Epidemiology and Psychiatric The role of ageing in the wish to be dead: Sciences disentangling age, period and cohort effects in cambridge.org/eps suicide ideation in European population M. Cabello1,2,3 , L. A. Rico-Uribe2, J. C. Martinez-Ávila2, A. Sánchez-Niubò2,4, Original Article F. F. Caballero5,6, G. Borges7 , B. Mellor-Marsá2,7, J. M. Haro2,7, M. Prina8,9, S. Koskinen10 and J. L. Ayuso-Mateos1,2,3 Cite this article: Cabello M et al (2021). The role of ageing in the wish to be dead: 1 disentangling age, period and cohort effects in Department of Psychiatry, Universidad Autonoma de Madrid, Madrid, Spain; 2Instituto de Salud Carlos III, Centro suicide ideation in European population. de Investigación Biomédica en Red de Salud Mental (CIBERSAM), Madrid, Spain; 3Instituto de Investigación Epidemiology and Psychiatric Sciences 30, e17, Sanitaria Princesa (IIS-Princesa), Madrid, Spain; 4Parc Sanitari Sant Joan de Déu, Fundació Sant Joan de Déu, 1–9. https://doi.org/10.1017/ Sant Boi de Llobregat, Barcelona, Spain; 5Centro de Investigación Biomédica en Red de Epidemiología y Salud S2045796021000020 Pública, Madrid, Spain; 6Department of Preventive Medicine, Public Health, and Microbiology, Universidad Autónoma de Madrid, Madrid, Spain; 7National Institute of Psychiatry Ramon de la Fuente, Mexico City, Mexico; Received: 5 July 2020 8 Social Epidemiology Research Group. Health Service and Population Research Department, Institute of Revised: 11 January 2021 Psychiatry, Psychology & Neuroscience, King’s College London, London, UK; 9Global Health Institute, King’s Accepted: 17 January 2021 College London, London, UK and 10National Institute for Health and Welfare (THL), Helsinki, Finland Key words: Cohort; Europe; period; suicide ideation Abstract Author for correspondence: Aims. To investigate potential age, period and birth cohort effects in the prevalence of suicide Jose Luis Ayuso-Mateos, ideation in European ageing population. E-mail: joseluis.ayuso@uam.es Methods. A total of 50 782 community-dwelling adults (aged + 50) from 20 different European countries were collected in the Survey Health Ageing and Retirement study. A multilevel logistic regression model of repeated measures was modelled to assess the effects of age and other variables, including the variability of observations over three levels: birth cohort groups, time period assessment and individual differences. Results. The larger effect of variability was attributed to individual-level factors (57.8%). Youngest-old people (65–79 years) showed lower suicide ideation than middle-aged people (50–64 years). No significative differences were found for suicide ideation between middle- aged people and oldest-old (80 + years). Only 0.85% and 0.13% of the total variability of sui- cide ideation accounted for birth cohort and period effects, respectively. Cohorts born between 1941 and 1944 possessed the lowest estimates of suicide ideation. Conversely, suicide ideation started to rise with post-War generations and reached a significant level for people born from 1953–1957 to 1961–1964. Regarding the time period, participants assessed in 2006–2007 showed a lower likelihood of suicide ideation. The rest of the cohorts and period groups did not show any significant effect on the prevalence of suicide ideation. Conclusions. Our results suggest that age and suicide ideation relationship is not linear in middle and older age. The European Baby boomers born from 50s to mid-60s might report higher suicide ideation than their ancestors. This scenario would imply a greater need for mental healthcare services for older people in the future. Introduction The expression of suicide ideation has been longitudinally associated with completed suicide both in psychiatric and general population samples (Hubers et al., 2016). Suicide ideation is higher among people younger than 50 years in Europe (Bernal et al., 2007). However, people aged 50 years and older have a higher prevalence of completing suicide (Eurostat, 2020). In addition, suicide ideation in middle and old age is associated with functioning problems (Briggs et al., 2018), loneliness, poor social support (Almeida et al., 2012), and most import- © The Author(s), 2021. Published by antly with an amount of psychological suffering that can be managed by existing treatments Cambridge University Press. This is an Open (O’Riley et al., 2014). Access article, distributed under the terms of Although there are well-known risk factors for suicide ideation among middle and old age the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0), people, the role of ageing in this group is still unclear. Some studies have found that people which permits unrestricted re- use, younger than 65 years show higher suicide ideation than people older than 65 years distribution and reproduction, provided the (Cabello et al., 2019), whereas other studies have reported that suicide ideation increased original article is properly cited. with age among people older than 70 years (Stolz et al., 2016; Fässberg et al., 2020). These results might suggest a U-Shaped relationship between age and suicide ideation, increasing sui- cide ideation in the middle-aged, decreasing in the young-old and increasing again over