Sleep.

19(9):685-690
© 1996 American Sleep Disorders Association and Sleep Research Society

Sleeping Position, Orientation, and Proximity

in Bedsharing Infants and Mothers

Downloaded from https://academic.oup.com/sleep/article-abstract/19/9/685/2749884 by PYEONGTAEK UNIVERSITY user on 21 March 2020

Christopher Richard, Sarah Mosko, James McKenna and Sean Drummond

Sleep Disorders Center, University of California, Irvine Medical Center,

Orange, California, U.S.A.

Summary: The impact of mother-infant bedsharing on infant sleeping position, orientation, and proximity to the
mother was assessed in 12 breast-feeding Latino mother-infant pairs. Six routinely bedsharing and six routinely
solitary-sleeping pairs slept 3 nights in the sleep laboratory. The first night matched the routine home condition,
followed by 1 bedsharing night and I solitary-sleeping night in random order. During bedsharing infants were never
placed prone, regardless of their routine sleeping condition. On the bedsharing night, mothers and infants spent
most of the night oriented toward each other; seven of 12 infants remained oriented toward their mothers the entire
night. While sleeping in a face-to-face orientation, most pairs slept most of the time less than 30 em apart with
appreciable amounts of time at less than 20 em. This orientation and proximity should facilitate sensory exchanges
between mother and infant which, we hypothesize, influence the infant's sleep physiology and nocturnal behavior.
We conclude that bedsharing minimizes the use of the prone infant sleeping position, probably in part to facilitate
breast feeding. By promoting nonprone positions, bedsharing may protect some infants from sudden infant death
syndrome (SIDS), since prone sleeping is a known risk factor for SIDS. The large percentage of the night that
mothers spent oriented toward their infants suggests that a higher degree of maternal vigilance may also result from
bedsharing. Key Words: Bedsharing-Prone sleep-Infant sleep-Breast feeding-SIDS.

