BACKGROUND: Quantification of suicidal risk in specific populations is important for the adoption of targeted prevention and harm reduction measures. Though there remains little systematic evidence, risk of suicide attempts for bulimia nervosa (BN) and binge-purging anorexia nervosa (AN-bp) appears higher than restrictive AN (AN-r); risk in binge eating disorder (BED) is still unclear. The aim of this meta-analysis was to compare proportions of suicide attempts in eating disorder (ED) subgroups. METHODS: A literature search using combinations of key-words for ED and suicide attempts was performed. Studies reporting proportions of suicide attempters in at least two ED groups, diagnosed according to DSM-IV or -5 and ICD-10 diagnostic criteria were considered. ED subgroups were analyzed in pairs using a binary random effect model for proportions. Publication bias, meta-regression, and sensitivity analyses were performed. RESULTS: In BN, attempted suicide was more frequent (21%) than in AN (12.5%), but the difference was statistically significant only when BN was compared with AN-r (9-10%). In BED, the proportion of suicide attempts was as high as in AN (10-12%). CONCLUSIONS: Though limited by heterogeneity across the studies in terms of methodology and aims, inability to control for relevant confounding variables, exclusion of ED not otherwise specified, this study supports suicide attempts as a major issue in EDs, especially in binge-purging subtypes, i.e. BN and AN-bp. Similar suicidal proportions were observed in AN and BED. The reasons for a greater proportion of attempted suicide in binge/purging subtypes need to be explored in future studies.
BACKGROUND: Quantification of suicidal risk in specific populations is important for the adoption of targeted prevention and harm reduction measures. Though there remains little systematic evidence, risk of suicide attempts for bulimia nervosa (BN) and binge-purging anorexia nervosa (AN-bp) appears higher than restrictive AN (AN-r); risk in binge eating disorder (BED) is still unclear. The aim of this meta-analysis was to compare proportions of suicide attempts in eating disorder (ED) subgroups. METHODS: A literature search using combinations of key-words for ED and suicide attempts was performed. Studies reporting proportions of suicide attempters in at least two ED groups, diagnosed according to DSM-IV or -5 and ICD-10 diagnostic criteria were considered. ED subgroups were analyzed in pairs using a binary random effect model for proportions. Publication bias, meta-regression, and sensitivity analyses were performed. RESULTS: In BN, attempted suicide was more frequent (21%) than in AN (12.5%), but the difference was statistically significant only when BN was compared with AN-r (9-10%). In BED, the proportion of suicide attempts was as high as in AN (10-12%). CONCLUSIONS: Though limited by heterogeneity across the studies in terms of methodology and aims, inability to control for relevant confounding variables, exclusion of ED not otherwise specified, this study supports suicide attempts as a major issue in EDs, especially in binge-purging subtypes, i.e. BN and AN-bp. Similar suicidal proportions were observed in AN and BED. The reasons for a greater proportion of attempted suicide in binge/purging subtypes need to be explored in future studies.