REGULAR ARTICLE

The Mortality Rate from Anorexia Nervosa C. Laird Birmingham, MD, MHSc1,2* Jenny Su, BSc1 Julia A. Hlynsky, BSc1 Elliot M. Goldner, MD, MHSc1,2 Min Gao, PhD2

ABSTRACT Objective: We determined the standardized mortality ratio (SMR) in our anorexia nervosa (AN) patient population.

Results: Of 954 patients, 326 diagnosed with AN completed an assessment over the 20 years. The SMR was 10.5 (95% confidence interval [CI] ¼ 5.5–15.5) for AN.

Method: We used a cross-sectional design to study an inception cohort (1981–2000) drawn from the provincial tertiary care eating disorders program at St. Paul’s Hospital (British Columbia, Canada). All patients who completed their initial assessment for an eating disorder were included in the study. Vital status, date and cause of death from British Columbia Vital Statistics Agency, date of assessment, date of birth, and diagnosis at the time of assessment were collected for each patient.

Discussion: Some studies in the literature report that AN has the highest mortality rate of any psychiatric disorder in young females. However, others dispute this fact and report an SMR lower than the normal population mortality (SMR ¼ 0.71). Contrary to some reports in the literature, our study confirms a high mortality rate within the AN population. ª 2005 by Wiley Periodicals, Inc.

Introduction The standardized mortality ratio (SMR; the ratio of observed to expected deaths) for anorexia nervosa (AN) has been reported to be between 0.71 and 17.8 (Ben-Tovim et al., 2001; Crisp, Callender, Halek, & Hsu, 1992; Deter & Herzog, 1994; Eckert, Halmi, Marchi, Grove, & Crosby, 1995; Emborg, 1999, 2001; Herzog et al., 2000; Keel et al., 2003; Korndorfer et al., 2003; Lee, Chan, & Hsu, 2003; Lowe et al., 2001; Moller-Madsen, Nystrup, & Nielsen, 1998; Nielsen et al., 1998; Strober, Freeman, & Morrell, 1997; Tolstrup et al., 1985; Wentz, Gillberg, Gillberg, & Rastam, 2001) (Table 1). The factors that correlate with a higher estimate are age, case severity, study period, and whether other eating disorders with a lower mortality were evaluated separately. Another important factor is the diagnostic classification used. Over the years, standards of diagnosis have changed considerably. The introduction of criteria outlined in the 4th ed. of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV;

Accepted 11 August 2004 *Correspondence to: C. Laird Birmingham, MD, Eating Disorders Program, St Paul’s Hospital, 1081 Burrard Street, Vancouver, British Columbia, Canada, V6Z 1Y6. E-mail: [email protected] 1 Department of Psychiatry, University of British Columbia, Vancouver, British Columbia, Canada 2 Centre for Health Evaluation and Outcome Sciences, University of British Columbia, Vancouver, British Columbia, Canada Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/eat.20164 ª 2005 Wiley Periodicals, Inc.

Int J Eat Disord 38:2 143–146 2005

Keywords: anorexia nervosa; standardized mortality ratio; psychiatric disorders (Int J Eat Disord 2005; 38:143–146)

American Psychiatric Association [APA], 1994) created a new subtype of eating disorders, that is, eating disorders not otherwise specified (EDNOS). The mortality rate before DSM-IV will be lower because it is based on a composite of AN and EDNOS (which has a lower mortality rate than AN alone) (Figure 1).

Methods We studied a cohort of 954 consecutive patients who were referred to the only adult tertiary care eating disorders program in the province of British Columbia (BC), Canada, located at St. Paul’s Hospital in Vancouver, over a 20-year period (November 1981 to July 2000). Complete psychosocial and medical assessments were performed on all patients. The diagnosis was made according to criteria in the 3rd ed. of the Diagnostic and Statistical Manual of Mental Disorders (DSM-III; APA,1980) in use at the time of the assessment. To assess the vital status of each patient, a search based on name, gender, and date of birth was conducted through the BC Vital Statistics Agency, a branch of the BC Ministry of Health. For each death record match, the date and cause of death by ICD-10 code were reported. The number of deaths due to all causes was used as the observed death in the calculation of the SMRs. Calculations of SMRs were made based on indirect methods of determining standardizations. The expected number of deaths was obtained by applying the age, gender, and

143

BIRMINGHAM ET AL. TABLE 1. Standardized mortality rates of previous epidemiology studies on anorexia nervosa (data are arranged in decreasing order of SMR)