the oldest-old. However, this available literature has several limitations that should be considered. First, the use of different measurements. Some measures include thoughts that life is not worth 2 M. Cabello et al. living (passive suicide ideation) (Stolz et al., 2016; Briggs et al., and country can be consulted in supplementary material (online 2018), other instruments assess thoughts of taking own life Supplementary Table S1). Wave-to-wave retention rates of Wave 1 (Cabello et al., 2019), and finally others mix both concepts participants was higher than 55% in all the countries (Bergmann (Almeida et al., 2012). Secondly, the inclusion of samples with et al., 2017). Data collection was conducted in the respondent’s different age groups. Some studies include age-heterogeneous house using a computer-assisted personal interview. All inter- samples (Briggs et al., 2018; Cabello et al., 2019), whereas others viewers were systematically trained by local group leaders who only include older people (Almeida et al., 2012; Fässberg et al., received a common train-the-trainer workshop. The study proto- 2020). Finally, a third potential factor is probably related to the col was originally built in English and then translated using a lack of distinction among age, cohort and period effects in most forward-backward methodology. Special efforts were conducted of the existing studies. Cohort effects refer to the existence of to assure all the translations kept functional equivalence both in common cultural and societal values that are shared by people concepts and phrases used across countries. Further information born in similar years. Period effects, in turn, are the environmen- on the study protocol development can be consulted elsewhere tal factors that are related to living in specific years and that affect (Börsch-Supan et al., 2008). The study methods were approved all population age groups (for example, an economic recession or by the University of Mannheim’s internal review board until medical and technological advances). Cohort and period effects 2011 and by the Ethics Council of the Max-Planck-Society for have been largely analysed for depression and suicide deaths the Advancement of Science since 2011. Further details of data (Phillips, 2014; Sullivan et al., 2020). However, only one study collection, sampling methods and study design used are explained to our knowledge has identified that suicide ideation trends in elsewhere for Wave 1 (Börsch-Supan et al., 2005), Wave 2 the last two decades could be partially attributed to cohort and (Börsch-Supan et al., 2008), Wave 4 (Malter and Börsch-Supan, period effects (Twenge et al., 2019). In this study, people born 2013) and Wave 5 (Malter and Börsch-Supan, 2015). after the 1980s reported higher suicide ideation rates than previ- ous generations. In addition, an increasing suicide ideation trend was observed since 2015 across all age groups. Although this study Measures is valuable, it did not include a large proportion of older adults, Suicidal ideation and it was limited to the US population. A binary item from the EURO-Depression scale (Prince et al., Therefore, the present study is aimed at filling the existing gap 1999) was used. The question included ‘In the last month, have in literature, by investigating longitudinally whether suicidal idea- you felt that you would rather be dead?’. Responses were coded tion trends in European ageing populations might respond to, as positive if any mention of wishing to be dead or suicidal feel- age, time period, or birth cohort effects. ings were reported and negative if there were no such feelings. Besides clarifying previously mentioned inconsistent results, disentangling the role of age, period, and cohort effects on the Health status prevalence of suicide would hold a number of different implica- A common latent health score based on self-reported health items tions for suicide prevention. Firstly, given that approximately related to impairments in body functions, limitations in Activities 35% of adults with suicide ideation continue to report suicide of Daily Living and Instrumental Activities of Daily Living, and ideation 10 years later (Borges et al., 2008), the existence of a pos- measured tests of cognitive performance and walking speed was sible cohort birth effect among younger generations may help to created. This variable allowed us to control the effect that the gen- predict a greater need for elderly mental health care as these gen- eral level of health might have on the presence of suicide ideation erations age. On the other hand, an increasing prevalence of sui- over time. Different statistical techniques (including Factor cide ideation in the last few years due to a period effect might Analysis, Bayesian multilevel Item Response Theory and reveal that effective interventions should be launched to stem Machine Learning methods) were employed. The metric has the increasing ‘epidemic’ of suicide ideation affecting all age shown a good performance in terms of predictive ability of mor- groups. Finally, if the change is mainly due to the ageing process tality (de la Fuente et al., 2018) and morbidity (Caballero et al., and the prevalence of suicide ideation lessens