592
European Journal of Public Health
20 Arendt JN. Does education cause better health? A panel data analysis using school reforms for identification. Econ Educ Rev 2005;24:149–60.
27 Ju¨rges H, Reinhold S, Salm M. Does schooling affect health behavior? Evidence from the educational expansion in Western Germany. Econ Educ Rev 2011;30:862–72.
21 Brunello G, Fabbri D, Fort M. Years of Schooling, Human Capital and the Body Mass Index of European Females. Berlin: Deutsches Institut fuu¨r Wirtschaftsforschung, 2010.
28 Reinhold S, Ju¨rges H. Secondary school fees and the causal effect of schooling on health behavior. Health Econ 2010;19:994–1001.
22 Braga M, Bratti M.The Causal Effect of Education on Health and Health-Related Behavior: Evidence from a Compulsory Schooling Reform. Health, Econometrics and Data Group (HEDG), Department of Economics, University of York, 2012.
29 Guyatt GH, Oxman AD, Kunz R, et al. GRADE guidelines: 7 Rating the quality of evidence–inconsistency. J Clin Epidemiol 2011;64:1294–302.
23 Lager ACJ, Torssander J. Causal effect of education on mortality in a quasiexperiment on 1.2 million Swedes. Proc Natl Acad Sci USA 2012;109:8461–6. 24 Meghir C, Palme M, Simeonova E. Education, Health and Mortality: Evidence from a Social Experiment. Stockholm: Department of Economics, Stockholm University, 2013. 25 Fischer M, Karlsson M, Nilsson T. Effects of compulsory schooling on mortality: evidence from Sweden. Int J Environ Res Public Health 2013;10:3596–618.
30 Clark D, Royer H. The effect of education on adult mortality and health: evidence from Britain. Am Econ Rev 2013;103:2087–120. 31 Erikson R, Torssander J. Clerics die, doctors survive: a note on death risks among highly educated professionals. Scand J Public Health 2009;37:227–31. 32 Deaton A. The Great Escape: Health, Wealth, and the Origins of Inequality. Princeton: Princeton University Press, 2013.
26 James J. Health and Education Expansion. Bath, UK: Department of Economics, 2013.
......................................................................................................... European Journal of Public Health, Vol. 25, No. 4, 592–597 ß The Author 2014. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved. doi:10.1093/eurpub/cku213 Advance Access published on 19 December 2014
Cumulative contextual and individual disadvantages over the life course and adult functional somatic symptoms in Sweden Per E. Gustafsson1, Anne Hammarstro¨m1, Miguel San Sebastian2 1 Division of Social Medicine, Department of Public Health and Clinical Medicine, Umea˚ University, Umea˚, Sweden 2 Division of Epidemiology and Global Health, Department of Public Health and Clinical Medicine, Umea˚ University, Umea˚, Sweden Correspondence: Per E. Gustafsson, Division of Social Medicine, Department of Public Health and Clinical Medicine, Umea˚ University, SE-901 85 Umea˚, Sweden, Tel: +46 90 785 35 90, Fax: +46 90 776 883, e-mail: [email protected]
Background: Disadvantage, originating in one’s residential context or in one’s past life course, has been shown to impact on health in adulthood. There is however little research on the accumulated health impact of both neighbourhood and individual conditions over the life course. This study aims to examine whether the accumulation of contextual and individual disadvantages from adolescence to middle-age predicts functional somatic symptoms (FSS) in middle-age, taking baseline health into account. Methods: The sample is the age 16, 21, 30 and 42 surveys of the prospective Northern Swedish Cohort, with analytical sample size n = 910 (85% of the original cohort). FSS at age 16 and 42, and cumulative socioeconomic disadvantage, social adversity and material adversity between 16 and 42 years were operationalized from questionnaires, and cumulative neighbourhood disadvantage between 16 and 42 years from register data. Results: Results showed accumulation of disadvantages jointly explained 9–12% of FSS variance. In the total sample, cumulative neighbourhood and socioeconomic disadvantage significantly predicted FSS at age 42 in the total sample. In women, neighbourhood disadvantage but not socioeconomic disadvantage contributed significantly, whereas in men, socioeconomic but not neighbourhood disadvantage contributed significantly. In all analyses, associations were largely explained by the parallel accumulation of social and material adversities, but not by symptoms at baseline. Conclusion: In conclusion, the accumulation of diverse forms of disadvantages together plays an important role for somatic complaints in adulthood, independently of baseline health.
