• occurrence of suicide events [ Time Frame: 6 months ] [ Designated as safety issue: Yes ]
  
• reduction of suicidal ideation [ Time Frame: 6 months ] [ Designated as safety issue: Yes ]
  
• reduction in neuropsychological measures of impulsivity [ Time Frame: 6 months ] [ Designated as safety issue: Yes ]
  

Suicide is a significant public health problem. Depression, alcoholism (abuse or dependence), and a prior suicide attempt are risk factors for suicide. However, little information exists to guide clinicians in the choice of antidepressant medication for patients with comorbid major depression and alcoholism who are at risk for suicidal acts.

There is evidence that selective serotonin reuptake inhibitors (SSRI) may reduce impulsive aggression, and therefore lower the risk for suicidal behavior. We will test the hypothesis that fluoxetine, an SSRI, will be associated with fewer suicide events (defined as suicidal acts or increases in suicidal ideation necessitating a change in management), decreased suicidal ideation and decreases in neuropsychological measures of impulsivity compared to bupropion. The non-serotonergic drug, bupropion will improve energy and hopelessness. We expect the two drugs to be equally efficacious in reducing global depression severity. We will compare fluoxetine with bupropion in a 6-month randomized, controlled study of major depressive disorder and comorbid alcoholism in patients who have a prior history of suicide attempt. Patients requiring alcohol detoxification will be excluded. Patients will also receive weekly psychotherapy.

We will study 42 subjects with a current major depressive episode, comorbid alcoholism and a history of a past suicide attempt (21 subjects per year) in a randomized, double-blind, controlled trial, stratified by alcoholism type (1 vs 2). All subjects with alcohol dependence will be evaluated for risk of alcohol withdrawal prior to randomization. We will include patients who have suicidal ideation. However, patients with a suicidal plan or intent will only be enrolled as inpatients if the research team and the independent treatment team on the inpatient research unit agree that this is clinically reasonable. For example, if ECT or antipsychotics are indicated, the patient will not be enrolled. The study will provide six months of antidepressant pharmacotherapy as well as psychotherapy focused on alcohol relapse prevention. Patients will also be encouraged to attend daily Alcoholics Anonymous meetings. The outcome measures will be: 1) occurrence of suicide events; 2) reduction of suicidal ideation; 3) reduction in neuropsychological measures of impulsivity.