ARTICLE
Hypoxic and Hypercapnic Events in Young Infants
During Bed-sharing
AUTHORS: Sally A. Baddock, PhD,a,b Barbara C. Galland,
PhD,a David P.G. Bolton, MRCP, PhD,a Sheila M. Williams,
DSc,c and Barry J. Taylor, MBChB, FRACPa
Departments of aWomen’s and Children’s Health, and cPreventive
and Social Medicine, University of Otago, Dunedin, New Zealand; and
bSchool of Midwifery, Otago Polytechnic, Dunedin, New Zealand
KEY WORDS
co-sleeping, infant, oxygen desaturation, rebreathing,
thermoregulation, sudden infant death syndrome, sleep
ABBREVIATIONS
AS—active sleep
BS—bed-sharing
CI—confidence interval
CO2—carbon dioxide
CS—cot-sleeping
HR—heart rate
QS—quiet sleep
Sao2—arterial oxygen saturation
SIDS—sudden infant death syndrome
Dr Baddock participated in the design and coordination of the
study, conception of the behavioral data analysis, carried out the
data collection, and drafted the manuscript; Dr Galland
participated in the conception and design of the study and
helped to draft the manuscript; Dr Bolton participated in the
design of the study and helped edit the manuscript; Dr Williams
conducted the statistical analysis and helped edit the
manuscript; and Dr Taylor participated in the conception and
design of the study, obtained the funding and helped edit the
manuscript. Dr Baddock is guarantor for the study.
www.pediatrics.org/cgi/doi/10.1542/peds.2011-3390
doi:10.1542/peds.2011-3390
Accepted for publication Apr 20, 2012
Address correspondence to Sally Baddock, PhD, School of
Midwifery, Otago Polytechnic, Private Bag 1910, Dunedin 9054,
New Zealand. E-mail: sally.baddock@op.ac.nz
PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275).
Copyright © 2012 by the American Academy of Pediatrics
FINANCIAL DISCLOSURE: The authors have indicated they have
no financial relationships relevant to this article to disclose.
FUNDING: Funding was obtained from the Health Research
Council of New Zealand, University of Otago Research Grant, and
a Fanny Evans Scholarship. The funders had no role in study
design; in the collection, analysis, and interpretation of data; or
in the writing of the report or the decision to submit the article
for publication.
PEDIATRICS Volume 130, Number 2, August 2012
WHAT’S KNOWN ON THIS SUBJECT: Sudden infant death
syndrome remains the major cause of postneonatal death in
developed countries. Although infant-parent bed-sharing following
antenatal smoking or maternal consumption of alcohol on the
bed-sharing night increases the risk of death, the mechanism is
not known.
WHAT THIS STUDY ADDS: Bed-sharing infants experienced more
oxygen desaturations and episodes of carbon dioxide rebreathing
than cot-sleeping infants but showed appropriate behavioral and
physiologic responses. A deficit in these responses in vulnerable
infants could link to increased risk of sudden infant death
syndrome.
abstract
OBJECTIVES: To identify desaturation events (arterial oxygen satura-
tion [Sao2] ,90%) and rebreathing events (inspired carbon dioxide
(CO2) .3%), in bed-sharing (BS) versus cot-sleeping (CS) infants.
METHODS: Forty healthy, term infants, aged 0 to 6 months who regu-
larly bed-shared with at least 1 parent .5 hours per night and 40 age-
matched CS infants were recruited. Overnight parent and infant
behavior (via infrared video), Sao2, inspired CO2 around the infant’s
face, and body temperature were recorded during sleep at home.
RESULTS: Desaturation events were more common in BS infants (risk
ratio = 2.17 [95% confidence interval: 1.75 to 2.69]), associated partly
with the warmer microenvironment during BS. More than 70% of desat-
urations in both groups were preceded by central apnea of 5 to 10 sec-
onds with no accompanying bradycardia, usually in active sleep. Apnea
.15 seconds was rare (BS infants: 3 events; CS infants: 6 events), as
was desaturation ,80% (BS infants: 3 events; CS infants: 4 events).
Eighty episodes of rebreathing were identified from 22 BS infants and 1
CS infant, almost all preceded by head covering. During rebreathing,
Sao2 was maintained at the baseline of 97.6%.
CONCLUSIONS: BS infants experienced more oxygen desaturations pre-
ceded by central apnea, partly related to the warmer microenviron-
ment. Rebreathing occurred mainly during bed-sharing. Infants
were at low risk of sudden infant death syndrome and maintained nor-
mal oxygenation. The effect of repeated exposure to oxygen desatura-
tion in vulnerable infants is unknown as is the ability of vulnerable
infants to respond effectively to rebreathing caused by head covering.
Pediatrics 2012;130:237–244
237
Bed-sharing (BS) has special significance
in many cultures and is widely practiced
for perceived benefits in parent/infant
bonding, encouraging breastfeeding,
and reducing maternal sleep disrup-
tion1–3 and infant stress.4 The proximity
of mother and infant facilitates increased
interactions, waking, and breastfeed-
ing.5–9 Similar benefits also occur dur-
ing early skin-to-skin contact.10 However,
many factors are associated with an in-
creased risk of sudden infant death syn-
drome (SIDS) during infant–parent BS,
including maternal smoking during pre-
gnancy,11 parental consumption of alcohol
or sedative drugs,12 excessive parental
tiredness,13 sleeping on a couch,12,14
young infant age,15 and excess bed-
ding.13 In this context, it is important to
determine behavioral and environmen-
tal factors that affect infant physiology
and that could increase the risk of SIDS.
