OBJECTIVE: Prone sleeping is a major risk factor for the sudden infant death syndrome and is associated with lower blood pressure and impaired arousability from sleep, both of which may be signs of cerebral hypoxia. However, the impact of sleep position on cerebral oxygenation during infancy remains unknown. We assessed the effects of sleeping position, sleep state, and postnatal age on cerebral oxygenation by measuring tissue oxygenation index (TOI) during the first 6 months of infancy.
SUBJECTS AND METHODS: Seventeen healthy term infants (8 girls and 9 boys) were recruited as study participants. Infants were studied at ages 2 to 4 weeks, 2 to 3 months, and 5 to 6 months by use of daytime polysomnography, with additional measurements of blood pressure (Finometer, FMS Finometer Medical Systems, Amsterdam, Netherlands) and tissue oxygenation index (TOI) (NIRO 200 spectrophotometer, Hamamatsu Photonics KK, Tokyo, Japan).
RESULTS: In infants who slept in the prone position, TOI was lower in both quiet sleep (QS) and active sleep (AS) at age 2 to 4 weeks and in QS at age 2 to 3 months (P < .05). TOI was lower in AS compared with QS in infants aged 2 to 4 weeks (P < .05). When the infants reached 5 to 6 months of age, TOI was greater in AS (P < .05), as there was a profound decrease in TOI during QS (P < .05) over this period. No relationship was identified between blood pressure and TOI at any age.
CONCLUSIONS: In healthy infants cerebral oxygenation is reduced during sleep in the prone position. This reduction may underpin the reduced arousability from sleep exhibited by healthy infants who sleep prone, a finding that provides new insight into potential risks of prone sleeping and mechanisms of sudden infant death syndrome.
WHAT'S KNOWN ON THIS SUBJECT:
Prone sleeping is a major risk factor for the sudden infant death syndrome. In the prone position blood pressure is often lower and arousability is depressed.
WHAT THIS STUDY ADDS:
Normative data on the effects of sleeping position, sleep state, and postnatal age on cerebral oxygenation were collected. Infants sleeping in the prone position had lower cerebral oxygenation, which may contribute to the decreased arousability and increased risk for sudden infant death syndrome associated with this sleeping position.
Despite its dramatic decline in recent years, sudden infant death syndrome (SIDS) remains the leading cause of mortality in infants aged between 1 month and 1 year in the Western world.1,2 SIDS is presumed to occur during sleep1; however, which sleep state, active sleep (AS) or quiet sleep (QS), places an infant at greatest risk is yet unknown. Despite intensive research the exact mechanisms that underpin SIDS also remain uncertain. The pattern of cardiovascular changes seen during physiologic recordings of future SIDS victims suggests that an impairment of protective autonomic cardiovascular control mechanisms in association with a failure to arouse from sleep are involved.3,–,7 In support of this hypothesis, autopsy studies have revealed brainstem abnormalities in SIDS victims in areas responsible for cardiorespiratory control and arousal from sleep.8
Sleeping in the prone position has been identified as a major risk factor for SIDS.9,–,11 Previously we have demonstrated that when healthy term infants sleep in the prone position they are 3 times more difficult to arouse from sleep, a characteristic that is most marked at 2 to 3 months of age, the age when SIDS risk is greatest.12 Results of our recent studies have shown that the arousal process itself is altered by sleep position, with cortical arousal being promoted in this vulnerable position in healthy term infants.13 These findings contrast with the results of previous studies, which have shown that cortical arousal was reduced in infants who died from SIDS.6 In addition, our study results have shown that the prone position also has a significant effect on blood pressure (BP) and its control, with healthy term infants exhibiting a fall in BP despite an elevated heart rate, again most marked at 2 to 3 months of age and in QS.14,15 What is yet to be ascertained is if this reduction in BP when infants sleep prone is reflected in reduced cerebral blood flow and oxygenation. Any cerebral oxygenation impairment may underpin the reduced arousal responses we and others have previously reported for infants sleeping in this position.
To address this issue we used continuous measurement of BP in combination with near-infrared spectroscopy to measure the cerebral tissue oxygenation index (TOI, %) in infants who were sleeping prone and supine. TOI represents the mixed oxygen saturation in all cerebral vascular compartments and has not been used previously in infant sleep studies designed to investigate oxygenation changes associated with risk factors for SIDS. We hypothesized that TOI would be reduced in the prone position, and that this effect would be most marked at the age of 2 to 3 months, which coincides with the increased vulnerability to SIDS at this age.
