Citation

Keel PK, Dorer DJ, Eddy KT, Franko D, Charatan DL, Herzog DB. Arch. Gen. Psychiatry 2003; 60(2): 179-183.

Copyright

(Copyright © 2003, American Medical Association)

DOI

10.1001/archpsyc.60.2.179

PMID

12578435

Abstract

BACKGROUND: Anorexia nervosa, but not bulimia nervosa, has one of the highest mortality rates among psychiatric disorders. However, potential predictors of mortality, such as comorbid psychiatric illnesses, remain unclear. We sought to determine mortality ratios and predictors of fatal outcome in women diagnosed as having anorexia or bulimia nervosa.
METHODS: Women (N = 246) diagnosed as having either DSM-IV anorexia nervosa (n = 136) or bulimia nervosa (n = 110) between January 1, 1987, and December 31, 1991, participated in a prospective longitudinal study. Vital status was determined by ongoing contact and a National Death Index search as of December 1998 (overall ascertainment, 99.8%) and telephone contact as of October 2000 (ascertainment, 95.0%).
RESULTS: Eleven women died. Standardized mortality ratios were elevated for all causes of mortality (11.6; 95% confidence interval, 5.5-21.3) and suicide (56.9; 95% confidence interval, 15.3-145.7) in anorexia nervosa but not for death (1.3; 95% confidence interval, 0.0-7.2) in bulimia nervosa. Predictors of mortality in anorexia nervosa included severity of alcohol use disorder during follow-up (P<.001). Hospitalization for an affective disorder before baseline assessment seemed to protect women from a fatal outcome (P<.001).
CONCLUSIONS: Physicians treating patients with anorexia nervosa should carefully assess patterns of alcohol use during the course of care because one third of women who had alcoholism and died had no history of alcohol use disorder at intake.

Language: en

Keywords

Alcoholism; Anorexia Nervosa; Bulimia; Cause of Death; Comorbidity; Diagnosis, Dual (Psychiatry); Feeding and Eating Disorders; Female; Follow-Up Studies; Humans; Life Tables; Longitudinal Studies; Probability; Prospective Studies