Insomnia drives changes in suicide ideation: A latent difference score model of community adults over a brief interval

Abstract

Insomnia is robustly associated with suicidal behavior (Bernert, Kim, Iwata, & Perlis, 2015), but limitations in existing studies hinder nuanced understanding of this relationship. The current study aimed to address limitations by utilizing a longitudinal design and advanced statistical modeling. Participants who endorsed lifetime experience of suicidal behavior were recruited through Amazon’s Mechanical Turk (N = 589) and completed self-report online surveys at six time points over a 15-day period. Latent difference score modeling was utilized to investigate whether levels and/or changes in insomnia symptoms drive subsequent changes in suicide ideation, or vice versa. Results revealed that previous level of insomnia was predictive of positive changes in suicide ideation (e.g., level of insomnia at Wave 2 predicted lagged increases in suicide ideation at Wave 3). This relationship was not bidirectional (i.e., suicide ideation exerted no effects on insomnia). Additionally, only previous level, and not previous changes, in insomnia were predictive of changes in suicide ideation. Our results help clarify the nature of the relationship between insomnia and suicide ideation as one that is unidirectional, thereby offering evidence of insomnia as a variable risk factor for suicide ideation. These findings yield clinical implications, including the importance of screening for insomnia symptoms, and provide support for exploring the potential effectiveness of insomnia treatments to target suicide ideation. Moreover, our study design and methodology establish a foundation for more rigorous and nuanced investigations of imminent suicide risk in future studies, which ultimately can promote better clinical practice in the reduction in suicidal behavior.