Co-Sleeping and Kids: What Research Shows About Safety and Development

Co-sleeping — sleeping in proximity to an infant or young child — is one of the most common parenting practices in the world. Cross-cultural research consistently finds that the majority of the world’s parents practice some form of infant-parent sleep proximity, whether bed-sharing, room-sharing, or sleeping on adjacent mats. In the United States, the practice exists at high rates despite official recommendations against it, creating a situation where the most common sleep arrangement among new parents is also the one most consistently warned against by their pediatricians. Understanding why requires actually reading the research — both the AAP’s guidelines and the evidence base James McKenna and others have developed that complicates the simple “never bed-share” message. This article covers both honestly, without minimizing real risks or ignoring the context those risks exist in.

Key Takeaways

• The AAP’s safe sleep guidelines recommend against bed-sharing specifically, citing SIDS risk — but the evidence base behind this recommendation is more nuanced than the recommendation itself, and context matters enormously.
• James McKenna’s mother-infant cosleeping research at Notre Dame has documented that breastfeeding mother-infant pairs who bed-share show different arousal patterns than solo-sleeping infants — a finding with safety implications the AAP guidance does not fully incorporate.
• The greatest SIDS risk in sleep-related infant deaths is not the parental bed per se but specific hazardous features: soft sleep surfaces, pillows, parental alcohol or drug use, heavy bedding, and sofa/chair sleeping.
• Cultural practice differences across global populations produce strikingly different SIDS rates even where bed-sharing rates are high — a finding that complicates simple causal claims about bed-sharing and infant death.
• The distinction between planned, safer bed-sharing and hazardous surface sleeping (sofas, recliners) is critical but often absent from both AAP guidance and popular parenting discourse.
• Room-sharing without bed-sharing is the AAP’s recommended compromise and is supported by evidence showing it reduces SIDS risk while avoiding the highest-risk sleep surfaces.

The AAP Safe Sleep Guidelines: What They Say and Why

The Current Recommendations

The American Academy of Pediatrics updates its safe sleep guidelines periodically; the most recent major revision was in 2022. The core recommendations: infants should sleep on their back, on a firm flat surface, in a separate sleep space (bassinet, crib, or play yard) that meets safety standards. Room-sharing — having the infant’s sleep space in the parents’ bedroom — is recommended for at least the first six months. Bed-sharing is not recommended.

The AAP also specifically identifies the highest-risk sleep environments: soft sleep surfaces (sofas, armchairs, soft mattresses), sleeping with parents who have consumed alcohol, marijuana, opioids, or other sedating substances, sleeping with adults who are very tired, and any sleep surface with soft bedding, pillows, or loose items that could obstruct an infant’s airway. The 2022 update was notably more specific about the hierarchy of risk than earlier versions — a change that implicitly acknowledges that not all bed-sharing scenarios carry equivalent risk.

The stated basis for the bed-sharing recommendation is epidemiological data showing that a disproportionate number of sleep-related infant deaths involve bed-sharing. This is real data. The question of what it tells us — and what it doesn’t — requires going further into the evidence than the headline recommendation.

What the Evidence Base Actually Shows

The epidemiological data on sleep-related infant death and bed-sharing has a significant methodological limitation that the research community acknowledges but that the public guidance rarely highlights: most studies of sleep-related infant death do not distinguish between bed-sharing with a sober, non-smoking, breastfeeding parent on a firm surface and bed-sharing on a sofa with a parent who has been drinking. These are counted in the same category in most surveillance data.

This limitation is not a conspiracy or an oversight — it reflects the genuine difficulty of reconstructing exact sleep conditions after an infant death. But it means that the risk estimate attached to “bed-sharing” in epidemiological studies is a composite of scenarios with dramatically different risk profiles, and using that composite number to advise against all bed-sharing treats very different situations as equivalent.

Peter Blair and colleagues at the University of Bristol conducted one of the most methodologically careful analyses of this distinction, published in the British Medical Journal in 2013. Using careful reconstruction of sleep circumstances, they found that among infants without specific risk factors (parental smoking, alcohol use, extreme prematurity, formula feeding), the risk associated with bed-sharing was substantially lower than population-level estimates. For infants with multiple risk factors — particularly parental alcohol use and soft sleep surfaces — risk was dramatically higher. The risk is not uniformly distributed.

