Validation of actigraphy sleep metrics in children aged 8 to 16 years: considerations for device type, placement and algorithms.

BACKGROUND: Actigraphy is often used to measure sleep in pediatric populations, despite little confirmatory evidence of the accuracy of existing sleep/wake algorithms. The aim of this study was to determine the performance of 11 sleep algorithms in relation to overnight polysomnography in children and adolescents. METHODS: One hundred thirty-seven participants aged 8-16 years wore two Actigraph wGT3X-BT (wrist, waist) and three Axivity AX3 (wrist, back, thigh) accelerometers over 24-h. Gold standard measures of sleep were obtained using polysomnography (PSG; Embletta MPRPG, ST + Proxy and TX Proxy) in the home environment, overnight. Epoch by epoch comparisons of the Sadeh (two algorithms), Cole-Kripke (three algorithms), Tudor-Locke (four algorithms), Count-Scaled (CS), and HDCZA algorithms were undertaken. Mean differences from PSG values were calculated for various sleep outcomes. RESULTS: Overall, sensitivities were high (mean ± SD: 91.8%, ± 5.6%) and specificities moderate (63.8% ± 13.8%), with the HDCZA algorithm performing the best overall in terms of specificity (87.5% ± 1.3%) and accuracy (86.4% ± 0.9%). Sleep outcome measures were more accurately measured by devices worn at the wrist than the hip, thigh or lower back, with the exception of sleep efficiency where the reverse was true. The CS algorithm provided consistently accurate measures of sleep onset: the mean (95%CI) difference at the wrist with Axivity was 2 min (-6; -14,) and the offset was 10 min (5, -19). Several algorithms provided accurate measures of sleep quantity at the wrist, showing differences with PSG of just 1-18 min a night for sleep period time and 5-22 min for total sleep time. Accuracy was generally higher for sleep efficiency than for frequency of night wakings or wake after sleep onset. The CS algorithm was more accurate at assessing sleep period time, with narrower 95% limits of agreement compared to the HDCZA (CS:-165 to 172 min; HDCZA: -212 to 250 min). CONCLUSION: Although the performance of existing count-based sleep algorithms varies markedly, wrist-worn devices provide more accurate measures of most sleep measures compared to other sites. Overall, the HDZCA algorithm showed the greatest accuracy, although the most appropriate algorithm depends on the sleep measure of focus.

Sleep and pre-bedtime activities in New Zealand adolescents: differences by ethnicity.

AIM: To describe the screen and nonscreen activities adolescents engage in one hour before bedtime and associations with sleep quantity and quality, including differences by ethnic group. DESIGN: Cross-sectional survey. PARTICIPANTS: 4,192 adolescents aged 13-17 years (52% boys); 71% NZ European, 13% Maori, 8% Asian, 6% Pacific, and 2% other ethnic groups. MEASURES: Participants completed questions about sleep timing, quality (Pittsburgh Sleep Quality Index), and chronotype (Morningness-Eveningness Scale for Children). Seventeen questions captured pre-bedtime activities. RESULTS: Overall, 39% slept less than the recommended hour of sleep (<8 h) and 57% reported poor sleep quality. Asian teenagers reported shorter sleep duration than New Zealand (NZ) Europeans (-45 min [95% CI: -58 to -32]) primarily from later bedtimes (1 hour), with higher odds of long sleep latency, but less disturbed sleep and a more "eveningness" chronotype. Bedtimes were later in Maori and Pacific adolescents (15 and 41 min, respectively) than NZ Europeans. Most screen activities were negatively associated with sleep quantity and quality. For nonscreen activities, snacking and drinking caffeinated beverages and alcohol were significantly associated with shorter sleep (-8, -28, and -20 min, respectively), whereas interacting with family and friends and exercise/sports before bed were associated with longer sleep (P < 0.001). Time with family, exercise, schoolwork, and household chores were all associated with better sleep quality (P < 0.001). Ethnic differences were apparent for several pre-bedtime activities. DISCUSSION: Ethnic differences related to subjective sleep parameters exist in NZ adolescents. Observed variations in sleep patterns and presleep activities suggest that sleep health messages should be tailored for different ethnic groups.

Effects of prior hypoxia exposure, endotoxin and sleep state on arousal ability to airway obstruction in piglets: implications for sudden infant death syndrome.

