RESEARCH REPORT

Where Will the Baby Sleep? Attitudes and Practices of New andExperienced Parents Regarding Cosleeping with Their Newborn Infants

HELEN L. BALLDepartment of Anthropology and UDSC

University of DurhamDurham DH1 3HN, England

ELAINE HOOKERDepartment of Anthropology and UDSC

University of DurhamDurham DH1 3HN, England

PETER J. KELLYCentre for Health & Medical Research

University ofTeessideMiddlesborough. TS1 3BA. England

An evolutionary perspective on human infant sleepphysiology suggests that parent-infant cosleeping,practiced under safe conditions, might be beneficialto both mothers and infants. However, cosleeping isnot part of mainstream parenting ideology in theUnited States or the United Kingdom, and little evi-dence is available to indicate whether, and underwhat circumstances, parents sleep with their new-born infants. We present data from an anthropologi-cal investigation into the practices and attitudes ofnew and experienced parents of newborn infants re-garding parent-infant sleeping arrangements in acommunity in the northeast of England. Despite nothaving contemplated cosleeping prior to the birth,new parents in our sample found it to be a conven-ient nighttime caregiving strategy, and one whichwas practiced regularly. Infants slept with both theirparents, some being habitual all-night cosleepers,but commonly beginning the night in a cnb andsleeping with their parents for several hours follow-ing the early morning feed, [infant sleep, newborn,cosleeping, new parents\

An evolutionary perspective on human infant sleepphysiology suggests that parent-infant cosleepingpracticed under safe physical and social circum-

stances might provide a variety of (as yet unexplored)psychosocial and physiological benefits to both mothersand infants. For example, based on laboratory studies ofmother-infant bed sharing, McKenna and colleagues havesuggested that the types of physiological changes mother-infant cosleeping induces in the infant's sleep architecture,arousals, and breast-feeding patterns may help to protectsome infants from SIDS (cot death), and they provide asubstantial amount of physiological, behavioral, and cross-cultural evidence supporting this hypothesis (e.g., Davies1985; McKenna 1990a, 1990b; McKenna and Mosko 1990;McKenna et al. 1994; McKenna et al. 1997; Mosko et al.1993; Mosko etal. 1996; Mosko etal. 1997a, 1997b; Rich-ard et al. 1996). However, cosleeping is not part of main-stream British or American parenting ideology, and issuesrelating to the potential benefits and risks of cosleeping arerarely addressed in antenatal classes, advice pamphlets, orpublications for new parents. When cosleeping is addressedit tends to be presented as if it were a unitary phenome-non—practiced in the same manner across all cultural

144 AMERICAN ANTHROPOLOGIST • VOL. 101, No. I • MARCH 1999

circumstances, for the same reasons, producing the samelikely outcomes, which are typically assumed to be nega-tive (e.g., Mitchell and Scragg 1993). McKenna (1995) ar-gues against this perspective and discusses why cosleepingcannot be treated in binary terms (i.e., practiced or notpracticed), or considered a discrete or uniform variable.Rather, he suggests, outcomes (whether benign, beneficial,or risky) will vary depending on how and why it is prac-ticed, and whether breast-feeding is part of the cosleepingexperience for infants. Furthermore, research-based dataon whether, and under what circumstances, new parentsdecide to sleep with their newborn infants, and how theyoperationalize their decision, are almost nonexistent. Thispaper presents data from an anthropological investigationinto the practices and attitudes of new and experienced par-ents of newborn infants regarding parent-infant sleepingarrangements. The data presented were generated as part ofa community-based research project conducted in theNorth Tees Health District in the northeast of England. Theproject was designed to investigate attitudes and practicestoward nighttime parenting strategies and to obtain base-line data on the prevalence of cosleeping and the circum-stances under which it occurred in a sample of parents in-terviewed before and after their baby's birth (see Hookerand Ball 1998).

Recent generations of British and American parentshave been advised by pediatric health professionals thatsleeping with their babies is "wrong" (e.g., Spock andRothenberg 1985; Sturgess 1977; Wright 1972). Until thisdecade it was common for pediatricians, child psycho-therapists, and other "baby experts" to dismiss parent-in-fant cosleeping (or bed sharing) as poor parenting practicewith detrimental side effects for both parents and children(e.g., Ferber 1986; Janes and Rodway 1974). Professionalobjections included predictions that cosleeping may fosterdependency, be habit-forming (even addictive), be sexu-ally arousing, be overstimulating, be frightening for chil-dren (who may observe their parents having sex), be harm-ful to the parents' relationship (if they refrain from sex),reflect poor limit setting, and contribute to sleep problems(Lozoff et al. 1984; Schachter et al. 1989). Parents werealso warned that not only was cosleeping with newbornsill-advised, it was downright dangerous—25 years agoeven having the baby in the same bedroom was considereda bad idea:

May I make an earnest plea that, whenever it is possible, nonew baby should sleep in the same room as his parents? Thereare several reasons for this, one of them is the fact that moth-ers are liable to be anxious and want 'to just have a look athim' all the time—or worse, to lean over and pick him up.This can be lethal because Mother is tired when she cuddlesthe baby and she is likely to lie down on her bed . . . and fallasleep. Babies have been suffocated that way. . . . So if you dohave any other room, put the baby in it. . . . He is not going to

die before the dawn just because he's alone in there! [Wright1972:57-58]

