Suicide in Older Adults: How Do We Detect Risk and What Can We Do About It?

Older adults have higher rates of suicide than younger adults in many industrialized nations.^1,2 Any discussion of this increasingly important issue must be qualified by 2 caveats. First, the age at which a person is considered to be "older" varies across cultures and from one era to the next. Age 65 is the traditional but arbitrary retirement age in many industrialized nations. This group is heterogeneous, but few studies have determined whether the clinical risk markers for persons aged 65 to 70 years are identical to those aged 80 to 85 years.

Second, in many nations, rates of at tempted suicide are highest in younger women,³ but rates of completed suicide are highest in older men.^1,2 Just as the demographics of nonlethal and lethal suicide attempts are different, their clinical risk markers probably differ.^4,5 The extent to which research on at tempted suicide might inform efforts to prevent completed suicide is arguable.⁶ When researchers conflate suicidal behavior and completed suicide, they are likely to reach inaccurate con clusions that could misdirect treatment and prevention efforts.⁶ In this article we focus primarily on completed suicide.

Demographics of suicide Age, sex, race, and national differences in suicide rates
As shown in the Figure 1, suicide rates for men in the United States increase with age, but women's rates peak in midlife and remain stable or decline slightly thereafter. Suicide rates in white men aged 85 years and older are almost 6 times the nation's age-adjusted rate. In the United States, whites have higher rates than blacks across the life span.

In the United States, the mid-Atlantic states and New England have lower suicide rates than do the mountain states. Acknowledging minor annual variations, the crude rates in the mountain states are nearly twice those of the mid-Atlantic region. These differences could be attributed in part to the density of available mental health services.⁷ Other possible explanations include differences in religious practice and de nomination, ethnicity, community cohesiveness or social capital, and availability of firearms.

It should not be assumed that all risk factors for suicide are similar in strength across sociodemographic groups or geographic locales. For example, alcohol dependence is a more potent risk factor in Sweden than it is in France,⁸ and it might well be a more potent driver of risk in one region of the United States than in another. Despite the need for demographically and geographically sensitive risk assessments, available data are of little guidance to the clinician.

Marital status
Durkheim⁹ reported that suicide rates for people who are single, divorced, or widowed are higher than for married people. Widowhood is a more potent risk factor for suicide in younger people than in older adults,¹⁰ perhaps because the death of a spouse is often an un expected event in young adulthood.¹¹ Nonetheless, widowhood confers risk in older adults as well. The clinical as sessment of marital status should in clude inquiries about current marital status, relationship satisfaction, presence of conflict, and past marital history. Married individuals with previous histories of divorce or widowhood might be at heightened risk, although there are no firm data on this issue.

Living arrangements
Residential choice is determined by local cultural and subcultural norms, housing prices, and population density. Therefore, it is not surprising that some studies indicate that living alone does not confer suicide risk^12,13 and others suggest that it does.¹⁴ Living alone might confer risk in some contexts be cause it increases loneliness and com promises adherence to mental health treatments. Living with others may amplify risk if these arrangements engender family discord.

After determining whether a patient lives alone, the clinical interviewer should follow up with questions about recent changes in living arrangements, reasons for any changes, and satisfaction with living arrangements.

Socioeconomic status
Our data suggest that older adults who have less education¹⁵ and lower income¹⁶ may be at increased risk. Reasons for this disparity include less access to good-quality and affordable health care; subtle bias on the part of service providers; lower levels of mental health literacy; more stigmatized attitudes toward the receipt of mental health care; and chronic exposure to life event stress ors and chronic strains, such as the daily consequences of poverty.^17,18 By itself, poor access to good-quality services probably could not explain the socioeconomic disparity as observed in a study of older patients receiving anti depressant treatment from a university-affiliated tertiary care facility because access to care in this study was controlled by design.¹⁹

The assessment of socioeconomic status should include questions about educational attainment, annual household income, changes in income, reasons for any changes in income, and sat isfaction or dissatisfaction with in come. Socioeconomic status is often tied to neighborhood influences, which could, in turn, independently increase risk of poor mental health outcomes.²⁰ Clinicians may wish to inquire about perceived neighborhood poverty or affluence, safety, and the patient's comfort with the surrounding neighborhood.

