Abstract
Background: The purpose was to examine the association between aspects of alienation and symptom load among adolescents. Furthermore an integrated purpose was to construct and validate an index of alienation. Methods: Cross-sectional data from 5205 school children aged 11–15 years from a random sample of schools in Denmark were used. Data stems from the Danish contribution to the cross-national study Health and Behaviour in School-aged Children (HBSC). Alienation was measured with a new index fulfilling four criteria: (i) theoretical foundation, (ii) inter-correlation between items, (iii) correlation between each of the index's items and the outcomes and (iv) no differential item functioning. The final index included three indicators of alienation: helplessness, feeling left out of things and lack of confidentiality with parents. Symptom load was measured by HBSC Symptom Checklist and divided into physical and psychological symptoms respectively. High symptom load was defined as experiencing at least one symptom on a daily basis. Results: The odds-ratio (OR) for high symptom load increased with the degree of alienation. For students with all three indicators of alienation, the OR for high physical symptom load was 2.49 (1.05–5.87). The OR for high psychological symptom load for the corresponding degree of alienation was 6.50 (3.11–13.56). Conclusion: The index of alienation fulfilled psychometric criteria for scalability. Furthermore the analyses showed a graded and significant association between alienation and high symptom load. This suggests alienation to be taken into account in future health interventions among adolescents. In school settings this may be done using principles of empowerment.
Introduction
Physical and psychological symptoms are common among adolescents.1–3 A high symptom load is not only unpleasant for the adolescent. Symptoms in childhood seem to track into adulthood as children with recurrent headache and abdominal pain have an increased risk for frequent headache, multiple physical symptoms and psychiatric morbidity in adulthood.4,5 It is therefore important to identify predictors and mechanisms behind a high symptom load among adolescents.
The association between poor well-being and poor health is well-known,1,6,7 but difficult to interpret as well-being is an unspecific concept offering only limited explanation of the processes behind the association. Psychological alienation is a related and appropriate concept as it refers to specific processes in the transaction between the individual and the social environment. These processes may affect health negatively.
Seeman (1959) conceptualized alienation and identified six dimensions on the basis of how they had been used empirically and theoretically.8 There is no theoretical structure between the six dimensions and presence of all six dimensions is not required. These six dimensions focus on the personal experiences: powerlessness, meaninglessness, normlessness, self-estrangement, cultural estrangement and social isolation.8,9 Alienation has been measured in various ways and no golden standard exists, although most scales take their starting point in Seeman's conceptualization.10
Alienation is a well-known predictor of deviant behaviour such as drug use, alcohol use, crime, truancy and suicide.11–16 Only few empirical studies focus on alienation and health. These studies find an association between school alienation and psychosomatic symptoms17 as well as risk behaviour such as smoking, alcohol use, drunkenness, less exercising and eating unhealthy food.12
This study examines the association between alienation towards oneself and those closest to oneself, and the outcome symptom load. An integrated aim was to explore whether items from an internationally standardized questionnaire on adolescent health could be combined into a coherent and valid measurement of alienation.
Methods
Data come from the Danish contribution to Health Behaviour in School-aged Children (HBSC) 1998—a cross-sectional WHO collaborative study. The overall purpose of HBSC is to increase understanding of young people's health and lifestyles in their social context.18
Sample
The study included a random sample of 64 schools of which 55 participated. The sample comprised all students from the fifth (mean age 11.8 years), seventh (13.8 years) and ninth grade (15.8 years), n = 5788. The students completed the internationally standardized HBSC questionnaire about socio-demographic factors, health-related behaviours, physical and mental health, self-concept, wellbeing and social relations in the course of a lesson. The participation rate was 99% of the students present and 87% of the students formally enrolled in the target classes, n = 5205 (2624 boys and 2581 girls).
There is no formal agency for ethical approval of school surveys in Denmark. Therefore we asked the school leader, the board of students and the board of parents in each of the participating schools for assessment and approval of the study. The study was anonymous.
Measurements
The dependent variable symptom load was measured by the HBSC Symptom Check List (HBSC-SCL).19,20 The items were: ‘In the last 6 months: how often have you had the following? Headache, stomach-ache, back ache, feeling low, irritated or bad temper, feeling nervous, difficulties in getting to sleep, feeling dizzy’. (‘About every day’, ‘More than once a week’, ‘About every week’, ‘About every month’, ‘Rarely or never’). We separated symptom load into physical (headache, stomach ache, back pains and dizziness) and psychological symptoms (feeling low, irritated/bad temper, nervous and sleeping difficulties). This dividing is supported by previous studies.19,21 Further, we dichotomized physical and psychological symptoms into experiencing at least one of the four symptoms in question ‘about every day’ (hereafter daily symptoms) vs. less often.
The analyses included four covariates predicting symptom load among school-children: sex,19,21–23 family socio-economic position (SEP),22 bullying1 and loneliness.24 Age group and family structure were not included in the final models as they showed none (age group) or only modest (family structure) confounding effect of the association between alienation and the outcome measures and no significant association in neither of the adjusted models.
SEP was measured by items on father's and mother's occupation and coded according to the standards of the Danish National Institute of Social Research into social class I (highest) to V and group VI (outside labour market). We classified the students by the highest ranking parent.
