92 European Journal of Public Health nativity and country of origin. Ethn Heal 2017;0:1–21. Available at: https://doi.org/ non-migrants in both the countries of origin and destination. Int Health 10.1080/13557858.2017.1346785. 2018;10:258–67. Doi:10.1093/inthealth/ihy026. 35 Verstraeten SPA, van den Brink CL, Mackenbach JP, van Oers HAM. The health 36 Hickling FW, Rodgers-Johnson P. The incidence of first contact schizophrenia in of Antillean migrants in the Netherlands: a comparison with the health of Jamaica. Br J Psychiatry 1995;167:193–6. ......................................................................................................... The European Journal of Public Health, Vol. 30, No. 1, 92–98  The Author(s) 2019. Published by Oxford University Press on behalf of the European Public Health Association. This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs licence (http:// creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial reproduction and distribution of the work, in any medium, provided the original work is not altered or transformed in any way, and that the work is properly cited. For commercial re-use, please contact journals.permissions@oup.com doi:10.1093/eurpub/ckz125 Advance Access published on 13 June 2019 ......................................................................................................... Socioeconomic inequalities in suicide mortality in Downloaded from https://academic.oup.com/eurpub/article/30/1/92/5549577 by guest on 10 October 2023 European urban areas before and during the economic recession Carme Borrell1,2,3,4, Laia Palència2,3,5, Marc Marı́ Dell’Olmo2,3,6, Joana Morrisson7, Patrick Deboosere8, Mercè Gotsens3,9, Dagmar Dzurova10, Claudia Costa11, Michala Lustigova10, Bo Burstrom12, Maica Rodrı́guez-Sanz2,3,4,13, Lucia Bosakova14,15, Nicolas Zengarini16, Klea Katsouyanni17,18, Paula Santana11 1 Agència de Salut Pública de Barcelona, Barcelona, Spain 2 CIBER Epidemiologı́a y Salud Pública (CIBERESP), Madrid, Spain 3 Institut d’Investigació Biomèdica (IIB Sant Pau), Barcelona, Spain 4 Department of Experimental and Health Sciences, Universitat Pompeu Fabra, Barcelona, Spain 5 Service of Health Information Systems, Agència de Salut Pública de Barcelona, Barcelona, Spain 6 Service of Environmental Quality and Intervention, Agència de Salut Pública de Barcelona, Barcelona, Spain 7 Institute of Health Equity at the Research Department of Epidemiology and Public Health, University College London, London, UK 8 Department of Sociology, Interface Demography, Vrije Universiteit Brussel, Brussels, Belgium 9 Service of Prevention and Attention of Drug Use, Agència de Salut Pública de Barcelona, Barcelona, Spain 10 Department of Social Geography and Regional Development, Faculty of Science, Charles University, Prague, Czechia 11 Department of Geography and Tourism, Centre of Studies in Geography and Spatial Planning (CEGOT), University of Coimbra, Coimbra, Portugal 12 Department of Public Health Sciences, Karolinska Institutet, Stockholm, Sweden 13 Unit of Research, Training and Communication, Agència de Salut Pública de Barcelona, Barcelona, Spain 14 Department of Health Psychology, Medical Faculty, P. J. Safarik University in Kosice, Kosice, Slovak Republic 15 Olomouc University Social Health Institute (OUSHI), Palacky University in Olomouc, Olomouc, Czech Republic 16 Department of Public Health, University of Turin, Turin, Italy 17 Department of Hygiene, Epidemiology and Medical Statistics, Medical School, National and Kapodistrian, University of Athens Medical School, Athens, Greece 18 Department of Population Health Sciences, and Department of Analytical, Environmental & Forensic Sciences, School of Population Health & Environmental Sciences, King’s College, London, UK Correspondence: Carme Borrell, Agència de Salut Pública de Barcelona, Plaça Lesseps 1, 08023 Barcelona, Spain, Tel: +34 93 238 4545, Fax: 34 93 2173197, e-mail: cborrell@aspb.cat Background: Few studies have assessed the impact of the financial crisis on inequalities in suicide mortality in European urban areas. The objective of the study was to analyse the trend in area socioeconomic inequalities in suicide mortality in nine European urban areas before and after the beginning of the financial crisis. Methods: This ecological study of trends was based on three periods, two before the economic crisis (2000–2003, 2004–2008) and one during the crisis (2009–2014). The units of analysis were the small areas of nine European cities or metropol- itan areas, with a median population ranging from 271 (Turin) to 193 630 (Berlin). For each small area and sex, we analysed smoothed standardized mortality ratios of suicide mortality and their relationship with a socioeconomic deprivation index using a hierarchical Bayesian model. Results: Among men, the relative risk (RR) comparing suicide mortality of the 95th percentile value of socioeconomic deprivation (severe deprivation) to its 5th percentile value (low deprivation) were higher than 1 in Stockholm and Lisbon in the three