Cover story

Co-sleeping: Can we ever put the issue to rest?

By Jane E. Anderson, MD

Supporters and opponents of an increasingly popular practice have strong opinions about whether bed sharing is good or bad for babies and their parents. Examining the attitudes behind these opinions and the evidence for co-sleeping's alleged benefits and pitfalls will help you tailor your advice to the specific needs of families in your practice.

Dare to mention the word "co-sleeping," and an emotionally charged debate about its pros and cons ensues. Because it pits various philosophies of parenting against one other, the issue of where a child should sleep often creates controversy, between parents, between parents and relatives, and between parents and physicians. Intensifying the predicament is the recent report from the US Consumer Product Safety Commission (CPSC), which recommends that children younger than 2 years of age never be allowed to sleep in bed with adults.¹ The American Academy of Pediatrics, too, has taken a strong stand against bed sharing, cautioning against "routine co-sleeping in a family bed."² Given these warnings and a trend among parents toward increased co-sleeping, pediatricians need to be up-to-date on current scientific knowledge about the medical risks and benefits of co-sleeping, and use a family- and culturally sensitive approach in counseling parents.

To co-sleep or not to co-sleep? That is the question

Co-sleeping means that an infant is sleeping either in contact with another individual or in close proximity to that person. The concept of co-sleeping incorporates a variety of sleeping arrangements, ranging from infants and children sleeping in the same bed with their parents to children and parents merely sharing the same room. Because concerns about co-sleeping focus on infants who spend most of the night sleeping in the same bed as their parents, however, the following discussion applies primarily to this group. Much of the world has practiced this type of co-sleeping for thousands of years, as "A long tradition" describes.

In choosing whether or not to co-sleep, parents may be following a particular parenting philosophy or have other reasons for their decision.

Philosophies. All parenting philosophies, each of which is promoted by a group of "experts," are based on similar premises: Every child has needs; parents are there to meet those needs; if the parent does not meet the child's needs no one else will or can; and if the child's needs are not met, he or she will have psychologic problems. These premises add to parents' burdens as they seek advice that will help them raise an emotionally well-adjusted child.

At one end of the spectrum are advisors who promote "child-guided parenting" or "attachment parenting." These observers believe that fulfilling the infant's needs requires parents to be close to the infant constantly and the mother to nurse on demand to assure the infant of her availability. In his books, William Sears, MD, promotes this philosophy of parenting, and in Nighttime Parenting he advocates co-sleeping to meet infants' needs. In The Baby Book he and his coauthor, Martha Sears, describe how parents Mark and Kelly allowed their child to "cry it out": "Baby cried in a feeling of mistrust and cried the sensitivity right out of the parents....A distance developed between parents and baby. More was cried out of the family than only night waking."³ The book's chapter on sleep says, "The important issue in nighttime parenting is not where your baby sleeps, but how responsive you are to baby's nighttime needs and how open you are to finding the best sleeping arrangement for your family."² Nonetheless, passages like the one cited earlier make clear that these authors believe the ideal sleep arrangement is co-sleeping for an extended period of time.

At the opposite end of the philosophical spectrum are advocates of "parent-guided parenting." These observers hold that one of the first things infants can learn is to comfort themselves. Parenting experts in this camp maintain that the baby should be allowed to cry and then comfort herself so as to develop a sense of self-esteem and competence. Without this opportunity, they say, the infant will not view herself as capable. Thus, Gary Ezzo in On Becoming Baby Wise and Richard Ferber, MD, in Solve Your Child's Sleep Problems are strong advocates of parents' regulating their children's sleep patterns and sleep environment. Ferber states, "Even if you and your child seem happy about his sharing your bed at night, and even if he seems to sleep well there, in the long run this habit will probably not be good for either of you, and you should consider making some changes in the nighttime routines."⁴ Similarly, Ferber writes, "Although taking your child into bed with you for a night or two may be reasonable if he is ill or very upset about something, for the most part this is not a good idea. We know for a fact that people sleep better alone in bed. Studies have shown that the movements and arousals of one person during the night stimulate others in the same room to have more frequent wakings....Sleeping alone is an important part of his learning to be able to separate from you without anxiety and to see himself as an independent individual. This process is important to his early psychological development. In addition, sleeping in your bed can make your child feel confused and anxious rather than relaxed and reassured." More recently, however, Ferber has been quoted as saying, "I wish I hadn't written those sentences....It is a blanket statement that is just not right. There's plenty of examples of co-sleeping where it works out just fine. What's really important is that the parents work out what they want to do."⁵

Thus, although advocates of child-guided and parent-guided philosophies both aim to meet the needs of the infant and child, they disagree on how to interpret those needs. In our practice, we have noticed that most parents alternate between philosophies, depending on the issue. I believe that either philosophy can work out well for the child as long as parents agree with each other and are consistent in their approach to a problem. Too often, parents who have different backgrounds and cultures, or different temperaments and expectations, interpret each other's philosophy as being "uncaring." In working with families, we emphasize that mothers and fathers should discuss their individual parenting philosophies to avoid misinterpreting the other's stance as unloving. We also keep the parents' philosophy in mind in formulating anticipatory guidance on sleep issues.