Bullying was measured by the item: ‘How often have you been bullied in school this term?’ (‘I haven't’ been bullied in school this term, ‘Once or twice’, ‘Sometimes’, ‘Once a week’, ‘Several times a week’). Loneliness was measured by the item: ‘Do you ever feel lonely?’ (‘Yes, very often’, ‘Yes, rather often’, ‘Yes, sometimes’, ‘No’).
Developing and validating an index of alienation
As no alienation index was included in the questionnaire an integrated purpose was to develop and validate an index of alienation towards oneself, and those closest to oneself. Four criteria had to be fulfilled to accept the construct validity of the index. The purpose was to confirm the conceptualization in Seemans alienation concept as a useful structure for measuring alienation. First the included items had to correspond with Seeman's conceptualization. Secondly the items had to be intercorrelated. Thirdly all items had to be correlated with the outcome measures. Finally no differential item functioning was allowed, i.e. none of the items could have an independent effect on the outcome measure besides the effect of the index itself, as this would cause systematic bias in relation to certain subgroups of students in the analyses.
Six items (table 1) were chosen as potential components of the index based on Seeman's conceptualization of alienation8,9: ‘Loneliness’ is one way in which the dimensions social isolation and cultural estrangement manifest themselves.8,9 Likewise ‘feeling left out of things’ represents aspects of these two dimensions. One can feel left out of things by social relations but also because you don't share the values of the surroundings. The items on lack of confident relations with parents and friends also refer to the dimension social isolation. ‘Helplessness’ refers to the dimension powerlessness. Finally ‘lack of educational plans for the future’ refers to the dimension meaninglessness as it may indicate low confidence in education as important for future life chances.
The six potential components of the alienation index
| Variable (corresponding theoretical dimension) | Phrasing of item | Response categories (‘vs.’ shows dichotomization) | Dichotomized |
|---|---|---|---|
| Lonelinessa (social isolation; cultural estrangement) | Do you feel lonely? | ‘Yes, very often’ and ‘Yes, rather often’ vs. ‘Yes, sometimes’ and ‘No’ | Yes |
| Feeling left out of things (social isolation; cultural estrangement) | How often do you feel left out of things? | ‘Always’ and ‘Often’ vs. ‘Sometimes’, ‘Rarely’ and ‘Never’ | Yes |
| Helplessness (powerlessness) | How often do you feel helpless? | ‘Always’ and ‘Often’ vs. ‘Sometimes’, ‘Rarely’ and ‘Never’ | Yes |
| Lack of confident relations with one's parents (social isolation) | How easy is it for you to talk to the following persons about things that really bother you? Mother, Father | ‘Very difficult’, ‘Difficult’ vs. ‘Easy’, ‘Very easy’, ‘Don't have or see this person’ | Yes |
| Lack of confident relations with one's friendsb (social isolation) | How easy is it for you to talk to the following persons about things that really bother you? Friends of the same sex, friends of the opposite sex | ‘Very difficult’, ‘Difficult’, ‘Easy’, ‘Very easy’, and ‘Do not have or see this person | No |
| Lack of educational plans for the futurec (meaninglessness) | What do you think you will be doing when you finish high school? | ‘College or University’, ‘Vocational or technical school’ and ‘Apprenticeship/trade’ vs. ‘Working’ and ‘Unemployed’. ‘Do not know’ was coded as missing | Yes (for use in correlation analyses) |
| Variable (corresponding theoretical dimension) | Phrasing of item | Response categories (‘vs.’ shows dichotomization) | Dichotomized |
|---|---|---|---|
| Lonelinessa (social isolation; cultural estrangement) | Do you feel lonely? | ‘Yes, very often’ and ‘Yes, rather often’ vs. ‘Yes, sometimes’ and ‘No’ | Yes |
| Feeling left out of things (social isolation; cultural estrangement) | How often do you feel left out of things? | ‘Always’ and ‘Often’ vs. ‘Sometimes’, ‘Rarely’ and ‘Never’ | Yes |
| Helplessness (powerlessness) | How often do you feel helpless? | ‘Always’ and ‘Often’ vs. ‘Sometimes’, ‘Rarely’ and ‘Never’ | Yes |
| Lack of confident relations with one's parents (social isolation) | How easy is it for you to talk to the following persons about things that really bother you? Mother, Father | ‘Very difficult’, ‘Difficult’ vs. ‘Easy’, ‘Very easy’, ‘Don't have or see this person’ | Yes |
| Lack of confident relations with one's friendsb (social isolation) | How easy is it for you to talk to the following persons about things that really bother you? Friends of the same sex, friends of the opposite sex | ‘Very difficult’, ‘Difficult’, ‘Easy’, ‘Very easy’, and ‘Do not have or see this person | No |
| Lack of educational plans for the futurec (meaninglessness) | What do you think you will be doing when you finish high school? | ‘College or University’, ‘Vocational or technical school’ and ‘Apprenticeship/trade’ vs. ‘Working’ and ‘Unemployed’. ‘Do not know’ was coded as missing | Yes (for use in correlation analyses) |
a: Excluded from index due to differential item functioning in relation to psychological symptoms.
b: Excluded from index due to insignificant correlation with both outcome measures.
c: Excluded from index due to insignificant correlations with the other five items.
