A randomized, placebo-controlled trial of paroxetine in nursing home residents with non-major depression

Depression is common across a broad spectrum of severity among nursing home residents. Previous research has demonstrated the effectiveness of antidepressants in nursing home residents with major depression, but it is not known whether antidepressants are helpful in residents with less severe forms of depression. We conducted a randomized double-blind placebo-controlled 8-week trial comparing paroxetine and placebo in very old nursing home residents with non-major depression. The main outcome measure was the primary nurse's Clinical Impression of Change (CGI-C). Additional outcome measures were improvement on the interview-derived Hamilton Depression Rating Scale (HDRS) and Cornell Scale for Depression (CS) scores. Twenty-four subjects with a mean age of 87.9 were enrolled and twenty subjects completed the trial. Placebo response was high, and when all subjects were considered, there were no differences in improvement between the paroxetine and placebo groups. Two subjects that received paroxetine developed delirium, and subjects that received paroxetine were more likely to experience a decrease in Mini Mental State Exam scores (P =.03). There were no differences in serum anticholinergic activity between groups. In a subgroup analysis of 15 subjects with higher baseline HDRS and CS scores, there was a trend toward greater improvement in the paroxetine group in an outcome measure that combined the CGI-C and interview-based measures (P =.06). Paroxetine is not clearly superior to placebo in this small study of very old nursing home residents with non-major depression, and there is a risk of adverse cognitive effects. Because of the high placebo response and the trend towards improvement in the more severely ill patients, it is possible that a larger study would have demonstrated a significant therapeutic effect for paroxetine as compared with placebo. The study also illustrates the discordance between patient and caregiver ratings, and the difficulties in studying very elderly patients with mood disorders.