Table 1 lists books of advice about sleep that represent the two major parenting philosophies.

TABLE 1

Sleeping "by the book"

Child-centered philosophy

Sears W, Sears M: The Baby Book: Everything You Need to Know About Your Baby—from Birth to Age Two. Boston, MA, Little Brown & Co, 1993

Strongly advocates co-sleeping and discourages letting infants cry it out

Tine T: The Family Bed. New York, NY, Avery, 1987

General and historical information on co-sleeping

Parent-centered philosophy

Cohen G (ed): American Academy of Pediatrics Guide to Your Child's Sleep: Birth Through Adolescence. New York, NY, Villard, 1999

Not particularly helpful for practical sleep issues

Cuthbertson J, Schevill S: Helping Your Child Sleep Through the Night: A Guide for Parents of Children from Infancy to Age Five. New York, NY, Main Street Books, 1985

Organized into chapters by the age of the child. Practical information on helping children sleep

Ezzo G, Bucknam R: On Becoming Baby Wise: Learn How Over 500,000 Babies Were Trained to Sleep Through the Night the Natural Way. Sisters, OR, Multnomah, 1998

A very regimented, structured approach. When followed too closely, can cause failure to thrive in breastfed infants. Contains helpful general principles

Ferber R: Solve Your Child's Sleep Problems. New York, NY, Simon & Schuster, 1986

A structured approach to helping infants sleep, often called "ferberizing"

Weissbluth M: Healthy Sleep Habits, Happy Child: A Step-by-Step Program for a Good Night's Sleep. New York, NY, Fawcett, 1999

A behavioral approach to sleep based on parental observation of the infant's behavioral cues. Divided into chapters according to child's age

Reasons for co-sleeping. Parents who co-sleep often say it is the "natural way" to parent an infant. In the introduction to The Family Bed, Herbert Ratner, editor of Child and Family Quarterly, writes, "Nature...has accumulated a built-in wisdom born of a vast clinical experience....The mother is nature's 'prepared environment' for the newborn! We cannot afford to ignore this wisdom."⁶ Parents also choose co-sleeping for the physical and emotional closeness it brings, ease of breastfeeding, or alleviation of nighttime problems and children's fears. Some parents co-sleep simply because the cramped quarters they live in make this the only choice, while others say that co-sleeping "just happened." Other parents maintain that co-sleeping is part of making the family bed a focal point for other family activities, such as reading together, playing board games, or talking. Again, from The Family Bed: "We enjoy being together when we are eating. We enjoy sitting together on the couch, reading. We enjoy going to church together. We enjoy sitting huddled together around a campfire, singing, talking. We receive encouragement to do all of these things together. The family that plays together, and prays together, will stay together. Yet when it comes to sleeping together, we bounce headlong into opposition. Why?"⁶

Reasons for maintaining separate beds. For at least the past 50 years cultural norms in our society have promoted separate sleeping for infants and parents. Expectant parents typically anticipate the birth of their infant by preparing a nursery, complete with crib and changing table. Indeed, many parents in the US never consider co-sleeping.

In a survey in 1990 in Worcester, MA, of 303 parents of 2- and 3-year-olds, 44% of children never shared a bed with their parents, and 55% slept in their bed only occasionally.⁷ Although 11% of parents said their child always co-slept and 14% said the child co-slept several times a week, of those who reported bed sharing, more than three quarters said the child co-slept for only part of the night. The prevalence of co-sleeping did not vary with the age or sex of the child, though other research shows that boys tend to co-sleep longer than girls. Frequent co-sleeping was more common among nonwhite families and single mothers than other groups. The co-sleeping status remained stable over time; a repeat survey showed that 65% of families who initially co-slept were still doing so after one year.⁷

The preference for separate sleeping in this country may be changing, however, at least in some areas of the country. In our practice in San Francisco, most new parents initially choose to sleep with their infants. Some new mothers and fathers soon change their minds about co-sleeping because a thrashing or noisy infant interferes with their own sleep. Other reasons for avoiding co-sleeping are spousal opposition, interference with marital intimacy, pressure from relatives, or fear of spoiling or suffocating the infant.