The six potential components of the alienation index
| Variable (corresponding theoretical dimension) | Phrasing of item | Response categories (‘vs.’ shows dichotomization) | Dichotomized |
|---|---|---|---|
| Lonelinessa (social isolation; cultural estrangement) | Do you feel lonely? | ‘Yes, very often’ and ‘Yes, rather often’ vs. ‘Yes, sometimes’ and ‘No’ | Yes |
| Feeling left out of things (social isolation; cultural estrangement) | How often do you feel left out of things? | ‘Always’ and ‘Often’ vs. ‘Sometimes’, ‘Rarely’ and ‘Never’ | Yes |
| Helplessness (powerlessness) | How often do you feel helpless? | ‘Always’ and ‘Often’ vs. ‘Sometimes’, ‘Rarely’ and ‘Never’ | Yes |
| Lack of confident relations with one's parents (social isolation) | How easy is it for you to talk to the following persons about things that really bother you? Mother, Father | ‘Very difficult’, ‘Difficult’ vs. ‘Easy’, ‘Very easy’, ‘Don't have or see this person’ | Yes |
| Lack of confident relations with one's friendsb (social isolation) | How easy is it for you to talk to the following persons about things that really bother you? Friends of the same sex, friends of the opposite sex | ‘Very difficult’, ‘Difficult’, ‘Easy’, ‘Very easy’, and ‘Do not have or see this person | No |
| Lack of educational plans for the futurec (meaninglessness) | What do you think you will be doing when you finish high school? | ‘College or University’, ‘Vocational or technical school’ and ‘Apprenticeship/trade’ vs. ‘Working’ and ‘Unemployed’. ‘Do not know’ was coded as missing | Yes (for use in correlation analyses) |
| Variable (corresponding theoretical dimension) | Phrasing of item | Response categories (‘vs.’ shows dichotomization) | Dichotomized |
|---|---|---|---|
| Lonelinessa (social isolation; cultural estrangement) | Do you feel lonely? | ‘Yes, very often’ and ‘Yes, rather often’ vs. ‘Yes, sometimes’ and ‘No’ | Yes |
| Feeling left out of things (social isolation; cultural estrangement) | How often do you feel left out of things? | ‘Always’ and ‘Often’ vs. ‘Sometimes’, ‘Rarely’ and ‘Never’ | Yes |
| Helplessness (powerlessness) | How often do you feel helpless? | ‘Always’ and ‘Often’ vs. ‘Sometimes’, ‘Rarely’ and ‘Never’ | Yes |
| Lack of confident relations with one's parents (social isolation) | How easy is it for you to talk to the following persons about things that really bother you? Mother, Father | ‘Very difficult’, ‘Difficult’ vs. ‘Easy’, ‘Very easy’, ‘Don't have or see this person’ | Yes |
| Lack of confident relations with one's friendsb (social isolation) | How easy is it for you to talk to the following persons about things that really bother you? Friends of the same sex, friends of the opposite sex | ‘Very difficult’, ‘Difficult’, ‘Easy’, ‘Very easy’, and ‘Do not have or see this person | No |
| Lack of educational plans for the futurec (meaninglessness) | What do you think you will be doing when you finish high school? | ‘College or University’, ‘Vocational or technical school’ and ‘Apprenticeship/trade’ vs. ‘Working’ and ‘Unemployed’. ‘Do not know’ was coded as missing | Yes (for use in correlation analyses) |
a: Excluded from index due to differential item functioning in relation to psychological symptoms.
b: Excluded from index due to insignificant correlation with both outcome measures.
c: Excluded from index due to insignificant correlations with the other five items.
Three items fulfilled all validation criteria: ‘helplessness’, ‘feeling left out of things’, and ‘lack of confident relations with one's parents’. In the last item we used the answer regarding the parent whom the student thought it was easiest to talk to.
Statistical analyses
All analyses were carried out using SAS version 8.2.
Step one was statistical validation of the alienation index: we used gamma correlations to test whether the potential items were correlated with one another as well as with the outcome measures. Items fulfilling the three first validation criteria were dichotomized and combined into a criteria index. The test for differential item functioning used logistic regression analysis to assess the association between this index and the outcome measures, adjusted for each of the items in the index one at a time. If an item was significantly associated with the outcome measure besides its influence in the index it was a source of differential item functioning and was removed from the index. To remain in the index, an item should become insignificant when adjusted for.
Step two was logistic regression and gamma correlations to test whether potential confounders were associated with both alienation and daily physical and psychological symptoms. The confounding effect of each of the covariates was tested using logistic regression.
Step three was multivariate logistic regression analyses with stepwise backwards selection to investigate the association between alienation and daily physical and psychological symptoms respectively. Interaction was tested for by including interaction terms in the model fitting. In relation to sex we tested for interaction between sex and the other variables. Furthermore to be sure that no second order interactions were present and to see the pattern of association for boys and girls separately we also conducted sex segregated analyses. The pattern of association was similar for boys and girls and no significant interactions were found. Therefore and to avoid an inappropriate loss of statistical power, the final analyses combined boys and girls.
As the models for daily physical and psychological symptoms were adjusted for different confounders the analyses included a different number of students, for physical symptoms 4388 students (84.3%) and psychological symptoms 4867 (93.8%).
Results
Six items founded in four of Seeman's alienation dimensions formed the basis for the development and validation of an alienation index (table 1). One item did not fulfil the criteria of inter-correlation (‘lack of educational plans for the future’) and one was not correlated with the outcome measures (‘lack of confident relations with one's friends’). Further, ‘loneliness’ caused differential item functioning in relation to the outcome ‘daily psychological symptoms’. Therefore loneliness was excluded from the final index and treated as a potential confounder in the final analyses.
