Suicide Risk among Eating Disorders Patients

Suicide is the second leading cause of death (after cardiac disease) among persons with anorexia nervosa (AN), and a major cause of death among people with eating disorders (Int J Eat Disord 2007; 40: 293). Results of studies in Spain and Italy provide new information on the risk of suicide among these patients.

Suicide and personality

A collaborative team led by Dr. Laura Forcano and colleagues in Barcelona, and Dr. Cynthia Bulik at the University of North Carolina at Chapel Hill, recently evaluated the relationship between suicide and personality among 172 patients with AN (Compr Psychiatry 2011; 52:352). The team examined the prevalence of lifetime suicide attempts in women with AN, and then factored in eating disorder symptoms, general psychopathology, and personality in a healthy control group and then across groups with restricting and purging subtypes of AN.

The study group was diagnosed according to DSM-IV criteria by experienced psychologists and psychiatrists. The mean age was 24.7 years; the mean age at onset of illness was 19.7 years; and the mean duration of illness was 5 years. Mean weekly number of binge-eating episodes was 0.8 and the number of mean weekly vomiting episodes was 2.3. The mean body mass index (kg/m²) was 15.6. Most patients were single and unemployed, and had completed primary or secondary schooling. The control group included 146 female undergraduate students (mean age 21.1 years and mean BMI 21.1). In contrast to those in the patient group, most controls were engaged to be married and employed, and had completed their secondary education.

To assess the lifetime incidence of suicide attempts, all participants were asked, “Have you ever attempted suicide?” during structured clinical face-to-face interviews. To be qualified as a suicide attempt, the suicidal gesture had to be self-destructive with intent to die. Study participants also completed a series of questionnaires, including the Symptom Checklist- Revised and the Eating Disorders Inventory-2.

Different patterns among the groups

The prevalence of suicide attempts differed significantly across the three groups studied. The prevalence was 0% among controls, 8.65% in the restricting AN group, but 25% in the purging AN group. Depression measures were elevated in those with suicide attempts. Among the women in the restricting AN group, those who attempted suicide scored significantly higher on Phobic Activity, measured by the Symptom Checklist-Revised, than those who did not.

The authors noted that the higher prevalence of suicide attempts among the purging AN patients supported the theory that suicide attempts in AN patients are largely concentrated among those who have purging as part of their symptom profile. Purging behaviors may signal greater psychopathology and impulsivity, as well as more dysfunctional expression of anger, which may increase the risk for suicide.

Patients with at least one suicide attempt during their lifetime also had greater depressive symptoms, as measured by the Symptom Checklist-90; this finding echoes results of previous studies. The authors noted that phobic anxiety was the one measure that distinguished between restricting AN patients with and without suicide attempts. The Phobic Anxiety scale measures symptoms associated with phobic anxiety states, such as feeling fear in open spaces, being afraid to leave one’s home, and feeling nervous when left alone. The authors also suggest that clinicians be particularly vigilant for possible suicidal ideation in persons with restricting AN who show symptoms of agoraphobia. Among restricting AN patients, greater anxiety may signal greater suicide risk.

A meta-analysis of suicide risk

A meta-analysis of 40 suicide risk studies among eating disorders patients showed that patients with AN and BN share a number of risk factors for suicide and that although the suicide risk among patients with AN has fallen over the past decades, these patients are still at higher risk for suicide than are persons in the general population (Acta Psychiatr Scand 2011; 124:6). The reduction was attributed to improved detection of patients in need of treatment, improved diagnostic criteria and reduced stigma against receiving treatment for mental disordersalong with an expansion of services for people with eating disorders.

Antonio Preti, MD, and colleagues at Centro Medico Gennuruxi, Cagliari, Italy, did a PubMed/Medline search for studies including more than 40 persons, where follow-up lasted at least 5 years. The researchers identified 40 studies of patients with AN, 16 studies of BN, and 3 of binge eating disorder. Of 16,342 patients with AN, 245 suicides were reported over a mean follow-up of 11.1 years (the suicide rate was 0.124 per 100 person-years).

Higher-than-expected suicide rates were found among patients with BN, but the rates were still lower than those among AN patients. The crude suicide rate for BN patients was 0.20%, and the average suicide rate was per study was 0.20%. The suicide rate was 0.030 per 100 person-years. Among the general population, the crude suicide rate was 0.21%, with a suicide rate of 0.004 per 100 person-years.

The lower suicide mortality rates for persons with BN were puzzling, since the patients in both AN and BN groups share numerous risk factors. In some samples, higher suicide attempt rates were reported among BN patients than among restrictive AN patients. One explanation might be that patients with AN could be more exposed to the lethal outcome of suicide attempts because of the complications of starvation. Another is the stigma against suicide present in some countries.

The authors recommend that follow-up data on patients include data on suicide, and that patients lost at follow-up also be reported. In addition, the authors called for a worldwide effort to study suicide among patients with BN through establishment of a consortium to study the risk of suicide in patients with eating disorders.