Exploring nightly variability and clinical influences on sleep measures: insights from a digital brain health platform.

BACKGROUND

Digital technology offers a convenient way to continuously monitor sleep and assess night-to-night variability, particularly in aging populations where traditional self-reported sleep assessments may be limited.

AIMS: This study aimed to investigate nightly variability in sleep measures obtained via a ring oximeter sensor in older adults and to explore the influence of demographic and cognitive factors on the stability of these metrics.

METHODS

The study included 62 participants (mean age 74, 67.7 % women, 90.3 % White) from the Boston University Alzheimer's Disease Research Center (BU ADRC) cohort. Each participant wore a SleepImage Ring for at least three consecutive nights. Thirty-four continuous sleep measures, such as mean SpO2 and apnea-hypopnea index within unstable sleep, were analyzed. Night-to-night variability was assessed using intraclass correlation coefficients (ICC) based on a two-way random-effects model. Subgroup analyses examined variability by sex, age, and cognitive status. Group-level changes were assessed using one-way repeated measures ANOVA.

RESULTS

Seven sleep measures demonstrated high stability across nights (ICC: 0.70-0.88), with average heart rate being the most stable, followed by mean SpO2 and apnea-hypopnea indices. Sleep latency exhibited the highest variability. Stability improved between the second and third nights compared to the first and second nights. Women and participants under 75 years old showed greater stability in several metrics, while cognitively intact individuals exhibited more consistent breathing-related measures.

CONCLUSION

At least three nights of monitoring are required for reliable estimates of key sleep metrics. Expanding studies with larger samples and extended monitoring periods could further elucidate sleep variability as a potential non-invasive marker for general health.