Sleep Spindle Features and Neurobehavioral Performance in Healthy School-Aged Children.

PURPOSE: In adults, central fast-frequency sleep spindles are involved in learning and memory functions. The density of local spindles is higher than global spindles, emphasizing the importance of local plastic neural processes. In children, findings on the association of spindles with cognition are more variable. Hence, we aim to study whether the local spindles are also important for neurobehavioral performance in children. METHODS: We studied the correlations between local (occurring in only one channel: Fp1, Fp2, C3, or C4), bilateral, and diffuse (occurring in all four channels) spindles and neurobehavioral performance in 17 healthy children (median age 9.6 years). RESULTS: Local spindles were not as frequent as bilateral spindles (P-values < 0.05). Central spindle types had significant correlations with sensorimotor and language functions (e.g., the density of bilateral central spindles correlated positively with the Object Assembly in NEPSY, r = 0.490). Interestingly, frontopolar spindles correlated with behavior (e.g., the more bilateral the frontopolar spindles, the less hyperactive the children, r = -0.618). CONCLUSIONS: In children, the local spindles, but also more widespread central spindles, seem to be involved in the cognitive processes. Based on our findings, it is important that ageadjusted frequency limits are used in studies evaluating the frequencies of spindles in children.

Local Differences in Computational Sleep Depth Parameters in Healthy School-aged Children.

OBJECTIVE: Slow wave sleep in children reflects several processes, such as sleep pressure, synaptic density, and cortical maturation. Deep sleep in children is abundant and our aim was to discover whether examining electroencephalography (EEG) mean frequency would help separate these processes. METHODS: Sleep EEG of 28 generally healthy 7- to 11-year-old children (14 first graders, 14 third graders, 14 girls, 14 boys) was analyzed. Median non-rapid eye movement (NREM) sleep EEG frequency (median sleep depth, in Hz) and the amount of computational deep sleep using the thresholds of 2 Hz and 4 Hz (DS2% and DS4%, respectively) were calculated from the frontopolar, central, and occipital EEG derivations. RESULTS: Median NREM sleep frequency was lower in the left frontopolar area than more posteriorly in the whole study group, in the third graders and in the girls. In the left hemisphere, the amount of DS4% was higher frontopolarly than occipitally in the third graders and in the girls. The amount of DS2% was higher frontopolarly than centrally in all groups except in the first graders. In the whole study group, DS4% declined smoothly across the NREM episodes, whereas DS2% centered in the first NREM sleep episode. DISCUSSION: The median NREM sleep EEG frequency results might denote earlier frontal maturation in girls than in boys. Interestingly, we found frontopolar predominance in slow mean EEG frequency in both hemispheres, even if frontal slow wave activity is found to enhance until adolescence. As with infants, it seems that slower sleep EEG frequencies do not reflect sleep pressure as well as <4 Hz activity in school-aged children either. CONCLUSION: Our analysis method suggests that in addition to slow wave activity, EEG frequency analysis might be useful in differentiating between the different sleep related processes in children.

Sleep apnea reduces the amount of computational deep sleep in the right frontopolar area in school-aged children.

OBJECTIVE: Obstructive sleep apnea (OSA) causes different symptoms in children, even though polysomnographic parameters that assess sleep quality may remain normal. Our spectral analysis of NREM sleep revealed local deep sleep reductions in adult OSA patients. We hypothesize that our method would also reveal local changes in pediatric OSA patients. METHODS: Polysomnographies were part of a larger study evaluating snoring in school-aged children. All right-handed children with OSA with matched peers (n=10+10) were included. The median sleep depth (in Hz) and the amount of deep sleep <4Hz (DS%) were extracted for the whole NREM sleep time and for the first four NREM sleep episodes from frontopolar, central and occipital EEG-channels. RESULTS: The main findings were that NREM sleep was lighter and DS% decreased in the right frontopolar area (p-values 0.034 and 0.019) in the OSA group when compared with the control group. CONCLUSION: Local sleep quality changes might provide new insights to evaluate the effects of pediatric OSA as our method revealed a local computational deep sleep decrease in the right frontopolar area in the OSA group. SIGNIFICANCE: The presented findings might implicate delayed local cortical development in children's OSA, which may account for the cognitive problems found in pediatric OSA.