Inequalities in mortality according to educational level in two large Southern European cities

Transcription

1 International Epidemiological Association 1999 Printed in Great Britain International Journal of Epidemiology 1999;28:58 63 Inequalities in mortality according to educational level in two large Southern European cities Carme Borrell, a Enrique Regidor, b Luís-Carlos Arias, c Pedro Navarro, b Rosa Puigpinós, d Vicente Domínguez b and Antoni Plasència a,d Background In Spain, studies on social inequalities in mortality based on individuals are few due to the poor quality of information on occupation in death certificates. This study looks at the differences in mortality according to educational level, using individual information obtained through the linkage between the Death Register and the Municipal Census, in the cities of Madrid and Barcelona, Spain. Methods Results Conclusion Keywords Accepted 2 June 1998 The study populations were residents of Madrid and Barcelona aged 24 years, who died in 1993 and Indicators obtained for each city and educational level were: age- and sex-specific mortality rates, and life expectancy at 25 years. Poisson regression models were fitted to obtain the relative risk (RR) of death for each educational level with respect to the reference level (higher education completed), adjusted for age. The mortality rate was lower among individuals with higher educational levels, while life expectancy at 25 years was higher. In both cities men and women with no education showed the highest mortality in all age groups, with very high RR in the youngest age group (RR for men aged years = 7.08 in Madrid and 6.02 in Barcelona, whereas in women these RR were 6.33 and 5.63 respectively). In Barcelona the greater part of the overall mortality difference for the group aged years was due to AIDS (acquired deficiency syndrome, 33.4% in men and 59.3% in women). The present study has found higher mortality (mainly from AIDS) among individuals with no academic qualifications thus drawing attention to the need to implement policies aimed at reducing these inequalities. Mortality, social class, causes of death, social inequalities in health The study of social inequalities in health has been extensively developed in certain Northern European countries, 1 3 but hardly at all in Southern Europe. 4 6 In Spain, studies based on individuals are few, due to the absence, or poor quality, of information on socioeconomic characteristics in death certificates, and more restrictive legislation regarding the use of individual data than in Northern and Central European countries. 7,8 As a result, most investigations on this topic are based on ecological studies, 9 with a few cross-sectional studies in geographical areas where occupation is recorded on the death certificate with a reasonable degree of reliability. 10 a Municipal Institute of Health, Pl. Lesseps 1, Barcelona, Spain. b Departamento de Medicina Preventiva y Salud Pública, Facultad de Medicina, Universidad Complutense de Madrid, Spain. c Regió Sanitària Costa de Ponent, Institut Català de la Salut, Spain. d Institut de Salut Pública de Cataluña, Spain. The increasing concentration of the population in urban areas has brought important inequalities with some urban areas having considerable marginal population groups. 11 In Madrid and Barcelona, the two largest cities in Spain, ecological studies have detected inequalities in mortality in different city areas related to socioeconomic indicators The present study assesses mortality differences according to educational level using individual-based information in the cities of Madrid and Barcelona. Methods The study population consisted of deaths which occurred during among those aged 24 years who were resident in the cities of Madrid ( deaths) and Barcelona ( deaths). Madrid, the capital of Spain, is located in the centre of the country and has a population of inhabitants. 58

2 INEQUALITY IN MORTALITY AND EDUCATION 59 Table 1 Population aged 25 years, mortality rates per 1000 inhabitants and life expectancy at 25 years according to educational level, Madrid and Barcelona, Educational Mortality Life Mortality Life level Population rates expectancy Population rates expectancy Madrid No education Primary education Secondary or higher education Barcelona No education Primary education Secondary or higher education Barcelona, located in the north east, is the second largest city, with a population of inhabitants. Both cities are evolving from being industrial to predominantly tertiary areas. The population at risk, according to age, sex and the highest completed level of study comes from the 1991 