School dropout: a major public health challenge: a 10-year prospective study on medical and non-medical social insurance benefits in young adulthood, the Young-HUNT 1 Study (Norway)
- Karin A A De Ridder1,2,
- Kristine Pape1,
- Roar Johnsen1,
- Steinar Westin1,
- Turid Lingaas Holmen3,
- Johan Håkon Bjørngaard1,4
- 1Department of Public Health and General Practice, Norwegian University of Science and Technology, Trondheim, Norway
- 2Department of Physical Medicine and Rehabilitation, Levanger Hospital, Nord-Trøndelag Health Trust, Levanger, Norway
- 3HUNT Research Center, Department of Public Health and General Practice, Norwegian University of Science and Technology, Levanger, Norway
- 4Forensic Department and Research Centre Bröset, St. Olav's University Hospital Trondheim, Trondheim, Norway
- Correspondence to Karin A A De Ridder, Department of Public Health and General Practice, Norwegian University of Science and Technology, 7491 Trondheim, Norway;
Contributors All authors are responsible for the reported research and have made substantial contributions to the conception and design of the study, acquisition of data, analysis and interpretation of data, the draft of the article or revisions of important content in the article. All authors have seen and approved the final version of the manuscript.
- Accepted 25 November 2011
- Published Online First 6 February 2012
Background School and work participation in adolescence and young adulthood are important for future health and socioeconomic status. The authors studied the association between self-rated health in adolescents, high school dropout and long-term receipt of medical and non-medical social insurance benefits in young adulthood.
Methods Self-rated health in adolescence was assessed in 8795 adolescents participating in the Norwegian Young-HUNT Study (1995–1997). Linkages to the National Education Database and the National Insurance Administration allowed identification of school dropout and receipt of long-term medical and non-medical benefits during a 10-year follow-up (1998–2007). The data were explored by descriptive statistics and by multinomial logistic regression.
Results A total of 17% was registered as being high school dropouts at age 24. The predicted 5-year risk of receiving benefits between ages 24–28 was 21% (95% CI 20% to 23%). High school dropouts had a 5-year risk of receiving benefits of 44% (95% CI 41 to 48) compared with 16% (95% CI 15 to 17) in those who completed high school (adjusted for self-rated health, parental education and sex). There was a 27% school dropout rate in adolescents who reported poor health compared with 16% in those who reported good health. The predicted 5-year risk of receiving any long-term social insurance benefits in adolescents who reported poor health was 33% (95% CI 30 to 37) compared with 20% (95% CI 19 to 21) in those who reported good health.
Conclusion The strong association between poor self-rated health in adolescence, high school dropout and reduced work integration needs attention and suggests preventive measures on an individual as well as on a societal level.
- Student dropouts
- health status
- longitudinal studies
- public health
- sickness absence
- adolescents CG
School dropout and lack of work integration among the young are a major concern in most Western countries.1–3 While the social gradient in health and disease is well studied,4 ,5 research on the importance of health for school attainment is limited.6 Most research on health and work is limited to adults, with little attention to adolescents and young adults.7
Long-term medical social insurance benefits in early adulthood are indicators of individual vulnerability, with possible health, social and economic consequences throughout the life course. Medical benefits are conditionally associated with ill health. Severe health impairments in childhood and adolescence are important for subsequent work disability.8 However, non-medical social insurance benefits, like unemployment and social insurance, support could also be an indicator of ill health. Chronic disease in childhood and adolescence affects work participation.9 Young people with medical and non-medical social insurance benefits have a weak attachment to the labour force and school and are in danger of marginalisation and social exclusion.10 In addition, social exclusion may negatively influence health.11
Low educational achievement is associated with lower work participation, higher risk of work disability and general health impairment.8–10 Health has been studied as a causal mechanism in the development of social gradient in adult life.12 ,13 However, less attention has been given to its influence on the educational and occupational processes which take place during the transition from adolescence to young adulthood—a period in life where health may have an substantial impact.14 It is important to consider impaired health and educational achievement in a life-course perspective and assess factors of early life experience.15 If factors already present in adolescence have a substantial impact on adult health and life chances, they represent major public health challenges both within and beyond the healthcare systems.16
We studied the association between high school dropout and the long-term reception of medical and non-medical social insurance benefits throughout a period of 10 years using a prospective design. Adolescents' self-rated health was taken into account.
All students attending middle and secondary school (13–19 years old) in the county of Nord-Trøndelag, Norway, were invited to participate in the Young-Hunt 1 Study in 1995–1997. Young-Hunt is the adolescent part of the HUNT Study (The Nord-Trøndelag Health Study, http://www.ntnu.no/hunt). A comprehensive questionnaire was completed by 8950 adolescents (90% response rate) during a class hour. Data from Young-Hunt 1 were linked to information about social insurance benefits from the National Insurance Administration in the period 1998–2007 and educational data from the Norwegian National Education Database (http://www.ssb.no/mikrodata/). Participants were linked to their biological parents through a national Family Register code.
