Two of the most commonly cited risk factors for suicide are older age and male gender. Among the industrialized countries that provide statistics on suicide, all report that men commit suicide more frequently than women (see Mościcki, this issue). In a majority of those countries, suicide rates rise progressively with age; in all but one, the highest suicide rates occur for men age 75 and older (Table 1; based on countries reporting statistics to the World Health Organization from 1988 to 1991 [WHO, 1990, 1991]). Despite the dramatic effects of gender and age on suicide risk, relatively little attention has been paid to the risk factors for and clinical profiles of late-life suicide. Given the progressive aging of populations around the industrialized world, and assuming that the relationship between older age and suicide will persist, the absolute number of elderly suicides is likely to increase dramatically in coming decades.

This article presents the epidemiology of suicide with a special focus on suicides among the elderly, and discusses the known risk factors for suicide within a framework designed to encourage a systematic approach to theory testing and prevention. Throughout the world, suicide rates are highest among the elderly. The risk factors for suicide can be classified as distal or proximal, and, within these broad categories, as sociodemographic, psychiatric, biological, familial, and situational. Mental and addictive disorders are the major risk factors for suicide in all age groups. Other risk factors include male gender, disrupted marital status, prior suicide attempt, reduced brain stem serotonergic activity, family history of psychiatric disorder or suicide, a firearm in the home, and a recent, severely stressful life event. Since risk factors for suicide rarely occur in isolation, prevention efforts are more likely to succeed if multiple risk factors are targeted.

Rates of suicide differ markedly as a function of age and gender. In a great majority of countries that report suicide statistics, rates are higher in the elderly than in younger age groups. Few data are available to determine whether this age-related pattern of rates is reflected in a differential prevalence of other factors known to be associated with increased suicide risk. Using the psychological autopsy method, researchers have established the distribution of psychiatric diagnoses in samples of adolescents, young adults, and elders who have committed suicide, and in victims across the life course in community-based populations. Review of those studies reveals that affective disorders, substance use disorders, and their comorbidity are common in completed suicides of all ages. Although methodologic issues and cohort and period effects limit the conclusions that can be drawn, comparison by age within and between studies further suggests that depressive illness is more common and substance abuse less prevalent in later-life victims. That the aging process itself (biological, psychological, and social) may account for the observed pattern has important implications for further research and the design of intervention strategies. There is a need for more coordinated and comprehensive study of psychopathological correlates of suicide across the life course in a broad range of cultures.

There is an accumulating body of research suggesting that suicidal behavior may be associated with abnormalities of the central serotonin system. Other monoaminergic, peptidergic, and neuroendocrine systems have been implicated as well. A review of studies that examine neurobiological variables in postmortem tissue of suicide victims and controls reveals that investigators in most instances have viewed age as a confound to be controlled rather than considering it as a variable of interest. However, the close associations between aging and increased suicide rates, and the knowledge that the functional integrity of many of these same systems changes with normal and abnormal aging processes, raise the possibility that biological aging contributes to suicide risk, and that the underlying neurobiology of suicide in the elderly differs from that of younger people. The few available studies that did examine the association of age with neurobiological measures indicate directions for future research into the role that aging may play in determining the biological bases of suicide risk.

Personality data gathered in a psychological autopsy study indicate that completed suicides obtain higher neuroticism scores than age- and gender-matched controls. Older suicide victims obtained lower openness to experience (OTE) scores than both younger suicide victims and normal controls. Although the role of neuroticism and negative affect in psychopathology has been discussed extensively, OTE has been accurded relatively little attention in the psychiatric literature. The apparent role of OTE in completed suicide warrants its closer examination. In this article, I offer the testable hypotheses that persons low in OTE are at risk for taking their own lives because their affective dampening, cognitive certainty, diminished behavioral repertoire, and rigidly defined self-concept have decreased their capacity to adapt to the expectable age-associated changes in role, health, and function that accumulate over time. Concrete thinking and excessive focus on proximal, low-level goals place them at risk for descending into states of suicidal meaninglessness in times of stress or crisis. This state of awareness increases the desirability of suicide as an action oriented solution to the stressors of aging. Other lines of investigation are suggested. These include research on OTE in attempted suicide; OTE and the neurobiology of suicidal behavior; OTE and gender differences in suicidal behavior; and clinical intervention designed to increase the degree to which one is open to experience.

This article explores how a life-course perspective and narrative methodology can be used to study risk factors for late-life suicide. A life-course approach to aging and suicide requires consideration of age as both social and personal construction. “On-” and “off-time” events and their impact on adjustmenta are used to illustrate these social and personal constructions. Cohort, period, and histrorical events have potentially profound effects on risk for suicide, yet the study of these effects is difficult because they are so often confounded in longitudinal study. Lifelong personality characteristics that are not life-threatening in earlier life may be of greater risk in later life depending on life circumstances such as physical dependencies. A life-story or narrative approach offers an alternative method for incorporating these complicated factors when studying late-life suicide. The psychological autopsy can be considered a type of “narrative” used by various individuals to gain understanding about a suicide.

