Suicide and Self-Injury
Halla Hafermann, B.A.
Research Fellow
National Institute of Mental Health
Washington, District of Columbia, United States
Dede K. Greenstein, Ph.D.
Psychologist
National Institute of Mental Health
Bethesda, Maryland, United States
Laura Waldman, LCSW
Study Coordinator
National Institute of Mental Health
Bethesda, Maryland, United States
Elizabeth D. Ballard, Ph.D.
Staff Scientist
National Institute of Mental Health
Bethesda, Maryland, United States
Carlos Zarate, M.D.
Principal Investigator
National Institute of Mental Health
Bethesda, Maryland, United States
Even as 50,000 Americans died by suicide in 2021, researchers have yet to fully understand factors affecting suicide risk. The Three Step Theory by Klonsky and May posits that suicidal ideation (SI) occurs when hopelessness and psychological pain (PP) are present. However, to our knowledge, hopelessness and PP have not been evaluated relative to the recency of a suicide crisis (SC) and there is minimal information as to how the factors may interact in the context of an SC. We hypothesized that hopelessness, PP, and their interaction would be associated with the recency of an SC. Second, we hypothesized that hopelessness, PP, and their interaction would be associated with SI at baseline and at a 6 month follow up.
Participants were volunteers in a suicide risk study and grouped by recency to an SC: recent SC (SC-R) (within last two weeks) (N = 14), past SC (SC-P) (crisis in lifetime but not in last year) (N = 42), ideators (chronic SI) (N = 23) and clinical controls (mood disorder with no SI history) (N = 43). Follow-up evaluations were conducted at 6 months (N= 52). Hopelessness, PP, and SI levels were assessed by the Beck Hopelessness Scale, Mee-Bunney Psychological Pain Scale, and Scale for Suicidal Ideation (SSI), respectively. We used ANOVA and Tukey HSD corrected post-hoc tests to evaluate recency group differences in hopelessness and PP. For each pairwise combination of the SC-R, SC-P, and ideator groups, we used logistic regression to evaluate the relationship between group membership and the interaction of PP and hopelessness. We used Spearman correlations to examine SI relationships with hopelessness and PP at baseline and follow up across all groups and a regression to evaluate the relationship between SI and the interaction of PP and hopelessness.
We found recency group differences in hopelessness (F(2,75) = 10.93, p < 0.001) and PP (F(2,76) = 7.23, p = 0.001), with higher hopelessness (p < 0.001) and PP (p < 0.001) in the SC-R group than the lifetime group, and higher hopelessness (p = 0.015) in the ideator group than the SC-P group. We did not detect significant hopelessness-by-PP interaction effects for any of the three pairwise combinations.
At baseline, SI was significantly correlated with hopelessness (rho = 0.71, p < 0.001) and PP (rho = 0.43, p = 0.002). At follow up, the SI/hopelessness relationship remained significant (rho = 0.63, p < 0.001), but the SI /PP relationship weakened to a trend (rho = 0.27, p = 0.052). There was a significant hopelessness-by-PP interaction effect for SI at baseline (t = 5.07, p < 0.001) such that the relationship between hopelessness and SI increased as PP increased. There was no significant interaction at follow up.
Our findings indicate associations between crisis recency, hopelessness, and PP, supporting them as critical factors in an SC. The findings replicate previous work linking hopelessness and PP with SI at baseline and indicate their possible predictive ability of SI at 6 month follow up. Finally, we demonstrated a significant interaction between PP and hopelessness on SI. Due to the complex nature of suicide, it is unlikely that one factor will be the sole predictor of a SC. Instead, by understanding the interactions between multiple risk factors, we can hope to unravel the multifactorial indicators of a suicide crisis.