On March 14, 1932, George Eastman, founder of the Eastman Kodak Company in Rochester, NY, and one of the leading philanthropists of his time, killed himself at the age of 77 with a bullet to the chest. His suicide note read simply, “My work is done. Why wait?”
Eastman’s death was a great shock to his friends and community, who had known him as a man of power, generosity, and self-determination. News reports from the time suggest that many took comfort in the interpretation that his suicide was as straightforward as Mr Eastman’s note implied. As expressed by close associates, “George Eastman played the game to the last. By his own hand he lived his life, and by his own hand he ended it.”1 The reality is more complex and involves physical health and emotional factors, personality vulnerabilities and social stressors, depression and demoralization. Far from being the “exceptional” case of suicide in an older man, George Eastman’s death is emblematic in many respects of suicide in later life, providing illustrative examples of risk factors in each key domain and indicators of opportunities for prevention.
Prevalence rates of suicidal ideation and suicide attempts decrease with increasing age.2 In contrast, incidence rates of completed suicide increase with age for men and women worldwide.3 The United States is somewhat atypical in that suicide rates for women peak at midlife and remain stable thereafter, while the suicide rates for men rise dramatically through old age. The risk is particularly great for white men. Rates of suicide among African Americans peak in young adulthood, drop somewhat through midlife, remain stable to about age 70, and then rise again; at each point, however, the rate remains lower than that of whites. Asian Americans show a pattern similar to that of whites while rates for Hispanic persons are in the mid-range.
Just as for younger persons, older adults who are married are at lower risk for suicide than those who are single, separated, divorced, or widowed.2 George Eastman never married and, except for his attendants, lived alone. Despite being one of the richest men in America, he shared the demographic features of the highest risk group.
CHARACTERISTICS OF SUICIDE Beyond demographic characteristics, risk and protective factors for suicide in older adults have been much more clearly defined in recent years by a series of methodologically rigorous, case-controlled psychological autopsy (PA) studies.4-12 These studies indicate the importance of factors in 5 domains that correspond loosely to Axes I through V of psychiatry’s multiaxial diagnostic system (Figure).
Axis I: major psychiatric illness
Older adults who take their own lives typically do so in the midst of active psychiatric illness. Studies show that 77% to 95% of elderly persons who completed suicide had an Axis I diagnosis.2 Affective disorders are by far the most common (present in 63% to 86% of cases). Alcohol use disorders, other drug disorders, anxiety, and nonaffective psychosis play a far smaller role. Interestingly, no carefully conducted PA study has found dementia to be a risk factor for suicide, perhaps because of the low sensitivity of the PA method to detect early dementia when affective symptoms are most common and cognitive and behavioral changes are easily missed.
Axis II: personality traits
Too few data are available from studies on which to base conclusions about the role of personality disorders in late-life suicide. Instead, emphasis has been placed on personality traits that may make older adults vulnerable to suicide in the face of other risk factors. High neuroticism and low openness to experience (preference for the routine and familiar, a constricted range of intellectual interests, and blunted affective and hedonic responses) have been associated with completed suicides in older adults.13 An introverted style that prevents the development of support networks to mobilize in times of need may also be a factor.14
Axis III: medical illness
Physical illness may increase the risk of suicide in older people, even when the effects of depression are accounted for.9 Illnesses that have been most closely associated with increased risk are cancer, chronic obstructive pulmonary disease, and neurologic illness (in particular, seizure disorders). Data are suggestive for visual impairment, bone fractures, and GI disease.15-17 Many suicides in the face of medical illness are no doubt mediated by depression. The emergence of suicidal ideation, even in the context of terminal illness, should signal the need for screening and assessment for a comorbid affective disorder.
Axis IV: stressful life events
Life events tend to cluster in the days and weeks before suicide in older adults just as for younger persons. The events tend to be those associated with aging, such as bereavement, other interpersonal losses, and social isolation.18 Controlled PA studies have consistently shown that social isolation, family discord, and financial strains are associated with suicide in later life.5,6,10
Axis V: functional status
Whereas Axis V is typically reserved for emotional functioning, we broaden the construct here to acknowledge the importance of physical functioning in late-life suicide. Case-controlled PA studies show significantly greater impairment in the conduct of instrumental and basic activities of daily living for suicide decedents than for controls.8 As previously noted, cognitive impairments have not been directly linked to suicide in later life,4,12 but most clinicians would argue for careful monitoring for the emergence of suicidal ideation whenever mental status changes are evident.
