An Infant Sleep Electroencephalographic Marker of Thalamocortical Connectivity Predicts Behavioral Outcome in Late Infancy

Abstract

Infancy represents a critical period during which thalamocortical brain connections develop and mature. Deviations in the maturation of thalamocortical connectivity are linked to neurodevelopmental disorders. There is a lack of early biomarkers to detect and localize neuromaturational deviations, which can be overcome with mapping through high-density electroencephalography (hdEEG) assessed in sleep. Specifically, slow waves and spindles in non-rapid eye movement (NREM) sleep are generated by the thalamocortical system, and their characteristics, slow wave slope and spindle density, are closely related to neuroplasticity and learning. Recent studies further suggest that information processing during sleep underlying sleep-dependent learning is promoted by the temporal coupling of slow waves and spindles, yet slow wave-spindle coupling remains unexplored in infancy. Thus, we evaluated three potential biomarkers: 1) slow wave slope, 2) spindle density, and 3) the temporal coupling of slow waves with spindles. We use hdEEG to first examine the occurrence and spatial distribution of these three EEG features in healthy infants and second to evaluate a predictive relationship with later behavioral outcomes. We report four key findings: First, infants’ EEG features appear locally: slow wave slope is maximal in occipital and frontal areas, whereas spindle density is most pronounced frontocentrally. Second, slow waves and spindles are temporally coupled in infancy, with maximal coupling strength in the occipital areas of the brain. Third, slow wave slope, spindle density, and slow wave-spindle coupling are not associated with concurrent behavioral status (6 months). Fourth, spindle density in central and frontocentral regions at age 6 months predicts later behavioral outcomes at 12 and 24 months. Neither slow wave slope nor slow wave-spindle coupling predict behavioral development. Our results propose spindle density as an early EEG biomarker for identifying thalamocortical maturation, which can potentially be used for early diagnosis of neurodevelopmental disorders in infants. These findings are complemented by our companion paper that demonstrates the linkage of spindle density to infant nighttime movement, framing the possible role of spindles in sensorimotor microcircuitry development. Together, our studies suggest that early sleep habits, thalamocortical maturation, and behavioral outcome are closely interwoven. A crucial next step will be to evaluate whether early therapeutic interventions may be effective to reverse deviations in identified individuals at risk.HighlightsSlow waves and spindles occur in a temporally coupled manner in infancySlow wave slope, spindle density, and slow wave-spindle coupling are not related to concurrent behavioral developmentSpindle density at 6 months predicts behavioral status at 12 and 24 monthsSlow wave slope and slow wave-spindle coupling are not predictive of behavioral development

The sleeping brain’s connectivity and family environment: characterizing sleep EEG coherence in an infant cohort

“…This may explain differences in the functional as well as the clinical relevance of these oscillations which have repeatedly been reported. For example, sleep spindle density has been identified as a predictor of behavioral outcomes in infancy, whereas slow wave activity showed no such effect 73 . Furthermore, increases in sleep EEG coherence in children with childhood onset schizophrenia as compared to healthy controls were more prominent in the delta than in the sigma band 74 , 75 .…”

Lack of association between behavioral development and simplified topographical markers of the sleep EEG in infancy

“…Time spent in the behavioral stage of sleep is a crucial part of a child's development, and the neurophysiological activity captured with the sleep EEG can provide valuable insights into the child's brain connectivity. Previous research has shown a correlation between EEG patterns and behavioral and cognitive outcomes in healthy children ( Buchmann et al, 2011 ; Ednick et al, 2009 ; Reynaud et al, 2018 ) and infants ( Jaramillo et al, 2021 ; Schoch et al, 2021 ). The non-rapid-eye-movement (NREM) sleep EEG, and therein specifically the oscillatory activity of slow wave activity (SWA, spectral power in the 1–4.5 Hz frequency range), undergoes drastic changes across childhood and the period of brain development, for example, with a shift in the predominant location of SWA from posterior to anterior regions of the scalp ( Kurth et al, 2010 ).…”

Sleep Spindles using Electroencephalography signals

“…It also provides information on the amplitude, PSD, and spectrogram of any detected sleep spindle. It is anticipated that Deep-spindle would find widespread use in many therapeutic settings, therefore reducing time and labor-intensive manual annotation (Jaramillo et al, 2023). Because variations in spindle characteristics (such as density, frequency, shape, and spatial distribution) help identify a number of sleep disorders, including sleep apnea, insomnia, narcolepsy, restless leg syndrome, and parasomnias, sleep spindles are significant from a clinical standpoint.…”