An Internet of Things Automated Cognitive Behavioral Therapy for Insomnia that Improves Cognition in Older Adults

Abstract and Summary We will address the public health issues of insomnia and Alzheimer’s Disease (AD), which costs the US healthcare system >$400 billion/year, with an additional ~$411 billion/year cost in lost productivity. Due to the strong role that sleep deprivation and insufficient slow wave sleep (SWS) play in conversion to mild cognitive impairment (MCI) and AD, the proposed business development research will address both public health issues. Previously, we showed that our patented technology could accurately measure sleep stages; and we successfully completed a randomized controlled trial (RCT) on 60–90-year old’s (n = 54) showing that we could augment conventional Cognitive Behavioral Therapy for Insomnia (CBTi). Our technology resulted in more remission to insomnia than conventional care when administered by a clinician. We are now seeking funding to conduct a validation study of a fully automated version of the technology that increases adherence to CBTi and is applicable to an older and/or non-adherent population. Called SleepSpace AI, the new system will improve adherence to CBTi using cues delivered via an unlocked smart phone, smart light bulbs, and sounds emitted from the phone. It will also provide personalized feedback based on user data to augment recommended sleep schedules, be easily accessible to deliver weekly interactive modules with a customized AI avatar, will be fully controllable via voice commands, and uniquely, it will focus on improving sleep quality and sleep amount. We will demonstrate the technology’s feasibility at improving sleep quality and cognition using a larger sample size and upgraded study design that includes sound and light sham conditions and active controls from the market leader in digital therapeutics. We will internally test this modified system and seek pre-submission feedback from the FDA (Specific Aim 1). We will then conduct an RCT on 180 participants aged 65+ years old who self-report moderate to severe insomnia based on an Insomnia Severity Index (ISI) score >= 15. Participants will commit to engage in an 11-week study that will include one-week baseline where sleep and cognition are evaluated, followed by random assignment to six-weeks of: 1) SleepSpace AI, 2) Competitor Digital Therapeutic (i.e. Somryst), and 3) Sleep Hygiene Control. (Specific Aim 2). Treatment will be followed by a one-week and four-week assessment of sleep and cognition. We hypothesize that SleepSpace AI will be superior to the leading competitor and Sleep Hygiene Control, as measured by the ISI (primary outcome). We also hypothesize that it will be superior to both conditions on sleep quality (SWS), sleep amount, survey assessments of cognition, and objective measures of cognition (secondary outcomes). A more effective, fully automated, digital CBTi that integrates with IoT devices will result in a safe, easy-to-use, effective, and affordable treatment for insomnia that could improve cognition in older adults. If proven effective in the planned study, this scalable technology can serve as a primary prevention strategy for MCI/ADRD, with the potential to be generalized to an MCI population that may typically struggle with following cumbersome CBTi directives.