Death and relapse in anorexia nervosa: survival analysis of 151 cases

The traditional historic-prospective follow-up study of anorexia nervosa patients may be well served by recent methods of statistical analysis, the so-called models for the analysis of survival data. These models are particularly suited to the sort of incomplete observations usually produced a longitudinal studies. They include methods for estimating the probability of death and relapse as a function of time. This makes possible powerful comparisons among studies and among subsamples in single studies. In the present study, 151 patients were followed-up after 4-22 yrs. The hazard of death was on average 0.5% per year, the hazard of relapse 3% per year. With both measures, the risk declined steadily after therapeutic contact.