Rates of co-sleeping are much higher in other countries than in the United States. In a comparative survey of 56 Japanese and 61 American parents, 59% of Japanese children slept with their parents three or more times a week, compared with only 15% of American children. In addition, all of the Japanese children who co-slept did so for the entire night, compared with only 11% of the American children who co-slept.⁸

Table 2 summarizes why parents do or don't favor co-sleeping.

TABLE 2

Why parents do or don't co-sleep

Do

Decreases nighttime problems and fears

Is "natural"

Makes frequent breastfeeding easy

Promotes physical and emotional closeness with infant

Don't

Can result in suffocation

Disturbs adults' sleep

Interferes with marital intimacy

Leads to disapproval by relatives or friends

May "spoil" the infant

Opposed by spouse

Benefits of co-sleeping

The anthropologist James McKenna has been the most prominent researcher to investigate sleep from the premise that co-sleeping is normal and healthy for infants. "Cross-cultural studies of infancy, human evolutionary data, and recent psychobiological studies of primate development," he writes, "suggest that the sensory-rich social sleep environment with which human infants evolved may confer certain physiological or psychological advantages on infants that solitary sleep environments do not provide."⁹ Although most research on infant sleep and disorders of sleep and arousal have been performed with infants' sleeping apart from their parents, studies performed by McKenna and others suggest some of the benefits of co-sleeping.

Makes breastfeeding easier. McKenna and colleagues found in their study of 35 mother-infant pairs that infants who normally co-slept at home "breastfed three times longer during the night than infants who routinely slept separately."¹⁰ This increased time was mainly due to a twofold increase in the number of feedings, though feeding duration also increased. Since human milk is low in fat and protein compared with the milk of other mammals, it is important and necessary for the human infant to nurse frequently. Although controversial, several studies demonstrate that breastfeeding protects against sudden infant death syndrome (SIDS), and that the protective effect is dose related.^11-14 Frequent nursing also benefits the mother by maintaining higher levels of prolactin and extending lactational amenorrhea.

Increases likelihood of supine position. Infants who sleep with their mothers, especially to make it easier to nurse, are much more likely to sleep on their backs and sides than on their stomachs.¹⁵ The evidence continues to accumulate that this sleeping position decreases incidence of SIDS.

Has favorable physiologic effects. During the last 20 years, researchers have documented the relationship between the physical proximity of the caretaker and physiologic changes in the infant. Infant monkeys that were separated from their mothers for as brief a time as three hours demonstrated a decrease in body temperature, an increase in the release of adrenocorticotropic hormone, cardiac arrhythmias, sleep disturbances, and compromise of the cellular immune response, as well as a decline in serum levels of immunoglobulins. These negative physiologic effects of separation may be greater in human infants, who are less neurologically mature than other primates at birth.

Another study demonstrated that infants who "roomed in" spent more time in quiet sleep than infants who slept in the hospital nursery.¹⁶ Other investigations showed that infants placed in incubators actually lost 1.5 more degrees of body temperature than newborns who were placed skin-to-skin with their mothers immediately after birth.¹⁷ McKenna has performed extensive research on co-sleeping mother-infant pairs, studying them in both co-sleeping and separate sleeping environments in the laboratory. Investigators made polysomnographic recordings, including electroencephalogram, electroculogram (which monitors eye movements during sleep), electromyogram, air flow, respiratory effort, and electrocardiogram, on 35 breastfeeding mother and infant pairs. McKenna demonstrated that infants who co-sleep experience more arousals, more frequent shifts from one stage of sleep to another, and less time in non-rapid eye movement (non-REM) sleep than infants who sleep apart from their mothers. McKenna hypothesizes that co-sleepers might be less prone to SIDS than infants who sleep alone because, according to some SIDS researchers, increased arousals and motor activity make it easier to terminate a cardiorespiratory crisis during sleep, especially deep sleep.

Interestingly, McKenna also found that movement of the mothers helped to synchronize mother-infant sleep patterns, allowing the infant less time in deep sleep, from which arousals are more difficult.¹⁸ The authors postulate that "under otherwise safe conditions, the observed changes in stage 3-4 sleep and arousals associated with bed sharing might be protective to infants at risk for SIDS because of a hypothesized arousal deficit. The responsivity of the mother to infant arousals during bed sharing might also be protective." In letters to the editor after publication of this article, however, readers criticized the authors for extrapolating from their data, as they presented no evidence related to SIDS. It is also difficult to know what the influence of an additional person in the bed, such as the father, might be.