The three items in the final index (‘helplessness’, ‘feeling left out of things’ and ‘lack of confident relations with one's parents’) shown in figure 1 represent aspects of three of Seeman's alienation dimensions: social isolation, cultural estrangement and powerlessness. Each item counted one if indication of alienation was present (figure 1).
The index of alienation
Table 2 shows the prevalence of alienation, physical and psychological symptoms and the covariates. The overall prevalence of the determinant, including all three degrees of alienation, was 23.9%. The prevalence of experiencing daily physical symptoms was 9.4%, daily psychological symptoms 13.9%.
The prevalence of determinant, outcome variables and covariates in absolute numbers and percentages
| Frequency | % | |
|---|---|---|
| Determinant and outcome | ||
| Alienation score | ||
| 0 | 3904 | 75.0 |
| 1 | 1083 | 20.8 |
| 2 | 124 | 2.4 |
| 3 | 34 | 0.7 |
| Missing | 60 | 1.2 |
| Physical symptoms | ||
| No | 4490 | 86.3 |
| Yes | 487 | 9.4 |
| Missing | 228 | 4.3 |
| Psychological symptoms | ||
| No | 4218 | 81.0 |
| Yes | 723 | 13.9 |
| Missing | 264 | 5.1 |
| Covariates | ||
| Sex | ||
| Boys | 2624 | 50.4 |
| Girls | 2581 | 49.6 |
| Socio-economic position | ||
| 1 | 386 | 7.4 |
| 2 | 954 | 18.3 |
| 3 | 1027 | 19.7 |
| 4 | 1361 | 26.1 |
| 5 | 650 | 12.5 |
| Outside labour market | 260 | 5.0 |
| Missing | 567 | 10.9 |
| Been bullied | ||
| Have not | 2554 | 49.1 |
| Once or twice | 1284 | 24.6 |
| Sometimes | 902 | 17.3 |
| Once a week | 180 | 3.5 |
| Many times a week | 249 | 4.8 |
| Missing | 36 | 0.7 |
| Feel lonely | ||
| No | 3110 | 59.8 |
| Sometimes | 1729 | 33.2 |
| Rather often | 238 | 4.6 |
| Very often | 105 | 2.0 |
| Missing | 23 |
| Frequency | % | |
|---|---|---|
| Determinant and outcome | ||
| Alienation score | ||
| 0 | 3904 | 75.0 |
| 1 | 1083 | 20.8 |
| 2 | 124 | 2.4 |
| 3 | 34 | 0.7 |
| Missing | 60 | 1.2 |
| Physical symptoms | ||
| No | 4490 | 86.3 |
| Yes | 487 | 9.4 |
| Missing | 228 | 4.3 |
| Psychological symptoms | ||
| No | 4218 | 81.0 |
| Yes | 723 | 13.9 |
| Missing | 264 | 5.1 |
| Covariates | ||
| Sex | ||
| Boys | 2624 | 50.4 |
| Girls | 2581 | 49.6 |
| Socio-economic position | ||
| 1 | 386 | 7.4 |
| 2 | 954 | 18.3 |
| 3 | 1027 | 19.7 |
| 4 | 1361 | 26.1 |
| 5 | 650 | 12.5 |
| Outside labour market | 260 | 5.0 |
| Missing | 567 | 10.9 |
| Been bullied | ||
| Have not | 2554 | 49.1 |
| Once or twice | 1284 | 24.6 |
| Sometimes | 902 | 17.3 |
| Once a week | 180 | 3.5 |
| Many times a week | 249 | 4.8 |
| Missing | 36 | 0.7 |
| Feel lonely | ||
| No | 3110 | 59.8 |
| Sometimes | 1729 | 33.2 |
| Rather often | 238 | 4.6 |
| Very often | 105 | 2.0 |
| Missing | 23 |
The prevalence of determinant, outcome variables and covariates in absolute numbers and percentages
| Frequency | % | |
|---|---|---|
| Determinant and outcome | ||
| Alienation score | ||
| 0 | 3904 | 75.0 |
| 1 | 1083 | 20.8 |
| 2 | 124 | 2.4 |
| 3 | 34 | 0.7 |
| Missing | 60 | 1.2 |
| Physical symptoms | ||
| No | 4490 | 86.3 |
| Yes | 487 | 9.4 |
| Missing | 228 | 4.3 |
| Psychological symptoms | ||
| No | 4218 | 81.0 |
| Yes | 723 | 13.9 |
| Missing | 264 | 5.1 |
| Covariates | ||
| Sex | ||
| Boys | 2624 | 50.4 |
| Girls | 2581 | 49.6 |
| Socio-economic position | ||
| 1 | 386 | 7.4 |
| 2 | 954 | 18.3 |
| 3 | 1027 | 19.7 |
| 4 | 1361 | 26.1 |
| 5 | 650 | 12.5 |
| Outside labour market | 260 | 5.0 |
| Missing | 567 | 10.9 |
| Been bullied | ||
| Have not | 2554 | 49.1 |
| Once or twice | 1284 | 24.6 |
| Sometimes | 902 | 17.3 |
| Once a week | 180 | 3.5 |
| Many times a week | 249 | 4.8 |
| Missing | 36 | 0.7 |
| Feel lonely | ||
| No | 3110 | 59.8 |
| Sometimes | 1729 | 33.2 |
| Rather often | 238 | 4.6 |
| Very often | 105 | 2.0 |
| Missing | 23 |
| Frequency | % | |
|---|---|---|
| Determinant and outcome | ||
| Alienation score | ||