Having a rich social support network, and, in particular, friends or relatives in whom one can confide is associated with lower suicide risk.17 Similarly, religious practice and higher ratings on spiritual values are associated with lower suicide rates.17,19
George Eastman experienced a slow course of deteriorating physical function and increasing pain over the last 2 years of his life.20 Suffering with a degenerative spinal disease, he experienced increasing pain and difficulty in ambulation and self-care. For a man whose favorite activities were hunting, camping, and work, the results of his infirmities, including passivity, increased restriction to his home, and dependency, were intolerable. As he became more ill, he withdrew from friends and colleagues into a more socially isolated state.
Unmarried and with few surviving extended family members, several weeks before his own death his despondency was exacerbated by the death of a lifelong friend and business associate whose illness had confined him to bed for the last several years. In the months before his death, Eastman confided to his associates that he felt there was nothing left to live for and on occasion talked of suicide. He withdrew from his business and social ties, and was observed to “shuffle along in great pain, inexplicably weepy and depressed, dragging one foot behind him.”1,20 We lack sufficient data with which to construct a confident retrospective diagnosis of depression, but Eastman may well have been typical of the modal elder suicide in this respect as well. Eastman thus shared many of the characteristics of older people at risk for suicide. He was an older white male, was single, and had a constricting social network. He was physically ill, functionally impaired, in pain, and depressed, with suicidal thoughts and access to a handgun.21 What options would his primary care or mental health provider have to manage this high-risk situation?
Approaches to Prevention
Up to three quarters of older adults who took their own lives did so within 30 days of seeing a primary care provider; 25% to 50% saw their provider in the last week of life.2 Combined with the fact that older adults are underrepresented in the case loads of mental health specialists, these data underscore the importance of developing and implementing clinical management and prevention strategies in the primary care setting.
The US Preventive Services Task Force does not recommend routine screening for suicidal ideation and risk among adults in primary care.22 Nonetheless, the strong association between depressive disorders, stressful life events, social isolation, and suicide in older adults indicates a logical sequence of inquiry. For those seniors experiencing significant stressors or functional impairment, periodic screening for depression using short, simple measures, such as the 9-item Physician’s Health Questionnaire23 (PHQ-9) or the 15-item Geriatric Depression Scale-Short Form24 (GDS-S), is indicated.
Any older adult with clinically significant depressive symptoms should be questioned about suicidal thoughts with a high-sensitivity question, such as “Have you been feeling that life is not worth living?” A positive response should lead to more detailed inquiry about the presence of active thoughts of, plans for, and intent for suicide.
The assessment should be balanced with an evaluation of risk factors in each of the 5 axes noted above.25 These data form the basis for clinical decision making, in collaboration with the patient and caregivers, about whether hospitalization is necessary or whether further assessment and treatment can be managed safely on an outpatient basis. A thorough assessment leading to a comprehensive multiaxial diagnosis, followed by treatment of the associated psychiatric, medical, and social morbidities should resolve the crisis.
Management of acute suicidality in older adults does not differ in most respects from that in younger people (see, for example, the American Psychiatric Association’s Practice Guideline for the Assessment and Treatment of Patients With Suicidal Behaviors26). However, preventive intervention takes on even greater urgency in older people. Although suicidal ideation and attempts are less common in older adults than in younger adults, rates of completed suicide are higher in the older group for several reasons.
Once a self-damaging act is initiated, older adults are more likely to die than younger people because they are more likely to live alone and escape rescue. Also, they have less physical reserve, making any injury more potentially lethal. Finally, older people with suicidal intent are more adept planners and more determined in their acts, choosing more immediately lethal means and implementing them with greater lethality of intent.27
Because suicidal ideation and behavior are so often fatal in older people, prevention requires early and aggressive intervention for persons identified as being at high risk. Every effort should also be made to identify and prevent the development of more urgent or emergent states in older people who are not yet acutely symptomatic but who have characteristics that put them at risk for suicidal states.
The Institute of Medicine, in its landmark report on prevention of mental illnesses, advocated the use of the terms “indicated,” “selective,” and “universal” to describe distinct levels of preventive interventions.28 Their application to suicide prevention is illustrated in the Table, including examples of their operationalization.