Municipal Census, an administrative register not subject to statistical secrecy. Everyone in the Spanish population is registered in the municipality where they reside. The census information is revised every 5 years through the active collection of data by the statistical office of each municipality; between these major revisions, the information is continually updated to incorporate data on births and deaths. The educational level of deceased individuals has been obtained through record linkage between the Death Register and the Municipal Census. Furthermore, in Barcelona, cause of death has also been obtained, made possible by the fact that the city statistical office collaborates with the National Death Registry in the coding of causes on death certificates. The variables studied were age, sex, city of residence, educational level and cause of death, coded according to the International Classification of Diseases, 9th Revision. 15 Educational level was categorized as follows: illiterate or no education (0 4 years of schooling), primary education (5 11 years) and secondary or higher education ( 12 years). Information on educational level was not recorded in 1.5% of deaths in Madrid (n = 758), and in 9% of cases in Barcelona (n = 2908), due to problems of record linkage or missing data. Indicators obtained for each city were: age- and sex-specific mortality rates, and life expectancy at 25 years for each educational level obtained by the life-table method. 16 For Barcelona the results are also presented in terms of the main causes of death. Poisson regression models 17 were fitted using data grouped by age and educational level for each city and sex in order to obtain the relative risk (RR) of death for each educational level with respect to the reference level (secondary or higher education completed), adjusted for age. The dependent variable was the logarithm of the mortality rate, and the independent variables were age and educational level. An interaction term was introduced and found to be statistically significant, for which reason the results are presented by age group. For the Barcelona deaths in each age group, the RR of cases who were illiterate or had no education with respect to those with secondary or higher education were estimated for each cause of death, as well as the contribution of each cause of death to the excess mortality among individuals without schooling compared to those with secondary or higher education. Results In Madrid 29% of men and 40% of women aged 24 years had not completed any type of schooling. In Barcelona, these percentages were 18% and 27% respectively (Τable 1). Although in young people the proportion without schooling was smaller, the differences between the two cities persisted in all age groups. It may also be observed in Table 1 that, in both cities, the mortality rate was lower in individuals with higher educational level whilst their life expectancy at 25 years was greater. In both cities, men and women with no education had a higher mortality in all age groups, with the RR values being extremely high in the youngest age group (Τable 2). Specifically, men aged years with no education in Madrid had an RR of death of 7.08 (95% CI : ), and in Barcelona of 6.02 (95% CI : ). Among women the corresponding RR were 6.33 (95% CI : ) and 5.63 (95% CI : ), respectively. Over the age of 34 years the effect in those with no education attenuates with age. In individuals with primary education, the years age group also had excess mortality compared with those with secondary or higher education. In men, the magnitude of the RR of death was 3.43 (95% CI : ) in Madrid and 2.50 (95% CI : ) in Barcelona; in women they were 2.69 (95% CI : ) and 3.00 (95% CI : ), respectively. In the remaining age groups no statistically significant differences were found for women, except for the group aged years, where women with no education or only primary education showed higher mortality in Madrid and lower mortality in Barcelona than women with secondary or higher education. Similarly, the RR of death among men in Barcelona aged 44 years was not statistically significant. Table 3 presents the magnitude of the mortality effect by cause of death and the contributions of the main causes of death to the excess or deficit in mortality in individuals with no education in Barcelona. The RR were higher in the group aged years, the values being notable for AIDS (acquired immunodeficiency syndrome) in men (5.4) and in women (10.5), and for cirrhosis (76.1) and drug overdoses (9.3) in men. In the other age groups RR were lower, even demonstrating an