The prone sleeping position is thought to be a sig-
nificant risk factor for sudden infant death syndrome
(SIDS) (1,2). Although the mechanism(s) through
which prone sleeping promotes SIDS is unknown,
some of the physiological correlates of this sleeping
position are known (see 3 for review). For example,
prone sleeping has been associated with fewer arous-
als, increased sleep duration, and increased amounts of
non-REM sleep (4). In addition, Chiodini and Thach
proposed that lethal rebreathing of CO 2 accounts for
some prone-related SIDS deaths as a result of lying
face down into bedclothes (5). Several factors appear
to interact with prone sleeping to further increase SIDS
risk, e.g. use of natural fiber mattresses, swaddling,
recent illness, and overheating of the sleeping room
(6). These findings support the concept that child care
practices and environmental factors can interact to cre-
ate substantial SIDS risks. Cosleeping may be one
't, child care practice that has the potential to affect SIDS
susceptibility (7,8).
Parent-infant cosleeping is today the routine sleep-
ing arrangement for most of the world's nonindustrial-
Accepted for publication July 1996.
Address correspondence and reprint requests to Christopher Rich-
ard, Sleep Disorders Center, Building 22C Rt. 23 UCIMC, 101 The
City Drive, Orange, CA 92668, U.S.A.
685
ized cultures and for 100% of contemporary hunter-
gatherer groups (9). Based in part on these observa-
tions, cosleeping is widely accepted by anthropologists
to represent the sleep environment of pre-historic hom-
inids (10-13). Therefore, it is likely that cosleeping is
the environment within which infant ontogeny and
physiology evolved, and it is a reasonable assumption
that cosleeping is of some adaptive (i.e. selective) val-
ue to the infant. Recognition of this basic human (and
nonhuman primate) behavior has been lacking in pre-
vious physiological or behavioral studies of infant
sleep. Instead, sleep studies have been modeled exclu-
sively on the recent western cultural practice of soli-
tary sleeping, i.e. placing infants in a room alone. In-
fant cosleeping occurs with different degrees of prox-
imity to and contact with the parent, varying from bed-
sharing (sharing the same bed or sleeping surface) to
sleeping on separate surfaces in the same room. .
Of the various cosleeping environments, bedsharing
is the most different from solitary sleeping because of
the complex auditory, visual, tactile, thermal, and ol-
factory stimuli resulting from the close proximity of
the parent(s). In ongoing studies of cosleeping, we
found several physiological and behavioral conse-
quences of bedsharing. While bedsharing, infants have
less deep non-REM sleep and more light non-REM
686 C. RICHARD ET AL.
sleep than when they sleep alone, and increased syn-
chronicity in sleep stages occurs between the mother
and infant during bed sharing (14,IS). Bedsharing is
also associated with more frequent arousals, many of
which temporarily overlap arousals of the bedsharing
mother (14). Furthermore, bedsharing infants breast-
feed more often and for longer durations than solitary-
sleeping infants (16,17). We report here on the differ-
ences in the amount of prone sleeping in the infant
between bed sharing and solitary-sleeping environ-
ments.
METHODS
Subjects were part of a larger ongoing study aimed
at characterizing the environment and sleep patterns of
the bedsharing infant. Twelve mother-infant pairs that
were either routinely solitary-sleeping (RS, n = 6) or
routinely bed sharing (RB, n = 6) were recruited from
a postnatal community health clinic. The mothers were
exclusively Latina, nonsmoking, and breast-feeding
and had normal pregnancies. The infants were between
11 andAS weeks old and had normal gestations, un-
complicated deliveries and S-minute Apgar scores of
2:8. Both members of the pair were examined by a
physician and were healthy at the time of the record-
ings. The mothers were screened for sleep disorders
by an interview with a physician trained in sleep med-
icine and by the sleep recording on the first (the ad-
aptation) night of the study.
Each pair slept 3 consecutive nights in a clinical
