Health Education Research, Vol. 16, No. 2, 201-214,
April 2001
© 2001 Oxford University Press
Health behavior-based selection into educational tracks starts in early adolescence
Department of Public Health, University of Turku, Lemminkäisenkatu 1, 20520 Turku,
1 Tampere School of Public Health, University of Tampere, 33014 Tampere and
2 National Research and Development Centre for Welfare and Health (Stakes), Box 220, 00531 Helsinki, Finland
 |
Abstract |
|---|
 Top Abstract IntroductionData and methodsResultsDiscussionReferences |
|---|
Health behaviors and educational tracks of an individual are here presumed to have a strengthening influence on each other during the developmental process, through which individuals gradually reach their adult health and social position. This longitudinal study of a Finnish nationally representative sample of 12 year olds born in 1970 (N = 1009) examined the associations of health behaviors at ages 12 and 14 with educational track at age 16. The dependent variable, educational track, classified the respondents into five successive categories, thought to predict their adult social position. Selection into different educational tracks according to health behaviors was obvious already at age 12, when frequency of tooth brushing, consumption of sweets, coffee drinking and level of participation in physical exercise predicted educational track independently of sociodemographic background. At age 14, the independent predictors were smoking, frequency of tooth brushing and coffee drinking. At both ages, sociodemographic factors had independent associations with educational track. It seems that certain health-related behaviors in early adolescence are indicators of a person's possibilities to benefit from a country's educational supply. Both sociodemographic background and health-related behaviors influence the process of selection into educational tracks leading to social position and health in adulthood.
|
Introduction |
|---|
|
TopAbstractIntroductionData and methodsResultsDiscussionReferences |
|---|
In modern specialized occupational structures, entering employment and advancement in one's career is strongly dependent on educational qualifications. The educational success of young people is closely related to their health-related behaviors in such a way that a way of life including various health-compromising behaviors is typical of those who do not see education as a giver of a good life or who do not have the resources needed for success in their educational careers (Nutbeam et al., 1989
; Glendinning et al., 1995). The same set of health-endangering behaviors is typical of members of lower social classes in adulthood (Macintyre, 1986). There is evidence from Finland that by the age of 16, the lifestyle profiles have already differentiated so that adolescents with health- damaging behaviors are selected into educational tracks leading to a low educational level in adulthood (Koivusilta et al., 1998). A recent study by Rönkä (Rönkä, 1999) showed that problems of social functioning tend to interact and co-occur throughout peoples' life careers, leading to marginalization among those who do not have supportive experiences in their families or in school.
There is evidence from longitudinal studies that life circumstances in early life are associated with health later in life (Power and Matthews, 1997) and educational achievement (Hollo, 1999). Also, the association of health behaviors with educational success and health has been confirmed in follow-up studies (Glendinning et al., 1994, 1995). However, since the health behaviors that have a fundamental influence on public health, especially smoking and alcohol use, begin to develop into habits in adolescence, it is important to find out precisely when the differentiation towards various behavioral patterns occurs. There is a great need for basic research using longitudinal designs on the adoption of health-related behaviors during the sensitive phase preceding important educational decisions, i.e. that of early adolescence.
Smoking has a central role in dividing adolescents into those who feel happy and do well in school, and those who experience poor success and reluctance to take part in further education after the compulsory phase. Smoking has been considered a marker of a broader lifestyle, typical of which are rejection of an achievement ideology, rebelliousness and the adoption of non-conventional values in society (West, 1991; Glendinning et al., 1995; Donovan et al., 1991). However, as the prevalence of different behaviors varies with age, for younger adolescents some other behaviors may reflect the same phenomenon of lack of success in school, which smoking reflects later.
People adopt behaviors for a variety of reasons, such as curiosity, desire to experiment and media images of behaviors (Friedman, 1989; Aloise-Young et al., 1996). Peer influence is especially important in adolescence (Norton et al., 1998). Among factors leading to health-related behaviors, great explanatory power has been ascribed to stress. Inappropriate behaviors are said to come from inadequate coping styles, such as helplessness or hopelessness, altered perceptions of risk and vulnerability, and an undue willingness to behave in risky ways (Rutter and Quine, 1994). In adolescence, pressure to succeed academically is one major stressor, along with other developmental tasks and social role expectations (Hurrelmann and Maggs, 1995). The stressfulness of this phase of life is depicted by the occurrence of various symptoms (Koivusilta et al., 1998). Young people coming from different social environments confront these stressors with different amounts of social, personal and economic resources, such as social support, problem-solving facilities and money to engage in popular activities (Taylor and Repetti, 1997; Caspi et al., 1998; Bosma et al., 1999). Some personality traits, such as aggressiveness, strongly influence a young person's ability to cope constructively in life (Rönkä and Pulkkinen, 1995).
Educational decisions and success are strongly influenced by the parents' social class. Children from higher social class families reach the highest levels of education (Erikson and Jonsson, 1996; Haven, 1998). This depends partly on economic resources, but also on attitudes towards education. Depending on the society, the relative significance of these factors may be different. In Nordic welfare societies, economic resources might influence educational choices less than, for example, in the US. Well-educated parents usually regard education as a valuable life resource, and transmit to their children the cultural capital needed in adapting to school values and in navigating through the educational system (Bourdieu and Passeron, 1977). They also tend to have a great confidence in their children's success and in their own capacity to get them through the more demanding tracks (Erikson and Jonsson, 1996). However, educational achievement is also influenced by personality characteristics, e.g. constructiveness, stability and compliance in childhood and adolescence, which indicate high self-control of emotions, have been found to predict good educational and work achievement in middle age (Pulkkinen et al., 1999).
