Primary: to identify physiologic indicators of venlafaxine treatment response using quantitative EEG (QEEG) cordance, and to determine if cordance changes are specifically associated with response to venlafaxine;

Secondary: to determine if cordance changes early in the course (i.e., prior to improvement in clinical symptoms) of venlafaxine (or another antidepressant if venlafaxine is not clinically indicated for a particular patient) are predictive of later clinical response.

After a one-week single-blind placebo lead in, subjects will be randomly assigned to either venlafaxine or placebo for 8 weeks. They will undergo 6 QEEG studies (end of wash-in, and 48 hours, 1 week, 2 weeks, 4 weeks, and 8 weeks after randomized treatment), with examiner and self-ratings of mood, anxiety, and clinical status at the time of each recording (Ham-D, MADRS, Ham-A, SCL-90, Beck, LIFE, and CGI) to assess improvement. Any subjects with significant deterioration in mood and/or suicidal ideation during the 8 week trial will be dropped from the study and placed in open treatment.

At the end of 8 weeks, code will be broken and all subjects will be maintained/re-assigned to open-label treatment with venlafaxine for an additional 10 months if they wish. However, if the subject's primary physician believes that another clinically available antidepressant would be indicated instead of venlafaxine (due to history of prior non-response to venlafaxine, etc.), the indicated antidepressant medication will be administered. The antidepressant medication recommended by the primary physician will be provided free of charge for a one-year period. Tricyclic antidepressants and monoamine oxidase inhibitors will not be included due to the greater possibility of serious clinical sequelae with these older medications. The open-label phase will consist of regular monitoring by the laboratory at intervals of three days and one week after beginning a new antidepressant medication, and then monthly clinical visits (or more frequently if clinically indicated) with QEEG recordings and assessments of mood and clinical status as above by the laboratory psychiatrist to ensure that the subject is getting appropriate care from his or her primary physician. Drug dose will be adjusted using standard clinical practice by the subject's primary physician in the community, and if the subject remains on venlafaxine, the dosage may be increased as high as 225 mg/day during this phase.

Subjects will have one additional follow-up QEEG at the end of the open-label phase or when significant clinical improvement is detected (defined as resolution of DSM-IV symptoms, or Ham-D < 9). After the subject's depression resolves, he or she will continue to be monitored and given medication free of charge for the remainder of the one-year period, but will be seen clinically only by the primary physician in the community. A study psychiatrist will be available for consultation in cases of clinical necessity until the primary physician can be contacted. Subjects for whom venlafaxine is not clinically indicated and/or subjects who refuse the placebo portion of the study may be allowed to bypass the placebo-controlled phase and proceed directly to the open-label phase.

Inclusion Criteria:

• All subjects will meet DSM-IV criteria for depression on the basis of a SCID-P interview, with subjects in the placebo controlled phase of the study having a score on the 17-item Ham-D > 18 (with item #1 > 2). Subjects will meet criteria both at recruitment, and after a one-week single blind placebo wash-in. - Study includes outpatients only.

Exclusion Criteria:

• All subjects will have no serious medical illness.

• The investigators will exclude patients also meeting criteria for the following groups of axis I diagnoses:

  • delirium or dementia
  • substance-related disorders
  • schizophrenia or other psychotic disorders
  • eating disorders.

• In addition, patients meeting criteria for cluster A or B axis II diagnoses will be excluded.

• Subjects with a history of current or past active suicidal ideation, or suicide attempts will be excluded from the placebo-controlled phase of the study.