Religious practice
One study showed that religious practice mitigates suicide risk among older adults.¹⁴ The relationship between religious practice and suicide was attenuated in analyses that controlled for the presence of alcohol use disorders, suggesting that the attitudes of many religious groups toward drinking might help decrease suicide risk. Other possible explanations include religious proscriptions against suicide, the tendency of people from the same religious denomination to live in proximity to eachother, the ability of faith communities to provide opportunities for socialization and the exchange of health-relevant information, and the availability of pastoral counseling.

Religious practice is probably pro tective only if it matches local cultural norms, however. Suicide may be more prevalent in the United States among members of relatively peripheral religious groups or cults than among members of denominations with deeper ties and established traditions. People who participate in a religious activity in a supportive cultural context might still be at risk, particularly if their involve ment in the faith community is minimal or peripheral.¹⁴

Physicians, including psychiatrists, tend to have lower participation in religious practices than the general population.²¹ Although they may not be less spiritual,²² mental health practitioners might underestimate the importance of conducting a careful assessment of religious practice in their patients.²³ Such assessments should include questions about the nature and frequency of attendance at religious services and functions, social involvement with a faith community, the extent to which he or she derives meaning and fulfillment from his religious practice and feels comfortable practicing his faith in the local context, and recent changes in religious practice or in feelings of support from the local faith community. In addition, clinicians should inquire whether a patient's religious commitment or beliefs would deter suicidal behavior, and under which specific circumstances.

Clinical risk markers
Few studies have been conducted with sufficient rigor to conclude that any single clinical risk indicator is a truly independent risk factor.^24,25 Nonetheless, the emerging literature points to risk markers that can be assessed reliably and rapidly (Table). Personality traits may also serve as risk markers,²⁴ but these findings have not been translated into clinically useful assessment tools.

Suicidal behavior (include suicide ideation expression/communication of a wish to die)³⁶

Advice for the Practicing Clinician Assessment
Recommendations for risk assessment are based on the fundamental premise that suicide is multidetermined. No single risk marker is paramount, and no single intervention will decrease risk in all at-risk individuals. Risk assessment is best accomplished through clinical interview of the patient in the context of a meaningful therapeutic relationship. Clinicians should never make a judgment about suicide risk based solely on a patient's response to a self-report instrument.

It is important to inquire about suicide ideation but research on the best phrasing and timing of these questions has not been conducted. Most measures of suicide ideation are neither developed nor standardized with older adult samples.^26,27 A recent exception is the Geriatric Suicide Ideation Scale (GSIS),²⁸ a 31-item self-report assessment measure developed and validated specifically for older adults.

To facilitate the identification of at-risk patients, clinicians might consider using cut-scores on widely used measures of depression.²⁹ Acknow ledging the limitations of relying exclusively on self-report data, we showed that 5 items on the Geriatric Depression Scale (GDS)³⁰ distinguish older adults reporting higher versus lower levels of suicide ideation.²⁹ Tapping hopelessness, worth lessness, emptiness, and un hap piness, these items mainly pertain to the "absence of positives" (eg, hope, self-worth, happiness). Similarly, others have suggested that it is important for clinicians to ask patients to identify positive events and positive attributes, such as the presence of a supportive social network and people in whom they can confide.^28,30

Concerns with assessing suicide risk include false positives: incorrectly judging a patient to be at risk for completed suicide; and false negatives: incorrectly judging a suicidal older adult to be at no or low risk for suicide. False positives could lead to unnecessary treatment, adverse effects, and financial costs. Patients who have been said to elicit hatred and other strong reactions from clinicians³¹ are probably more likely to be falsely identified as at-risk for death.

False negatives can lead to serious physical injury and/or death. These errors are more likely to occur with "difficult-to-reach" patients: persons who have little or no history with the mental health care system, attribute their symptoms to medical issues or life events,³² have difficulty in verbalizing their feelings, and believe that suicide is a noble, private, and ethical act.