sleep laboratory. The first night was spent in the pair's
routine condition as an adaptation night. The following
2 nights were randomly assigned to 1 night each of
bedsharing and solitary sleeping. For solitary sleep,
the infant was placed in a crib in a room adjacent to
the mother's room with both doors ajar to allow the
mother to respond to her infant ad lib and perform all
care-giving interventions. On the bedsharing night
(BN) the pair shared a hospital twin bed also used by
the mother on the solitary night (SN). All the mothers
were blind to the goals of the study and positioned
their infants in bed without instruction. Both members
of the pair went to bed and arose at their normal times.
Standard polysomnographic and infrared videotape
recordings were carried out on both members of the
pair each night; the videotape recordings provided the
data reported here. Video recordings were analyzed for
the amount of time spent in each of four body posi-
tions; i) supine, ii) prone, iii) right side, and iv) left
side. On the BN, the position data were also catego-
rized with regard to whether or not a member of a pair
was facing toward the other member of the pair.
Whenever they were face-to-face, the proximity of the
pair's faces was derived by measuring the distance on
Sleep, Vol. 19, No.9, 1996
the video screen between their nares and computing
the actual distance using an object of known size
placed in the video field near the pair's heads. A digital
clock placed in the field of view allowed computation
of times to the nearest minute. "'
Time spent in gross body movements was excluded
Downloaded from https://academic.oup.com/sleep/article-abstract/19/9/685/2749884 by PYEONGTAEK UNIVERSITY user on 21 March 2020
from the analyses since it included transitions between
sleeping positions. Minor movements involving short"
lasting limb or head repositioning were not excluded.
In addition, infant data collected on the BN before the
mother entered bed were excluded. The data for each
group were expressed as either the mean (± SEM)
number of minutes or percent of nonmovement time
(NMT) for that night and analyzed with repeated-mea-
sures ANOVA to assess main effects (routine home
condition and laboratory condition) and with nonpara-
metric Mann-Whitney U tests to evaluate differences
between routine conditions (RB on the BN vs. RS on
the SN). Significance was assigned when p < O.OS.
RESULTS
The amount of time spent in each of the four body
positions, expressed as a percent of NMT, is given in
Fig. 1. Mothers never placed their infants prone while
bedsharing, regardless of their routine (home) sleeping
condition. In contrast, three RB infants and one RS
infant were placed prone on their SN, and one addi-
tional RS infant was placed prone on the BN before
the mother got into bed; when she did enter the bed
she repositioned the infant to supine. In many cases,
infants spent most of their time in just one or two of
the position categories. Because of this infrequent re-
positioning, the variability of group means for both
mothers and infants was high, and therefore there were
no significant differences between the groups. The
most common body position for infants was clearly
the supine position, but for mothers there was more
variability in position although the prone position was
almost as rare in mothers as it was in infants.
The substantial amount of time many RB infants
spent on their left sides during the BN (Fig. 1) is main-
ly a reflection of the mothers' placement of all RB
infants on the same side of the bed (i.e. the open side,
away from the wall) during bedsharing; all RB moth-
ers began the BN by placing infants on their left side
or supine on the right side of the bed. On the BN, RS
mothers exhibited no clear preference in placing their
infants with regard to side of the bed.
For both RB and RS pairs on the BN, both members
spent large portions of NMT facing their bedsharing
partner (Table 1). More than half of the infants (7/12)
faced the mother 100% of the night; in fact, only two
infants spent less than 7S% of the BN facing their
mothers (range 23-100%). There was no difference
 BODY POSITION IN BEDSHARING INFANTS 687