In a 1-year period, 6 of 11 suicides in a rural Wisconsin county were committed by people over the age of 60. At the request of the local coroner, the American Medical Association sent a team of investigators to perform psychological autopsies. Family member survivors, friends, and attending physicians were interviewed. The majority of suicide victims had evidence of major depressive episodes, delusions that they had a terminal physical illness, and knowledge of family/friends who had committed suicide. Most had seen their physician within 48 hours before their death. Continuing education efforts focused on increasing a physician's abilities to recognize depression and suicidal ideation have ensued, based on this experience.

The limited attention paid to old-age suicide in Germany is likely related to society's lack of interest in the specific problems of old people. This is despite the fact that over half of all German suicides are committed by persons 65 and older. In reviewing what is known about elderly suicide based on retrospective studies, I submit that suicide in later life is not always based on pathologic mental processes. There may be a number of risk factors unique to elderly suicides, in addition to depressive disorders that involve broader cultural issues among the many motives likely present in any one suicide. Today's emphasis on individualized values may lead to a feeling of meaninglessness in the suicidal elderly. The uncertainty and fear of the inability to influence their own dying and a certain weariness of life are also likely unique risk factors for the elderly. Thus the reasons or motives for a suicide may be best considered by understanding the entire life situation and the biographical aspects of the person. Prevention of elderly suicides requires a number of approaches ranging from social assistance and improved training of care providers to more acceptance and valuing of older persons by society.

Although individuals aged 65 and over accounted for 12% of the total population of Japan in 1990, suicides in this age group consisted of 29% of all suicides. The elderly population of Japan is expected to grow rapidly to 24% of the total population by the year 2020, and suicide prevention for the elderly is an urgent mental health problem. Among a total of 1,216 elderly patients who were admitted to the Department of Psychiatry at Tokyo Metropolitan Geriatric Hospital between April 1980 and March 1990, 50 were hospitalized immediately after suicide attempts in order to study their psychosociomedical problems. Because early diagnosis of depression and initiation of proper treatment are indispensable—even if patients do not show obvious depressive symptoms—those who develop persistent somatization and/or delirium should be considered highly suicidal and given special attention. With the cooperation of mental health professionals, it is necessary to educate general practitioners, the public, and the elderly themselves about characteristics of psychiatric disorders and various problems associated with aging. Further research on other elderly Japanese populations, as well as research that examines suicide completion, is needed to confirm the findings of the present study.

It is estimated that in the year 2020, approximately 25% of the Japanese population will be over 65 years of age. Moreover, suicide is a significant public health problem in Japan, where more than 6,000 elders take their lives each year. The authors compare late-life suicide in urban Kawasaki with suicide among the elderly in rural Higashikubiki over a 12-year period, from 1979 through 1990. The suicide rates in Kawasaki were lower than for Japanese elders as a whole, whereas those in rural Higashikubiki were extraordinarily high. The most frequent method used in both areas was hanging, and none of the victims died of gunshot wounds. In Higashikubiki, almost two thirds of victims lived in a three-generation family and none lived alone. The change of the family system from the traditional extended family to the nuclear family is accelerating, especially in rural areas. The authors suggest that the greatly elevated suicide rates among the elderly in Higashikubiki, and in rural regions of Japan more generally, result from these rapidly occurring changes in traditional social structure.

This article reviews British government policies regarding the reduction of elderly suicide. Epidemiologic trends in elderly suicide indicate males aged 75 and older have the highest rates, with frequent methods being hanging and overdose. In contrast, rates among elderly who have immigrated from the Indian subcontinent in the past 10 to 30 years are significantly lower. Characteristics of elderly suicides and deliberate self-harm in the United Kingdom suggest that depression, social isolation and loss, and physical illness all are risk factors. Because most elderly suicides have contact with a primary healthcare provider in the month prior to the suicide, an effort to educate general practitioners and psychiatrists about treatable depression in the elderly is being made. Other policies also are being implemented: reduction of means of suicide, better linkage among services for the elderly for enhanced appropriate care, and clinical audits of suicide. Research associated with these policy efforts may further elucidate clinically relevant risk factors.

