Usefulness of amenorrhoea in the diagnoses of eating disorder patients.

We investigated the usefulness of amenorrhoea as a criterion in eating disorder diagnoses. Menstrual history, eating disordered behavior and weight history of 251 female eating disorder inpatients were assessed on admission to hospital. Menstrual status of 88 (35%) could not be assessed (80 taking hormonal contraception, 5 (< 16 years) had never menstruated, one hysterectomized, 2 postmenopausal). Of the remaining 163; 90 had secondary amenorrhoea (no periods for > 3 months), 19 irregular periods and 54 regular cycles. Some patients with recent changes in energy status, who warranted inpatient treatment for their eating disorder did not develop amenorrhoea until later during their admission. Menstrual disturbance is not limited to anorexia nervosa, 24% of patients with an eating disorder not otherwise specified (EDNOS) and 15% of bulimic patients had amenorrhoea/oligomenorrhoea on admission. The best predictors (82% cases, 83% noncases, R2 = 0.41) of secondary amenorrhoea at admission, were; current BMI < or = 18, and having rules for exercising. Menstrual status of women with an eating disorder diagnosis cannot always be assessed, is mostly measuring low body weight and exercise and is not useful in planning or initiating treatment. There is insufficient evidence to support the use of amenorrhoea as a criterion for any eating disorder.