The European Journal of Public Health Advance Access originally published online on June 14, 2006
The European Journal of Public Health 2007 17(1):21-26; doi:10.1093/eurpub/ckl087
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Child and adolescent health |
Health inequalities among adolescents—the impact of academic orientation and parents' education
Curt E. I. Hagquist** Karlstad University SE-651 88 Karlstad, Sweden
Correspondence: Curt E. I. Hagquist, Karlstad University, SE-651 88 Karlstad, Sweden, tel.: +46 54 700 2536, fax: +46 54 700 2523, e-mail: curt.hagquist{at}kau.se
Received April 12, 2006, accepted May 15, 2006
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
Background: Traditionally, the socio-economic position of adolescents has been measured using information about parents' occupation, parents' level of education, or household income. Since the adolescence is a developmental stage characterised by a search for and a move into individual life tracks a shift of focus from socio-economic position of origin to socio-economic position of destination is justified. Academic orientation may be used as a rough indicator of future social position. The purpose of the study was to elucidate the link between academic orientation and parents' education on the one hand and subjective health and health-related behaviour among adolescents on the other. Methods: The study was based on cross-sectional questionnaire data collected in 1999 and 2003 among 1828 18-year-old students in year 2 of upper secondary school in a Swedish city. The data were analysed using contingency tables and logistic regression. Results: Subjective health and health-related behaviour was strongly linked to academic orientation but not directly to parents' education. The pattern is unambiguous, poor health and health-damaging behaviour being significantly higher among students in non-theoretical programmes than among students in theoretical programmes. Conclusion: Academic orientation is a useful concept in order to detect health inequalities and a powerful way of identifying adolescents at higher risk. The unequal distribution of health and health-damaging behaviour according to academic orientation among adolescents turns out to be an important challenge for public health work.
Keywords: academic orientation, adolescents, health inequalities, health-related behaviour, social class
While there is overwhelming evidence of socio-economic differences in health in early childhood and adulthood, the situation among adolescents is ambiguous and the patterns are less visible.1–3 In fact, there are indications that adolescence may be characterised by equalisation of health.3 The complexity of the relationship in adolescence between socio-economic position of origin and health is conveyed in papers based primarily on British data,1,2 as well as in research reports including US studies.4 The presence or absence of socio-economic differences in adolescence seems to depend on what aspects of health are focused.2,4 However, reviewing the literature, West found no clear link between socio-economic position among adolescents and their psychological well-being2, while Starfield in her own data and in most published studies recognised a social class gradient in perceived well-being.4
For health-related behaviour, the reports on socio-economic distribution among adolescents also diverge. For instance, socio-economic patterns paralleling those in adult populations have been found in some studies,5 while inconsistent patterns have been reported in others.6
Traditionally, the socio-economic position of adolescents has been measured using information about parents' occupation, parents' level of education, or household income.3 In addition to these classical socio-economic indicators, alternative indicators or proxies have been elaborated and used in some surveys. Such measures include self-assessment of the socio-economic status of the family,7 and measurements of self-reported family affluence.8 Significant relationships to self-rated health among adolescents have been shown for both of these measurements.7,8
Adolescence is a developmental stage characterised by a search for and a move into individual life tracks which justifies a shift of focus from socio-economic position of origin to socio-economic position of destination. Young adults' own occupations and their educational attainment are more strongly linked to health than is socio-economic position of origin.9 Similar patterns have been found for adolescent smoking behaviour10 and adolescent health,11 showing strong relationships to the adolescents' socio-economic position but not to their parents' socio-economic position.
Since most adolescents have not yet achieved a socio-economic position as individuals it is logical to focus on their academic orientation. This concept may be viewed as a rough indicator of future social position, which can be used to study the distribution of health among different groups of adolescents. The concept of academic orientation may include aspects of school achievement and educational intentions.12,13 For the purpose of the present analysis of older adolescents, current programme affiliation at upper secondary school is used as an indicator of academic orientation.