BACKGROUND: Respiratory tract infections may be an important component in many deaths attributed to sudden infant death syndrome (SIDS), although the mechanism of involvement remains unclear. OBJECTIVES: The hypothesis was tested that prolonged hypoxia and a thermogenic state (simulating a fever due to respiratory tract infection) would impair respiratory responsiveness to airway obstruction during sleep. METHODS: Thirty nine piglets aged 5-7 days were exposed to 24 h of moderate hypoxia and/or a low dose of endotoxin derived from Salmonella abortus equi. Responsiveness to complete and subtotal upper airway obstruction was tested during non-rapid eye movement (NREM) and rapid eye movement (REM) sleep. The end-point for airway obstruction tests was taken as the first protective response, either arousal or initiation of mouth breathing. Responsiveness was assessed as response time and response threshold (measured as respiratory effort, i.e. esophageal pressure swing). RESULTS: All animals demonstrated a thermogenic state following endotoxin delivery (drop in ear temperature of 5.8 +/- 0.2 degrees C and a small but significant increase in rectal temperature). Response time to subtotal airway obstruction was reduced during the heat conserving phase of the fever (thermogenesis; 2.8 +/- 0.5 s compared to 4.3 +/- 0.7 s during pre-endotoxin tests), but markedly increased during the recovery period (20.3 +/- 5.1 compared to 14.0 +/- 2.5 s pre-endotoxin) in NREM sleep. Response threshold was not significantly affected by either endotoxin or hypoxia in NREM sleep. Respiratory responsiveness to subtotal obstruction was markedly reduced during REM sleep (response time 40.3 +/- 10.9 s compared to 14.7 +/- 2.2 s in NREM; response threshold -14.0 +/- 1.3 mm Hg compared to -11.7 +/- 1.0 mm Hg in NREM). CONCLUSIONS: This study has demonstrated in a neonatal animal model that respiratory responsiveness to airways obstruction is delayed during recovery from fever. The findings may have implications for the human infant recovering from a respiratory illness.

Bed-sharing and the infant's thermal environment in the home setting.

AIMS: To study bed-sharing and cot-sleeping infants in the natural setting of their own home in order to identify differences in the thermal characteristics of the two sleep situations and their potential hazards. METHODS: Forty routine bed-sharing infants and 40 routine cot-sleeping infants aged 5-27 weeks were individually matched between groups for age and season. Overnight video and physiological data of bed-share infants and cot-sleeping infants were recorded in the infants' own homes including rectal, shin, and ambient temperature. RESULTS: The mean rectal temperature two hours after sleep onset for bed-share infants was 36.79 degrees C and for cot-sleeping infants, 36.75 degrees C (difference 0.05 degrees C, 95% CI -0.03 to 0.14). The rate of change thereafter was higher in the bed-share group than in the cot group (0.04 degrees C v 0.03 degrees C/h, difference 0.01, 0.00 to 0.02). Bed-share infants had a higher shin temperature at two hours (35.43 v 34.60 degrees C, difference 0.83, 0.18 to 1.49) and a higher rate of change (0.04 v -0.10 degrees C/h, difference 0.13, 0.08 to 0.19). Bed-sharing infants had more bedding. Face covering events were more common and bed-share infants woke and fed more frequently than cot infants (mean wake times/night: 4.6 v 2.5). CONCLUSIONS: Bed-share infants experience warmer thermal conditions than those of cot-sleeping infants, but are able to maintain adequate thermoregulation to maintain a normal core temperature.

Endotoxin effects on markers of autonomic nervous system function in the piglet: implications for SIDS.

The hypothesis was tested in 30 newborn piglets that the effects of a low dose of endotoxin (1 microg i.v. bolus; Salmonella abortus equi) would impair autonomic nervous system function. Two tests of autonomic function were performed following external warming (pre-endotoxin) and during endotoxin-generated thermogenesis: (1) analysis of heart rate variability in the time and frequency domains and (2) baroreflex sensitivity measured following intravenous injection of the vasoactive drugs nitroprusside and phenylephrine. Beat-to-beat heart rate variability (SDDeltaRR) fell by 2.2 ms from 7.0 ms before fever (p < 0.05). Low-frequency spectral power fell by 2.4 ms(2) from 4.1 ms(2) before fever (p < 0.05). The sensitivity of the baroreflex to changes in blood pressure induced by the vasoactive drugs decreased during fever by 0.72 ms/mm Hg for the nitroprusside test (p < 0.0005) and by 0.31 ms/mm Hg for the phenylephrine test (p < 0.005). These results indicate that in the piglet the balance of autonomic tone is altered and autonomic responsiveness reduced during the thermogenic phase of a fever. These findings are consistent with known risk factors for sudden infant death syndrome.

