First night effect in children and adolescents undergoing polysomnography for sleep-disordered breathing

To establish whether there is a first night effect (FNE) in children and adolescents with suspected obstructive sleep apnoea undergoing polysomnography (PSG) and whether this affects sleep and breathing, furthermore, to determine the extent to which age may influence the sleep and cardiorespiratory parameters.

One hundred and thirty-one children and adolescents (age classes-A: 2-6 years n=37; B: 7-12 years n=60; C: 13-17 years n=34) underwent PSG on 2 consecutive nights (I and II) under identical conditions for suspected sleep-related respiratory disorders. One hundred and five patients including 3 patients with obstructive sleep apnoea syndrome (OSAS) treated by adenotonsillectomy and 18 OSAS patients receiving nCPAP-therapy had no PSG-abnormalities (Group 1-A: n=28; B: n=53; C: n=24). A further 26 patients (Group 2) had clinically and polysomnographically confirmed untreated OSAS (A: n=9; B: n=12; C: n=5).

There were no statistically significant differences between children with no PSG-abnormalities (Group 1) and those with OSAS (Group 2) in terms of sleep parameters (arousal indices excluded), oxygen saturation (SaO(2)) and heart rate (HR), and these parameters have, therefore, been pooled for the entire group (n=131) in the 3 age classes A, B and C. In the second and third age classes, sleep efficiency on the first night was reduced. In all age classes, there was significantly more wakefulness during the first night. In the second and third age ranges, the proportion of NREM 1 in the first night was significantly higher, with a correspondingly reduced proportion of NREM 4 in the third age group. In all age classes, REM sleep was significantly less during the first night, but REM latency was comparable on both nights. On the first night, the mean HR was higher. There were significant differences in apnoea/hypopnoea-index (AHI), electroencephalogram (EEG)-arousal-index (AI) and motoric arousal index (jerk index, JI) between Groups 1 and 2. In neither group, were there any significant differences in AHI, mean SaO(2) or number of EEG-arousals between nights 1 and 2. Only in the age class A, in Group 2 (n=9) was the number of motoric arousals significantly higher on the first night. Comparison of the age classes A, B, and C revealed that most polysomnographic parameters were age-dependent. Increasing age was found to correlate with a higher proportion of NREM 1, especially on the first night. Also, there was an age-dependent increase in NREM 2 on both nights, a decrease in NREM 3 on the first night, and a decrease in NREM 4 on both nights. In older children, we also found a lower proportion of REM sleep on the first night and a lower HR on both nights. In Group 1, we found a lowered AHI, AI and JI (for JI significant only on the first night) in older patients. No such age dependence of AHI, AI and JI was seen in OSAS patients (Group 2).

In children and adolescents, there is an FNE comparable with that described in adults. In OSAS children and also in children with no PSG-abnormalities, there is night-to-night-variability in sleep parameters, but not in respiratory parameters. An adaptation night is, therefore, necessary when sleep architecture is to be studied, but not when only the nocturnal respiratory pattern is investigated. Sleep parameters, HR and arousal indices are all age-dependent.