Ethics approval for this project was obtained from the Southern Health and Monash University human research ethics committees. No monetary incentive was provided for participation, and written parental consent was obtained before commencement of the study.
Seventeen healthy term infants (8 girls and 9 boys) born at 38 to 42 weeks' gestational age were studied. Infants had normal birth weights ranging from 2900 to 4615 g (mean 3666 ± 105 g) and Apgar scores ranging from 9 to 10 (median 9) at 5 minutes. All infants were born to nonsmoking mothers, routinely slept in a supine position at home, and were breast fed. Each infant was studied longitudinally by use of daytime polysomnography on 3 occasions over the first 6 months of life, at age 2 to 4 weeks, when the risk of SIDS is low; at 2 to 3 months, when the risk of SIDS peaks; and 5 to 6 months, when the risk of SIDS is again low.
Electrodes required for definition of sleep and to record other physiologic variables were attached during a routine morning feed. Measurement procedures comprised electroencephalogram, electro-oculogram, submental electromyogram, electrocardiogram, thoracic and abdominal breathing movements (Resp-ez bands, Pro-Tech Services, Inc, Mukilteo, WA, USA), arterial blood oxygen saturation (SPo2) (Masimo Radical Oximeter, Masimo Corporation, Irvine, CA, USA), and abdominal skin temperature (ADInstruments, Sydney, Australia).
BP was measured by using a noninvasive photoplethysmographic cuff (Finometer, FMS, Finapres Medical Systems, Netherlands) placed around the infant's wrist, a method previously validated in preterm infants16,17 and used by our group in both term and preterm infants.14,15,18,19
Cerebral Oxygenation Measurements
Cerebral oxygenation (tissue oxygenation index, TOI, %) was continuously recorded by using spatially resolved spectroscopy (NIRO 200 spectrophotometer, Hamamatsu Photonics KK, Tokyo, Japan). To enable a quantitative measure of TOI, spatially resolved spectrometry employs continuous-wave light emission coupled with light detectors at multiple distances. The NIRO-200 system transmits light at 3 wavelengths in the near-infrared range at 775, 810, and 850 nm, delivered via a fiber optic bundle terminating in an emission probe. A detection probe sited 4 cm from the emission probe incorporates 2 aligned photograph detectors, which are separated by 4 mm. Both the emission and detection probes were placed over the frontal region of the head and covered with light-proof dressing. The spatially resolved spectroscopy algorithm computes TOI by using slopes of near-infrared light attenuation versus the different distances of the 2 photograph-detectors from the emission probe. By fitting these data to a modified diffusion equation of light transport in tissue, the ratio of concentrations of oxyhemoglobin to total hemoglobin, and hence the absolute average TOI, is computed continuously.20,21
All physiologic data were recorded at a sampling frequency of 512 Hz by using a Compumedics E-Series sleep recording system with ProFusion PSG 2 software (Compumedics Limited, Abbortsford, Vic, Australia). At the completion of the study, data were exported via European Data Format to Chart 5.2 analysis software (ADInstruments, Sydney, Australia). Sleep state was defined as QS or AS by using electroencephalographic patterns, heart rate (HR), breathing patterns and behavioral criteria.22,23
Polysomnography was performed between 0900 and 1600 hours in a sleep laboratory. For the course of the study, noise and light levels were kept to a minimum and infants were studied under constant room temperature (22–23°C). Once the infant was in stable sleep, BP measurements were made in baseline epochs of 1 to 2 minutes for the duration of each sleep cycle, with a total of 10 to 12 minutes being recorded in each infant for each sleep state. Infants were closely monitored to ensure there were no changes in behavior, sleep state, or HR induced by BP cuff inflation. Infants were positioned for sleep in both the prone and supine positions, with the initial position randomized. Sleeping position was changed between morning and afternoon sleep periods that were interrupted by a midday feed.
Beat-to-beat TOI, mean arterial pressure (MAP), and HR values were calculated (Chart 5.2). When movement artifacts disrupted physiologic signals, data that lay 1.5 times the interquartile range outside the first and third quartiles were removed from subsequent analyses.15 For each infant, TOI, MAP, and HR were averaged over a 1-minute period for QS and AS, and data were pooled for each study and sleeping position.
Statistical analysis was performed using SigmaStat (Systat Software Inc, Richmond, CA, USA). The effects of sleep state and position on TOI, BP, and HR were assessed using a 2-way repeated measures analysis of variance with Student-Newman-Keuls posthoc analysis at each age. The effects of age on TOI, BP and HR within each sleep state and position were assessed using a 1-way analysis of variance with Student-Newman-Keuls posthoc analysis to identify specific sources of differences. Linear regression analysis was performed to assess the relationship between TOI and MAP, separated for sleep state, sleeping position, and age. Results are presented as mean ± SEM, with significance taken at the P < .05 level.