periods. In Barcelona, the RR was 2.06 (95% credible interval: 1.24–3.21) in the first period, decreasing in the other periods. No significant changes were observed across the periods. Among women, a positive significant association was identified only in Stockholm (RR around 2 in the three periods). There were no significant changes across the periods except in London with a RR of 0.49 (95% CI: 0.35–0.68) in the third period. Conclusions: Area socioeconomic inequalities in suicide mortality did not change significantly after the onset of the crisis in the areas studied. ......................................................................................................... Socioeconomic inequalities in suicide mortality 93 Introduction members). The study population consisted of the individuals residents in these areas during the different periods. ne of the most important causes of premature mortality and O mortality among the young population in Europe is suicide. Data Eurostat collects approximately 60 000 suicides yearly for the European Union (EU). Moreover, suicide mortality is higher in The data used in the present study were the following: populations with a low socioeconomic position.1–6 A systematic review analysing area socioeconomic disadvantage and suicide (1) Mortality, corresponding to the number of deaths due to suicide behaviour in Europe described a significant association between and intentional self-harm (International Classification of Diseases-ICD-9n revision: E950–E959, ICD 10: X60–X84) by socioeconomic disadvantage and suicide behaviour.7 The geograph- area, age-group and sex. Mortality data were obtained mainly ical units of analysis in the studies reviewed ranged from small from mortality registries. neighbourhoods to large cities but only a small number of studies (2) Population, corresponding to the number of inhabitants living focussed on the impact on urban areas.7–11 in the area. Most of the urban areas had population data for the During the financial crisis that started in 2008, suicide mortality whole period or for at least 2 years. Population data were increased in many countries, including European countries, stratified by age (5-year groups) and sex and were obtained a finding that has been demonstrated in several systematic Downloaded from https://academic.oup.com/eurpub/article/30/1/92/5549577 by guest on 10 October 2023 from census or population registries. reviews.12–14 However, the impact of the financial crisis on (3) Socioeconomic data, obtained from the census, consisting of several socioeconomic inequalities in suicide mortality has been less well indicators selected to identify the level of deprivation of the areas. analysed, with discrepant results, and has been even less in urban areas where the majority of the population lives.15–18 Measures Therefore, the objective of this study was to analyse the trend in area socioeconomic inequalities in suicide mortality in nine The suicide mortality indicator used for the analysis was the European urban areas in two periods: before (2000–2003, 2004– standardized mortality ratio (SMR). For descriptive purposes, we 2008) and after (2009–2014), the beginning of the financial crisis. calculated the indirectly standardized rate, calculated by multiplying Our hypothesis was that these inequalities increased in the cities and the SMR for the city/metropolitan area by the crude rate in the metropolitan areas analysed during the economic recession because standard population which was the population of the EU-28 in people living in areas with higher socioeconomic deprivation tend to the year 2007. be those most affected by the crisis,13,19 especially in the countries We built a composite socioeconomic deprivation index most negatively impacted by the crisis. using principal component analysis within each city. This analysis included the following variables for 2001 (or 2002 in the case of Berlin): the percentage of unemployment (16 years, economically active population), the percentage of manual Methods workers (16 years), the percentage of people with primary Design, units of analysis and study population education as the highest attainment [International Standard Classification of Education (ISCED) 0 and 1, except for London, This ecological study of trends was based on three periods, two which was ISCED 0, 1, 2] (25–64 years) and the percentage of periods before the economic crisis and one during the crisis (2000– people with university education (25–64 years). The percentage 2003, 2004–2008 and 2009–2014). The units of analysis were the small of manual workers was not available for Stockholm and the areas of nine European urban areas (Athens metropolitan area, index was performed with the other three indicators. The index Barcelona city, Berlin-Brandenburg metropolitan region, Brussels was the first component of the principal component analysis metropolitan area, Lisbon metropolitan area, Greater London, performed. Prague city, Stockholm metropolitan area and Turin city). The median population of