The presence of adults in bed, as well as
the use of adult bedding (eg, heavy du-
vets), expose infants to a different envi-
ronment compared with sleeping alone in
a cot. A component of this is a warmer
microenvironment, thus raising periph-
eral temperature of BS infants.16 The
rectal–shin temperature difference, an
indicator of thermal comfort,17 is smaller
in BS infants, indicating a thermolytic
state, whereas in cot-sleeping (CS) in-
fants it is larger, indicating a state of
thermogenesis.16 Patterns of thermo-
regulation, sleep state, and breathing
are tightly linked,18 and elevated body
temperatures (hyperthermia) can trig-
ger periodic breathing in infants.19
Most studies of infant sleep physiology
are of solitary sleeping infants, with few
studies describing infants’ responses to
the BS environment. Polysomnographic
studies of infants at low risk of SIDS
report decreased time in quiet sleep
(QS)20 and increases in arousals,21 peri-
odic breathing,22 inspired carbon dioxide
(CO2),23 heart rate (HR),24 peripheral16
and axillary temperature,25 and synchro-
nicity of mother–infant sleep cycles26
238 BADDOCK et al
during BS. The aim of this study was to
investigate the usual sleep practices of
infants in their own home to identify
and compare desaturation events (ar-
terial oxygen saturation [Sao2] ,90%),
and rebreathing episodes (inspired
CO2 .3%) between BS infants and CS
infants that could improve under-
standing of the impact of the different
environments of BS and CS on infant
physiology.
METHODS
Participants
The power calculation was based on
a study identifying the frequency of
high-risk behavior in CS infants.27 From
this data, we predicted 50% of BS
infants and 20% of CS infants were
likely to experience a potentially dan-
gerous event. By using a binomial dis-
tribution, it was suggested that 40 to 50
infants were needed to show this dif-
ference at the 5% level of significance
with 80% power.
Forty infants who regularly slept in the
parental bed with 1 or both parents for
a minimum of 5 hours per night (BS)
and 40 age-matched CS infants who
slept in a cot or bassinette for at least
5 hours per night were studied, as
previously described.5 Briefly, BS fami-
lies were recruited through local post-
natal groups and media advertising. CS
infants matched for age and season of
study were recruited from the postnatal
ward of a local hospital. Infants were
aged 0 to 6 months and were .37
weeks’ gestation (except 2 infants in
each group who were 28 and 32 weeks’
gestation, respectively).
The study was approved by the South-
ern Regional Health Authority Ethics
Committee, New Zealand (protocol 97/
04/036). Informed consent was ob-
tained from the parents.
Study Protocol
Infants were monitored over 2 con-
secutive nights during their usual sleep
practice (bed-share or cot-sleep) in
their home. The details of recordings
have been described previously.16 Briefly,
the first night involved video recording
and the second, both video and physio-
logic recording. The latter data included
electrocardiogram, Sao2, and HR (Nellcor
N-200, Nellcor, Inc, Hayward, CA), re-
spiratory pattern via chest and ab-
dominal bands (Respitrace model 150;
Respitrace Co, NY), airflow via nasal
thermistors, and shin and rectal tem-
perature. The level of CO2 in the mi-
croenvironment around the infant’s
face was measured (NORMOCAP 200
OXY, Datex Instrumentation Corp, Helsinki,
Finland) from air sampled through a
catheter attached to midway between
the inner canthus of the infant’s eye
and the lateral edge of the nostrils.
All leads were secured to allow moth-
ers to handle infants freely during
the night. All physiologic signals from
the recording units were relayed
through a portable polygraphic system
(BabyLog Link; Christchurch Hospital,
New Zealand) and stored on a laptop
computer. A small surveillance camera
(Panasonic CEC-C38, Osaka, Japan) and
infrared light source allowed record-
ings on a video recorder (Panasonic AG-
TL700, Osaka, Japan), set to “long play”
(to enable 15 hours of recording on a
3-hour videotape), of the infant’s sleep
position, movements, and parent–infant
interactions, as previously described.5
Mothers completed a questionnaire pro-
viding demographic information.
Data Analyses
Desaturation Events
A desaturation event was defined as
Sao2 ,90%28,29 lasting for $1 second.
The events were reviewed in association
with electrocardiogram, respiratory
pattern, airflow, and inspired CO2
data, and those attributed to move-
ment artifact were discarded. For each
desaturation event, the following mea-
sures were taken: the Sao2 nadir and

duration, simultaneous HR (from elec-
trocardiogram), and inspired CO2. Apneas
associated with desaturation events
were identified and the duration (from
the end of the preceding inspiration
until the end of the next inspiration)
measured. Apnea was described as
central if there was a flattened trace
($3 seconds) for both chest and ab-
dominal movement as well as nasal
airflow, and as obstructive if respiratory
effort was evident without nasal airflow.
