Treatment of depression in the elderly: Effect of physical illness on response

Objectives. To determine the response to treatment of different subgroups within a sample of physically ill elderly depressed patients. Design. Acute geriatric medical inpatients with depression, randomly assigned to an X-week double-blind placebo-controlled trial of fluoxetine. Main outcome measure. Response rate as defined by the 17-item Hamilton Depression Rating Scale. Results. Data were analysed in subgroups according to diagnosed physical problems and concomitant medication. A logistic regression analysis was performed to identify subgroups valid for separate analysis. Those reaching at least 5 weeks of treatment showed a significant improvement compared with the placebo group if they had serious life-threatening disease, cerebrovascular disease, were not demented, or were either on no analgesics or on analgesics stronger than paracetamol. Conclusion. While the response to treatment in these subgroups was encouraging, general physicians must not be led to believe that the answer to depression lies only in pharmacological intervention, just as the belief that the risk-benefit ratio of such treatment where indicated is too high must be discouraged. (C) 1997 John Wiley & Sons, Ltd.