We propose to conduct under randomized double-blind conditions the first controlled study on the pharmacologic treatment of non-bipolar Psychotic Major Depression (PMD) in geriatric patients. PMD is a significant problem in late-life: About one third of elderly depressives requiring hospitalization present with PMD, often associated with significant but reversible cognitive impairment. Many of these patients are treated with ECT. Randomized double-blind studies have demonstrated that the response of younger PMD patients to either amoxapine (AMO) or a combination of amitriptyline plus perphenazine (PRZ) is excellent, alleviating the need for ECT in most. However, these results have not been replicated in, and cannot be extrapolated to, older PMD patients. In the absence of both controlled data and clinical consensus on the pharmacologic treatment of PMD in late life, we propose to study the effectiveness and safety of three treatments: (1) nortriptyline alone (NTP); (2) a combination of NTP plus PRZ; or (3) AMO alone. Over 5 years, 150 non-bipolar inpatients, age 60 and over, presenting with either a first or a recurrent episode of major depression with psychotic features (DSM- III-R) will be randomized under double-blind conditions to one of these three treatments. The efficacy and toxicity of these three treatments will be compared over 5 weeks of inpatient treatment followed by 3 weeks of outpatient treatment. Using methods for the analysis of both contingency tables and longitudinal categorical ordinal data, as well as repeated measures analysis for continuous data, we will address three questions: (i) To what extent does non-bipolar PMD respond to pharmacotherapy in late-life? (ii) Is the symptomatologic improvement different among the three treatment groups? (iii) Are the adverse effects different among the three treatment groups? We hypothesize that the response to either NTP+PRZ or AMO will be substantial and significantly higher than to NTP alone. Treatment toxicity -- in particular extrapyramidal symptoms and tardive dyskinesia -- will be most severe with NTP+PRZ, least severe with NTP, and intermediate with AMO. Confirmation of these hypotheses would be of immediate and significant benefit to a large number of severely but reversibly impaired elderly.