For the 30 seconds preceding the desa-
turation event, the mean HR, sleep state,
and rectal and shin temperatures were
measured. QS was indicated by periods
of regular breathing accompanied by
low HR variation, while active sleep
(AS) was indicated by periods of ir-
regular breathing and HR.18
Rebreathing Events
Sections of recording in which CO2 was
.3% were identified (above the level
known to stimulate ventilation30). The
rebreathe event duration encompassed
the rise and fall of that event. The dura-
tion of each rebreathe event, and mean
and maximum CO2 levels, were mea-
sured. A free of movement artifact 30-
second sample free of movement artifact
that included the maximum level of CO2
was identified, and mean values were
calculated for Sao2, HR, respiration rate,
and CO2. A 30-second baseline CO2 period
was identified in the same behavioral
sleep state preceding the rebreathe event.
Co2 Cumulative Exposure
TheBabyLogsoftwarewasusedtoidentify
all episodes in which CO2 averaged over
5 seconds exceeded either 2%, 3%, or 4%
for a minimum of 3 seconds. Episodes
during awake time or due to movement
artifact were deleted, and a cumulative
time was calculated for each level of CO2.
Statistical Analyses
Although BS and CS infants were
matched for age and season of study,
PEDIATRICS Volume 130, Number 2, August 2012
data for both members of the pair were
not always available. The data were
therefore analyzed as 2 groups, and
a form of regression analysis, adjusted
for infant age and season, was used to
take the matching into account for all
comparisons. Desaturation events were
analyzed by using Poisson or logistic
regression to estimate the relative risk
or odds ratio for the BS group for the
variables based on counts or categories,
with the CS group used as the reference.
All comparisons were adjusted for infant
age, season of study, and multiple events
per infant but not for breastfeeding
because all BS infants and 35 of the 40 CS
infants were breastfed. Almost all re-
breathe events were observed in the BS
infants, and thus a group comparison
was not possible. All rebreathe events
were analyzed as 1 group. Regression
analysis, adjusting for multiple events
per infant, was used to investigate
changes in HR, breathing rate, and SaO2
associated with rebreathe events. Sep-
arate analyses of infants of smoking
mothers was not possible due to the
small number of smokers.
RESULTS
BS infants were aged 0 to 6 months, and
CS infants were matched for age and
season of study. There were no signif-
icant differences between groups with
respect to mean 6 SD for the following
characteristics: infant study weight
(BS infants: 6580 6 1301 g; CS infants:
6594 6 1265 g), gestation (BS infants:
39.8 6 2.7 weeks; CS infants: 39.5 6 2.7
weeks), birth weight (BS infants: 3583 6
735 g; CS infants: 3562 6 671 g), or
maternal age (BS infants: 28 6 5.6
years; CS infants: 30 6 5.5 years). The
ratio of the number of males compared
to females was 0.6 for BS infants, 0.55
for CS infants. All mothers had some
tertiary education, most infants were
breastfed (BS infants: 100%; CS infants:
88%), and maternal smoking in the sec-
ond trimester was less common among
ARTICLE
BS mothers (8% vs 25%) than among
CS mothers. The majority of mothers
were of European descent, with 10%
of BS mothers and 18% of CS mothers
identifying as Maori (indigenous New
Zealanders). Maternal alcohol consump-
tion was minimal in both groups, ranging
from rarely to 3 glasses of wine or beer
per week; 17 of 40 BS mothers and 15 of
40 CS mothers reported no alcohol
consumption during pregnancy.
Desaturation Events
There were more desaturation events in
BS infants: 255 events were identified
from 39 BS infants and 123 events from
40 CS infants. There was no difference
on logistic regression between BS and
CS infants with respect to the charac-
teristics of events. Most events (85% in
each group of infants) comprised a
drop in Sao2 to between 90% and 85%.
A drop to ,80% was rare (BS infants:
3 events; CS infants: 4 events). Duration
of events ranged from 2 to 16 seconds,
with 70% lasting ,5 seconds. A central
apnea of 5 to 10 seconds preceded 70%
of desaturation events. Few apneas of
.15 seconds were recorded (3 BS in-
fants; 6 CS infants). Events mainly oc-
curred in AS in both groups (AS: BS 61%
and CS 87%; QS: BS 39% and CS 13%)
with CS infants less likely than BS in-
fants to begin an event in QS (odds ratio:
0.2 [95% confidence interval [CI]: 0.05–
0.79]). CIs are large due to the effect of
multiple observations per infant.
Table 1 shows that the estimated mean
number of desaturation events (per
infant per night) were 6.8 for BS infants
and 3.1 for CS infants, giving a relative
risk of 2.17 (adjusted for sleep time).
The ratio was practically the same
(2.15) in the 51 infants for whom tem-
perature data were available (the lesser
numbers due to parents’ reluctance
to have infant rectal temperature mea-
sured). Adjusting for the mean overnight
rectal–shin temperature difference re-
duced the risk ratio from 2.15 to 1.54,
239
TABLE 1 Number of Desaturation Events per Night in BS Infants and CS Infants, Including the TABLE 3 Characteristics of Rebreathe
Effect of Overnight Temperature Events
Event No. of Infants Estimated Group Mean Risk Ratio P Characteristic Median Range
(Range) (95% CI)
Event duration, min 4 1–59
BS CS Maximum event CO2, % 4 3–5.7
Mean event CO2, % 2.5 1.7–4.7
No. of desaturation events/night Total event duration/infant, min 19 3–154
All 79 6.8 (0–44) 3.1 (0–16) 2.17 (1.75–2.69) ,.001
n = 23 infants, n = 80 observations.
