In this revised competing renewal of MH37869-16, we propose a randomized, double-blind, placebo-controlled study to test a method for the rapid treatment of geriatric major depression and for probing treatment response variability. We hypothesize that therapeutic sleep deprivation (TSD) will accelerate response to paroxetine (PX), as compared with TSD (+ placebo) or with PX alone. In open pilot studies, we have observed a rapid response rate of 69 percent (9/13 subjects), with Hamilton depression ratings of 10 or less by 14 days, in elderly depressed patients treated with the combination of TSD (one night) and paroxetine (20 mg QHS). By contrast, in other studies of bereavement- related or recurrent major depression, we have observed rapid response rates to placebo of 15 percent, to nortriptyline of 25-32 percent, and to paroxetine alone of 26 percent, suggesting that the use of combined (TSD + medication) may double or triple the rate of rapid response as compared with placebo or drug monotherapy, respectively. With respect to treatment response variability in geriatric major depression, we hypothesize that metabolic activity in cortical areas (prefrontal cortex and ventral anterior cingulate gyrus) will decrease to normal levels in patients showing an antidepressant response to TSD and will remain decreased after recovery sleep in patients responding rapidly to antidepressant treatment. (Data from non-depressed control subjects will be collected for comparison.) By contrast, we predict that glucose metabolism will remain elevated in non-responders and unchanged in non-responders who may be hypometabolic at baseline. Our pilot data show a reduction in cingulate metabolism after TSD in patients but not controls; the reduction persists after successful treatment with paroxetine. We will recruit 108 elderly depressed outpatients with current major depression into a 14-day randomized, placebo-controlled, double-blind, parallel-group study of TSD + PX, TSD + Placebo, and PX without TSD. All subjects will have pre-treatment MRI scans and twenty of 36 subjects in each treatment condition will also participate in PET studies of treatment response variability, together with 20 normal elderly control subjects. The percentage of patients meeting criteria for rapid response after 14 days of treatment in each of the three conditions will be contrasted in the intent-to-treat sample using contingency table analysis. Measures of subject expectancy, vascular risk factors, cerebral atrophy and white-matter hyperintensity, and cognitive status will be used as covariates in survival analyses of treatment response variability. Correlational analyses will be used to determine the association between changes in depression severity and alterations in regional glucose metabolic rates. Thus, this study aims to develop strategies to accelerate treatment response in geriatric major depression, to improve the early discrimination of non-responders, to model the functional neuroanatomy of treatment response variability, and ultimately to reduce heterogeneity of treatment response in geriatric depression.
Showing the most recent 10 out of 215 publications
