Prognosis of Anorexia Nervosa

To the Editor: In his review of outcome studies concerning patients with anorexia nervosa, Hans-Christoph Steinhausen, M.D., Ph.D. (1), concluded that the disorder “did not lose its relatively poor prognosis in the 20th century” (p. 1284). Such a conclusion is pessimistic, as Theander (2) warned: “A weather forecast cannot influence the predicted events themselves. On the other hand, a prognosis concerning human problems may influence the further course considerably. A pessimistic prognosis for a line of business may start a negative trend.” Indeed, this influential Swedish researcher was the first to publish long-term outcome data that appear to be in line with Steinhausen’s conclusion but also “may restore hope and confidence in patients and their relatives, as well as their therapists, as they supply clear evidence that recovery may eventually occur even after a very long and severe illness” (2). For this reason, Theander rightly cautioned against flippant use of the term “chronicity” and proposed discussion of a “long-lasting” or “protracted” eating disorder (3).

In this respect, Dr. Steinhausen’s notion of chronicity should be seriously questioned (1). In fact, he used it as a synonym for “poor outcome.” This may lead to illogical and confusing labeling of young adolescents as patients with “chronic” illness (Figure 1), even if their follow-up was for less than 4 years! I am also puzzled by his use of chronicity as a prognostic factor (Table 5) and his tautological conclusion that “chronicity leads to poor outcome” (p. 1288). In view of the expectation that his otherwise very instructive review will be often quoted, I wished that Dr. Steinhausen had been more prudent in his phrasing. In fact, his own data have shown that with increasing duration of follow-up, the rate of recovery can rise to more than 70%. The prognostic picture of anorexia nervosa should include this more optimistic conclusion, in reference to Theander’s wise words, although I do not underestimate the seriousness of the disorder.