Respiratory responses to hypoxia/hypercapnia in small for gestational age infants influenced by maternal smoking.

AIM: To determine any variation in the respiratory responses to hypoxia/hypercapnia of infants born small for gestational age (SGA) to smoking and to non-smoking mothers. METHODS: A total of 70 average for gestational age (AGA) infants (>36 weeks gestation, >2500 g, >25th centile for gestational age, and no maternal smoking), and 47 SGA infants (<10th centile for gestational age) were studied at 1 and 3 months of age, in quiet and active sleep. Respiratory test gases were delivered through a Perspex hood to simulate face down rebreathing by slowly allowing the inspired air to be altered to a CO(2) maximum of 5% and O(2) minimum of 13.5%. The change in ventilation with inspired CO(2) was measured over 5-6 minutes of the test. The slope of a linear curve fit relating inspired CO(2) to the logarithm of ventilation was taken as a quantitative measure of ventilatory asphyxial sensitivity (VAS). RESULTS: There was no significant difference in VAS between the AGA and SGA infants (0.25 v 0.24). However within the SGA group, VAS was significantly higher (p = 0.048) in the infants whose mothers smoked during pregnancy (0.26 (0.01); n = 24) than in those that did not (0.23 (0.01); n = 23). The change in minute ventilation was significantly higher in the smokers than the non-smokers group (141% v 119%; p = 0.03) as the result of a significantly larger change in respiratory rate (8 v 4 breaths/min; p = 0.047) but not tidal volume. CONCLUSIONS: Maternal smoking appears to be the key factor in enhancing infants' respiratory responses to hypoxia/hypercapnia, irrespective of gestational age.

Prone versus supine sleep position: a review of the physiological studies in SIDS research.

A number of physiological studies, published over the last 10 years, have investigated the links between prone sleeping and sudden infant death syndrome (SIDS). This review evaluates those studies and derives an overview of the different affects of sleeping prone or supine in infancy. Generally, compared with the supine, the prone position raises arousal and wakening thresholds, promotes sleep and reduces autonomic activity through decreased parasympathetic activity, decreased sympathetic activity or an imbalance between the two systems. In addition, resting ventilation and ventilatory drive is improved in preterm infants, but in older infants (>1 month), there is no improvement in ventilation, and in 3-month-old infants, the position is adverse in terms of poorer ventilatory drive (in active sleep only). The majority of findings suggest a reduction in physiological control related to respiratory, cardiovascular and autonomic control mechanisms, including arousal during sleep in the prone position. Since the majority of these findings are from studies of healthy infants, continued reinforcement of the supine sleep recommendations for all infants is emphasized.

Ventilatory responses to rebreathing in infants exposed to maternal smoking.

UNLABELLED: This study assessed the effect of maternal smoking during pregnancy on ventilatory and waking responses of infants to a respiratory challenge. This challenge mimicked the time-course and concentration of gases that an infant would experience rebreathing face-down into soft bedding. Control (C; n = 97) and smokers' infants (SM; n = 96) were studied at ages I and 3 mo. Asphyxial gas (hypercapnia/hypoxia) was delivered to the supine sleeping baby via a hood by slowly altering the inspired air: CO2 maximum 5% and O2 minimum 13.5%. Respiratory pattern was recorded by inductive plethysmography as the sum of ribcage and abdominal movements. The change in ventilation with inspired CO2 was measured over 5-6 min of the test. The slope of a linear curve fit relating inspired CO2 to the logarithm of ventilation was taken as a quantitative measure of ventilatory asphyxial sensitivity (VAS). Protective responses were graded from 1: no waking and an estimated arterial carbon dioxide tension (PaCO2) > or = 60 mmHg (least protective), to 4: fully awake (most protective). The results showed VAS was higher in SM infants than controls: +0.03 (p = 0.04). The oxygen saturation (SaO2) of SM infants fell -0.4% (p = 0.02) more than SaO2 of controls despite a greater tidal volume increase: +13.0% (p = 0.04). Overall protective responses were the same between groups, but grade 1, although rare, was found in 7 SM infants and only 4 control infants; this difference was not significant. CONCLUSION: The study did not confirm the postulate that infants of mothers who smoked during pregnancy have a reduced ventilatory response or raised waking thresholds.

