Prognostic factors and outcome in anorexia nervosa: a follow-up study.

BACKGROUND AND AIMS: Anorexia nervosa is an eating disorder characterized by food restriction, irrational fear of gaining weight and consequent weight loss. High mortality rates have been reported, mostly due to suicide and malnutrition. Good outcomes largely vary between 18 and 42%. We aimed to assess outcome and prognostic factors of a large group of patients with anorexia nervosa. Moreover we aimed to identify clusters of prognostic factors related to specific outcomes.

METHODS: We retrospectively reviewed data of 100 patients diagnosed with anorexia nervosa previously hospitalized in a tertiary level structure. Then we performed follow-up structured telephone interviews.

RESULTS: We identified four dead patients, while 34% were clinically recovered. In univariate analysis, short duration of inpatient treatment (p = 0.003), short duration of disorder (p = 0.001), early age at first inpatient treatment (p = 0.025) and preserved insight (p = 0.029) were significantly associated with clinical recovery at follow-up. In multiple logistic regression analysis, duration of first inpatient treatment, duration of disorder and preserved insight maintained their association with outcome. Moreover multiple correspondence analysis and cluster analysis allowed to identify different typologies of patients with specific features. Notably, group 1 was characterized by two or more inpatient treatments, BMI ≤ 14, absence of insight, history of long-term inpatient treatments, first inpatient treatment ≥30 days. While group 4 was characterized by preserved insight, BMI ≥ 16, first inpatient treatment ≤14 days, no more than one inpatient treatment, no psychotropic drugs intake, duration of illness ≤4 years.

CONCLUSIONS: We confirmed the association between short duration of inpatient treatment, short duration of disorder, early age at first inpatient treatment, preserved insight and clinical recovery. We also differentiated patients with anorexia nervosa in well-defined outcome groups according to specific clusters of prognostic factors. Our study might help clinicians to evaluate prognosis of patients with anorexia nervosa.