Suicide is a devastating event for the individuals, their families and the community. Understanding the risks of suicide behavior in older adults and exploring different interventions to reduce suicide ideation is an important clinical and research issue. This symposium will focus on correlates of baseline suicide ideation, explore a newly developed stress/diathesis model of suicidal behavior in late-life, and describe a randomized trial of peer companionship and its effect on depression, suicide ideation and psychological and structural connectedness.

Dr. Dimitris Kiosses will discuss the incidence of suicide ideation and its correlates in 81 older adults with major depression and cognitive impairment. Specifically, Dr. Kiosses will examine the effects of clinical (depression symptoms and depression severity) and neuropsychological variables (overall cognitive impairment, memory deficits, and executive dysfunction) as well as measures of social support on suicide ideation in older adults with cognitive deficits, up to the level of moderate dementia. He will also discuss the different profiles of older adults with suicide ideation with varying degrees of cognitive impairment.

Dr. Katalin Szanto will discuss a newly developed stress/diathesis model of suicidal behavior in late-life that may help to identify modifiable risk factors. Suicide follows an altered decision process. There is accumulating evidence that the suicidal diathesis involves cognitive deficits and maladaptive decision-making. The discussion of the model will be based on evidence from the literature, as well as from recent cognitive and behavioral data from over 100 depressed elderly who attempted suicide and demographically similar controls groups. The behavioral data provide new insight into facets of impulsivity, prominent in suicide attempters, and link these to objective characteristics of suicide attempt (planning and medical lethality).

Dr. Yeates Conwell will present the theoretical basis and the innovative design of the Senior Connection (TSC), a CDC-funded randomized trial of peer companionship for adults ages ≥ 60 years identified through screening in primary care as having feelings of loneliness and/or being a burden on others. Subjects are randomized to receive either care as usual (CAU) or the TSC intervention – assignment to a peer companion trained and supervised by a community-based aging services agency – and followed for 2 years. Outcomes of primary interest include depressive symptoms, suicidal ideation, and measures of psychological and structural connectedness. Dr. Conwell will also present preliminary results characterizing the subjects and their responses to CAU and TSC.

Dr. George Alexopoulos will serve as the discussant.