• Percent of subjects, in each treatment arm, with remission of major depressive episode [ Time Frame: yearly ] [ Designated as safety issue: No ]
  
• Time to remission of major depressive episode [ Time Frame: yearly ] [ Designated as safety issue: No ]
  
• Number of venous thromboembolic events in the 3 treatment arms [ Time Frame: monthly ] [ Designated as safety issue: Yes ]
  
• Endometrial hyperplasia cases in each treatment arm [ Time Frame: monthly ] [ Designated as safety issue: Yes ]
  
• Incident breast CA cases in the 3 treatment arms [ Time Frame: monthly ] [ Designated as safety issue: Yes ]
  
• Irregular vaginal bleeding in the 3 treatment arms [ Time Frame: monthly ] [ Designated as safety issue: Yes ]
  
• Undesired breastfeeding discontinuation in the 3 treatment arms [ Time Frame: monthly ] [ Designated as safety issue: Yes ]
  

• Infant development among 6.5 month old children of mothers with PPMD, as assessed by Bayley Scales of Infant development [ Time Frame: yearly ] [ Designated as safety issue: No ]
  
• Infant serum concentrations of estradiol in 3 treatment arms [ Time Frame: monthly ] [ Designated as safety issue: Yes ]
  

This study aims to advance our therapeutic armamentarium by evaluating the efficacy of estradiol (E2) therapy for Postpartum Major Depression (PPMD), which has received minimal research attention in America. The design of the proposed study is an 8 week randomized double-blind clinical trial of SERT vs. E2 vs. Placebo. Responders enter a continuation phase with the blind intact through 6.5 months postpartum. The primary aims of this investigation are to: 1) Test the efficacy of E2 compared to placebo for the treatment of PPMD. Sertraline will be included as an active comparator. We have powered the study to test for differences among the three groups and also test for differences between the E2 and placebo group. We will test the hypothesis that E2 will be significantly more effective than placebo and that SERT will be significantly more effective than placebo. 2) Evaluate developmental outcomes in infants exposed to the disorder, PPMD, and the medications (SERT, exogenous E2 or Placebo) which may be transmitted to the infants through breastfeeding. All infants in this study will have exposure to mothers with depression. We will assess maternal depression, mother-infant serum SERT and E2 levels and relate them to mother-infant interactional quality and infant developmental outcomes on the Bayley Scales of Infant Development. These data will enhance the sophistication of risk-benefit analyses for pharmacotherapy during lactation.

Inclusion Criteria:

• Ages 18-40 years
• Postpartum Major depression with pregnancy or postpartum onset, intake by 3 months of delivery.
• SIGH-ADS score >= 18 and DSM-IV Diagnosis of major depression, current, outpatient care
• Medically healthy, including normal lipid profile, renal function, liver function, thyroid function, and CBC

Exclusion Criteria:

• Current use of other therapies for depression, such as antidepressants, psychotherapy, bright light therapy, and herbal remedies such as Hypericum St. John's Wort
• DSM-IV diagnoses of bipolar 1 or 2 disorder or any psychotic episode; substance abuse within last 6 months
• Previous adverse reaction to sertraline or provera
• No pediatric care: No pediatrician with whom to coordinate breastfeeding and infant care
• Use of medications for medical disorders, except for treatment of hypothyroidism or inhalers for asthma or progestin-only contraceptives
• Heavy smoking (>10 cigarettes per day) or intent to resume heavy smoking
• hyperlipidemia
• hypertension
• personal history thromboembolic event, hypercoagulability, or first degree relatives with thromboembolic events.
• Current or past personal history of breast, uterine, or ovarian cancer.
• First degree family history of premenopausal breast cancer or bilateral breast cancer; >3 family members with postmenopausal breast cancer.
• Current migraine headache disorder that is either complex in nature or is accompanied by hypertension or obesity.
• Arterial vascular disease and/or heart disease: increased risk of stroke.
• Liver disease: increased risk of biliary stones, cholestatic jaundice and benign hepatic lesions with E2 treatment.
• Pregnancy
• Infants born <=31 weeks of gestation
• Imminent suicidality and/or homocidality: in need of higher level of care than is provided in this study.