The purpose of the study was to elucidate the link between academic orientation and parents' education on the one hand and subjective health and health-related behaviour among adolescents on the other.
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Target group
Almost all students in Sweden attend a programme for 3 years at upper secondary school after they have finished the compulsory school. The present study is based on data collected in 1999 and 2003 among students in year 2 of the upper secondary school in a Swedish city. The targeted population consisted of 1376 students in 1999 and 1513 students in 2003 who attended six different schools. Excluding those school classes which did not participate, the attrition rates in 1999 and 2003 were 16 and 22%, respectively.
By pooling the data from 1999 and 2003, the analysis covered 1828 students who were 18 years old or would reach that age during the year of investigation. These students represented 80.2% of the entire sample. In general, there were no or very small internal loss on the questions used in the analyses; for parents' education it was 8.2%.
Data collection
Data were collected using a self-administered questionnaire for completion anonymously in the classroom. Participation was voluntary. The data collection was carried out in accordance with research ethics principles in humanistic–social science research stipulated by the Swedish Research Council. The questionnaire covered a broad range of topics, tapping information about subjective health, health-related behaviour, school environment, etc.
Instruments
General perceived health was measured with the question:
- Would you say your health is? (Excellent; Quite good; Poor)
Psychosomatic complaints were measured with the following items: Had difficulty in concentrating, had difficulty in sleeping, suffered from headaches, suffered from stomach aches, felt tense, had little appetite, felt sad, and felt giddy. The response categories for all of these eight items, which are in the form of questions, were: never, seldom, sometimes, often, and always.
Tobacco habits were measured in the following ways:
- Smoking: Do you smoke? (No, never; No, merely tried; No, have given up; Yes, only sometimes; Yes, only at parties; Yes, only on weekends; Yes, almost every day; Yes, every day).
- Using snuff (= Swedish snus, i.e. moist tobacco tucked under the lip): Do you take snuff? (No, never; No, merely tried; No, have given up; Yes, every day; Yes, almost every day; Yes, only sometimes).
Alcohol consumption (binge drinking) and drug use were measured in the following ways:
- Alcohol: How often do you drink alcohol equivalent to at least half a bottle of spirits or a whole bottle of wine or four cans of strong beer or six cans of medium-strength beer on the same occasion? (Don't drink alcohol; Once a week; Twice a month; Once a month; A few times a year; More seldom; Never).
- Drugs: Have you ever used narcotics (for example hashish, marijuana, amphetamines, ecstasy, LSD, heroin) (Yes; No).
Physical activity during leisure time was measured with one question:
- How often do you take exercise in your free time, so that you get out of breath or sweaty? (Every day; 4–6 times a week; 2–3 times a week; Once a week; Once a month; Less than once a month; Never).
In addition to the outcome variables previously described, three independent variables were used for the contingency table analysis:
- Sex (boy; girl).
- Parents' highest educational level: (Both parents, compulsory; Neither higher than upper secondary school; At least one a graduate).
- Academic orientation: Upper secondary school programme (Theoretical = mainly natural and social sciences; Non-theoretical = vocational, individual, and other non-theoretical programmes).
Two additional independent variables were used for the logistic regression analysis:
- Parents' country of birth: (Both Sweden; Others).
- Year of investigation (1999; 2003).
Analysis
A composite measure of psychosomatic health was constructed based on a summation of the responses to the eight items on psychosomatic complaints. The psychometric properties of this measure were examined with the Rasch model14,15 with a focus on the operating characteristics of the items. At a general level of analysis, the data on psychosomatic complaints met measurement criteria of invariant comparison and proper ordering of the data. As a result of the Rasch analysis each person is placed on a logit scale generated by the analysis. The person location values are the person parameters estimates, which in turn comprise transformed raw scores. Low values on the logit scale indicate a relatively low degree of psychosomatic complaints while high values indicate a relatively high degree of psychosomatic complaints. Using percentile values based on the entire sample as cut-off points, students at and above/below certain levels was compared for different classifications of individuals.