Recently a new wave of interest in cosleeping (alsocalled bed sharing, or "the family bed") has prompted freshdiscussion in both popular and scientific media (Jackson1992; McKenna et al. 1993; Wright 1997). While it isoften cited that most "traditional" societies, and somenotable "developed" or "industrialized" ones such as Japan(Caudill and Plath 1966) and Korea (Lee 1992), practicecosleeping as the cultural norm, anecdotal evidence sug-gests that British and American parents who sleep withtheir infants/children view themselves as a much malignedminority. Little research has actually been conducted,however, on the prevalence of cosleeping in American andWestern European nations (see Rath and Okum 1995).Where such data exist they tend to either contrast cosleep-ing prevalence among families with different ethnic back-grounds (and hence different cultural traditions regardingchildrearing) in a given community (e.g., Gantley et al.1993; Lozoff et al. 1984; Schachter et al. 1989), or contrastprevalence of cosleeping in case-control studies of clinicaland nonclinical samples from pediatric sleep disorder orpsychotherapy clinics (e.g., Forbes et al. 1992; Hanks andRebelsky 1977; Kaplan and Poznanski 1974). The result isa confusing array of literature on cosleeping that juxtaposesdata on children of different ages, obfuscates cosleepingwith infants (as custom or cultural practice) and cosleepingwith toddlers (as a reactive attempt to address sleep prob-lems), and confounds normal variation in parenting prac-tice with extreme examples associated with trauma andfamily psychotherapy.

With an article in Medical Anthropology (1986)McKenna reignited interest in the cosleeping debate whenhe hypothesized that the behavioral and physiological rela-tionship between a cosleeping mother and child may play aprotective role in helping some infants to resist some formsof SIDS (cot death). Subsequently elaborated upon in a de-tailed series of articles (McKenna 1990a, 1990b; McKennaand Mosko 1990), and reinforced with experimental data(McKenna and Mosko 1993, 1994; McKenna et al. 1994;McKenna et al. 1997a, 1997b; Mosko et al. 1996; Moskoet al. 1997a, 1997b; Richard et al. 1996), these ideas havestruck a chord with many parents and professionals (e.g.,Davies 1994; McKenna et al. 1993; Wright 1997).McKenna drew attention to a series of positive reasons forparent-infant cosleeping (particularly in the first sixmonths) that emerge from evolutionary, developmental,and cross-cultural perspectives. His laboratory researchdemonstrates that maternal sleep contact promotes infantarousals while at the same time lengthening the totalamount of time infants (and sometimes mothers) sleep(Mosko et al. 1997a, 1997b), and mothers and infants whosleep together experience synchronous arousals and coor-dination of sleep stages. Routinely cosleeping infants sleep

BALL. HOOKER, AND KELLY / PARENT-NEWBORN COSI.EEHING 145

in a supine position to facilitate breast-feeding (McKennaet al. 1997), and breast-feeding occurs more frequently(both factors being associated with lower risk of SIDS insome studies) (McKenna et al. 1997). The argument thatthe normal sleep environment for healthy newborns is withtheir mothers receives support from the analysis of thecomposition of human milk—its low energy value causinginfants to feed on demand throughout the day and night;comparative data from closely related primate species(great-ape infants sleep on their mothers) and cross-cul-tural evidence indicates that cosleeping is the species-typi-cal pattern for humans (Blurton-Jones 1972).

A review of the academic literature prior to 1987 indi-cates that there was little published research about theprevalence and practice of parent-child cosleeping at thattime, let alone parent-infant cosleeping. An oft-cited cross-cultural survey (Barry and Paxson 1971), reporting on thepresence or absence of cosleeping as cultural "norm" in186 societies worldwide, described 65% of the 127 socie-ties surveyed as practicing cosleeping (for the remaining35% infants slept in the same room, but not the same bedas the parent(s)—for a further 61 societies infants sharedthe parent's room, but it was unknown whether they alsoshared the same bed). Caudill and Plath (1966) observedthat it was common practice in Japanese culture for chil-dren to share their parents' bed, this style of nighttime par-enting emphasizing the nurturing aspects of Japanese fam-ily life. The observation that cosleeping was the norm inother cultures prompted Lozoff et al. (1984) to investigatethe sleep practices of urban families in the United States;however, they restricted their analyses to children over sixmonths old. In the decade since McKenna's hypothesiswas published several surveys of cosleeping have been un-dertaken (e.g., Hayes et al. 1996; Johnson 1991; Morelli etal. 1992; Schachter et al. 1989), but again the predominantfocus has been on toddlers and preschoolers, not newborns.A notable exception was an ethnographic investigation ofinfant care practices in Welsh and Bangladeshi families inCardiff (Gantley et al. 1993), which found that it was cus-tomary for Bangladeshi mothers to sleep with their infants,while babies born to Welsh and English parents weresometimes placed in cots in their parents' rooms for a pe-riod of two or three months, and then encouraged (close tothe peak age for SIDS) to "get used" to sleeping alone,where possible in their own rooms. The authors also ob-served less pressure in Bangladeshi families for babies to"sleep through the night," whereas for Welsh parents, thedemands of either or both working outside the home re-sulted in considerable emphasis on the infant's sleepingtime. In contrast to the notion of vulnerability in infants ex-pressed by Bangladeshi families, there was among Welshand English mothers a clear push toward encouraging ba-bies to be independent. Gantley et al.'s descriptions ofWelsh parental behavior resonate distinctly with the find-ings, published 30 years earlier, of the Newsons' study of

infant care practices in Nottingham (Newson and Newson1966).

Although Gantley et al. included no statistics on the fre-quency of cosleeping families in their sample, such datawere published following a survey on infant care practicesin Birmingham (Farooqi 1994). This study determinedthat, based on the responses of 374 mothers who com-pleted a questionnaire issued at a large District GeneralHospital, 36% of Asian infants slept in their parents' bedcompared with 11% of white infants. Furthermore, 33% ofwhite infants were reported to sleep in a separate bedroom,compared with only 4% of Asians.

