The influence of peripheral glucose on sleep brain oscillations

Previous research indicates a coupling of sleep-associated brain activity to peripheral glucose levels, such that ripple activity, a sign of increased hippocampal memory processing, is followed within 10 min by a dip in peripheral glucose levels. Whether there is respective negative feedback to the brain of peripheral glucose levels, or associated metabolic signaling, during sleep is not clear.

We tested the effects of systemic glucose administration as well as the effects of fasting (for 6 h vs. ad libitum feeding) on sleep and oscillatory signatures of memory processing during sleep, i.e. hippocampal ripples, cortical spindles and slow oscillations (SOs), using electroencephalography (EEG) from skull and local field potential recordings (LFP) from CA1 in male rats. Continuous monitoring of interstitial glucose concentrations in a subset of animals allowed the assessment of the temporal relationship between glucose fluctuations and oscillatory events.

Hippocampal ripples and sleep spindles were accompanied by a transient decrease in peripheral glucose concentration. Neither glucose injection nor prior fasting influenced the macro-architecture of sleep during the subsequent 6 hours. Glucose injection (vs. water) increased spindle density. Prior fasting, compared with ad libitum feeding, increased numbers of ripples, spindles, SOs and of co-occurring SO-spindle events. Fasting also influenced SO-spindle phase-amplitude coupling, such that spindles occurred later during the SO upstate in fasted animals compared to those with ad libitum food access.

Our findings indicate a relatively tight-paced glucose-to-brain feedback loop during sleep that may affect memory processing in thalamocortical and hippocampal networks.