.85)0.93 (0.91, 0.95)
 976,98322.70.89 (0.88, 0.91)0.96 (0.95, 0.98)
 10(ref)79,07825.01.001.00
Parental education    
 High(ref)117,88045.41.001.00
 Low634,40314.40.24 (0.24, 0.25)0.55 (0.54, 0.55)
Parental SEI    
 Non-manual high(ref)48,49949.11.001.00
 Non-manual intermediate160,49231.00.55 (0.54, 0.56)0.80 (0.79, 0.81)
 Non-manual low110,87219.90.33 (0.32, 0.34)0.69 (0.67, 0.70)
 Skilled workers171,51110.60.16 (0.16, 0.17)0.44 (0.43, 0.45)
 Unskilled workers190,0919.70.15 (0.14, 0.15)0.41 (0.41, 0.42)
 Farmers3213911.90.17 (0.17, 0.18)0.40 (0.39, 0.42)
 Others38,67922.40.39 (0.38, 0.40)0.69 (0.68, 0.71)

The multivariate models included adjustments for performance on the intelligence test, height, parental education and SEI group, country of birth, municipality, and conscription center (Table 1). Even though height is an important risk factor for low education (14), it did not confound the relationship between obesity and the probability of high education. In the fully adjusted Cox regression model, the association between BMI and probability of attaining high education became weaker but remained highly significant (HR for obese subjects, 0.48; 95% CI, 0.44, 0.52) (Table 1). The adjusted relationship showed a decreased likelihood of attaining high education for underweight as well as overweight and obese men. The associations between BMI and attained education were somewhat weaker when the outcome was redefined as at least 13 years of education, corresponding to at least 1 year of university studies (HR for obese subjects, 0.63; 95% CI, 0.60, 0.66). Accordingly, obese men had an HR of 0.63 for starting a university education and an HR of 0.48 for graduating compared with normal-weight men.

To control for changes in BMI over calendar time, birth year-adjusted z scores for BMI and height were created. We compared fully adjusted Cox regression analyses including deciles of birth year-adjusted z scores of BMI with fully adjusted Cox regression analyses including crude deciles of BMI. The estimated associations in the analyses did not differ, indicating that year of birth did not confound our results (Table 2).

Table 2.  HRs for attaining high education in relation to BMI deciles, classified with or without birth year-adjusted z scores, among 752,283 Swedish men
Crude BMI decilesHR (95% CI)*Birth year-adjusted BMI z scores in decilesHR (95% CI)*
  • HR, hazard ratio; CI, confidence index; SEI, socio-economic index; (ref), reference level for analysis.

  • *

    HRs were estimated with Cox regression models controlled for age and adjusted for intelligence test performance, height, parental education, parental SEI, country of birth, municipality, and conscription center.

10.91 (0.89, 0.93)10.91 (0.89, 0.94)
20.95 (0.93, 0.97)20.95 (0.93, 0.97)
3(ref)1.003(ref)1.00
40.97 (0.95, 0.99)40.98 (0.96, 1.00)
50.99 (0.97, 1.01)50.99 (0.97, 1.01)
61.01 (0.99, 1.04)61.01 (0.99, 1.03)
70.99 (0.97, 1.01)70.99 (0.96, 1.01)
80.96 (0.94, 0.99)80.96 (0.94, 0.98)
90.88 (0.85, 0.90)90.87 (0.85, 0.89)
100.65 (0.64, 0.67)100.64 (0.62, 0.66)

The logistic regression analyses included 925,898 men. The fully adjusted ORs for attaining high education were 0.93 (95% CI, 0.91, 0.95) for underweight men, 0.68 (95% CI, 0.67, 0.70) for overweight men, and 0.47 (95% CI, 0.43, 0.50) for obese men. When conducting fully adjusted logistic regression analyses on the 752,283 observations included in the main analyses, very similar results emerged: ORs were 0.96 (95% CI, 0.94, 0.98) for underweight men, 0.69 (95% CI, 0.67, 0.71) for overweight men, and 0.47 (95% CI, 0.44, 0.51) for obese men.

We identified 93,374 men born from 1972 to 1973 who had additional register information on school marks from ninth grade (age, 15 years). In Table 3, we have stratified the ninth grade mean marks by BMI and the intelligence performance scores from age 18 years. At any given intelligence level, men obese at age 18 had lower mean school marks at age 15 than men with normal weight at age 18. We repeated the Cox regression models on this smaller cohort with additional adjustments for school marks in ninth grade and type of educational program (theoretical or practical) in school years 10 to 12 (Table 4). The adjusted HRs (same adjustments as in the main model described above) were very similar compared with the main model, HR for obese men being 0.41 (95% CI, 0.34, 0.49). We entered the adjustments in the model according to their order of appearance during life course. In the first multivariate model including adjustment for parental factors and country of birth, the HR for attaining high education was 0.33 (95% CI, 0.28, 0.40) for obese men. We additionally adjusted for ninth grade school marks, increasing the obesity HR to 0.56 (95% CI, 0.47, 0.68). The HR for high educational attainment could not be further explained by inclusion of educational program years 10 to 12, intelligence test performance, height, municipality, or conscription center.

