Dream recall in cognitively healthy older adults: associations with blood p ‐tau217 levels, APOE ε4 carriage, and cognitive decline

Among cognitively normal older adults, not remembering dreams is linked to higher blood levels of p-tau217, carrying the APOE ε4 gene variant, and faster cognitive decline over 10 years. In a study of 1,049 people (average age 74.7), 31% did not recall dreams. Those with elevated p-tau217 were about half as likely to remember dreams, and APOE ε4 carriers had 39% lower odds of dream recall. Non-recallers showed a steeper annual decline in cognitive composite scores compared to recallers. Dream recall status was not tied to baseline cognitive performance. The findings suggest that absent dream recall in later life may signal early neurodegeneration in the default mode network, potentially preceding dementia.

Study at a glance

Abstract

Abstract Background Evidence suggests that dreaming is subserved by the default mode network (DMN) (Domhoff, 2022. The Neurocognitive Theory of Dreaming ). β‐amyloid plaques and phosphorylated tau ( p ‐tau) protein tangles accumulate and contribute to abnormal functionality in the DMN long before clinical Alzheimer's disease, and similarly early abnormal DMN functionality is shown by apolipoprotein E ( APOE ) gene ε4 allele carriers. Given this, we reasoned that dream recall in cognitively normal older adults may be associated with blood p ‐tau levels, APOE ε4 carriage, and future cognitive decline. Method Data were for 1049 cognitively unimpaired individuals (mean age=74.7 years; 64% women) in the Spanish Vallecas Project. Dream recall was ascertained by asking “Do you remember your dreams?” (yes/no). Blood p ‐tau217 levels >0.247 pg/mL were considered high, and ≥1 ε4 alleles indicated APOE ε4 carriage. Cognition was assessed with a modified Preclinical Alzheimer Cognitive Composite (PACCm) annually over 10 years, with z‐scores <‐1 considered relatively low baseline cognition. Our analyses used general linear models, as well as a linear mixed‐effect model to compare the longitudinal cognitive trajectories of dream recallers and non‐recallers. Analyses were controlled for factors including sleep characteristics, medications, depression, and memory test scores. Result Thirty‐one percent of participants did not remember dreams. Higher p ‐tau217 levels and APOE ε4 carriage were both associated with a lower likelihood of remembering dreams (OR=0.52, 95%CI=0.35‐0.79, p = .002 and OR=0.61, 95%CI=0.43‐0.86, p = .006, respectively), independently of memory test scores. Further, individuals who did not remember dreams at baseline showed faster decline in PACCm scores (β non‐DR =‐0.03) over 10 years than dream recallers (β DR =‐0.02; β Age*Dream Recall =0.011, SE=0.005, p = .036), despite dream recall status not being associated with relatively low cognition at baseline (OR=0.94, 95%CI=0‐60‐1.48, p = .788). Conclusion Dream recall status was associated with blood p ‐tau217 levels, APOE ε4 carriage, and future cognitive decline among older individuals cognitively healthy at baseline. Not remembering dreams in later life may be an early indicator of neurodegeneration in the DMN. This would help explain an unexpected and overlooked finding more than 30 years ago that lower dream recall frequency predicted incident dementia (Persson & Skoog, J Geriatr Psychiatry Neurol 1992;5:172‐178), a decade before the DMN was discovered.