With temperature data 51 8.5 (0–44) 3.8 (0–16) 2.15 (1.69–2.73) ,.001
Adjusted for rectal–shin 51 7.7 (0–44) 3.6 (0–16) 1.54 (1.17–2.02) .002
temperature difference CO2 Cumulative Exposure
Adjusted for age, season, and multiple events per infant.
The cumulative time infants were ex-
posed to .2%, .3%, or .4% CO2 is
indicating that a large part of the dif- event), a decrease to ,70% of base- illustrated in Fig 2. Twenty-seven BS
ference in desaturation events per line occurred for 1 BS event. HR did not infants were exposed to at least 2%
night was associated with the warmer drop to ,90 beats/min during any CO2 compared with only 2 CS infants,
temperature of the BS infants (smaller desaturation. and similarly marked differences were
rectal–shin temperature difference). seen at levels .3%; 1 BS infant was
Furthermore, we calculated that for a 1 Rebreathing Events exposed to .4% CO2 for 4 minutes and
degree decrease in rectal–shin temper- 1 CS infant for 13 minutes.
Eighty rebreathe events were identified
ature difference (ie, warmer microenvi- from 22 BS infants and 1 CS infant
ronment), desaturation events increased DISCUSSION
(Tables 3 and 4). Although the median
by 60% (95% CI: 31–96). duration of an event was 4 minutes, the This study found that infants who bed-
The desaturation events were not as- maximum lasted 59 minutes. The mean share experienced more episodes of
sociated with elevated inhaled CO2 or level of CO2 varied from 1.7% to 4.7%. mild oxygen desaturation associated
with significant changes in HR. The The median total duration of rebreathing with the warmer microenvironment
level of CO2 30 seconds before desatu- per infant per night was 19 minutes, and responded more frequently to re-
ration events was ,0.5% for .70% of but 1 infant (who spent most of the breathe stimuli compared with infants
events in each group (Table 2). A small night with head covered by over- who slept in a cot. More severe desat-
change in HR of 630% from baseline bedding) reached 154 minutes. Mean uration events (,80%) and longer
occurred during the majority of desat- respiration rate increased from 32 apneic events (.15 seconds) were
uration events (HR increased in 175 of to 38 breaths per minute (P , .001), rare but occurred in both groups of
378 events and decreased in 188 of 378 and mean SaO2 was maintained at infants. Infants in this study were
events), whereas an increase of .30% the baseline level of 97.6%, while heart healthy and, based on demographic
from baseline occurred in 4 events rate increased slightly (126 to 129 characteristics, at low risk for SIDS,
(BS infants: 3 events; CS infants: 1 beats/min). and they responded appropriately to
The majority of events (70%) were as- the potential stressors. Epidemiologic
TABLE 2 No. (%) of Desaturation Events sociated with head covering. Of the 5 BS evidence identifies an increased risk
Associated With Varying Levels of associated with BS for vulnerable in-
CO2 at the Infant’s Face infants who spent some time sleeping
prone, events occurred only in 1 while fants, and we hypothesize that an infant
CO2 (%) BS (n = 30) CS (n = 24)a P
face-down into a pillow and another who does not adjust to the microenvi-
0–0.4 200 (78) 84 (71) — ronment created during BS could be at
0.5–0.9 39 (15) 25 (21) — with face into the mother’s chest. Events
1.0–1.4b 9 (4) 8 (7) .79 also occurred in 1 side-sleeping infant increased risk.
1.5–1.9b 2 (1) 1 (1) — with face into a tri-pillow, and in 1 in-
2–2.9b 5 (2) 0 — Rebreathing
.3 0 0 —
fant while the mother breathed directly
Total no. 255 118 — over the infant’s face. Rebreathing BS infants were exposed to more re-
of events events were recorded in 1 CS infant. breathe events than CS infants, com-
Multiple logistic regression adjusted for season, age, and This infant was swaddled in a muslin monly associated with head covering by
multiple events per infant was used for comparison.
a CO data missing for 5 CS desaturation events. No desa-
2
wrap and repeatedly pushed the wrap bedding or mothers’ clothing (during
turation events were identified in the remaining 9 BS over the face (Fig 1). Two CS infants supine or side positioning), and less
infants and 16 CS infants.
b Categories combined into 1 category for analysis due to slept prone for the entire night, but no frequently by the infant’s face into a pil-
low numbers. rebreathing events were identified. low (during prone or side positioning).

240 BADDOCK et al

TABLE 4 Baseline and Rebreathing Events: Mean 6 SD Values for SaO2, Breathing, and HR
Event Baseline
CO2 = 0.5%
SaO2, % 97.6 6 1.1
Breaths per minute 32 6 7
HR, beats/min 126 6 12
n = 23 infants, n = 80 observations. Data are presented as mean 6 SD values for 30-second periods during baseline
conditions and 30-second periods during rebreathing events.
FIGURE 1
CS infant with head covered by muslin wrap.
FIGURE 2
Percentage of infants exposed to varying levels of CO2 over the night’s recording. Numbers in pa-
rentheses represent the range of minutes of duration of exposure to the corresponding level of Co2.