Participation in research: informed consent, motivation and influence.

OBJECTIVE: To investigate the process and quality of informed consent, motivation and influence in parents who were invited to enroll their baby in a research project. METHODOLOGY: A mixed quantitative/qualitative questionnaire was sent to a cohort invited to participate in a physiological research project on sudden infant death syndrome (SIDS) at the Dunedin Public Hospital, Dunedin, New Zealand. Separate questionnaires were used for parents who participated (94) and those who declined to participate (103). Response rates were 69% and 47%, respectively. RESULTS: All consenting parents felt they understood the purpose and procedure of the study. The majority (90%) thought the information about the study was very good; 6.5% felt more detail was required. Eighty-five per cent found the verbal explanation the most useful source of information. All participated for altruistic reasons such as to aid SIDS research. Although 27% had concerns about safety of the tests, after the tests all responders felt happy with the safety of the tests. Inconvenience was the main reason (53%) for declining to participate. Twenty-eight per cent of declining parents were concerned about the safety of the tests. CONCLUSION: Of those who responded to the questionnaire, the process for obtaining informed consent in the SIDS studies was satisfactory. Parents' motives for participating were mostly altruistic. The role of recall bias and selection bias may make the implications of this study unclear.

Ventilatory sensitivity to mild asphyxia: prone versus supine sleep position.

AIMS: To compare the effects of prone and supine sleep position on the main physiological responses to mild asphyxia: increase in ventilation and arousal. METHODS: Ventilatory and arousal responses to mild asphyxia (hypercapnia/hypoxia) were measured in 53 healthy infants at newborn and 3 months of age, during quiet sleep (QS) and active sleep (AS), and in supine and prone sleep positions. The asphyxial test mimicked face down rebreathing by slowly altering the inspired air: CO(2), maximum 5% and O(2), minimum 13.5%. The change in ventilation with inspired CO(2) was measured over 5-6 minutes of the test. The slope of a linear curve fit relating inspired CO(2) to the logarithm of ventilation was taken as a quantitative measure of ventilatory asphyxial sensitivity (VAS). Sleep state and arousal were determined by behavioural criteria. RESULTS: At 3 months of age, prone positioning in AS lowered VAS (0.184 prone v 0.269 supine, p = 0.050). At newborn age, sleep position had no effect on VAS. Infants aged 3 months were twice as likely to arouse to the test than newborns (p = 0.013). Placing infants prone as opposed to supine increased the chances of arousal 1.57-fold (p = 0.035). CONCLUSION: Our findings show 3 month old babies sleeping prone compared to supine have poorer ventilatory responses to mild asphyxia, particularly in AS, but the increased prevalence of arousal is a protective factor.

Factors affecting heart rate variability and heart rate responses to tilting in infants aged 1 and 3 months.

Heart rate variability (HRV) and heart rate (HR) responses following a 60 degree head-up tilt were measured in 60 infants at 1 and 3 mo of age to investigate the effects on these of age, sleep state, sleep position, and mother's smoking status. HRV was determined from Poincaré plots of 500 sequential RR intervals to measure overall variability derived from the SDRR of this plot, and instantaneous variability derived from the SDdeltaRR. HR responses to the tilt were measured as changes in RR interval length from rest to immediately following the tilt and again once a stable pattern was reached. SDRR and SDdeltaRR increased 20 and 40%, respectively, with age (p < 0.0001), SDRR was higher in active sleep (AS) than quiet sleep (QS, +72%, p < 0.0001) but both measures of variability (SDRR and SDdeltaRR) were lower in the prone position compared with supine (-18%, p < 0.0001). However, several findings were dependent on the basal RR interval, thus the age effect disappeared once RR interval was taken into account, sleep state remained an important factor and the lower variability when prone now became a difference of -3% (p = 0.034). The tilt generally provoked a reflex tachycardia followed by a bradycardia and settling to a stable HR level below, at, or above baseline within 30 s. The more unusual responses were no HR change, sustained tachycardia or sustained bradycardia (15% of total). These were more likely to occur in younger infants (p = 0.008) and in AS (p < 0.0001). No changes were seen in any of the cardiac indices related to maternal smoking status. The findings confirm several reports indicating that prone sleeping damps some physiologic responses. The data emphasize the need to consider basal heart rate, and sleep position as well as sleep state in autonomic function testing during infant sleep.