Given the dearth of detailed recent information regard-ing the behavior of parents of newboms, the present studywas designed to investigate the nighttime strategies of par-ents with small infants in an economically depressed,postindustrial community in the northeast of England. Ourparticular interest was in where newborn infants slept andhow parents coped with nighttime caregiving. In this paperwe present and compare data on the expectations and expe-riences of nighttime caregiving for parents having theirfirst baby with those of experienced parents. Other aspectsof the data will be presented elsewhere (Hook et al. inpress; Ball et al. n.d.).

Methods

Using a prospective study design we investigated theopinions and practices of parents regarding nighttimecaregiving, before and after the birth of their infants, viasemi-structured interviews and focus groups. Ethics ap-proval was obtained from North Tees Health Authority andwe contacted parents-to-be through antenatal classes, clin-ics, and the maternity wards at North Tees Hospital. Par-ents-to-be were approached personally by one of the inves-tigators (EH), either in the antenatal ward of North TeesHospital, or at one of the antenatal classes run by NorthTees midwives and Health Visitors. The purpose of thestudy was explained to parents in general terms (no spe-cific mention was made of cosleeping), and confidentialityof all information obtained was assured. If parents agreedto participate their permission was obtained on a consentform and the initial interview was conducted. At the initialinterview, background data on the parents and householdwere obtained (e.g., ages, marital status, occupations, edu-cational qualifications, medical problems, smoking, familycomposition, etc.), together with information on the cur-rent pregnancy, and parents' expectations and intentionsregarding infant feeding, sleep arrangements, infant illness,and sources of parenting information. A contact numberwas obtained in order to arrange follow-up interviewswhen the infant was expected to be two to four months old.This procedure allowed us to gather data on both parents'expectations and opinions prior to the birth, and to monitoractual parental child-care practices after the birth. After

146 AMERICAN ANTHROPOLOGIST • VOL. 101, No. 1 • MARCH 1999

Table 1. Background data on participants.

NParity rangeBoys: girlsDelivery

Mean ace motherMean age fatherMarital status

SmokingMother's education

Mean age of baby at post-natal inlerview

New Parents

231

11:12 (all singletons)16 vaginal7 c-section

25.2 (range 15-33)27.9 (range 19-34) 2 missing

16 married3 living with partner

3 single1 living apart from partner

21 nonsmoking parents9 no postsecondary

14 some postsecondary10.04 wks (range 4-25)

Experienced Parents

172-10

11:8 (inc. boy twins x 1, mixed twins x 1)11 vaginal6 c-section

31.2 (range 23-42)32.9 (range 24-46)

17 married

14 nonsmoking parents10 no postsecondary

7 some postsecondary8.23 wks (range 3-17)

sufficient time had elapsed, parents were recontacted and(if available and willing) were reinterviewed by the originalinterviewer. Recontact interviews were conducted with oneor both parents in their own homes. At this second inter-view, parents were asked about the birth, feeding arrange-ments, sleeping arrangements, baby's environment (e.g.,smoking, etc.), baby's temperament, baby's illnesses, ad-vice received from health professionals, parents' assess-ment of how they were coping, changes they had instigatedin infant care, and the effect of the baby on their relationshipand family. All initial and recontact interviews were con-ducted by the same researcher (EH). Written notes weremade on the interview forms, then coded, entered into aspreadsheet for quantification, and written up as field notes.

Results

Sixty mothers or couples participated in intensive inter-views prior to the birth of their infant and 40 participated inrecontact interviews. Five original subjects declined to bereinterviewed, 14 were uncontactable (phone numbers hadchanged to a cable company in the interim and new num-bers were unobtainable), and one couple experienced apremature stillbirth. Among the 40 families recontacted,two sets of twins had been born. For 23 of the mothers re-contacted this was their first baby. Background data on theparticipants by parenting experience are presented in Table1. First-time parents were, on average, five years youngerthan the experienced parents. 837c were married or livingtogether, while 100% of the experienced parents were mar-ried. Two of the first-time parents were teenage mothersliving with their own parents. 19% of new and 187c of ex-perienced parents included at least one smoker. 6l7r offirst-time mothers and 42%- of experienced mothers hadsome postsecondary education. Approximately 30-35% ofdeliveries for both primiparous and multiparous mothers

were by cesarean section, an unusually high percentage,probably a reflection of the fact that we recruited somemothers who were in the antenatal ward because of poten-tial delivery complications.

The majority of both new (74%) and experienced (59%)mothers intended to try breast-feeding. None of the newparents intended to sleep with their baby (although 4thought it might happen), while over a third (357c) of expe-rienced parents acknowledged that they would, or might,cosleep. Furthermore, 30% of new parents and477<? of ex-perienced parents never anticipated bringing the baby intotheir bed for any reason. Situations imagined by parentswho thought that they might do so included ease of breast-feeding, when baby was ill, and morning cuddles. Newparents were significantly more fearful of perceived nega-tive consequences of sleeping with their infants than wereexperienced parents. 52% of new parents rated fear ofoverlaying as their primary reason against cosleeping,while two sets of parents were fearful of promoting badhabits, one mother citing an example of a friend who wasstill sleeping with her 5-year-old daughter. Another mothercommented that she had read a news report about a babywho, when sleeping with its parents, had "been nearlystrangled" by becoming tangled in the mother's long hair.Among the experienced parents, 4 sets were concernedwith overlaying, 1 set about promoting bad habits, and 2about their own sleep being affected.

