Introduction ... Sociology of Health & Illness à 2010 Foundation for the Sociology of Health & Illness/Blackwell Publishing Ltd. Published by ... divisions, at least in societies that experience post-scarcity (Bourdieu 1984, Giddens 2006). Indeed ...
Sociology of Health & Illness Vol. 33 No. 3 2011 ISSN 0141–9889, pp. 372–383 doi: 10.1111/j.1467-9566.2010.01285.x
Subjective social status and health in young people Sakari Karvonen1 and Ossi Rahkonen2 1
Department of Social and Health Policy and Economics, National Institute for Health and Welfare, Helsinki, Finland 2 Department of Public Health, University of Helsinki, Finland
Abstract
Health inequalities according to objective socioeconomic position (SEP), have been well-documented. Yet, in young people the associations are negligible. Recently, research on the association of subjective social status (SSS), and adult health has begun to accumulate. Studies on young people are rare and describe societies with large income inequalities. Here, we investigated the association between SSS and health, while controlling for own and familial SEP. The study population consisted of 15-year-olds (N = 2369) who have grown up in a context of low social inequalities. Data were derived from surveys carried out in 2004 in 29 secondary schools in Helsinki. The SSS was measured with an indicator specific to and validated for adolescents (a societal ladder). Outcome measures were self-rated health, health complaints, presence of limiting longstanding illness (LLI) and GHQ-12 caseness (indicating psychiatric morbidity). The SSS associated strongly with all health measures. Adjusting for objective socioeconomic measures attenuated the associations; although they all remained statistically significant apart from LLI among girls. The subjective assessment contributes to health inequalities in young people largely independent of objective SEP. Subjective ratings most probably capture aspects of social hierarchy that are more subtle and less well represented than in conventional measures.
Keywords: subjective social status, young people, health, socioeconomic status Introduction Health inequalities among young people have been less well-established than in other age groups (Rahkonen et al. 1997, West and Sweeting 2004). The relative lack of socioeconomic inequalities in youth has been attributed to two factors. West (1997) suggests that in adolescence, social processes (such as lifestyles) and institutions (such as compulsory education) effectively mix young people from different backgrounds. This, in turn, can be expected to lead to low inequalities in health. We have previously pointed out (Rahkonen and Karvonen 2004) that the lack of inequalities might be an artefact due to poor measures of social position in young people. Conventional measures of socioeconomic position are either based on parents’ material characteristics or describe young people themselves, such as by school attainment. However, the former measures may not adequately reflect social differences emergent in adolescence, while the latter ones include measures of positions that have not yet fully materialised, such as level of education. 2010 The Authors. Sociology of Health & Illness 2010 Foundation for the Sociology of Health & Illness/Blackwell Publishing Ltd. Published by Blackwell Publishing Ltd., 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA
Subjective social status and health in young people
373
This debate focuses mostly on so-called objective measures of socioeconomic position (SEP). Objective measures refer to indicators that can be recorded regardless of an individual’s own perception. Such indicators have included measures such as education, occupation, income and wealth. In addition to such objective measures, some scholars would even argue that the element of social comparison is essential to the embodiment of social divisions, at least in societies that experience post-scarcity (Bourdieu 1984, Giddens 2006). Indeed, in these types of societies the importance of social comparison as a determinant of health inequalities has lately raised much interest, and the hypothesis of relative income inequalities has generated much health research (Wilkinson 1996, Wilkinson and Pickett 2006). Studies exploring the association between health and subjective social status (SSS) have started to accumulate, but most of these focus on adult populations (Adler et al. 2000, Macleod et al. 2005, Singh-Manoux et al. 2005, Franzini and Fernandez-Esquer 2006, Demakakos et al. 2008). Goodman et al. (2001) showed in one of the early studies among young people, that low SSS was associated with depressive symptoms in the US. Further, Piko and Fitzpatrick (2001) have reported a relationship between poor psychosocial health and low SSS in a sample of Hungarian adolescents. A recent longitudinal study showed that both lower baseline SSS and a decrease in SSS predicted subsequent poor self-rated health over time even after adjusting for baseline health status and objective measures of parental socioeconomic status (Goodman et al. 2007). These studies originate from societies that are characterised by relatively large social inequalities. For example according to the OECD in 2000 the gini coefficient for measuring income inequalities was 35.7 for the US (OECD 2005). Traditionally, the Nordic countries have witnessed small income inequalities due to an effective, universal social policy based on a relatively high tax rate and comprehensive income transfers (Andersen et al. 2007). In the OECD statistics all four Nordic countries (excluding Iceland) rank among the five countries with the lowest inequalities; in Finland the gini coefficient was as low as 26.1 in the year 2000 (OECD 2005). Since young people may be better able to sensitively place themselves and their families in the social hierarchy in societies with large, and thus in many ways more visible, inequalities, the question arises as to whether SSS is related to health in societies in which there are fewer inequalities or where they are less obvious. The aims of this article were therefore: (1) to analyse whether SSS is associated with health status among 15-year-olds in Finland; and (2) to examine whether the association can be explained via objective measures of social position.
