Don't Retract Pack

Solo vs. Shared Sleep: What Is Safe?

By Patricia Donohue-Carey

For more than ten years, I have been helping expectant families prepare for birth and early parenting. During this period I have become accustomed to hearing strong and conflicting positions on many topics, including epidural anesthesia, circumcision, and the best age for weaning. But no subject has been more challenging than that of bedsharing.

Biologic Model versus Cultural Message
Nighttime solitary infant sleep is not practiced in traditional societies, and even during daylight hours it is the exception. Babies are kept near their mothers. Shared nighttime sleep may take the form of bedsharing (actually sharing the same sleep surface) or cosleeping (when the baby is within arms reach of its mother, but not on the same sleep surface).

According to Katherine Dettwyler, adjunct professor of anthropology at Texas A&M University, Many people in the United States assume that non-Western cultures cosleep or bedshare because limited resources prevent them from creating separate sleep areas for their children. This is simply not true. Mothers in non-Western cultures traditionally sleep with their children to monitor them and keep them safe, to facilitate breastfeeding, and simply to be near them.1 If shared sleep is the behavioral template from earliest human history, why, of late, are some voices seeking to erode its legitimacy?

The American cultural values of independence and control explain a great deal of the societal encouragement of parent-infant separation and the priority placed on parental convenience. (Think of the many products designed to spell parents from their children during daylight hours, such as the swing, infant seat, playpen, jumper, activity center, and walker.) Since bedsharing literally embodies maternal-infant interaction in order to meet a child's nighttime needs, it may appear both out of sync and just plain unattractive.

These overarching cultural messages have long been apparent. What is less clear is the percentage of US infants who have slept in cribs versus adult beds versus a bit of both. Breastfeeding rates decreased dramatically after World War II, and the primary motivation for keeping one's baby nearby at night was considerably diminished.

Solitary sleep for babies gradually became normal; eventually, for many parents, it seemed preferable. Distance between baby and mother was deemed good, even healthy. Sleep had come to mean sex, and concerns that bedsharing might threaten marital intimacy took on a priority greater than the traditional wisdom of being close to one's baby in the night. Somewhere along the line, fear of attaching too deeply to one's child also became a preoccupying, if subconscious, message. Threaded throughout these elements was the long-held fear that babies who shared a bed with their parents might become victims of suffocation through overlying.

But as breastfeeding rates in America began climbing again, reaching nearly 62 percent in 1982, the pragmatic value of bedsharing was revived.2

The Government Weighs In
In September 1999, the Consumer Product Safety Commission (CPSC) recommended against placing children under two years of age in an adult bed, based on its analysis of information on the deaths of children under the age of two for the years 1990 to 1997.3,4 The CPSC warned that using an adult bed as a young child's sleep environment put the child at risk for overlying, strangulation, and suffocation. Young children, the agency advised, should be put only in safety-approved cribs.

Numerous critics (among them pediatricians, anthropologists, and at least one key individual within the agency) quickly pointed out the many shortcomings of the study on which the CPSC recommendation was based:

Inability to Apply Findings and Compare Risks: The study did not allow for statistical inferences because its admittedly anecdotal data did not control for demographic variables such as race, ethnicity, age, family unit structure, and social/economic status. Without such controls, a study cannot define who is at risk. Its conclusions can describe and analyze patterns of injury and death within the population comprising the database, but they cannot be statistically applied to the general population. The data also lacked the means to compare the risk of adult beds with the risk of other sleep environments; including the safety-approved cribs the CPSC was recommending.

Problems with the Diagnosis of Overlying: Diagnostic information from the CPSC databases on suffocation due to overlying was criticized because of the dependence on subjective descriptive data from death-scene investigations. In the absence of physical signs of injury, SIDS (sudden infant death syndrome) deaths can be indistinguishable from deaths due to overlying. Therefore, some death certificates listing overlying as the cause of death may actually represent SIDS deaths. Moreover, in some parts of the country, death certificates are completed by persons (e.g., coroners) who may have little medical training, while in others, specialists in forensic science complete death certificates.

Death-scene investigation forms also vary widely from state to state. Some make detailed inquiries into deaths in cribs as well as adult beds; others ask numerous questions about adult beds but few about cribs. In the latter template, the inference is that inquiry needs to be pursued only when death occurs in the adult bed. Thus, personal and social biases can also impact diagnosis.