Despite the fact that they did not plan to sleep with theirnew babies, all the first-time parents planned to keep theirnewborns close by them at night, 100% indicating parents'bedroom, and particularly "next to parents' bed" as theplace where the baby would sleep. Carry cots and Mosesbaskets were the most frequently cited items that infantswould sleep in, although a couple of other options (cot,swinging crib) were mentioned. Parents' reasons for initiallykeeping the baby in their room centered around two

BALL, HOOKER, AND KELLY / PARENT-NEWBORN COSLRLPING 147

themes: safety and security; for instance, one mother com-mented that she thought her baby would "feel more secure"if it slept near her. New parents anticipated that the infantwould sleep in their room for only a few weeks or months.The longest estimate was of nine months—the majority ofnew parents, however, assumed that they would move theirinfant into another room at some point between one andthree months of age. The parents of almost all the first-timeinfants (except the expectant teenage mothers) already hadthe infant's nursery prepared. However, most indicatedthat they would be flexible in their arrangements and rec-ognized that they had to wait and see what things would belike before making any firm decisions. Several of the first-time mothers, and particularly the younger mothers,seemed to have thought very little about what they woulddo with their baby after it was born, seemingly having dif-ficulty contemplating beyond labor.

When asked to reflect upon why they had chosen a par-ticular sleeping arrangement for their infant, and wheretheir ideas had come from, most new parents claimed tohave trusted their instincts: "it was a natural decision," "itfeels like the right thing to do," "feels comfortable," "moreconvenient." Other popular sources of ideas included fam-ily and friends and literature. Advice from parent-craftclasses was cited only once as a source of ideas for wherethe baby could sleep. When questioned further, severalnew parents commented that although feeding at night hadbeen discussed at parent-craft/antenatal classes, no men-tion was made of any other aspects of nighttime parenting.When we conducted focus groups at parent-craft classes,however, parents consistently volunteered "lack of sleep"as the primary negative influence the impending birthwould have on their lives. Nighttime parenting was, there-fore, something they anticipated with trepidation. Al-though parents had been asked near the beginning of theinterview whether they anticipated that their baby wouldsleep in their bed, and all the parents expecting their firstchild had replied in the negative, we rephrased the questionat the end of the interview and asked parents if they everimagined bringing the baby into their bed. The answers wereceived were strikingly different—only 5 of the new par-ents were adamant that they could not imagine ever bring-ing their baby into their bed. Many parents anticipatedbringing their baby into bed if it was ill, for breast-feeding,or to help it go to sleep, but imagined that they would re-turn it to its crib before they fell asleep themselves. Someparents were comfortable with the idea of falling asleepwith their babies, and one young father (age 19) com-mented that if he ever had to take care of the baby by him-self at night he "would definitely feel safer if the baby wasin the bed" with him.

For the experienced parents, less imagination and moreretrospection was involved in planning for the presentbaby. All but 2 sets of experienced parents anticipated hav-ing their baby in their bedroom at night initially, and ex-

pectations of how long this would last were much greaterthan the expectations of the first-time parents. No experi-enced parents thought the infant would sleep in their roomfor less than four months, most anticipated six months,with a couple of families planning on one to one and a halfyears. Of the 2 who were not planning to have the baby intheir room, 1 mother with one previous child stated that thebaby would be in its own room, and she "didn't want asound monitor either." This mother reported that her firstchild had slept and been fed in its nursery as soon as itcame home, but that it had also been kept in intensive carefor two weeks after birth and she didn't feel she hadbonded with the first baby. With regard to the preparationof a nursery, few of the experienced parents had gone to theeffort of preparing a special room in advance, and approxi-mately one third anticipated that the baby would move intoa sibling's room. This group of parents were evenly di-vided about establishing a routine and sticking rigorouslyto it, half generally expecting that they would follow thepattern established with their previous child(ren), the otherhalf anticipating "playing it by ear." One experiencedmother who was expecting twins commented in responseto our query about her plans that she was prepared to beflexible because "you can't dictate to babies!"

When we recontacted the parents in our sample two tofour months after the birth, we found them pursuing a var-ied array of nighttime parenting strategies. New parents'expectations turned out to be (as surmised) a far cry fromtheir later experience. Although none of the new parentsanticipated bed sharing with their infants, at the postnatalinterview 70% were found to be sleeping in bed with theirinfants at least occasionally. Of these 16 newborns andtheir parents, 2 habitually slept in their parents' bed allnight every night, another 11 habitually slept in their par-ents' bed every night for part of the night, while a further 5slept in their parents' bed occasionally (once per week orless). Although all of the new parents had envisaged thatthe baby would sleep in their bedroom for at least a month,at the time of the recontact interview three (13%) infantswere sleeping in a separate room, one after two nights, an-other after one week, and the third after two weeks. Theparents remarked that these infants were "too noisy," andone or more of the parents was unable to sleep with thebaby in the room. The remainder of the babies were in theparents' room in either their own crib or basket, their par-ents' bed, or a combination of the two. The infant of one ofthe teenage mothers slept next to the mother's bed onweekends and next to the grandparents' during the week.This was explained by the grandmother as an arrangementto help her daughter cope with schoolwork, but was clearlyresented by the daughter herself, telling her mother duringthe interview "she might as well be your baby, Mam!"