The aim of this study is to find out whether health behaviors in early adolescence predict the educational track an individual will follow after the compulsory education. More specifically, we want to identify which particular behaviors at ages 12 and 14 are associated with the educational track at age 16, and whether the adoption of these behaviors is a response to stress. The conceptual model of the study is outlined in Figure 1. The associations of the health-related lifestyle in early adolescence with educational track at age 16 are assessed with adjustment for sociodemographic background. On the right are the presumed consequences of the process, an individual's educational level and health in adulthood.
|
|
Data and methods |
|---|
|
TopAbstractIntroductionData and methodsResultsDiscussionReferences |
|---|
Study design
The data used in this study were drawn from the Adolescent Health and Lifestyle Survey data base, which is a national monitoring system of adolescent health and health behaviors (Rimpelä et al., 1988
). Data were collected from a national sample of Finns, born between 20 and 25 July 1970. The sample was drawn from the Population Register Centre in February 1983, when the cohort members were 12 years old. The questionnaire was sent to the whole baseline sample, which was then followed at ages 14 and 16 (in 1985 and 1987). In order to get the maximum number of respondents for the analyses, two data sets were formed. Data1 consisted of those who responded both in 1983 and in 1987, and Data2 consisted of those who responded in 1985 and in 1987 (Figure 2). The first inquiry was always sent in February and was followed by two further inquiries to those who had not responded to the first one. Responding was voluntary and the purpose of the study was explained in a covering letter.
|
Dependent variable
In the Finnish educational system, after 9 years of compulsory schooling (basic education), a division is made at age 16 into upper secondary schools and vocational or other schools (mostly vocationally oriented educational institutions) (Statistics Finland, 1995
). The channel to academic degrees and thus supposedly to higher social positions has traditionally gone through upper secondary schools. The dependent variable, educational track at age 16, was formed by classifying the respondents into five successive categories, expected to predict their social position in adulthood. The first category consisted of those who were presumed to have the poorest social prospects, i.e. the lowest probability of reaching a high social position in adulthood. The fifth category consisted of those who were presumed to have the best social prospects, i.e. the highest probability of reaching a high social position in adulthood. The categories of educational track were formed according to the type of school the respondents were attending and to their achievement, measured by the pupil's own assessment of his or her position in the class, according to the average grades in the preceding end-of-term school report. Information about school achievement was missing in 4% of the respondents in Data1 and in 5% in Data2. Replacing the missing values by the most typical average grades in each school type did not change the magnitude of the associations between educational track and the independent variables. The first of the ordered categories, i.e. the ones with the poorest social prospects, consisted of those who were not attending school at age 16 (6 % in Data1 and 7% in Data2, n1 = 47, n2 = 49). The second category consisted of pupils in vocational or other schools, who had at most average school achievement (28% in both data sets, n1 = 210, n2 = 209). The third category consisted of pupils in vocational or other schools with above average school achievement (10% in Data1 and 9% in Data, n1 = 76, n2 = 68). The fourth category consisted of pupils in upper secondary schools with at most average school achievement (26% in both data sets, n1 = 196, n2 = 189). The fifth category, i.e. those with the best social prospects, consisted of pupils in upper secondary schools with above average school achievement (30% in both data sets, n1 = 223, n2 = 218).
Independent variables
Independent variables were divided into three main groups (see below). A more detailed description of the independent variables has been published earlier (Koivusilta et al., 1999, 1998).
Sociodemographic background
Sociodemographic background was, at both ages, described by six variables: father's or other guardian's education (high, middle, low), father's or other guardian's occupation (upper white- collar workers, lower white-collar workers, farmers, blue-collar workers), the urbanization level of the place of residence (capital area and large towns, small towns and villages, sparsely populated rural municipalities), the geographical region of residence (South, South-West, Central-West, East, North), family type (nuclear family, non-nuclear family) and gender (male, female).
Health behaviors
Health behaviors included in the study were known to be risk factors for diseases that cause a major burden to public health, such as cardiovascular disease, certain cancers, dental caries, etc. (Rimpelä et al., 1988). The names of the variables as well as their categories are shown in Table I. The same value categories were used at ages 12 and 14, except for alcohol use and smoking. At age 12, the categories of alcohol use were: none (did not drink alcohol), other (drank alcohol). At age 14, the categories were: none, controlled drinking (drank at most twice a month and never got drunk), uncontrolled drinking (drank at least weekly, as well as those who drank more seldom but got drunk sometimes). At age 12, the categories of smoking were: never tried, had smoked once, had smoked more. (No one smoked daily.) At age 14, the categories were: never tried, experimental or occasional, daily.
|
Health
Health variables were categorized as in Table I
. The variable psychosomatic symptoms formed a sum-index constructed from the frequency of experiencing eight symptoms (abdominal pains, anxiety, irritability or fits of anger, difficulties in sleeping, headache, tremor of hands, fatigue or weakness, fits of dizziness) during the preceding 6 months. Cut-off points were set to leave at most 30% of respondents in the lower and upper ends of the distribution.
