Individual Sleep Problems Are Associated With an Accelerated Decline in Multiple Cognitive Functions in Older Adults

• Published in: Journal of sleep research Vol. 34; no. 5; pp. e70067 - n/a
• Main Authors: Sarsembayeva, Dina, Schreuder, Marieke J., Huisman, Martijn, Kok, Almar, Wagner, Maude, Capuano, Ana W., Hartman, Catharina A.
• Format: Journal Article
• Language: English
• Published: England 01.10.2025
• Subjects: Aged; Aged, 80 and over; Aging - physiology; Cognition - physiology; cognition disorders; cognitive aging; Cognitive Dysfunction; dementia; Female; Humans; insomnia; Longitudinal Studies; Male; Middle Aged; Neuropsychological Tests; nonlinear mixed models; older adults; prospective cohort study; sleep initiation and maintenance disorders; Sleep Wake Disorders - complications; prospective cohort study; older adults; sleep initiation and maintenance disorders; dementia; nonlinear mixed models; cognition disorders; cognitive aging; insomnia
• ISSN: 0962-1105; 1365-2869; 1365-2869
• DOI: 10.1111/jsr.70067

ABSTRACT Poor sleep is a known risk factor of cognitive disorders, but the role of individual sleep problems in age‐related cognitive changes remains unclear. This study used two complementary statistical models to estimate nonlinear trajectories of decline in four domains of cognitive functioning in the age period between 55 and 100 years depending on the severity of problems with falling asleep, night awakenings, and early morning awakenings, and short/long sleep duration. The sample included 5132 older adults (M = 67 years, 48% male) from the Longitudinal Aging Study Amsterdam (LASA), assessed 4–10 times every 2–3 years. Sleep problems were self‐reported, and cognitive functioning was measured with the 15‐Word test (reflecting episodic memory as immediate and delayed recall), Coding task (information processing speed) and Mini‐mental State Examination/MMSE (global cognition). Data were analysed using quadratic and piecewise changepoint mixed models. The piecewise models provided more precise and interpretable findings. Decline in information processing speed accelerated significantly earlier in participants with short sleep duration (regression coefficient (B) = −2.3[95% confidence interval (CI): −3.86; −0.81]; p < 0.01) and faster with more severe early morning awakenings (B = −0.07 [−0.1; −0.03]; p < 0.01). Decline in immediate recall accelerated earlier in those with short sleep (B = −2.8 [−4.44; −1.14]; p < 0.01) and severe problems with falling asleep (B = −1.22 [−2.06; −0.39]; p = 0.01). Decline in delayed recall was faster with long sleep (B = −0.06 [−0.08;‐0.03]; p < 0.01). Decline in global cognition accelerated faster in those with short/long sleep duration (B = −0.07 [−0.13; −0.01]/−0.10 [−0.18; −0.03]; p < 0.01) and severe night awakenings (B = −0.04 [−0.07;−0.02]; p < 0.01). To conclude, this study showed that some sleep problems can differentially predict earlier acceleration of deterioration in specific cognitive functions in older adults.