The difficult patient
Suicide risk (see Table) should be addressed directly, calmly, and empathically.³³ When indicated, a critical care protocol should be developed³³ and day treatment, partial hospitalization, and inpatient care considered.The need for continuity of care is ab solutely critical. Suicide risk among recently hospitalized adults,^7,34,35 including older adults,^13,36,37 is high shortly after discharge and remains high for many years.³⁸ Discharge stress, incomplete treatment, and the tendency of some patients to conceal their suicide plans from care providers in order to facilitate prompt discharge may in crease risk. Poor communication³⁹ and poor continuity of care have been shown to elevate risk in relatively young (mean age less than 50 years) patients.^7,40 No study has attempted to identify predictors of suicide after discharge in an older adult sample, but alarmingly low levels of treatment engagement following discharge from gero psychiatric units have been documented.⁴¹ If hospitalization decreases risk, then discharge may be viewed as the withdrawal of a protective factor³⁵ and should be handled with caution.⁴¹ It can be argued that the siloed health care delivery system undermines this effort.⁴¹

The difficult-to-reach patient
A qualitative study of people who completed suicide without seeking medical advice in their final month of life revealed that many were older men who were characterized as "habitual nonconfiders."⁴² In one study, nearly 10% of the older adults who went on to complete suicide denied having thoughts of suicide when asked directly beforehand.⁴³ Clearly, the expression of suicide ideation is neither a necessary nor sufficient condition for completed suicide in older adults; the same is true for a history of suicidal behavior-an undeniably important risk marker. How ever, it is important to keep in mind that people who have never previously engaged in suicidal behavior account for 58%12 to 75%⁴⁴ of completed suicides in older age.Persons who deny previous suicidal behavior may be at heightened risk by virtue of the presence of other risk factors. Seasoned clinicians might have difficulty in eliciting good clinical in formation from some at-risk older patients, given the tendency of some, particularly men,^45,46 to minimize depressive symptoms.^32,47-49 As a consequence, mood disorders and related symptoms are less likely to be diagnosed and treated.

Clinicians must gather collateral in formation from relatives, friends and acquaintances, clergy, and other care providers, including primary care physicians. These persons are more likely to report the presence of some mood disorder symptoms than at-risk patients; however, they are less able to report suicide ideation.

⁴⁹

The phenomenology of mood disorders in older adults can mask suicide risk. Neurovegetative symptoms are non specific indicators of an array of phy sical diseases and treatment side effects. They are not specific indicators of mood disorders, let alone suicide risk. Clinicians should be particularly attentive to symptoms of depression among those presenting for treatment of memory, physical, and functional problems. Subsyndromal conditions should be monitored, especially in patients who may be at risk by virtue of family history or other risk markers (see Table).

No discussion of the difficult-to-reach patient would be complete without some mention of sociodemograph ic issues. Primary care physicians may be less likely to recognize and diagnose mood disorders in men than in women.^50,51 Whether the same is true of psychiatrists is unknown. It is also not known whether ethnic factors affect the interactions between patients and psychiatrists, although patient-physician communication^18,52 and the diagnosis of mood disorders⁵³ are more difficult when primary care physicians and patients are of different races or ethnic backgrounds. Either way, sensitivity and empathy are instrumental in developing good clinical rapport that can help facilitate effective treatment.

Conclusion
Using solid data on clinical risk markers, interventions can be developed that specifically target the symptom or symptoms that appear to confer risk. This clinical research strategy, successfully used in 2 prior studies,^54,55 is not foolproof. Many at-risk patients will not avail themselves of treatment, and few treatments are tailored to patients' needs and preferences, leading to poor adherence levels.⁵⁶

Clinical initiatives are most effective with patients who are conflictually engaged with the health care system but fail to reach disengaged patients outside the system. Prevention programs can indirectly lead to lower suicide rates by modifying attitudes about aging, shifting cultural norms concerning the acceptability of suicide, restricting access to lethal means, and increasing rates of mental health literacy and health-promoting behaviors.^57-59

However, many ostensibly "universal" interventions will not reach all at-risk people,⁶⁰ some universal initiatives (screening, firearms control) are vul nerable to political gridlock, and these interventions are unlikely to exert much of an influence on those at greatest risk. Of course, the long-term goal is not to bring the suicide rate down to zero, an unrealistic goal in a free society, but rather to do whatever must be done to decrease the number of people who believe that suicide is their only option. This will require better treatments and better prevention programs, targeting individuals, families, and communities.Dr Duberstein is professor of psychiatry, University of Rochester Medical Center in New York. He reports no conflicts of interest concerning the subject matter of this article.

