Issues of time, development and (natural) course have always attracted me, not only now at the turn both of the century and the second millenium where there is a proliferation of (sometimes obscure) books on time.
There is a dearth of treatment studies in AN, in sharp contrast to BN. However, Russell and co-workers have extended their classic controlled trial on family therapy and individual supportive therapy to a 5-year follow-up . Psychological treatments are rightly viewed as the major treatment modality in eating disorders. Nevertheless it is one of the ®rst clinical trials to look at the long-term course of such therapies in AN. Four groups of patients were compared: early onset (418 years) with short (43 years) or long (53 years) duration of illness and later onset AN and low weight BN (bulimic anorexia). Clear differences between the four groups emerged, with the long duration group and the BN group showing least improvement. The most striking fact was that some long-term bene®ts of psychological treatments were still discernible after 5 years: family therapy was better for early onset short duration AN and individual therapy for late onset AN. Thus, for adolescent patients family therapy is the treatment of choice even viewed from the long run! As for prognosis, the status 15 months after presentation (including a roughly 3-month weight restoration period) seems to be most important. For the majority of cases (52/80; 65 per cent) the category at the end of treatment is predictive of follow-up status, be it good, intermediate or poor. This would favour the static continuum hypothesis, even over a period of about 5 years after outpatient treatment. However, amelioration of status at follow-up (25/80; 31 per cent) is more common than deterioration (3/80) in this sample. Even for poor outcome at end of treatment there is a 38 per cent chance of improvement by 5 years later including a 19 per cent chance of moving to the good outcome group. So we might discern early and late responders (within roughly 1 year