Uncertain Rates of Impaired Arousal: Only two death certificates in the study listed alcohol consumption as a factor in the overlying diagnosis. The study's authors, however, noted that Ă’death certificates often provide limited information in this regard, and therefore it is not known whether alcohol consumption was a contributing factor in other cases.5 Asked about the study, Dettwyler stated, "Rates of drug abuse are higher in the United States than in any traditional societies where people breastfeed and cosleep." Given that an earlier report (mentioned in the study) had found that alcohol consumption was involved in a significant number of overlying cases, it appears very likely that the rate of alcohol or drug-related impairment may be higher than was reported and that this would affect the real rate of death from overlying. One can only imagine the additional guilt and related disincentive that a parent of a suffocated child would experience in divulging drug-impaired behavior to a death-scene investigator.

Missing Information: Several key variables were not consistently featured in the study's data and therefore could not be analyzed. For instance: Were young infants placed prone (tummy down) for sleep? Was the mattress soft and/or sagging? Was the baby found lying upon a soft surface, such as a pillow? How many persons were bedding together? Were young children sleeping with the infant who died? Did the mother smoke during her pregnancy or at the time of the child's death? Was either parent (or both) significantly obese or suffering from extreme fatigue? Was there any history of abuse or previous infant death in the family?

The CPSC researchers did not frame the significance of this missing data, choosing instead to focus solely on the location of these children at the time of death. Thus, advising against bedsharing became the main message the CPSC delivered to the American public.

Public Health Message or Product Promotion?
In May 2002 the CPSC, in conjunction with the Juvenile Product Manufacturers Association (JPMA), released a second recommendation against putting a baby to sleep in an adult bed. This time the announcement was part of a new "national safety campaign." The campaign, promoted first through an agency media release, includes a video news release, pamphlets, and posters, and targets new and expectant parents, daycare providers, hospitals, and health departments. Retail outlets of infant products will also participate in disseminating the message that infants should sleep only in safety-approved cribs.

Neither CPSC acting chairman Thomas Moore, in his statements in a CNN article, nor the agency's own announcement to the media made mention of the anecdotal nature of the data the agency used for this recommendation.6,7 In fact, the CPSC's advice was based on the same anecdotal data as the 1999 study and was, therefore, characterized by the same inherent limitations described above. Unlike the CPSC's 1999 statement, the 2002 recommendation refrained from making an overt pronouncement against bedsharing. It did, however, mention the risk of a child dying due to overlying as an inherent "hidden hazard" of an adult bed and instructed the public to put babies into safety-approved cribs for sleep; thereby giving the critical, if unintended, impression that bedsharing cannot be safely practiced.

The CPSC's internal memorandum of May 22, 2002, issued by Joyce McDonald of the agency's Division of Hazard Analysis, however, clearly states that the reports from the databases are anecdotal, not statistical in nature.8 In response to questions submitted by e-mail, the division's director, Russell Roegner, confirmed that the same three databases used in the 1999 study had again been used for the May 2002 analysis and that the data were anecdotal, not statistical. As it did in 1999, the CPSC has chosen to highlight the location at the time of death rather than the presence of discrete risk factors in these cases. While still not forming a complete picture, this additional information (had it been available) would have better described the situations surrounding these deaths.

I, and many others, would have been grateful for the CPSC's national safety campaign message had it read, "Do not put a baby or young child alone in an adult bed." Leaving a young child unsupervised in an adult bed is not a practice endorsed by anyone on either side of the debate about bedsharing.

Where Is the CPSC Spending Its Energies?
One of the criticisms leveled at the CPSC warnings about bedsharing has been that this issue is fundamentally a parenting matter, not a product concern. In 1999 Mary Sheila Gall, then vice-chair of the agency, made just this point when she characterized the CPSC's recommendation as overreaching by a federal regulatory agency.9

One might suppose that the CPSC has turned its attention (overtly in 1999 and somewhat more subtly in 2002) to warnings about bedsharing because it has ensured that infant products over which it has regulatory authority are demonstrably safe. Unfortunately, such an assumption is incorrect. In her book It's No Accident: How Corporations Sell Dangerous Baby Products (Common Courage Press, 2001), E. Marla Felcher exposes what few parents would guess the fact that infant products can be sold without safety standards in place and without field (actual use) testing. While some infant products have mandatory safety standards guiding their design, many rely on voluntary standards with which the manufacturers are not required to comply. Furthermore, some infant products, such as front and back infant carriers and baby swings, lack even voluntary standards.10 Not until June 2002 were voluntary standards for cradles and bassinets developed and seriously considered for adoption by infant product manufacturers.11 So while a particular adult bed might not be safe for a baby, neither might a new infant product purchased at the local retailer where a poster for the CPSC/JPMA national safety campaign is on display.