as the person ages, 2017). Values ranged from 0 to 100, where higher values reflected suicide ideation prevention efforts might be particularly indicated a higher health status. Further information on this variable can be for specific lifetime periods. consulted elsewhere (Caballero et al., 2017). Methods Country A total of 20 countries were included. Not all the countries parti- Sample and study design cipated in all the study assessments. There were some countries The Survey of Health, Ageing and Retirement in Europe (SHARE) such as the Netherlands, Spain, Belgium, the Czech Republic, is a cross-national panel survey including European community- France, Germany, Italy, Sweden and Switzerland that participated dwelling population aged 50 and older (Börsch-Supan et al., in all the study waves, whereas others such as Hungary or 2013). A total of 50 782 participants from 20 European countries Portugal only participated in Wave 4. The number of participants that participated at least once from Wave 1 to Wave 5 were by wave in each country can be consulted in the supplementary included. Wave 1 was conducted in 2004–2005. Wave 2 was material (online Supplementary Table S2). assessed during 2006–2007. Wave 4 information was mainly col- lected in 2011 and Wave 5 was carried out in 2013. Data from Covariates Wave 3 were excluded as they did not include any suicide idea- Age which was grouped into three categories (50–64, 65–79 and tion measure. Household response rates in baseline ranged 80 and older), gender and level of education, divided into four from 40.3% (Belgium) to 97.5% (France) with an overall individ- categories (less than primary education, primary, secondary, ual response rate of 60.1%. Household response rates by Wave and tertiary or higher education) and marital status grouped Epidemiology and Psychiatric Sciences 3 into four categories (single, married-cohabiting, separated- periods effects on suicide ideation. Predicted OR and confidence divorced, and widow) were also included. intervals (95%) were generated from 500 simulated distributions, estimating the values of suicide ideation for each cohort and per- iod level. Results were graphically presented. Analyses All statistical analyses were done using R (R Core Team, 2013). All graphs were created using ggplot2 package (Wickham, 2016). Descriptive characteristics of the sample were described, using Mixed-effect model was fitted using the glmer function in the mean and standard deviation for continuous variables, and fre- lme4 package (Bates et al., 2015). Simulations of random effects quencies and percentages for categorical variables. Chi-squared were calculated using the REsim function in the merTools pack- tests were used in order to assess the association between categor- age (Knowles and Frederick, 2016). P values lower than 0.05 ical variables and the presence of suicidal ideation in Wave 1. were considered significant. All the analyses were conducted Similarly, the Wilcoxon test was used for estimating differences with the available information with no data imputation. between suicide ideation groups in continuous variables. The measurement occasions were compilated as period effects including the following four groups of years: 2004–2005, 2006– Results 2007, 2011 and 2013. Over this time period, a total of 15 consecu- The overall sample consisted of 50 782 participants that were tive birth cohorts were made every 4 years, beginning with those interviewed at least once (n = 22 621; 44.5% were men). born in 1905–1908 and ending with those born in 1961–1964. We Descriptive statistics by the presence of suicide ideation are pro- selected cohorts based on 4 years period to ensure a homogeneous vided in Table 1. A total of 4031 participants (7.94%) showed sui- and sufficient number in all groups. Prevalence rates of suicide cide ideation the first time they were interviewed. Prevalence ideation were calculated by birth cohort and period groups and of suicide ideation across gender, age, level of education, country graphically presented. groups and health status can be consulted in Table 1. Unadjusted Next, to better disentangle the effects of age, time period, and prevalence rates of suicide ideation were descriptively higher for birth cohort, we performed a mixed-effect model. Suicide ideation cohorts born during the first two decades of the 20th century was recorded in SHARE as a dichotomous variable, for this reason than the cohorts born since 1941 (Fig. 1). Two cohorts 1909– a logistic approach was applied. The proposed model was a logis- 1912 and 1913–1916, reported descriptively an increasing trend tic mixed model of repeated measures which could be described of suicide ideation in the year 2013 (Fig. 1). as follows: The results of the mixed-effect regression model are presented in Table 2. As can be seen in the fixed part of the model, suicide ideation  sex + educational level + age youngest-old people (65–79 years) showed a lower likelihood of + health status + country, suicide ideation (OR = 0.88; CI 95% = 0.81, 0.96) than middle-aged people (50–64 years) (Table 2). In contrast, people with participant  N(0, s2p ), birth cohort  N(0, s2bc ) and aged 80 + years reported similar levels of suicide