The focus of this study is to gather preliminary data regarding the effects of a psychological therapy-Problem Solving Therapy-and an antidepressant medication-sertraline-on 1) cerebral perfusion (CP), 2) brain derived neurotrophic factor (BDNF), and 3) measures of cognitive function in subjects with late life major depression (LLMD). This research goal will be achieved by recruiting 38 individuals over the age of 65 with LLMD. The primary outcomes will be change in CP, change in BDNF, and change in cognitive measures from baseline to the end of 12 weeks of either therapy. We will also examine predictors of treatment outcome including severity of executive dysfunction, baseline BDNF concentrations, and baseline CP measures. The baseline neuropsychological testing, brain imaging, and depression assessment will be obtained in a companion study (PI S. Mackin; CHR #H42689-32681-01) that is IRB approved and is already in progress. In the current study a baseline serum BDNF level will be added to Dr. Mackin's protocol. Patients will then receive either 12 weeks of Problem Solving Therapy or antidepressant treatment with sertraline. Both treatments are evidence based and commonly administered in our clinic. Outcome variables will be measures of depression severity, the BDNF serum concentration, cerebral perfusion using a MRI arterial spin labeling (ASL) technique and cognitive changes in memory and executive dysfunction. This is a preliminary or pilot study. The primary objectives are to determine if the methods appear feasible and to determine if change in BDNF or CP occur after treatment and secondarily to determine if there are changes in cognitive functioning. The study is not powered to show differences between treatments. The hypotheses are 1) PST will result in increased perfusion in frontal regions of the brain but that frontal perfusion will not change with sertraline; 2)sertraline will result in an increase in BDNF but PST will not. Change in cognitive measures of memory, learning, and executive dysfunction will be examined on an exploratory basis.

Subjects: subjects will have non-psychotic unipolar major depression based on DSM IV criteria (Structured Clinical Interview for DSM IV; SCID)and be 65 years or older and will have depression of moderate or greater depression severity (Hamilton Depression Rating Scale score > or = 19. Depression severity will be assessed with the clinician rated Hamilton Depression Rating Scale (HDRS) and the subject rated Quick Inventory of Depressive Symptoms (QIDS). Ratings will be performed weekly for the first month and then every 2 weeks until the end of the trial.

Neuroimaging: The MRI studies included in this protocol are commonly utilized in clinical practice. In addition to conventional structural MRI, this protocol will utilize techniques developed to evaluate cerebral blood flow (arterial spin labeling), white matter integrity (diffusion tensor imaging), and brain biochemistry (MR spectroscopy) and will be performed using a 4 Tesla magnet. Imaging will be performed at the VAMC/UCSF Center for Imaging of Neurodegenerative Disease (CIND) under the direction of Michael Weiner, M.D. Brain derived neurotropic factor (BDNF): a blood sample will be drawn pretreatment to determine the serum concentration of BDNF. This test will be repeated post-treatment. The assay for BDNF will be performed in the laboratory of Synthia Mellon, Ph.D at UCSF. Serum will be assayed for BDNF in duplicate, using a commercial BDNF ELISA assay kit (R&D Systems, Minneapolis, MN, USA).

Cognition will be assessed with a battery of neuropsychological tests including the Stroop Color-Word Test, the Trail Making Test: A and B; Dementia Rating Scale I/P; Boston Naming Test, and the Hopkins Verbal Learning Test-Short.

Treatment assignment: If patients have a preference for PST or sertraline, they will receive that treatment. If they have no preference, they will be randomized to one or the other until 19 subjects have been assigned to each treatment.

Sertraline, a selective serotonin reuptake inhibitor (SSRI), will be administered as the antidepressant. It will be started at 25 mg/day for one week and then increased to 50 mg and continued for 3 weeks. At the end of 4 weeks if the patient has had limited response, the dose will be increased to 100 mg/day. At 8 weeks if response is limited, dose will be increased to 150 mg/day. At any time, the dose can be lowered for tolerability reasons.

Problem Solving Therapy. Patients receiving PST will be seen weekly for 12 individual 45 minute sessions. Problem Solving Therapy (Arean, Raue, and Julian; UCSF unpublished manuscript, 2003) consists of 12 weekly sessions to teach participants a five-step problem-solving model. This model is taught over the first five weeks of treatment. Subsequent sessions are dedicated to refining PST skills. In the last two PST sessions, participants create a relapse prevention plan using the PST model. There will be two therapists in the study both trained to perform PST and with experience in prior studies of PST for MDD in older adults.

Data will be analyzed by Drs Nelson and Mackin in consultation with Kevin Delucchi, Ph.D.(Department of Biostatistics/Psychiatry; UCSF). All analyses will be performed within treatment groups (psychotherapy or drug treatment) and will compare baseline values with post-treatment values in patients completing at least 8 weeks of treatment. All analyses will begin by graphically and numerically summarizing all measures to assess the distribution of scores. To statistically control for appropriate associations of other demographic variables (such as age, sex, and education) and clinical variables (such as social support, medical comorbidity, physical frailty, and medication use) with the outcome measure, we will use linear regression modeling methods. Missing data, if any, will be carefully described and analyzed. The methodology to be used assuming missing at random (MAR) which is a reasonable assumption for this type of data and the relatively low levels of missing data we anticipate based on prior research with this population. Because we have specific a priori hypotheses regarding BDNF and the composite score for executive dysfunction, we will not correct for multiple comparisons. Analysis of the cerebral blood flow data, however, will employ corrections for multiple comparisons. As stated the aim of the study is to gather preliminary data regarding change from baseline to the study endpoint within the sertraline or within the PST treatment group. Attrition is estimated at 20% allowing for 15 patients to complete each treatment arm.

Inclusion Criteria:

• clinical diagnosis of DSM IV non-psychotic unipolar major depression
• depression severity > or = to 19 on HDRS
• age 65 years or older
• able to provide informed consent

Exclusion Criteria:

• history of psychosis
• failure to respond to sertraline or PST during the episode
• allergic to sertraline