Ball31 has also reported more head
covering. The basis for the rebreathe
events may be the head covering per
se32–35 or the mother breathing over the
infant.23 In response, infants may either
increase ventilation effectively33 or re-
move themselves from the rebreathe
situation or prompt action by the
mother.36 Mothers also spontaneously
remove bedding during head covering
events,36 indicating the importance of
a responsive mother during BS. Re-
breathing may be hazardous due to the
potential asphyxial microenvironment
PEDIATRICS Volume 130, Number 2, August 2012
Rebreathing Event, Difference 95% CI
CO2 = 3%
97.6 6 1.2 0.06 6 0.9 –0.2 to 0.3
38 6 8 5.6 6 6.0 4 to 7
129 6 11 2.9 6 6.5 1 to 5
or the cumulative impact of repeated
exposure to a mildly hypoxic environ-
ment,37 although, conversely, low levels
of CO2 may be protective because of the
stimulatory effect of the CO2 on venti-
lation.23 Rebreathe events were iden-
tified in only 1 CS infant, who was
swaddled in a muslin wrap and who
pulled the wrap over her face several
times. Furthermore, there were some
periods when the head could be cov-
ered but with relatively little accumu-
lation of CO2, most likely explained by
the existence of air channels between
ARTICLE
the bedding and the infant’s face.38
P Infants often remained head covered
for extended periods of time (up to
.637 59 minutes) and maintained oxygen
,.001
saturation.
.003
We and others have shown that in-
creased head covering in BS infants
compared with CS infants is due to in-
creased mobility of the bedding.31,36 We
demonstrated here that infants main-
tained their SaO2, in part due to an in-
creased rate of breathing, as also
shown by Ball.31 The ventilatory re-
sponse to asphyxia is mainly an in-
crease in tidal volume,38 but we could
not measure this because we used an
uncalibrated respitrace system. Our
data indicate that normal homeostatic
respiratory responses to the re-
breathe environment protected these
infants.
However, infants with hypoxemia re-
sulting from an infection may not re-
spond as well,39 and whether the
degree of chronic hypoxia that some BS
infants are exposed to is sufficient
to blunt ventilatory responses is un-
known. The vulnerability of the non-
responsive infant may lie in intrinsic
abnormalities of formation, or imma-
turity of development of certain brain
regions that subserve respiratory and
cardiovascular control,40–43 which may
be due to or aggravated by external
factors such as maternal smoking
during pregnancy44,45 or prematurity.
Desaturation Events
Desaturation events to ,90% were
seen more often in BS infants than in
CS infants, although the importance of
this event as a risk factor for SIDS is
unclear. Episodic apnea46 and periodic
breathing47 occur in normal infants
and are not observed more frequently
in future SIDS victims,48 albeit that an
increase in obstructive apneas has
been reported.48 We note that almost
all of the desaturation events in our study
were preceded by central apnea and
241
were accompanied by the observation
of a warmer microenvironment, as
similarly observed by Richard et al.22
Oxygen saturation levels of ,80% were
not common and occurred in 2% of
the infants (3 events in BS infants with
4 events in CS infants). Even in these
more severe events, desaturation was
not associated with bradycardia. The
lowest HR recorded in any event was
90 beats/min. This finding is consis-
tent with other studies of healthy
infants, and apnea without bradycar-
dia is considered a normal event.47,49,50
Richard and Mosko24 have reported
a higher HR with less HR variability in
BS infants compared with CS infants,
partly explained by the higher axillary
temperature. It is of interest that ani-
mal studies have shown that repeated
episodes of hypoxia blunt arousal
responses.37 We suggest that frequent
desaturations in vulnerable BS infants
could be a contributory factor in their
risk for SIDS.
Thermal Microenvironment
Thick bedding on adult beds raises
the in-bed temperature while infant–
mother contact reduces the surface
area available for heat loss,51 leading
to increased infant peripheral tem-
perature.16,52 Warmer temperatures
can trigger periodic breathing in pre-
mature infants at near term age19 but
not in very premature infants at 5 days
of age.53 Body heat loss (and thus en-
ergy expenditure) rather than body
temperature is also associated with
increased frequency and duration of
apneic events in premature infants.54
Skin-to-skin contact in premature in-
fants has been shown to have a greater
effect on bradycardia and desatura-
tions than hyperthermic conditions
induced by elevated incubator tem-
perature.55 In that study, skin-to-skin
contact did not raise core tempera-
ture, but the semi-upright infant posi-
tion during skin-to-skin contact was
suggested to contribute to the unstable
242 BADDOCK et al
respiration. Whether this also occurs
in older infants is unknown, but it is
probably less of a concern in healthy
infants with maturation of cardiore-
spiratory control.
Sleep Architecture
BS infants have less QS than CS infants,
possibly affording protection due to
increased arousability in AS.20 However,
depression of the hypoxic ventilator
response occurs in AS,56 and desatu-
ration events were more common in
our study in AS. Interestingly, BS infants
had a larger percentage of desatura-
tion events in QS than did CS infants.
Richard et al22 reported increased
apneas in QS in BS infants associated
with elevated axillary temperature in
QS only, suggesting a link between
desaturations, temperature, and sleep
state.