In 1986, investigators reported the responses of premature infants in the intensive care unit who were provided with a teddy bear that rhythmically breathed.¹⁹ They noted that premature infants found, touched, and cuddled with the bear. In addition, infants placed with a breathing bear had more quiet and regular sleep than infants placed with a nonbreathing bear. Investigators concluded that "the experience with the breathing bear apparently facilitates neurobehavioral development in these fragile infants."²⁰

Other studies have demonstrated that infants who co-sleep have higher glucose levels, less cardiac variability, and cry less to initiate feedings than infants who sleep alone. Infants who co-sleep are also exposed to higher carbon dioxide levels than other infants because of maternal exhalation. Some studies suggest that this might increase the infant's respiratory drive, though other investigators worry that it might contribute to the risk of SIDS.

Risks of co-sleeping

In discussing the risks of co-sleeping, it's important to distinguish between SIDS and infant suffocation by "overlying." These events are different, yet it is very difficult for autopsy to distinguish them, since neither leads to pathognomonic pathologic findings.

SIDS. Because the cause of SIDS is unknown, probably multifactorial, and difficult to distinguish from rare metabolic, cardiac, or respiratory regulatory dysfunction, it must be acknowledged that an infant who might succumb to SIDS in a crib might also do so while co-sleeping. Further complicating the issue, the position in which an infant is found is not necessarily the position in which he or she died, since parents do move during sleep. Advocates of co-sleeping like to remind people that SIDS initially was called "crib death," indicating children were more likely to die when sleeping alone than when co-sleeping. Any research evaluating the risk of suffocation while co-sleeping must carefully distinguish between infants dying from SIDS while co-sleeping and infants who were truly suffocated by another sleeper.

Researchers who have reinvestigated deaths initially thought to have been caused by SIDS have found that some of these deaths might more accurately be attributed to factors in the environment, such as asphyxiation by an object in the crib or bassinet.²¹ This idea is supported by a report on 30 cases of accidental asphyxia in children between 1 and 36 months of age, none of which related to suffocation by parents. Investigators determined that "necropsy findings in cases of asphyxia due to wedging may be quite nonspecific and may be identical to those found in cases of SIDS."²² Thus, the authors say, a death scene investigation is crucial to accurately determining the cause of death.

Environmental factors that appear to increase the risk of SIDS include high room temperatures, maternal smoking, and low socioeconomic status. How do these relate to co-sleeping? Infants who co-sleep have been noted in several studies to have higher rectal temperatures than children who don't co-sleep. Investigators evaluated 24 co-sleeping infants with overnight continuous recordings of rectal temperatures and compared these data with data from 34 matched infants who slept alone. The co-sleeping infants had significantly higher rectal temperatures (0.1° C) than the infants who slept alone.²³ Another study evaluated the axillary temperature of 26 3-month-old breastfeeding infants who were observed both co-sleeping and sleeping separately and noted that the co-sleeping infants had increased temperatures only during non-REM sleep.²⁴ Researchers also evaluated the development of nighttime temperature patterns in a small sample of Asian babies and found that, despite higher body temperatures when co-sleeping, the infants were at lower risk for SIDS than babies who did not co-sleep.²⁵ Thus, it is not clear whether the increased body temperature of co-sleeping infants constitutes increased risk or protection against SIDS.

Prenatal maternal smoking was identified early as a risk factor for SIDS, and more recently, postnatal maternal smoking has been recognized as a separate risk factor. Many of the early studies of SIDS failed to control for maternal smoking and initially identified co-sleeping as a risk factor because of the higher rates of SIDS among ethnic groups that co-slept. When the data were reanalyzed to account for increased incidence of smoking in various ethnic groups, however, the risk of SIDS from co-sleeping disappeared. It seems now that the increased risk of SIDS with maternal smoking stems from deficient infant hypoxic arousal responses.^26,27 It is evident that mothers who smoke should not co-sleep with their infants. Care must be taken not to classify SIDS that occurs in this situation as parental suffocation, however.

Suffocation by overlying. Even though it is difficult to determine whether an infant has died from suffocation or from SIDS, certain risk factors are associated with overlying, namely alcohol or drug use and co-sleeping on a water bed or couch.

