Psychologic components of the premenstrual syndrome. Evaluating the research and choosing the treatment.

Methodologic problems that impede the evaluation of the psychologic components of the premenstrual syndrome (PMS) include: (1) the lack of a consensus on the definition of PMS (resulting in significant variations between studies with respect to severity, timing and course of the symptoms); (2) the use of retrospective, rather than prospective, methods of data collection (which increases the likelihood of culturally based response bias); (3) sampling errors; (4) failure to assess and/or control for experimental bias introduced as a function of the subject's perception of the purpose of the evaluation; and (5) inadequate attention paid to timing within the cycle of data collection. Treatment of the psychologic components of PMS requires an individualized approach. Given our current knowledge of this syndrome (or syndromes), there is no way to predict with a high level of assurance which patients will benefit most from behavioral, pharmacologic, nutritional or hormonal therapy. The most urgent priorities in future research on PMS treatment are controls for placebo effects and more sophisticated approaches to sample selection.