The purpose of this study was to investigate the prevalence and comorbidity of current mental disorders defined by DSM-III-R among elderly suicide victims and to compare them with the mental disorders among younger victims. Using a psychological autopsy method, we collected comprehensive data on all suicides in Finland during 1 year. Retrospective Axis I-III consensus diagnoses were assigned to a random sample consisting of 43 victims aged 60 years or over and 186 victims aged under 60 from the nationwide suicide population. At least one Axis I diagnosis was made for 91% of the elderly victims. Major depression as the principal diagnosis was more common among the elderly victims. Almost all elderly female victims were major depressives. Psychiatric comorbidity was more common among elderly male than among elderly female victims. More of the elderly victims (88%) than the younger (36%) received Axis III diagnoses. Suicide among the elderly without a diagnosable mental disorder and somatopsychiatric comorbidity seems to be rare.

The primary objectives of this study were to examine recent stressful life events and to explore possible social interaction factors across age groups in a population of suicides. Life events during the last 3 months preceding suicide (32 items) and factors related to social interaction (6 items) among 219 suicide victims aged 60 years and older were compared with those of 803 victims aged 20 to 59 years. The study population comprised all adult suicides who had life events interview data assessed as reliable by the interviewers in a nationwide total suicide population (1,397) of 1 year in Finland.

Differences in life events were found across age groups: Family discord, loss (separation, death), financial trouble, job problems, unemployment, and residence change were more common among younger victims, and somatic illness was more common among elderly victims. Age groups were similar with regard to proportions of persons living alone, availability of confidants and friends with common interests, and reports of loneliness. Similarly, few differences were found among persons aged 60–74 compared to age 75+ with regard to rates of life events, living alone, opportunities for social interaction, and complaints of loneliness. Loneliness was reported for persons who experienced loss among both younger and older suicides. In terms of age differences, younger men were more likely to have experienced job-related events,whereas older men were more likely to have had a somatic illness.

Somatic illness appeared to be the most important stressor in elderly suicides, particularly for men. Living alone and diminished opportunity for social interaction were not common factors in late-life suicides. Controlled studies with age matched comparisons are needed to further investigate how life events and social interaction factors vary in their risk for suicide across the life course.

Recent major political changes in Eastern European countries might have affected their suicide rates. For this article, suicide statistics available to the World Health Organization were used to compare data from eight Eastern European countries to those from seven countries in Northern or Western Europe. Comparisons were made between 1987 and 1991/92 data using total suicide rates for each country, rates by gender, and rates for the elderly (age 75 and older). The total rates indicated an increase in suicide in Eastern European countries and a decrease in other European countries. The ratio of male-to-female suicides in the Eastern European countries increased during this time as well, more than in other European countries. Among those over the age of 75, however, rates of suicide in Eastern European countries decreased; this pattern was less clear in the European countries chosen for comparison. The article discusses the role of economic, cultural, and health service factors affecting these trends.

In the cosmopolitan city of Singapore the annual suicide rates in the general population from 1985 to 1991 remained fairly constant, with a mean of 15.3 per 100,000. It was highest among Indians (19.5 per 100,000), followed by Chinese (16.2 per 100,000) and Malays (2.3 per 100,000). The suicide rates were higher in elderly people (aged 65 years and over) than in younger age groups (10 to 64 years) and in males than in females. For the elderly, the mean annual suicide rate for this period was 52.0 per 100,000. However, it was highest among Chinese, with 59.3 per 100,000, followed by Indians at 33.9 per 100,000, and, again, lowest among Malays, with 3.0 per 100,000. Possible sociocultural factors are proposed to account for differences in suicide rates among these ethnic groups.

Australia is a multiethnic society with 25% of its elderly population born outside the country. Rates and methods, gender, and country of birth are given for all suicides aged 65 and over in Australia during the 12-year period of 1979–1990. There was a marked heterogeneity in rates and methods of suicide among the various migrant groups. The data also showed that the suicide rates of migrants were mostly higher than in their country of origin. Migrants born in countries with high suicide rates generally had high rates in Australia, and vice versa. Possible reasons for this finding are discussed. Factors influencing rates and methods of suicide in elderly migrants appeared to have much in common with factors operating in migrants of all age groups.

Beyond its empirical importance documented through epidemiological data, suicide is of symbolic importance due to its close relationship with the values of life and death, privacy and openness, responsibility and reliance, and suffering and resistance. It appears, in daily life circumstances, to reveal or remind us of the limits of our perception of others' intimate world and of our fundamental solitude in the face of life and death. The issues associated with suicide, particularly with elderly suicide, must be confronted. Researchers, clinicians, and planners, among others, must take care not to close the chapter prematurely on suicide by pretending to have discovered “the” final cause of suicide, in a kind of scientific exorcism.

Suicide is an action that calls for explanation. Many who kill themselves leave written words behind, attempting to explain, but many others do not. For those outside the experience, there remains a need to understand. Self-inflicted death may make sense occasionally, however painful it may be to those left behind. Most often, it is called a tragedy, death before one's time.