Socioeconomic status must also be carefully considered in evaluating research on SIDS and suffocation associated with co-sleeping. Many families of modest means have no choice but to co-sleep because of their cramped living space. Since other factors associated with lower socioeconomic status, including high rates of prematurity, poor prenatal nutrition, and low birth weight, may increase the rates of SIDS in poorer communities, researchers must be careful to identify and adjust for existing risk factors for SIDS.

How co-sleeping affects sleeping behavior

Few studies have evaluated the effect of co-sleeping on children's sleep and behavioral problems. An evaluation of 186 urban families with children 6 to 48 months of age found that regular co-sleeping was associated with increased night waking or bedtime protests.²⁸ Among families who co-slept, Caucasian parents were more likely than black parents to consider the child's sleep behavior a problem. Much more difficult to ascertain, however, is whether co-sleeping actually causes the sleep problems or merely allows parents to become more aware of sleep problems than when the child sleeps separately. Other possibilities are that children who have sleep problems are much more likely than other children to become co-sleepers, or that co-sleeping is the normal parental response to existing sleep problems.

Attack on co-sleeping

A 1999 report from the CPSC, based on its own data, concludes that no child younger than 2 years of age should be allowed to sleep in an adult bed.¹ The report has generated intense reactions from both the lay and the medical community One example is an essay by McKenna (see "In defense of Maya's mother"), which discusses the flaws in the CPSC's data and the report's failure to acknowledge that parents have a right to choose how they want to take care of their infants. The authors of the report, who are employees of the CPSC, acknowledge flaws in their research, such as the incomplete and anecdotal nature of their databases, which depend on information from death certificates that mention a specific consumer product and consumer complaints, as well as media articles and medical emergency services reports. The way the data were collected did not permit the authors to determine how many children co-slept, and hence were at risk of dying while doing so, or the total number of children who died while co-sleeping. These incomplete databases should lead to underreporting of infant deaths associated with co-sleeping. Because of the way the authors analyze the data, however, they may have overreported such deaths. According to the report, 515 children younger than 2 years who were placed to sleep on adult beds died during an eight-year period (1990-1997). Of these, 121 deaths, or 15 a year, were attributed to overlying by a parent or other family member. The remaining 394 deaths, the authors determined, resulted from suffocation or strangulation when the child's head was caught in a structure of the bed.

The most significant flaw in the CPSC's research is how it ascertained the cause of death. Of the infants who reportedly died because of parental suffocation, the authors say they excluded deaths that probably were caused by SIDS, but included those caused by overlying, indicated in the databases by descriptions such as "mother fell asleep while nursing," "sleeping mother overlay child's body," and infant "accidentally rolled over by mother." We have no way of knowing if these infants really died of SIDS instead of by overlying, as the CPSC claims. In addition, investigators did not collect any information about parental drug, alcohol, or cigarette use. Even more important, the report does not include any information about the number of infants who die in their cribs because of an unsafe sleeping environment. The lay press is appropriately criticizing the government for using flawed research to advocate so strongly that infants should not sleep in an adult bed and for not acknowledging data that demonstrate the possible benefits of co-sleeping.^5,29

After the CPSC report was published, the British Medical Journal presented research from Great Britain that evaluated the sleeping habits of 325 babies between 1 week and 1 year of age who died from SIDS, compared with those of 1,300 healthy infants. The authors concluded that "usual bed sharers" were three times as likely to die of SIDS as "non-bed sharers," but that all of the increased risk could be accounted for by parental smoking. More than four of five (84.2%) parents of infants who died of SIDS were smokers. The authors state, "Our results suggest that...perhaps it is not bed sharing per se that is hazardous but rather the particular circumstance in which bed sharing occurs."³⁰

Sleeping in a crib does not, in and of itself, of course, make the sleeping environment safe. A review of infant deaths caused by mechanical suffocation in the US between 1980 and 1997 examined data on the deaths of 2,178 infants younger than 13 months, most younger than 6 months. Eight hundred seventy nine (40%) of the deaths were attributed to wedging incidents, 22% of which were in a crib. Because of the way the data are presented, it is difficult to assess exact numbers, but at the very least 273 infants died in cribs during the study period, about 16 deaths a year—almost the same number as the parental suffocations revealed by the CPSC analysis. The authors admit to being surprised that so many infants died in cribs: "Few crib deaths were expected, given that the crib regulation has been in effect since 1973."³¹ The regulation calls for no more than 1 inch of space between the crib mattress and the side of the crib and crib railings no more than 2 ³/₈ inches apart.

So what's a pediatrician to do?