Table 3.  School marks in 9th grade (age 15) by intelligence test score and BMI category (age 18) among 93,374 men born in 1972 to 1973
Intelligence test scoreUnderweight (N = 6427) [mean (SD) n]Normal weight (N = 75,453) [mean (SD) n]Overweight (N = 9459) [mean (SD) n]Obese (N = 2035) [mean (SD) n]
  1. SD, standard deviation.

12.08 (0.46) 2352.11 (0.48) 17882.12 (0.42) 3701.96 (0.49) 123
22.34 (0.51) 5392.38 (0.48) 50562.30 (0.46) 9342.20 (0.43) 249
32.52 (0.51) 6132.63 (0.50) 74172.53 (0.49) 11822.35 (0.47) 312
42.75 (0.52) 8742.85 (0.51) 108362.73 (0.51) 15922.52 (0.49) 364
53.01 (0.52) 15403.12 (0.51) 186392.98 (0.50) 23532.84 (0.49) 484
63.29 (0.52) 10053.40 (0.50) 129273.26 (0.50) 13793.02 (0.51) 227
73.58 (0.49) 8343.62 (0.48) 97293.51 (0.49) 9463.31 (0.48) 167
83.75 (0.48) 4913.82 (0.46) 59393.66 (0.45) 4573.50 (0.45) 79
93.96 (0.48) 2964.05 (0.43) 31223.91 (0.46) 2463.79 (0.42) 30
Table 4.  HRs for attainment of high education in relation to BMI among a subset of men (N = 93,282) with information on school marks
BMINPrevalence of high education (%)Unadjusted HRHR (95% CI)*HR (95% CI)HR (95% CI)HR (95% CI)§HR (95% CI)
  • HR, hazard ratio; CI, confidence index; SEI, socio-economic index.

  • *

    HRs were estimated with Cox regression models controlled for age and were adjusted for parental education, parental SEI at birth, and birth country.

  • Additionally adjusted for 9th grade mean school marks.

  • Additionally adjusted for 9th grade mean school marks and for educational program years 10 to 12.

  • §

    Additionally adjusted for 9th grade mean school marks, for educational program years 10 to 12, and for intelligence test performance, municipality, and conscription center.

  • HRs were estimated with Cox regression models and were adjusted for parental education, parental SEI at birth, birth country, intelligence test performance, height, municipality, and conscription center (same adjustments as in the rightmost column of Table 1).

Underweight642320.610.93 (0.88, 0.99)0.95 (0.90, 1.01)1.07 (1.01, 1.13)1.07 (1.01, 1.13)1.03 (0.97, 1.09)0.95 (0.90, 1.00)
Normal(ref)75,37622.001.001.001.001.001.001.00
Overweight945211.780.50 (0.47, 0.53)0.60 (0.57, 0.64)0.75 (0.71, 0.80)0.77 (0.72, 0.82)0.76 (0.72, 0.81)0.67 (0.63, 0.71)
Obesity20315.810.24 (0.20, 0.29)0.33 (0.28, 0.40)0.56 (0.47, 0.68)0.58 (0.49, 0.70)0.57 (0.48, 0.68)0.41 (0.34, 0.49)

Figure 1 shows the associations between categories of BMI and prevalence of high education (Fig. 1, left) or adjusted HR for high education (Fig. 1, right) stratified by intelligence test score. The statistical interaction between intelligence and obesity was significant (p = 0.0055). The crude prevalence and adjusted HR of high education among those with low score on the intelligence test (values, 1 to 3) did not change on the absolute scale according to BMI. However, the crude prevalence and adjusted HR of those with high intelligence (score, 7 to 9) were affected by BMI on the absolute scale. The higher the intelligence was, the larger the absolute consequences of being obese.

image

Figure 1. Associations between categories of BMI and high education stratified by intelligence test performance scores in 752,283 Swedish men. * Unadjusted prevalence of high education (left) and corresponding fully adjusted HRs with 95% CIs (right).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

There are four main findings. First, obese men in Sweden are doing worse in the educational system than their normal-weight counterparts, even though adjustments were made for intelligence, parental education, and parental SEI group. Strong associations between obesity and education in a longitudinal setting are novel findings. Second, few obese individuals start a university education compared with their normal-weight counterparts, but even fewer graduate. Third, at any given intelligence level, men obese at age 18 years had lower school marks in ninth grade than men who were normal weight at age 18 years. Fourth, our results indicate that obese pupils may have received lower marks than would be expected based on their intelligence test performance, because school marks in ninth grade better explained the residual association between obesity and educational attainment than the intelligence performance test at age 18.

