Impaired memory consolidation in children with obstructive sleep disordered breathing.

Kiran Maski,Hannah Holbrook,Eliot S Katz,Kush Kapur,Robert Stickgold,Erin Steinhart,Andrea Romigi

doi:10.1371/journal.pone.0186915

Abstract

Memory consolidation is stabilized and even enhanced by sleep (and particularly by 12–15 Hz sleep spindles in NREM stage 2 sleep) in healthy children but it is unclear what happens to these processes when sleep is disturbed by obstructive sleep disordered breathing. This cross-sectional study investigates differences in declarative memory consolidation among children with primary snoring (PS) and obstructive sleep apnea (OSA) compared to controls. We further investigate whether memory consolidation group differences are associated with NREM stage 2 (N2) sigma (12–15 Hz) or NREM slow oscillation (0.5–1 Hz) spectral power bands. In this study, we trained and tested participants on a spatial declarative memory task with cued recall. Retest occurred after a period of daytime wake (Wake) or a night of sleep (Sleep) with in-lab polysomnography. 36 participants ages 5–9 years completed the protocol: 14 with OSA as defined by respiratory disturbance index (RDI) > 1/hour, 12 with primary snoring (PS) and 10 controls. OSA participants had poorer overall memory consolidation than controls across Wake and Sleep conditions [OSA: mean = -18.7% (5.8), controls: mean = 1.9% (7.2), t = -2.20, P = 0.04]. In contrast, PS participants and controls had comparable memory consolidation across conditions (t = 0.41; P = 0.38). We did not detect a main effect for condition (Sleep, Wake) or group x condition interaction on memory consolidation. OSA participants had lower N2 sigma power than PS (P = 0.03) and controls (P = 0.004) and N2 sigma power inversely correlated with percentage of time snoring on the study night (r = -0.33, P<0.05). Across all participants, N2 sigma power modestly correlated with memory consolidation in both Sleep (r = 0.37, P = 0.03) and Wake conditions (r = 0.44, P = 0.009). Further observed variable path analysis showed that N2 sigma power mediated the relationship between group and mean memory consolidation across Sleep and Wake states [Bindirect = 6.76(3.5), z = 2.03, P = 0.04]. NREM slow oscillation power did not correlate with memory consolidation. All results retained significance after controlling for age and BMI. In sum, participants with mild OSA had impaired memory consolidation and results were mediated by N2 sigma power. These results suggest that N2 sigma power could serve as biomarker of risk for cognitive dysfunction in children with sleep disordered breathing.

Highlights

Obstructive sleep disordered breathing (SDB) refers to partial or complete upper airway collapse that can exist along a continuum including snoring, upper airway resistance syndrome, and obstructive sleep apnea (OSA)
Given our findings that participants with OSA had reduced memory consolidation across both Wake and Sleep conditions compared to controls and had reduced NREM stage 2 (N2) sigma power compared to both control and primary snoring (PS) groups, we further explored whether group differences existed between N2 sigma and memory consolidation across the Wake condition
We found that OSA participants had less N2 sigma power compared to other groups, N2 sigma power correlates with amount of snoring on the study night, and memory consolidation is predicted by N2 sigma power

Summary

Introduction

Obstructive sleep disordered breathing (SDB) refers to partial or complete upper airway collapse that can exist along a continuum including snoring, upper airway resistance syndrome, and obstructive sleep apnea (OSA). The Respiratory Disturbance Index (RDI) is a more sensitive index of sleep disordered breathing that includes respiratory effort related arousals (any reduction in nasal airflow resulting in a 3-second arousal from sleep) in addition to obstructive apneas and hypopneas. While the ICSD specifies that OSA can be defined in adults using RDI [2], similar specifications are not offered for pediatric OSA. This is problematic, as SDB with AHI < 1 can still impact children’s academic performance and behavior [3,4,5]. In order to define clinically relevant disease, more data are needed to link specified levels of pediatric SDB to adverse outcomes

Methods

Results

Discussion

Conclusion