1066 A Pilot Randomized Controlled Trial of Digital Cognitive Behavioral Therapy for Insomnia in Older Adults with Suicidal Ideation

Abstract Introduction Insomnia is more common among older adults and may increase risk for suicide in this population. Treating insomnia may reduce suicide risk. Cognitive behavioral therapy for insomnia is the first-line recommended therapy for insomnia and is available in a digital format (dCBT-I). This study piloted dCBT-I in older adults with both insomnia and suicidal ideation. Main outcomes included feasibility, tolerability, and intervention effect. Methods Older adults with suicidal ideation recruited from outpatient clinic referrals and internet advertising were randomized to either dCBT-I treatment (TX, N=13) or to waitlist control with crossover treatment (WLC, N=11). Treatment was provided through Sleep Health Using the Internet (SHUTi). Outcomes included the Columbia Suicide Severity Rating Scale (CSSRS), the Insomnia Severity Index (ISI), and the Center for Epidemiologic Studies Depression Scale (CESD) and treatment effects were estimated using linear mixed effects regression models. Tolerability was assessed using a self-report questionnaire. Results Mean participant age was 69.1 years and did not differ between groups (p=0.26). Out of N=19 participants that completed baseline measures (11 TX, 8 WLC), N=12 (6 TX, 54.5%; 6 WLC, 75%) completed dCBT-I or monitoring. All N=6 WLC participants crossed over to dCBT-I and completed treatment, and N=1 TX participant dropped out after dCBT-I, leaving N=11 total dCBT-I completers. The most common side effects were sleepiness (N=8/11), anxiety (N=7/11), feeling tired (N=7/11), depression (N=6/11), and apathy (N=6/11). Two participants (8.3%) discontinued due to worsening mental health. No falls, suicide attempts, or preparatory acts for suicide were reported. Baseline insomnia severity averaged 14.7 (range 6-26), depression severity averaged 25.9 (range 7-49), and suicidal ideation averaged 2.18 (range 0-5); there were no group differences at baseline. Regression models showed significant decreases in suicidal ideation (-1.1, p=0.048), insomnia (-5.6, p=0.02), and depression (-9.4, p=0.043) across groups, and a trend toward a group difference in insomnia severity before crossover (-5.9, p=0.087). Conclusion Older adults with suicidal ideation tolerated dCBT-I without severe adverse effects, although dropout was high. Future work to reduce attrition and increase engagement with dCBT-I may increase access to insomnia treatment in this population. Support (if any) K24AG055602; R25MH112473