Altogether 155 individuals were excluded from this study. Causes for exclusion were disability pension at age 18 or 19 (66), start of a social insurance process before 1998 that resulted in disability pension (35), death during follow-up (46), born after 1983 (4) and age–school level mismatch (4).
Norwegian social insurance system
The National Insurance Administration keeps a complete record of social insurance benefits, and the data are available in Statistics Norway's database (http://www.ssb.no/mikrodata/). In Norway, people who are employed can receive sickness benefits for health-related work disability. Medical benefits included sickness benefits, medical and vocational rehabilitation and temporary and permanent disability pension. Sickness benefits can be granted for a period up to 52 weeks. At the time the data were registered, the period could be extended up to 3 years with reduced payment when medical or vocational rehabilitation occurred. Young adults have to be at least 18 years old to be eligible for medical rehabilitation compensation. Vocational rehabilitation requires that the recipient is 19 years or older and is registered as a job applicant. Disability pension requires longstanding reduced work ability by at least 50% due to disease, injury or impairment and requires that vocational rehabilitation has been tried.
Unemployment benefits are granted to individuals with a minimum of 50% unemployment and who are actively seeking employment, attending job courses or are enrolled in job training. Because prior work participation is necessary, students do not qualify for this type of benefit. Social insurance support can be given to people who do not qualify for any other benefits and who are unable to financially support themselves.
Long-term social insurance benefits
We constructed three categories: (1) long-term medical benefits (regardless of any other non-medical benefit); (2) only long-term non-medical benefits and (3) no long-term benefits.
Long-term medical social insurance benefits are defined as one of following benefits: (1) medical benefits which per definition were long term (permanent and temporary disability pension, medical and vocational rehabilitation) and (2) sickness benefits received at least 180 days in 1 year. Long-term non-medical social benefits are defined as one of the following benefits: (1) unemployment benefits received at least 180 days in 1 year or (2) social insurance support for >1 month in 1 year.
Whether a participant completed high school or not was registered in the calendar year when he or she turned 24 years old. Those not registered as having completed secondary school at age 24 were defined as ‘high school dropouts’.
The question in Young-Hunt “How is your health at the moment?” had four response alternatives and was dichotomised into two groups: ‘very good/good’ and ‘not very good/poor’.
Parental education level
The biological parents' education level was registered in 1995 and divided into three categories: compulsory education (primary school and lower secondary school), intermediate education (upper secondary school and post-secondary non-tertiary education), and tertiary education (undergraduate, graduate and postgraduate education).
The percentage of those who were receiving long-term medical and non-medical benefits was assessed according to age in the follow-up period (n=8795). This percentage was identified separately for those who had and those who had not completed high school at age 24. The follow-up period lasted for an average of 8.2 years (range 6–10 years) in the age interval from 19 to 28 years old.
The percentage of the cohort who received long-term medical and non-medical benefits was presented in two successive 4-year periods in order to describe the development over time (n=8795). In the period 19–23 years of age, the ‘early period’, the mean follow-up time was 4.6 years (range 2–5 years), and in the period 24–28 years of age, the ‘late period’, the mean follow-up time was 3.6 years (range 1–5 years).
The estimated percentage of the cohort who received long-term medical and non-medical benefits between ages 24–28 was assessed according to poor self-rated health and high school completion (n=8641).
Complete case data (n=8339) were analysed using multinomial logistic regression since the outcome variable has three nominal categories. The outcome variable was the receipt of long-term social insurance benefits (medical, non-medical and no benefits) in the period 24–28 years old. Mean follow-up time in the 5-year period between the ages of 24 and 28 was 3.6 years (range 1–5 years). In the unadjusted model, high school completion and follow-up time were included. The adjusted model also included self-rated health, sex and the mother's education level. The models were used to estimate 5-year risks and marginal changes in the risk of medical and non-medical social insurance benefit receipt between the ages of 24 and 28. A marginal change describes how 1 unit change in an independent variable (eg, high school completion or not) alters the risk of a current outcome (long-term social insurance benefits). The risk difference was estimated with the covariates at their mean and the follow-up time at 5 years. Analyses were conducted with STATA V.11.1 (StataCorp LP).
Each participant and also the parents/legal guardians of the participants younger than 16 years old gave their written consent to participate in the Young-Hunt Study.
Description of the participants
The descriptive statistics of the Young-Hunt 1 population is presented in table 1. At the age of 24, 17% of the participants had not completed high school.