Statistical procedures
The associations between educational track and the independent variables were analyzed using polychotomous logistic regression models (Agresti, 1990). Because of the ordinal nature of the dependent variable, cumulative logistic models were used. The statistical model has been described earlier (Koivusilta et al., 1999). As a first stage of analysis, tests were done to find out whether interactions existed between gender and any other independent variables. Since no interactions were significant at either age, gender was taken into the analyses as a confounding variable. A stepwise procedure was then carried out for both data sets. First, those behavior and health variables were selected that showed univariate associations with educational track (
2-test). Then, a model including these variables, together with all the sociodemographic background variables was fitted to the data sets. In the final models, only the variables showing significant explanatory power were included. Cumulative odds ratios (COR) with 95% confidence intervals (CI) were calculated for them so that the categories giving approximately the same CORs were combined. Since cumulative models examine the relationships between the categories of an ordinal-type dependent variable, the COR expresses the incidence of upper categories of the dependent variable as compared to the lower categories on various values of the independent variables. P <0.05 was used as the cut-off point of significance.
Since the logistic functions are computed using only the cases which have values for every variable, there was a need to retain as many observations for each variable as possible in the analyses. For the variables father's education and father's occupation, missing values were replaced by their most probable values. As it was supposed that social circumstances in early childhood have far-reaching effects on people's life careers (Barker, 1991), a missing value (e.g. at age 16) was in the first place replaced by the nearest previous value found in the data for this person (a value at age 14). If this measurement was also lacking, then the earlier one (at age 12) was used. When this could not be done, the preceding values were replaced by later ones. The associations of father's education and occupation with educational track were similar before and after the replacement of missing values.
A sensitivity analysis was carried out to study the influence of bias caused by non-response on the associations between educational track and two independent variables, smoking and tooth brushing. These were selected for analysis, because only in them did the drop-out rates from the first inquiry in 1983 to the last inquiry in 1987 differ statistically significantly according to variable category. The missing 257 cases (N – n = 1009 – 752) were generated having such values for smoking, tooth brushing and educational track which made it possible to compare two hypothetical situations—the one where the missing cases would behave educationally like respondents did and the one where missing cases with health-endangering behaviors were found in the same educational tracks as were those with health-promoting behaviors. The sensitivity analysis showed that if missing cases behaved educationally like respondents did, the association between smoking and educational track would not vanish (P < 0.0001). If smoking non-respondents actually belonged to the same educational tracks as did non-smokers, the association between smoking and educational track would be slightly insignificant (P = 0.057). For tooth brushing, the associations were highly significant also after generating the new hypothetical cases.
The analyses were done using the PR program in the BMDP statistical software (Dixon, 1990).
|
Results |
|---|
|
TopAbstractIntroductionData and methodsResultsDiscussionReferences |
|---|
Most of the respondents continued their education after the compulsory phase. Attending upper secondary schools and good attainment was more typical of girls, while boys more often went to vocational or other schools and got average or below average level reports. The distributions of behavior and health variables for the two age groups are shown in Table I
.
At age 12, the amount of psychosomatic symptoms and all the behaviors, except for alcohol use, were statistically significantly associated with educational track in univariate analysis. In the stepwise logistic regression analysis, six behaviors were selected into the set of independent predictors: tooth brushing, consumption of sweets, milk fat and coffee, and both types of exercise. The selection of the symptom variable into the model rendered the smoking variable insignificant. All the background variables were significantly associated with educational track in univariate analysis. When a final model containing all the independent behavioral and health predictors of educational track was fitted into the data, together with the background variables, father's occupation and use of milk fat were no longer statistically significant (Table II). The CORs in the final model show that, at age 12, daily participation in organized physical exercise, very active or very passive participation in unorganized exercise, less frequent than daily tooth brushing, daily coffee drinking and daily use of sweets, as well as a high amount of psychosomatic symptoms increased the risk of belonging to educational tracks with poor social prospects. Also, male gender, father's low education, non-nuclear family type, living in a sparsely populated rural municipality or living in the Central-West part of Finland increased this risk.
|
At age 14, participation in organized physical exercise was the only behavior which was not univariately associated with educational track. Four behaviors were selected into the set of independent predictors: tooth brushing, smoking, and consumption of coffee and milk fat. Both the health variables, as well as alcohol use and participation in unorganized exercise lost their significance when the coffee variable was selected into the model. Finally, the selection of the smoking variable made the significance of alcohol use vanish altogether. All the background variables were significantly associated with educational track in univariate analysis. When the final model of independent predictors, together with the background variables, was fitted into the data, use of milk fat, father's occupation, both of the regional variables and family type lost their significance (Table III
). The CORs of the model show that, by this age, smoking had become an important predictor of educational track in that the probability of an educational track with poor social prospects increased along with the amount of smoking. This probability also increased with a low tooth brushing frequency and with daily coffee drinking, as well as with male gender and father's low education.
|
|
Discussion |
|---|
|
TopAbstractIntroductionData and methodsResultsDiscussionReferences |
|---|
The proportion of respondents in education was similar to the respective proportion in official Finnish statistics for the late 1980s. The proportion of those attending upper secondary schools was slightly higher in the study material than in the population of the same age (Havén, 1998
). The boys' lower response rates led to a slightly higher proportion of girls in the material when compared to the entire Finnish population of the same age (Statistics Finland, 1989). In other Finnish studies, non-response in young men has been associated with poor school performance and negative health behaviors (Pietilä et al., 1995; Rimpelä et al., 1997). This kind of selection would mean that the associations found in this data were weaker than they would have been, had the response rates in boys been as high as in girls.