The data were subjected to contingency table analysis (tables 1–3) and multinomial logistic regression (table 4).
|
The differences in proportions across different levels of parents' education and school programme reported in the contingency tables were examined with two-tailed z-tests conducted separately for boys and girls.
The multinomial logistic regression analysis was conducted on all eight outcome measures using two different kinds of models: (i) a model including only the main effects of the five independent variables; (ii) a model including all main effects and an interaction term (parents education * school programme). The goodness of fit of both kinds of models was tested by means of the total likelihood ratio test statistic (marginal G2). The partitioned likelihood ratio test statistic (conditional G2) was used in order to compare the fit of the main effect models compared to the models including an interaction term.
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
In table 1 the proportion of adolescents in theoretical and non-theoretical programmes is shown in relation to the educational levels of their parents.
Table 1 shows significant relationships between the educational level of the parents and the academic orientation of the students. Students whose parents had attended a university were much more likely to be in a theoretical programme than those whose parents had no university background.
In table 2 health and health-related behaviour among students is shown in relation to the educational levels of their parents.
|
Table 2 shows that there were no significant differences in health when the students were classified according to parents' educational levels.
Two health-related behaviours among girls and one among boys showed significant relationships with parents' educational levels, i.e. the highest proportions of health-damaging behaviour occurred among students whose parents lacked university background: Smoking habits and physical activity among female students were clearly differentiated by the educational level of their parents, and so was use of snuff among male students.
In table 3 health and health-related behaviours among students are shown distributed by students' academic orientation.
|
Table 3 shows differences in health when the students were classified according to their academic orientation. The proportion of students who reported excellent health was significantly higher among those in theoretical programmes. Psychosomatic complaints show a similar pattern in the sense that students in theoretical programmes reported less problems than students in non-theoretical programmes.
All five health-related behaviours examined showed significant links to the students' academic orientation, i.e. the highest proportions of health-damaging behaviour occurred among students in non-theoretical programmes. In particular, smoking habits were clearly differentiated by academic orientation. The proportion of daily smokers was three times higher among students in non-theoretical programmes than among those in theoretical programmes. Large differences according to academic orientation were also found for physical inactivity. The proportion of inactive girls was twice as high among those in non-theoretical programmes as among those in theoretical programmes.
In table 4 multinomial logistic regression analysis of health and health-related behaviour among students is shown, with a focus on the effect of academic orientation and parents' educational level, using sex, year of investigation, and parents' country of birth primarily as control variables.
Table 4 shows that the marginal G2-values are non-significant for all eight models which were analysed, indicating good fit for the main effect models. The table shows that the odds ratios for health, psychosomatic complaints, and all five variables of health-related behaviour are significant with respect to academic orientation. In these multivariate analyses the odds of having relatively bad health or health-damaging behaviour were significantly higher among students in non-theoretical programmes than among those in theoretical programmes. For instance, the odds of having poor health were more than three times higher among students in non-theoretical programmes than among those in theoretical programmes. The odds of drug use and physical inactivity were twice as high in the non-theoretical group of students as in the theoretical group. The differences were even greater for daily smoking, where the odds were more than four times higher among students in non-theoretical programmes than among those in theoretical programmes. Except for daily use of snuff, there were not any significant main effects for parents' educational level.
|
Including an interaction term (parents education * school programme) provides similar results as for the main effect models with respect to the values of the marginal G2, i.e. good fit for all models (not reported in the table). Comparing the main effects models with the models also including an interaction term shows that the latter models do not improve the fit for seven out of eight models, which is indicated by non-significant conditional G2-values. In the daily-smoking case, however, inclusion of an interaction term improved the fit, which is indicated by a significant conditional G2-value (P < 0.05).