3 60 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY Table 2 Relative risks (RR) and 95% confidence intervals (CI) of dying according to educational level, Madrid and Barcelona, Secondary or higher education Primary education No education Sex and age group No. of deaths RR RR 95% CI RR 95% CI Madrid Barcelona inverse relationship between educational level and mortality due to breast cancer and lung cancer among women of all age groups. In the group aged years, the largest contribution to the difference in overall mortality was due to AIDS (33.4% in men and 59.3% in women). In the group aged years, among men the differences were mainly due to cancer, while in women AIDS was the main cause. In the group aged years the contribution of lung cancer among men is notable, as is that of cerebrovascular disease among women. After 65 years the causes which contributed in greatest proportion were the diseases of the respiratory system among men and cardiovascular diseases among women. Breast and lung cancer in women had a negative contribution in most age groups. Discussion This individual-based study describes the existence of inequalities in mortality according to educational level among the population aged 24 years in the two largest cities in Spain. This is one of the first studies of this type in Spain, made possible by the automated linkage between the death register and the municipal censuses of these two cities. 18 It has shown that the illiterate and those with no education have higher mortality and a lower life expectancy than people with secondary or a higher education. These differences are more important for the younger population than for the middle-aged and the elderly population, while some causes of death contribute more than others. The results obtained agree with those published in the literature, where the existence of higher mortality in population groups with lower educational levels has been described, most notably the studies in the US of Kitagawa and Hauser, 19 Feldman et al., 20 Orcutt, 21 and more recently that of Pappas et al. which describes the increase in these inequalities between 1960 and In Barcelona, AIDS deaths are mainly responsible for the differences in overall mortality by educational level in the group

4 INEQUALITY IN MORTALITY AND EDUCATION 61 Table 3 Number of deaths by cause and age group, relative risk of dying (RR) among individuals with no education with respect to those with secondary or higher education and contribution to excess mortality which each of the causes represents among individuals with no education with respect to those with secondary or higher education. Barcelona, No. of RR % a No. of RR % a No. of RR % a No. of RR % a Cause of death (ICD-9) deaths deaths deaths deaths Infectious diseases, including AIDS ( , 279.5) b AIDS (279.5) b Cancer ( ) b b b 28.0 Lung cancer(162) b b 9.7 Cardiovascular diseases ( ) b b Ischaemic heart disease ( ) Cerebrovascular disease ( ) b b Diseases of the respiratory system ( ) b b 43.9 Diseases of the digestive system ( ) b b Cirrhosis (571) b b External causes (E800 E999) b b Traffic accidents (E810 E829) Drug overdose (E850 E858) b b All causes ( ) b b b 100 Infectious diseases, including AIDS ( , 279.5) b b AIDS (279.5) b b Cancer ( ) b 12.9 Lung cancer (162) b b 2.7 Breast cancer (174) b b 4.6 Cardiovascular diseases ( ) b b 62.9 Ischaemic heart disease ( ) b 16.5 Cerebrovascular disease ( ) b b 18.8 Diseases of the respiratory system ( ) b b 9.1 Diseases of the digestive system ( ) b 10.2 Cirrhosis (571) External causes (E800 E999) b b 3.8 Traffic accidents (E810 E829) Drug overdose (E850 E858) All causes ( ) b b 100 a This percentage is calculated as follows: (mortality rate in individuals with no education mortality rate in individuals with secondary or higher education in the cause of death being dealt with)/difference between the total mortality rates for the two population groups. b The relative risk 95% confidence interval does not include 1. aged years. Although in Madrid it was not possible to obtain deaths by cause and educational level, the results would probably be similar. Two pieces of indirect evidence support this statement: firstly, in both cities RR estimates in this age group are very high, and secondly, a great increase in mortality due to AIDS and drug overdose in young adults has been detected in both cities during the second half of the 1980s and the beginning of the 1990s, such that they have become the primary causes of death In many developed societies drug abuse alone is the primary cause of death in young adults. 27 In Madrid and Barcelona drug abuse can also be considered, indirectly, the first cause of death between 25 and 34 years of age, since the majority of AIDS cases which appear are related with injected drug use, as is true for the majority of cases occurring in the rest of Spain. 25,28,29 Even so, the results from the present study suggest that this phenomenon affects mainly those with lower educational level. These results had been found by ecological studies in Barcelona, in which the existence of higher mortality

