IQ as a predictor of antidepressant responses to imipramine.

Abstract

Although the use of IQs to predict scholastic achievement is widely accepted, their use as predictors of response to drugs is infrequently encountered. But a recent study s u e gested that IQs were one of several predictors of response to the antidepressant drug irnipramine ( 2 ) . It was reported that patients who were assessed by clinical impression to be less intelligent than average may respond better to imipramine than patients above average in intelligence. The present study was designed to evaluate the potential of relatively low and high 1Qs to predict response to imipramine. Each of 21 females and 8 males, hospitalized for depression at the clinical facilities of the National Institute of Mental Health, Bethesda, Maryland, was administered a complete WAIS (Full Scale IQs ranged from 80 to 133) on admission during which time they were free of medication. They were then separated into two groups, i.e., one group of responders ( n = 13) and another group of nonresponders ( n = 16) . All subjects were treated with imipramine HCL, 150/300 mg/day, for a minimum of 24 days. Subjects not responding to doses of 150/200 mg/day after 2 wk. treatment were increased to 250/300 mg/day unless side effects supervened. Three subjects who did not tolerate doses of 150 mg or higher during the final days of treatment were not included in the study. A trained nursing research team rated the patients cwice daily on a modified 15-point scale ( 1 ) . Responders (M = 107.08, SD = 11.44) showed lower but nonsignificant ( t = 1.35, d f = 27, p > .05) Full Scale IQs than nonresponders ( M = 112.75, SD = 10.32). Computation of a critical value ( 3 ) indicated that a Full Scale IQ of less than 106 was a responder to imipramine whereas 106 or greater was a nonresponder. Use of an IQ of 106 as a cutting point enabled us to predict 6 of 13 responders and 11 of 16 nonresponders ( p = .22, binomial test). A biserial correlation (rbls = -.311) of IQ and response to imipramine was not significant. These results require cross-validation and suggest that, although responders have lower mean 1Qs than nonresponders when IQs are assessed psychometrically, relatively low and high IQs are poor predictors of response-nonresponse to imipramine.