Authors Norring and Sohlberg (1993) Deter and Herzog (1994) Eckert, Halmi, Marchi, Grove, and Crosby (1995) Keel et al. (2003) Su and Birmingham (2001) Lee, Chan, and Hsu (2003) Lowe et al. (2001) Herzog et al. (2000) Jorgenson (1992) Moller-Madsen, Nystrup, and Nielsen (1996) Pagsberg and Wang (1994) Emborg (2001) Patton (1988) Tolstrup et al. (1985) Crisp, Callender, Halek, and Hsu (1992) Aberdeen Cohort Theander (1970) Korndorfer et al. (2003) Ben-Tovim et al. (2001) Strober, Freeman, and Morrell (1997) Wentz, Gillberg, Gillberg, and Rastam (2001)

n

48 84 76 136 954 88 84 136 132 1,035 50 2,763 460 151

63 94 208 95 95 51

Mean Length of Follow-Up Period (Years)

Population

Diagnoses

6.0 11.8 10.0 8.6 8.2 4 21.0 11.0 11.7 8.0 6.2 10.3 7.2 22.9

Patient Patient Patient Patient Patient Patient Patient Patient Population Patient Population Patient Patient Patient

DSM-III-R DSM-III-R DSM-IV DSM-IV DSM-IV DSM-III-R DSM-IV DSM-IV DSM-III-R ICD-8 ICD-10 ICD-8 undefined undefined

6.25 10.7 6.6 7.4 5.2 3.4 16.0 5.1 16.7 6.9 6 8.4 3.04 11.9

17.8 14.4 13.0 11.6 10.5 10.5 9.8 9.6 7.15 6.96 6.92 6.69 6.17 4.80

22.1 36.2 27.1 5 10–15 10

Population Patient Population Patient Patient Population

DSM-III-R undefined DSM-III-R DSM-IV DSM-III undefined

12.7 25.5 8.0 3.16 0 0

4.71 2.93 0.71

Crude Mortality Rate (%)

SMR

Note: n ¼ number of subjects; patient ¼ subjects defined as being treated/hospitalized at some stage during study; population ¼ subjects drawn from a general population, no hospital screening; crude mortality rate ¼ number of deaths as a percentage of the subject population; SMR ¼ standardized mortality ratio, i.e., the ratio of observed to expected deaths.

year-specific mortalities of the general BC population to the corresponding cumulative person-years of the study cohort.

Results The cohort consisted of 474 subjects with bulimia nervosa (BN; 49.7%), 326 subjects with AN (34.2%), and 154 subjects with EDNOS (16.1%) (Table 2). The mean follow-up interval for our study cohort was 8.7 years (SD ¼ 5.2). The mean follow-up intervals for AN, BN, and EDNOS, were, respectively, 7.3 (SD ¼ 4.9), 10.3 (SD ¼ 5.0), and 7.0 (SD ¼ 4.9) years. At follow-up, 25 of the 954 patients (2.6%) had died (Table 3). Of the 25 patients, 17 came from the AN subject population. Causes of death for the 17 patients with AN were suicide (n ¼ 7), pneumonia (n ¼ 2), hypoglycemia (n ¼ 2), liver disease (n ¼ 2), cancer (n ¼ 2), alcohol poisoning (n ¼ 1), and subdural hemorrhage (n ¼ 1). The SMR for AN was 10.5 (95% confidence interval [CI] ¼ 5.5–15.5). Only one death was reported for EDNOS, which is of unknown cause. The SMR of EDNOS was 1.1 (95% CI ¼ 0.0–3.2).

144

Discussion Our study confirms the association of AN with a clinically important increase in mortality rate. Suicide was the most common cause of mortality in AN. There is little reported about the morbidity or mortality associated with EDNOS because it was not an accepted diagnosis until the DSMIV criteria were published. Our subjects were tertiary care referrals, so the SMRs of population-based and primary care cohorts may be different. A recent study from the Mayo Clinic found that AN did not have an increased mortality rate above that reported for controls (Korndorfer et al., 2003; Palmer, 2003; Sullivan, 2003). The Mayo Clinic study used unselected (nonreferred) patients who had lived in Rochester, Minnesota, for at least a year and met criteria for AN as outlined in the 3rd Rev. ed. of the Diagnostic and Statistical Manual of Mental Disorders (DSM-III-R; APA, 1987). Korndoffer et al. (2003) stated in their article that the relatively benign prognosis compared with previous findings is related to the selection of patients from the general population, creating a much milder clinical spectrum. Our cohort was made up of patients with eating disorders referred to our tertiary care program for AN. In addition, the Mayo Clinic study may include patients with EDNOS. EDNOS did not exist in the DSM-III-R Int J Eat Disord 38:2 143–146 2005

MORTALITY FROM AN FIGURE 1. Standardized mortality ratio (SMR) of anorexia nervosa. The studies are arranged in decreasing order of SMR. To the left of the dashed line use DSM-IV criteria; to the right, use earlier criteria (i.e., DSM-III-R, DSM-III, ICD-8, ICD-10).

