Real-time digital assessment of sleep-cognition dynamics in dementia care dyads

Sleep disturbance is common in persons living with dementia (PLWD) and may contribute to day-to-day variability in cognitive functioning, with implications for daily activities and caregiving. However, traditional cognitive assessments may not capture these short-term fluctuations. Ecological momentary cognitive testing (EMCT) enables repeated, real-world assessment of cognitive performance. This study examined day-to-day associations between sleep and cognitive performance in PLWD-caregiver dyads.

Baseline data from PLWD-caregiver dyads enrolled in a behavioral sleep trial were analyzed. PLWD completed smartphone-based EMCT tasks assessing processing speed and response inhibition three times daily for 7 days, along with the Mini-Mental State Examination (MMSE). Caregivers reported daily sleep outcomes for themselves and PLWD. Pearson correlations and mixed-effects models adjusted for PLWD age, gender, education, and phone type.

EMCT performance correlated significantly with MMSE scores (r = 0.51-0.58, all p < 0.01). Higher EMCT scores were associated with shorter time in bed and greater percentage of good sleep in both PLWD and caregivers. Mixed-effects models revealed significant sleep-by-session interactions: the association between PLWD sleep and processing speed was stronger in the evening than the morning, while longer caregiver time in bed was more strongly associated with PLWD cognitive performance in the morning.

EMCT captured within-person, time-varying associations between dyad sleep and cognitive performance in PLWD, highlighting its utility for identifying daily fluctuations relevant to functioning and caregiving. These findings support digital, dyadic approaches to monitoring sleep and cognitive function in dementia care.