Infant Sleeping Positions

90

Downloaded from https://academic.oup.com/sleep/article-abstract/19/9/685/2749884 by PYEONGTAEK UNIVERSITY user on 21 March 2020

.~ §
§
E
E
E
~

I~
o - RB-BN
~
==
== ~. JI =-
RS-BN RB-SN RS-SN
• Prone m Supine ~ Right Side Ii Left Side

FIG. 1. Summary data (mean ::':: SEM) for each of the four statistical groups in each of four sleeping positions. Notice the relatively
large amount of supine sleeping and the total lack of prone sleeping in the infants on the bedsharing night (BN). None of the four sleeping
positions was significantly affected by routine sleeping arrangements (RB vs. RS) or by laboratory condition (BN vs. SN). Abbreviations
used: RB, routine bedsharing sleepers; RS, routine solitary sleepers; SN, solitary night.

between RB and RS infants in how they were oriented mothers spent more time facing toward their infants
toward their mothers (F = 0.38). ANOVA results con- than away (F = 8.23; p < 0.05, p < 0.001, respec-
firmed that infants spent significantly more time facing tively).
their mothers than facing away (F = 17.81), regardless The last row of Table 1 gives the amount of NMT
of their routine sleeping condition (RB, p < 0.005; RS, pairs spent oriented face-to-face. For 10 of the 12
P < 0.01). Infants sleeping in the supine position also pairs, more than half of NMT was spent in face-to-
tended to remain oriented toward their mothers. Four face orientation (range 14-91%). There were no sig-
of the six RB infants spent 100% of their supine time nificant differences between the RB and RS groups for
with their heads turned toward the mother, one spent this variable.
44% of its supine time facing the mother, and only one Analysis of face-to-face proximity was carried out
spent none of its supine time in that orientation (mean in response to our observation that bedsharing pairs
74 ± 17.4% SEM). RS infants also exhibited substan- commonly slept at very close range. Face-to-face dis-
tial variability in orientation during supine sleep (range tances were grouped into four bins of 10 cm and are
18.3-100% facing the mother, mean 53.6 ± 16.2%). presented in Fig. 2. There were no significant differ-
Eleven of the 12 mothers spent more than half the ences between RB and RS pairs (F = 0.89), but there
night oriented toward their infants (range 37-98%). As were differences in the amount of time spent at the
with the infants, there was no main effect for routine four distances (F = 9.91). Pairs spent significantly
sleeping condition (F = 1.38), but both RB and RS more of their face-to-face time between 11 and 20 cm
apart as compared to :::; 10 cm (p < 0.005) or compared
to >30 cm (p < 0.01), but there was no difference
TABLE 1. Bedsharing orientation (% nonmovement between the times spent at 11-20 cm and 21-30 cm.
time) Three of the six RB infants and two of the RS infants
Routinely spent at least some time less than 10 cm from their
Routinely solitary- mothers.
bedsharing sleeping
pairs pairs
61.4 ::':: 6.0
DISCUSSION
Mother facing infant 73.7 7.3
~J Infant facing mother 87.4 9.8 81.3 ::':: 12.3 Parent-infant cosleeping is thought to represent the
I Mother and infant facing each other 68.2 8.2 55.5 ::':: 9.3
usual sleeping arrangement throughout human evolu-
Sleep, Vol. 19, No.9, 1996
688 C. RICHARD ET AL.