The sensitivity analysis only slightly pointed to the direction that the unexplored behavior among the smoking non-respondents could be a factor of uncertainty in interpreting the associations found in this study. In the light of other evidence on the associations between health-related behaviors and educational career (Glendinning et al., 1995), we found this fear as being largely unnecessary. Responses to questions about health-related behaviors have been found to somewhat depend on the social desirability of those behaviors in families and among peers (Hebert et al., 1995). Generally, however, there is evidence that the reliability and validity of self-reported behavior measures are good in adolescence (Brener et al., 1995; Lintonen and Rimpelä, 2000).
In forming the dependent variable, it was assumed that its categories reflect people's positions on an ordinal scale. The ordering is based on the knowledge that the selected educational track and school achievement predict an individual's social position in adulthood (Halsey et al., 1980; Havén, 1998). The selected educational track does not straightforwardly predict the educational level to be attained, as changes from one track to another or dropping out may occur. In Finland, in 1987, 6% of the students in daytime full-time upper secondary schools dropped out, although they could later apply for vocational education. In vocational schools, the drop-out rate was 8%. Despite these drop-out rates, the overall predictability of adult social position is thought to be very strong (Havén, 1998). It remains to be seen how the introduction of integrated upper secondary schools and polytechnics into the Finnish school system in the mid-1990s will influence educational decisions in adolescence (Statistics Finland, 1998).
Differences in health among socioeconomic and educational groups are a persistent feature of societies (Mackenbach et al., 1997). This study showed that already as early as at age 12, health behaviors distinguish a group of children whose educational career is likely to be less favorable. Among these behaviors are heavy consumption of coffee, milk fat and sweets, infrequent tooth brushing, and either excessive participation in physical exercise or total lack of exercise taken alone or with friends or family members. The association of a `reasonable' amount of exercise with success in education may be two-way. Exercise may increase the energy needed in schoolwork, and, vice versa, having energy makes it possible to engage actively both in schoolwork and exercise. Both exercise participation and educational success reflect activity and motivation to work hard and attain goals. The adolescents who take part in organized exercise daily may be the competing ones who do not direct their time and energy into educational pursuits. Very frequent exercise might be related to psychosocial problems if total devotion to one activity leads to isolation from normal everyday social relationships. On the other hand, lack of exercise has been found to be associated with psychosocial problems, such as loneliness, shyness, hopelessness and overall difficulty in being included in groups (Page and Tucker, 1994). These factors are likely to play a role in the development of an unsuccessful educational career.
The role of tooth brushing as a powerful indicator of educational track is in accordance with a previous finding that dental hygiene in mid-adolescence predicts educational level in early adulthood (Koivusilta et al., 1998). As physical appearance is among the most valued characteristics of adolescents (Prokhorov and Perry, 1993), striving for increasing personal attractiveness is a strong motive behind many health-related activities. Young people who do not take care of their dental hygiene have given up one possibility to obtain social acceptance and peer prestige (Evans et al., 1995), and this may be an indication of other problems in their lives. Tooth brushing probably also reflects the families' ability to take care of their children. It is known that in families where children at an early age are taught to brush their teeth regularly, other health-enhancing behaviors are also practiced (Paunio, 1993). Thus, this finding may indicate the importance of a safe and stable family background, which both emphasizes health promotion and provides resources for educational success.
Coffee drinking has been considered a marker of a lifestyle promoting atherosclerosis, although not a risk factor (Schwartz et al., 1994). Adults who drink more coffee have more saturated fatty acids and cholesterol in their diet, and are more often smokers than are those who drink less coffee (Aro et al., 1989; Puccio et al., 1990). Thus, some young people seem to have a high probability of both ending up in a low social class and suffering from diet-related diseases, such as cardiovascular disease, certain forms of cancer and dental caries. Excessive use of sweets also increases the risk of dental caries among these individuals. At age 12, the predictive value of the milk fat variable was due to its connection with family background. This is natural, since the diets of children and younger adolescents are still largely influenced by their parents' socioeconomic position (James, 1997).
At age 14, the picture sharpened to show a smaller number of essential indicators of educational track, i.e. smoking, coffee drinking and tooth brushing. In a Finnish longitudinal study, these factors, as measured at ages 16 and 18, were among the strongest predictors of a low educational level at the age of 24–30 years (Koivusilta et al., 1998). Thus, already at age 14, the well-known association of smoking with poor educational attainment and disassociation with the school world is obvious (Donovan et al., 1991; Glendinning et al., 1995). By this age, previous successes and failures in school have moulded a person's opinions regarding his or her talents and capacities to benefit from education. Psychosomatic symptoms predicted educational track at age 12, when smoking had not yet become a widespread habit. This points to the interpretation that smoking may be an attempt to relieve stress among adolescents who do not feel able to respond to expectations from family, school and peers (Hurrelmann and Maggs, 1995). This could apply to other health-endangering behaviors as well, in that the immediate pleasure offered by psychoactive substances is used to escape from problems, anxiety and bad feelings (Wills, 1986).
