Major depressive disorder (MDD) is a common psychiatric illness with high cost to society and individual patients. One reason for the high cost is that most patients endure lengthy and ultimately unsuccessful empiric antidepressant trials before a successful medication is identified by trial-and-error. Care would be improved if a biomarker could determine, early in the course of treatment, whether a particular antidepressant would likely lead to response, remission, or treatment failure. Physicians could rapidly change treatments to an antidepressant which the biomarker indicated would be likely to help the patient. We have identified quantitative electroencephalographic (QEEG) changes that emerge early in the course of treatment with selective serotonin reuptake inhibitors (SSRIs) that appear to predict later response and remission in a general adult patient population. Demographic trends in the United States suggest that improved care for MDD will be essential for a growing number of elderly with late-life depression. While the consequences of prolonged trial-and-error periods to find a successful treatment are particularly inauspicious for elders with late-life depression, this patient group has not been included in the past studies which demonstrated the use of this biomarker approach in a general adult population. We propose a 12-week treatment trial to evaluate a practical biomarker for predicting outcome based on data from the first week of antidepressant treatment, with a focus only on depression in late life (age ≥65).

There are three study Hypothesis:

H1) ATR prediction of treatment outcome in older subjects will show >70% accuracy.

H2) The predictive accuracy of the model will be enhanced by including clinical, socio-demographic, and genetic predictors.

H3) The accuracy of ATR prediction will not show a significant dependence on subject gender.