Material and methods The Helsinki Health Promotion Survey data were collected as part of a wider school-based study programme, the Finnish School Health Promotion Survey (Roos et al. 2001, Ellonen et al. 2008). The larger study covers approximately 80 per cent of Finnish 8th and 9th graders (14- and 16-year-old pupils). The questionnaires were distributed to pupils during two school lessons that were supervised by a teacher, who ensured that the students worked on the survey undisturbed, but who was not allowed to interfere with the students’ answers. The anonymous questionnaires were returned in closed envelopes at the end of the lesson. The data collection was carried out in 2004 from a 75 per cent random sample of 35 mainstream schools, representing young people in their final year of compulsory education. Three schools declined to take part, which resulted in a response rate of 80 per cent (N = 2369, 51% females, 29 schools) (Karvonen et al. 2001, Turtiainen et al. 2007). 2010 The Authors Sociology of Health & Illness 2010 Foundation for the Sociology of Health & Illness/Blackwell Publishing Ltd
374
Sakari Karvonen and Ossi Rahkonen
Measures of socioeconomic position Parental educational level was assessed so that the mother’s level of education was applied in cases where the father’s data were missing. The question was: ‘What is the highest educational degree that your mother ⁄ father has achieved?’ There are four options that reflect meaningful educational categories in Finnish society and are roughly equivalent to years of education (9, 11, 12 and 15, see Table 1). Approximately one in ten of the respondents came from families with only the basic level of education, whereas about 40 per cent (43% of boys and 37% of girls) came from higher educated families, which may be due to a bias in the data (Table 1). The distribution of education was in line with the 25–64 year old Helsinki population (City of Helsinki 2008). The actual rate of employment (75% among those aged 15–64), also corresponds to the figures recorded by young people; four fifths of the young people reported both their parents to be employed. Parental employment status was investigated with the question: ‘Have your parents been unemployed or laid-off during the past year?’ The options were: ‘Neither’, ‘One of my parents’ and ‘Both parents’. School performance and pocket money were used to measure own objective social position. School performance was recorded by asking the respondents to describe the average of their marks on a scale of four (fail) to ten (excellent), which is the standard scale in the Finnish school: ‘What was the average of your school marks (all subjects) in the last report?’ Girls had statistically significantly better marks than boys with 24 per cent of girls, but only 10 per cent of boys, reporting an average of nine or more. Boys, on the other hand, reported having more pocket money than girls; 35 per cent of boys and 27 per cent of girls declared that they had €18 or more at their disposal per week. Respondents were asked: ‘How much money do you have at your disposal on average in a week (pocket money or other income that you may use as you wish)?’ with seven options ranging from ‘less than one euro’ to ‘more than 35 euros’. The youth version of the MacArthur scale of SSS developed by Goodman et al. (2001) was used to measure the subjective assessment of the family’s social status. Following their
Table 1. Subjective and objective social status measures (%) by gender among 15-year-olds (N = 2223)
Parental educational level
Parents’ employment status
School performance
Weekly pocket money
Subjective social status of the family
Basic Vocational Upper secondary University degree Both employed One parent unemployed Both parents unemployed Poor (4–7) Good or fair (7–8.9) Excellent (9–10) Less than 10€ 10–17€ More than 18€ Low (steps 1 to 3) Average (steps 4 to 7) High (steps 8 to 10)
Boys
Girls
12 22 24 43 79 21 2 18 72 10 37 29 35 4 56 40
16 25 22 37 74 23 3 10 66 24 43 30 27 4 64 32
p for gender difference p = 0.02
p = 0.33
p < 0.001
p = 0.001
p = 0.001
2010 The Authors Sociology of Health & Illness 2010 Foundation for the Sociology of Health & Illness/Blackwell Publishing Ltd
Subjective social status and health in young people
375
measure the instruction included also a description of markers of high (e.g. ‘the most money’), and low (e.g. ‘little education’), position. The complete instruction was as follows: ‘Imagine that this ladder pictures how Finnish society is made up. At the top of the ladder are the people who are the best off – they have the most money, the highest amount of schooling and the jobs that bring the most respect. At the bottom of the ladder are the people who are the worst off – they have the least money, little or no education, no job or jobs that no one wants or respects’. Here a 10-rung ladder scale was used, with higher scores indicating higher perceived rank, to assess perceptions of status in the context of the Finnish society: ‘Think about your own family. In which place would your family be? Please tick the rung that best reflects the position of your family in these ladders’. A pilot test was run in two school classes during Spring 2004. The students filled out a twopage questionnaire that included the SSS measure. They were also instructed to give feedback on the questions posed. This was generally very supportive, even