Predictors of Amitriptyline Response in Outpatient Depressives

A step-search multiple regression procedure was applied to data obtained from 120 amitriptyline-treated and 138 placebo-treated depressed nonpsychotic outpatients in order to: 1) assess the magnitude of the drug's effect in an outpatient sample after the possibly contaminating effects of relevant sources of patient heterogeneity have been statistically removed and 2) identify predictors of response to amitriptyline, placebo, or both. Patient, illness, and treatment attributes deemed of possible relevance were employed as potential predictors. Patients were drawn from two city hospital clinics and the private practices of both general practitioners and private psychiatrists. The double blind drug trial employed was of 4 weeks' duration, and improvement was assessed through both a patient global rating of change and a pretreatment-post-treatment difference in physicianevaluated depressive psychopathology (PDS improvement). The drug effect was highly significant. Its magnitude was left unchanged for global improvement and slightly enhanced for PDS improvement when possibly contaminating influences of other predictors were statistically controlled through the regression procedure. For both improvement measures, drug-placebo differences were greater in more severely than in less severely ill patients. For all patients, irrespective of treatment agent, acute illness 6 months) responded better to treatment than did long term illness. For the global measure there were indications that the more adequate personal and social resources, as well as the more favorable and realistic attitudes, of the higher socioeconomic class patient were instrumental in effecting greater drug-placebo differences than occurred for patients of lower socioeconomic class affiliation. Finally, in replication of previous results, placebo patients were found to show greater global improvement when their physicians reported liking them less.