School dropout and social insurance benefits
At age 26, 585 participants of 6451 (9%) received medical and non-medical social insurance benefits and more than half of the benefits were medical benefits (n=373, 6%). Among high school dropouts, 23% received long-term benefits at the age of 26, 15% received medical benefits and 8% received non-medical benefits. Among the participants who had completed high school (referred to as high school completers), 7% received long-term benefits at age 26 and 4% of which were medical benefits. The large difference in benefit receipt between high school completers and high school dropouts was observed in the entire follow-up period (from age 19 to 28) (figure 1A and B).
The predicted 5-year risk of receiving any social insurance benefits between the ages of 24 and 28 was 21.4% (95% CI 20.2% to 22.7%). Specified by type of benefit, the predicted 5-year risk for medical benefits was 11.4% (95% CI 10.4% to 12.4%) and for non-medical benefits, it was 10.0% (95% CI 9.1% to 11.0%). High school dropouts had an unadjusted risk difference in 5-year risk of 15% points (95% CI 13% to 18%) for medical benefits and 15% points (95% CI 12% to 18%) for non-medical benefits compared with high school completers. Adjusted for self-rated health, sex, follow-up time and education level of the mother, the predicted risk difference in 5-year risk for medical and non-medical benefits was 15% points (95% CI 12% to 18%) and 13% points (95% CI 10% to 16%), respectively (table 2). This resulted in an adjusted 5-year predicted risk to receive any social insurance benefit of 44.2% (95% CI 41.0% to 47.7%) for high school dropouts and 15.7% (95% CI 14.7% to 17.0%) for high school completers. Men received statistically significant fewer benefits in the adjusted model, especially medical benefits (p<0.001), but there was no evidence for statistical interaction between sex and school dropout (p>0.1). Analyses adjusting for the highest education level of either of the parents gave the same results.
Social insurance benefits in relation to age
During the follow-up period, the distribution of medical and non-medical social insurance benefits changed. In the late period (age 24–28), the percentage of the participants who received medical benefits was larger than in the early period (age 19–23) (9.5% vs 6%), while the percentage of those who received non-medical benefits had decreased. Most of those who had received medical benefits in the early period also did so in the late period (66%, 344) (figure 2). On the other hand, most participants who received non-medical benefits in the early period did not receive any benefits in the late period (62%, 474). The majority of those who had not received long-term benefits in the early period did not receive any benefits in the late period.
Adolescents who reported poor health had a higher school dropout rate (26.5%) compared with adolescents who reported good health (16.1%). Additional regression analyses revealed that high school dropout and self-rated health were strongly associated (p<0.001), also when adjusted for sex and maternal education. More adolescents with poor health received social insurance benefits between the ages of 24 and 28 (33.2%, 95% CI 29.8% to 36.6%) compared with adolescents with good health (20%, 95% CI 18.7% to 21.2%). The risk increase between good and poor health was substantial and of the same size for both medical and non-medical benefits, independently of high school completion (p<0.001) (table 2). There was no evidence of statistical interaction between high school dropout and self-reported health (p=0.8). Figure 3 shows the predicted 5-year risk of receiving medical and non-medical social insurance benefits according to self-rated health and high school completion.
We observed a strong association between poor self-rated health, high school dropout and the risk of receiving medical and non-medical benefits in young adulthood. Those receiving benefits, particularly medical benefits, between 19 and 23 years of age had a substantially higher risk of being a benefits recipient at 24–28 years of age.
This is a large prospective study with a high response rate, low number of missing cases and complete follow-up information from the registers. When defining long-term benefits based on large registers, choices and limitations must be made. The 6-month limit we set for long-term benefits might be somewhat high, but at least the proportions of long-term benefits are not overestimated. This study shows strong associations, but we should be cautious to draw conclusions related to causation. We do not know whether high school dropout itself is the cause of long-term medical and non-medical benefits in young adulthood or whether high school dropout and long-term social insurance benefits are joint consequences of other individual or contextual factors, like health, family or school (confounding). Yet, our findings may be an indication that ill health increases vulnerability to social exclusion in the transition from adolescence to adulthood being that the health variable was registered while the participants were still in school.
We found a strong association between high school dropout and long-term medical and non-medical benefits, which remained after adjustment for self-rated health, sex, age and education level of the mother. The results showed that self-rated health in adolescence is an additive and independent risk factor for receiving long-term social insurance benefits in young adulthood. Low education level has been described as an important risk factor for disability pension at all ages, also in young adulthood.17 Gravseth et al 8 and Bjerkedal et al 18 studied the impact of both chronic disease in childhood and high school dropout in a young population. They found a higher risk of receiving disability pension in young adulthood in relation to high school dropout than in relation to chronic diseases in childhood. Kristensen and Bjerkedal9 found a higher risk of unemployment in young adulthood in relation to high school dropout and chronic disease in childhood.