The inclusion of an interaction term for parents' education and school programme in the analysis of daily smoking does not imply a significant interaction effect when the lowest level of parents' education is contrasted to the highest level. However, the contrast between parents' with upper secondary school background to parents' with university background shows a significant interaction effect: among students in non-theoretical programmes the odds for smoking is higher for those whose parents' have a upper secondary school background compared to those whose parents' have a university background while the relationship is the opposite among students in theoretical programmes.
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
The analysis focused on the current distribution of subjective health and health-related behaviour among 18-year-old adolescents classified according to parents' education and academic orientation. The contingency table analysis carried out separately on each of these independent variables indicated a stronger link to subjective health and health-related behaviour for academic orientation than for parents' education. Applying these two variables simultaneously using logistic regression analysis, the previous effects of parents' education disappeared for all health variables and for four of five variables of health-related behaviour. In contrast, in the logistic regression analysis all the effects of academic orientation remained significant. The pattern was unambiguous, the odds of having poor health as well as the odds of health-damaging behaviours being significantly higher among students in non-theoretical programmes than among those in theoretical programmes.
At face value the absence of an effect of parents' education may seem surprising, given the strong link reported between social class and health in adult populations.16 However, given the literature review mainly based on British data previously referred to2 the outcomes seem less unexpected. The results are also consistent with findings that adolescents' smoking behaviour shows a stronger link to current social class than to social class background.10,17 Moreover, although no direct effect was manifested in the logistic regression analysis, there were probably mediating effects. Thus the selection of students into different programmes at upper secondary school was strongly related to the educational level of the parents, placing students with highly educated parents into theoretical programmes.
A key question is why the social class and health relationships among adolescents do not parallel those among adults? Is this because of how social class is conceptualised, operationalised, and measured, or may it be due to lifestyles that bridge the gaps between adolescents with different social class backgrounds?
Equalisation of health among adolescents has so far been discussed primarily in connection with social class of origin, while it make sense to consider the possible equalisation processes also from the view of social class of destination. Equalisation processes may well operate in different ways depending on what concept of social position is analysed. These possible processes are not necessarily mutually exclusive, but may operate simultaneously although in opposite directions. Peer group influences in combination with lifestyles bridging class borders may make the adolescent group more homogeneous with respect to health, while academic orientation may work in the opposite way, making the group more heterogeneous.
Not only the patterns of social class inequalities seem complex among adolescents, but also the mechanisms surrounding the hypothesised process of health equalisation as well as possible causal pathways. Theoretically, social class differences in health among adolescents and among adults may be an artefact, or due to health selection or causality, while in reality both selection and causality are likely to occur simultaneously, thereby blurring the distinction between cause and effect.
Extending the life course perspective beyond adolescence, Finnish studies show that poor perceived health and health-compromising behaviour in adolescence is accompanied by low educational level in adulthood.18 However, compared to health-related behaviour, health reveals a weaker independent relationship with future educational level after adjustment for school achievement and socio-demographic background.19 Another Finnish follow-up study of psychosomatic symptoms from adolescence to adulthood indicates slightly different patterns for females compared to males: among the former group low SES was likely to be both a cause and a consequence, while among the latter group more the health selection hypothesis was supported.20
Regardless of the directions of the link between educational track and health, the focus on academic orientation turns out to be a powerful way of identifying adolescents at higher risk. From a public health perspective, the unequal distribution of health and health-damaging behaviour among adolescents should affect the allocation of resources for health prevention and health promotion, giving priority to less favoured groups of adolescents. Also from a measurement point of view academic orientation is a useful concept since it can be operationalised in a way that enables classification of all students at their age of 15–16.
Finally, a few methodological remarks. Although the results reveal distinct patterns, due to the cross-sectional nature of the data, the analysis is descriptive, solely permitting hypotheses about possible pathways and causal effects. Further, if applied on a bigger sample the analysis of academic orientation may be elaborated further by analysing possible differences within the groups of students in the theoretical programmes and the non-theoretical programme groups, respectively.
| Key points
|
|
Acknowledgments |
|---|
The paper is based on data that were collected by financial support from the City of Karlstad, Sweden.
