94 Suicidal behaviours on all the continents among the elderly

Despite tremendous variability across countries, suicide rates among the elderly, especially elderly males, remain globally the highest. On average, suicide rates increase with age, with the global suicide rate among those aged 75 and over being approximately three times higher than the rate among youth under 25 years of age. This chapter gives an overview of the epidemiology of fatal and non-fatal suicidal behaviour among people over 65 years of age on the five continents. The purpose is to facilitate a better appreciation of the extent of the differences among nations (where possible data from continents are provided), collectively with an example from a country from each of the continents: South Africa, China, Australia, Italy, and Brazil.

Lack of epidemiological data from many nations on mortality and morbidity makes cross-cultural comparisons regarding suicidal behaviours in the elderly a challenging task. Estimation of the actual prevalence of suicide among the older persons should be approached with particular caution due to high frequency of under-reporting, and lack of formal recognition of passive suicidal behaviours (e.g. self-starvation, refusal to take life-sustaining medication, etc.) in reported suicide statistics (De Leo and Diekstra 1990).

However, available data indicates that, despite tremendous variability across countries, suicide rates in the elderly remain globally the highest (Bertolote and Fleischman 2005). On average, suicide rates increase with age, with the global suicide rate among those aged 75 and over being approximately three times higher than the rate among youth under 25 years of age.

However, the pattern of suicide rates increasing with age is not found in every nation. Today, approximately one-third of the countries reporting mortality data to the World Health Organization present higher rates in young adults (males) than in subjects over 65 years of age. In addition, over the last thirty years, countries of different cultures have registered an increase in youth and young adults' suicide that has been paralleled by a decline in elderly rates (Alte da Veiga and Saraiva 2003; De Leo and Evans 2004; Pritchard and Hansen 2005). The decrease has been particularly relevant in Anglo-Saxon countries, possibly in relation to increased economic security of the elderly, changing attitudes toward retirement, improved social services and better psychiatric care (De Leo 1998, 1999; De Leo and Spathonis 2003). The decline has not been confined to elderly males only. For example, in Australia, women who are aged 65–74 years today, have rates of suicide that are four times lower than those in 1965 (De Leo 2006). In a number of Latin American countries, as well as in some Asian nations, suicide rates in old age present less favourable trends. In these countries, economic, social changes and the collapse of traditional family structures may have contributed to the high elderly suicide rates (Pritchard and Baldwin 2000; De Leo 2003; Rudmin et al. 2003).

Worldwide, suicide is most prevalent among male subjects in the 75 years and older age group (De Leo and Heller 2004). Particularly in the Western world, this seems to contrast with the poor health and social status experienced by elderly women that result from more compromised psychophysical conditions secondary to greater longevity, poverty, widowhood, and abandonment. To explain this difference, it has been suggested that women might benefit from better established social networks, grater self-sufficiency in activities of daily living, and commitment to children and grandchildren (De Leo 2003).

Fatal and non-fatal suicidal behaviour exhibit opposite tendencies with respect to age; while suicide rates peak in the elderly in most nations, attempted suicide decreases with advancing age virtually everywhere else (De Leo et al. 2001; De Leo and Spathonis 2004). In the elderly, estimated ratios between attempts and completions vary from 4:1 (McIntosh 1992) to 2:1 (De Leo et al. 2001), whereas in young persons, they can reach the level of 100–200:1 (McIntosh 1992). Results of the WHO/EURO Multicentre Study on Suicidal Behaviour indicated the highest prevalence of suicide attempts in younger age groups, with the frequency decreasing with age for both sexes. In fact, subjects between the age of 15 and 34 years accounted for 50 per cent of all attempts recorded in hospitals, while the elderly (65+) comprised only 9 per cent of all suicide attempters (De Leo et al. 2001; Schmidtke et al. 2004). Similar age-related trends in non-fatal suicidal behaviour were observed in the recent WHO/SUicide PREvention-Multisite Intervention Study on Suicidal Behaviours (WHO/SUPRE-MISS) (Fleischmann et al. 2005). In this investigation only 1 per cent of all attempters were over the age of 65.