Dependency on health-damaging ways of stress relief is emphasized in situations where families experience problems which make parents unable to encourage and support their children. Among these may be economic problems (Mayhew and Lempers, 1998) or difficulties in family functioning (Sweeting and West, 1995). Smoking may be a ticket of admission into a group of understanding other persons who share the feeling of not fitting into the school atmosphere or being incapable of fulfilling the expectations directed at them. When people with like behaviors join together, the implication of these behaviors may be further strengthened. Thus, some ways of acting may easily become permanent habits (Fergusson et al., 1995). The permanence of the smoking habit is further strengthened by the fact that pharmacological dependence on nicotine develops alongside social and psychological motives at an early stage (Jarvis, 1994).
The fact that of the two closely inter-related variables, smoking, and not alcohol use, was selected into the logistic model, may signify that smoking reflects the essential features of the school-relevant lifestyle better than alcohol use. It is possible that, while smoking may indicate contranormative attitudes and rebellion, drinking is confined to recreational contexts and relaxation. Getting acquainted with alcohol may also be a part of maturation in such a way that those who have early experiences with alcohol reach maturity in their attitudes towards education earlier than their slowly maturing peers. However, the significance of various behaviors is not constant over time (Karisto et al., 1993). Different connotations may be related to drinking among adolescents in the year 2000 than previously.
Among sociodemographic variables, gender and father's education were the most important independent predictors of educational track. Thus, this study also confirms the finding that children need a high social class background to get the most benefit from the educational opportunities offered in a society (Mehan, 1992). The fact that father's education overtook his occupation as the strongest background factor may indicate that other factors outrun the importance of economic resources in explaining the association between social origin and education. Less well-educated parents may lack the faith in schooling as a provider of a good future and are content with their son or daughter getting a vocational education, while highly educated parents often want their children to prepare for university studies (Erikson and Jonsson, 1996). On a practical and everyday level, parents differ in their abilities to help their children with their schoolwork. The impact of family background on achievement is emphasized at certain critical moments when decisions about future studies are to be made (McNeal, 1999). A young person who lacks educational models may learn early to appreciate other types of culture than that represented by school (Willis, 1977).
The better academic attainment of girls is shown also in other studies in Nordic countries (Rimpelä et al., 1990; Emanuelsson and Svensson, 1985). One explanation given for this phenomenon is that girls are often more disciplined and hard-working than boys (Svensson, 1971). Moreover, they are more concerned with the status, rather than the economic value, of education than boys (Shavit and Blossfeld, 1996). Regional factors were, in this study, associated with educational track only in univariate analysis. Their effects may be hidden behind educational structures, since the share of less-well-educated fathers is higher in rural regions.
Family type was independently associated with educational track only in 12 year olds. Many studies have shown the significance of family type for educational track. Not living with both parents increases the risk of conduct disorders, health-damaging behaviors and problems in school (Dawson, 1991; Mulkey et al., 1992; Norton et al., 1998). In families where the interaction between children and both of their parents has been disturbed, children lack the social capital needed to achieve success at school (Teachman et al., 1996). At age 12, regional factors also independently predicted educational track. This is consistent with the fact that, in Finland, regional differences in education have prevailed up to present. Rural people have generally a lower educational level than people living in towns (Havén, 1998). The finding that neither family type, nor the regional factors, or father's occupation had independent predictive value at age 14 may mean that, at this age, some strong behaviors mediated the influence of these factors on educational track.
This study gave an opportunity to dig deep into the roots of educational career. There was evidence prior to the study that the health-related lifestyle at age 16 is a strong predictor of the educational level attained in early adulthood (Koivusilta et al., 1998). We now found out that health behaviors operate as selection factors into educational tracks at quite early stages of life. Since educational track and school achievement strongly predict an individual's social position in adulthood (Haven, 1998), lacking the strength or readiness to fully invest in education leads to many kinds of difficulties in the future. Although social background creates a starting point for the development of a child's behaviors and educational career, and also structures his or her later life (Wadsworth, 1997), many behaviors also have significance independently of background. By looking at some strong behaviors, information may be obtained about young peoples' strengths and resources to control their own life. Engaging in certain health-damaging behaviors most probably illustrates feelings or doubts that life is not sufficiently under control to justify taking care of one's health, making plans for the future, and having the strength and support to realize these plans.