Sleep Position
Prone sleeping infants up to 5 months of
age and supine infants at 2.5 months are
very limited in their ability to remove
blankets from over their head.57 In this
study, few infants were placed prone to
sleep in either group, but the micro-
environment could be similar to BS in
terms of the warmer microenviron-
ment and potential for rebreathing.34,58
Increased arousals21 may not be suffi-
cient to protect the infant if the head is
also covered by bedding. We have pre-
viously reported if the infant’s head is
covered, the mother is involved in
uncovering of the head ∼60% of the
time, either initiating action herself or
after prompting by the infant.36 This
suggests a mechanism for the risk of
BS with a mother impaired by alcohol
or drugs12 who may be unable to re-
spond to her infant.
Strengths
We had a large sample size, and mon-
itored 40 BS infants and 40 CS infants,
producing ∼10 hours of overnight
data per study on family behavior,
environmental factors, and infant phys-
iology. Although the sample size for
temperature analysis was reduced to
51 infants, it remained considerably
larger than similar published stud-
ies.12,8,31,59 We monitored infants in
their home, where families might be
more likely to carry out usual nighttime
practices. Our design allowed obser-
vation of infants in their usual sleep
arrangement rather than a crossover
design that requires infants to sleep
both in their usual arrangement and
an assigned arrangement.7,8,31 There
is evidence to indicate differences in
behavior between regular bed-sharers
and occasional bed-sharers.21 Lastly,
we have reported previously that
the sensors did not affect infant
behavior.16
Limitations
The sample population was largely
at low risk of SIDS. The low number
of maternal smokers among the bed-
sharers, together with the high stan-
dard of maternal education and high
breastfeeding rate, suggests our sam-
ple did not cover the full spectrum of
bed-sharers in which social depriva-
tion, low maternal education, smoking,
and lack of breastfeeding are other key
factors in sudden deaths associated
with BS.60 The main reasons for BS in
our group were related more to par-
enting style5 rather than to cultural or
economic influences.
CONCLUSIONS
The presence of the mother and other
bed-partners, and the physical envi-
ronment of the adult bed clearly led to
a different sleep environment for the BS
infant compared with the CS infant,
resulting in beneficial and potentially
compromising situations. Infant ho-
meostatic responses and frequent ma-
ternal interactions seemed to keep
these low-risk infants safe. However, we

suggest that it is potentially hazardous
for an infant to sleep in the same bed as
their parent, if the infant and/or mother
are unresponsive. We acknowledge that
BS is a practice valued by many; thus, it
is important to identify the specific dan-
gers related to this practice. Keeping the
infant smoke-free at all times, including
REFERENCES
1. Abel S, Park J, Tipene-Leach D, Finau S,
Lennan M. Infant care practices in New
Zealand: a cross-cultural qualitative study.
Soc Sci Med. 2001;53(9):1135–1148
2. Tipene-Leach D, Abel S, Finau SA, Park J,
Lenna M. Maori infant care practices: im-
plications for health messages, infant care
services and SIDS prevention in Maori com-
munities. Pac Health Dialog. 2000;7(1):29–37
3. Ball HL. Reasons to bed-share: why parents
sleep with their infants. J Reprod Infant
Psychol. 2002;20(4):207–222
4. Morgan BE, Horn AR, Bergman NJ. Should
neonates sleep alone? Biol Psychiatry.
2011;70(9):817–825
5. Baddock SA, Galland BC, Bolton DP, Williams
SM, Taylor BJ. Differences in infant and
parent behaviors during routine bed shar-
ing compared with cot sleeping in the home
setting. Pediatrics. 2006;117(5):1599–1607
6. Ball HL. Breastfeeding, bed-sharing, and
infant sleep. Birth. 2003;30(3):181–188
7. McKenna JJ, Mosko SS, Richard CA. Bed-
sharing promotes breastfeeding. Pediat-
rics. 1997;100(2 pt 1):214–219
8. Pollard K, Fleming P, Young J, Sawczenko A,
Blair P. Night-time non-nutritive sucking in
infants aged 1 to 5 months: relationship
with infant state, breastfeeding, and bed-
sharing versus room-sharing. Early Hum
Dev. 1999;56(2–3):185–204
9. Gettler LT, McKenna JJ. Evolutionary per-
spectives on mother-infant sleep proximity
and breastfeeding in a laboratory setting.
Am J Phys Anthropol. 2011;144(3):454–462
10. Moore ER, Anderson GC, Bergman N. Early
skin-to-skin contact for mothers and their
healthy newborn infants. Cochrane Data-
base Syst Rev. 2007;(3):CD003519
11. Scragg R, Mitchell EA, Taylor BJ, et al; New
Zealand Cot Death Study Group. Bed shar-
ing, smoking, and alcohol in the sudden
infant death syndrome. BMJ. 1993;307(6915):
1312–1318
12. Blair PS, Sidebotham P, Evason-Coombe C,
Edmonds M, Heckstall-Smith EM, Fleming P.
PEDIATRICS Volume 130, Number 2, August 2012
during pregnancy, is perhaps the most
important message, as well as avoiding
consumption of alcohol or sedating
drugs on the BS night; keeping the
infant’s face clear; and avoiding the use
of thick insulation as overbedding.
Studies with high-risk infants are re-
quired to advance understanding of
Hazardous cosleeping environments and
risk factors amenable to change: case-
control study of SIDS in south west England.