.........................................................................................................
Introduction oth physical and mental health are shaped by the social conditions
Bunder which people live their lives. Social determinants of health
are found not only in the direct environment of the home and work but also in the surrounding context, such as the neighbourhood.1 Moreover, not only one’s current social conditions matter for health but also circumstances from one’s past life course.2 Drawing on these two approaches to social determinants of health, the present report seeks to investigate whether accumulation of neighbourhood and individual disadvantages over the life course is related to selfreported health, in the form of functional somatic symptoms (FSS), in a middle-aged Swedish population.
The parallel observations that contextual factors may impact on concurrent health in both young people3,4 and adults,5 and that personal living conditions from earlier life may have long-term health effects,6 have in recent years converged into integrated research studying the health effects of contextual determinants from the past life course.7,8 A handful of studies suggest that the health impact of unfavourable neighbourhood conditions seems to accumulate over the life course,9,10 with less independent importance of context at specific life course period.11 This notion corresponds to what within life course research is called a ‘cumulative risk model’, where the total accumulation of adverse exposures across the life course corresponds to a greater risk for ill health in adulthood.12
Downloaded from http://eurpub.oxfordjournals.org/ by guest on November 15, 2015
.........................................................................................................
Contextual and individual disadvantages
Methods
FSS at age 16 and 42 The FSS measure covered symptoms within the areas of cardiopulmonary/autonomic, gastrointestinal, musculoskeletal and general symptoms, similar to other FSS measures.13 It was operationalized identically at age 16 and 42 as a summative score of 10 questionnaire items, each coded 0–2, and covering symptoms during the last 12 months. First, eight symptoms (headache or migraine; stomach ache other than gastritis or gastric ulcer; nausea; backache, hip pain or sciatica; fatigue; breathlessness; dizziness; overstrain) were each assessed by the question ‘Do you have (or have you during the last 12 months had) any of the following:’ with response options ‘No’ (0), ‘Yes, light’ (1) and ‘Yes, severe’ (2) for each symptom. Second, palpitations were measured by the question ‘How often have you had nervous problems during the past 12 months’, where participant indicated specific symptoms, including palpitations, with response options ‘never’ (0), ‘sometimes’ (1) and ‘always’ (2). Third, sleeplessness came from the question ‘Have you had sleeping difficulties during the past 12 months?’ with the response options coded as ‘Never’ (0), ‘Sometimes’ (1) and ‘often’ or ‘always’ (2). All 10 items were summed up into a score with range 0–20. Internal consistency for the FSS measure was Cronbach’s = 0.79 at age 42, and = 0.70 at age 16.
Sample and procedures
Cumulative ND over the life course
The Northern Swedish Cohort is based on all school-leavers of the ninth grade of the Swedish compulsory school (aged 16 years), in the municipality of Lulea˚, in 1981 (N = 1083). For details of the sample and procedures, see Refs. 26 and 29. For this report, the baseline survey at age 16 as well as follow-up surveys at age 21 (1986), 30 (1995) and 42 (2007) were used. At age 42, there were 1001 participants (93.4% of n = 1071 of the original cohort still alive), for whom n = 910 had complete data (85.0%). At each survey, participants completed a comprehensive selfadministered questionnaire about health and social circumstances, which was used to operationalize cumulative disadvantages and health. Data were also retrieved from Statistics Sweden to create aggregate measures of cumulative neighbourhood disadvantage (ND). Ethical approval was granted by the Regional Ethical Review Board in Umea˚.