ideation than wave  N(0, s2wave ) as random effects middle-aged people (OR = 0.95; CI 95% = 0.91, 1.01). In addition, Therefore, the random part of the model included three levels: higher suicide ideation was related to being a woman, having a the person, the birth cohort and the time period. Random effects lower level of education and lower health status. Population at a personal level were added in order to capture variability due with married or cohabiting marital status reported lower suicide to intrinsic participant characteristics. Cohort birth was added as ideation than singles (OR = 0.72; CI 95% = 0.65, 0.79). Being a random effect to explain the existence of cultural and societal widowed (OR = 1.21; CI 95% = 1.09, 1.35) and separated/divorced changes that make people born in specific years more vulnerable (OR = 1.27; CI 95% = 1.13, 1.42) were associated with higher odds to suicide ideation. Finally, the period effect was also included as of suicide ideation than being single. Significant differences in sui- random to account for variability due to the specific circum- cide ideation estimates were also found across countries (Table 2). stances that all participants experienced over the study time. Regarding the random part of the model, most of the variation in The identification problem, which is the linear dependence suicide ideation was explained by inter-participant characteristics, between age, period and cohort effects was managed according 57.8% of the total variance (σ 2 = 1.40; CI 95% = 1.24–1.55). to Reither’s indications (2015). We decided to include periods Birth-cohort random effect explained a smaller percentage 0.85% and cohorts as random since they are contextual variables, that but still had an effect on the variability in suicide ideation (σ 2 = are independent of individual-level variables. These random 0.02; CI 95% = 0.0019–0.0023), whereas the time period level effects followed a normal distribution with a mean of zero and explained only 0.13% of the variation (σ 2 = 0.003; CI 95% = variance to be estimated, the larger the variance, the higher the 0.001–0.006) (Table 2). Predicted estimates for suicide ideation impact in the observed variability. The fixed part of the model across birth cohort groups are shown in Fig. 2. People born in included sex, educational level, age, health status and country as 1941–1944 (OR = 0.89; CI 95% = 0.80–0.99), showed a decreased fixed effects. Age was treated as fixed because it was considered likelihood of reporting suicide ideation (Fig. 2). In addition, people as a biological factor that is related to individual-level character- born between 1953 and 1955 (OR = 1.17; CI 95% = 1.05–1.31), istics and whose effects would be constant across individuals. 1956–1960 (OR = 1.21; CI 95% = 1.08–1.35) and 1961–1964 After running the Restricted Maximum Likelihood Model, (OR = 1.32; CI 95% = 1.16–1.49) showed a higher likelihood of estimated coefficients for fixed effects were transformed into reporting suicide ideation. Predicted intervals and values of time odds ratios (ORs), with variance components estimations provid- period effect are displayed in Fig. 3. In comparison with other ing an idea of influence in variability since total variance can be time periods, participants assessed during the 2006–2007 period described as follows, s2Total = s2p + s2cb + s2wave + 1. reported a decreased likelihood of suicide ideation (OR = 093; CI Finally, considering the results of the previous modelled rela- 95% = 0.87–0.99). The rest of the birth cohorts and period levels tionships, we simulated the posterior distributions of birth and did not show any significant effect on suicide ideation. 4 M. Cabello et al. Table 1. Characteristics of the SHARE participants interviewed in the baseline Suicide ideation No (n = 46 751) Yes (n = 4031) χ2/Wilcoxon test p Sex, n (%) 287.214 <0.001 Men 21 339 (45.6) 1282 (31.8) Women 25 412 (54.4) 2749 (68.2) Age, 497.89 <0.001 50–64 25 712 (55.0) 1682 (41.7) 65–79 17 137 (36.7) 1633 (40.5) +80 3908 (8.3) 716 (17.8) Marital status 834.15 <0.001 Single 2470 (5.3) 232 (5.8) Married-cohabiting 34 327 (73.4) 2181 (54.1) Separated-divorced 3158 (6.8) 392 (30.4) Widow 6792 (14.5) 1226 (30.4) Education, n (%) 578.855 <0.001 Less than primary 1597 (3.4) 346 (8.6) Primary 10 571 (22.8) 1291 (32.3) Secondary 25 325 (54.6) 1948 (10.4) Tertiary 8926 (19.2) 416 (10.4) Health status, Mean (S.D.) 71.82 (15.93) 56.20 (16.70) 2797.205 <0.001 Country 517.231 <0.001 Austria 1477 (3.2) 60 (1.5) Belgium 3323 (7.1) 325 (8.1) Czech Republic 2501 (5.3) 190 (4.7) Denmark 1484 (3.2) 93 (2.3) Estonia 5984 (12.8) 531 (13.2) France 2525 (5.4) 336 (8.3) Germany 2735 (5.9) 163 (4.0) Greece 2421 (5.2) 163 (4.0) Hungary 2563 (5.5) 384 (9.5) Ireland 1049 (2.2) 36 (0.9) Israel 2207 (4.7) 226 (5.6) Italy 2325 (5.0) 156 (3.9) Luxembourg 1429 (3.1) 139 (3.4) Netherlands 2658 (5.7) 126 (3.1) Poland 2077 (4.4) 295 (7.3) Portugal 1770 (3.8) 185 (4.6) Slovenia 2491 (5.3) 186 (4.6) Spain 2024 (4.3) 263 (6.5) Sweden 2823 (6.0) 121 (3.0) Switzerland 885 (1.9) 53 (1.3) Notes: Baseline year varied depending on the country. Baseline assessment in Austria, Belgium, Switzerland, Germany, Denmark, Spain, France, Greece, Italy, Netherlands, Sweden, Israel was in 2004. It was in 2007 in the Czech Republic, Ireland and Poland. It was in 2011 in Estonia, Hungary, Portugal, Slovenia and it was Luxembourg in 2013 Epidemiology and Psychiatric Sciences 5 Fig. 1. Unadjusted prevalence of suicide ideation by Birth Cohort groups over study time. Notes: Wave 1 (2004–2005), Wave 2 (2006–2007), Wave 4 (2011) and Wave 5 (2013). Birth cohort is denoted by the first year of each cohort. Discussion Decreasing trends of suicide ideation were observed since the earliest 20s, with cohorts born between 1941 and 1944 reporting This study is, to our knowledge, the first to disentangle age, the lowest likelihood of suicide ideation. This result is congruent cohort and period effects on the prevalence of suicide ideation with other studies that have shown that people born just before among people aged 50 years and older in Europe, shedding and during World War II were the generation with the lowest light on the ongoing debate about the role of ageing in suicide prevalence of suicide deaths and depression (Phillips, 2014; ideation. To do so, we analysed the prevalence trends of suicide Sullivan et al., 2020). Some authors argue that people from this ideation, including the age effect and the variability effects of generation could have learned to adapt personal aspirations to birth cohorts from 1905 to 1961, and of four time periods persevere through difficult times and be more resilient to frustra- between 2004 and 2013. In this study, the main effects of suicide tions (Phillips, 2014). Other authors have also supported that ideation prevalence were primarily due to individual-level factors. selective mortality could have largely affected this generation, In comparison with youngest-old (65–79 years), middle-aged i.e. those with health problems died younger and hence they are (50–64 years) reported higher suicide ideation, regardless of per- not presented in middle and old aged population studies (Hasin iod and cohort effects. Middle-aged might be a period associated and Link, 1988). Further studies should explore the reasons why with economic, health and social burdens that could rise the so many people from this generation have shown low levels of sui- prevalence of suicidal ideation (Jo et al., 2017). In contrast, no dif- cide ideation. ferences were found between people aged 80 years and older and On the other hand, an increasing trend of suicide ideation was those in middle-aged. Oldest-old people might more likely have to observed for post-War generations (called baby boomers), reach- face with difficulties such as physical diseases and bereavement ing the significant level for people born from 1953 to 1964. This than youngest-old people (Koo et al., 2017), which may increase finding is also congruent with previous studies that report that their likelihood of suicide ideation. Our results support that the baby boomers are showing higher psychopathology than their relationship between age and suicide ideation may not be linear ancestors (Phillips, 2014). According to Durkheim (2005) societal in elderly people, and suggest that different age groups should and cultural changes that occurred from the post-war generations be separately analysed in the study of suicide ideation in older onwards might have weakened societal bonds, making these gen- people. erational groups more vulnerable to suicide. Other authors have In line with previous literature, married reported a lower like- added that baby boomers were the first ones who imposed a lihood of suicide ideation than single people (Smith et al., 1988). youth culture where ageing was marginalised. Therefore, it is pos- In contrast, being single was associated with a lower level of sui- sible that some of them became victims of the culture they created cidal ideation than being widowed and separated-divorced. themselves (McCue and Balasubramaniam, 2016). Although fur- Our study has also revealed that there was a small but still rele- ther studies are necessary, our findings suggest that baby-boomers vant cohort effect in the prevalence of suicide ideation. are more affected by suicide ideation. Therefore, there will 6 M. Cabello et al. Table 2. Parameter estimates of the multilevel mixed-effects models assessing the prevalence of suicide ideation in the SHARE participants Suicide ideation Mixed model Fixed effects OR (CI 95%) p Intercept 2.108 (1.72, 2.58) <0.001 Gender (male) 0.82 (0.78, 0.86) <0.001 Age (ref. 50–64) 65–79 0.88 (0.81,0.96) 0.004 80+ 0.95 (0.91, 1.01) 0.078 Marital status (ref. single) Married-cohabiting 0.72 (0.65, 0.79) <0.001 Separated-divorced 1.27 (1.13,1.42) <0.001 Widow 1.21 (1.09, 1.35) <0.001 Education (ref. no education) Primary 0.94 (0.85, 1.04) 0.25 Secondary 0.84 (0.76, 0.93) 0.001 Tertiary 0.75 (0.68, 0.85) <0.001 Health Satus 0.95 (0.9, 0.95) <0.001 Country (ref. Austria) Belgium 2.01 (1.79, 2.26) <0.001 Czech Republic 1.93 (1.71, 2.17) <0.001 Denmark 0.89 (0.76, 1.04) 0.15 Estonia 1.13 (0.99, 1.28) 0.057 France 2.77 (2.46, 3.12) <0.001 Germany 1.32 (1.15, 1.51) 0.001 Greece 0.97 (0.81, 1.16) 0.761 Hungary 2.17 (1.84, 2.55) <0.001 Ireland 0.73 (0.50, 1.07) 0.105 Israel 1.57 (1.35, 1.84) <0.001 Italy 0.85 (0.74, 0.97) 0.017 Luxembourg 1.82 (1.45, 2.27) <0.001 Netherlands 1.13 (0.99, 1.30) 0.097 Poland 1.87 (1.60, 2.19) <0.001 Portugal 1.47 (1.20, 1.81) <0.001 Slovenia 1.10 (0.94, 1.29) 0.24 Spain 1.40 (1.23, 1.59) <0.001 Sweden 0.98 (0.85, 1.14) 0.82 Switzerland 1.49 (1.29, 1.73) <0.001 Random effects Variance Bootstrap 95% CI Participant 1.40 1.24–1.55 Birth Cohort 0.02 0.0019–0.023 Period 0.003 0.001–0.006 probably be a higher need for elderly mental health and social Finally, our study also showed that a very small amount of sui- community support in Europe in the future since this generation cide ideation changes might be attributed to period effects. A is already entering the old age. small period effect was observed for participants assessed during Epidemiology and Psychiatric Sciences 7 Fig. 2. Predicted ORs and 95% CIs of suicide ideation for each birth cohort. Note: Births cohorts are denoted by the first year of each cohort. Fig. 3. Predicted ORs and 95% CIs of suicide ideation for each time period. Notes: Wave 1 covered the 2004–2005 period; Wave 2 was conducted in the 2006–2007 period; Wave 4 was in 2011 and Wave 5 was in 2013. 8 M. Cabello et al. the 2006–2007 years, who reported a lower prevalence of suicide particularly high-risk time period of suicide ideation in compari- ideation in comparison with other time occasions. Some authors son with early elderly (65–78 years). However, the protective have highlighted the optimism experienced during the 2006–2007 effect of older age might not apply for people age 80+. Our results years in the European Union due to the inclusion of new coun- have also suggested that there is a certain cohort effect, with baby tries and to good economic stability which met or even exceeded boomers being more likely to report a wish to die than the gener- expectations (Schrogl et al., 2008). Conversely, no increases of sui- ation born just during War World II, showing this later gener- cide ideation were observed during the economic downturn per- ation the lowest levels of suicide ideation. Thus, it is possible iod in the 2011 and 2013 years. Although we do not have a clear that a greater need for community mental health services by interpretation for this result, it is possible that our cross-country older European people could follow in the near future. approach blurred some differences, since some European coun- Supplementary material. The supplementary material for this article can tries were greatly affected by the economic recession, whereas be found at https://doi.org/10.1017/S2045796021000020 others were not. In addition, the effect of a selection bias cannot be ruled out either, since people may be less likely to participate in Acknowledgements. The authors gratefully acknowledge the participation a study if they were affected by the events. of the interviewed participants, without whom this study would not have The great variety of countries covered, the use of a longitudinal been possible. panel design – which allowed us to simultaneously manage age, period and cohort effects – controlling for important effects Financial support. This work was supported by the Ageing Trajectories of Health: Longitudinal Opportunities and Synergies (ATHLOS) project which such as general health or country and the use of a measure that was funded by the European Union’s Horizon 2020 (No. 635316). The present recorded the presence of suicide ideation in the last month – work also received support from the Instituto Carlos III-FIS (project PI16/ which reduced the likelihood of recall bias effects – are probably 00218) and by The European Union Regional Development Fund (ERDF) the main strengths of this study. However, we acknowledge that ‘A Way to Build Europe’ (project PI16/00218). The first four Waves of the the results of this study should also be interpreted with caution SHARE study have been mainly funded by the European Commission by due to the following limitations. Firstly, although European coun- the 5th framework programme (project QLK6-CT-2001-00360 in the thematic tries share common cultural and societal values, it is possible that programme Quality of Life)), the 6th framework programme (projects the aim of searching a global European birth and period cohort SHARE-I3, RII-CT-2006-062193, COMPARE, CIT5-CT-2005-028857, and effect may have hidden the variability that each birth cohort SHARE-LIFE (CIT4-CT-2006-028812)) and by the 7th framework programme, and period group had in the different countries. Unfortunately, projects SHARE-PREP (No 211909), SHARE-LEAP (No 227822) and M4 (No 261982). Additional funding from the German Ministry of Education and our data did not allow specific country analyses due to the vari- Research, the Max Planck Society for the Advancement of Science, the US ability of time periods that each country included in the study. National Institute on Aging (U01_AG09740-13S2, P01_AG005842, Secondly, our study did not include information of severity and P01_AG08291, P30_AG12815, R21_AG025169, Y1-AG-4553-01, IAG_BSR06-11, covered a concept of passive suicide ideation, which has some- OGHA_04-064, HHSN271201300071C), and from various national funding sources times been considered as a less severe form of suicide ideation is gratefully acknowledged (see www.share-project.org). in clinical settings. However, wishes of death or that life is not worth living have been strongly associated with psychopathology Conflict of interest. None declared. and they did not seem to be normative in older populations (Van Ethical standards. The authors assert that all procedures contributing to Orden et al., 2015). Thirdly, the use of a self-reported measure this work meet the ethical standards of the relevant national and institutional might have introduced biases related to underreporting wishes committees on human experimentation and the Helsinki Declaration of 1975, of death due to social desirability. Fourthly, household response as revised in 2013. rates were globally higher than 45% in all the study waves. However, there were some countries with rates lower than 40%. Availability of Data and Materials. This paper uses data from SHARE Waves 1, 2, 4, and 5 (DOIs:10.6103/SHARE.w1.700, 10.6103/SHARE.w2.700, This fact has to be carefully considered in the interpretation of 10.6103/SHARE.w4.700, and 10.6103/SHARE.w5.700), which are available the study results. Nonetheless, it is important also to consider via the SHARE project’s website. that response rates in the SHARE study are in line with the num- bers of comparable European surveys conducted in the same per- iod (Koen et al., 2018). Fifthly, the method on how to best deal References with point identification problem is still an open debate in the lit- Almeida OP, Draper B, Snowdon J, Lautenschlager NT, Pirkis J, Byrne G, erature with not a single answer (O’Brien, 2017). We acknowledge Sim M, Stocks N, Flicker L and Pfaff JJ (2012) Factors associated with sui- there are other approaches based on mathematical solutions that cidal thoughts in a large community study of older adults. The British could have been applied (Tu et al., 2012). However, we managed Journal of Psychiatry 201, 466–472. the point identification problem according to explicit theory- Bates D, Mächler M, Bolker B and Walker S (2015) Fitting linear based assumptions that have also been shown as valid (Reither mixed-effects models using lme4. Journal of Statistical Software. Articles et al., 2015). Moreover, we categorised age, periods and cohorts 67, 1–48. into unequal width groups in order to also decrease the possible Bergmann M, Kneip T, De Luca G and Scherpenzeel A (2017) Survey par- linear dependence between age, period and cohort effects. ticipation in the survey of health, ageing and retirement in Europe Finally, although the study included a follow-up period of 8 (SHARE), Wave 1–6. Based on Release 6.0.0 (March 2017), SHARE years (2004–2013), middle ages could be underrepresented for Working Paper Series 31-2017. Munich: Munich Center for the Economics of Aging, pp. 1–46. earlier generations. Bernal M, Haro JM, Bernert S, Brugha T, de Graaf R, Bruffaerts R, Lépine In spite of all these limitations, our study is the first to analyse JP, de Girolamo G, Vilagut G and Gasquet I (2007) Risk factors for sui- the role of ageing including the age, cohort and period effects on cidality in Europe: results from the ESEMED study. Journal of Affective the prevalence of suicide ideation in the European population. Disorders 101, 27–34. The findings in our study support that the relationship between Borges G, Angst J, Nock MK, Ruscio AM and Kessler RC (2008) Risk factors age and suicide ideation is non-linear. Middle-age may be a for the incidence and persistence of suicide-related outcomes: a 10-year Epidemiology and Psychiatric Sciences 9 follow-up study using the National Comorbidity Surveys. Journal of efforts? Survey Methods: Insights from the Field, 1–12. https://surveyin- Affective Disorders 105, 25–33. sights.org/?p=9673 (Accessed 11 April 2020). Börsch-Supan A, Hank K and Jürges H (2005) A new comprehensive and Koo YW, Kõlves K and De Leo D (2017) Suicide in older adults: differences international view on ageing: introducing the “survey of health, ageing between the young-old, middle-old, and oldest old. International and retirement in Europe”™. European Journal of Ageing 2, 245–253. Psychogeriatrics 29, 1297. Börsch-Supan A, Brugiavini A, Jürges H, Kapteyn A, Mackenbach J, Malter F and Börsch-Supan A (2013) SHARE Wave 4: Innovations & meth- Siegrist J and Weber G (2008) First results from the Survey of Health, odology. Munich: Munich Center for the Economics of Aging, Max Planck Ageing and Retirement in Europe (2004–2007): Starting the longitudinal Institute for Social Law and Social Policy. dimension. Mannheim, Germany: Mannheim Research Institute for the Malter F and Börsch-Supan A (2015) SHARE Wave 5: Innovations & meth- Economics of Aging. odology. Munich: Munich Center for the Economics of Aging, Max Planck Börsch-Supan A, Brandt M, Hunkler C, Kneip T, Korbmacher J, Malter F, Institute for Social Law and Social Policy. Schaan B, Stuck S and Zuber S (2013) Data resource profile: the survey of McCue RE and Balasubramaniam M (eds) (2016) Rational Suicide in the Elderly: health, ageing and retirement in Europe (SHARE). International Journal of Clinical, Ethical, and Sociocultural Aspects. Cham, Switzerland: Springer. Epidemiology 42, 992–1001. O’Brien RM (2017) Mixed models, linear dependency, and identification in Briggs