Altogether, the interplay of lifestyle and education forms a selection process which allocates individuals into educational tracks which later lead to different positions in relation to health and social status in adulthood (Figure 1). During this process, on the one hand, health-enhancing behaviors increase educational resources while, on the other hand, good performance at school strengthens the willingness to maintain and promote one's health. This process has a bearing on health differences among social classes. Further studies are needed to elucidate the early antecedents of the difficulties which at later ages will be reflected in health-endangering behaviors. Among these `fundamental causes' (Link and Phelan, 1996) may be economic circumstances in early life (Lundberg, 1993), childhood illnesses or developmental defects (Power et al., 1986; Hollo, 1999), some personality traits (Turner et al., 1998; Lawrence and Bennett, 1992), intelligence, proneness for learning, characteristics of family life (Sweeting and West, 1995) and some school-related factors. They may also be hidden in the structures of societies, such as the consequences of income distributions (Wilkinson, 1997). It must, however, be remembered that the whole personal developmental process has a bearing on the formation of behavioral and educational careers of individuals (Pulkkinen et al., 1999). Altogether, this study strengthens the evidence gained from many previous studies that the origin of socioeconomic health differences is in an early age, and that education and behaviors influencing health are closely intertwined.
|
Notes |
|---|
The final version of this article was edited while A. V. worked at the Max Planck Institute for Demographic Research at Rostock, Germany
|
Acknowledgments |
|---|
The authors thank Hans Helenius for statistical advice. This study was supported by the Academy of Finland. This study was supported by the Ministry of Social Affairs and Health, Finland.
|
References |
|---|
|
TopAbstractIntroductionData and methodsResultsDiscussionReferences |
|---|
Agresti, A. (1990) Categorical Data Analysis. John Wiley, New York.
Aloise-Young, P. A., Hennigan, K. M. and Graham, J. W. (1996) Role of the self-image and smoker stereotype in smoking onset during early adolescence: a longitudinal study. Health Psychology, 15, 494–497.[ISI][Medline]
Aro, A., Pietinen, P., Uusitalo, U. and Tuomilehto, J. (1989) Coffee and tea consumption, dietary fat intake and serum cholesterol concentration of Finnish men and women. Journal of Internal Medicine, 226, 127–132.[ISI][Medline]
Barker, D. J. P. (1991) The foetal and infant origins of inequalities in health in Britain. Journal of Public Health Medicine, 13, 64–68.
Bosma, H., van de Mheen, H. D. and Mackenbach, J H. (1999) Social class in childhood and general health in adulthood: questionnaire study of contribution of psychological attributes. British Medical Journal, 318, 18–22.
Bourdieu, P. (1984) Distinction. A Social Critique of the Judgement of Taste. Harvard University Press, Cambridge, MA.
Bourdieu, P. and Passeron, J. C. (1977) Reproduction in Education, Society and Culture. Sage, London.
Brener, N. D., Collins, J. L., Kann, L., Warren, C. W. and Williams, B. I. (1995) Reliability of the Youth Risk Behavior Survey Questionnaire. American Journal of Epidemiology, 141, 575–580.
Caspi, A., Moffitt, T. E., Wright, B. R. E. and Silva, P. A. (1998) Early failure in the labor market: childhood and adolescent predictors of unemployment in the transition to adulthood. American Sociological Review, 63, 424–451.
Dawson, D. A. (1991) Family Structure and children's health and well-being: data from the 1988 National Health Interview Survey on Child Health. Journal of Marriage and the Family, 53, 573–584.
Dixon, W. J., Brown, M. B., Engelman, L. and Jennrich, R. I. (eds) (1990) BMDP Statistical Software Manual: Vols 1 and 2. University of California Press, Berkeley, CA.
Donovan, J. E., Jessor, R. and Costa, F. M. (1991) Adolescent health behavior and conventionality–unconventionality: an extension of problem-behavior theory. Health Psychology, 10, 52–61.[ISI][Medline]
Emanuelsson, I. and Svensson, A. (1985) Does the Level of Intelligence Decrease? A Comparison between Thirteen-Year-Olds Tested in 1961, 1966 and 1980. Report 1985:02. Department of Education and Educational Research, University of Göteborg, Göteborg.
Erikson, R. and Jonsson, J. O. (1996) Introduction. Explaining class inequality in education: the Swedish test case. In Erikson, R. and Jonsson, J. O. (eds), Can Education be Equalized? The Swedish Case in Comparative Perspective. Westview Press, Oxford.
Evans, N., Gilpin, E., Farkas, A. J., Shenassa, E. and Pierce, J. P. (1995) Adolescents' perceptions of their peers' health norms. American Journal of Public Health, 85, 1064–1069.[ISI]
Fergusson, D. M., Lynskey, M. T. and Horwood, L. J. (1995) The role of peer affiliations, social and individual factors in continuities in cigarette smoking between childhood and adolescence. Addiction, 90, 647–659.[ISI][Medline]
Friedman, H. L. (1989) The health of adolescents: beliefs and behaviour. Social Science and Medicine, 29, 309–315.
Glendinning, A., Shucksmith, J. and Hendry, L. (1994) Social class and adolescent smoking behaviour. Social Science and Medicine, 38, 1449–1460.
Glendinning, A., Hendry, L. and Schucksmith, J. (1995) Lifestyle, health and social class in adolescence. Social Science and Medicine, 41, 235–248.
Halsey, A. H., Heath, A. F. and Ridge, J. M. (1980) Origins and Destinations. Family, Class and Education in Modern Britain. Clarendon Press, Oxford.
Haven, H. (ed.) (1998) Koulutus Suomessa. Statistics Finland. SVT 1998:1. Paino-Center, Helsinki.