the areas ranged from 271 (Turin) to 193 630 (Berlin) (table 1). These urban areas were selected by the Data analysis EURO-HEALTHY project (a project funded by Horizon2020), to The SMR is dependent on population size since its variance is cover different areas within the EU with a diverse cultural, geograph- inversely proportional to the expected values; thus areas with a ical and economic background (http://www.euro-healthy.eu/ small population size tend to show highly unstable estimates. To Table 1 Description of the nine European urban areas: number and type of small areas, population year and first, second and third quartiles of the population by small area, for men and women Urban area Name Number of Type of Population used small areas small areas Year Men Women Total p25 p50 p75 Total p25 p50 p75 Athens metropolitan area Athens 40 Municipalities 2001 1 577 172 18 565 29 745 35 489 1 710 446 20 136 32 163 39 965 Barcelona city Barcelona 1491 Census tracts 2000 697 563 365 457 577 796 497 418 517 648 Berlin-Brandenburg metropolitan Berlin 30 Districts 2002 2 927 616 66 326 96 176 129 157 3 047 188 68 041 97 454 130 560 region Brussels metropolitan area Brussels 145 Neighbourhoods 2001 464 364 2727 4004 5707 505 673 3058 4288 6172 Lisbon metropolitan area Lisbon 188 Parishes 2001 1 275 813 2694 5437 8962 1 386 314 2938 5835 9904 Greater London London 983 Census tracts 2001 3 597 120 3442 3810 4284 3 725 283 3526 3960 4382 Prague city Prague 57 Districts 2001 549 652 1010 2206 15 001 610 466 1024 2100 14 838 Stockholm metropolitan area Stockholm 1299 Census tracts 2001 897 487 218 560 1050 936 977 232 599 1104 Turin city Turin 2678 Census tracts 2000 425 782 88 129 196 465 987 96 142 215 94 European Journal of Public Health smooth the SMR, we used the hierarchical Bayesian model proposed analysed by small areas with a median population of less than 6000 by Besag, York and Mollié.20 This model takes two types of random for men and women. The areas of analysis were districts for Berlin effects into account, spatial and heterogeneous: the former takes (median population of approximately 100 000 for men and women) account of the spatial structure of the data while the latter deals and municipalities for Athens (median population of approximately with non-structural (non-spatial) variability. We estimated 30 000 for men and women). smoothed SMR (sSMR) for both sexes and periods using the The socioeconomic indicators used to build the socioeconomic following model: deprivation index for each small area are presented in the Supplementary table. Variations in some indicators between the Oi  PoissonðEi i Þ cities included in the analyses were large, for example, the median logði Þ ¼  þ Si þ Hi ðmodel1Þ percentage of the population with primary education as their highest attainment ranged from 37.6% (Lisbon) to 0.9% (Berlin). Of note, where, for each area i, Oi is the number of observed cases, Ei the the figure for London (54.0%) was not strictly comparable, as expected cases, i, the sSMR, with respect to the European explained in the Methods section due to having included population, Si the spatial effect and Hi the heterogeneous effect. ISCED 0–2. The table also shows the correlation coefficients Expected cases were calculated by indirect standardization taking between the indicators and the socioeconomic deprivation index. the suicide mortality rates of the EU-28 in 2007 as reference The highest correlations were observed with university educa- Downloaded from https://academic.oup.com/eurpub/article/30/1/92/5549577 by guest on 10 October 2023 (using the year approximately in the middle of the period), by age tion followed by manual workers in most of the cities/metropolitan (using 5-year groups). areas. In addition, to analyse the trend in socioeconomic inequalities, we The description of mortality data is presented in table 2. The first fitted an ecological regression model, which included the period included 4 years (2000–2003), and the second and the third socioeconomic deprivation index (D), the period (through two 5 years (2004–2008 and 2009–2013) for most of the cities. Suicide dummy variables P2 and P3) and their interaction: mortality rates (crude and age-adjusted) were higher among men Oit  PoissonðEit it Þ than women. Suicide mortality rates were lower in Athens, Barcelona and London than in other cities. During the crisis, suicide death logðit Þ ¼  þ 1 Di þ 2 P2t þ 3 P3t rates increased mainly in Athens: among men, the age-standardized þ4 P2t Di þ 5 P3t Di þ Sit þ Hit ðmodel2Þ mortality rate increased from 4.37 per 100 000 inhabitants in the second period to 7.07 in the third. These rates among women were where, for each area i and period t (t = 1 for the first period, t = 2 for 1.03 and 1.24, respectively. the second period and t = 3 for the third period), Oit is the number of The RR of association between the socioeconomic deprivation observed cases, Eit the expected cases, it the sSMR with respect to the index and suicide mortality for the three periods for men and European