Dr. Steinhausen Replies

To the Editor: The comment by my distinguished colleague Dr. Vandereycken is much appreciated. However, I feel that his reasoning is faulty. In my article, I warned that both the variability in my findings regarding prognostic factors and the likely nature of the data precluded any delineation of rules as to individual prognosis in a patient suffering from anorexia nervosa. Unfortunately, Dr. Vandereycken did not take this warning seriously and used a statement by Theander to argue against the potentially detrimental effects of outcome research findings for individual prognosis.

Furthermore, Dr. Vandereycken is mistaken regarding the findings presented in Figure 1 that show a subgroup of patients with adolescent onset of anorexia nervosa who developed a chronic course. It is reasonable to calculate the number of patients who had a chronic or protracted course less than 4 years after adolescent onset. This has been done repeatedly in various outcome studies that formed the basic data set of my review.

Similarly, “chronicity” has not been used only as a prognostic factor in various outcome studies. The association with outcome is also not tautological. Although the probability seems to be lower, protracted illnesses may show some improvement. In fact, the data shown in my Figure 1 indicated that with increasing duration of follow-up, the outcome of anorexia nervosa does improve.

In quoting my research on the outcome of subjects with adolescent-onset anorexia (1), Dr. Vandereycken kindly provided me with the opportunity to emphasize the variability of outcome findings, as documented in my review. However, the statistical analyses of the entire body of outcome studies in the last century, including the high rates of mortality, chronicity, and impairment and the relatively stable pattern across five decades, lead to the conclusion that anorexia nervosa did not lose its relatively poor prognosis in the 20th century.