Conference Agenda

Introduction: Several studies have identified the association between sleep disturbances, particularly insomnia and experiencing nightmares, and suicide risk in adolescents (Fernandes et al., 2021). Mechanisms for the relationship between sleep disturbance and suicide risk are still being explored. A handful of studies have examined defeat and entrapment as possible mechanisms in this relationship. However, existing studies have focused on college student samples or lifetime suicide ideation as the outcome variable (Bradford et al., 2022; Russell et al, 2018). The current study examined defeat and entrapment as mediators in the relationships between insomnia and current SI, and nightmares and current SI, in a community sample of adolescents. It was expected that nightmares and insomnia symptoms would indirectly associate with current suicide ideation severity through defeat and entrapment.

Methods: Data were collected from 425 high school students (M_age = 15.64, SD=1.23; 50.8% female; 72.4% white). Adolescent participants were recruited from high schools in the south-central U.S; researchers visited schools and administered self-report questionnaires. Measures assessed defeat and entrapment, insomnia symptoms, nightmares, and current suicide ideation severity.

Results: Two serial multiple mediation models were run using PROCESS with insomnia or nightmares as the independent variable, defeat and then entrapment as the mediators, and SI severity as the dependent variable. Mediation models were significant and showed that both insomnia symptoms and nightmares were indirectly related to suicide ideation severity through defeat and entrapment. The full model for insomnia accounted for 43% of the variance, and the model with nightmares accounted for 42.5%.

Conclusions: These results provide additional evidence that current sleep disturbances are associated with current suicide ideation severity and provide additional support for defeat and entrapment as possible mechanisms in this relationship. Sleep disturbances in adolescents are a modifiable risk factor for suicide and should be studied further as targets for intervention.

Introduction: Death-oriented implicit associations with the self are a risk factor for suicidal ideation (SI) and reflect underlying cognitive vulnerabilities not captured by self-report. The death/suicide implicit association test (d/s-IAT) objectively measures death-self oriented implicit associations and evinces criterion validity for SI (Millner et al., 2018). Poor sleep is associated with worsening SI and may compound the effects of other risk factors (Liu et al., 2020; Russell et al., 2019). As such, lower total sleep time (TST) may amplify death-self oriented vulnerabilities, driving worsening SI over time. However, the compounding (i.e., average) and variable (i.e., change) effects of sleep problems on death-self orientation and SI is unknown.

Methods: This study measured the relationship between d/s-IAT scores and SI moderated by weekly average within-person stability and variability in TST in 132 college-students with a past month history of SI. Participants used actigraphy during a 28-day daily-diary ambulatory phase, completed self-report assessments (i.e., suicidal intent), and completed d/s-IAT each week.

Results: Results of multilevel mixed effects models indicated that average weekly TST moderated the relationship between weekly d/s-IAT and SI; the association was strongest at low (-1 SD) levels of TST (B=.55, SE=.13, p<.001), non-significant at high (+1 SD) levels of TST (B=-.05, SE=.13, p=.683). Variability of weekly TST moderated the relationship between weekly d/s-IAT and SI (B =-.004,SE =.001, p<.001); the association was strongest at low (-1 SD) levels of TST variability (B=.42, SE=.12, p < .001), non-significant at high (+1 SD) levels of TST variability (B=.03, SE=.13, p=.826).

Discussion: Stable lower TST amplifies the effect of death-oriented implicit associations captured by the d/s-IAT, driving worsening SI over time. However, lower (not higher) variability in TST compounded the effects of death-oriented implicit associations on SI suggesting short-term improvements in sleep (i.e., higher variability) may support SI stabilization.

Introduction: The middle of the night is a period characterized by increased risk for suicide, yet little is known about mechanisms underlying this effect. This presentation will describe the results from two independent investigations of the association among sleep and suicide in military personnel using real-time ecological momentary assessment (EMA). In both studies, daily/nightly-level associations among sleep and suicide risk are explicated, as well as mechanisms of action for this effect.

Method: Sample 1 includes 34 Veterans in a residential treatment unit for PTSD who completed daily EMA of sleep, cognitive control, and suicide risk for 28 days. Sample 2 includes 90 active-duty U.S. military personnel and Veterans following an acute suicidal crisis who completed daily EMA of sleep, PTSD, depression, suicidal beliefs and suicide risk for 28 days. Mixed effects models evaluate associations among these variables.

Results: In both samples, the middle of the night (between 0000 and 0400) was characterized by significant elevations in suicidal urges on the Suicide-Visual Analogue Scale (S-VAS), with a highest standardized incidence ration of 14 .43 (at 0200) in Sample 1 and 14.646 (also at 0200) in sample 2. In sample 1, both nocturnal wakefulness and sleep quality predicted worsened next-day S-VAS (p < .05), which was partially mediated by self-reported cognitive control accounting for 7% of the variance in the effect. In sample 2, sleep quality and nightmares were both associated with worsened next-day S-VAS, which was partially mediated by PTSD symptoms and depression severity.

Conclusions: Self-reported sleep is an important predictor of next-day intensity of suicidal urges, which is mediated, in part, by deficits in cognitive control, PTSD symptoms, and depression symptoms. These findings highlight opportunities to potentially design and deliver real-time sleep interventions to military personnel in an effort to circumvent suicide risk escalation that night and the next day.

Introduction: Multiple randomized clinical trials (RCTs) demonstrate that brief cognitive behavioral therapy (BCBT) reduces suicide attempts and suicidal ideation. One proposed pathway involves the weakening of interdependent relationships among cognitive-affective states.

Methods: Using data from a recent RCT, we compared BCBT, which includes some insomnia-focused procedures, and present-centered therapy (PCT), a problem-solving focused therapy that does not address insomnia, on session-by-session trajectories of insomnia severity and suicidal ideation assessed with the Insomnia Severity Index (ISI) and the first 5 items of the Scale for Suicide Ideation (SSI-5).

Results: Mixed effects modeling showed that changes in ISI and SSI-5 scores significantly differed between treatments. In BCBT, insomnia and suicidal ideation improved rapidly within the first 3 weeks and then stabilized thereafter. In PCT, insomnia initially worsened before returning to baseline, and suicidal ideation showed little change. Between-group differences in ISI and SSI-5 scores were observed by week 2: ISI, F(1,493)=4.6, p=.033, and SSI-5, F(1,490)=10.4, p=.001. Dynamical systems analyses indicated ISI and SSI-5 scores were coupled in PCT but not BCBT, indicating insomnia and suicidal ideation were interconnected over time in PCT but changes in one were unrelated to changes in the other in BCBT. Eigenvalue decomposition analysis revealed negative determinants in both groups (PCT=-0.046, BCBT=-0.011), indicating patients in both treatments showed small, diminishing adjustments when returning to equilibrium, but patients in BCBT returned to equilibrium nearly twice as fast, suggesting greater resilience to disruptions.

Conclusions: Overall, these findings suggest the de-coupling of insomnia from suicidal ideation is a possible mechanism through which BCBT reduces suicide attempt risk.