5 62 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY due to overdose and AIDS was described for wards with lower socioeconomic level. Further, in recent years inequalities between wards with low and high socioeconomic levels have increased, as shown by increased mortality in recent years. 30,31 AIDS and drug overdoses are also the primary causes of death among men aged years in Madrid and in Barcelona, although the RR of death is lower than in men aged years. Over 40 years of age, transmission of the human immunodeficiency virus (HIV) infection in the majority of AIDS cases in Barcelona arises as a consequence of homosexual relations, 28 the practice of which is more common among individuals with higher educational levels. Mortality due to AIDS and overdose has previously been related to socioeconomic level. 32,33 Wallace et al. have described how in areas of low socioeconomic level in large US cities, AIDS and drug overdose are widespread, and are related to the violence and urban decay which has occurred in many of these areas. 34,35 Over 45 years of age in men the contributions of lung cancer, respiratory system diseases, cerebrovascular disease and cirrhosis to inequalities are notable. Some of these diseases are related to tobacco consumption, which is higher among men of lower social classes. 36 In women, attention should be drawn to the negative contributions of breast cancer and lung cancer in the group aged years. It has been previously described that breast cancer is more common in women of higher socioeconomic groups, mainly due to reproductive factors. 37 With reference to lung cancer, it should be noted that currently in Barcelona it is still the women in higher social classes who smoke most, 36 a fact which parallels the situation in Spain as a whole and in other Southern European countries. 38 These negative contributions favour the reduction of RR among women with no education in the groups aged and years. The decline with age of the RR of death among individuals with no education reproduces another of the usual findings in studies of this topic. Various hypotheses have been proposed to explain these results. One is the selective mortality during middle age of individuals with poorer health in lower socioeconomic strata. 39,40 Another explanation is the acquisition with age of increased control over life circumstances which tends to diminish the harmful effects of exposure to adverse environmental factors. 41 Finally, the fact of having no schooling denotes a much worse socioeconomic situation in young people than in older people, since nowadays access to basic education is far greater. 42 It should be mentioned that in Barcelona, although the educational level was not recorded for 9% of cases, these appeared to be randomly distributed among all the districts, both rich and poor, suggesting that there was no evidence of any bias by educational level. In summary, the present study has found higher mortality (mainly from AIDS and drug overdose) among individuals with no academic qualifications thus drawing attention to the need to implement policies aimed at reducing these inequalities. Acknowledgements The authors would like to thank Enrique Montoliu of the Statistics Department of the Madrid City Council for having facilitated data relating to the city of Madrid. References 1 Townsend P, Davidson N, Whitehead M. Inequalities in Health: The Black Report and the Health Divide. London: Penguin Books, Syme SL, Berkman LF. Social class, susceptibility and sickness. Am J Epidemiol 1976;104: Feinsten JS. The relationship between socioeconomic status and health. A review of the literature. Milbank Q 1993;71: Ramis-Juan O, Sokou K. Social health inequalities in South European countries: is it a different problem? In: Fox J (ed.). Health Inequalities in European Countries. Aldershot: European Science Foundation. Gower, Regidor E, Gutiérrez-Fisac JL, Rodríguez C. Diferencias y desigualdades en salud en España. Madrid: Díaz de Santos, Navarro V, Benach J y la Comisión científica de estudios de las desigualdades sociales en salud en España. Desigualdades Sociales en Salud en España. Madrid: Ministerio de Sanidad y Consumo y The School of Hygiene and Public Health, The Johns Hopkins University, Valkonen T. Problems