| 0 | 3904 | 75.0 |
| 1 | 1083 | 20.8 |
| 2 | 124 | 2.4 |
| 3 | 34 | 0.7 |
| Missing | 60 | 1.2 |
| Physical symptoms | ||
| No | 4490 | 86.3 |
| Yes | 487 | 9.4 |
| Missing | 228 | 4.3 |
| Psychological symptoms | ||
| No | 4218 | 81.0 |
| Yes | 723 | 13.9 |
| Missing | 264 | 5.1 |
| Covariates | ||
| Sex | ||
| Boys | 2624 | 50.4 |
| Girls | 2581 | 49.6 |
| Socio-economic position | ||
| 1 | 386 | 7.4 |
| 2 | 954 | 18.3 |
| 3 | 1027 | 19.7 |
| 4 | 1361 | 26.1 |
| 5 | 650 | 12.5 |
| Outside labour market | 260 | 5.0 |
| Missing | 567 | 10.9 |
| Been bullied | ||
| Have not | 2554 | 49.1 |
| Once or twice | 1284 | 24.6 |
| Sometimes | 902 | 17.3 |
| Once a week | 180 | 3.5 |
| Many times a week | 249 | 4.8 |
| Missing | 36 | 0.7 |
| Feel lonely | ||
| No | 3110 | 59.8 |
| Sometimes | 1729 | 33.2 |
| Rather often | 238 | 4.6 |
| Very often | 105 | 2.0 |
| Missing | 23 |
Table 3 presents crude and adjusted odds ratios for daily physical and psychological symptoms respectively. There was a graded and significant association between alienation and physical symptoms even after adjustment for the confounders sex, SEP, bullying and loneliness which were all strongly associated with physical symptoms. The odds ratio for daily physical symptoms was 2.49 (95% CI 1.05–5.87) among students experiencing all three indicators of alienation (score = 3) compared to students with no indicators of alienation. In the sex stratified analyses the corresponding odds ratios were 3.24 (0.82–12.76) for boys and 2.21 (0.23–6.77) for girls (OR not shown in tables).
Crude and adjusted OR (95% CI) for experiencing daily physical and psychological symptoms respectively by the degree of alienation
| Crude OR for daily physical symptoms | Adjusted OR for daily physical symptoms (n = 4388) | Crude OR for daily psychological symptoms | Adjusted OR for daily psychological symptoms (n = 4867) | |
|---|---|---|---|---|
| Alienation score | ||||
| 0 | 1 | 1 | 1 | 1 |
| 1 | 1.74 (1.40–2.16) | 1.37 (1.08–1.75) | 2.47 (2.07–2.95) | 2.18 (1.82–2.62) |
| 2 | 4.11 (2.70–6.26) | 2.10 (1.28–3.48) | 5.59 (3.83–8.17) | 3.74 (2.51–5.58) |
| 3 | 5.78 (2.78–12.03) | 2.49 (1.05–5.87) | 11.0 (5.47–22.11) | 6.50 (3.11–13.56) |
| Sex | ||||
| Boys | 1 | 1 | 1 | 1 |
| Girls | 1.73 (1.42–2.09) | 1.56 (1.23–1.94) | 1.57 (1.34–1.84) | 1.57 (1.33–1.85) |
| Socio-economic position | ||||
| 1 | 1 | 1 | – | – |
| 2 | 1.90 (1.14–3.18) | 2.05 (1.20–3.48) | – | – |
| 3 | 1.89 (1.14–3.15) | 1.90 (1.12–3.24) | – | – |
| 4 | 2.09 (1.27–3.42) | 2.02 (1.21–3.39) | – | – |
| 5 | 2.13 (1.26–3.62) | 1.95 (1.13–3.38) | – | – |
| Outside labour market | 3.63 (2.04–6.45) | 3.28 (1.81–5.98) | – | – |
| Been bullied | ||||
| Have not | 1 | 1 | 1 | 1 |
| Once or twice | 0.98 (0.76–1.26) | 0.87 (0.65–1.15) | 1.27 (1.03–1.57) | 1.24 (1.00–1.54) |
| Sometimes | 1.74 (1.36–2.23) | 1.47 (1.11–1.95) | 2.06 (1.67–2.55) | 1.87 (1.51–2.33) |
| Once a week | 2.39 (1.56–3.66) | 1.79 (1.11–2.90) | 3.12 (2.18–4.49) | 2.53 (1.72–3.71) |
| Many times a week | 3.87 (2.78–5.38) | 2.42 (1.63–3.59) | 5.46 (4.08–7.31) | 4.15 (3.04–5.67) |
| Feel lonely | ||||
| No | 1 | 1 | – | – |
| Sometimes | 1.84 (1.50–2.25) | 1.48 (1.17–1.87) | – | – |
| Rather often | 3.71 (2.63–5.23) | 2.14 (1.42–3.25) | – | – |
| Very often | 4.88 (3.07–7.74) | 2.13 (1.20–3.78) | – | – |
| Crude OR for daily physical symptoms | Adjusted OR for daily physical symptoms (n = 4388) | Crude OR for daily psychological symptoms | Adjusted OR for daily psychological symptoms (n = 4867) | |
|---|---|---|---|---|
| Alienation score | ||||
| 0 | 1 | 1 | 1 | 1 |
| 1 | 1.74 (1.40–2.16) | 1.37 (1.08–1.75) | 2.47 (2.07–2.95) | 2.18 (1.82–2.62) |
| 2 | 4.11 (2.70–6.26) | 2.10 (1.28–3.48) | 5.59 (3.83–8.17) | 3.74 (2.51–5.58) |
| 3 | 5.78 (2.78–12.03) | 2.49 (1.05–5.87) | 11.0 (5.47–22.11) | 6.50 (3.11–13.56) |
| Sex | ||||
