Electroencephalographic sleep macrostructure and sleep spindles in early infancy.
Level 3 - non-randomized controlled study
Prospective observational cohort study establishing normative values.
PubMed 34755881 · doi:10.1093/sleep/zsab262
What was done
Healthy term infants recruited at birth underwent daytime sleep electroencephalography (EEG) at 4-5 months of age. Researchers performed sleep staging to analyze five macrostructural features and annotated sleep spindles to extract seven quantitative parameters. Metrics were analyzed across sex, recording time (morning vs. afternoon), age, and progression from the first to the second sleep cycle.
What was found
In 91 infants (41% female), median (IQR) sleep macrostructure values were: total sleep duration 49.0 (37.8-72.0) min (n = 77), first sleep cycle duration 42.8 (37.0-51.4) min, REM percentage 17.4% (9.5-27.7%) (n = 68), and latency to REM 36.0 (30.5-41.1) min (n = 66). First-cycle spindle characteristics included a median density of 6.6 (5.7-8.0) spindles/min (n = 77), frequency of 13.0 (12.8-13.3) Hz, duration of 2.9 (2.6-3.6) s, spectral power of 7.8 (4.7-11.4) µV², brain symmetry index of 0.20 (0.16-0.29), and synchrony of 59.5% (53.2-63.8%) (n = 91). After controlling for gestational and postnatal age and nap timing, males had 24.5% lower spindle spectral power (p = .032) and 24.2% higher brain symmetry index (p = .011) than females. Spindle spectral power decreased in the second cycle compared to the first (p < .001). No other significant associations were found with recording time, age, or sex.
Why it matters
This study establishes quantitative baseline reference ranges for EEG sleep architecture and spindle characteristics during early infancy. These normative values provide a reference point for future studies evaluating neurodevelopmental delays and sleep abnormalities in infants.
Limits
The study is restricted to daytime nap recordings at a single developmental time point (4-5 months), which may not generalize to overnight sleep architecture or other ages. The sample size is modest (n = 91) with data attrition across sub-analyses (complete macrostructure data available in 66-77 infants). Long-term clinical or neurodevelopmental outcomes were not evaluated.