in the measurement and international comparisons of socio-economic differences in mortality. Soc Sci Med 1993; 36: Informe de un grupo de trabajo de la Sociedad Española de Epidemiología. La Medición de la Clase Social en Ciencias de la Salud. Barcelona: SG Editores, Rodríguez JA, Lemkow L. Health and social inequities in Spain. Soc Sci Med 1990;31: Regidor E, Gutiérrez-Fisac JL, Rodríguez C. Increased socioeconomic differences in mortality in eight Spanish provinces. Soc Sci Med 1995; 41: Flynn P. Medición de la salud en las ciudades. In: Ashton J (ed.). Ciudades Sanas. Barcelona: Masson, SA, Pérez-Dominguez FJ, Gallardo V, García M, Ruiz R. Análisis de la mortalidad en el sector 7 de la Comunidad de Madrid durante Atención Primaria 1993;12: Arias A, Rebagliato M, Palumbo MA et al. Desigualdades en salud en Barcelona y Valencia. Med Clin (Barc) 1993;100: Borrell C, Arias A. Socio-economic factors and mortality in urban settings: the case of Barcelona (Spain). J Epidemiol Community Health 1995;49: OPS/OMS. Clasificación Internacional de Enfermedades, 9ª Revisión. Washington: OMS, Shyrock HS, Siegel JS. The Methods and Materials of Demography. New York: Academic Press, Breslow NE, Day NE. Statistical Methods in Cancer Research. Vol. II: The Design and Analysis of Cohort Studies. Lyon: International Agency for Research in Cancer, Arias LC, Borrell C. Desigualdades en la mortalidad según la educación en la ciudad de Barcelona. Med Clin (Barc) 1998;110: Kitagawa EM, Hauser PM. Differential Mortality in the United States: A Study in Socioeconomic Epidemiology. Cambridge: Harvard University Press, Feldman JJ, Makuc DM, Kleinman JC, Cornoni-Huntley J. National trends in educational differentials in mortality. Am J Epidemiol 1989;129: Orcutt H. Measuring socioeconomic mortality differentials over time. Demography 1989;26: Pappas G, Queen S, Haddem W, Fisher G. The increasing disparity in mortality between socioeconomic groups in the United States, 1960 and N Engl J Med 1993;329: Montellà N, Borrell C, Brugal MT, Plasència A. Evolución de la mortalidad en los jóvenes de la ciudad de Barcelona: Med Clin (Barc) 1997;108:

6 INEQUALITY IN MORTALITY AND EDUCATION De la Fuente L, Barrio G, Vicente J, Bravo MJ, Santacreu J. The impact of drug-related deaths on mortality among young adults in Madrid. Am J Public Health 1995;85: Noguer I, Castilla J, Gutiérrez MA et al. Epidemiología del Sida en España. Madrid: Ministerio de Sanidad y Consumo, Sánchez J, Rodríguez B, De La Fuente L et al. Opiates or cocaine: mortality from acute reactions in six major Spanish cities. J Epidemiol Community Health 1995;49: Selwyn PA. Injection drug use mortality, and AIDS Epidemic (editorial). Am J Public Health 1991;81: García-Olalla P, Caylà JA, Jansà JM et al. Sida a Barcelona. Vigilància Epidemiològica. Barcelona: Institut Municipal de la Salut, European Centre for the Epidemiological Monitoring of AIDS: AIDS surveillance in Europe. Quarterly Rep 1996;49: Torralba L, Brugal MT, Villalbí JR, Tortosa MT, Toribio A, Valverde JL. Mortality due to acute adverse drug reactions: opiates and cocaine in Barcelona, Addiction 1996;91: Borrell C, Plasència A, Pasarín I, Ortun V. Widening social inequalities in mortality: the case of a southern European city (Barcelona). J Epidemiol Community Health 1997;51: Simon PA, Hu DJ, Kerndt PR. Income and AIDS rates in Los Angeles County. AIDS 1995;9: Wilkinson J, Lawes G, Unell I, Bradbury J, Maclean P. Problematic drug use and social deprivation. Public Health 1987;101: Wallace R, Fullilove M, Fullilove R, Gloud P, Wallace D. Will AIDS be contained within US minority populations? Soc Sci Med 1994;39: Wallace R, Wallace D. Inner-city disease and the public health of the suburbs: the sociogeographic dispersion of point-source infection. Env A Planning 1993; Nebot M, Borrell C, Ballestín M, Villalbí JR. Prevalencia y características asociadas al consumo de tabaco en población general en Barcelona entre 1983 y Rev Clin Española 1996;196: Kelsey Jl, Horn-Ross PL. Breast cancer: magnitude of the problem and descriptive epidemiology. Epidemiol Rev 1993;15: Graham I. Smoking prevalence among women in the European Community, Soc Sci Med 1996;43: House JS, Kessler RC, Herzog R et al. Age, socioeconomic status and health. Milbank Q 1990;68: Jefferys M. Social inequalities in health. Do they diminish with age (editorial)? Am J Public Health 1996;86: Thoits PA. Stress, coping and social support process: where are we? what next? J Health Soc Behav 1995;80:53 79 (extra issue). 42 Reijneveld SA, Gunning-Schepers LJ. Age, health and the measurement of the socio-economic status of individuals. Eur J Public Health 1995;5: