Why are suicide rates rising in young men but falling in the elderly?—a time-series analysis of trends in England and Wales 1950–1998
Introduction
Suicide is one of the principal causes of premature mortality in young adults in industrialised countries. The last 50 years have seen striking changes in its occurrence in England and Wales (see Fig. 1) (Charlton et al., 1992; McClure, 2000). In males aged <45 years rates have doubled, whilst they have declined substantially in females and older males. Similar increases in young male suicide have occurred in other industrialised nations although, in contrast to the pattern in England and Wales, many of these countries have also experienced rises in young female suicide (Charlton et al., 1993, Cantor, 2000). Explanations for these trends remain largely speculative.
Over 100 years ago, Durkheim recognised that societal changes, such as economic crises or periods of war, contribute to changing patterns of suicide (Durkheim, 1952). He viewed suicide rates as ‘social facts’ and argued that they may be influenced by, amongst other factors, the extent to which individuals are integrated within society. Durkheim argued that social integration could be achieved through family support together with religious, political and work affiliations (Durkheim, 1952). Sociological and epidemiological inquiry over the course of the last century has confirmed and refined Durkheim's theories, identifying the importance of patterns of unemployment, divorce, alcohol misuse and religion in explaining national differences and trends in suicide (Gunnell et al., 1999a; Lester, Curran, & Yung, 1991; Makela, 1996; Lester, 1997; Neeleman et al., 1997). In the last decade, it has been theorised that income inequality may too be an important additional influence on levels of social integration (Wilkinson, 1996; Kawachi & Kennedy, 1997).
The differing age- and sex-specific trends in suicide suggest that the societal processes underling them may be different in the various sociodemographic groups. Alternatively, common adverse influences may be offset by varying distributions of protective factors. With some exceptions (Morrell et al., 1993; Gunnell et al., 1999a; Makela, 1996), however, most previous assessments of the causes of recent changes in national suicide rates have investigated factors associated with time trends in overall population suicide rates (i.e. all age-groups combined) (Weyerer & Wiedenmann, 1995; Low et al., 1981; Lester & Yang, 1991). Such analyses cannot elucidate factors underlying the markedly varying time trends in different age–sex groups. Furthermore, many studies have assessed the effects on suicide rates of changes in single factors such as unemployment (Crombie, 1990; Gunnell et al., 1999a) or alcohol intake (Yip, Callanau, & Yuen, 2000). In other studies, associations between two or more factors have been examined but important variables, such as unemployment or alcohol consumption (Stack, 1990; Leenaars, Yang, & Lester, 1993), have been neglected. Inconsistent findings have emerged from these studies and in view of the degree of correlation between many proposed ‘risk factors’ it is challenging to identify the factors underlying recent trends.
Focussing on the two age/sex groups with the most divergent suicide trends in post-war Britain, i.e. 25–34 and 60+ year old men and women (see Fig. 1), the principle aim of this research was to identify the social, health and economic factors associated with recent patterns. Where possible we have used age- and sex-specific risk factor data. The main factors we have examined are those consistently associated with suicide risk in prospective person-based research and in time-series analyses (see Table 1). We also examined associations with three other measures—income inequality, gross domestic product (GDP) and measures of state support for the elderly. We investigated the influence of income inequality (measured using the GINI coefficient) as it has been hypothesised that increased inequality contributes to reduced social integration and increased mortality (Wilkinson, 1996; Kawachi & Kennedy, 1997). We hypothesised that changing levels of income inequality may influence suicide rates through a similar mechanism. We examined associations with GDP to investigate whether changes in a nation's economic performance, which have been shown to influence population levels of life satisfaction (Di Tella, MacCulloch, & Oswald, 2001), may also affect patterns of suicide. Lastly, as it has been suggested that declining suicide rates in older people may be due to increased levels of state income support and welfare provision (Hoxey & Shah, 2000), we used markers of these changes to assess their association with suicide rates in the elderly.
Section snippets
Suicide data
We obtained routinely available mortality data for England and Wales for the period 1950–1998 from the Office of National Statistics (ONS). For consistency across time we have used suicide data only—deaths coded E950-9 in the International Classification of Diseases (ICD) ICD 8 & 9 and E970-9 in ICD7. The category undetermined deaths—those given a coroner's open verdict and ICD9 coded E980–E989—was introduced in 1968. Rather than include undetermined deaths for only part of the period
Suicide trends
Fig. 1 shows age-specific trends in suicide in England and Wales over the last 50 years. The most striking feature over the years displayed is the reduction in suicide rates in males aged 55 and over. The rises in suicide in 15–24 and 35–44 year old males seen in the 1970s and 1980s have recently levelled off, but rates are still rising in 25–34 year old males and this group now has the highest rate of all age–sex groups. In females, with the exception of 15–24 year olds, there have been marked
Main findings
Four main findings emerged from our analyses. First, the increases in young male suicide in England and Wales in the last 30 years have paralleled rises in a number of risk factors for suicide in this age group, namely unemployment, divorce, alcohol and drug abuse, and declines in marriage. In keeping with the findings of Charlton and colleagues in their assessment of post-war suicide trends up to 1990 (Charlton et al., 1993) one of the factors most consistently associated with patterns of
Conclusions
This analysis indicates that no single factor can be clearly implicated as underlying recent trends in suicide. The causes of suicide are complex and multifactorial. Population trends are therefore likely to be influenced not only by social changes and patterns of health and healthcare, but also by trends in the lethality of popular methods used for suicide. Importantly the data presented here underline the striking differences in effects of societal changes on suicide risk at different stages
Acknowledgements
We wish to acknowledge: NHS South and West R&D for funding this project; IMS Health, Pinner, Middlesex for access to UK prescribing data. Professor Frankel was supported by the Leverhulme Trust during this period.