| Boys | 1 | 1 | 1 | 1 |
| Girls | 1.73 (1.42–2.09) | 1.56 (1.23–1.94) | 1.57 (1.34–1.84) | 1.57 (1.33–1.85) |
| Socio-economic position | ||||
| 1 | 1 | 1 | – | – |
| 2 | 1.90 (1.14–3.18) | 2.05 (1.20–3.48) | – | – |
| 3 | 1.89 (1.14–3.15) | 1.90 (1.12–3.24) | – | – |
| 4 | 2.09 (1.27–3.42) | 2.02 (1.21–3.39) | – | – |
| 5 | 2.13 (1.26–3.62) | 1.95 (1.13–3.38) | – | – |
| Outside labour market | 3.63 (2.04–6.45) | 3.28 (1.81–5.98) | – | – |
| Been bullied | ||||
| Have not | 1 | 1 | 1 | 1 |
| Once or twice | 0.98 (0.76–1.26) | 0.87 (0.65–1.15) | 1.27 (1.03–1.57) | 1.24 (1.00–1.54) |
| Sometimes | 1.74 (1.36–2.23) | 1.47 (1.11–1.95) | 2.06 (1.67–2.55) | 1.87 (1.51–2.33) |
| Once a week | 2.39 (1.56–3.66) | 1.79 (1.11–2.90) | 3.12 (2.18–4.49) | 2.53 (1.72–3.71) |
| Many times a week | 3.87 (2.78–5.38) | 2.42 (1.63–3.59) | 5.46 (4.08–7.31) | 4.15 (3.04–5.67) |
| Feel lonely | ||||
| No | 1 | 1 | – | – |
| Sometimes | 1.84 (1.50–2.25) | 1.48 (1.17–1.87) | – | – |
| Rather often | 3.71 (2.63–5.23) | 2.14 (1.42–3.25) | – | – |
| Very often | 4.88 (3.07–7.74) | 2.13 (1.20–3.78) | – | – |
Crude and adjusted OR (95% CI) for experiencing daily physical and psychological symptoms respectively by the degree of alienation
| Crude OR for daily physical symptoms | Adjusted OR for daily physical symptoms (n = 4388) | Crude OR for daily psychological symptoms | Adjusted OR for daily psychological symptoms (n = 4867) | |
|---|---|---|---|---|
| Alienation score | ||||
| 0 | 1 | 1 | 1 | 1 |
| 1 | 1.74 (1.40–2.16) | 1.37 (1.08–1.75) | 2.47 (2.07–2.95) | 2.18 (1.82–2.62) |
| 2 | 4.11 (2.70–6.26) | 2.10 (1.28–3.48) | 5.59 (3.83–8.17) | 3.74 (2.51–5.58) |
| 3 | 5.78 (2.78–12.03) | 2.49 (1.05–5.87) | 11.0 (5.47–22.11) | 6.50 (3.11–13.56) |
| Sex | ||||
| Boys | 1 | 1 | 1 | 1 |
| Girls | 1.73 (1.42–2.09) | 1.56 (1.23–1.94) | 1.57 (1.34–1.84) | 1.57 (1.33–1.85) |
| Socio-economic position | ||||
| 1 | 1 | 1 | – | – |
| 2 | 1.90 (1.14–3.18) | 2.05 (1.20–3.48) | – | – |
| 3 | 1.89 (1.14–3.15) | 1.90 (1.12–3.24) | – | – |
| 4 | 2.09 (1.27–3.42) | 2.02 (1.21–3.39) | – | – |
| 5 | 2.13 (1.26–3.62) | 1.95 (1.13–3.38) | – | – |
| Outside labour market | 3.63 (2.04–6.45) | 3.28 (1.81–5.98) | – | – |
| Been bullied | ||||
| Have not | 1 | 1 | 1 | 1 |
| Once or twice | 0.98 (0.76–1.26) | 0.87 (0.65–1.15) | 1.27 (1.03–1.57) | 1.24 (1.00–1.54) |
| Sometimes | 1.74 (1.36–2.23) | 1.47 (1.11–1.95) | 2.06 (1.67–2.55) | 1.87 (1.51–2.33) |
| Once a week | 2.39 (1.56–3.66) | 1.79 (1.11–2.90) | 3.12 (2.18–4.49) | 2.53 (1.72–3.71) |
| Many times a week | 3.87 (2.78–5.38) | 2.42 (1.63–3.59) | 5.46 (4.08–7.31) | 4.15 (3.04–5.67) |
| Feel lonely | ||||
| No | 1 | 1 | – | – |
| Sometimes | 1.84 (1.50–2.25) | 1.48 (1.17–1.87) | – | – |
| Rather often | 3.71 (2.63–5.23) | 2.14 (1.42–3.25) | – | – |
| Very often | 4.88 (3.07–7.74) | 2.13 (1.20–3.78) | – | – |
| Crude OR for daily physical symptoms | Adjusted OR for daily physical symptoms (n = 4388) | Crude OR for daily psychological symptoms | Adjusted OR for daily psychological symptoms (n = 4867) | |
|---|---|---|---|---|
| Alienation score | ||||
| 0 | 1 | 1 | 1 | 1 |
| 1 | 1.74 (1.40–2.16) | 1.37 (1.08–1.75) | 2.47 (2.07–2.95) | 2.18 (1.82–2.62) |
| 2 | 4.11 (2.70–6.26) | 2.10 (1.28–3.48) | 5.59 (3.83–8.17) | 3.74 (2.51–5.58) |
| 3 | 5.78 (2.78–12.03) | 2.49 (1.05–5.87) | 11.0 (5.47–22.11) | 6.50 (3.11–13.56) |
| Sex | ||||
| Boys | 1 | 1 | 1 | 1 |
| Girls | 1.73 (1.42–2.09) | 1.56 (1.23–1.94) | 1.57 (1.34–1.84) | 1.57 (1.33–1.85) |
| Socio-economic position | ||||
| 1 | 1 | 1 | – | – |
| 2 | 1.90 (1.14–3.18) | 2.05 (1.20–3.48) | – | – |
| 3 | 1.89 (1.14–3.15) | 1.90 (1.12–3.24) | – | – |
| 4 | 2.09 (1.27–3.42) | 2.02 (1.21–3.39) | – | – |
| 5 | 2.13 (1.26–3.62) | 1.95 (1.13–3.38) | – | – |
| Outside labour market | 3.63 (2.04–6.45) | 3.28 (1.81–5.98) | – | – |
| Been bullied | ||||
| Have not | 1 | 1 | 1 | 1 |