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
1 West P. (1988) Inequalities? Social class differentials in health in British youth. Soc Sci Med. 27:291–6.[CrossRef][Web of Science][Medline]
2 West P. (1997) Health inequalities in the early years: is there equalisation in youth? Soc Sci Med. 44:833–58.[CrossRef][Web of Science][Medline]
3 West P and Sweeting H. (2004) Evidence on equalisation in health in youth from the West of Scotland. Soc Sci Med 59:13–27.[CrossRef][Web of Science][Medline]
4 Starfield B, Riley AW, Witt WP, Robertson J. (2002) Social class gradients in health during adolescence. J Epidemiol Community Health 56:354–61.
5 Conrad KM, Flay BR, Hill D. (1992) Why children start smoking cigarettes: predictors of onset. Br J Addict 87:1711–24.[CrossRef][Web of Science][Medline]
6 Tuinstra J, Groothoff JW, Van Den Heuvel WJA, Post D. (1998) Socio-economic differences in health risk behaviour in adolescence: do they exist? Soc Sci Med 47:67–74.[CrossRef][Web of Science][Medline]
7 Piko B and Fitzpatrick KM. (2001) Does class matter? SES and psychological health among Hungarian adolescents. Soc Sci Med 53:817–30.[CrossRef][Web of Science][Medline]
8 Currie CE, Elton RA, Todd J, Platt S. (1997) Indicators of socioeconomic status for adolescents: the WHO health behaviour in school-aged children survey. Health Educ Res 12:385–97.
9 Rahkonen O, Arber S, Lahelma E. (1995) Health inequalities in early adulthood: a comparison of young men and women in Britain and Finland. Soc Sci Med 41:163–71.[CrossRef][Web of Science][Medline]
10 Glendinning A, Shucksmith J, Hendry L. (1994) Social class and adolescent smoking behaviour. Soc Sci Med 38:1449–60.[CrossRef][Web of Science][Medline]
11 Glendinning A, Love JG, Hendry LB, Shucksmith J. (1992) Adolescence and health inequalities: extensions to Macintyre and West. Soc Sci Med 35:679–87.[CrossRef][Web of Science][Medline]
12 Hagquist C. (2000) Socioeconomic differences in smoking behaviour among adolescents. The role of academic orientation. Childhood 7:467–78.
13 Tucker JS, Ellickson PL, Klein DJ. (2003) Predictors of the transition to regular smoking during adolescence and young adulthood. J Adolesc Health 32:314–24.[CrossRef][Web of Science][Medline]
14 Andrich D. (1988) Rasch Models for Measurement. (Sage Publications, Newbury Park).
15 Hagquist C. (2001) Evaluating composite health measures using Rasch-modelling: an illustrative example. Soc Prev Med 46:369–78.
16 Townsend P and Davidson N. (1992) The Black Report. (Penguin Books, London) In: Inequalities in health.
17 Vereecken CA, Maes L, De Bacquer D. (2004) The influence of parental occupation and the pupils' educational level on lifestyle behaviours among adolescents in Belgium. J Adolesc Health 34:330–8.[Web of Science][Medline]
18 Koivusilta L. (2000) Health-related selection into educational tracks. A mechanism producing socio-economic health differences. (University of Turku, Turku, Finland).
19 Koivusilta L, Rimpelä A, Vikat A. (2003) Health behaviours and health in adolescence as predictors of educational level in adulthood: a follow-up study from Finland. Soc Sci Med 57:577–93.[CrossRef][Web of Science][Medline]
20 Huurre T, Rahkonen O, Komulainen E, Aro H. (2005) Socioeconomic status as cause and consequence of psychosomatic symptoms from adolescence to adulthood. Soc Psychiatry Psychiatr Epidemiol 40:580–7.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. Havas, H. Bosma, C. Spreeuwenberg, and F. J. Feron Mental health problems of Dutch adolescents: the association with adolescents' and their parents' educational level Eur J Public Health, November 3, 2009; (2009) ckp172v1. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||