Literature converges in depicting the elderly as subjects whose suicide attempts usually involve the highest suicide intent scores; their acts are often carefully planned and are less impulsive than in younger individuals. Methods tend to be more violent, in general, and therefore have less opportunity for rescue (Pearson and Brown 2000; Caine and Conwell 2001; De Leo et al. 2001, 2002; Conwell et al. 2002). Also, alcohol is less involved in elderly suicidal behaviours compared to their younger counterparts (Neulinger and De Leo 2001). It has been reported that in old age, 85 per cent of subjects take their own lives on the first attempt (Suominen et al. 2004). Among factors contributing to the high rate of death subsequent to an attempt are: decreased healing abilities, frailty compounded by inability to survive the physical injury, and social isolation (Conwell et al. 2002).

There have been relatively few population studies looking at the prevalence of suicidal ideation in old age. Differences in methodology (e.g. timeframe, type of questions, and age of surveyed participants) make comparisons difficult (Forsell et al. 1997; Scocco et al. 2001). However, results indicate that between 2.3 per cent (Jorm et al. 1995) and 15.9 per cent (Skoog et al. 1996) of individuals over the age of 65 had experienced suicidal ideation in the month preceding the interview. In an Italian study by Scocco and De Leo (2002), almost one in ten participants over the age of 65 reported death wishes or suicidal ideation over the last 12 months. In the Gold Coast Survey on Suicidal Behaviour (De Leo et al. 2005), the lifetime prevalence of suicide ideation was 5.3 per cent in subjects aged 65–74 years, and 5.6 per cent in those aged 75+, percentages that are approximately half the frequency expressed by people below the age of 55 (in the 55–64 group the percentage was 8.1). In Taiwan, Yen et al. (2005) found that 16.7 per cent of respondents (65–74 years old) had thoughts of suicide within the past week. This study is of special interest, given that the majority of research on the subject has been conducted in Western countries, and there is scarcity of data regarding the prevalence of suicidal ideation among the elderly in other cultures (Chou et al. 2005; Fujisawa et al. 2005).

International studies of the prevalence of suicidal ideation and suicide attempts across the lifespan lead to the conclusion that social and cultural factors might contribute to the observed variations in suicide rates among nations (De Leo and Spathonis 2003). Differences in socio-economic status, social systems, and types of health care, as well as ethnicity, religion, and traditions, can provide important insights to a better understanding of suicide and its prevention (Hawton et al. 2001; Rudmin et al. 2003; Berk et al. 2005; Wu and Bond 2006).

In order to facilitate a better appreciation for the extent of the differences in suicide rates among nations, data from all contients are given where available (Table 94.1), as well as an example from a country from each of the five continents. Data on elderly suicide rates in China are reported and discussed here due to the unique gender-, age- and urban/rural patterns of suicide in the country. The brief discussion on elderly suicide in South Africa underlines the role of sociocultural and political factors contributing to suicidal behaviour among the elderly.

Furthermore, data from the WHO/ SUPRE-MISS investigation allows for some unique insight regarding the prevalence of non-lethal suicidal behaviours among the elderly in participating countries.

Injuries, including suicide, are common and increasing in many developing countries, including Africa; however, their incidence and trends are understudied and poorly known (Nordberg 2000). Most countries in Sub-Saharan Africa do not have compulsory registration of deaths, and in many cultures in the region, suicide remains a taboo subject, making it unlikely that the relatives of people dying by suicide will report the true cause of death (Gureje et al. 2007). Several studies have looked at suicide among the elderly in South Africa (which are reviewed below), and there is survey data regarding the prevalence of suicide ideation and attempts across the lifespan in the general populations of Ethiopia (Alem et al. 1999; Kebede and Alem 1999) and Nigeria (Gureje et al. 2007). Studies conducted in Ethiopia indicated that the majority of self-reported suicide attempts occurred when the respondents were between 15 and 24 years of age, with the frequency of such behaviour decreasing with age (Alem et al. 1999). Moreover, the lifetime prevalence of suicide ideation was decreasing with age; for example, suicide ideation among individuals aged 60 years and older was 68 per cent lower than young adults under the age of 25 (Kebede and Alem 1999). Gureje and his colleagues (2007) found that in Nigeria, respondents in the youngest age group (18–34 years old) were significantly more likely to have ever thought about suicide, made a suicide plan or attempted suicide than respondents over the age of 65.