TABLE 2. Subject characteristics Age at Initial n All ED patients AN BN EDNOS

954 326 474 154

Female

Assessment (SD) 26.1 24.7 26.1 28.5

(8.6) (9.6) (7.4) (9.3)

n 927 312 466 149

Male

Average Age Average Age (Years) (SD) n (Years) (SD) 26.0 24.6 26.0 28.6

(8.6) (9.6) (7.4) (9.1)

27 14 8 5

29.3 27.9 34.2 24.8

(9.0) (7.5) (5.9) (15.9)

Note: n ¼ number of subjects; ED ¼ eating disorder; AN ¼ anorexia nervosa; BN ¼ bulimia nervosa; EDNOS ¼ eating disorder not otherwise specified. TABLE 3.

Eating disorder cohort mortality (SMR) Cases n

AN BN EDNOS All

criteria that were used by the Mayo Clinic to define their patient population. This would tend to lower the mortality rate of the Rochester patients towards that of our EDNOS patients (SMR ¼ 1.1). The strengths of our study are the large sample size, the long follow-up period, and the nature of the cohort. As the sole tertiary-care center for adult eating-disordered patients in the province of BC, all provincial referrals are directed to St. Paul’s Hospital.

326 474 154 954

Deceased

Female (%)

n

Male (%)

Age at Death (M ± SD)

Time to Death (M ± SD)

312 466 148 927

17 7 1 25

16 (94.1) 6 (85.7) 1 (100.0) 23 (92.0)

36.3 ± 10.7 39.9 ± 11.3 46.3 37.7 ± 10.7

6.2 ± 4.8 4.3 ± 4.2 11.0 5.9 ± 4.7

(95.7) (98.3) (96.8) (97.2)

SMR (95% Cl) 10.5 2.0 1.1 4.1

(5.5–15.5) (0.5–3.5) (0–3.2) (2.5–5.7)

Note: AN ¼ anorexia nervosa; BN ¼ bulimia nervosa; EDNOS ¼ eating disorder not otherwise specified; SMR ¼ standardized mortality ratio. Int J Eat Disord 38:2 143–146 2005

145

BIRMINGHAM ET AL.

The limitations of our study include the following: the eating disorder diagnoses were based on the three revisions of the DSM that were published over the 20-year interval of our study (DSM-III, 1980; DSMIII-R, 1987; DSM-IV, 1994); and the method of ascertaining death by record linkage may underestimate the mortality rate, as our sample is an open cohort and subjects may have moved outside the province.

References American Psychiatric Association. (1980). Diagnostic and statistical manual of mental disorders (3rd ed.) Washington, DC: Author. American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd Rev. ed.) Washington, DC: Author. American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. Ben-Tovim, D.I., Walker, K., Gilchrist, P., Freeman, R., Kalucy, R., & Esterman, A. (2001). Outcome in patients with eating disorders: A 5-year study. Lancet, 357(9264), 1254–1257. Birmingham, C., Su, J., Hlynsky, J., Goldner, E., & Gao, M. (2005). The mortality rate from anorexia nervosa. IJED, 38(2), 1–4. Crisp, A.H., Callender, J.S., Halek, C., & Hsu, L.K. (1992). Long-term mortality in anorexia nervosa. A 20-year follow-up of the St George’s and Aberdeen cohorts. British Journal of Psychiatry, 161, 104–107. Deter, H.C., & Herzog, W. (1994). Anorexia nervosa in a long-term perspective: Results of the Heidelberg-Mannheim Study. Psychosomatic Medicine, 56(1), 20–27. Eckert, E.D., Halmi, K.A., Marchi, P., Grove, W., & Crosby, R. (1995). Ten-year follow-up of anorexia nervosa: Clinical course and outcome. Psychological Medicine, 25(1), 143–156. Emborg, C. (1999). Mortality and causes of death in eating disorders in Denmark 1970–1993: A case register study. International Journal of Eating Disorders, 25(3), 243–251. Emborg, C. (2001). [Mortality and causes of death in patients with eating disorders in Denmark, 1970–1993]. Ugeskrift for laeger, 163(25), 3476–3480. Fichter, M.M., & Quadflieg, N. (1999). ergSix-year course and outcome of anorexia nervosa. International Journal of Eating Disorders, 26(4), 359–385. Herzog, D.B., Greenwood, D.N., Dorer, D.J., Flores, A.T., Ekeblad, E.R., Richards, A., Blais, M.A., Keller, M.B. (2000). Mortality in eating disorders: a descriptive study. Int J Eat Disord, 28(1), 20–26. International Classification of Disease. (1993). Mental and behavioral disorders. Diagnostic criteria for research (10th ed.). Geneva: World Health Organization. Jorgensen, J. (1992). The epidemiology of eating disorders in Fyn County, Denmark, 1977–1986. Acta Psychiatr Scand, 85(1), 30–34.

