In a recent study published in JAMA Neurology, researchers investigated whether slow-wave sleep (SWS) proportions reduced as individuals aged and whether intra-individual decreases were linked to dementia risk.
Inadequate sleep might be a risk factor for dementia. Sleep deprivation, in particular, stimulates the amyloid buildup and tau protein secretion and distribution, aggregating in Alzheimer's disease (AD) in animal models. Sleep aids in removing metabolic toxic waste from brain cells (glymphatic clearance).
However, the function of slow-wave-type sleep in dementia development is unclear, necessitating more studies to understand better dementia pathogenesis and aid in developing treatment therapies.
About the study
In the present prospective cohort study, researchers investigated the association between SWS loss, aging, and dementia risk. They also investigated whether hippocampal volumes indicating early neurodegeneration and AD genetic risk [i.e., apolipoprotein E4 (APOE ε4) allele] were linked to SWS loss.
The study included 346 Framingham Heart Study (FHS) participants aged ≥60 years who completed two polysomnography (PSG) overnight assessments between 1995 and 1998 and between 1998 and 2001. None of the participants had dementia or neurological disorders such as multiple sclerosis during the second polysomnography. The data were analyzed between January 2020 and August 2023.
The study exposure was alterations in SWS proportions determined based on overnight sleep assessments, and the study outcome was new-onset any-cause dementia risk evaluated for ≤17 years following the second polysomnography. Sleep staging sores were determined in 0.5-minute epochs using the Rechtshaffen and Kales method. The Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV), was used to diagnose dementia.
Hippocampal volumes were determined as a fraction of intracranial volume by magnetic resonance imaging (MRI) of the brain before the second polysomnography. To assess AD risk gene status, 23 AD-linked single-nucleotide polymorphisms (except APOE) were combined to develop a polygenic risk score.
Cox proportional regression modeling was performed to determine the hazard ratios (HRs), adjusting for age, gender, Omni and Offspring cohorts, smoking status, at least one apolipoprotein E4 allele positivity, and the usage of medications such as anxiolytics and antidepressants to improve sleep.
Moderation and sensitivity analyses were performed by censoring the initial two follow-up years, adjusting for the sleep apnea-hypopnea index (AHI) and oxygen desaturation during the initial PSG, and assessing physical activity levels using the Physical Activity Index (PAI).
In addition, data were adjusted for the vascular burden based on the Framingham Stroke Risk Profiles, Epworth Sleepiness Scale (FSS) scores, and the total sleep duration at the initial PSG and between the two PSGs. Further, the researchers explored interactions by gender and the percentage of rapid eye movement (REM) sleep.
The mean participant age was 69 years; 52% (n=179) were women, and 84% (n=291) were white. The mean duration between brain MR imaging and the second polysomnography was two years, and the mean SWS proportion at study initiation was 18%. Aging was linked to SWS reductions in the overnight sleep evaluations (mean annual reduction of 0.6 units).
Over 17 follow-up years, 52 new-onset dementia cases were reported, of which 44 were AD-related dementia. A percent annual reduction in SWS was linked to 27% and 32% elevations in all-cause and AD-related dementia risks, respectively. SWS loss accelerated with aging in AD genetic risk presence but was unrelated to hippocampal volumes close to the initial PSG.
The rate of SWS loss accelerated nominally at ≥60 years of age, peaking at 75 to 80 years of age, followed by slowing. In contrast, REM sleep proportion and total sleep duration remained stable. Waking up after sleep onset and apnea-hypopnea index scores increased, whereas the efficiency of sleep maintenance decreased. Individuals experiencing a reduction in SWS proportion showed a higher likelihood of suffering from cardiovascular diseases, being an APOE ε4 carrier, and taking drugs affecting sleep.
Sensitivity analyses yielded similar results, indicating the robustness of the primary findings. The mean reduction in SWS proportion was two-fold higher in new-onset dementia cases (−1.0 annually) versus non-cases (−0.5 annually). Gender and REM sleep did not modify the link between SWS loss and new-onset dementia.
The causal mediation assessment showed that SWS percentage changes mediated 17.0% of the effects of apolipoprotein E4 positivity (vs. ε3) on new-onset dementia risk. However, the indirect effects were non-significant (HR, 1.1).
Overall, the study findings showed that SWS proportion decreased with age and AD genetic risk, with higher decreases associated with new-onset dementia risk, indicating that SWS loss may be a modifiable risk factor for dementia. The APOE ε4 allele was independently linked to a greater proportion of SWS loss.
However, regardless of apolipoprotein ε4 status, increased SWS loss was linked to an increased risk of dementia. Future studies could investigate the association between SWS loss and tau or amyloid β accumulation and whether SWS enhancement could prevent neurodegeneration and cognitive decline among high-risk individuals.