The two habitual all-night cosleepers were both breast-feeding infants, and 9 of the 11 combination cosleeperswere, or had initially been, breast-fed. The 5 babies who

148 AMERICAN ANTHROPOLOGIST • VOL. 101. No. 1 • MARCH 1999

never slept in their parents' bed were all bottle fed frombirth, and the association between breast-feeding andcosleeping for new parents was significant (%2 = 10.08,p < 0.001). The ease of breast-feeding in bed, and its natu-ral relationship with cosleeping, was frequently brought upby new parents during interviews: mothers remarked thatthey were surprised how easy they found it to breast-feedlying down—some having been shown how to do this inthe hospital, particularly after a c-section delivery. New fa-thers occasionally initiated bringing the baby into bed, butgenerally it was the breast-feeding mothers who imple-mented this strategy, causing some new fathers a degree ofconcern (see Ball and Hooker et al. n.d.). One new father,whose wife had to return to hospital due to persistentbleeding after five weeks, coped with the nighttime care ofhis daughter by having her in bed. Generally, however, itwas mothers who periodically had to cope with the newbaby at night on their own (fathers on business trips, work-ing night shifts, etc.), and they described the prevalence ofall-night cosleeping as increasing at these times.

All of the habitual and combination cosleepers, and 3out of the 5 occasional cosleepers, expressed surprise athow much easier it was to take care of their infants whenthey coslept. Of these 14 sets of new parents, 10 hadclaimed at the antenatal interview that they would notcosleep, and 3 had rejected the idea of ever bringing theirbaby into bed, including one couple who subsequentlyslept all night, every night with their newborn.

Amongst the experienced parents, 35% anticipated thatthey would sleep with their infants, while 59% actually didso. Two newborns were habitual all-night, every nightcosleepers (in both cases parents had coslept with a pre-vious baby), with another 3 being combination cosleeperswho habitually spent part of the night in the parents' bed.Additionally, 1 of the sets of twins were regular combina-tion cosleepers. A further 4 occasional cosleepers had sleptwith their parents a handful of times, or had been broughtinto bed when ill or unsettled. Seven sets of experiencedparents told us that they had never (or "not really") let theirbabies sleep in their bed. All of the infants of experiencedparents who were habitual or combination cosleepers werealso breast-fed. One experienced father insisted that hisbaby went straight into his own room from birth, while an-other who would have preferred such an arrangement wasoverruled by his wife. In general, however, the experiencedfathers were happy to have their babies both in their bed-room and in their bed. In one instance (the cosleepingtwins) the mother had a previous child with whom she hadcoslept, but for the father these were his first offspring. Al-though he soon adjusted to having the babies sleep in thebed (alternating one asleep on his chest and one by themother's side), for the first week he described sleepingwith "one foot firmly planted on the floor" for fear of roll-ing on them.

Our results indicate that for both new and experiencedparents in our sample cosleeping was a relatively commonpractice. Generally infants were brought into bed with bothparents, but parents having to cope with the baby alone atnight (both mothers and fathers) described cosleeping as astrategy they commonly employed. Many of the parentswe talked to practiced regular part-night cosleeping withthe infant beginning the night in a crib or cot and moving tothe parents' bed sometime during the course of the night,generally between midnight and 4 a.m., and remaining thereuntil morning. Despite preconceptions to the contrary andinitial fears of overlaying, new parents who coslept consid-ered it an effective and easy means of obtaining a goodnight's sleep, and experienced parents who had coslept witha previous infant were willing to repeat the practice.

Discussion

The first-time parents in our sample reinforced the find-ings of other researchers in reporting how ill-prepared theyfelt for the reality of coping with the demands of a newbornbaby (e.g., Monk et al. 1996). Several parents acknowl-edged they found it extremely difficult to cope in the firstfew postnatal weeks, especially at night. Although ada-mant at the prenatal interview that they would not cosleepwith their infant, few realized how frequently infants wakeduring the night, and many adopted cosleeping as part oftheir coping strategy. Parents with no prenatal intention tocosleep (primarily first-time parents) were found to be tak-ing their babies into bed with them for a variety of reasonsincluding ease of feeding, desire to monitor infant, parentalneed for more sleep, desire for closeness, and inability tosettle baby alone.

The finding from our small sample that 70% of new, andalmost 60% of experienced, parents with newborn infantswere cosleepers is interesting in comparison with previousAmerican studies. Morelli et al. (1992) reported that of 18families who were interviewed, none of the parents sleptwith their newborns on a regular basis, most sharing aroom temporarily, with 58% of babies sleeping in separaterooms by the age of three months. This finding contrastedwith data from interviews with 14 Mayan mothers, all ofwhom slept in the same bed with their infant for over a year(Morelli et al. 1992). A survey of U.S. mothers of 126 olderinfants (from six months) and children up to four years old(Lozoff et al. 1984) determined that cosleeping was a rou-tine and recent practice in 35% of white and 707c of blackfamilies (defining a child as a cosleeper if parent(s) andchild slept more than once in the same bed together duringthe previous month as part of normal family practice).Black children, particularly, were frequent all-nightcosleepers. In our sample there was negligible ethnic diver-sity (we interviewed only one Bangladeshi family in whichthe mother coslept downstairs with the infant, while the fa-ther coslept upstairs with their older children). Lozoff et al.

BALL, HOOKER, AND KELLY / PARENT-NI;WBORN CO.SI.HI-:•:PIN<; 149

reported a suggestive trend in both black and white fami-lies for cosleeping to be increased in father-absent house-holds, a finding which is reflected in our data with regardto temporary parental absence, but not restricted to fathers.