Cumulative ND was operationalized in accordance with a previous report.10 The basis was register data on the neighbourhoods (SmallArea Market Statistics areas) which at least one participant was registered in, on 31 December 1980, 1986, 1995 and 2007. For all residents in these neighbourhoods, a number of variables were retrieved and used to operationalize ND. First, aggregate measures of ND were calculated at age 16, 21, 30 and 42. At each age, eight neighbourhood indicators were selected: percentages of residents with (i) ‘Low income’, (ii) ‘High income’ (reverse coded), (iii) ‘Housing allowance’, (iv) ‘Wealth’ (reverse coded), (v) ‘Non-employment’, (vi) ‘Single-parent household’, (viia) ‘Low occupational status’ (only available for 1981 and 1986), (viib) ‘Low educational attainment’ (only available for 1995 and 2007), (viiia) ‘High occupational status’ (reverse coded; only available for 1981 and 1986) and (viiib) ‘High educational achievement’ (reverse coded; only available for 1995 and 2007). All indicators were Z-transformed and the mean of all indicators was calculated, creating four age-specific ND scores. Internal consistency for the age-specific scores was Cronbach’s = 0.89 (1981); = 0.81 (1986); = 0.86 (1995); and = 0.89 (2007) at the neighbourhood level, and = 0.93 (1981); = 0.88 (1986); = 0.85 (1995); and = 0.86 (2007) at the individual level.
Measures Cumulative ND was based on register data corresponding to age 16, 21, 30 and 42, and all other measures were based on selfadministered questionnaires completed at the corresponding ages. See table 1 for descriptive statistics of all measures.
Table 1 Descriptive statistics of key variables in the total sample and by gender (N = 910; 438 women and 472 men) Variable
FSS at age 42, M(SD) FSS at age 16, M(SD) ND, M(SD) SED, n (%) 0 occasions 1 occasion 2 occasion 3 occasion 4 occasion Social adversity, M(SD) Material adversity, M(SD)
a: P-value from t-test. b: P-value from 2-test.
Total sample
By gender
Range
Estimate
Women
Men
P-value
0–18 0–16 2.02 to 2.65 0–4
4.21 (3.32) 3.34 (2.53) 0.01 (0.67)
4.78 (3.51) 3.71 (2.51) 0.05 (0.66)
3.69 (3.03) 3.01 (2.51) 0.03 (0.68)
European Journal of Public Health
20 Arendt JN. Does education cause better health? A panel data analysis using school reforms for identification. Econ Educ Rev 2005;24:149–60.
27 Ju¨rges H, Reinhold S, Salm M. Does schooling affect health behavior? Evidence from the educational expansion in Western Germany. Econ Educ Rev 2011;30:862–72.
21 Brunello G, Fabbri D, Fort M. Years of Schooling, Human Capital and the Body Mass Index of European Females. Berlin: Deutsches Institut fuu¨r Wirtschaftsforschung, 2010.
28 Reinhold S, Ju¨rges H. Secondary school fees and the causal effect of schooling on health behavior. Health Econ 2010;19:994–1001.
22 Braga M, Bratti M.The Causal Effect of Education on Health and Health-Related Behavior: Evidence from a Compulsory Schooling Reform. Health, Econometrics and Data Group (HEDG), Department of Economics, University of York, 2012.
29 Guyatt GH, Oxman AD, Kunz R, et al. GRADE guidelines: 7 Rating the quality of evidence–inconsistency. J Clin Epidemiol 2011;64:1294–302.
23 Lager ACJ, Torssander J. Causal effect of education on mortality in a quasiexperiment on 1.2 million Swedes. Proc Natl Acad Sci USA 2012;109:8461–6. 24 Meghir C, Palme M, Simeonova E. Education, Health and Mortality: Evidence from a Social Experiment. Stockholm: Department of Economics, Stockholm University, 2013. 25 Fischer M, Karlsson M, Nilsson T. Effects of compulsory schooling on mortality: evidence from Sweden. Int J Environ Res Public Health 2013;10:3596–618.
30 Clark D, Royer H. The effect of education on adult mortality and health: evidence from Britain. Am Econ Rev 2013;103:2087–120. 31 Erikson R, Torssander J. Clerics die, doctors survive: a note on death risks among highly educated professionals. Scand J Public Health 2009;37:227–31. 32 Deaton A. The Great Escape: Health, Wealth, and the Origins of Inequality. Princeton: Princeton University Press, 2013.