Before advising on whether the child should sleep alone or share a bed with his parents, the pediatrician must take into account many factors, such as the infant's needs, the cultural milieu of the family, the current sleep environment, and parental use of drugs, alcohol, and cigarettes. Personal parental issues may include maternal guilt about returning to work, conflicts between parents over parenting styles and philosophies, fatigue and unmet sleep requirements, expectations of infant behavior, interpretation of infant needs and temperament, and pressure from relatives. In addition to assessing the infant's temperament, development, and medical health, the pediatrician needs to determine the expectations and goals of the parents and how well they communicate with each other. Too often, sleep issues, particularly the practice of co-sleeping, become a "family secret" because the parents think the practice is frowned on in our culture. The CPSC recommendations give parents one more reason to hide their bed sharing from pediatricians, increasing the risk that we will lose the opportunity to counsel parents on ways they can co-sleep safely.

Pediatricians need to ask families about sleep and be willing to spend the necessary time to address their concerns. Although we give families educational handouts about sleep, like the accompanying Parent Guide on sleep safety, we find that most families' concerns about sleep issues require additional time at the well visit or on the telephone. We encourage parents to evaluate carefully their expectations and goals so they can decide together on a sleep plan that feels comfortable for both of them. Our goal is to help parents avoid middle-of-the-night arguments about sleep problems, a common scenario. We let parents know that infants' sleep patterns are usually established by 6 months of age, so if they choose to co-sleep, they need to decide how long that will feel comfortable. We also discuss how parents may see their expectations and goals change as they realize how much effort is required to change an infant's sleep patterns or as their own sleep needs become more important.

After assessing parental expectations and all the factors surrounding infant and parental sleep, pediatricians should stress that no matter where the baby is put to sleep, the environment should be safe. Parents need to know that the question of where their child should sleep has no right or wrong answer. If they favor co-sleeping and the infant's needs do not argue against it, co-sleeping is the right choice for them. Parents who are not attracted to co-sleeping should be made to feel equally confident about their choice.

REFERENCES

1. Nakamura Suad, Wind M, Danello MA: Review of hazards associated with children placed in adult beds. Arch Pediatr Adolesc Med 1999;153:1019

2. Cohen, GJ (ed): Guide to Your Child's Sleep. Elk Grove Village, IL, American Academy of Pediatrics, 1999

3. Sears W, Sears M: The Baby Book. Boston, MA, Little, Brown and Company, 1993

4. Ferber R: Solve Your Child's Sleep Problems. New York, NY, Simon and Schuster, 1986

5. Seabrook J: Annals of parenthood: Sleeping with the baby. The New Yorker, Nov 8, 1999, p 56

6. Thevenin T: The Family Bed. East Rutherford, NJ, Avery Publishing Group, 1987

7. Madansky D, Edelbrock C: Cosleeping in a community sample of 2- and 3-year-old children. Pediatrics 1990;86:197

8. Latz S, Wolf AW, Lozoff B: Cosleeping in context: Sleep practices and problems in young children in Japan and the United States. Arch Pediatr Adolesc Med 1999;153:339

9. McKenna J, Thoman E, Anders T, et al: Infant-parent co-sleeping in an evolutionary perspective: Implications for understanding infant sleep development and the sudden infant death syndrome. Sleep 1993;16(3):263

10. McKenna J, Mosko S, Richard C: Bedsharing promotes breastfeeding. Pediatrics 1997;100:214

11. Fredrickson D, Sorenson J, Biddle A, et al: Relationship of sudden infant death syndrome to breast-feeding duration and intensity. Am J Dis Child 1993;147:460(Abstract)

12. Mitchell E, Taylor B, Ford R et al: Four modifiable and other major risk factors for cot death: The New Zealand Study. J Paediatr Child Health 1992;Suppl 1:S3

13. Gilbert RE, Wigfield RE, Fleming PJ, et al: Bottle feeding and the sudden infant death syndrome. BMJ 1995;310:88

14. Hauck F, Kemp J: Bedsharing promotes breastfeeding and AAP Task Force on Infant Positioning and SIDS. Pediatrics 1998;102:662

15. McKenna J, Mosko S: Evolution and infant sleep: An experimental study of infant-parent co-sleeping and its implications for SIDS. Acta Pediatr (suppl) 1993;389:31

16. Keefe M: Comparison of neonatal nighttime sleep-wake patterns in nursery versus rooming-in environments. Nurs Res 1987;36:140