A further survey of cosleeping in urban Hispanic-American young children (6-48 months old) (Schachter etal. 1989), following the protocol and definitions of Lozoffet al. (1984), found frequent all-night cosleeping was sig-nificantly more common in the Hispanic-American samplethan Lozoff et al.'s sample of white U. S. families (21% vs.6%). While 25-28% of children in the age brackets 13-24,25-36, and 37^8 months were frequent all-night cosleep-ers, contrary to our findings very few parents coslept withtheir infants under one year of age (6.7% in 6-12 mo. cate-gory). The finding that frequent all-night cosleeping wasrare for infants of 6 to 12 months was consistent with thewhite middle-class American data and also with the Japa-nese data of Caudill and Plath (1966). A recent survey inrural New England (Hayes et al. 1996) of parent-infantcosleeping in a sample of 51 3-5-year-old children in-cluded retrospective questions about sleeping in infancy.Using response categories of "always," "often," and "some-times" to classify cosleepers, and "rarely" or "never" forsolitary sleepers, the authors determined that although ap-proximately half of the children were cosleepers at ages 3to 5, all but one of the 51 children had been placed to sleepin a separate bed in infancy. The practice of infants remain-ing in the parents' bed after breast-feeding was recorded inthis survey but was not classified by these authors ascosleeping. This discrepancy raises an important point re-garding research into cosleeping in infancy: parents oftenrespond to questions regarding the place where their infantsleeps at night by identifying the place where the infantstarts the night, or where the infant "is supposed to sleep."As almost any parent will confirm, newborns have rarelyread the rule book and commonly fail to conform to ex-pected patterns of behavior, particularly regarding where(and when) they "should" sleep. Thus many infants whospend large portions of the night asleep with their parents(i.e., those who start the night elsewhere but are broughtinto the parental bed during the night for whatever reason)fail to be identified as cosleepers unless researchers spe-cifically ask parents if their baby is moved during thecourse of the night. The data presented in this paper wouldlook remarkably different if we had taken an infant's initialnighttime sleep site as the indicator of whether or not itcoslept (9% for new parents vs. 70%; 12% for experiencedparents vs. 59%) and would be very similar to figures gen-erated for white infants in Birmingham (UK) (Farooqi1994). An illustrative case can be made from a study con-ducted in Toledo, Ohio (Chessare et al. 1995). that at-tempted to determine prevalent infant sleep positions viaquestionnaires issued at pediatricians' offices to parentswith infants under seven months. One of the questionsasked parents to indicate where their baby slept, specifying

a series of tick boxes. The results from this question indi-cated that 42% indicated "crib in parents' room," 6% "inparents' bed," 50% "own room," and 2% "other." Taken atface value these results indicate that only a few infantssleep in their parents' bed. Without documenting whetherthe baby was moved during the night, however, the resultsof neither this study nor that by Farooqi in Birmingham(UK) are very useful in helping us understand the preva-lence or practice of parent-infant cosleeping in these com-munities.

The relationship observed between breast-feeding andcosleeping is an obvious one—cosleeping breast-feedingmothers are able to nurse their infants with a minimalamount of disruption, and without either of them fullywaking up, a finding compatible with the reports of Mayanmothers (Morelli et al. 1992) who claimed that they gener-ally did not notice feeding their babies in the night. Obser-vations of cosleeping breast-feeding infants indicate thatthey nurse more frequently and for longer periods thanbreast-feeding infants who do not cosleep (McKenna et al.1997), but nonetheless routinely cosleeping mothers sleepas much as solitary sleeping mothers and rate their sleepmore positively (Mosko et al. 1996; Mosko et al. 1997b).New mothers who had discovered the ease with which theycould nurse at night when cosleeping indicated that theypersisted with breast-feeding for much longer periods thanthey might have done otherwise.

Another important finding highlighted by this researchis the fact that newborn cosleepers in our sample (with bothfirst-time and experienced parents) generally slept withboth their parents. This is relevant in light of research intothe physiological and behavioral correlates of cosleepingin newborns, which has thus far examined only mother-in-fant cosleeping (e.g., McKenna et al. 1990, 1994; Saw-czenko 1997; Taylor 1997; Young 1997; although see Balland Hooker 1997,1998).

We conclude, from this research, that the reality of par-enting a newborn infant causes first-time parents to imple-ment nighttime caregiving strategies that they had not pre-viously contemplated. Bringing the baby into their bed tosleep was described as an "intuitive" strategy by many newparents. In fact, many parents explained their change ofopinion regarding cosleeping with their newborn because"it just felt like the right thing to do." Experienced parents,who had much more realistic expectations regarding in-fants and sleep, generally followed the strategies that hadworked with their previous child(ren). The majority hadpreviously coslept with other infants and sometimes har-bored severe anger or resentment toward third parties(health professionals, relatives, strangers) who voiced theopinion that parents and infants should not cosleep, feelingpnssionately that cosleeping with their infant was "natu-ral." Trevathan and McKenna (1994) summarized theresults of 59 studies that illustrate why parent-infant sleepcontact "feels like the right thing to do," ranging from the

150 AMERICAN ANTHROPOLOGIST • VOL. 101. No. MARCH 1999

benefits of bonding and attachment, through frequent suck-ling, sensory cues that regulate breathing, physiological ef-fects of touch (especially skin-to-skin contact), to thesoothing effects on infants of vestibular stimulation andmaternal heartbeat. Although new parents were generallyunaware of the range of developmental, psychological, andphysiological benefits accruing from parent-infant sleepcontact, they were able to articulate that cosleeping withtheir newborn reduced their anxiety regarding its safety atnight, soothed their infant, minimized the effect of night-feeds on parental sleep, and enhanced their feelings of"closeness" with their baby. The results of this research in-dicate that, despite receiving advice to the contrary (andholding opinions to the contrary in the prenatal period),new parents in our sample experimented with a variety ofinfant sleeping arrangements in the first few postnatalweeks. Once they had experienced cosleeping, the benefitsto both themselves and their infants became obvious, andcosleeping emerged as a regular pattern of behavior. Con-trary to the opinion of Davies (1994) that cosleeping (bedsharing) is unfamiliar to the white ethnic majority of theUnited Kingdom, the results of this sample of new and ex-perienced parents from the northeast of England lead us topredict that parent-infant cosleeping is a more prevalentpractice in Britain than has been generally recognized.

