The primary aim of this study is to examine the effects of co-administered wake therapy followed by light treatment on mood, and secondarily on circadian rhythms, to test the hypothesis that critically-timed chronotherapy improves mood by correcting phase disturbances in melatonin and sleep in women with Premenstrual Dysphoric Disorder.

The design is a randomized cross-over contrasting Late Wake Therapy plus morning bright light (LWT+Am BWL)vs. Early Wake Therapy plus evening bright light (EWT+PM BWL)administered in the luteal phase of two separate menstrual cycles, and preceded by 2 evaluation months. To lessen the patient's burden, the 1-night EWT or LWT and the following 7-day BWL interventions will be conducted at home, given at a fixed point in each menstrual cycle, from day 1 to 7 after the mid-cycle luteinizing hormone(LH) surge (ovulation). We anticipate that LWT+7 days of AM BWL (vs. EWT+PM BWL) will produce much greater mood benefits and larger physiological responses, than the one-time light pulses used in our earlier phase-shift studies.

Inclusion Criteria:

• Age: 18-45 years.
• Women with regular ovulatory menstrual cycles 26-32 days in length (for at least the previous six months).
• A history of a depressive (but not bipolar) mood disorder, but not an ongoing episode (symptom free for the last 12 months).
• Patients must meet DSM-IV criteria for Premenstrual Dysphoric Disorder (that includes irritability).
• Objective ratings: mean HRSD < 7 for follicular phase (day 5-10 of cycle after menses); mean HRSD > 14 for premenstrual (luteal) phase (6 days prior to onset of menses onward).
• Subjective ratings: mean Beck Depression Inventory < 5 follicular phase; > 10 premenstrual (luteal) phase, or
• Daily ratings: minimal symptoms (mean less than 50 on 100mm scale) follicular phase; at least a 30% increase in mean affective symptom ratings, premenstrual (luteal) phase.
• By clinical assessment and ratings, the patient has reported a history (for at least the last six months) of recurrent, moderate to severe premenstrual mood symptoms that impair some aspect of social or occupational functioning and that remit within a few days after the onset of menses. This pattern is prospectively documented with subjective and objective ratings over a 2-3 month interval. Patients must demonstrate a consistency of symptoms and a long enough duration of symptoms (7-10 days) to allow for study.
• Subjects willing to endure the rigors of a long-term (up to 6 months) research study.

Exclusion Criteria:

• Subjects with significant medical illness including hepatic (abnormal liver function tests), neurological, renal, cardiac, pulmonary, hematologic, gastrointestinal, or metabolic disorders.
• Subjects who are lactating, are within 6 months postpartum, or have an irregular sleep- wake cycle, e.g., from having very young children in the home.
• Subjects who are using hormonal contraception (within six months prior to the study).
• Subjects using other medication within one month of initiating the study or anytime during the study.
• Subjects with significant psychiatric disorder (schizophrenia, bipolar disorder, anxiety disorders, eating disorders, personality disorders, sleep disorders). An ongoing major depressive episode within the last year is reason for exclusion, although a previous history of a depressive episode is not (using DSM-IV diagnostic criteria for a major depressive episode).
• Subjects with a recent history (within the past year) of drug or alcohol abuse.
• Subjects with clinically significant abnormal laboratory values.
• Subjects with irregular menstrual cycles (cycle lengths vary greater than 3 days).
• Subjects unlikely to cooperate with the requirements of the study.
• Subjects needing frequent or continuous use of any medication, including nicotine (> 5 cigarettes daily).
• Subjects whose prospective ratings do not show cyclic variation in association with the menstrual cycle (as per inclusion criteria).
• Subjects with an irregular sleep schedule, extreme chronotypes or a sleep-wake cycle that does not correspond to the environmental light-dark cycle (e.g., subjects within 2 weeks of transmeridian travel, night shift workers, or those with significant advanced or delayed sleep phase syndromes). To enhance precision of the timing of the light stimulus on circadian phase (temporal resolution), we will exclude women with habitual sleep onset times after midnight or wake times after 9 am.

The principal investigator and clinical rater will remain blind to treatment condition until completion or discontinuation of participant from study. As patients will be coded, one statistician and technician and supervisor running assays also will be blind to patient name and treatment condition. Members of the Data Safety Monitoring Committee will review records in which patient names are coded, unless the code needs to be broken in the event of an adverse outcome. All other lab personnel will not be blind to treatment as the study coordinator and research associate prepare patient equipment and instruct patients on proper completion of the treatment protocol, the data manager and quality assurance officer enter all data into the database (including treatment condition). Due to the nature of the treatment (sleep, light intervention), patients will not be blind to condition.