Development of a culturally tailored sleep intervention using artificial intelligence for filipino immigrants

The National Institute for Minority Health and Health Disparities awarded New York University $203,993 on September 1, 2026, under the Minority Health and Health Disparities Research program (CFDA 93.307) to develop and pilot test a generative AI-powered sleep chatbot tailored for Filipino Americans.

The intervention delivers personalized sleep advice grounded in cognitive behavioral therapy for insomnia via text messaging. The chatbot provides cost-effective, self-paced delivery to address the elevated prevalence of short sleep duration and poor sleep quality among Filipino American adults compared to other population groups. Preliminary research demonstrated high feasibility and acceptability, with participants reporting improvements in sleep duration and quality. The funded work will conduct a single-arm, pre-post pilot study to further test the intervention.

Work is performed in New York, New York, with a period of performance extending from September 1, 2026, through February 29, 2028. The assistance type is a Project Grant.

Abstract

Poor sleep is a prevalent public health problem that is associated with increased morbidity and healthcare utilization. Epidemiological data indicate that asian american adults are at increased risk of experiencing short sleep and poor sleep quality. This risk is particularly substantial among filipino americans, one of the largest subgroups of asian americans. Our research using a nationally representative cohort of U.S. Adults found that filipino american adults had the highest prevalence of short sleep duration among all population groups examined. They also reported greater difficulty falling asleep and staying asleep than other asian american subgroups. Despite the causal link between poor sleep and adverse physical and mental health, no intervention to date has addressed sleep issues among filipino americans. Existing non-pharmacological sleep interventions are effective in improving sleep outcomes, but they are primarily constrained by high costs, lack of accessibility. Emerging research has shown that artificial intelligence (AI)-driven lifestyle interventions can significantly improve sleep duration and quality. Integrating AI technologies into sleep interventions is highly promising because they are cost-effective, engaging, and offer an automated, self-paced delivery modality. To address these gaps, our team developed a generative AI-powered sleep chatbot that delivers personalized advice grounded in cognitive behavioral therapy for insomnia (CBT-I) via a text messaging APP. In our preliminary study, the chatbot demonstrated high feasibility and strong acceptability. Participants reported improvements in multiple sleep outcomes after chatbot use, including longer sleep duration and better sleep quality. Building on this work, we will conduct a single-arm, pre-post study to pilot test our generative AI-based sleep intervention among filipino americans with poor sleep. Specifically, we will examine the feasibility, acceptability, usability, and engagement of the intervention, as well as pre-post changes in sleep outcomes. Using a convenience sampling strategy, we will enroll 40 filipino americans aged 30 to 70 years who have short sleep duration and poor quality of sleep. The study will last for four weeks, including a 1-week baseline assessment, a 2-week intervention period, and a 1-week post-intervention assessment. During the 2-week sleep intervention, participants will engage with an AI chatbot through texting to receive personalized education and guidance for sleep promotion. A debriefing session will follow the intervention to gather each participant's feedback and identify areas for improvement. This proposed study represents a pioneering effort to harness generative AI to improve sleep health among asian americans. Findings from this study will lay the groundwork for further refinement and efficacy testing of our AI-based sleep intervention on a larger scale. More importantly, the AI-based sleep intervention holds potential for expansion to other population groups that experience poor sleep.