About 10,000 people die by suicide in South Africa every year, and the problem of suicide is increasing, especially among young people (Meel 2006). In the first national longitudinal study of suicide trends in the country over the period of 1968–1990, Flisher et al. (2004) reported that 1.1 per cent of all deaths were attributable to suicide, with the white males having the highest suicide rate in the country (33.1 per 100,000). Suicide rates in the elderly showed a significant variation depending on sex and race: the white male suicide rate was the highest (41.2 per 100,000), while the rate among the male Coloureds of Asian and African ancestry was significantly lower (7.7 per 100,000). Similarly, the elderly female suicide rate was the highest among Whites (7.1 per 100,000) and the lowest among the Coloureds (0.9 per 100,000).

To date, there is scarcity of data on the epidemiology of non-fatal suicidal behaviour in the general population in South Africa, and among the elderly in particular. Data from the WHO/SUPRE-MISS study showed a trend similar to the one found in other countries; the majority of suicide attempters in Durban were young, and the elderly only accounted for 0.3 per cent of all attempters (Fleishmann et al. 2005).

Since the 1990s, South Africa has been undergoing serious political transition, transforming the social world of the country, with the elderly appearing as the least adaptable to these challenges. Due to the ethnic variation of the South African population, any analysis of suicide trends in the country, including elderly suicide, has to take into consideration the relevant cultural and socio-economic factors, including decreased resilience and unmet expectations regarding quality of life among the Whites, cultural taboos against suicide among the Blacks, and religious sanctions against suicide among the Asians and the Coloureds (Flisher and Parry 1994; Flisher et al. 1997; Scribante et al. 2004).) Also, the high suicide rates may be related to the increase in the elderly population, leading to reduced resources available for this age group, paralleled by lack of suicide prevention initiatives, discrepancies in government expenditures on healthcare resources and social security (Burrows et al. 2003; Flisher et al. 2004).

Given the geographical, cultural, political and socio-economic diversity of Asian countries, often accompanied by scarcity of research and epidemiological data on suicide in many nations, it is difficult to give a concise overview of elderly suicide in the continent. Available evidence suggests that there are significant differences in age-related suicide patterns between Asian and Western countries. In Western countries, rates of suicide are high in the 15–25 age groups and highest in the elderly, while in Asia those under the age of 30 often have the highest suicide rates (Vijayakumar 2005; Vijayakumar et al. 2005a, b). Nonetheless, for the period 1999–2001, suicide rates among elderly males in China (59.2 per 100,000 in 1999) and South Korea (55.4 per 100,000 in 2000), and elderly females in China (63.1 per 100,000 in 1999), Hong Kong, SAR (26.6 per 100,000 in 1999), Japan (26.0 per 100,000 in 1999) and South Korea (23.0 per 100,000 in 2000) were among the highest globally (De Leo and Evans 2004). Given the unique suicide patterns of China, with regard to gender, age and urban/rural characteristics, a detailed analysis of elderly suicide in the country is presented in the next section.

China accounts for more than 30 per cent of the world's number of suicides; however, national suicide mortality rates have been decreasing from 22.9 per 100,000 in 1991 to 15.4 per 100,000 in 2000 (Yip 2001; Yip et al. 2005). When compared to Western nations, China displays significant disparities in age- and sex-specific rates in rural and urban parts of the country (Phillips et al. 2002). There are reported similarities among Chinese elderly suicides of Beijing, Taiwan, Singapore and Hong Kong, SAR; however, the urban suicide rate is lower in Beijing than in other Asian cities (Yip and Tan 1998; Yip et al. 1998; Tsoh et al. 2005; Chan et al. 2006).