R, Tobin K, Kenny RA and Kennelly SP (2018) What is the preva- age-period-cohort models. Statistics in Medicine 36, 2590–2600. lence of untreated depression and death ideation in older people? Data O’Riley AA, Van Orden K and Conwell Y (2014) Suicidal ideation in late life. from the Irish Longitudinal Study on aging. International Psychogeriatrics In Pachana NA and Laidlaw K (eds), The Oxford Handbook of Clinical 30, 1393–1401. Geropsychology: International Perspectives. Oxford, UK: Oxford University Caballero FF, Soulis G, Engchuan W, Sanchez-Niubo A, Arndt H, Press, pp. 267–284. Ayuso-Mateos JL, Haro JM, Chatterji S and Panagiotakos DB (2017) Phillips JA (2014) A changing epidemiology of suicide? The influence of birth Advanced analytical methodologies for measuring healthy ageing and its cohorts on suicide rates in the United States. Social Science & Medicine 114, determinants, using factor analysis and machine learning techniques: the 151–160. ATHLOS project. Scientific Reports 7, 43955. Prince MJ, Reischies F, Beekman AT, Fuhrer R, Jonker C, Kivela S, Lawlor Cabello M, Miret M, Ayuso-Mateos J, Caballero FF, Chatterji S, BA, Lobo A, Magnusson H and Fichter M (1999) Development of the Tobiasz-Adamczyk B, Haro JM, Koskinen S, Leonardi M and Borges EURO–D scale–a European Union initiative to compare symptoms of depres- G (2019) Cross-national prevalence and factors associated with suicide sion in 14 European centres. The British Journal of Psychiatry 174, 330–338. ideation and attempts in older and young-and-middle age people. Aging R Core Team (2013) R: A Language and Environment for Statistical comput- & Mental Health 24, 1533–1542. ing. Vienna: R Core Team. http://www.R-project.org/. de la Fuente J, Caballero FF, Sánchez-Niubó A, Panagiotakos DB, Prina Reither EN, Masters RK, Yang YC, Powers DA, Zheng H and Land KC AM, Arndt H, Haro JM, Chatterji S and Ayuso-Mateos JL (2018) (2015) Should age-period-cohort studies return to the methodologies of Determinants of health trajectories in England and the United States: an the 1970s? Social Science & Medicine 128, 356–365. approach to identify different patterns of healthy aging. The Journals of Schrogl K, Mathieu C and Peter N (2008) Yearbook on Space Policy 2006/ Gerontology: Series A 73, 1512–1518. 2007: New Impetus for Europe. Vienna: Springer Science & Business Media. Durkheim E (2005) Suicide: A Study in Sociology. London: Taylor & Francis. Smith JC, Mercy JA and Conn JM (1988) Marital status and the risk of sui- Eurostat (2020) Suicide death rate by age group. Available at https://ec.europa. cide. American Journal of Public Health 78, 78–80. eu/eurostat/web/products-datasets/-/tps00202 (Accessed 12 April 2020). Stolz E, Fux B, Mayerl H, Rásky É and Freidl W (2016) Passive suicide idea- Fässberg MM, Vanaelst B, Jonson M, Sterner TR, Ahlner F, Wetterberg H, tion among older adults in Europe: a multilevel regression analysis of indi- Rydén L, Kern S, Sigström R and Zettergren A (2020) Epidemiology of vidual and societal determinants in 12 countries (SHARE). Journals of suicidal feelings in an ageing Swedish population: from old to very old Gerontology Series B: Psychological Sciences and Social Sciences 71, 947–958. age in the Gothenburg H70 birth cohort studies. Epidemiology and Sullivan KJ, Liu A, Dodge HH, Andreescu C, Chang CH and Ganguli M Psychiatric Sciences 29, e26, 1–14. (2020) Depression symptoms declining Among older adults: birth cohort Hasin D and Link B (1988) Age and recognition of depression: implications analyses from the rust belt. The American Journal of Geriatric Psychiatry for a cohort effect in major depression. Psychological Medicine 18, 683–688. 28, 99–107. Hubers AA, Moaddine S, Peersmann SH, Stijnen T, Van Duijn E, Van der Tu Y, Krämer N and Leec W (2012) Addressing the identification problem in Mast RC, Dekkers OM and Giltay EJ (2016) Suicidal ideation and subse- age-period-cohort analysis: a tutorial on the use of partial least squares and quent completed suicide in both psychiatric and non-psychiatric popula- principal components analysis. Epidemiology (Cambridge, Mass.) 23, 583–593. tions: a meta-analysis. Epidemiology and Psychiatric Sciences 27, 186–198. Twenge JM, Cooper AB, Joiner TE, Duffy ME and Binau SG (2019) Age, Jo A, Jeon M and Oh H (2017) Age-differentiated risk factors of suicidal idea- period, and cohort trends in mood disorder indicators and suicide-related tion among young and middle-aged Korean adults. Osong Public Health outcomes in a nationally representative dataset, 2005–2017. Journal of and Research Perspectives 8, 201. Abnormal Psychology 128, 185–199. Knowles JE and Frederick C (2016) Mertools: tools for analyzing mixed effect Van Orden KA, O´Riley AA, Simning A, Podgorski C, Richardson TM and regression models. R Package Version 0.3.0. https://CRAN.R-project.org/ Conwell Y (2015) Passive suicide ideation: an indicator of risk among older package=merTools (Accessed 19 May 2020). adults seeking aging services? The Gerontologist 55, 972–980. Koen B, Loosveldt G, Vandenplas C and Stoop I (2018) Response rates in the Wickham H (2016) ggplot2: Elegant Graphics for Data Analysis. New York, European Social Survey: increasing, decreasing, or a matter of fieldwork NY: Springer.