Dr Heisel is assistant professor of psychiatry and epidemiology and biostatistics, University of West ern Ontario. He discloses that he is the author of one of the measures of the Geriatric Suicide Ideation Scale and that he has contrib uted to research on suicide ideation screening items for the Geriatric Depression Scale; both are mentioned in this article.

Preparation of this article was partially supported by Grant K24MH072712 from the United States Public Health Service (P.D.), and a Young Investigator Award from the American Founda tion for Suicide Prevention (M.J.H.).

References:

References1. Bertolote JM. Suicide in the world: an epidemiological overview 1959-2000. In: Wasserman D, ed. Suicide: An Unnecessary Death. London: Martin Dunitz; 2001:3-10.
2. Krug EG, Dahlberg LL, Mercy JA, et al, eds. World Report on Violence and Health. Geneva: World Health Organization; 2002. Available at: http://www.who.int/violence_injury_prevention/violence/world_report/en/full_en.pdf. Accessed October 5, 2006.
3. Kessler RC, Borges G, Walters EE. Prevalence of and risk factors for lifetime suicide attempts in the National Comorbidity Survey. Arch Gen Psychiatry. 1999;56:617-626.
4. Tsoh J, Chiu HF, Duberstein PR, et al. Attempted sui cide in elderly Chinese persons: a multi-group, con trolled study. Am J Geriatr Psychiatry. 2005:13:562-571.
5. Useda JD, Duberstein PR, Beckman A, et al. Differences in attempted versus completed suicide: informant report of personality traits in adults 50 years of age and older. J Consult Clin Psychol. In press.
6. Duberstein PR, Conwell Y. Personality disorders and completed suicide: A methodological and conceptual review. Clin Psychol Sci Pract. 1997;4:359-376.
7. Desai RA, Dausey DJ, Rosenheck RA. Mental health service delivery and suicide risk: the role of individual patient and facility factors. Am J Psychiatry. 2005;162:311-318.
8. Norstrom T. Alcohol and suicide: a comparative analysis of France and Sweden. Addiction. 1995;90:1463-1469.
9. Durkheim E. Suicide: A Study in Sociology. New York: Free Press; 1962.
10. Luoma JB, Pearson JL. Suicide and marital status in the United States, 1991-1996: is widowhood a risk factor? Am J Public Health. 2002;92:1518-1522.
11. Cohler BJ, Jenuwine MJ. Suicide, life course, and life story. Int Psychogeriatr. 1995;7:199-219.
12. Rubenowitz E, Waern M, Wilhelmson K, Allebeck P. Life events and psychosocial factors in elderly suicides: a case-control study. Psychol Med. 2001;31:1193-1202.
13. Beautrais AL. A case control study of suicide and attempted suicide in older adults. Suicide Life Threat Behav. 2002;32:1-9.
14. 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.
15. 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.
16. 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.
17. Link BG, Phelan J. Social conditions as fundamental causes of disease. J Health Soc Behav. 1995;special no.:80-94.
18. Van Ryn M, Fu SS. Paved with good intentions: do public health and human service providers contribute to racial/ethnic disparities in health? Am J Public Health. 2003;93:248-255.
19. Cohen A, Houck PR, Szanto K, et al. Social inequalities in response to antidepressant treatment in older adults. Arch Gen Psychiatry. 2006;63:50-56.
20. Kubzansky LD, Subramanian SV, Kawachi I, et al. Neigh borhood contextual influences on depressive symptoms in the elderly. Am J Epidemiol. 2005;162:253-260.
21. Neeleman J, Persaud R. Why do psychiatrists neglect religion? Br J Med Psychol. 1995;68:169-178.
22. Smith DP, Orlinsky DE. Religious and spiritual experience among psychotherapists. Psychotherapy: Theory, Research, Practice, Training. 2004;41:144-151.
23. Kehoe NC, Gutheil TG. Neglect of religious issues in scale-based assessments of suicidal patients. Hosp Community Psychiatry. 1994;45:366-369.