Face-To-Face Proximity
I
180~----~----~------~----~ '1

Downloaded from https://academic.oup.com/sleep/article-abstract/19/9/685/2749884 by PYEONGTAEK UNIVERSITY user on 21 March 2020

1
rIJ 1 I

.....
~

=
.-S=

< 11 em 11-20 em 21-30 em >30 em

RB RS
FIG. 2. Mean (± SEM) number of minutes mothers and infants spent at specific distance categories in the face-to-face orientation. There
were no significant group differences (RB vs. RS). The amounts of time spent at 11-20 cm and at 21-30 cm were not significantly different
from each other but both were significantly different from the time spent at <11 cm and at >30 cm. Abbreviations as in Fig. I.

tion, presumably adapting the infant (and parent) to
that environment (11). This raises the question of
whether the relatively recent practice (150-200 years),
in certain western industrialized societies, of placing
infants in solitary sleep arrangements could have po-
tentially deleterious effects on infant health and de-
velopment. Preliminary studies have already identified
several physiological and behavioral differences be-
tween bedsharing infants and solitary-sleeping infants
(14-18). This research is predicated, in part, on the
hypothesis that bed sharing may afford some protection
from SIDS (see 7,19) via infant responses to sensory
stimuli present in the bedsharing environment and/or
maternal behaviors that are facilitated by bedsharing.
In the current study, infants of Latina mothers were
never placed in the prone position during bedsharing
regardless of whether those mothers routinely bed-
shared at home or routinely slept apart from their in-
fants. This is an important observation since prone
sleeping has been found to be a significant risk factor
for SIDS (1,2).
The use of the prone position for infant sleep varies
widely between countries and cultures. Data compiled
from seven different geographic areas (all western-
type, industrialized societies) indicated that the inci-
dence of prone sleeping in those areas ranged from 31
to 65% in 1988 (2). Willinger et al. (20) reported that
the proportion of infants sleeping prone in the U.S.A.
was 74% prior to 1993. However, that study did not
Sleep, Vol. 19, No.9, 1996
examine the relationship between positioning and eth-
nicity which, along with the low number of subjects
in the current study, makes comparisons with the find-
ings of Willinger et al. tenuous. Farooqi et al. (21)
presented data supporting the idea that choice of infant
sleeping position may have an ethnic component.
Their study in the U.K. revealed that 12% of the
Asians born outside the u.K. and 24% of Asians born
in the U.K. chose the prone position, whereas 31 % of
white mothers placed their infants prone. Therefore,
our results may not be applicable to other ethnic
groups or societies, since our subject population was
made up exclusively of Latinos living in southern Cal-
ifornia. In addition, since breast-feeding may influence
a mother's choice of infant position, as well as her
orientation and proximity to the infant, our data may
not generalize to non-breast-feeding pairs. Interesting-
ly, 98% of the Asian mothers in Farooqi et al.'s study
put their infants to bed in the parents' room (34% in
the parents' bed), while only 65% of infants of white
mothers were placed in the mother's bedroom. This
supports the idea that cosleeping has an ethnic com-
ponent as well and that cosleeping may be correlated
with a reduction in the use of the prone position for
infants.
Since bedsharing and breast-feeding are closely as-
sociated, we hypothesize that a mother's choice of in-
fant sleep position may be driven, in part, by the sleep-
ing arrangement and the ease of breast-feeding in non-
 BODY POSITION IN BEDSHARING INFANTS
prone positions. If bedsharing/breast-feeding do con-
tribute to the mother's choice of nonprone infant
positions, then this constellation of child care practices
could be seen as providing an environment that is ben-
eficial to the infant, at least in terms of minimizing the
high-risk prone position. In contrast, it has recently
been suggested that bedsharing may instead pose an
independent risk for SIDS (8). However, recent data
from the same group indicated that bedsharing inter-
acts with maternal smoking to increase SIDS risk with-
out posing a significant independent risk (22).
Bedsharing mothers and, especially, infants spent
most of the night facing their bed partner, resulting in
substantial portions of the night in a face-to-face ori-
entation. More than half of the infants spent the entire
night facing their mothers. That the infants in this
study did not turn over on their own (with one excep-
tion) suggests that infant orientation at 3-4 months is
mainly determined by the mother. However, the ob-
servation that six of 10 infants exhibiting supine sleep
spent more than half their supine sleep facing their
mothers may indicate some degree of infant preference