though some amendments had to be made regarding questions not reported on in this article. As the original distribution was quite skewed, the 10-point scale was collapsed into three groups composed of the three lowest rungs, the three highest rungs and the four in the middle (Table 1). Boys placed their families somewhat higher than girls, which corresponds with their reporting of parental educational level; 40 per cent of boys and 32 per cent of girls saw their families as situated in the three highest steps of the ladder (Table 1). The poorest positions, on the other hand, were rarely selected. Health measures The health measures used were self-rated health, health complaints, presence of limiting longstanding illness and the general health questionnaire (GHQ-12). Earlier studies have shown them to be relevant in describing aspects of young people’s health. Self-rated health was used as a general measure of subjective health, while health complaints capture psychosomatic symptoms present in everyday life (Bowling 1997). Limiting longstanding illness was selected to measure more objectively diagnosed illnesses, disabilities or injuries (Bowling 1997). The GHQ-12 was used as a measure of psychological distress (Goldberg 1972). Self-rated health was examined by asking the respondents to rate their general health on a four-point scale with the question: ‘What do think about your state of health? Is it ‘‘very good’’, ‘‘good’’, ‘‘average’’ or ‘‘poor or very poor’’?’ Responses were dichotomised into average or poor (henceforth described as ‘poor health’) and better than average. Health complaints included a list of eight psychosomatic symptoms such as ‘headache’, ‘neck and shoulder pain’ or ‘irritation’, and were assessed via a four-point frequency scale ranging from ‘almost every day’ to ‘rarely or never’. Limiting long-standing illness (LLI), was examined by asking the respondent to report whether they had limiting illnesses, disabilities or injuries diagnosed by a physician. The GHQ-12 includes questions on psychological distress symptoms, such as ability to concentrate, feeling depressed and losing sleep over worries. The respondents assessed whether each has ‘lately’ been ‘better than usual’, ‘same as usual’, ‘worse than usual’ or ‘much worse than usual’. The binary method of scoring (0 0 1 1), was used which is based on the presence (coded 1), or absence (coded 0), of symptoms (range 0–12). The cut-off for caseness was four symptoms or more. Methods First, distributions of the SSS and objective measures of socioeconomic position were calculated by gender. Then we explored the association between gender and health 2010 The Authors Sociology of Health & Illness 2010 Foundation for the Sociology of Health & Illness/Blackwell Publishing Ltd
376
Sakari Karvonen and Ossi Rahkonen
measures as well as between parental educational level and health measures. Parental educational level was selected to illustrate associations by the objective measures. As the descriptive results showed gender to be associated strongly with most measures, all consecutive analyses were performed separately for males and females. Next, the interrelationships between the different social status measures (subjective and objective), were explored by means of Pearson’s correlation analysis. As the correlations remained generally low, it was decided to include all four in subsequent stages of the analysis. Next, the patterning of the four health measures according to SSS was assessed by means of the chi-square statistic. In addition to descriptive tabulations and correlational analyses, the associations were examined by logistic regression analyses using the SPSS 17.0 for Windows. These were performed to assess whether the bivariate association between SSS and health was explained (i.e. removed), in multivariate analyses including objective social status. Apart from school performance, the independent variables were included as categorical variables in the models. For the SSS, p-values and 95 per cent confidence intervals are reported.
Results Girls showed poorer health than boys in all measures except limiting longstanding illness (Table 2). Approximately 10 per cent of all young people reported suffering from a limiting longstanding illness. Self-rated health was associated with parental educational level such that young people from families with the highest level of education showed consistently better health and they also reported fewer health complaints, less limiting longstanding illness and a lower frequency of GHQ caseness. Although those with the highest SSS generally had the best health and those with the lowest the worst, the relationships were not neatly graded (Table 2). Correlations between the SSS and objective measures of social position ranged from 0.11 to 0.27 (Table 3). The highest association was between the SSS and parental educational level among girls and lowest between SSS and school performance among boys. Among the objective measures the inter-relationships were generally low with the exception of those
Table 2. Health measures (%) by gender and parental educational level among 15-year-olds (N = 2115)
Poor or average perceived health Liming longstanding illness: Yes Four or more health complaints per week GHQ-caseness: Yes 1 2
Gender
Parental educational level
Boys Girls p1
Basic Upper University level Vocational secondary degree p2
16
22
11
11
17 18
