Severe premenstrual syndrome (PMS) is characterized by disabling physical and psychological symptoms that occur during the luteal phase of the menstrual cycle. Allopregnanolone has been implicated in the pathophysiology of mood disorders, stress, and possibly PMS and premenstrual dysphoric disorder. 1 There is some evidence that response to treatment of PMS correlates with decreased allo levels. 2 We previously found in a small pilot study that women with PMS who improved with selective serotonin reuptake inhibitor (SSRI) treatment had significantly lower allo levels at treatment end point than the unimproved subjects. 3 However, a major limitation of that study was that pretreatment allo levels were not evaluated, and associations between SSRI treatment and changes in allo could not be determined. The aims of the present study were to identify changes in allo levels after SSRI treatment and determine whether the changes in allo levels were related to improvement in PMS symptoms. Based on our previous pilot study, we hypothesized that high allo levels decreased with SSRI treatment and that the changes were associated with symptom improvement. We also hypothesized that low allo levels at baseline increased, as previously shown in patients with depression, and that the changes were associated with improvement of dysphoric symptoms. 4

This was a prospective study of 46 women with PMS, whose conditions were diagnosed with clearly defined criteria including daily symptom ratings and treated with sertraline as described elsewhere. 5 All participants who had serum samples collected within 8 days before menses both before and after sertraline treatment and met the criteria for this study were included. Inclusion criteria included regular menstrual cycles in reference range, a positive result for urine test indicating probable ovulation, and general good health. Exclusions included any hormone use, other treatments for PMS, any major Axis I psychiatric diagnosis currently or in the past year, lifetime diagnosis of bipolar disorder or psychosis, and alcohol or drug abuse. Flexible regimens were used; all but 4 subjects had luteal phase dosing (14 days before estimated menses through 2 days after the onset of menses) with sertraline dosages of 50 or 100 mg/d. All subjects signed consent forms approved by the university institutional review board.