Previous studies have mainly focused on disability pension and/or unemployment. We believe that this is inadequate as a measure of social exclusion in young people since the incidence of work disability is marginal in young adults.17 In our cohort, <2% received disability pension during follow-up, and we chose to exclude two-thirds of them because they received disability pension at the start of their adulthood. Taking into account all types of long-term medical benefits, the number of people at risk for marginalisation or social exclusion was considerably larger. For each junior high school class with 26 pupils, four pupils would not complete high school, and each calendar year from age 22, one of the four would receive long-term social insurance benefits. This picture should give rise to concern and is, indeed, a public health challenge.
We observed that medical benefits were received more often than non-medical benefits. This is in accordance with reports from OECD concerning work, sickness and disability, in which it was stated that the use of medical social insurance benefits in Norway is extensive compared with other OECD countries.1 ,2 In the current cohort, it appears that once young people have received long-term medical benefits, their risk of receiving even more medical benefits over time increases. Other studies have described that long-term sickness leave increased the risk of disability pension, though mostly in adult populations.1 ,19–22 The majority of young people in our study who received non-medical benefits in their early twenties returned to work or school, but the percentage still receiving benefits at a later time was larger than among non-receivers. These findings are in line with Raaum et al,10 who described greater risk for social exclusion for those out of work or school for >1 year or in the case of even shorter breaks if participation was the norm. In the present study, we cannot distinguish between the process of marginalisation and social exclusion but we found an increased risk of receiving long-term benefits for high school dropouts. Dropping out of high school could be the first formal registration of an ongoing process of marginalisation, that later develops into social exclusion, already in young adulthood.
Individuals who receive medical benefits have a medical diagnosis per definition. We wanted to know if people who reported poor health might be more vulnerable to exclusion from school or work. We found that adolescents with poor self-rated health more often dropped out of school and more often received social insurance benefits, both medical and non-medical. In adulthood, poor self-rated health is reported to be associated with higher morbidity,23 use of healthcare24 and mortality.25 Self-rated health is relatively stable through adolescence and related to physical health complaints, disability, mental health and health behaviour.26 Chronic disease in childhood has been found to be associated with lower achieved education level and higher risk for unemployment and disability pension, but the studies are based on registered benefits in childhood and this means the research is focused on a decidedly more vulnerable population.8 ,9 ,18 Educational achievement and familial socioeconomic status have been suggested to minimise the effect of adolescent health on educational attainment.27 We found, however, that adolescent health was strongly associated with social status in young adulthood and especially with lower educational level, consistent with the study of Huurre et al.28 Torsheim et al 29 described that a stress moderator, sense of coherence30 and its association with school-related stress and health complaints were not stable across ages in adolescence: the association with school-related stress grew weaker with age, whereas the association with health complaints became stronger. Seen together, these studies and the present findings may shed new light on the role of health on socioeconomic status and possible mechanisms explaining this: What starts as a distress reaction to poor school achievement or social adversity in childhood or adolescence may later lead to ill health, which then becomes an independent risk factor for marginalisation and social exclusion during the transition to adulthood.
High school dropout and problematic work integration are extensive problems that affect many young people and have important individual and social consequences. Research related to the association between health, high school dropout and work integration of young adults is scarce, while publicity on the topic is often dramatic and moralistic and could be a contributor to further stigmatisation of the adolescents who experience such problems. Future research should focus on the influence of health on school dropout and work integration. New and relevant information could promote a more balanced perspective and provide a basis for preventive measures or interventions in an early phase.
What is already known on this subject
School and work participation in adolescence and young adulthood are important for future health and socioeconomic status.
Severe health impairments in childhood and adolescence affect work participation and are important for subsequent work disability.
Low educational achievement is associated with lower work participation, higher risk for work disability and general health impairment.
What this study adds
School dropout and reduced work integration in young adulthood are strongly associated and affect a high proportion of young people.
Poor self-rated health in adolescence is associated with high school dropout and the risk of receiving medical and non-medical benefits in young adulthood.
This association calls for a more balanced perspective on the lack of work integration in young adulthood and should inform preventive measures in an early phase.
The Nord-Trøndelag Health Study (The HUNT Study) is a collaboration between HUNT Research Centre (Faculty of Medicine, Norwegian University of Science and Technology NTNU), Nord-Trøndelag County Council, Central Norway Health Authority and the Norwegian Institute of Public Health.
Funding This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None.
Patient consent The collection of the data was in 1995–1997, and the whole county of Nord-Trøndelag was invited. All participants signed a consent form approved by Regional Committee for Medical Research Ethics and the Norwegian Data Inspectorate.
Ethics approval The Young-Hunt Study was voluntary and approved by the Regional Committee for Medical Research Ethics and the Norwegian Data Inspectorate. The current study was approved by the Regional Committee for Medical Research Ethics (reference 2010/1727-5).
Provenance and peer review Not commissioned; externally peer reviewed.