A study of the epidemiological profile of suicides in Beijing over the period of 1987–1996 indicated that suicide rates were increasing with age. Rural males over the age of 75 had the highest suicide rates (109.1 per 100,000), twelve times higher than among rural youths (15–24 years old) (Yip 2001). Among urban males of 75 years of age and older, rates were approximately three times lower (38.3 per 100,000).

In parallel with age trends in non-fatal suicidal behaviour observed in other countries, the WHO/SUPRE-MISS study results showed that the majority of suicide attempters in the Chinese site (i.e. Yuncheng) belonged to the younger age groups (the mean age of females was 30 and 33 for males) (Fleischmann et al. 2005), while the older attempters over the age of 65 accounted for 4.2 per cent of all recorded cases.

In China, the problem of elderly suicide is particularly serious, due to the lack of effective welfare systems (including pension programmes), increasing poverty of the elderly, and sociocultural transitions taking place in the country (He and Lester 2001; Ji et al. 2001; Yip et al. 2005). Traditionally, the younger generations respect and take care of the elderly; however, the one-child policy introduced in the 1970s, and the ongoing disintegration of the three-generational family in most parts of the country, have led to increasing social isolation among the elders.

In Australia and New Zealand, similarly to the other countries in the New World (i.e. Canada and United States), suicide mortality among males 65 years of age has been declining since the late 1980s, following an increase in the early 1980s (De Leo and Evans 2004). Suicide rates of elderly females have been steadily declining over the last two decades, with the trend most notable in New Zealand. In both countries, suicide rates among the elderly males and females are relatively low compared to the elderly suicide rates (65–74 and 75+ age groups) in other parts of the world (for details see Table 94.1, and a detailed analysis of elderly suicide in Australia is presented in the next section). Unfortunately, due to the lack of suicide mortality and morbidity data in the Pacific Island countries, it is not possible to ascertain the incidence of elderly suicide in that region. However, data available from Fiji, Guam and Western Samoa indicates that from 1988–1992, suicide rates were the highest in the 15–24 age group in both males and females, and declining with age (Booth 1999). A similar pattern was also reported in Micronesia from 1960–1987 (Booth 1999).

According to official data in 2005, Australia's national suicide rate was 10.3 per 100,000, and suicide comprised 1.6 per cent of all deaths (ABS 2007). In 2005, the suicide rate among the elderly in the 65–74-year-old group was below the national rate (i.e. 9.6 per 100,000), and the suicide rate in those aged 75+ was higher (11.8 per 100,000). However, from the mid-1960s to 2002, mortality rates for those 75+ years of age in both genders have halved (De Leo and Heller 2004).

In Australia, among other New World Anglo-Saxon countries, suicide mortality rates in older males (75+) have been in decline since late 1980s, but sharper declines have been observed in elderly females, especially those aged 65–74 (Cantor et al. 1999; De Leo and Evans 2004). Simultaneously, there has been an increase in youth and young adult suicides (Snowdon and Hunt 2002; Pritchard and Hansen 2005).

Apart from improved financial status and psychosocial assistance, the decrease in elderly suicide rates in Australia and other Anglo-Saxon countries has been attributed to improved health, increased life expectancy, and higher levels of activity, as well as to reduced feelings of redundancy among the older generation. Today, ‘the over 65s, relative to previous generations, are more affluent and healthier than ever before’ (Pritchard and Hansen 2005, p. 22). It has also been speculated that limited access to lethal means used in suicide could have played a role in this process (Pritchard and Hansen 2005; De Leo 2006).

Results from a large population survey conducted in Queensland indicated that the lifetime prevalence of suicidal ideation and behaviour was the highest in the 25–44-year-old group, and declined with increasing age (De Leo et al. 2005), a result similar to studies performed in Europe and the US (e.g. Diekstra and Gulbinat 1993; Kessler et al. 1999). In the survey, 6.1 per cent of respondents in the age group 65–74 years reported having seriously considered suicide in the course of their lives, 2.4 per cent reported having planned it, and 2.3 per cent reported having attempted suicide (results for the 75+ age group were 6.9, 0.3 and 1.6 per cent, respectively) (De Leo et al. 2005).