Vasoconstriction following spontaneous sighs and head-up tilts in infants sleeping prone and supine.

The cutaneous vasoconstrictor responses following a 60 degrees head-up tilt and a spontaneous sigh were measured in 36 infants at 1 and 3 months age to investigate the effects of age, sleep state and sleep position on these responses. The vasoconstrictor response was determined by a measure of cutaneous blood flow using a laser Doppler flowmeter. The mean reduction in blood flow (vasoconstriction) was 52% following the tilt, and 33% following the sigh. Prone positioning 1-month-old infants as compared to supine, reduced the degree of vasoconstriction following the tilt (P=0.027) and sigh (P=0.026). The supine to prone reduction was: tilt, -11% in quiet sleep (QS) (from 55.1 to 49.1% vasoconstriction) and -18% in active sleep (AS) (from 52.0 to 42.9%) and; sigh, -26% in QS (35-26%), and -15% in AS (31-26%). The degree of vasoconstriction following the sigh was significantly greater in 3- compared to 1-month-old infants (+26%, P=0.040). The mean response to the tilt in the older age group was 12% greater but this did not reach significance (P=0.069). Sleep state did not affect the degree of vasoconstriction but influenced transmission of the response so that latency to minimal vasoconstriction was 1 s shorter in AS than QS. This study provides data on two simple measures of sympathetic activity during sleep that have not previously been described in any detail in infant studies, and add more evidence that autonomic activity is reduced in the prone position compared to supine during sleep.

The effect of cold stimulation to the face on the metabolic rate of the febrile piglet.

Sudden infant death syndrome has been associated with winter climates, infection, and overwrapping of babies. The hypothesis has been tested in this laboratory that two different causes of increased metabolic rate, high core temperature (via the van't Hoff or 'Q10' effect) and face-cooling, might synergistically induce hyperthermia. This proved not to be the case. We now report on a 'febrile' state adding Salmonella abortus equi pyrogens. The combination of face-cooling and pyrogen administration to 14 already hot piglets produced an increase in oxygen consumption of 47% in 6 of the animals (19% overall). Face-cooling alone caused a 6.5% fall in oxygen consumption, and injection of pyrogens alone had no effect on oxygen consumption. We conclude that there may be a danger of life-threatening hyperthermia in the combination of a cold face and febrile state.

Sleep position, autonomic function, and arousal.

AIMS: To investigate and compare heart rate variability (HRV) and responses of heart rate and arousal to head-up tilting in infants sleeping prone and supine. METHODS: Thirty seven healthy infants aged 2-4 months were studied. HRV was measured for 500 beats while they were in a horizontal position. Subjects were then tilted 60 degrees head-up, and heart rate recorded over 1 minute and arousal responses observed. Data were collected during both quiet and active sleep for both prone and supine sleep positions. RESULTS: HRV, as assessed by the point dispersion of Poincaré plots, was significantly reduced in the prone position for both sleep states. Sleep position did not influence the changes in heart rate seen during a head-up tilt. Full awakening to the tilt was common in active sleep but significantly less so in the prone position (15% of prone tests vs 54% supine). Full awakening to the tilt rarely occurred during quiet sleep in either sleep position. CONCLUSION: This study provides some evidence that blunted arousal responses and/or altered autonomic function are a feature of the prone sleeping position. Decreased HRV may be a sign of autonomic impairment. It is seen in many disease states and in infants who later die of sudden infant death syndrome (SIDS).

Arousal responses of the newborn piglet to airways obstruction and rebreathing during normothermia and hyperthermia.

Most experimental studies of arousal to respiratory stressors have been performed under neutral thermal conditions. The present study focussed on arousal responses with the added challenge of heat stress. The subjects were two groups of 10 sedated piglets aged 4-6 days and 2-6 days respectively. Respiratory stimuli (partial and total airways obstruction (AO) or rebreathing (RB) expired gases) were applied during non-rapid-eye-movement (NREM) sleep. Measurements of heart rate, SaO2, intra-pleural pressure or inspired CO2 and O2 were recorded during tests in normothermia and hyperthermia. Hyperthermia significantly shortened the time to arousal. Thus, arousal times from partial and total AO and from RB were: in normothermia, 40.3 +/- 5.8, 9.2 +/- 0.7 and 94.9 +/- 14.7 s respectively; in hyperthermia, 17.8 +/- 3.7, 7.6 +/- 0.9 and 68.5 +/- 9.1 s respectively. Cardio-respiratory variables at arousal were similar in both thermal states. Seven non-sedated piglets were challenged with RB stimuli during normothermia only to determine the influence of sleep state on arousal. Arousal in REM sleep was delayed approximately 2-3-fold (NREM, 41.2 +/- 8.2 s; REM, 88.5 +/- 18.1 s), occurring at a lower SaO2 and higher inspired CO2. It is concluded that arousal from NREM sleep under adverse conditions of hyperthermia shortens the arousal time from asphyxial stimuli induced by AO and RB with no change in the arousal threshold.

