Decreased Delta/Beta ratio index as a state-independent signature of sleep state misperception in insomnia disorder

Introduction: Sleep State Misperception (SSM) is the tendency ofInsomnia Disorder (ID) patients to overestimate Sleep Latency (SL) andunderestimate Total Sleep Time (TST). Literature exploring topographicalcomponents in ID with SSM is scarce and does not allow us to fullyunderstand the potential mechanisms underlying SSM. This studyaimed to evaluate the existence of sleep EEG topography alterations inID patients associated with SSM compared to Healthy Controls (HC),focusing on the Sleep Onset (SO) and the whole night.Method: Twenty ID patients (mean age: 43.5 ± 12.7; 7M/13F) and18 HCs (mean age: 41.6 ± 11.9; 8M/10F) underwent a night of polysomnographyand completed sleep diaries the following morning uponawakening. Two SSM indices, referring to the misperception of SL(SLm) and the TST (TSTm), were calculated by comparing objectiveand subjective sleep indices extracted by polysomnography and sleepdiary. According to these indices, the entire sample was split into4 sub-groups: ID + SLm and HC-SLm; ID + TSTm and HC-TSTm.Results: Considering the SO, the two-way mixed-design ANOVAshowed a significant main effect of Groups pointing to an increased betaactivity involving anterior and temporoparietal cortical derivations(p ≤ 0.023) and to decreased delta/beta ratio in the whole scalp topography(p ≤ 0.040). Moreover, we found a significant (p ≤ 0.024) interactioneffect for the sigma and beta bands. Post Hoc tests showed highersigma and beta power in anterior and temporo-parietal sites during theSO period in IDs + SLm compared to HC–SLm. Considering the wholenight, the unpaired t-test revealed in IDs + TSTm significantly lowerdelta power during NREM (p ≤ 0.035) and lower delta/beta ratio indexduring NREM and REM sleep (p ≤ 0.034) compared to HCs–TSTm.Finally, we found diffuse significant negative correlations between SSMindices and the delta/beta ratio during the SO (r ≥  0.589; p ≤ 0.011),NREM (r ≥  0.424; p ≤ 0.042), and REMsleep (r ≥  0.372; p ≤ 0.042).Conclusion: The main finding of the present study suggests that higherSL overestimation and TST underestimation are both related to diffusecortical hyperarousal, interpreted as a state-independent electrophysiologicalcorrelate of the SSM, both during the SO and the whole night.