NoteAcknowledgments. This project was funded by the Univer-

sity of Durham and the Centre for Health and Medical Re-search, University of Teesside. We are grateful for the coop-eration of North Tees Area Health Trust, particularly the staffof the antenatal wards at North Tees Hospital, and all theNorth Tees midwives and health visitors who contributed tothe study. We are particularly grateful to the parents and theirnewborns who participated in the research, and to colleagueswhose comments improved this manuscript.

References CitedBall, H. L., and E. Hooker

1997 The North Tees Cosleeping Project: The First 3 Years.Paper presented at Workshop on Mother/Infant Interactionsand the Effects upon Development, Morbidity and Mortal-ity, Institute of Child Health, Bristol, U.K., June 18.

1998 What Happens to Infants Who Sleep with Their Par-ents? American Journal of Physical Anlhropology (ab-stract) Suppl. 26:67.

Ball, H. L.. E. Hooker, and P. J. KellyN.d. Parent Infant Cosleeping. Fathers' Roles and Perspec-

tives.Barry, H., and L. M. Paxson

1971 Infancy and Early Childhood: Cross-Cultural Codes 2Ethnology 10:466-508.

Blurton-Jones, N. G.1972 Comparative Aspects of Mother-Child Contact. //;

Ethological Studies of Child Behaviour. N. G. Blurton-Jones. cd. Pp. 305-329. Cambridge: Cambridge UniversityPress.

Caudill.W., andD.W. Plath1966 Who Sleeps by Whom? Parent-Child Involvement in

Urban Japanese Families. Psychiatry 29:344-366.Chessare, J. B., C. E. Hunt, and C. Bourguignon

1995 A Community-Based Survey of Infant Sleep Position.Pediatrics 96:893-896.

Davies, D. P.1985 Cot Death in Hong Kong: A Rare Problem? Lancet

2.1346-1349.1994 Ethnicity and SIDS: What Have We Learnt? Early

Human Development 38:215-220.Farooqi, S.

1994 Ethnic Differences in Infant Care Practices and in theIncidence of Sudden Infant Death Syndrome in Birming-ham. Early Human Development 38:209-213.

Ferber, R.1986 Solve Your Child's Sleep Problems. New York: Si-

mon and Schuster.Forbes, J. F., D. S. Weiss, and R. A. Folen

1992 The Cosleeping Habits of Military Children. MilitaryMedicine 157:196-200.

Gantley, M., D. P. Davies, and A. Murcott1993 Sudden Infant Death Syndrome: Links with Infant

Care Practices. British Medical Journal 306:16-19.Hanks, C. C , and F. G. Rebelsky

1977 Mommy and the Midnight Visitor: A Study of Occa-sional Cosleeping. Psychiatry 40:277-280.

Hayes. M. J., S. M. Roberts, and R. Stowe1996 Early Childhood Co-Sleeping: Parent-Child and Par-

ent-Infant Nighttime Interactions. Infant Mental HealthJournal 17:348-357.

Hooker, E., and H. L. Ball1998 Parent-Infant Cosleeping: Attitudes and Practices in

North Tees, England. American Journal of Physical An-thropology Suppl. 26:136.

Hooker E., H. L. Ball, and P. J. KellyIn press Sleeping Like a Baby: Attitudes and Experiences of

Cosleeping in Northeast England. Medical Anthropology.Jackson,D.

1992 Three in a Bed. New York: Avon Books.Janes, E., and A. Rodway

1974 The Bounty Baby Book. 10th edition. London: BountyServices.

Johnson, C. M.1991 Infant and Toddler Sleep: A Telephone Survey of Par-

ents in One Community. Developmental and BehavioralPediatrics 12:108-114.

Kaplan, S. L., and E. Poznanski1974 Child Psychiatric Patients Who Share a Bed with a

Parent. Journal of American Academy of Child Psychiatry13:344-356.

Lee, K.1992 Pattern of Night Waking and Crying of Korean Infants

from 3 Months to 2 Years Old and Its Relation with Vari-ous Factors. Developmental and Behavioral Pediatrics13:326-330.

Lozoff. B.. A. W. Wolf, and N. S. Davis1984 Cosleeping in Urban Families with Young Children in

the United States. Pediatrics 74:171 -182.

BALL, HOOKER, AND KELLY / PARI.NT-NI.WBORN OIM.I I-.PING 151

McKenna, J. J.1986 An Anthropological Perspective: On The Sudden In-

fant Death Syndrome (SIDS). Medical Anthropology10:8-22.

1990a Evolution and Sudden Infant Death Syndrome (SIDS):Part 1: Infant Responsivity to Parental Contact. Human Na-ture 1:145-177.

1990b Evolution and Sudden Infant Death Syndrome (SIDS):Part II: Why Human Infants? Human Nature 1:179-206.

1995 The Potential Benefits of Infant-Parent Cosleeping inRelation to SIDS Prevention: Overview and Critique ofEpidemiological Bed-Sharing Studies with Notes on FutureResearch Directions. In SIDS: New Trends in the Nineties.Tori O. Rognum, ed. Pp. 256-265. Oslo: Scandinavian Press.