BMJ. 2009;339:b3666
13. Blair PS, Fleming PJ, Smith IJ, et al. Babies
sleeping with parents: case-control study of
factors influencing the risk of the sudden
infant death syndrome. CESDI SUDI research
group. BMJ. 1999;319(7223):1457–1461
14. Tappin D, Ecob R, Brooke H. Bedsharing,
roomsharing, and sudden infant death
syndrome in Scotland: a case-control study.
J Pediatr. 2005;147(1):32–37
15. Carpenter RG, Irgens LM, Blair PS, et al.
Sudden unexplained infant death in 20 re-
gions in Europe: case control study. Lancet.
2004;363(9404):185–191
16. Baddock SA, Galland BC, Beckers MG, Taylor
BJ, Bolton DP. Bed-sharing and the infant’s
thermal environment in the home setting.
Arch Dis Child. 2004;89(12):1111–1116
17. Clark RP, Edholm OG. Man and His Thermal
Environment. London, UK: Edward Arnold
Ltd; 1985
18. Tappin DM, Ford RP, Nelson KP, et al.
Breathing, sleep state, and rectal temper-
ature oscillations. Arch Dis Child. 1996;74
(5):427–431
19. Berterottière D, D’Allest AM, Dehan M,
Gaultier C. Effects of increase in body
temperature on the breathing pattern in
premature infants. J Dev Physiol. 1990;13
(6):303–308
20. Mosko S, Richard C, McKenna J, Drummond
S. Infant sleep architecture during bed-
sharing and possible implications for SIDS.
Sleep. 1996;19(9):677–684
21. Mosko S, Richard C, McKenna J. Infant
arousals during mother-infant bed sharing:
implications for infant sleep and sudden
infant death syndrome research. Pediat-
rics. 1997;100(5):841–849
22. Richard CA, Mosko SS, McKenna JJ. Apnea
and periodic breathing in bed-sharing and
solitary sleeping infants. J Appl Physiol.
1998;84(4):1374–1380
ARTICLE
the specific mechanism(s) leading to
their increased vulnerability.
ACKNOWLEDGMENTS
We thank the families that participated
in this study and Charrissa Makowhar-
emahihi and Amanda Phillips for re-
search assistance.
23. Mosko S, Richard C, McKenna J, Drummond
S, Mukai D. Maternal proximity and infant
CO2 environment during bedsharing and
possible implications for SIDS research.
Am J Phys Anthropol. 1997;103(3):315–328
24. Richard CA, Mosko SS. Mother-infant bed-
sharing is associated with an increase in
infant heart rate. Sleep. 2004;27(3):507–511
25. Richard CA. Increased infant axillary tem-
peratures in non-REM sleep during mother-
infant bed-sharing. Early Hum Dev. 1999;55
(2):103–111
26. McKenna JJ, Mosko SS. Sleep and arousal,
synchrony and independence, among moth-
ers and infants sleeping apart and together
(same bed): an experiment in evolutionary
medicine. Acta Paediatr Suppl. 1994;397:
94–102
27. Waters KA, Gonzalez A, Jean C, Morielli A,
Brouillette RT. Face-straight-down and face-
near-straight-down positions in healthy,
prone-sleeping infants. J Pediatr. 1996;128
(5 pt 1):616–625
28. Brouillette RT, Morielli A, Leimanis A, Waters
KA, Luciano R, Ducharme FM. Nocturnal
pulse oximetry as an abbreviated testing
modality for pediatric obstructive sleep ap-
nea. Pediatrics. 2000;105(2):405–412
29. Urschitz MS, Wolff J, Von Einem V, Urschitz-
Duprat PM, Schlaud M, Poets CF. Reference
values for nocturnal home pulse oximetry
during sleep in primary school children.
Chest. 2003;123(1):96–101
30. Haddad GG, Leistner HL, Epstein RA, Epstein
MA, Grodin WK, Mellins RB. CO2-induced
changes in ventilation and ventilatory pat-
tern in normal sleeping infants. J Appl
Physiol. 1980;48(4):684–688
31. Ball H. Airway covering during bed-sharing.
Child Care Health Dev. 2009;35(5):728–737
32. Campbell AJ, Bolton DP, Williams SM, Taylor
BJ. A potential danger of bedclothes cov-
ering the face. Acta Paediatr. 1996;85(3):
281–284
33. Chiodini BA, Thach BT. Impaired ventilation
in infants sleeping facedown: potential
243
 significance for sudden infant death syn-
drome. J Pediatr. 1993;123(5):686–692
34. Bolton DP, Taylor BJ, Campbell AJ, Galland
BC, Cresswell C. Rebreathing expired gases
from bedding: a cause of cot death? Arch
Dis Child. 1993;69(2):187–190
35. Paluszynska DA, Harris KA, Thach BT. In-
fluence of sleep position experience on
ability of prone-sleeping infants to escape
from asphyxiating microenvironments by
changing head position. Pediatrics. 2004;
114(6):1634–1639
36. Baddock SA, Galland BC, Taylor BJ, Bolton
DPG. Sleep arrangements and behavior of
bed-sharing families in the home setting.
Pediatrics. 2007;119(1). Available at: www.
pediatrics.org/cgi/content/full/119/1/e200
37. Johnston RV, Grant DA, Wilkinson MH, Walker
AM. Repetitive hypoxia rapidly depresses
arousal from active sleep in newborn lambs.