24. Duberstein PR, Heisel MJ, Conwell Y. Suicide. In: Agronin ME, Maletta GJ, eds. Principles and Practice of Geriatric Psychiatry. Philadelphia: Lippincott Williams & Wilkins; 2005:393-405.
25. Heisel MJ, Duberstein PR. Suicide prevention in older adults. Clin Psychol: Science and Practice. 2005;12:242-259.
26. Beck AT, Kovacs M, Weissman A. Assessment of suicidal intention: the Scale for Suicide Ideation. J Consult Clin Psychol. 1979;47:343-352.
27. Latham AE, Prigerson HG. Suicidality and bereavement: complicated grief as psychiatric disorder presenting greatest risk for suicidality. Suicide Life Threat Behav. 2004;34:350-362.
28. Heisel MJ, Flett GL. The development and initial validation of the geriatric suicide ideation scale. Am J Geriatr Psychiatry. 2006;14:742-751.
29. Heisel MJ, Flett GL, Duberstein PR, Lyness JM. Does the geriatric depression scale (GDS) distinguish between older adults with high versus low levels of suicidal ideation? Am J Geriatr Psychiatry. 2005;13:876-883.
30. Yesavage JA, Brink TL, Rose TL, et al. Development and validation of a geriatric depression screening scale: a preliminary report. J Psychiatr Res. 1982-1983;17:37-49.
31. Maltsberger JT, Buie DH. Countertransference hate in the treatment of suicidal patients. Arch Gen Psychiatry. 1974;30:625-633.
32. Eaton WW, Neufeld K, Chen LS, Cai G. A comparison of self-report and clinical diagnostic interviews for depression: diagnostic interview schedule and schedules for clinical assessment in neuropsychiatry in the Baltimore epidemiologic catchment area follow-up. Arch Gen Psychiatry. 2000;57:217-222.
33. Brown GK, Bruce ML, Pearson JL. High-risk management guidelines for elderly suicidal patients in primary care settings. Int J Geriatr Psychiatry. 2001;16:593-601.
34. Qin P, Agerbo E, Mortensen PB. Suicide risk in re lation to socioeconomic, demographic, psychiatric, and familial factors: a national register-based study of all suicides in Denmark, 1981-1997. Am J Psychiatry. 2003;160:765-772.
35. Meehan J, Kapur, Hunt IM, et al. Suicide in mental health in-patients and within 3 months of discharge. Br J Psychiatry. 2006;188:129-134.
36. Hawton K, Harriss L. Deliberate self-harm in people aged 60 years and over: characteristics and outcome of a 20-year cohort. Int J Geriatr Psychiatry. 2006;21:572-581.
37. Erlangsen A, Mortensen PB, Vach W, Jeune B. Psy chiatric hospitalization and suicide among the very old in Denmark: population-based register study. Br J Psy chiatry. 2005;187:43-48.
38. Hawton K, Harriss L. Deliberate self-harm in people aged 60 years and over: characteristics and outcome of a 20-year cohort. Int J Geriatr Psychiatry. 2006;21:572-581.
39. Hendin H, Haas AP, Maltsberger JT, et al. Problems in psychotherapy with suicidal patients. Am J Psychiatry. 2006;163:67-72.
40. King EA, Baldwin DS, Sinclair JM, et al. The Wessex Recent In-Patient Suicide Study, 1: case-control study of 234 recently discharged psychiatric patient suicides. Br J Psychiatry. 2001;178:531-536.
41. Tew JD Jr. Post-hospitalization transitional care needs of depressed elderly patients: models for improvement. Curr Opin Psychiatry. 2005;18:673-677.
42. Owens C, Lambert H, Donovan J, Lloyd KR. A qualitative study of help-seeking and primary care con sultation prior to suicide. Br J Gen Pract. 2005;55:503-509.
43. Waern M, Beskow J, Runeson B, Skoog I. Suicidal feelings in the last year of life in elderly people who commit suicide. Lancet. 1999;354:917-918.
44. Conwell Y, Duberstein PR, Conner KR, et al. Case control study of illness burden and service utilization prior to suicide in adults over 49 years of age. Submitted to Am J Geriatr Psychiatry. 2006.