McKenna, J. J., and S. S. Mosko1990 Evolution and Sudden Infant Death Syndrome (SIDS):

Part III: Infant Arousal and Parent-Infant Co-Sleeping. Hu-man Nature 1(3):291-330.

1993 Evolution and Infant Sleep: An Experimental Study ofInfant-Parent Co-Sleeping and Its Implications for SIDS.Acta Paediatrica Supplement 389:31-36.

1994 Sleep and Arousal, Synchrony and Independence,among Mothers and Infants Sleeping Apart and Together(Same Bed): An Experiment in Evolutionary Medicine.Acta Paediatric Supplement 397:94-102.

McKenna, J. J., S. S. Mosko, C. Dungy, and J. McAninch1990 Sleep and Arousal Patterns of Co-Sleeping Human

Mother/Infant Pairs: A Preliminary Physiological Study withImplications for the Study of Sudden Infant Death Syndrome(SIDS). American Journal of Physical Anthropology 83:331-347.

McKenna, J. J., S. S. Mosko, and C. A. Richard1997 Bedsharing Promotes Breastfeeding. Pediatrics 100(2):

214-219.McKenna, J. J., S. S. Mosko, C. Richard, S. Drummond, L.Hunt, M. B. Cetel, and J. Arpaia

1994 Experimental Studies of Infant-Parent Co-Sleeping:Mutual Physiological and Behavioral Influences and TheirRelevance to SIDS. Early Human Development 38: 187-201.

McKenna, J. J., E. B. Thoman, T. F. Anders, A. Sadeh, V. L.Schechtman, and S. F. Glotzbach

1993 Infant-Parent Co-Sleeping in an Evolutionary Per-spective: Implications for Understanding Infant Sleep De-velopment and the Sudden Infant Death Syndrome. Sleep16:263-281.

Mitchell, E. A., and R. Scragg1993 Are Infants Sharing a Bed with Another Person at In-

creased Risk of Sudden Infant Death Syndrome? Sleep16(4):387-389.

Monk, T. H., M. J,Essex, N. A. Smider, M. H. Klein, K. K.Lowe, and D. J. Kupfer

1996 The Impact of the Birth of a Baby on the Time Struc-ture and Social Mixture of a Couple's Daily Life and ItsConsequences for Weil-Being. Journal of Applied SocialPsychology 26:1237-1258.

Morelli, G. A., B. Rogoff, D. Oppcnheim, and D. Goldsmith1992 Cultural Variations in Infants' Sleeping Arrange-

ments: Questions of Independence. Developmental Psy-chology 28(4):604-613.

Mosko, S., J. J. McKenna, M. Dickel, and L. Hunt1993 Parent-Infant Cosleeping: The Appropriate Context

for the Study of Infant Sleep and Implications for SuddenInfant Death Syndrome (SIDS) Research. Journal of Be-havioral Medicine 16:589-611.

Mosko, S., C. Richard, and J. McKenna1997a Infant Arousals during Mother-Infant Bedsharing:

Implications for Infant Sleep and SIDS Research. Pediat-rics IOO(2O):841-849.

1997b Maternal Sleep and Arousals during Bedsharing withInfants. Sleep 201(2): 142-150.

Mosko, S., C. Richard, J. McKenna, and S. Drummond1996 Infant Sleep Architecture during Bedsharing and Pos-

sible Implications for SIDS. Sleep 19:677-684.Newson, J., and E. Newson

1966 Patterns of Infant Care in an Urban Community. Mid-dlesex, U.K.: Penguin.

Rath, F. H., and M. E. Okum1995 Parents and Children Sleeping Together: Cosleeping

Prevalence and Concerns. American Journal of Orthopsy-chiatry 65:411^18.

Richard, C , S. Mosko, J. McKenna, and S. Drummond1996 Sleeping Position, Orientation, and Proximity in Bed-

sharing Infants and Mothers. Sleep 19:685-690.Sawczenko, A.

1997 Observations of the Effects of Bedsharing and Mater-nal Actions upon Infant Night-Time Thermal Physiologyand CO2 Micro-Environment, in a Low SIDS Risk Group.Paper presented at Workshop on Mother/Infant Interactionsand the Effects upon Development, Morbidity and Mortal-ity, Institute of Child Health, Bristol, U.K., June 18.

Schachter, F. F., M. L. Fuchs, P. E. Bijur, and R. K. Stone1989 Cosleeping and Sleep Problems in Hispanic-American

Urban Young Children. Pediatrics 84:522-530.Spock, B., and M. B. Rothenberg

1985 Dr. Spock's Baby and Child Care. New York: PocketBooks.

Sturgess, R.1977 The Baby Book. London: Magnum Books.

Taylor, B.1997 Preliminary Observations of a Bed-Sharing Study,

Ethics and Methods of Analysis. Paper presented at Work-shop on Mother/Infant Interactions and the Effects uponDevelopment, Morbidity and Mortality, Institute of ChildHealth, Bristol, U.K., June 18.

Trevathan, W. R.. and J. J. McKenna1994 Evolutionary Environments of Human Birth and In-

fancy: Insights to Apply to Contemporary Life. Children'sEnvironments 11:88-104.

Wright, E.1972 The New Childhood. London: Tandem.

Wright, R.1997 Why Johnny Can't Sleep. Time, April 14:74-75.

Young, J.1997 Night-time Behaviour between Low SIDS Risk In-

fants and Their Mothers: A Longitudinal Study of RoomSharing and Bed Sharing. Paper presented at Workshop onMother/Infant Interactions and the Effecis upon Develop-ment. Morbidity and Mortality, Institute of Child Health,Bristol, U.K., June 18.