Hebert, J. R., Clemow, L., Pbert, L., Ockene, I. S. and Ockene, J. K. (1995) Social desirability bias in dietary self-report may compromise the validity of dietary intake measures. International Journal of Epidemiology, 24, 389–398.
Hollo, O. (1999) Alipainoisena syntynyt lapsi kymmenen vuoden iässä. Annales Universitatis Turkuensis. Ser. C. Tom. 149. Scripta lingua fennica edita. Painosalama, Turku.
Hurrelmann, K. and Maggs, J. L. (1995) Health impairments in adolescence: the biopsychosocial `costs' of the modern life-style. In Neubauer, G. and Hurrelmann, K. (eds), Individualization in Childhood and Adolescence. de Gruyter, Berlin.
Jarvis, M. J. (1994) A profile of tobacco smoking. Addiction, 89, 1371–1376.[ISI][Medline]
James, W. P., Nelson, M., Ralph, A. and Leather, S. (1997) The contribution of nutrition to inequalities in health. British Medical Journal, 314, 1545–1549.
Karisto, A., Prättälä, R. and Berg, M. A. (1993) The good, the bad, and the ugly. Differences and changes in health related lifestyles. In Kjærnes, U., Holm, L. Ekström, M., Fürst, E. L. and Prättälä, R. (eds), Regulating Markets, Regulating People. On Food and Nutrition Policy. Novus Forlag, Oslo.
Koivusilta, L., Rimpelä, A. and Rimpelä, M. (1999) Health-related lifestyle in adolescence—origin of class differences in health? Health Education Research, 14, 339–355.
Koivusilta, L., Rimpelä, A. and Rimpelä, M. (1998) Health-related lifestyle in adolescence predicts adult educational level—a longitudinal study from Finland. Journal of Epidemiology and Community Health, 52, 794–801.[Abstract]
Lawrence, B. and Bennett, S. (1992) Shyness and education: the relationship between shyness, social class and personality variables in adolescents. British Journal of Educational Psychology, 62, 193–202.
Link, B. G. and Phelan, J. C. (1996) Editorial: understanding sociodemographic differences in health—the role of funda- mental social causes. American Journal of Public Health, 86, 471–472.
Lintonen, T. (2000) The validity of drunkenness as a measure of adolescent problem drinking. Poster presented at the Conference. `Outcomes in Health Promotion', Tampere, Finland.
Lundberg, O. (1993) The impact of childhood living conditions on illness and mortality in adulthood. Social Science and Medicine, 36, 1047–1052.
Macintyre, S. (1986) The patterning of health by social position in contemporary Britain: directions for sociological research. Social Science and Medicine, 23, 393–415.
Mackenbach, J. P., Kunst, A. E., Cavelaars, A. E. J. M., Groenhof, F, Geurts, J. J. M. and EU Working Group on Socioeconomic Inequalities in Health (1997) Socioeconomic inequalities in morbidity and mortality in Western Europe. Lancet, 349, 1655–1659.[ISI][Medline]
Mayhew, K. P. and Lempers, J. D. (1998) The relation among financial strain, parenting, parent self-esteem, and adolescent self-esteem. Journal of Early Adolescence, 18, 145–173.[Abstract]
McNeal, R. B., Jr (1999) Parental involvement as social capital: differential effectiveness on science achievement, truancy and dropping out. Social Forces, 78, 117–144.
Mehan, H. (1992) Understanding inequality in schools: the contribution of interpretive studies. Sociology of Education 65, 1–20.[ISI]
Mulkey, L. M., Crain, R. L. and Harrington A. J. C. (1992) One-parent households and achievement: economic and behavioural explanations of a small effect. Sociology of Education, 65, 48–65.
Norton, E. C., Lindrooth, R. C. and Ennett, S. T. (1998) Controlling for the endogeneity of peer substance use on adolescent alcohol and tobacco use. Health Economics, 7, 439–453.[ISI][Medline]
Nutbeam D., Aarø L. and Catford J. (1989) Understanding childrens' health behaviour: the implications for health promotion for young people. Social Science and Medicine, 29, 317–325.
Page, R. M. and Tucker, L. A. (1994) Psychosocial discomfort and exercise frequency: an epidemiological study of adolescents. Adolescence, 29, 183–191.[ISI][Medline]
Paunio, P. (1993) Dental Health Habits in Early Childhood. Annales Universitatis Turkuensis. Ser D. Tom 127. Medica-Odontologica. Painosalama, Turku.
Pietilä, A. M., Rantakallio, P. and Läärä, E. (1995) background factors predicting non-response in a health survey of Northern Finnish young men. Scandinavian Journal of Social Medicine, 23, 129–136.[ISI][Medline]
Power, C. and Matthews, S. (1997) Origins of health inequalities in a national population sample. Lancet, 350, 1584–1589.[ISI][Medline]
Power, C., Fogelman, K. and Fox, A. J. (1986) Health and social mobility during the early years of life. Quarterly Journal of Social Affairs, 2, 397–413.
Prokhorov, A. V., Perry, C. L., Kelder, S. H. and Klepp, K. I. (1993) Lifestyle values of adolescents: Results from Minnesota Heart Health Youth Program. Adolescence, 28, 637–647.[ISI][Medline]
Puccio, E. M., McPhillips, J. B., Barrett-Connor, E. and Ganiants, T. G. (1990) Clustering of atherogenic behaviors in coffee drinkers. American Journal of Public Health, 80, 1310–1313.