population, Sit the spatial effect and Hit the heterogeneous women and each city or metropolitan area are shown in figures 1 effect. Finally, P2t and P3t took the following values: Pjt = 1 if j = t, and and 2. Among men, RR were higher than 1 in Stockholm and Lisbon 0 otherwise. The expected cases were calculated as in the previous in the three periods. In Barcelona, the RR was 2.06 (95% CI: 1.24– model. Changes between periods in the relationship between 3.21) in the first period, decreasing in the other periods. For the socioeconomic deprivation index and mortality were evaluated other cities/metropolitan areas, there were no significant associ- through the interactions included in model 2. Specifically, we ations. No significant changes were observed across the periods. studied the change between the first and second period (4) and Among women, a positive significant association was identified the second and third period (5  4). only in Stockholm (RR around 2 in the three periods). There were In the two models (models 1 and 2), an intrinsic conditional no significant changes during the periods except in London with an autoregressive prior distribution was assigned to the spatial effect, RR lower than 1 in the third period (RR 0.49, 95% CI: 0.35–0.68) which assumed that the expected value of each area coincided with which implies that the lower the socioeconomic deprivation of the the mean of the spatial effect of the adjacent areas and had a variance area the higher the risk of suicide mortality. of  s2, while the heterogeneous effect was represented using inde- The associations between the socioeconomic deprivation index pendent normal distributions with mean 0 and variance  h2. A and suicide sSMR can also be observed through maps. The uniform distribution U(0,1) was assigned to the standard Supplementary figure illustrates the example of men in Lisbon, deviations  s and  h. A normal vague prior distribution was showing similar geographical patterns for the socioeconomic depriv- assigned to the parameters , 1, 2, 3, 4 and 5. ation index and suicide sSMR. The other example illustrates the As the socioeconomic deprivation index scale is dimensionless maps for women in London, where the pattern shows an inverse and arbitrarily fixed, for each city/metropolitan area, we relationship. calculated the relative risk (RR) that compares suicide mortality of the 95th percentile value of socioeconomic deprivation (severe deprivation) to its 5th percentile value (low deprivation). RR Discussion estimates were obtained based on the mean of their subsequent distribution, along with corresponding 95% credible intervals This study shows that suicide mortality rates were higher for men (95% CIs). than for women, varied among urban areas, and increased signifi- To support the analysis of suicide sSMR, and the socioeconomic cantly only in Athens during the financial crisis. Suicide mortality deprivation index, these variables were represented on maps using a was higher in deprived areas of Stockholm for men and women and classification based on septiles. in London and Lisbon for men. Socioeconomic inequalities did not All analyses were performed using the INLA library of the R stat- change significantly in the three periods (with the exception of istical package. London among women where the RR decreased in the third period showing an inverse relationship between socioeconomic de- Results privation and suicide mortality). As shown by previous studies,6,9 in our study suicide rates differed The characteristics of the nine areas studied, some of which are cities in cities/metropolitan areas. Moreover, death rates were higher for and others metropolitan areas, are shown in table 1. By population men than for women. Previous publications have reported that men size, the largest is Greater London (7 322 403 inhabitants) and the make fewer suicide attempts than women but are more likely to be smallest Turin (891 769 inhabitants). Except for Athens metropol- successful.5,21 This could be explained by traditional gender roles: in itan area and Berlin, all the other cities and metropolitan areas were men a higher level of strength, risk taking behaviour, independence, Socioeconomic inequalities in suicide mortality 95 Table 2 Description of suicide mortality in nine urban areas: years, number of deaths and crude mortality rate (MR) and indirectly age-standardized mortality rate (ISR) by 100 000 inhabitants, for men and women in each study period First period Second period Third period Years Deaths Crude MR ISR Years Deaths Crude MR ISR Years Deaths Crude MR ISR Men Athens 2000–2003 272 4.34 4.55 2004–2008 323 4.26 4.37 2009–2013 521 7.05 7.07 Barcelona 2000–2003 264 9.17 9.08 2004–2008 408 10.75 10.60 2009–2013 378 9.82 9.60 Berlin 2002 582 19.88 20.15 2006 532 17.59 17.72 2011 469 16.00 15.81 Brussels 2001–2003 249 17.60 19.44 2004–2008 519 20.59 22.88 2009–2011 257 15.80 17.65 Lisbon 2000–2003 732 14.33 15.13 2004–2008 979 15.16 15.82 2009–2012 839 15.81 16.31 London 2000–2003 1097 7.58 8.65 2004–2008 1362 