0630 For Whom the Bell Tolls: Empirical Clustering of Insomnia and Suicidal Ideation

Introduction Insomnia is associated with risk for suicidal thoughts and behaviors. It is unclear if insomnia drives risk equivalently across all individuals, or whether there are subgroups of high-risk individuals for whom insomnia is more relevant. This exploratory archival analysis examined the relationship between insomnia and current suicidal ideation among subgroups of individuals with lifetime suicidal ideation. Methods Data were acquired from the Military Suicide Research Consortium on 2,617 individuals without missingness on measures of suicidal thoughts and behaviors, anxiety, insomnia, thwarted belongingness, traumatic brain injury, risky alcohol use, and symptoms of post-traumatic stress. Medoid (k-means) clustering algorithms identified subgroups. Robust Poisson models tested interactions between insomnia and cluster identity in association with current suicidal ideation. Models were unadjusted, adjusted for demographic factors, and adjusted for clinical characteristics. Results Clustering identified two subgroups: Cluster 1 (N=1,141) and Cluster 2 (N=1,494). Cluster 1 had fewer individuals with current suicidal ideation (29.2% vs. 66.6%) and in general had less severe symptoms of anxiety, PTSD, insomnia, traumatic brain injury, and suicidal thoughts and behaviors. Across clusters, insomnia severity was associated with risk for current suicidal ideation (PRR: 1.09 [1.03-1.14]). Although individuals in Cluster 1 were less likely to have current suicidal ideation (PRR 0.51 [0.44-0.58]), greater severity of insomnia was associated with greater risk (PRR: 1.11 [1.04-1.18]). By contrast, individuals in Cluster 2 had a greater risk of current suicidal ideation (PRR: 1.68 [1.51-1.88]) and greater severity of insomnia was associated with a diminished risk of current ideation (PRR 0.92 [0.88-0.97]). Conclusion In this exploratory study, individuals with less severe psychopathology were more likely to report current suicidal ideation with worsening insomnia, while those with severe psychopathology were less likely to report suicidal ideation with worsening insomnia. Insomnia may disproportionately increase suicide risk among lower risk individuals. Support (if any) This work was in part supported by the Military Suicide Research Consortium (MSRC), an effort supported by the Office of the Assistant Secretary of Defense for Health Affairs under Award No. (W81XWH-16-2-0003). Opinions, interpretations, conclusions, and recommendations are those of the author and are not endorsed by the MSRC or the Department of Defense.