45. Allen-Burge R, Storandt M, Kinscherf DA, Rubin EH. Sex differences in the sensitivity of two self-report depression scales in older depressed inpatients. Psychol Aging. 1994;9:443-445.
46. O’Connor DW, Rosewarne R, Bruce A. Depression in primary care, 1: elderly patients’ disclosure of depressive symptoms to their doctors. Int Psychogeriatr. 2001;13:359-365.
47. Gallo JJ, Rabins PV, Anthony JC. Sadness in older per sons: a 13-year follow-up of a community sample in Baltimore, Maryland. Psychol Med. 1999;29:341-350.
48. Lyness JM, Cox C, Curry J, et al. Older age and the underreporting of depressive symptoms. J Am Geriatr Soc. 1995;43:216-221.
49. McAvay GJ, Raue PJ, Brown EL, Bruce ML. Symptoms of depression in older home-care patients: patient and informant reports. Psychol Aging. 2005;20:507-518.
50. Callahan EJ, Bertakis KD, Azari R, et al. Depression in primary care: patient factors that influence recognition. Fam Med. 1997;29:172-176.
51. Stoppe G, Sandholzer H, Huppertz C, et al. Gender differences in the recognition of depression in old age. Maturitas. 1999;32:205-212.
52. Johnson RL, Roter D, Powe NR, Cooper LA. Patient race/ethnicity and quality of patient-physician communication during medical visits. Am J Public Health. 2004;94:2084-2090.
53. Comino EJ, Silove D, Manicavasagar V, et al. Agree ment in symptoms of anxiety and depression between patients and GPs: the influence of ethnicity. Fam Pract. 2001;18:71-77.
54. Bruce ML, Ten Have TR, Reynolds CF III, et al. Reducing suicidal ideation and depressive symptoms in depressed older primary care patients: a randomized controlled trial. JAMA. 2004;291:1081-1091.
55. Unützer J, Tang L, Oishi S, et al. Reducing suicidal ideation in depressed older primary care patients. J Am Geriatr Soc. In press.
56. Pyne JM, Rost KM, Farahati F, et al. One size fits some: the impact of patient treatment attitudes on the cost-effectiveness of a depression primary-care intervention. Psychol Med. 2004;35:839-854.
57. Kaplan RM. Two pathways to prevention. Am Psychol. 2000;55:382-396.
58. Knox KL, Conwell Y, Caine ED. If suicide is a public health problem, what are we doing to prevent it? Am J Public Health. 2004;94:37-45.
59. Levine M. Prevention and community. Am J Com munity Psychol. 1998;26:189-206.
60. Neeleman J. Beyond risk theory: suicidal behavior in its social and epidemiological context. Crisis. 2002;23:114-120.
61. 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.
62. Waern M, Runeson BS, Allebeck P, et al. Mental disorder in elderly suicides: a case-control study. Am J Psychiatry. 2002;159:450-455.
63. Rubio A, Vestner AL, Stewart JM, et al. Suicide and Alzheimer's pathology in the elderly: a case-control study. Biol Psychiatry. 2001;49:137-145.
64. Waern M, Rubenowitz E, Wilhelmson K. Predictors of suicide in the old elderly. Gerontology. 2003;49:328-334.
65. Conner KR, Duberstein PR, Conwell Y, et al. After the drinking stops: completed suicide in individuals with remitted alcohol use disorders. J Psychoactive Drugs. 2000;32:333-337.
66. 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.
67. Horton-Deutsch SL, Clark DC, Farran CJ. Chronic dyspnea and suicide in elderly men. Hosp Community Psychiatry. 1992;43:1198-1203.
68. Waern M, Rubenowitz E, Runeson B, et al. Burden of illness and suicide in elderly people: case-control study. BMJ. 2002;324:1355.
69. Quan H, Arboleda-Flórez J, Fick GH, et al. Association between physical illness and suicide among the elderly. Soc Psychiatry Psychiatr Epidemiol. 2002;37:190-197.
70. Malphurs JE, Cohen D. A statewide case-control study of spousal homicide-suicide in older persons. Am J Geriatr Psychiatry. 2005;13:211-217.
71. Conwell Y, Duberstein PR, Connor K, et al. Access to firearms and risk for suicide in middle-aged and old er adults. Am J Geriatr Psychiatry. 2002;10:407-416.