7.27 8.27 2009–2014 1682 6.88 7.80 Prague 2001–2003 418 25.23 25.55 2004–2008 574 19.92 20.04 2009–2014 796 21.86 22.01 Stockholm 2001–2003 449 16.56 17.92 2004–2008 706 14.99 16.22 2009–2011 418 14.41 15.59 Turin 2000–2003 183 10.71 10.29 2004–2008 231 10.81 10.28 2009–2013 271 12.56 11.84 Women Downloaded from https://academic.oup.com/eurpub/article/30/1/92/5549577 by guest on 10 October 2023 Athens 2000–2003 53 0.78 0.78 2004–2008 87 1.05 1.03 2009–2013 106 1.31 1.24 Barcelona 2000–2003 139 4.27 3.95 2004–2008 190 4.51 4.17 2009–2013 182 4.28 3.94 Berlin 2002 239 7.84 7.57 2006 194 6.23 6.01 2011 152 4.98 4.73 Brussels 2001–2003 136 8.86 9.17 2004–2008 277 10.27 10.78 2009–2011 138 8.04 8.55 Lisbon 2000–2003 233 4.19 4.18 2004–2008 382 5.37 5.29 2009–2012 265 4.49 4.37 London 2000–2003 439 2.94 3.21 2004–2008 479 2.48 2.72 2009–2014 535 2.14 2.35 Prague 2001–2003 140 7.65 7.21 2004–2008 201 6.47 6.09 2009–2014 216 5.60 5.33 Stockholm 2001–2003 224 7.93 8.23 2004–2008 360 7.39 7.70 2009–2011 214 7.19 7.53 Turin 2000–2003 95 5.08 4.62 2004–2008 90 3.84 3.47 2009–2013 98 4.13 3.69 Figure 1 Association between socioeconomic deprivation index and suicide mortality, relative risk (RR) and 95% credible intervals for men in nine urban areas. Notes: RR compares suicide mortality of the 95th percentile value of socioeconomic deprivation (severe deprivation) to its 5th percentile value (low deprivation). RR1: RR of first period; RR2: RR of second period; RR3: RR of third period. economic status, etc. are often stressed. As ‘bread-winners’, they increases the risk of suicidal thinking.22,23 Moreover, men more tend to feel a high degree of pressure and stress to maintain the often use highly lethal methods in suicidal behaviour and their family’s economic status. Moreover, reinforcement of these trad- suicide attempts are more serious, independently of the method itional gender roles often prevent men from seeking help when used, suggesting gender differences in intentionality associated they feel depressed or suicidal. Therefore, dominant masculinity with suicidal behaviour.24 96 European Journal of Public Health Downloaded from https://academic.oup.com/eurpub/article/30/1/92/5549577 by guest on 10 October 2023 Figure 2 Association between socioeconomic deprivation index and suicide mortality, relative risk (RR) and 95% credible intervals for women in nine urban areas. Notes: RR compares suicide mortality of the 95th percentile value of socioeconomic deprivation (severe deprivation) to its 5th percentile value (low deprivation). RR1: RR of first period; RR2: RR of second period; RR3: RR of third period. + indicates that RR3 is statistically significantly different from RR2 (no RR2 is statistically significantly different from RR1) Death rates from suicide tended to increase mainly in Athens for found in the pre-crisis period are similar to those described by both men and women after the start of the financial crisis but not in Gotsens et al.9 analysing area deprivation inequalities in several other settings; these results are consistent with those of other studies external causes of death (including suicide) in 15 cities/metropolitan conducted in Greece.25,26 This finding could be connected to the areas of Europe (7 of them are included in our study). However, a severity of the financial crisis in Greece compared with the rest of review by Cairns et al.7 found a significant association (in 25/27 EU, although other southern European countries were also strongly studies) between area socioeconomic disadvantage and suicidal affected by the financial crisis (http://ec.europa.eu/eurostat/ behaviour in Europe. The majority of studies of this review were statistics-explained/index.php/Unemployment_statistics#Recent_ performed in the UK and 20 of them analysed completed suicides. developments). Area-level deprivation had a stronger influence on suicide among In view of other studies published on the topic, our finding of no men than women. Although most of the studies were not centred on increase in suicide mortality in the other cities/metropolitan areas urban areas, several pathways are discussed by the authors to was unexpected. Previous reviews have shown that during past understand the associations:7 some studies explain their results in recessions12 and the recession of 2008,13,14 most studies reported terms of compositional factors because the associations disappeared an increase in suicide mortality, mainly among men of working when individual factors were adjusted; other studies found age and the unemployed. A review by Karanikolos et al.13 found contextual factors explaining the associations found, such as social 28 studies that focussed on mental health in different countries fragmentation, city size (associations were strongest in large cities), and all the studies reported worsening in a least one mental health rurality or regional unemployment levels; other studies found an indicator during the crisis. However, most of these studies focussed independent effect of area-level disadvantage