A recent analysis of epidemiology of suicide among the elderly in 19 European countries between 1999 and 2001 has revealed similar rates and trends in nations clustered according to their sociopolitical and geographical similarities (De Leo and Evans 2004, also see Table 94.1 for the most recent data available to the WHO). The highest suicides rates among males 65 years and over were found in Eastern European nations (i.e. Lithuania, Latvia, Russian Federation, Hungary and Bulgaria; between 56.8 and 94.8 per 100,000 in Bulgaria and Russia in 2000, respectively), and some parts of Western Europe (i.e. 61.3 per 100,000 in 2001 in Austria, 55.6 per 100,000 in 1999 in France, and 55.4 per 100,000 in 1999 in Switzerland). The lowest elderly male rates were reported in the Old World countries: Northern Ireland (7.8 per 100,000 in 2000), England and Wales (11.6 per 100,000 in 2000), Scotland (13.2 per 100,000 in 2000) and Ireland (17.6 per 1000,000 in 1999). The pattern of suicide rates in elderly females was somewhat different; although, the highest rates in Europe were still found in Eastern European countries (e.g. 28.2 per 100,000 in Hungary in 2002 and 24.5 per 100,000 in Lithuania in 2000), the lowest rates were reported in southern Europe: Greece (2.9 per 100,000 in 1999), Portugal (5.5 per 100,000 in 2000), and Italy (6.1 per 100,000 in 1999).

Suicide rates in Italy have been rather stable over the last two decades, with overall rates around 6 cases per 100,000 (ISTAT 2004). A detailed analysis of suicide trends over the period of 1887–1993 indicated a significant decrease in suicide rates for young people (15–24 years old), especially in females, and a rise in suicide among individuals over the age of 55 (De Leo et al. 1997). Over the last century, suicide rates of the elderly exhibited a rather stable trend with a marked increase at the end of the 1970s (with a peak of 23.1 per 100,000 in 1985), followed by relatively stable high rates of over 20 per 100,000 from 1985 onwards (ISTAT 2004).

Tatarelli et al. (1999) pointed out the non-uniformity of the nationwide distribution of suicides in Italy, an observation confirmed by recent analyses of elderly suicide mortality in the southern (Pavia et al. 2005) and northern parts of the country (Zeppegno et al. 2005).

De Leo et al. (1997) noted a discrepancy between the recent increase in the ageing Italian population and a relative lack of public health and social services to meet the needs of this group. The changing family structure, by moving away from the traditional patriarchal three-generational model providing protection, and emotional and economic stability, has resulted in loss of spontaneous support for the elderly. Conversely, the development of institutionalized social and health facilities, which is able to compensate at least in part for the loss of traditional support, is slow (Pritchard and Baldwin 2000). The structural changes in society and the family resulting from industrialization and urbanization (including lack of job opportunities) led to an altering in the social role played by the elderly, both for males and females, and their increasing isolation. As a result, the elderly males find that taking care of themselves has become increasingly challenging (as in other Latin American countries), while ageing females lose their traditional function in the family, and share the fate of loneliness and institutionalization with men.

The elderly suicide rates for both genders in Canada and United States have been declining since the end of the 1980s, and although suicide rates in elderly females in both North American countries are among the lowest globally, the male rates are elevated (De Leo and Evans 2004). Furthermore, suicide rates among the elderly (especially males) are among the highest across the lifespan, for example, 15.6 per 100,000 in 2002 for the 65+ age group (both males and females) in the US (McKeown et al. 2006), and 26.9 and 7.5 per 100,000 (for males and females, respectively) in British Columbia, Canada during 1981–1991 (Agbayewa et al. 1998). Unfortunately, there seems to be a scarcity of research and updated epidemiological data regarding suicide, including elderly suicide, in many South American countries.