Effects of Sleeping Position and Sleep State
Figure 1 illustrates an example of the effect of sleep position on TOI in an infant studied at 5 to 6 months of age, with lower TOI during prone sleeping. Data for all infants on the effects of sleeping position and sleep state on TOI are presented in Fig 2. Sleeping position had a significant effect on TOI, with lower TOI identified in infants sleeping in the prone position for both sleep states. Lower TOI when prone was found at all ages, and reached significance during both QS and AS in infants at the age of 2 to 4 weeks (5% and 4% lower in prone than supine, respectively; P < .05), and during QS at the age of 2 to 3 months (3% lower; P < .05). Sleep state also had a significant effect on TOI, which was significantly lower in AS than QS in infants 2 to 4 weeks old; this difference was evident in both supine and prone positions (2% and 2%, respectively, P < .05). At the age of 2 to 3 months infants showed no significant difference in TOI between sleep states, whereas at the age of 5 to 6 months TOI became higher in AS in both the supine and prone sleeping positions (3% and 3%, respectively, P < .05).
Cardiorespiratory parameters are presented in Table 1. HR was higher in AS compared with QS at age 5 to 6 months in both sleeping positions (P < .05). MAP was usually higher in AS compared with QS at each age and in both sleeping positions (P < .05), with the exception of the prone position at age 5 to 6 months. HR and MAP were usually higher in the prone position compared with the supine position; these differences reached significance for HR at both 2 to 3 months and 5 to 6 months of age in both sleep states, and for MAP at 2 to 4 weeks of age during AS (P < .05). There were no effects of sleeping position or sleep state on SPo2 or abdominal skin temperature at any of the 3 ages studied.
Effects of Postnatal Age
The effects of postnatal age on TOI are presented in Fig 3. During QS, a significant decrease in TOI was observed from 2 to 4 weeks to 5 to 6 months of age in both the prone and supine positions (P < .05), with the decrease being most profound between 2 to 4 weeks and 2 to 3 months of age. The largest decrease in TOI occurred in QS in the supine position, averaging 9% across the first 6 months of age. A lesser trend for TOI to decrease was also evident in AS, with the largest difference (5%) between 2 to 4 weeks to 2 to 3 months of age in the supine position.
MAP tended to increase with postnatal age; this increase reached statistical significance in QS in the prone sleeping position (Table 1). HR decreased significantly with postnatal age in both sleep positions and sleep states (Table 1).
Regression of TOI and MAP
Results of regression analyses performed for each sleep state, sleeping position, and age revealed no significant relationships between TOI and MAP.
During the period of infancy remarkable developmental increases occur in the cerebral metabolic requirements for oxygen, which may expose infants to a heightened risk of cerebral hypoxia. This study was the first to examine developmental changes in cerebral oxygenation longitudinally in healthy term infants during the first 6 months after birth, while taking into account the effects of both sleep state and sleeping position. Lower cerebral oxygenation was associated with the prone sleeping position and advancing postnatal age. In addition, sleep state had significant effects on cerebral oxygenation. TOI was initially significantly lower in AS compared with QS in infants at age 2 to 4 weeks, then exhibited no difference between sleep states at age 2 to 3 months, and subsequently was significantly higher in AS compared with QS at age 5 to 6 months.
TOI is an indicator of oxygen saturation in all cerebral vascular compartments, but TOI is most influenced by the venous compartment, which comprises ∼75% of total cerebral blood volume.24 When arterial oxygen saturation is maintained constant, as in the healthy infants we studied, the level of cerebral venous saturation as reflected by TOI indicates the degree of oxygen extraction, and therefore the balance between cerebral blood flow and cerebral oxygen consumption. Accordingly, a lower TOI associated with the prone sleeping position, advancing postnatal age, and AS at 2 to 4 weeks of postnatal age, represents a relative reduction of cerebral blood flow compared with cerebral oxygen consumption.