This study has several strengths. It is nationwide, population-based, and longitudinal. Furthermore, we have objective measures of height, weight, childhood socioeconomic position, parental education, and attained education. Because the estimated ORs from the logistic regression including 925,898 observations were similar to the estimated ORs from the logistic regression including the same 752,283 observations included in the main Cox regression analyses, the risk of selection bias seems unlikely. Because of this and the relatively high participation rate in the study, it is reasonable to believe that the results are generalizable to all men in Sweden born from 1952 to 1973. It is, however, a weakness that our study is limited to men, and the results cannot be generalized to women. In Sweden, military enlistment examinations are compulsory for men only, and similar data on women were not available.

There are several potential explanations for our results. First, obese individuals might perceive themselves, and other obese individuals, as being inferior and less valuable than normal-weight people (15). This might lead to low self-esteem among obese individuals, creating an invisible barrier for attainment of a higher education. In one study, obese adolescent boys were more likely to believe that they were poor students, more prone to expect that they would drop out of school, and also more likely to being held back a grade compared with their normal-weight peers (16). They did not, however, dislike school to any greater extent than their peers (16). It is a limitation that we were unable to adjust for psychological factors such as self-esteem in our study.

Second, Silventoinen et al. (17) suggested that common genetic factors affecting both BMI and education are likely explanations for the correlation between these two variables. Good school performance has been shown to be inversely associated with being obese in boys and girls in both developed and developing countries (18, 19). However, we have shown in our analyses that differences in intellectual capacity could explain only a small part of the differences in attained education between obese and normal-weight men.

There might also be medical complications behind the lower educational achievements of obese men. This is, however, unlikely in the present study because the men were relatively young (mean age, 27.3) when attaining a high education. Medical complications of obesity are much more common in higher ages.

A recent review on discrimination of obese subjects suggested that discrimination occurs in schools, health services, and employment settings (20). Discrimination against obese individuals in educational settings might lead to lower attained education. The first obstacle to obese children is peer rejection in school (21, 22). In addition, one study reported that teachers believe obesity to be under individual control, and the teachers agreed that obese children are untidy, more emotional, and less likely to succeed at work (23). Furthermore, teachers might expect less from obese pupils and might not encourage obese pupils as much as they encourage normal-weight pupils.

To further understand the determinants of attained education, we analyzed the impact of school marks in ninth grade. Our results indicated that obese pupils had received lower marks than would be expected based on their intelligence test performance. We are, however, unable to establish patterns of causality, and it is important to keep in mind that both BMI and intelligence test performance were measured after the school marks were set. However, BMI is tracked over time, and BMI at age 15 is strongly correlated with BMI 3 years later (24). It is, therefore, likely that most of the subjects who were obese at age 18 had developed the condition at age 15 years or earlier in childhood.

Studies from the U.S. have shown discrimination against obese adolescents in college admission and parents willingness to pay for their children's education (25, 26). However, this scenario is less likely in Sweden because university education is free of charge, and university admissibility criteria are based mostly on school marks. However, our result that few obese young men start a university education compared with normal-weight young men and even fewer men graduate indicates that discrimination is likely to occur in Swedish universities. Consequently, we are unable to exclude discrimination as a contributing factor behind the low educational achievements of the obese men in our study.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

We thank The Bank of Sweden Tercentenary Foundation for financial support of this study (Contract J2003-0543).

Footnotes
  • 1

    Nonstandard abbreviations: SEI, socioeconomic index; HR, hazard ratio; OR, odds ratio; CI, confidence interval.