References (66)
- et al.
Increasing rates of suicide in young men in England during the 1980sThe importance of social context
Social Science & Medicine
(1999) - et al.
Sex differences in suicide trends in England and Wales
Lancet
(1999) Catalytic converters and prevention of suicides
Lancet
(1998)Are there social correlates to suicide?
Social Science & Medicine
(1997)- et al.
Excess mortality of unemployed men and women during a period of rapidly increasing unemployment
Lancet
(1996) - et al.
Suicide and unemployment in Australia 1907–1990
Social Science & Medicine
(1993) - et al.
Suicide and unemployment in ItalyDescription, analysis and interpretation of recent trends
Social Science & Medicine
(1992) - et al.
Urban/rural and gender differentials in suicide ratesEast and west
Journal of Affective Disorders
(2000) - et al.
Alcohol and mortality among young men: longitudinal study of Swedish conscripts
British Medical Journal
(1988) Suicide during pregnancy and in the first postnatal year
British Medical Journal
(1991)
Suicide in Americaa test of Durkheim's theory of religious and family intergration, 1933–1980
American Journal of Sociology
Suicides 1950–82
Population Trends
Time series analysis of alcohol consumption and suicide mortality in the United States, 1934–1987
Journal of Studies on Alcohol
Suicide in the Western World
Suicide deaths in England and WalesTrends in factors associated with suicide deaths
Population Trends
Trends in suicide deaths in England and Wales
Population Trends
The analysis of time series : An introduction
Application of least squares regression to relationships containing autocorrelated error terms
Journal of the American Statistical Association
Quasi-experimentationDesign & analysis issues for field settings
Can changes in the unemployment rates explain the recent changes in suicide rates in developed countries?
International Journal of Epidemiology
Current patterns of trends in male mortality by Social Class
Population Trends
Suicide (originally published 1897)
Increased rates of depressionUpdate of epidemiological findings and analytical problems
Acta Psychiatrica Scandinavica
Mental disorders and suicide in Northern Ireland
British Journal of Psychiatry
Suicides after pregnancy in Finland, 1987–94Register linkage study
British Medical Journal
Suicide and unemployment in young people.Analysis of trends in England and Wales 1921–1995
British Journal of Psychiatry
Method availability and the prevention of suicide—a reanalysis of secular trends in England and Wales 1950–1975
Social Psychiatry and Psychiatric Epidemiology
Excess mortality of mental disorder
British Journal of Psychiatry
Trends in marriage and divorce in England and Wales1837–1987
Population Trends
Relative mortality from overdose of antidepressants
British Medical Journal
Recent trends in elderly suicide rates in England and Wales
International Journal of Geriatric Psychiatry
Cited by (208)
The ethics of flourishing or failing: Social, economic and environmental determinants of global mental health in an uncertain future
2020, Global Mental Health and NeuroethicsFemale labour force participation and suicide rates in the world
2017, Social Science and MedicineThe association between attempted suicide and stock price movements: Evidence from Taiwan
2017, Psychiatry ResearchCitation Excerpt :Suicidal behavior has been viewed as a multifaceted clinical phenomenon shaped by many factors. Epidemiological studies have shown many risk factors: age (Gunnell et al., 2003), gender (Lin and Lu, 2008), psychiatric history (Weiser et al., 2015; Smith et al., 2015), area (Singh and Siahpush, 2002; Tiesman et al., 2015), interpersonal relationships (Ping et al., 2003; Brent and Mann, 2006), marital status (Yeh et al., 2008; Yip et al., 2015), socioeconomic factors (Rehkopf and Buka, 2006; Knipe et al., 2015), season/climate (Lee et al., 2006; Bramness et al., 2015), ethnicity (Liu et al., 2011), disaster (Ohto et al., 2015), unemployment (Fountoulakis et al., 2013; Breuer, 2015), business cycles and economic recession (Luo et al., 2011; Reeves et al., 2012; Norstrom and Gronqvist, 2015). Among these risk factors, economic conditions have drawn considerable attention in recent years, because of the recent rash of financial crises.
More than just numbers: Suicide rates and the economic cycle in Portugal (1910-2013)
2016, SSM - Population HealthEpidemiology and mental illness in old age
2020, BJPsych AdvancesDo suicide characteristics differ by age in older people?
2018, International Psychogeriatrics