The micro-environment of the sleeping newborn piglet covered by bedclothes: gas exchange and temperature.

This study followed the thermal and gaseous micro-environment of the newborn piglet sleeping under two levels of bedding insulation. After 1 h in the piglet's thermal comfort zone, the head of each piglet was covered for 2 h with bedding either 12 mm or 21 mm thick. Body temperature rose rapidly, but the arterial gases showed no change in the direction of asphyxia; with the thicker covering PaO2 was unchanged and PaCO2 fell and pH rose. These changes were despite a small rise in environmental PCO2 and fall in PO2 but were also minimized by the piglet's ability to avoid profound hypocapnia by panting at a very low tidal volume. No significant changes in blood gases occurred with the thinner bedding; the temperature rose markedly but more slowly. It is suggested that human babies submerged under bedclothes for any length of time would be more likely to succumb to the effects of hyperthermia than of asphyxia.

Sleep state organization in the developing piglet during exposure to different thermal stimuli.

Sleep state changes in response to different thermal stimuli were investigated in newborn piglets between 2 and 10 days of age. Test animals were exposed to cold air (7-12 degrees C) and warm air (27-33 degrees C) around the face, while the remainder of the body was kept at first warm (normothermic) then hyperthermic. A separate group of animals was studied under normothermic conditions (control) for the duration of the study. Piglets showed typical changes in sleep state patterns characteristic of rapid maturation over the first 10 days of development. It was found that both the amount of rapid eye movement (REM) sleep and, in some cases, the duration of REM episodes increased in response to facial cooling regardless of rectal temperature. However, hyperthermia with warm air exposure caused a significant decrease in the amount of REM sleep but not in the duration of REM episodes. It is suggested that an infant placed to bed in a cold room or exposed to a draft might also experience a greater amount of REM sleep than an infant placed to sleep in a warm draft-free room.

Apnea and rapid eye movement sleep excess in the piglet during recovery from hyperthermia.

This investigation tested the hypothesis that artificially induced mild hyperthermia and recovery from hyperthermia in the developing newborn piglet (2-10 d of age) alter sleep state patterns and respiratory control. Continuous measurements of sleep state, respiratory pattern, carbon dioxide production and oxygen consumption were made before, during, and after a 3-h period of sustained hyperthermia. During hyperthermia, rectal temperatures were raised a mild 1-1.5 degree C above normal, well below the levels likely to cause severe physiologic distress in this species. This resulted in a disruption of the sleep state pattern characterized by a decrease in duration of rapid eye movement (REM) episodes, whereas immediately afterwards, during recovery, the amount of REM sleep increased. In some cases the amount of REM sleep in recovery more than doubled the basal level. Apneas were rarely observed during hyperthermia, but in recovery there was an increase in the total amount of time spent in apnea in both REM and non-rapid eye movement sleep states with a predominance in the REM state. Arousal responses to chemostimulation were not affected at this time. We conclude that the sleeping newborn piglet does indeed show marked changes in sleep state pattern (particularly REM sleep) and in the amount of apnea recorded during and immediately after only a mild increase in core temperature.

Rebreathing expired gases from bedding: a cause of cot death?

The reported association of cot death and sleeping prone could be due to rebreathing of expired gases. A mechanical model simulating the respiratory system of an infant, exhaling warm humidified air with an end tidal carbon dioxide of 5%, has been used to investigate this. Some commonly used bedding materials caused an accumulation of carbon dioxide of 7% to over 10% with the model lying face down. This phenomenon persisted even with the head inclined at 45 degrees, but only on very soft materials, and could be a cause of cot death in a baby unresponsive to asphyxial blood gas changes. A coir fibre mattress allowed complete dispersal of exhalate as did a rubber sheet between any mattress and the covering sheet.