on suicide on countries.27 Our study is based on urban areas and suicide rates behaviours, after analysing individual and contextual factors. are often higher in rural than in urban areas.28,29 Like the present Our results are contrary to our previous hypothesis. In some study, a previous study analysing trends in socioeconomic countries, the economic recession resulted in increased debt, inequalities in suicide mortality in Barcelona and the Basque reduced public spending, job loss and consequently high unemploy- Country, Spain, found no increase in suicide mortality.18 ment and sometimes an inability to afford housing costs followed by This study found minor results concerning socioeconomic eviction;30 these factors are related to poor health outcomes, inequalities in suicide mortality and the trend in these inequalities: including poor mental health.15 A possible explanation for our only two metropolitan areas showed socioeconomic inequalities results could be that the impact of the crisis on suicide behaviour among men and one among women and the trend did not change also affected the most privileged areas, as reported by studies across the periods studied. We also found stronger associations performed in Japan16,17 and South Korea31 where, following the among men than women, but for most cities/metropolitan areas 2008 economic downturn, suicide mortality in men increased the RR had a 95% CI that included the value of 1. The results more markedly among those who held managerial and professional Socioeconomic inequalities in suicide mortality 97 posts. Also, a study done in England and Wales did not find an Acknowledgements increase in socioeconomic inequalities in suicide mortality between 2001 and 2011.15 We thank Sani Dimitroulopoulou and Christina Mitsakou of Department of Environmental Hazards and Emergencies, Public Health England, UK for providing the UK data. Strengths and limitations The results show the pattern of socioeconomic inequalities in suicide Funding mortality in the cities/metropolitan areas and their changes over This study is a part of the EURO-HEALTHY project (Shaping time. These results are important because few studies have shown EUROpean policies to promote HEALTH equity) and has received how these inequalities have changed in urban areas. But the results funding from the European Union’s Horizon 2020 research and of the study should be interpreted with some limitations in mind. innovation programme under Grant Agreement No 643398. Our study focussed on differences in suicide mortality between Michala Lustigova was supported by Charles University (UNCE/ areas with different levels of deprivation based on the level of un- HUM 018, too). employment, proportion of manual workers, the number of residents with primary education and those with university Conflicts of interest: None declared. Downloaded from https://academic.oup.com/eurpub/article/30/1/92/5549577 by guest on 10 October 2023 education. These indicators of deprivation may have different meanings in the cities studied. Moreover, it is necessary to take into account that data for socioeconomic indicators were from Ethical approval 2001 and the deprivation of areas could have changed, however, This manuscript uses only secondary data, for this reason we do not probably the rank of areas by socioeconomic deprivation is enclose ethical approval. maintained. In addition, the number of areas per city varied considerably, which may have affected the robustness of estimates of suicide Key points mortality on area level (small number of deaths) and the level of differentials observed within each city. For some cities (Athens and  During the financial crisis that started in 2008, suicide Berlin) the number of areas was very small, a larger number of areas mortality increased in many countries. with a smaller population size would lead to a higher statistical  The study of the impact of the financial crisis on power and probably a greater gradient of difference across areas. socioeconomic inequalities in suicide mortality has been We were unable to include migration, a factor that is important less well analysed, with discrepant results. for suicide mortality. Suicide risk among immigrant populations  Suicide mortality was higher in deprived areas of Stockholm also depends on cultural factors, and on the suicide risk in the for men and women and in London and Lisbon for men. countries of origin and for these reasons suicide death rates may  Socioeconomic inequalities in suicide mortality did not change by country of origin.32 change significantly in the three periods. Although our study includes a trend analysis of 13–14 years, it  It is important to emphasise the importance of implement- would be interesting to study longer trends in order to be able to ing policies to tackle suicide mortality and inequalities in assess changes in those trends for a greater number of years. suicide. However, this potential limitation is partially overcome by the inclusion of two periods before the crisis. Finally, the