146

Keel, P.K., Dorer, D.J., Eddy, K.T., Franko, D., Charatan, D.L., & Herzog, D.B. (2003). Predictors of mortality in eating disorders. Archives of General Psychiatry, 60(2), 179–183. Korndorfer, S.R., Lucas, A.R., Suman, V.J., Crowson, C.S., Krahn, L.E., & Melton, L.J., 3rd. (2003). Long-term survival of patients with anorexia nervosa: A population-based study in Rochester, Minn. Mayo Clinic Proceedings, 78(3), 278–284. Lee, S., Chan, Y.Y., & Hsu, L.K. (2003). The intermediate-term outcome of Chinese patients with anorexia nervosa in Hong Kong. American Journal of Psychiatry, 160(5), 967–972. Lowe, B., Zipfel, S., Buchholz, C., Dupont, Y., Reas, D.L., & Herzog, W. (2001). Long-term outcome of anorexia nervosa in a prospective 21-year follow-up study. Psychological Medicine, 31(5), 881–890. Moller-Madsen, S.M., Nystrup, J., & Nielsen, S. (1998). [Mortality of anorexia nervosa in Denmark 1970–1987]. Ugeskrift for laeger, 160(38), 5509–5513. Nielsen, S., Moller-Madsen, S., Isager, T., Jorgensen, J., Pagsberg, K., & Theander, S. (1998). Standardized mortality in eating disorders—a quantitative summary of previously published and new evidence. J Psychosom Res, 44(3–4), 413–434. Norring, C.E., & Sohlberg, S.S. (1993). Outcome, recovery, relapse and mortality across six years in patients with clinical eating disorders. Acta Psychiatr Scand, 87(6), 437–444. Pagsberg, A.K., & Wang, A.R. (1994). Epidemiology of anorexia nervosa and bulimia nervosa in Bornholm County, Denmark, 1970–1989. Acta Psychiatr Scand, 90(4), 259–265. Palmer, R.L. (2003). Death in anorexia nervosa. Lancet, 361(9368), 1490. Patton, G.C. (1998). Mortality in eating disorders. Psychol Med, 18(4), 947–951. Steinhausen, H.C. (2002). The outcome of anorexia nervosa in the 20th century. American Journal of Psychiatry, 159(8), 1284– 1293. Strober, M., Freeman, R., & Morrell, W. (1997). The long-term course of severe anorexia nervosa in adolescents: Survival analysis of recovery, relapse, and outcome predictors over 10–15 years in a prospective study. International Journal of Eating Disorders, 22(4), 339–360. Sullivan, P.F. (2003). Discrepant results regarding long-term survival of patients with anorexia nervosa? Mayo Clinic Proceedings, 78(3), 273–274. Theander, S. (1970). Anorexia nervosa. A psychiatric investigation of 94 female patients. Acta Psychiatr Scand Suppl, 214, 1–194. Tolstrup, K., Brinch, M., Isager, T., Nielsen, S., Nystrup, J., Severin, B., olesen, N. S. (1985). Long-term outcome of 151 cases of anorexia nervosa. The Copenhagen Anorexia Nervosa Follow-Up Study. Acta Psychiatrica Scandinavica, 71(4), 380–387. Wentz, E., Gillberg, C., Gillberg, I.C., & Rastam, M. (2001). Ten-year follow-up of adolescent-onset anorexia nervosa: Psychiatric disorders and overall functioning scales. Journal of Child Psychology and Psychiatry, and Allied Disciplines, 42(5), 613–622.

Int J Eat Disord 38:2 143–146 2005