Pulkkinen, L. Ohranen, M. and Tolvanen, A. (1999) Personality antecedents of career orientation and stability among women compared to men. Journal of Vocational Behavior, 54, 37–58.
Rimpelä, A., Karvonen, S., Rimpelä, M. and Siivola, M. (1990) Socio-economic and Regional Differences in Health Habits of Young People in Finland in 1977–87—The Attainment of the Health for All by the Year 2000 Target `Equal Distribution of Health' (English summary). Publications of the National Board of Health, Finland. Health Promotion. Series Original Reports 1/1990. Valtion Painatuskeskus, Helsinki.
Rimpelä, M., Rimpelä, A., Rahkonen, O. and Teperi, J. (1988) The evolution of the Juvenile Health Habit Study 1977–1985. In Anderson, R., Davies, J. K., Kickbush, I. and McQueen, D. V. (eds), Health Behaviour Research and Health Promotion. Oxford University Press, Oxford.
Rimpelä, M., Rimpelä, A., Vikat, A., Ahlström, S., Huhtala, H. and Lintonen, T. (1997) Nuorten terveystapatutkimus: Tupakointi ja päihteet 1977–1997. Stakes Aiheita 28/1997. Stakesin monistamo, Helsinki.
Rutter, D. R. and Quine, L. (1994) Demographic inputs and health outcomes: psychological processes as mediators. In Rutter, D. R. and Quine, L. (eds), Social Psychology and Health: European Perspectives. Aldershot, Avebury.
Rönkä, A. (1999) The Accumulation of Problems of Social Functioning. Jyväskylä Studies in Education, Psychology and Social Research 148. University of Jyväskylä. Jyväskylä University Printing House, Jyväskylä, and ER-Paino Ky, Lievestuore.
Rönkä, A. and Pulkkinen, L. (1995) Accumulation of problems in social functioning in young adulthood: a developmental approach. Journal of Personality and Social Psychology, 69, 381–391.[ISI][Medline]
Schwartz, B., Bischof, H. P. and Kunze, M. (1994) Coffee, tea, and lifestyle. Preventive Medicine, 23, 377–384.[ISI][Medline]
Shavit, Y. and Blossfeld, H. P. (1996) Equalizing educational opportunity: do gender and class compete? In: Erikson, R. and Jonsson, J. O. (eds), Can Education Be Equalized? The Swedish Case in Comparative Perspective. Westview Press, Oxford.
Statistics Finland (1987) Classification of Occupations. Handbooks 14. Valtion Painatuskeskus, Helsinki.
Statistics Finland (1989) Statistical Yearbook of Finland. Official Statistics of Finland. Valtion painatuskeskus, Helsinki.
Statistics Finland (1995) Statistical Yearbook of Finland. Official Statistics of Finland. Valtion painatuskeskus, Helsinki.
Statistics Finland (1998) Statistical Yearbook of Finland. Official Statistics of Finland. Valtion painatuskeskus, Helsinki.
Svensson, A. (1971) Relative Achievement. School Performance in Relation to Intelligence, Sex and Home Environment. Almqvist & Wiksell, Stockholm.
Sweeting, H. and West, P. (1995) Family culture and health in adolescence: a role for culture in the health inequalities debate? Social Science and Medicine, 40, 163–175.
Taylor, S. and Repetti, R. L. (1997) Health psychology: what is an unhealthy environment and how does it get under the skin? Annual Review of Psychology, 48, 411–447.[ISI][Medline]
Teachman, J. D., Paasch, K. and Carver, K. (1996) Social capital and dropping out of school early. Journal of Marriage and the Family, 58, 773–784.
Turner, L. A., Pickering, S. and Johnson, R. B. (1998) The relationship of attributional beliefs to self-esteem. Adolescence, 33, 477–484.[ISI][Medline]
Wadsworth, M. E. J. (1997) Health inequalities in the life course perspective. Social Science and Medicine, 44, 859–869.
West, P. (1991) Rethinking the health selection explanation for health inequalities. Social Science and Medicine, 32, 373–384.
Wilkinson, R. G. (1997) Heath inequalities: relative or absolute material standards? British Medical Journal, 314, 591–595.
Willis, P. (1977) Learning to Labor. Columbia University Press, New York.
Wills, T. H. (1986) Stress and coping in early adolescence: relationships to substance use in urban school samples. Health Psychology, 5, 503–529.[ISI][Medline]
Wolfson, A. R. and Carskadon, M. A. (1998) Sleep schedules and daytime functioning in adolescents. Child Development, 69, 875–887.[ISI][Medline]
Received on February 23, 2000; accepted on June 1, 2000
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  C. Friestad and K.-I. Klepp Socioeconomic status and health behaviour patterns through adolescence: Results from a prospective cohort study in Norway Eur J Public Health, February 1, 2006; 16(1): 41 - 47. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Koivusilta, S. Honkala, E. Honkala, and A. Rimpela Toothbrushing as Part of the Adolescent Lifestyle Predicts Education Level J. Dent. Res., May 1, 2003; 82(5): 361 - 366. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||