Indicated preventive strategies
Indicated preventive interventions target individuals with detectable symptoms or syndromes known to be proximal risk factors for suicide in older people (eg, depressive disorders). These interventions aim to prevent the initiation of a potentially lethal suicidal act. No studies have examined the effectiveness of an indicated preventive strategy in which completed suicide in a population of older adults was the targeted outcome.
Based on findings from the Prevention of Suicide in Primary Care Elderly: Collaborative Trial (PROSPECT), Bruce and colleagues29 reported that the intervention—consisting of patient and provider education, a depression care specialist attached to the primary care practice, and standardized algorithm-driven care that included medications and psychotherapy as indicated—yielded a significantly greater proportional reduction in suicidal ideation than enhanced care as usual. Depression outcomes were significantly improved as well. However, because more than 70% of the sample were women, the study’s relevance to completed suicide prevention in the United States must be questioned. Nonetheless, models of collaborative care for depression and other mental disorders of later life delivered in the primary care setting are likely to be important contributors to reduction of suicide rates in older people.
Selective preventive strategies
Selective preventive interventions target high-risk groups defined by the presence of relatively more distal risk factors than active symptomatic states, such as bereavement or family discord, physical illness, or functional impairment. Again, no selective strategy has been tested in a controlled trial.
The Tele-Help/Tele-Check service piloted by De Leo and colleagues30 in Padua, Italy, offers an indication of potential effectiveness. This service provided telephone access by voice-activated systems placed within the homes of older adults at risk for poor physical and mental health outcomes and referred by social service agencies. Clients received telephone support twice weekly and as needed any time of the day or night. Over an 11-year period, significantly fewer suicides occurred in almost 19,000 service users than would have been expected in a matched population. But again, the effect was demonstrable only in women, an important caveat given the far higher rates of completed suicide in older men.
Universal preventive strategies
Universal approaches to prevention target the entire population irrespective of the risk status of any individual or group. One example of a universal preventive intervention that may have been effective in reducing late-life suicide was the Brady Handgun Violence Prevention Act of 1994. This “natural experiment” offered an opportunity to compare trajectories of suicide and homicide rate change in 32 states that newly instituted the law’s provisions with 18 “control” states that already had such measures in place.31
Analyses revealed that while there was no difference between intervention and control states in patterns of change for homicide rates or for suicide rates in younger and middle-aged adults, there was a significantly greater reduction in the intervention states in firearm suicides specifically for persons aged 55 years and older. Almost 75% of older adults who take their own lives, the great majority of whom are men, do so with a firearm. Therefore, this universal intervention may be effective for the group of high-risk older men who appear not to have benefited from indicated and selective approaches to suicide prevention.
Because older people are the fastest growing segment of the population, we anticipate that the absolute numbers of late-life suicides will probably rise dramatically in the next 2 decades. There is an urgent need for more research, development of policy, and implementation of empirically based prevention practices if we are to reduce, or even contain, the toll that suicide-related morbidity and mortality take on older people and their families. Given our increased understanding of the factors that place older people at risk for suicide, prevention strategies can be designed and tested in a rigorous manner. Preliminary findings show that indicated, selected, and universal approaches all have a role to play. A comprehensive approach that incorporates all 3 is most likely to result in meaningful and sustained reductions in elder suicide injuries and deaths.
Dr Conwell is professor of psychiatry and co-director of the Center for the Study and Prevention of Suicide at the University of Rochester School of Medicine and Dentistry in Rochester, NY. He reports that he has no conflicts of interest concerning the subject matter of this article.
1. George Eastman is mourned. Democrat & Chronicle. March 15, 1932:1.
2. Conwell Y. Suicide. In: Roose SP, Sackeim HA, eds. Late-Life Depression. New York: Oxford University Press; 2006:95-106.
3. World Health Organization. Distribution of suicides rates (per 100 000) by gender and age, 2000. Geneva: World Health Organization; 2002.
4. Waern M, Runeson B, Allebeck P, et al. Mental disorder in elderly suicides: a case-control study. Am J Psychiatry. 2002;159:450-455.
5. Rubenowitz E, Waern M, Wilhelmsson K, Allebeck P. Life events and psychosocial factors in elderly suicides: a case control study. Psychol Med. 2001;31:1193-1202.
6. Beautrais AL. A case control study of suicide and attempted suicide in older adults. Suicide Life Threat Behav. 2002;32:1-9.