| Once or twice | 0.98 (0.76–1.26) | 0.87 (0.65–1.15) | 1.27 (1.03–1.57) | 1.24 (1.00–1.54) |
| Sometimes | 1.74 (1.36–2.23) | 1.47 (1.11–1.95) | 2.06 (1.67–2.55) | 1.87 (1.51–2.33) |
| Once a week | 2.39 (1.56–3.66) | 1.79 (1.11–2.90) | 3.12 (2.18–4.49) | 2.53 (1.72–3.71) |
| Many times a week | 3.87 (2.78–5.38) | 2.42 (1.63–3.59) | 5.46 (4.08–7.31) | 4.15 (3.04–5.67) |
| Feel lonely | ||||
| No | 1 | 1 | – | – |
| Sometimes | 1.84 (1.50–2.25) | 1.48 (1.17–1.87) | – | – |
| Rather often | 3.71 (2.63–5.23) | 2.14 (1.42–3.25) | – | – |
| Very often | 4.88 (3.07–7.74) | 2.13 (1.20–3.78) | – | – |
The crude logistic regression analyses also showed a graded and significant association between alienation and daily psychological symptoms. When adjusted for sex and bullying the estimates decreased substantially, but the association remained graded and significant (table 3). Students with all three indicators of alienation (score = 3) had 6.50 (3.11–13.56) times greater odds for experiencing daily psychological symptoms compared to students with no indicators of alienation. The corresponding odds ratios in the sex-stratified analyses were 8.22 (2.64–25.59) for boys and 5.5 (2.13–14.47) for girls (not shown in tables).
The analyses for differential item functioning suggested that loneliness could be a mediator in the association between alienation and psychological symptoms as loneliness had an independent and significant effect besides its effect in the alienation index. Furthermore preliminary analyses showed that SEP did not confound the association between alienation and psychological symptoms. Neither was SEP significantly associated with psychological symptoms during the model fitting. Thus the final model was not adjusted for SEP and loneliness.
Discussion
This study had two main findings. First, three items coherent with Seeman's conceptualization of alienation8,9—‘helplessness’, ‘feeling left out of things’ and ‘lack of confidentiality with one's parents’—combined into a valid index for the measurement of alienation. Secondly, alienation was associated with experiencing daily physical and psychological symptoms. In both models the odds ratio estimates and their confidence intervals showed a significant and graded tendency suggesting that the association between alienation and high symptom load is real.
We have only identified one previous study investigating this association. Natvig et al. (1999) found that school alienation was associated with an increased odds ratio (OR = 2.15) for experiencing psychosomatic symptoms among 13- to 15-years-olds students.17 Nutbeam et al.12 likewise focused on alienation towards school but addressed the association with various kinds of risk behaviour (smoking, alcohol use, having been drunk, less exercising and eating unhealthy food). Alienation increased all these risk behaviours. Eder24 found an association between lack of social integration, and physical or psychological symptoms. This study does not address alienation specifically. Still its findings suggest that other psychosocial concepts related to dimensions of alienation are associated with high symptom load.
The questionnaire used in the HBSC survey did not include a previously validated alienation index. Due to pragmatic reasons, alienation was therefore measured by a new index. Thus our findings are not directly comparable with results of other studies. Alienation has been measured by various scales and indexes25 and no golden standard exists. This also applies to the number of alienation dimensions included. However, most alienation scales, as this study's, take their starting points in Seeman's conceptualization of alienation.10 Despite differences in items and alienation dimensions covered it is still possible to compare our results with other studies’ findings, on a conceptual level.