J Physiol. 1998;510(pt 2):651–659
38. Patel AL, Harris K, Thach BT. Inspired CO(2)
and O(2) in sleeping infants rebreathing
from bedding: relevance for sudden infant
death syndrome. J Appl Physiol. 2001;91(6):
2537–2545
39. Parkins KJ, Poets CF, O’Brien LM, Stebbens
VA, Southall DP. Effect of exposure to 15%
oxygen on breathing patterns and oxygen
saturation in infants: interventional study.
BMJ. 1998;316(7135):887–891
40. Duncan JR, Paterson DS, Hoffman JM, et al.
Brainstem serotonergic deficiency in sud-
den infant death syndrome. JAMA. 2010;303
(5):430–437
41. Broadbelt KG, Barger MA, Paterson DS,
et al. Serotonin-related FEV gene variant in
the sudden infant death syndrome is
a common polymorphism in the African-
American population. Pediatr Res. 2009;66
(6):631–635
42. Filiano JJ, Kinney HC. A perspective on neu-
ropathologic findings in victims of the sud-
den infant death syndrome: the triple-risk
model. Biol Neonate. 1994;65(3–4):194–197
244 BADDOCK et al
43. Paterson DS, Trachtenberg FL, Thompson
EG, et al. Multiple serotonergic brainstem
abnormalities in sudden infant death syn-
drome. JAMA. 2006;296(17):2124–2132
44. Richardson HL, Walker AM, Horne RSC.
Maternal smoking impairs arousal pat-
terns in sleeping infants. Sleep. 2009;32(4):
515–521
45. Campbell AJ, Galland BC, Bolton DP, Taylor
BJ, Sayers RM, Williams SM. Ventilatory
responses to rebreathing in infants ex-
posed to maternal smoking. Acta Paediatr.
2001;90(7):793–800
46. Hoppenbrouwers T, Hodgman JE, Arakawa
K, Harper R, Sterman MB. Respiration
during the first six months of life in normal
infants. III. Computer identification of breath-
ing pauses. Pediatr Res. 1980;14(11):1230–
1233
47. Horemuzova E, Katz-Salamon M, Milerad J.
Breathing patterns, oxygen and carbon di-
oxide levels in sleeping healthy infants
during the first nine months after birth.
Acta Paediatr. 2000;89(11):1284–1289
48. Franco P, Szliwowski H, Dramaix M, Kahn A.
Decreased autonomic responses to ob-
structive sleep events in future victims of
sudden infant death syndrome. Pediatr
Res. 1999;46(1):33–39
49. Hunt CE, Corwin MJ, Weese-Mayer DE, et al;
Collaborative Home Infant Monitoring Eval-
uation (CHIME) Study Group. Longitudinal
assessment of hemoglobin oxygen satura-
tion in preterm and term infants in the first
six months of life. J Pediatr. 2011;159(3):
377–383.e1
50. Ramanathan R, Corwin MJ, Hunt CE, et al;
Collaborative Home Infant Monitoring Eval-
uation (CHIME) Study Group. Cardiorespi-
ratory events recorded on home monitors:
comparison of healthy infants with those at
increased risk for SIDS. JAMA. 2001;285
(17):2199–2207
51. Karlsson H. Skin to skin care:heat balance.
Arch Dis Child Fetal Neonatal Ed. 1996;75
(2):F130–F132
52. Bystrova K, Matthiesen AS, Vorontsov I,
Widström AM, Ransjö-Arvidson AB, Uvnäs-
Moberg K. Maternal axillar and breast
temperature after giving birth: effects of
delivery ward practices and relation to
infant temperature. Birth. 2007;34(4):291–
300
53. Rieger-Fackeldey E, Schaller-Bals S, Schulze
A. Effect of body temperature on the pat-
tern of spontaneous breathing in extremely
low birth weight infants supported by
proportional assist ventilation. Pediatr Res.
2003;54(3):332–336
54. Tourneux P, Cardot V, Museux N, et al. In-
fluence of thermal drive on central sleep
apnea in the preterm neonate. Sleep. 2008;
31(4):549–556
55. Bohnhorst B, Gill D, Dördelmann M, Peter
CS, Poets CF. Bradycardia and desaturation
during skin-to-skin care: no relationship to
hyperthermia. J Pediatr. 2004;145(4):499–
502
56. Horne RS, Parslow PM, Harding R. Postnatal
development of ventilatory and arousal
responses to hypoxia in human infants.
Respir Physiol Neurobiol. 2005;149(1–3):
257–271
57. Skadberg BT, Markestad T. Consequences
of getting the head covered during sleep in
infancy. Pediatrics. 1997;100(2). Available
at: www.pediatrics.org/cgi/content/full/100/
2/E6
58. Kemp JS, Thach BT. A sleep position-
dependent mechanism for infant death on
sheepskins. Am J Dis Child. 1993;147(6):
642–646
59. McKenna J, Mosko S, Richard C, et al. Ex-
perimental studies of infant-parent co-sleeping:
mutual physiological and behavioral influ-
ences and their relevance to SIDS (sudden
infant death syndrome). Early Hum Dev. 1994;
38(3):187–201
60. James C, Klenka H, Manning D. Sudden in-
fant death syndrome: bed sharing with
mothers who smoke. Arch Dis Child. 2003;
88(2):112–113