quality of injury mortality statistics, including suicide, References may change in the different countries due to various factors, such as those described by Värnik et al.:33 routines for registering suicides, 1 Li Z, Page A, Martin G, Taylor R. Attributable risk of psychiatric and socio- the resources available for medical and legal inquests into causes of economic factors for suicide from individual-level, population-based studies: a death, adherence to regulations referring to causes of death, varying systematic review. Soc Sci Med 2011;72:608–16. proportions of suicide methods, cultural aspects, stigma associated 2 Milner A, Spittal MJ, Pirkis J, LaMontagne AD. Suicide by occupation: systematic with suicide, financial implication for relatives and the sociopolitical review and meta-analysis. Br J Psychiatry 2013;203:409–16. situation. However, we believe that these differences were unlikely to 3 Knipe DW, Carroll R, Thomas KH, et al. Association of socio-economic position have affected the inequalities found in the different cities/metropol- and suicide/attempted suicide in low and middle income countries in South and itan areas studied. South-East Asia—a systematic review. BMC Public Health 2015;15:1055. 4 Blakely TA, Collings SCD, Atkinson J. Unemployment and suicide. Evidence for a causal association? J Epidemiol Community Health 2003;57:594–600. Conclusion 5 Batty GD, Kivimäki M, Bell S, et al. Psychosocial characteristics as potential predictors of suicide in adults: an overview of the evidence with new results from The main conclusion of this study is that area socioeconomic prospective cohort studies/692/699/631/477/2811 review-article. Transl Psychiatry inequalities in suicide mortality did not change significantly after [Internet] 2018;8: doi:10.1038/s41398-017-0072-8. the onset of the crisis in the European cities/metropolitan areas 6 Lorant V, De Gelder R, Kapadia D, et al. Socioeconomic inequalities in suicide in studied. However, these inequalities should continue to be Europe: the widening gap. Br J Psychiatry 2018;212:356–61. monitored because these results could change in the long run. 7 Cairns JM, Graham E, Bambra C. Area-level socioeconomic disadvantage and Moreover, it is important to emphasise the importance of imple- suicidal behaviour in Europe: a systematic review. Soc Sci Med [Internet] menting policies to tackle suicide mortality and inequalities in 2017;192:102–11. suicide, including a broader view of the social determinants of 8 Bergen H, Hawton K, Waters K, et al. Premature death after self-harm: a multicentre health during the life course.34 cohort study. Lancet [Internet] 2012;380:1568–74. 9 Gotsens M, Marı́-Dell’Olmo M, Pérez K, et al. Socioeconomic inequalities in injury mortality in small areas of 15 European cities. Heal Place 2013;24:165–72. Supplementary data 10 Gotsens M, Marı́-Dell’Olmo M, Martı́nez-Beneito MA, et al. Socio-economic inequalities in mortality due to injuries in small areas of ten cities in Spain (MEDEA Supplementary data are available at EURPUB online. Project). Accid Anal Prev 2011;43:1802–10. 98 European Journal of Public Health 11 Santana P, Costa C, Marı́-Dell’Olmo M, et al. Mortality, material deprivation and 23 Wyllie C, Platt S, Brownlie J, et al. Men, Suicide and Society [Internet]. Samaritans, editor. urbanization: exploring the social patterns of a metropolitan area. Int J Equity Surrey, UK, 2012. 155 p. Available at: http://www.mengage.org.au/images/Men_Suicide_ Health 2015;14:55. and_Society_Research_Report_151112.pdf (30 June 2019, date last accessed). 12 Oyesanya M, Lopez-Morinigo J, Dutta R. Systematic review of suicide in economic 24 Szekely A, Mergl R, Koburger N, et al. What are reasons for the large gender recession. World J Psychiatry 2015;5:243–54. differences in the lethality of suicidal acts? An epidemiological analysis in four 13 Karanikolos M, Heino P, McKee M, et al. Effects of the global financial crisis on European countries what are reasons for the large gender differences in the health in high-income OECD countries: a narrative review. Int J Health Serv lethality of suicidal acts? An epidemiological analysis. PLoS One 2016;46:208–40. 2015;10:e0129062. 14 Parmar D, Stavropoulou C, Ioannidis JPA. Health outcomes during the 2008 25 Papaslanis T, Kontaxakis V, Christodoulou C, et al. Suicide in Greece 1992-2012: a financial crisis in Europe: systematic literature review. Br Med J 2008;1–11. time-series analysis. Int J Soc Psychiatry 2016;62:471–6. 15 Coope C, Gunnell D, Hollingworth W, et al. Suicide and the 2008 economic 26 Christodoulou C, Efstathiou V, Michopoulos I, et al. The economic crisis in Greece recession: who is most at risk? Trends in suicide rates in England and Wales 2001- and its impact on the seasonality of suicides in the Athens Greater Area. Psychiatry 2011. Soc Sci Med 2014;117:76–85. Investig 2017;14:16–20. 