26 James J. Health and Education Expansion. Bath, UK: Department of Economics, 2013.
......................................................................................................... European Journal of Public Health, Vol. 25, No. 4, 592–597 ß The Author 2014. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved. doi:10.1093/eurpub/cku213 Advance Access published on 19 December 2014
Cumulative contextual and individual disadvantages over the life course and adult functional somatic symptoms in Sweden Per E. Gustafsson1, Anne Hammarstro¨m1, Miguel San Sebastian2 1 Division of Social Medicine, Department of Public Health and Clinical Medicine, Umea˚ University, Umea˚, Sweden 2 Division of Epidemiology and Global Health, Department of Public Health and Clinical Medicine, Umea˚ University, Umea˚, Sweden Correspondence: Per E. Gustafsson, Division of Social Medicine, Department of Public Health and Clinical Medicine, Umea˚ University, SE-901 85 Umea˚, Sweden, Tel: +46 90 785 35 90, Fax: +46 90 776 883, e-mail: [email protected]
Background: Disadvantage, originating in one’s residential context or in one’s past life course, has been shown to impact on health in adulthood. There is however little research on the accumulated health impact of both neighbourhood and individual conditions over the life course. This study aims to examine whether the accumulation of contextual and individual disadvantages from adolescence to middle-age predicts functional somatic symptoms (FSS) in middle-age, taking baseline health into account. Methods: The sample is the age 16, 21, 30 and 42 surveys of the prospective Northern Swedish Cohort, with analytical sample size n = 910 (85% of the original cohort). FSS at age 16 and 42, and cumulative socioeconomic disadvantage, social adversity and material adversity between 16 and 42 years were operationalized from questionnaires, and cumulative neighbourhood disadvantage between 16 and 42 years from register data. Results: Results showed accumulation of disadvantages jointly explained 9–12% of FSS variance. In the total sample, cumulative neighbourhood and socioeconomic disadvantage significantly predicted FSS at age 42 in the total sample. In women, neighbourhood disadvantage but not socioeconomic disadvantage contributed significantly, whereas in men, socioeconomic but not neighbourhood disadvantage contributed significantly. In all analyses, associations were largely explained by the parallel accumulation of social and material adversities, but not by symptoms at baseline. Conclusion: In conclusion, the accumulation of diverse forms of disadvantages together plays an important role for somatic complaints in adulthood, independently of baseline health.
.........................................................................................................
Introduction oth physical and mental health are shaped by the social conditions
Bunder which people live their lives. Social determinants of health
are found not only in the direct environment of the home and work but also in the surrounding context, such as the neighbourhood.1 Moreover, not only one’s current social conditions matter for health but also circumstances from one’s past life course.2 Drawing on these two approaches to social determinants of health, the present report seeks to investigate whether accumulation of neighbourhood and individual disadvantages over the life course is related to selfreported health, in the form of functional somatic symptoms (FSS), in a middle-aged Swedish population.
The parallel observations that contextual factors may impact on concurrent health in both young people3,4 and adults,5 and that personal living conditions from earlier life may have long-term health effects,6 have in recent years converged into integrated research studying the health effects of contextual determinants from the past life course.7,8 A handful of studies suggest that the health impact of unfavourable neighbourhood conditions seems to accumulate over the life course,9,10 with less independent importance of context at specific life course period.11 This notion corresponds to what within life course research is called a ‘cumulative risk model’, where the total accumulation of adverse exposures across the life course corresponds to a greater risk for ill health in adulthood.12
Downloaded from http://eurpub.oxfordjournals.org/ by guest on November 15, 2015
.........................................................................................................