Brazil has one of the lowest suicide rates in the world (4 per 100,000 in 2000); however, in terms of absolute numbers of deaths, Brazil is among the ten countries with the highest number of suicides—over 6,000 per year (De Mello-Santos et al. 2005).

The analysis of suicide trends in Brazil during 1980–2000 shows that suicide rates have increased from the 45–54 age group onward, with individuals over the age of 65 representing the subpopulation with the highest suicide rates (De Mello-Santos et al. 2005). In 2000, the suicide rate for the age range 65–74 was 6.9 per 100,000 (12.1 in males and 2.5 in females), and 7.2 per 100,000 in people of 75+ years of age (14.2 in males and 2.1 in females).

While rates of fatal suicidal behaviour in Brazil follow the pattern characteristics for Latin American countries (i.e. the risk for suicide increasing with age), data from the WHO/SUPRE-MISS study showed that suicide attempters (in the catchment area of Campinas) concentrate mostly among young adults of both sexes. The elderly recorded only 1.2 per cent of all cases of non-fatal suicidal behaviour (Fleischmann et al. 2005).

Rodrigues and Werneck (2005) reported that, compared to younger generations, older Brazilian cohorts are at higher risk of suicide, in both genders. Worse, early life conditions experienced by today's elderly people (wars, urbanization, industrialization, and serious political changes) were suggested as a possible explanation for this phenomenon (Rodrigues and Werneck 2005).

Understanding suicide and suicidal processes poses specific challenges at different stages across the lifespan; it appears more difficult in the elderly, due to the relatively limited numbers of behaviours (i.e. less suicide attempts, less repeats, more masked intent), and to a perhaps increasing attitudinal divide between the old and the young (De Leo et al. 2001). The task of deepening the knowledge on elderly suicide and its prevention is even more challenging in the global perspective, given the lack of epidemiological data from many countries, the complexity of cultural, ethnic and socio-economic factors involved, and the generally insufficient attention that elderly suicide has attracted so far.

Most research on suicide in older adults has been conducted in Western countries, and only a few studies have tried to examine this phenomenon and its correlates across the five continents. Recently, this situation has been slowly changing with new information deriving from studies conducted in Asian countries, including China (Tsoh et al. 2005), Hong Kong, SAR (Chiu et al. 2004), India (Abraham et al. 2005), Japan (Awata et al. 2005), and Taiwan (Yen et al. 2005). Results of these investigations show that there are a number of similarities between the Western and Asian countries in regard to risk factors, including depression, physical illness and previous history of suicidal behaviour (De Leo et al. 2002; Snowdon and Baume 2002). However, other studies indicate that the psychosocial dynamics of suicide outside the Western context are different (Chan et al. 2004), but seem to be rapidly changing as a result of the Westernization of traditional cultures (Pritchard and Baldwin 2000, 2002; Chiu et al. 2004; Yip et al. 2005).

Cultural values and attitudes, as well as the legal and political context in many countries, lead to methodological problems in the study of elderly suicide, as it is a case of criminalization of suicide as well as of attempted suicide, for example in India (Fleischmann et al. 2005). Suicide remains a taboo subject, not only in Hong Kong, SAR (Chiu et al. 2004), but also in many other countries where negative attitudes towards persons with suicidal behaviours continue to exist (see Part 1 of this book).

Despite an epidemiological dimension that should not permit any neglect, suicide in old age still attracts very little attention from media (which focuses much more on youth suicide) and public health planners. As a consequence, elderly suicidal behaviour has been the target of a small and barely relevant number of preventive programmes. This may reflect some of the many nihilistic aspects of the ageism culture, which, for example, considers depression as a normal/natural feature of the old age, elderly suicide as understandable, rational, and unpreventable (Draper 2006), and older adults as second or third rank citizens. In the words of Snowdon and Baume: ‘it may be that the health strategists have formed a view that suicides in late life are often understandable, and little can be done to prevent them’ (2002, p. 261). Clearly, elderly suicide prevention should start from fighting this very perspective.