16 Wada K, Gilmour S. Inequality in mortality by occupation related to economic crisis 27 Sen CS, Stuckler D, Yip P, Gunnell D. Impact of 2008 global economic crisis on from 1980 to 2010 among working-age Japanese males. Sci Rep 2016;6:22255. suicide: time trend study in 54 countries. BMJ 2013;347:1–15. Downloaded from https://academic.oup.com/eurpub/article/30/1/92/5549577 by guest on 10 October 2023 17 Wada K, Kondo N, Gilmour S, et al. Trends in cause specific mortality across 28 Hirsch JK. A review of the literature on rural suicide: risk and protective factors, occupations in Japanese men of working age during period of economic stagnation, incidence, and prevention. Crisis 2006;27:189–99. 1980-2005: retrospective cohort study. BMJ 2012;344:e1191. 29 Santana P, Costa C, Cardoso G, et al. Suicide in Portugal: spatial determinants in a 18 Borrell C, Marı́-Dell’Olmo M, Gotsens M, et al. Socioeconomic inequalities in context of economic crisis. Heal Place [Internet] 2015;35:85–94. suicide mortality before and after the economic recession in Spain. BMC Public 30 Saltkjel T, Ingelsrud MH, Dahl E, Halvorsen K. A fuzzy set approach to economic crisis, Health 2017;17:772. austerity and public health. Part I. European countries’ conformity to ideal types during 19 Borrell C, Rodrı́guez-Sanz M, Bartoll X, et al. The suffering of the population in the the economic downturn. Scand J Public Health [Internet] 2017;45:41–7. economic crisis of the Spanish State. Salud Colect [Internet] 2014;10:95–8. Available 31 Chan CH, Caine ED, You S, et al. Suicide rates among working-age adults in South at: http://www.scielo.org.ar/scielo.php?script=sci_arttext&pid=S1851- Korea before and after the 2008 economic crisis. J Epidemiol Community Health 82652014000100008&lng=en&nrm=iso&tlng=en (2 July 2014, date last accessed). 2014;68:246–52. 20 Besag J, York J, Mollié A. Bayesian image restoration, with two applications in 32 Spallek J, Reeske A, Norredam M, et al. Suicide among immigrants in Europe—a spatial statistics. Ann Inst Stat Math 1991;43:1–59. systematic literature review. Eur J Public Health 2015;25:63–71. 21 Steele IH, Thrower N, Noroian P, Saleh FM. Understanding suicide across the 33 Värnik P, Sisask M, Värnik A, et al. Validity of suicide statistics in Europe in relation lifespan: a United States perspective of suicide risk factors, assessment & to undetermined deaths: developing the 2-20 benchmark. Inj Prev 2012;18:321–5. management. J Forensic Sci 2018;63:162–71. 34 Das-Munshi J, Thornicroft G. Failure to tackle suicide inequalities across Europe. Br 22 Pirkis J, Spittal MJ, Keogh L, et al. Masculinity and suicidal thinking. Soc Psychiatry J Psychiatry 2018;212:331–2. Psychiatr Epidemiol 2017;52:319–27. ......................................................................................................... The European Journal of Public Health, Vol. 30, No. 1, 98–104  The Author(s) 2019. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved. doi:10.1093/eurpub/ckz128 Advance Access published on 11 July 2019 ......................................................................................................... Social mobility, accumulation of disadvantages and health. An analysis with retrospective data from the GSOEP (2002–14) Anja Knöchelmann, Sebastian Günther, Irene Moor, Nico Seifert, Matthias Richter Institute of Medical Sociology, Medical Faculty, Martin Luther University Halle-Wittenberg, Halle (Saale), Germany Correspondence: Anja Knöchelmann, Institute of Medical Sociology, Medical Faculty, Martin Luther University Halle-Wittenberg, Magdeburger Str. 8, 06112 Halle (Saale), Germany, Tel: +49 (0) 345 557 1171, Fax: +49 (0) 345 557 1165, e-mail: anja.knoechelmann@medizin.uni-halle.de Background: Socioeconomic position (SEP) in different life stages is related to health-related quality of life (HRQoL). Yet, research on the relevance of life course processes is scarce. This study aims to analyse the association between accumulation of disadvantages, social mobility and HRQoL. Methods: Analyses were conducted using population-averaged panel-data models and are based on data from the German Socio-Economic Panel 2002–14, including retrospective biographical information, comprising 25 473 observations from 8666 persons. Intergenerational and intragenerational mobility included the occupational positions in childhood (parental position), first job and middle age. Accumulation of disadvantages was measured using an accumulation index. HRQoL was assessed using the Mental and Physical Component Summary Scores of the SF12v2. Results: Accumulation of disadvantages was the main predictor for the Physical Component Summary in mid-age. Men and women in a stable low SEP or with a steep downward mobility showed the least favourable physical HRQoL. This holds for intergenerational and intragenerational mobility. Mental HRQoL did not seem to be associated with accumulation or social mobility. Conclusion: The results show that physical HRQoL is related to social mobility and accumulation of (dis-)advantages. Further research is needed thoroughly analysing this association. .........................................................................................................