Contextual and individual disadvantages
Methods
FSS at age 16 and 42 The FSS measure covered symptoms within the areas of cardiopulmonary/autonomic, gastrointestinal, musculoskeletal and general symptoms, similar to other FSS measures.13 It was operationalized identically at age 16 and 42 as a summative score of 10 questionnaire items, each coded 0–2, and covering symptoms during the last 12 months. First, eight symptoms (headache or migraine; stomach ache other than gastritis or gastric ulcer; nausea; backache, hip pain or sciatica; fatigue; breathlessness; dizziness; overstrain) were each assessed by the question ‘Do you have (or have you during the last 12 months had) any of the following:’ with response options ‘No’ (0), ‘Yes, light’ (1) and ‘Yes, severe’ (2) for each symptom. Second, palpitations were measured by the question ‘How often have you had nervous problems during the past 12 months’, where participant indicated specific symptoms, including palpitations, with response options ‘never’ (0), ‘sometimes’ (1) and ‘always’ (2). Third, sleeplessness came from the question ‘Have you had sleeping difficulties during the past 12 months?’ with the response options coded as ‘Never’ (0), ‘Sometimes’ (1) and ‘often’ or ‘always’ (2). All 10 items were summed up into a score with range 0–20. Internal consistency for the FSS measure was Cronbach’s = 0.79 at age 42, and = 0.70 at age 16.
Sample and procedures
Cumulative ND over the life course
The Northern Swedish Cohort is based on all school-leavers of the ninth grade of the Swedish compulsory school (aged 16 years), in the municipality of Lulea˚, in 1981 (N = 1083). For details of the sample and procedures, see Refs. 26 and 29. For this report, the baseline survey at age 16 as well as follow-up surveys at age 21 (1986), 30 (1995) and 42 (2007) were used. At age 42, there were 1001 participants (93.4% of n = 1071 of the original cohort still alive), for whom n = 910 had complete data (85.0%). At each survey, participants completed a comprehensive selfadministered questionnaire about health and social circumstances, which was used to operationalize cumulative disadvantages and health. Data were also retrieved from Statistics Sweden to create aggregate measures of cumulative neighbourhood disadvantage (ND). Ethical approval was granted by the Regional Ethical Review Board in Umea˚.
Cumulative ND was operationalized in accordance with a previous report.10 The basis was register data on the neighbourhoods (SmallArea Market Statistics areas) which at least one participant was registered in, on 31 December 1980, 1986, 1995 and 2007. For all residents in these neighbourhoods, a number of variables were retrieved and used to operationalize ND. First, aggregate measures of ND were calculated at age 16, 21, 30 and 42. At each age, eight neighbourhood indicators were selected: percentages of residents with (i) ‘Low income’, (ii) ‘High income’ (reverse coded), (iii) ‘Housing allowance’, (iv) ‘Wealth’ (reverse coded), (v) ‘Non-employment’, (vi) ‘Single-parent household’, (viia) ‘Low occupational status’ (only available for 1981 and 1986), (viib) ‘Low educational attainment’ (only available for 1995 and 2007), (viiia) ‘High occupational status’ (reverse coded; only available for 1981 and 1986) and (viiib) ‘High educational achievement’ (reverse coded; only available for 1995 and 2007). All indicators were Z-transformed and the mean of all indicators was calculated, creating four age-specific ND scores. Internal consistency for the age-specific scores was Cronbach’s = 0.89 (1981); = 0.81 (1986); = 0.86 (1995); and = 0.89 (2007) at the neighbourhood level, and = 0.93 (1981); = 0.88 (1986); = 0.85 (1995); and = 0.86 (2007) at the individual level.
Measures Cumulative ND was based on register data corresponding to age 16, 21, 30 and 42, and all other measures were based on selfadministered questionnaires completed at the corresponding ages. See table 1 for descriptive statistics of all measures.
Table 1 Descriptive statistics of key variables in the total sample and by gender (N = 910; 438 women and 472 men) Variable
FSS at age 42, M(SD) FSS at age 16, M(SD) ND, M(SD) SED, n (%) 0 occasions 1 occasion 2 occasion 3 occasion 4 occasion Social adversity, M(SD) Material adversity, M(SD)
a: P-value from t-test. b: P-value from 2-test.
Total sample
By gender
Range
Estimate
Women
Men
P-value
0–18 0–16 2.02 to 2.65 0–4
4.21 (3.32) 3.34 (2.53) 0.01 (0.67)
4.78 (3.51) 3.71 (2.51) 0.05 (0.66)
3.69 (3.03) 3.01 (2.51) 0.03 (0.68)
