Between 2 and 9 percent of women of reproductive age experience severe symptoms during the luteal phase of the menstrual cycle that meet criteria for premenstrual dysphoric disorder (PMDD). The management of these women remains highly controversial, but treatment with antidepressants has been consistently effective. Most evidence exists for the selective serotonin reuptake inhibitors (SSRIs) fluoxetine and sertraline. Because patients only have symptoms intermittently and usually respond rapidly to SSRI therapy, intermittent rather than continuous SSRI treatment has been advocated. Halbreich and colleagues studied the effect of intermittent sertraline in 281 women who met criteria for severe PMDD.

Patients 24 to 45 years of age were recruited by advertisement and referral to 14 psychiatric or gynecologic centers in the United States and Canada. Patients were required to have had regular menstrual cycles and to have reported PMDD symptoms for at least two years. The diagnosis was confirmed by the standardized Daily Record of Severity of Problems for two consecutive menstrual cycles showing scores at least 75 percent higher during the five most symptomatic days in the week before menses compared with the midfollicular phase.

Patients with high scores on depression screening and those using oral contraceptives were excluded from the study. Other reasons for exclusion included history of significant psychiatric illness, use of antidepressants, and anovulatory state, including women who had undergone hysterectomy. All patients had previously responded adequately to treatment of their PMDD with antidepressants.

After all patients initially completed one cycle of placebo treatment, only those whose symptoms continued and who demonstrated compliance were entered into the final study. The 142 women randomly allocated to sertraline therapy began with 50 mg daily on the 14 days before the anticipated onset of bleeding. Physicians could increase the dosage to 100 mg per day if clinically indicated. All patients completed the Daily Record of Severity of Problems throughout the three cycles of the study.

The treated patients were comparable to those receiving placebo in all relevant variables. The typical patient was white, 36 years of age, well educated, married with one or two children, and had a mean cycle length of 27.5 days. Before treatment, the mean severity scores were 73 during the luteal phase and 31 in the midfollicular phase.

Women taking sertraline improved more than those taking placebo in the first cycle of treatment and sustained improvement for all three cycles monitored. The improvement was statistically significant in all measures of depression and anxiety and in the Daily Record of Severity of Problems. Women receiving sertraline did not show significantly greater improvement in physical symptoms such as headache, breast tenderness, or bloating when compared with women receiving placebo. Nevertheless, measures of the net impact on daily life showed significant advantage for sertraline treatment. By the third cycle, 63 percent of the patients taking sertraline and 46 percent of those taking placebo had scored within 10 percent of community norms for measures of life satisfaction. Women taking sertraline reported more headache, nausea, dry mouth, insomnia, and diarrhea than those taking placebo, and 8 percent discontinued sertraline therapy because of side effects.

The authors conclude that using sertraline during the luteal phase provides effective relief of PMDD symptoms. This result correlates with other studies suggesting that intermittent therapy with SSRIs can be effective in women with this condition.

editor's note: Few topics are more controversial or more difficult to treat objectively than the various types of premenstrual syndrome (PMS). The authors of this paper have published extensively on the topic and are well-known advocates for vigorous treatment of all the manifestations of PMS. All of the measures used, although validated, depend on subjective reports by patients or clinicians. With selected patients and enthusiastic, expert clinicians, the placebo effect is powerful and is unlikely to be replicable in everyday practice. The placebo groups in this study achieved significant improvements over baseline. Nevertheless, even in this very select situation, the reported mean gains in measures have wide ranges, indicating that individual women had very different experiences. What is the “bottom line” for us in practice? Probably that we do not have perfect treatments for every woman who has premenstrual symptoms but that empathic objective approaches are beneficial and that we need to pay almost as much attention to how we negotiate treatments for individual women as to what we recommend. Very few of the current therapies can be supported by systematic evidence-based reviews, but short courses of selective serotonin reup-take inhibitors every month could be useful in some women.—a.d.w.