Lack of effect of induced menses on symptoms in women with premenstrual syndrome.

BACKGROUND

No physiologic abnormality of the luteal phase has been consistently demonstrated in women with premenstrual syndrome (PMS). Using the progesterone antagonist mifepristone, we truncated the late luteal phase of the menstrual cycle in a blinded fashion to evaluate the relation of the events of the late luteal phase to the symptoms of PMS.

METHODS

Fourteen women with PMS were given mifepristone (12.5 or 25 mg per kilogram of body weight) by mouth on the seventh day after the surge of luteinizing hormone. On the sixth through the eighth days after the surge, they also received injections of either placebo or human chorionic gonadotropin (2000 IU). Seven women with PMS received placebo instead of both mifepristone and human chorionic gonadotropin. All the women completed daily questionnaires measuring a variety of mood-related and somatic symptoms.

RESULTS

Mifepristone consistently induced menses. The women receiving only mifepristone had plasma progesterone levels like those of the follicular phase (less than 3 nmol per liter) within four days, whereas all the other women had plasma progesterone levels characteristic of the luteal phase (greater than 8 nmol per liter) for at least seven days after treatment. In all three groups, the severity of symptoms was significantly higher after treatment than before, according to an analysis of variance with repeated measures. The level and pattern of the ratings of symptom severity were similar in all treatment groups.

CONCLUSIONS

Neither the timing nor the severity of PMS symptoms was altered by mifepristone-induced menses or luteolysis. The temporal association of typical PMS symptoms with an artificially induced follicular phase suggests that endocrine events during the late luteal phase do not directly generate the symptoms of PMS.