Effects of Sleeping Position on Cerebral Oxygenation
To our knowledge, this is the first study to demonstrate lower levels of cerebral oxygenation in the prone sleep position compared with supine position in early infancy. Because the lower TOI in the prone position was not associated with significant changes in cardiorespiratory parameters or body temperature, lower TOI may have arisen from position-dependent perturbations in cerebral arterial blood flow and venous drainage. Results of postmortem anatomic studies in SIDS victims have indicated that prone sleeping is likely to be associated with reduced cephalic blood flow, particularly in the vertebral arteries.25 Similarly, Doppler studies in infants indicate that increased head rotation causes obstruction of the ipsilateral jugular vein.26,27 Consequently, position-related vessel compression may also occur in the cerebral venous-drainage system, which may lead to cerebral venous congestion and hence a fall in TOI. Interestingly, the difference in TOI between the sleep positions in our study was most marked during the neonatal period and became insignificant with increasing age, possibly because of anatomic maturation of the head and neck conformation and a concomitant reduction of head-to-body ratio with age.
Clinically, low cerebral oxygenation may possibly contribute to the high risk for SIDS associated with prone sleeping. Notably, reductions in overall arousability from sleep occur in infants sleeping in the prone position 12,13, a deficit which has been proposed as one of the mechanisms underlying SIDS. In support of this idea future SIDS victims also displayed reduced overall arousability and decreased cortical arousals.6 In addition, infants who routinely slept in the prone position exhibited reduced cortical arousals in AS.28 In contrast, results of our studies in healthy term infants who routinely slept in the supine position showed increased cortical arousals when they were placed in the prone position for sleep, a finding that we speculate is protective in this vulnerable position, because these infants did not die.13 Future studies to test arousal thresholds together with cerebral TOI will be needed to delineate the relationship between arousability and cerebral oxygenation.
In contrast to our previous studies,14 in this study we did not identify lower BP in all infants in the prone versus supine sleeping positions. Nevertheless, a reduction in TOI with prone sleeping was clearly evident, which suggests that TOI is not determined by systemic cardiovascular changes, although reductions in TOI related to sleep position may be exacerbated should BP fall in this sleeping position.14
Effects of Postnatal Age on Cerebral Oxygenation
Our results of a reduction of cerebral oxygenation with age during early infancy, especially between the neonatal period and the age of 2 to 3 months, are consistent with another report of infants studied up to the postnatal age of 6 to 8 weeks.29 In this study we expanded these findings by studying over the first 6 months of life. We propose that the decrease in cerebral oxygenation with postnatal age is attributable to increased cerebral oxygen extraction, because the rapid postnatal increase in cerebral oxygen consumption30,31 outstrips the increment in cerebral blood flow and oxygen delivery. Cerebral oxygen consumption shows a maturational increase through infancy and childhood before a subsequent decline in adulthood.30,31 The lower level of cerebral metabolism in early infancy is likely to be attributable to the presence of fewer neurons and synapses32 and lower neural activity at this earlier age.33,34 In infants during the postnatal period, concordant increases in both cerebral oxygen consumption and cerebral blood flow have been reported,29,35 which indicate that cerebral blood flow is coupled to increasing metabolism in the developing brain. However, the progressive reduction in cerebral oxygenation before the age of 2 to 3 months, which has been observed by us and other investigators,29 suggests that the flow-metabolism coupling is either limited or inadequate at this age. Subsequently, the difference in cerebral oxygenation becomes insignificant between the ages of 2 to 3 months and 5 to 6 months, which suggests that maturation of flow-metabolism coupling occurs during this stage of development.
Effects of Sleep State on Cerebral Oxygenation
Neonatal AS has been proposed to be a state of endogenous brain activation essential for sustaining brain development at a time when wakefulness is limited.36,37 The high brain activity during AS results in higher cerebral oxygen consumption compared with QS, with values being similar to those of wakefulness during the newborn period.37,38 Thus, lower cerebral oxygenation during AS in the newborn period may reflect a state of high cerebral metabolic demand, which exacerbates the developmental deficit in cerebral blood flow and oxygen delivery.
Our data are supported by results of sleep studies in preterm infants, which have also suggested that cerebral blood flow does not increase during AS compared with QS, as it does during wakefulness.39 Accordingly, increased oxygen extraction is required to sustain cerebral oxygen consumption during AS, which results in reduced cerebral venous oxygenation and hence a fall in TOI. The fall in TOI we have demonstrated during AS is in agreement with the results of a previous study in term newborn infants who slept in prone or lateral positions, which showed reduced cerebral oxyhemoglobin and increased deoxyhemoglobin during AS.40
Our longitudinal study design enabled us to demonstrate that the difference in TOI between sleep states in infants became insignificant at the age of 2 to 3 months, with TOI in AS subsequently becoming higher than in QS by the age of 6 months. Developmental maturation of QS into staged nonrapid eye movement sleep41 is associated with a additional reduction of cerebral oxygen consumption in QS42 and could underpin a widening with postnatal age of the difference in cerebral oxygen consumption between AS and QS. Our finding of a concomitant, progressive increase in TOI in AS relative to QS with postnatal age suggests that cerebral blood flow increases to match the high metabolic demand of AS, consistent with a maturation of cerebral blood-flow–metabolism coupling after the age of 2 to 3 months.