for this orientation. Available data suggest that the ori-
enting of an infant toward its mother is driven, at least
in part, by the infant's response to olfactory stimuli
from the mother's breast or axilla (see below).
Face-to-face orientation may be.a factor in the fa-
cilitation of breast-feeding found to occur during bed-
sharing (16,l7). The most common orientation ob-
served in our study during bedsharing was the mother
on her side facing the baby and the infant on its side
or back facing the mother near breast level. In this
orientation, the baby can elicit breast-feeding easier
since the mother is already oriented toward the infant
and the mother can allow breast-feeding without hav-
ing to change either the infant's or her own position.
This orientation also allows mother and infant to fall
back to sleep after feeding without repositioning. Fa-
cilitation of breast-feeding, then, might be an example
of the adaptive value of bedsharing since there is ep-
idemiological evidence that breast-feeding decreases
SIDS risk (23,24), although this decrease in risk is not
universally accepted (25). Furthermore, this facilitation
serves to illustrate an association between bedsharing
and breast-feeding.
Bedsharing infants and mothers were also found to
sleep at short face-to-face distances. This close prox-
imity would promote many types of sensory exchang-
es and would increase the mother's ability to closely
monitor the infant's status. Visual, auditory, olfactory,
tactile, vestibular, and thermal stimuli to the infant
would all be enhanced by close face-to-face bedshar-
ing. Infants often awaken and elicit the mother to
breast-feed, and sensory cues from the mother may
playa role in this elicitation. Previous studies suggest
689
that olfactory stimuli, for example, can account for
some of the infant's orientation to its mother. Mac-
Farlane (26) found that infants preferentially orient to
their own mother's breast odors, and Cernoch and Por-
ter (27) reported that infants orient to their own moth-
er's axillary odors. The latter results were found only
Downloaded from https://academic.oup.com/sleep/article-abstract/19/9/685/2749884 by PYEONGTAEK UNIVERSITY user on 21 March 2020
in breast-feeding infants, which represents the feeding
behavior of our subjects. Those authors suggested that
their results may be explained by breast-feeding in-
fants spending much more time in close proximity to
the mother than do bottle-feeding infants.
While bedsharing in close face-to-face proximity,
other important sensory stimuli may be generated by
the mother's breathing onto the infant's face and head.
For example, we measured the peak expiratory CO 2
levels in air at various distances from women's nares
and found that levels ranged from 2.36% at 3 cm to
0.34% at 21 cm (28). Since CO 2 can stimulate respi-
ration and arousal, the mother's proximity to the infant
warrants investigation as a potentially important factor
in susceptibility to SIDS (29). Other stimuli inherent
in maternal breathing onto the infant that need to be
evaluated include: i) thermal stimuli, which exert dif-
ferential effects on sleep stages when applied to the
face versus the rest of the body (30); ii) tactile stimuli
from expiratory airflow on the face or head; iii) olfac-
tory stimuli from the mother's breath; and iv) auditory
stimuli which may influence infant respiratory patterns
(31,32).
In conclusion, in our sample of Latino infants, bed-
sharing minimized prone sleeping, possibly to facili-
tate breast-feeding. Since prone sleeping and the ab-
sence of breast-feeding appear to be ~gnificant risk
factors for SIDS (33), bedsharing might indirectly low-
er SIDS risk. In addition, we report that mother-infant
pairs commonly sleep facing each other at short dis-
tances, which should increase the variety and magni-
tude of sensory stimuli to the infant. Whether or not
such sensory stimuli significantly affect infant sleep
physiology or behavior, the high degree of mutual ori-
entation maintained by both members of the pair in-
dicates that they retain some degree of awareness of
the other's presence. This argues against a popular
misconception that bedsharing inherently carries a sig-
nificant risk of overlying and accidental suffocation
independent of that which might occur as a result of
intoxication or other factors rendering the adult less
responsive.
Acknowledgement: This research is funded by NIH
ROl 27482.
REFERENCES
1. Kattwinkel J, Brooks J, Myerberg D. Positioning and SIDS.
AAP task force on infant positioning and SIDS. Pediatrics
1992;89:1120-5.
Sleep, Vol. 19, No.9, 1996
690 C. RICHARD ET AL.
2. Willinger M, Hoffman HJ, Hartford RB. Infant sleep position
and risk for sudden infant death syndrome. Pediatrics 1994;93:
814-9.
3. Engleberts AC, de Jonge GA. Choice of sleeping position for
infants: possible association with cot death. Arch Dis Child
1990;65:462-7.
