Diagnosis of premenstrual tension syndrome: description and evaluation of a procedure for diagnosis and differential diagnosis

Abstract

A method for diagnosing premenstrual tension syndrome (PMS) with daily prospective symptom self-ratings is described. Statistical criteria based on non-parametric statistics are developed for how large a deviation from the ideal PMS pattern is acceptable before the patient is considered significantly different from the ideal PMS pattern. Eighty-two consecutive patients seeking help for PMS were asked to keep daily prospective symptom records for two menstrual cycles. Blood samples for estradiol and progesterone radioimmunoassay were taken at least once weekly. In 80 of the patients ratings in at least one ovulatory cycle were useful. All patients also completed an Eysenck Personality Inventory (EPI) and were asked for earlier psychiatric history (PH). By using the statistical criteria it was possible to divide the patients into three categories. Group 1, patients with only symptoms during the premenstrual phase ‘pure PMS’ (30%); Group 2, patients with a ‘premenstrual aggravation’ of negative symptoms existing during the whole menstrual cycle (56%); and Group 3, patients with no relationship between the symptom variations and different phases of the menstrual cycle (14%). Group 1 showed significantly lower neuroticism score on the EPI and had a significantly lower frequency of patients with earlier PH than both Groups 2 and 3.