TY - JOUR
T1 - Correlates of suicide attempts among self-injurers
T2 - A meta-analysis
AU - Victor, Sarah E.
AU - Klonsky, E. David
PY - 2014/6
Y1 - 2014/6
N2 - Suicide attempts (SAs) are common among those who engage in non-suicidal self-injury (NSSI). It is therefore important to determine which suicide risk factors are most predictive of SA among those who self-injure. Toward this aim, we conducted a systematic review of studies examining predictors of SA history among self-injurers. A total of 52 empirical articles provided data comparing self-injurers with and without SA. From these studies we focused our meta-analysis on the 20 variables that were evaluated with respect to SA history in five or more different samples. The strongest correlate of SA history was suicidal ideation. After suicidal ideation, the strongest predictors of SA history were NSSI frequency, number of NSSI methods, and hopelessness. Additional, moderate predictors of SA history included Borderline Personality Disorder, impulsivity, Post-Traumatic Stress Disorder, the NSSI method of cutting, and depression. Demographic characteristics, such as gender, ethnicity, and age, were weakly associated with SA history. Notably, some oft-cited risk factors for SA displayed small or negligible associations with SA among self-injurers, including histories of sexual and physical abuse, anxiety, substance use, and eating disorders. Findings have implications for conceptual models of the NSSI-SA relationship and the evaluation of suicide risk among self-injuring populations.
AB - Suicide attempts (SAs) are common among those who engage in non-suicidal self-injury (NSSI). It is therefore important to determine which suicide risk factors are most predictive of SA among those who self-injure. Toward this aim, we conducted a systematic review of studies examining predictors of SA history among self-injurers. A total of 52 empirical articles provided data comparing self-injurers with and without SA. From these studies we focused our meta-analysis on the 20 variables that were evaluated with respect to SA history in five or more different samples. The strongest correlate of SA history was suicidal ideation. After suicidal ideation, the strongest predictors of SA history were NSSI frequency, number of NSSI methods, and hopelessness. Additional, moderate predictors of SA history included Borderline Personality Disorder, impulsivity, Post-Traumatic Stress Disorder, the NSSI method of cutting, and depression. Demographic characteristics, such as gender, ethnicity, and age, were weakly associated with SA history. Notably, some oft-cited risk factors for SA displayed small or negligible associations with SA among self-injurers, including histories of sexual and physical abuse, anxiety, substance use, and eating disorders. Findings have implications for conceptual models of the NSSI-SA relationship and the evaluation of suicide risk among self-injuring populations.
KW - Attempted suicide
KW - Meta-analysis
KW - Review
KW - Self-harm
KW - Self-injury
KW - Suicide
UR - http://www.scopus.com/inward/record.url?scp=84898647172&partnerID=8YFLogxK
U2 - 10.1016/j.cpr.2014.03.005
DO - 10.1016/j.cpr.2014.03.005
M3 - Review article
AN - SCOPUS:84898647172
SN - 0272-7358
VL - 34
SP - 282
EP - 297
JO - Clinical Psychology Review
JF - Clinical Psychology Review
IS - 4
ER -