4. Kahn A, Groswasser J, Sottiaux M, Rebuffat E, Francot P, Dra-
maix M. Prone or supine body position and sleep characteristics
in infants. Pediatrics 1993;91:1112-5.
5. Chiodini BA, Thach BT. Impaired ventilation in infants sleeping
facedown: potential significance for sudden infant death syn-
drome. } Pediatr 1993;123:686-92.
6. Posonby A-L, Dwyer T, Gibbons LE, Cochrane JA, Wang Y-G.
Factors potentiating the risk of sudden infant death syndrome
associated with the prone position. N Engl J Med 1993;329:
377-82.
7. McKenna n. An anthropological perspective on the sudden in-
fant death syndrome (SIDS): the role of parental breathing cues
and speech breathing adaptations. Med Anthropol 1986; 10:9-
53.
8. Mitchell EA, Scragg R. Are infants sharing a bed with another
person at increased risk of sudden infant death syndrome? Sleep
1993;16:387-9.
9. Lozoff B, Brittenham G. Infant care: cache or carry. } Pediatr
1979;95:478-83.
10. Lancaster JB, Lancaster CS. Parental investment: The hominid
adaptation. In: Ortner D, ed. How humans adapt: a biocultural
odyssey. Washington, DC: Smithsonian Institution Press, 1982:
33-56.
II. Konner MJ. Evolution of human behavior development. In:
Monroe R, Whiting B, eds. Handbook of cross-cultural human
development. New York: Garland STPM Press, 1981:3-52.
12. Konner M, Worthman C. Nursing frequency, gonadal function,
and birth spacing among !Kung hunter-gatherers. Science 1980;
207:788-91.
13. Devore I, Konner M. Infancy in hunter-gatherer life: an etho-
logical perspective. In: White N, ed. Ethology and psychiatry.
Toronto: University of Toronto Press, 1974:254-68.
14. Mosko S, Richard C, McKenna J, Drummond S. Sleep and
arousals in cosleeping infants and mothers. Sleep Res 1995;24:
76.
15. Richard CA, Drummond S, Hunt L, et al. Polysomnographic
study of mother-infant cosleeping. Sleep Res 1994;24:191.
16. McKenna J, Mosko S, Richard C, et al. Experimental studies of
infant-parent cosleeping: mutual physiological and behavioral
influences and their relevance to SIDS (sudden infant death syn-
drome). Early Hum Dev 1994;38:187-201.
Sleep, Vol. 19, No.9, 1996
17. McKenna JJ, Mosko SS, Richard CA. Bedsharing promotes
breast feeding. Pediatrics 1996 (in press).
18. McKenna JJ, Mosko SS, Richard C, et al. Behavior and behav-
ioral interactions of solitary and social sleeping human mother-
infant pairs: implications for SIDS and infant development.
Sleep Res 1994;24:22.
19. McKenna JJ, Mosko S. Evolution and infant sleep: an experi-
mental study of infant-parent co-sleeping and its implications
Downloaded from https://academic.oup.com/sleep/article-abstract/19/9/685/2749884 by PYEONGTAEK UNIVERSITY user on 21 March 2020
for SIDS. Acta Paediatr 1993;389:31-6.
20. Willinger M, Hoffman HJ, Scheidt PC, et al. Infant sleep po-
sition and SIDS in the United States. A}DC 1993;147:460-1.
21. Farooqi S, Perry n, Beevers DG. Ethnic differences in sleeping
position and in risk of cot death. Lancet 1991;338:1455.
22. Scragg R, Stewart AW, Mitchell EA, Ford RPK, Thompson
JMD. Public health policy on bed sharing and smoking in the
sudden infant death syndrome. N Z Med } 1995;108:218-22.
23. Ford RPK, Taylor BJ, Mitchell EA, et al. Breast feeding and
the risk of sudden infant death syndrome. Int} Epidemiol1993;
22:885-90.
24. Hoffman HJ, Damus K, Hillman L, Krongrad E. Risk factors
for SIDS: results of the National Institute of Child Health and
Human Development SIDS Cooperative Epidemiological Study.
Ann NY Acad Sci 1988;533:13-30.
25. Gilbert RE, Wigfield RE, Fleming PJ. Bottle feeding and the
sudden infant death syndrome. 8M} 1995;310:88-90.
26. MacFarlane A. Olfaction in the development of social prefer-
ences in the human neonate. In: Parent-infant interaction. Ciba
Foundation Symposium 33. New York: Elsevier, 1975:103-13.
27. Cernoch JM, Porter RH. Recognition of maternal axillary odors
by infants. Child Dev 1985;56: 1593-8.
28. Mosko S, Richard C. McKenna J, Drummond S. Infant sleeping
position and CO 2 environment during cosleeping: the parent's
contribution. Pediatr Pulmonol 1995;18:394.
29. Kinney HC, Filiano JJ, Sleeper LA, Mandell F, Valdes-Dapena
M, White WF. Decreased muscarinic receptor binding in the
arcuate nucleus in sudden infant death syndrome. Science 1995;
269: 1446-50.
30. Galland BC, Peebles CM, Bolton DPG, Taylor BJ. Sleep state
organization in the developing piglet during exposure to differ-
ent thermal stimuli. Sleep 1993;16:610-9.
31. Stewart MW, Stewart LA. Modification of sleep respiratory pat-
terns by auditory stimulation: indications of a technique for pre-
venting sudden infant death syndrome? Sleep 1991;14:241-8.
32. Thoman EB, Graham SE. Self-regulation of stimulation by pre-
mature infants. Pediatrics 1986;78:855-60.
33. Mitchell EA, Taylor BJ, Ford RPK, et al. Four modifiable and
other major risk factors for cot death: the New Zealand study.
J Pediatr Child Health 1992;28(Suppl. 1):S3-8.