Barriers to Information
Expectant and new parents often can feel overwhelmed by the amount of information they are expected to acquire about birth options, caregiver's styles, the warning signs of preeclampsia, preterm labor, and newborn illness, the typical schedule of well-baby appointments and vaccinations, indicators of adequate infant fluid and calorie intake, and infant CPR. Concern about parental overwhelm, however, does not discourage learning in our society if the targeted message serves a tacit cultural priority. Imagine a governmental agency suggesting that it would be best to keep young children out of motor vehicles, since they were designed for adult transportation. An argument could be made that the information parents need to understand regarding car seat safety is too voluminous, making parental compliance unlikely and car travel for young children unacceptably risky. The American Academy of Pediatrics (AAP) website houses a 13-page document entitled "Car Safety Seats: A Guide for Families 2002" ( [1] carseatguide.htm), which includes three diagrams, answers to more than ten FAQs, more than 20 detailed reminders, and a table comparing more than 70 different infant and toddler safety seats.

According to the National Highway Traffic Safety Administration, car accidents are the number-one killer of children over the age of one. Additionally, more than 80 percent of car safety seats are thought to have been installed incorrectly, rendering them ineffective, and in approximately 30 percent of car crashes, car seats cannot prevent death or serious injury.12,13 Yet despite these sobering realities, our acknowledged and institutionally supported cultural priority is to travel with our children as safely as possible. We are willing to accept the hazards of car transportation. Why doesn't optimizing the safety of bedsharing elicit the same response from the CPSC to support families who wish to sleep together as safely as possible?

I have been told that the information I compiled in 1999 about how to maximize the safety of bedsharing is too lengthy and complex, that parents won't be able to fully comply; yet the safety checklist in question was one page in length, compared to the AAP's 13-page overview on car seat safety. Other examples of dismissive attitudes toward parent education about bedsharing are similarly instructive. A colleague once heard the obstetrician responsible for coordinating care for low-income women at their medical center state that the parents of the community in question were "uneducable" and that, therefore, cribs and cribs only should be recommended.14

In 1999 I asked SIDS researcher Fern Hauck at the University of Chicago and Phipps Cohe of the SIDS Alliance if they thought a safety-approved "family bed" could be designed. Both answered that it could be done, but immediately expressed concerns that once a family got such a bed home, they could make modifications to bedding or mattress or the bed's placement in the room that would render it unsafe. I pointed out that the very same dangerous modifications can be, and in fact are made to safety-approved cribs.

Individuals and agencies are quick to marginalize the benefits of bedsharing. Parents frequently comment that their children's nighttime care (particularly breastfeeding) is made easier by bedsharing. For a host of important health reasons, many individuals and agencies encourage mothers to breastfeed, and compelling data exist to suggest that bedsharing may very well reinforce breastfeeding. Yet these same parents are often told that by bedsharing, they are putting their children at risk.

Many families report that they get more sleep when bedsharing with their young children. Mothers who work outside the home often find that bedsharing helps them feel more connected to their infants. Attachment is critical to relationship and emotional health; yet these mothers may be told they are inappropriately compensating for daytime separation from their babies.

Pediatric pulmonologist and researcher James Kemp of the Washington University School of Medicine in St. Louis maintains that since the oft-mentioned benefits of bedsharing (enhanced breastfeeding and maternal-infant bonding) are not yet part of the body of epidemiological data, they are not admissible to the bedsharing dialogue.15 But studies of bedsharing mothers and infants have contributed compelling data suggesting that the relationship between bedsharing and breastfeeding is one of reinforcement and enhancement.16,17,18 These studies, taken together with childrearing practices in present-day traditional societies and the body of knowledge regarding lactation physiology, make a logical case supporting the frequently reported maternal perception that bedsharing enhances breastfeeding. In our e-mail exchange, Kemp said that he is in favor of more investigations of "primitive cultures where close sleeping is practiced safely" because "Americans do not know how to sleep safely in close proximity to their babies."

It is time to do away with the notion that a mother and baby's needs are somehow pitted against one another, that in their core interactions, such as eating and sleeping, what is good for one is inherently hazardous to the other. Our biological design is not so fickle or so fragile. It is time to realize that answers about infant sleep location do not come in a one-size-fits-all package. Unequivocal advocacy of either separate or shared sleep for all families, in all places, at all times, should be rejected. The issues are just not that simple. Yet that is often how recommendations, from individuals and agencies alike, are framed.

