Arousals during sleep are associated with cortical amyloid-β burden and cognition in healthy older adults

Arousals during sleep have been studied in sleep pathologies such as sleep apnoea and periodic limb movement disorder, or in interventional experimental protocols triggering arousals through external stimulations (mainly noise). Those studies generally pointed towards a detrimental role of arousals. Yet, the causes and consequences of spontaneous arousals in healthy individuals remain largely unknown. Recently, a bidirectional detrimental link between sleep-wake dysfunction and the pathogenesis of Alzheimer’s disease is emerging in the literature, including the abnormal accumulation of amyloid-β (Aβ) and tau protein in the brain. Here, we investigated whether spontaneous arousals during sleep were associated with Aβ and tau burden, as well as with cognitive performance.
We recorded night-time sleep of 101 cognitively normal older individuals devoid of sleep disorders (aged 51-69; mean 59 ± 5; 68 females) under EEG. We assessed 18F-flutemetamol and 18F-THK 5351 regional brain uptake using Positron Emission Tomography to infer their level of Aβ and tau protein burden, respectively. Validated automatic arousal detection was computed on all EEG recordings. All participants completed an extensive cognitive battery of neuropsychological tasks to assess memory, attentional, and executive functioning while well-rested.
Generalised linear mixed models analyses revealed that arousals concomitant to a muscular tone increase (arousals EMG) are significantly negatively associated with Aβ burden in the cortex (F(1,95) = 11.62, p = .001), while correcting for age, sex and total sleep time, whereas no such relationship was observed for tau (p = .47). Furthermore, global cognitive performances were positively linked with arousals EMG throughout the night (F(1,95) = 6.38, p = .01), the observed relationship being mainly driven by performances in the attentional (p = .005) and executive domain (p = .04), after adjusting for age, sex, education and total sleep time.
These findings suggest arousals during sleep may form a heterogeneous category of events composed of arousals evoked by external or pathological event, that may be deleterious, and spontaneous arousals that may be associated with better brain structure or cognition. Future research should try to further characterise arousal events to disentangle what has been so far taring everyone with the same brush.