We used construct validity as a criterion for accepting the index, i.e. no differential item functioning in relation to the outcome measure. This criterion left us with four items (the above three mentioned plus ‘loneliness’) in relation to the outcome ‘daily physical symptoms’ but only three items in relation to ‘daily psychological symptoms’ as the item ‘loneliness’ caused differential item functioning. Due to the clarity of the analyses we used only one index of alienation, consisting of only three indicators of alienation. The consequence of excluding loneliness from the index will, in the analysis for physical symptoms, be a slightly weaker reliability. This problem is largest when the number of participants is small and the risk of type-2 errors bigger. The consequence of including loneliness in the index would have been a systematic bias in relation to certain subgroups of students in the analysis for psychological symptoms. As the number of participants was high and thus the risk of type-2 errors small, we excluded loneliness from the final index of alienation.
From a conceptual point of view it may seem surprising that loneliness was removed from the alienation index. However, Seeman's alienation concept only identifies the most common ways in which alienation has been used empirically and theoretically8—it does not require a full coverage of all the dimensions. Therefore we see no crucial problems in excluding loneliness from the index from a conceptual perspective.
Our choice of not including the item ‘loneliness’ as a confounder in the analysis for daily psychological symptoms may have exaggerated the association between alienation and psychological symptoms. We did however also conduct an analysis adjusted for loneliness. This analysis showed smaller estimates but still a graded and significant association between alienation and psychological symptoms. Students with all three indicators of alienation present had an odds ratio of 2.89 (1.32–6.34) for daily symptoms compared to students with no indicators of alienation.
Our findings do not reveal the processes between alienation and symptom load. The increased symptom load among alienated students could be caused by alienation-related stress or other psychosomatic pathways between alienation and symptoms. Another possible pathway is that alienation leads to risk behaviours12 that causes an increased symptom load. Finally a possible explanation might be that an underlying structural factor (e.g. school environment) generates stress and confounds the association between alienation and symptom load. More studies are needed to identify the pathways between alienation and symptom load.
We expected age differences in the association between alienation and symptom load because both social relations and symptom load are linked to age19,21,24 and because the age span from 11 to 15 represents a considerable psychological development. We found no significant age differences, however, and age was therefore not included as a confounder. Still, the processes behind the association might differ between the three age groups.
Limitations: Firstly, the cross-sectional design of the study does not reveal the direction of the association between alienation and symptom load. Secondly, it was not possible to conduct an analysis of the non-respondents due to the anonymity of the study. Thus the risk of selection bias must be considered even though the response rate was high. If the non-respondents were more alienated and had a higher symptom load, we would see a stronger association between alienation and symptom load. Thirdly, the items of the index have been validated and used in several school surveys. Still, the risk of information bias must be considered as the applied items have not been validated in relation to the full range of Seeman's conceptualization of alienation. Since there is no requirement for a full coverage of all the dimensions it is a merit of the study that our index of alienation fulfils the criterion of construct validity. Another merit is the large representative sample and the anonymity of the students which stimulated honest answers.
Alienation may be a concept which should be taken into consideration when planning health interventions among adolescents. As powerlessness is one dimension of alienation one-way to prevent alienation could be to strengthen the students’ empowerment. Psychological empowerment has been defined as ‘… a subjective feeling of greater control over one's own life that an individual experiences following active membership in groups or organizations’ by Bracht et al.26 Empirical studies suggest that use of empowerment strategies decrease global alienation and social isolation, increase the adolescents’ social bonding with family, school and peer-group and their ability to play an assertive role in controlling one's participation in social activities.27–29
The principles behind empowerment of adolescents are (i) involvement in projects which the adolescents consider meaningful, (ii) critical awareness of the surrounding world, (iii) participation in problem identification, strategies to solve the problem and the social action to change the conditions causing problems, (iv) collaboration with adults and (v) building upon the strengths and successes of the adolescents.27,30–33 It is crucial that the participation of adolescents is genuine and not just symbolic.34 In The Youth Empowerment Strategies (YES!) Project32–33 the intervention model intends to stimulate health and wellness outcomes by means of empowerment elements as collaborative decision making, conflict resolution skills, involvement in designing and implementing positive project efforts and increased efficacy and political participation. Most empowerment interventions among adolescents focus on the community setting.27,28,35,36 However, the YES project focuses on the school as a small community engaging pupils in after school activities suggesting that the principles of empowerment may also be useful within the school and class setting, where adolescents spend a large amount of time.
The YES intervention model has been widely adopted as a mean to reduce and prevent tobacco use.33 Although there are no hard facts to document the effectiveness of the program, initial experiences suggest that it is possible to stimulate a positive empowerment cycle among adolescents. The number of intervention studies carried out in the school setting and focusing on overall empowerment is, however, limited.
Funding
Health Insurance Fund in Denmark (j.nr.: 11/216-98); The Danish Ministry of Health (5.kt.j.nr.: 1999-524/04-20); the Nordea Denmark Foundation.
Conflicts of interest: None declared.
An index of alienation including three indicators of alienation: helplessness, feeling left out of things and lack of confidentiality with parents fulfilled psychometric criteria for scalability.
Alienation is significantly associated with a high symptom load among adolescents.
Alienation may be a concept to be taken into account in future health interventions among adolescents. In school settings this may be done by use of principles of empowerment.
Acknowledgements
We thank the Principal Investigator for the Danish contribution to the HBSC project, Dr Pernille Due, for her permission to use the data set.
Comments