Frequency and correlates of lifetime suicidal ideation and suicide attempts among consecutively hospitalized youth with anorexia nervosa and bulimia nervosa: results from a retrospective chart review

Abstract

Background: Youth with eating disorders (EDs) face an increased risk of a premature suicide death. Precursors of completed suicide are suicidal ideation and suicide attempts, which need to be well understood to prevent suicide. However, epidemiological data on the lifetime prevalence and clinical correlates of suicidal ideation and suicide attempts (i.e., “suicidality”) are lacking for the vulnerable group of inpatient ED youth. Methods: This retrospective chart review was conducted at a psychiatric child and adolescent inpatient department, covering a 25-year period. Consecutively hospitalized youth with an ICD-10 diagnosis of anorexia nervosa (AN), restricting type (AN-R), binge-purging type (AN-BP), and bulimia nervosa (BN) were included. Data extraction and coding were standardized with trained raters extracting information from patient records according to a procedural manual and using a piloted data extraction template. The lifetime prevalence of suicidal ideation and suicide attempts was calculated for each ED subgroup, and clinical correlates of suicidality were analyzed via multivariable regression analyses. Results: In the sample of 382 inpatients aged 9–18 years (median age = 15.6, females = 97.1%; AN-R: n = 242, BN: n = 84, AN-BP: n = 56), 30.6% of patients had lifetime suicidal ideation (BN:52.4% ≈ AN-BP:44.6% > AN-R:19.8%, χ2(2,382) = 37.2, p < 0.001, Φ = 0.31), and 3.4% of patients reported a history of suicide attempts (AN-BP:8.9% ≈ BN:4.8% > AN-R:1.7%, χ2(2,382) = 7.9, p = 0.019, Φ = 0.14). Independent clinical correlates of suicidality were i) for AN-R a higher number of psychiatric comorbidities (OR = 3.02 [1.90, 4.81], p < 0.001), and body weight < 1st BMI percentile at hospital admission (OR = 1.25 [1.07,1.47], p = 0.005) (r2 = 0.20); ii) for AN-BP patients a higher number of psychiatric comorbidities (OR = 3.68 [1.50, 9.04], p = 0.004) and history of childhood abuse (OR = 0.16 [0.03, 0.96], p = 0.045) (r2 = 0.36), and iii) for BN patients a higher prevalence of non-suicidal self-injury (NSSI)(OR = 3.06 [1.37, 6.83], p = 0.006) (r2 = 0.13). Conclusions: About half of youth inpatients with AN-BP and BN had lifetime suicidal ideation, and one-tenth of patients with AN-BP had attempted suicide. Treatment programs need to address specific clinical correlates of suicidality, namely, low body weight, psychiatric comorbidities, history of childhood abuse, and NSSI. Trial registration: This study was not a clinical trial but a retrospective chart review based on routinely assessed clinical parameters. The study includes data from human participants; however: (1) no intervention and no prospective assignment to interventions were performed, and (2) no evaluation of intervention in participants was accomplished.