17. Fardig JA: A comparison of skin-to-skin contact and radiant heaters in promoting neonatal thermoregulation. J Nurse Midwife 1980;25:19

18. Mosko S, Richard C, McKenna J: Infant arousals during mother-infant bed sharing: Implications for infant sleep and sudden infant death syndrome research. Pediatrics 1997;100:841

19. Thomas E, Graham SE: Self regulation of stimulation by premature infants. Pediatrics 1986;78:855

20. McKenna J, Thoman E, Anders T, et al: Infant-parent co-sleeping in an evolutionary perspective: implications for understanding infant sleep development and the sudden infant death syndrome. Sleep 1993; 16(3):263

21. Bass M, Kravath R, Glass L: Death scene investigation in sudden infant death. N Engl J Med 1986;315:100

22. Byard R, Beal S, Bourne A: Potentially dangerous sleeping environments and accidental asphyxia in infancy and early childhood. Arch Dis Child 1994;71:497

23. Tuffnell CS, Peterson SA, Wailoo MP: Higher rectal temperatures in co-sleeping infants. Arch Dis Child 1996;75:249

24. Richard C: Increased axillary temperature in non-REM sleep during mother-infant bed-sharing. Early Human Dev 1999;55(2):103

25. Petersen SA, Wailoo MP: Interactions between infant care practices and physiological development in Asian infants. Early Human Dev 1994; 38:181

26. Scragg R, Mitchell E, Taylor BJ, et al: Bedsharing, smoking and alcohol in the sudden infant death syndrome: Results from the New Zealand cot death study. BMJ 1993;207:1312

27. Lewis KW, Bosque EM: Deficient hypoxia awakening response in infants of smoking mothers: Possible relationship to sudden infant death syndrome. J Pediatr 1995; 127(5):69129

28. Lozoff B, Askew GL, Wolf AW: Cosleeping and early childhood sleep problems: Effects of ethnicity and socioeconomic status. J Devel Behav Pediatr 1996;17:9

29. Dickinson A: Kids in the bed. Time, Oct 11, 1999, p 99

30. Blair P, Fleming P, Smith I, 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 (clinical research ed) 1999; 319:1457

31. Drago D, Dannfenberg A: Infant mechanical suffocation deaths in the United States, 1980-1997. Pediatrics 1999;103:e59

THE AUTHOR is Associate Clinical Professor of Pediatrics, UCSF/Mount Zion Medical Center, San Francisco, CA.

A long tradition

The first written reference to co-sleeping is in the Book of I Kings in the Old Testament when two prostitutes went to King Solomon in about 950 BC to resolve a dispute as to whose infant had died during the night. Interestingly, the women attributed the child's death to "overlying." The literature of the 1500s is replete with references to children co-sleeping; they were swaddled and remained under their mothers' constant attention and in the maternal bed until about the age of 2 years. Then the child was gradually weaned and began sleeping in a sibling's bed. Paintings of families in this era usually portray a bed in the background, since all family members slept in the large "kitchen." During this time the Trinity bed was developed in England. It consisted of a large bed on which the immediate family slept and two smaller beds (trundle beds), which rolled out from underneath for the older children, servants, or relatives. In the 17th century, the largest of all beds was designed by John Fosbrooke for the royal family. It held 102 people!

Beginning in the 18th century, people began to worry about the psychologic effect of co-sleeping on the child. Parents were urged to have children conceal their bodies from siblings, and only married couples were expected to share a bed. Even Puritans, however, observed the custom of "bundling," sleeping in the same bed regardless of gender while fully clothed. The purpose of bundling was to keep warm during the cold winters, and it was considered common courtesy to allow a stranger passing through to "bundle" with the family.

In the 19th century, an awareness of how germs are passed from one person to another and cause disease increased the desire for separate beds. In 1893, Scribner's Magazine carried one of the first advertisements for twin beds. It read, "Our English cousins are now sleeping in separate beds. The reason is: never breathe the breath of another." With the technologic advances of the 20th century, children began to be born in hospitals rather than at home, and infants were separated from their mothers in hospital nurseries. The creation and promotion of infant formula furthered the tendency of women to spend less time in bed with their babies, as did the entrance of mothers into the work force. In addition, homes became larger, accommodating separate rooms for the children.

Thus, it is only in the last 100 years that co-sleeping has not been the norm in the US, though it remains the conventional sleeping arrangement in most other countries. Proponents of co-sleeping point out that advances in technology, which separated parents and infants, have now provided products to replicate the co-sleeping environment, such as tape recordings of parental heartbeat, crib vibrators, and infant monitors.