Conclusions and Implications for SIDS
Several of the new findings of this study in healthy term infants at low risk for SIDS have implications that may eludicate the mechanism of SIDS in high risk infants. Of significance, a nadir in TOI occurred in infants at the age of 2 to 3 months, when SIDS risk is greatest. Our study also demonstrated that simply placing infants to sleep in the prone position results in a decrease in cerebral oxygenation; we suggest that this lowering of cerebral oxygenation in the prone position may underpin the previously reported reduced arousability of infants in this position.12 Considering the potential for cerebral hypoxia related to sleep state, the high oxygen extraction and low cerebral venous oxygenation during AS in early neonatal life indicates that during this period infants may have a diminished capacity for any additional increase in oxygen extraction should there be coincident hypoxic stress, hypotension, and cerebral hypoperfusion. In support of AS being a vulnerable state with a lower safety margin against cerebral hypoxia, imposed hypoxia is a particularly potent arousing stimulus in AS in early infancy.43,44 Later in postnatal life, a potential for deficits in cerebral oxygenation emerges in QS, because TOI decreases with age in this sleep state. This decrease in TOI coincides with the period when the probability of arousal was significantly decreased during QS compared with AS at both 2 to 3 months and 5 to 6 months of age.12
In summary, a heightened risk of cerebral hypoxia may be a critical mechanism by which the risk for SIDS is exacerbated. An important consideration is that impairment in cerebral oxygenation may be further exacerbated in infants at increased risk for SIDS, such as preterm infants and infants whose mothers smoked during pregnancy,11 both factors that may adversely affect the regulation of cerebral blood flow and oxygen delivery.
This project was supported by Scottish Cot Death Trust and the National Health and Medical Research Council of Australia. Dr Flora Wong was supported by the Kathleen Tinsley Fellowship, Monash University, and Dr Stephanie Yiallourou was supported by the Kaarene Fitzgerald Fellowship, SIDS and Kids, Victoria.
We acknowledge all the parents and infants who participated in this study and the support of staff of the Melbourne Children's Sleep Centre.
- Accepted November 24, 2010.
- Address correspondence to Rosemary S. C. Horne, PhD, Ritchie Centre, Level 5, Monash Medical Centre, 246 Clayton Rd, Clayton, Victoria, Australia 3168. E-mail:
Drs Wong and Witcombe contributed to the conception and design of the study, acquisition of the data, analysis and interpretation of the data, drafting and revision of the manuscript, and acquisition of funding for the project. Dr Yiallourou contributed to the conception and design of the study and acquisition of the data and reviewed the final manuscript. Ms Sophie Yorkston was a research assistant on the project and assisted with data collection and analysis. Ms Alicia Dymowski and Ms Lalitha Krishnan participated in this study as part of the bachelor of behavioural neuroscience honours degree program. Professor Walker contributed to the conception and design of the study, interpretation of the data, revision of the manuscript, and acquisition of funding for the project. Associate Professor Horne contributed to the conception and design of the study, acquisition of the data, interpretation of the data, drafting and revision of the manuscript, and acquisition of funding for the project.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
- SIDS =
- sudden infant death syndrome •
- QS =
- quiet sleep •
- AS =
- active sleep •
- BP =
- blood pressure •
- TOI =
- tissue-oxygenation index •
- HR =
- heart rate •
- MAP =
- mean arterial blood pressure
- Krous HF,
- Beckwith JB,
- Byard RW,
- et al
- Yiallourou SR,
- Walker AM,
- Horne RS
- Witcombe NB,
- Yiallourou SR,
- Walker AM,
- Horne RS
- Anders T,
- Emde R,
- Parmelee A
- Curzi-Dascalova L,
- Mirmiran M
- Pamphlett R,
- Raisanen J,
- Kum-Jew S
- Watson GH
- Cowan F,
- Thoresen M
- Roche-Labarbe N,
- Carp SA,
- Surova A,
- et al
- Levene MI,
- Lilford RJ
- England MA
- Roffwarg HP,
- Muzio JN,
- Dement WC
- Jenni OG,
- Borbely AA,
- Achermann P
- Copyright © 2011 by the American Academy of Pediatrics