  • The costs of publication of this article were defrayed, in part, by the payment of page charges. This article must, therefore, be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  • 1
    Sobal, J. (1991) Obesity and socioeconomic status: a framework for examining relationships between physical and social variables. Med Anthropol. 13: 231247.
  • 2
    Sonne-Holm, S., Sorensen, TI (1986) Prospective study of attainment of social class of severely obese subjects in relation to parental social class, intelligence, and education. Br Med J (Clin Res Ed) 292: 586589.
  • 3
    Wardle, J., Waller, J., Jarvis, MJ (2002) Sex differences in the association of socioeconomic status with obesity. Am J Public Health 92: 12991304.
  • 4
    Sarlio-Lahteenkorva, S., Lahelma, E. (1999) The association of body mass index with social and economic disadvantage in women and men. Int J Epidemiol. 28: 445449.
  • 5
    Gortmaker, S. L., Must, A., Perrin, J. M., Sobol, A. M., Dietz, WH (1993) Social and economic consequences of overweight in adolescence and young adulthood. N Engl J Med. 329: 10081012.
  • 6
    World Health Organization (1998) Obesity: Preventing and Managing the Global Epidemic: Result of a WHO Consultation on Obesity World Health Organization Geneva, Switzerland.
  • 7
    David, A. S., Malmberg, A., Brandt, L., Allebeck, P., Lewis, G. (1997) IQ and risk for schizophrenia: a population-based cohort study. Psychol Med. 27: 13111323.
  • 8
    Jiang, G. X., Rasmussen, F., Wasserman, D. (1999) Short stature and poor psychological performance: risk factors for attempted suicide among Swedish male conscripts. Acta Psychiatr Scand. 100: 433440.
  • 9
    Otto, U. (1976) Male youths: a sociopsychiatric study of a total annual population of Swedish adolescent boys. Acta Paediatr Scand. 264: 547.
  • 10
    Carlstedt, B. (2000) Cognitive abilities: aspects of structure, process and measurement. Goteborg Studies in Educational Sciences 148: 161. Universitas Gothoburgensis Goteborg, Sweden.
  • 11
    Harrison, G., Fouskakis, D., Rasmussen, F., Tynelius, P., Sipos, A., Gunnell, D. (2003) Association between psychotic disorder and urban place of birth is not mediated by obstetric complications or childhood socio-economic position: a cohort study. Psychol Med. 33: 723731.
  • 12
    SAS Institute I (2002) SAS/STAT User's Guide: Version 9.1 SAS Institute, Inc Cary, NC.
  • 13
    Zhang, J., Yu, KF (1998) What's the relative risk? A method of correcting the OR in cohort studies of common outcomes. JAMA 280: 16901691.
  • 14
    Magnusson, P. K., Rasmussen, F., Gyllensten, UB (2006) Height at age 18 years is a strong predictor of attained education later in life: cohort study of over 950 000 Swedish men. Int J Epidemiol. 35: 658663.
  • 15
    Wang, S. S., Brownell, K. D., Wadden, TA (2004) The influence of the stigma of obesity on overweight individuals. Int J Obes Relat Metab Disord. 28: 13331337.
  • 16
    Falkner, N. H., Neumark-Sztainer, D., Story, M., Jeffery, R. W., Beuhring, T., Resnick, MD (2001) Social, educational, and psychological correlates of weight status in adolescents. Obes Res. 9: 3242.
  • 17
    Silventoinen, K., Sarlio-Lahteenkorva, S., Koskenvuo, M., Lahelma, E., Kaprio, J. (2004) Effect of environmental and genetic factors on education-associated disparities in weight and weight gain: a study of Finnish adult twins. Am J Clin Nutr. 80: 815822.
  • 18
    Mikkila, V., Lahti-Koski, M., Pietinen, P., Virtanen, S. M., Rimpela, M. (2003) Associates of obesity and weight dissatisfaction among Finnish adolescents. Public Health Nutr. 6: 4956.
  • 19
    Mo-Suwan, L., Lebel, L., Puetpaiboon, A., Junjana, C. (1999) School performance and weight status of children and young adolescents in a transitional society in Thailand. Int J Obes Relat Metab Disord. 23: 272277.
  • 20
    Puhl, R., Brownell, KD (2001) Bias, discrimination, and obesity. Obes Res. 9: 788805.
  • 21
    Richardsson, S. A., Goodman, N., Hastorft, A. H., Dornbusch, SM (1961) Cultural uniformity in reaction to physical disabilities. Sociol Rev. 26: 241247.
  • 22
    Latner, J. D., Stunkard, AJ (2003) Getting worse: the stigmatization of obese children. Obes Res. 11: 452456.
  • 23
    Neumark-Stzainer, D., Story, M., Harris, T. (1999) Beliefs and attitudes about obesity among teachers and school health care providers working with adolescents. J Nutr Educ. 31: 39.
  • 24
    Guo, S. S., Wu, W., Chumlea, W. C., Roche, AF (2002) Predicting overweight and obesity in adulthood from body mass index values in childhood and adolescence. Am J Clin Nutr. 76: 653658.
  • 25
    Crandall, CS (1991) Do parents discriminate against their heavy-weight daughters? Pers Soc Psycol Bull. 21: 724735.
  • 26
    Crandall, CS (1991) Do heavyweight students have more difficulty paying for collage? Pers Soc Psychol. 17: 606611.