In defense of Maya's mother

By James J. McKenna, PhD

In her book I Know Why the Caged Bird Sings, Maya Angelou writes about how her mother encouraged her to bring her infant son into her bed. When Maya realizes she hasn't crushed her son, as she had feared she would, she hears her mother whisper, "See, you don't have to think about doing the right thing. If you're for the right thing, then you do it without thinking."

A recent report from the US Consumer Product Safety Commission (CPSC) would have us believe that Maya's mother, as well as hundreds of thousands of other mothers and fathers, are wrong; that they somehow are not "doing the right thing." In truth, the CPSC's sweeping recommendation—that all infants regardless of circumstances should sleep in cribs—was made "without thinking" on the basis of data so badly flawed that the renowned SIDS researcher Abraham Bergman calls it a classic example of "garbage in, and garbage out."

At the core of the CPSC study is the finding that 121 children died from 1990 to 1997 when a bed sharing adult rolled over and suffocated them. What is missing from the study, however, are crucial details of the actual bed sharing circumstances, including the infant's sleep position and whether the adult smoked, ingested drugs, suffered from depression, was sober, or was even aware that the baby was present in bed. All of these factors and others significantly increase the chances of an "overlay" or SIDS, quite independently of the use of the adult bed.

Of further concern, the study reports on the number of infants said to have died in adult beds but does not provide information on the total number who sleep in such arrangements and live. Thus, the relative risk is unknown. A third flaw is the study's dependence on what even the authors agree is the anecdotal nature of information they gathered from death certificates. Because death investigations and certification practices vary widely in the US, regional differences exist in how a term such as "overlying" is defined, for example. In addition, Bergman has observed that economic factors can come into play: Deaths of infants with identical pathologic findings are classified as overlying or suffocation if the child is from a family that is poor or from a minority, but is considered SIDS or interstitial pneumonia if the child is from a family that is white or middle class.

Should parents be counseled to take precautions to minimize catastrophic accidents in the bed sharing environment? Absolutely. And, of course, parents should take similar precautions when they place their infants in cribs, where an average of 50 children die by strangulation or suffocation each year. But in making their Draconian recommendation against bed sharing, CPSC officials failed to appreciate that the choice of sleeping arrangements reflects parents' rights and need to take care of an infant or child during the night in a way that they find most fulfilling. Such arrangements are about defining and building social relationships and often depend on whether the parents choose to feed their child with breast or bottle and what they want their infants to know about them and to experience emotionally. Bed sharing reflects how parents best believe they can protect their infants and show them affection, through nurturing gestures—spontaneous touches, caresses, and loving whispers—that my colleagues and I have had the privilege to document using infrared video cameras.

Our research also has shown that the commission is simply wrong to imply that sleeping mothers and fathers are unresponsive to the sounds, touches, cries, and needs of the children in their beds. Consistent with the views of Maya's mother, our studies show emphatically that even in the deepest stages of sleep, mothers respond within seconds to a strange noise, sudden movement, grunt, or cough of a co-sleeping child. Research also shows that bed sharing and breastfeeding mutually reinforce each other since they are an integrated, time-tested, biologic system that maximizes—not threatens—human infant survival as well as maternal health. The closer babies sleep to their mothers, the more they breastfeed. Interestingly, the data also show that both mother and infant actually sleep more when they sleep together than when they sleep in different rooms. Moreover, in self-appraisals, mothers who routinely bed share rate the quality of their sleep as high as, if not higher than, mothers who routinely sleep apart from their infants. And, as mothers know, bed sharing makes breastfeeding easier and more successful, for both the mother and child.

The controversy about co-sleeping may have a positive side because it has educated parents about the benefits of bed sharing and makes them aware of choices they didn't know were theirs to make. Indeed, perhaps someday we will join the rest of the world and regard infant-parent bed sharing as an appropriate and potentially rewarding choice, when practiced safely. Then scientists and parents alike will regard co-sleeping parents not as "products" to be managed by the CPSC, but as loving, nighttime protectors of their children.

Bed sharing with young children may not be for everyone, and shouldn't be for some, when safety is problematic. But what we need right now are more mothers like Maya's to help defend and speak up for the rights of mothers and fathers to sleep in bodily contact with their infants or children in the same bed, if that is their choice.

THE AUTHOR is Professor of Anthropology and Director, Mother-Baby Behavioral Sleep Laboratory, University of Notre Dame, South Bend, IN. This article originally was published in Brown University Child and Adolescent Behavior Newsletter 2000;16(2):1.