7. Harwood D, Hawton K, Hope T, Jacoby R. Psychiatric disorder and personality factors associated with suicide in older people: a descriptive and case-control study. Int J Geriatr Psychiatry. 2001;16:155-165.
8. Conwell Y, Lyness JM, Duberstein P, et al. Completed suicide among older patients in primary care practices: a controlled study. J Am Geriatr Soc. 2000;48:23-29.
9. Duberstein PR, Conwell Y, Conner KR, et al. Suicide at 50 years of age and older: perceived physical illness, family discord and financial strain. Psychol Med. 2004; 34:137-146.
10. Duberstein PR, Conwell Y, Conner KR, et al. Poor social integration and suicide: fact or artifact? A case-control study. Psychol Med. 2004;34:1331-1337.
11. Waern M, Rubenowitz E, Runeson B, et al. Burden of illness suicide in elderly people: case-control study. BMJ. 2002;324:1355.
12. Chiu HF, Yip PS, Chi I, et al. Elderly suicide in Hong Kong: a case-controlled psychological autopsy study. Acta Psychiatr Scand. 2004;109:299-305.
13. Duberstein PR. Are closed-minded people more open to the idea of killing themselves? Suicide Life Threat Behav. 2001;31:9-14.
14. Duberstein PR, Conwell Y, Seidlitz L, et al. Personality traits and suicidal behavior and ideation in depressed inpatients 50 years of age and older. J Gerontol B Psychol Sci Soc Sci. 2000;55:P18-P26.
15. Juurlink DN, Herrmann N, Szalai JP, et al. Medical illness and the risk of suicide in the elderly. Arch Intern Med. 2004;164:1179-1184.
16. Quan H, Arboleda-Florez J, Fick GH, et al. Association between physical illness and suicide among the elderly. Soc Psychiatry Psychiatr Epidemiol. 2002;37:190-197.
17. Turvey CL, Conwell Y, Jones MP, et al. Risk factors for late-life suicide: a prospective, community-based study. Am J Geriatr Psychiatry. 2002;10:398-406.
18. Conwell Y, Rotenberg M, Caine ED. Completed suicide at age 50 and over. J Am Geriatr Soc. 1990;38:640-644.
19. Tsoh J, Chiu HF, Duberstein PR, et al. Attempted suicide in elderly Chinese persons: a multi-group, controlled study. Am J Geriatr Psychiatry. 2005;13:562-571.
20. Brayer E. George Eastman: A Biography. Baltimore, Md: Johns Hopkins University Press; 1996. 21. Conwell Y, Duberstein PR, Connor K, et al. Access to firearms and risk for suicide in middle-aged and older adults. Am J Geriatr Psychiatry. 2002;10:407-416.
22. US Preventive Services Task Force. Screening for suicide risk: recommendation and rationale. Ann Intern Med. 2004;140:820-821.
23. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. 2001;16:606-613.
24. Almeida OP, Almeida SA. Short versions of the geriatric depression scale: a study of their validity for the diagnosis of a major depressive episode according to ICD-10 and DSM-IV. Int J Geriatr Psychiatry. 1999;14: 858-865.
25. Schulberg HC, Bruce ML, Lee PW, et al. Preventing suicide in primary care patients: the primary care physician’s role. Gen Hosp Psychiatry. 2004;26:337-345.
26. Practice guideline for the assessment and treatment of patients with suicidal behaviors. Am J Psychiatry. 2003;160(suppl 11):1-60.
27. Conwell Y, Duberstein PR, Cox C, et al. Age differences in behaviors leading to completed suicide. Am J Geriatr Psychiatry. 1998;6:122-126.
28. Institute of Medicine. Reducing Risks for Mental Disorders: Frontiers for Preventive Intervention Research. Washington, DC: National Academy Press; 1994.
29. Bruce ML, Ten Have T, Reynolds CF 3rd, et al. Reducing suicidal ideation and depressive symptoms in depressed older primary care patients: a randomized controlled trial. JAMA. 2004;291:1081-1091.
30. De Leo D, Dello Buono M, Dwyer J. Suicide among the elderly: the long-term impact of a telephone support and assessment intervention in northern Italy. Br J Psychiatry. 2002;181:226-229.
31. Ludwig J, Cook PJ. Homicide and suicide rates associated with implementation of the Brady Handgun Violence Prevention Act. JAMA. 2000;284:585-591.