Where an infant or young child should sleep needs to be determined after an individual family takes a careful look at its own values associated with shared sleep as well as any present risks to sharing sleep safely. Factors that can be changed (type of bed frame, linens and blankets, room temperature, placement of the bed in the room) need attention. Risks that cannot be readily eliminated (impaired arousal due to alcohol or other drug use, extreme exhaustion, reluctance of one partner to take responsibility for the baby's well-being in the adult bed) need to be understood as clear dangers. When such risks are not present, mothers need to be supported in their intuitive desire to be near their babies. The last thing that American mothers need is another degree of separation from their core mothering instincts.

Bedsharing is not going to go away. In addition to the reasons already discussed, many parents practice shared sleep because they regard separate sleep as a careless and insensitive way of behaving toward their infant. These parents need and deserve to have all the available information as the basis of their decision.

1. K. Dettwyler, e-mail exchanges, October 10 and 12, 1999, and June 5, 2002.
2. US Department of Health and Human Services, Division of Maternal and Child Health, Report of the Surgeon General's Workshop on Breastfeeding and Human Lactation, Washington, DC, June 1984, 18.
3. Consumer Product Safety Commission, "CPSC Warns against Placing Babies in Adult Beds: Study Finds 64 Deaths Each Year from Suffocation and Strangulation," [2] (September 29, 1999).
4. S. Nakamura et al., "Review of Hazards Associated with Children Placed in Adult Beds," Archives of Pediatric and Adolescent Medicine 153 (October 1999): 1019-1023.
5. Ibid.
6., "Safety Agency Warns on Babies in Adult Beds," [3] (May 3, 2002).
7. Consumer Product Safety Commission, "CPSC, JPMA Launch Campaign about the Hidden Hazards of Placing Babies in Adult Beds," [4] cpscpub/prerel/prhtml02/02153.html (May 3, 2002).
8. J. McDonald, Consumer Product Safety Commission memorandum, "Methodology for Adult Bed-Related Fatalities of Children 0-5 Years of Age" (May 22, 2002).
9. M. S. Gall, "Infant-Sleeping Study a Case of Agency's Overreaching," USA Today (October 12, 1999).
10. E. M. Felcher, "The US Consumer Product Safety Commission: The Paper Tiger of American Product Safety," a Project of Understanding Government, [5] (April 2002, 23).
11. E. M. Felcher, phone interview, June 2, 2002.
12. Phone interview with T. Herd, National Highway Traffic Safety Administration, Department of Media and Public Relations, May 30, 2002.
13. National Safety Council, "Buckle Up! Safety Tips on Choosing and Using the Best Car Seat for Your Baby," [6].
14. P. Donohue-Carey, "Good Night, Sleep Tight? Safety in the Family Bed," Childbirth Instructor (March-April 2000): 18-23.
15. J. Kemp, e-mail exchange, June 4, 2002.
16. J. McKenna et al., "Bedsharing Promotes Breastfeeding," Pediatrics 100, no. 2 (1997): 214-219.
17. S. Mosko et al., "Infant-Parent Co-Sleeping: The Appropriate Context for the Study of Infant Sleep and Implications for SIDS," Journal of Behavioral Medicine 16, no. 6 (1993): 589-610.
18. J. McKenna et al., "Sleep and Arousal Patterns of Co-Sleeping Human Mother-Infant Pairs: A Preliminary Physiological Study with Implications for the Study of the Sudden Infant Death Syndrome (SIDS)," American Journal of Physical Anthropology 82, no. 3 (1990): 331-347.


Research Links on CoSleeping vs. Crib Use
Back to Sleep Campaign: 800-505-CRIB; [7]
The Danny Foundation: 800-83-DANNY; [8]
Kids In Danger: 312-595-0649; [9]
National Safe Kids Campaign: 202-662-0600; [10]
US Consumer Product Safety Commission: 800-638-2772; [4]

Patricia Donohue-Carey, BS, LCCE, CLE, has worked as a perinatal health educator since 1991. Her interest in the cross-cultural aspects and spiritual significance of birth and infant care shaped her visit to Ireland (1992) and initiated her trip to Romania (2002). She lives with her husband and their two children in Pleasanton, California.


  1. You mention a one page checklist for safe cosleeping... could we have a link to it?

  2. Thank you for posting this! I am so glad to have FACTS to back up my deeply held belief that co-sleeping is not only okay, but beneficial.