Dr Needles questions whether the Frisch hypothesis might not hold if one would exclude the population with a "sick hypothalamus." There is no simple answer, because both menarche and the maintenance of regular periods are influenced by a variety of factors such as the family history, level of stress, amount and type of exercise, dietary intake, and many other events. Among them, the percentage of body fat can certainly be influential, but it is not, as has been statistically demonstrated, the determining factor.1 The best reply is to consider that loss of menses in anorexia nervosa is multifactorial. Hypothalamic dysfunction has classically been demonstrated in patients with anorexia nervosa by their immature secretory patterns of luteinizing hormone2 and their blunted response of luteinizing hormone to gonadotropin-releasing hormone.3 Almost a quarter century ago, Sherman et al3 showed normalization of the luteinizing hormone response to gonadotropin-releasing hormone when