James McKenna’s Mother-Infant Cosleeping Research

The Breastfeeding Bed-Sharing Pair

James McKenna, who directed the Mother-Baby Behavioral Sleep Laboratory at the University of Notre Dame, has spent decades studying the behavioral and physiological dynamics of bed-sharing mother-infant pairs in a laboratory setting. His work focuses specifically on breastfeeding mother-infant pairs, a population he argues is meaningfully different from the general population studied in epidemiological research.

McKenna’s central finding: breastfeeding mothers and infants who sleep together exhibit synchronized arousal patterns across the night — they wake and rouse at similar times, in ways that appear to be coordinated through physical contact, sound, and olfactory cues. Infants sleeping next to a breastfeeding mother make more micro-arousals across the night than infants sleeping alone, and these arousals are associated with the kind of breathing and heart rate regulation that may be protective against SIDS.

SIDS is believed to involve a failure of arousal — an infant who, when oxygen levels drop (due to position, airway obstruction, or other factors), fails to rouse sufficiently to self-correct. The theoretical protective mechanism McKenna proposes is that bed-sharing with a breastfeeding mother, by increasing the frequency of infant micro-arousals, may strengthen this arousal capacity during a developmental window when it is immature. This hypothesis has not been definitively confirmed as a causal protective mechanism, but it is biologically plausible and consistent with the cross-cultural data.

McKenna is explicit that his research supports a specific and narrow version of bed-sharing: breastfeeding, sober, non-smoking parent on a firm surface without heavy bedding. He does not claim that all bed-sharing is safe, and he does not advocate against the AAP’s warnings about hazardous sleep environments.

The Evolutionary Argument

McKenna’s work also draws on evolutionary biology: across all non-human primates, infants sleep in contact with their mothers. Human infants show physiological characteristics — high-pitched cries, frequent waking, need for constant feeding — that are consistent with an evolutionary history of continuous contact sleeping rather than isolated sleeping. The isolated infant sleep environment that Western pediatric guidance promotes is, in McKenna’s framework, a historically recent and evolutionarily novel context that the infant’s neurological and physiological systems were not designed for.

This argument is not, in itself, a safety claim — many evolutionarily novel practices are safe and beneficial. But it provides a biologically grounded reason to expect that the physiological dynamics of parent-infant sleep proximity would not be uniformly negative, and it motivates the kind of careful mechanistic research McKenna’s laboratory has conducted.

Understanding SIDS Data Across Cultures

The Cross-Cultural Paradox

The epidemiology of SIDS shows a pattern that is difficult to explain if bed-sharing is uniformly a primary risk factor: several countries and cultures with very high rates of bed-sharing have SIDS rates substantially lower than the United States, where bed-sharing rates are more variable and where official guidance discourages it.

Japan has among the highest bed-sharing rates in the developed world and among the lowest SIDS rates. Hong Kong has similarly high bed-sharing rates and very low SIDS rates. Research examining these cases consistently points to cofactors: Japanese and Hong Kong families have extremely low rates of maternal smoking, very high rates of breastfeeding, and cultural norms around sleep surfaces (futons on the floor, firm surfaces, minimal loose bedding) that reduce other risk factors even while bed-sharing is ubiquitous.

The United States data shows the opposite pattern in some subgroups: African American infants have disproportionately high rates of sleep-related infant death, a disparity that persists after controlling for bed-sharing rates. Research examining this disparity points to higher rates of soft sleep surfaces, sofas, parental alcohol use, and formula feeding — cofactors that cluster with bed-sharing in high-risk environments but are not inherent to bed-sharing itself.

What This Means for Interpreting Risk

Sleep Situation	SIDS Risk Level	Key Risk Factors Present
Back to sleep, crib/bassinet, firm surface, sober parents	Lowest	None — baseline recommendation
Room-sharing without bed-sharing, firm bassinet next to bed	Very low	None — AAP’s recommended approach
Bed-sharing, sober non-smoking breastfeeding parent, firm mattress, minimal bedding	Low-moderate (disputed; Blair et al. suggest low for no-risk-factor pairs)	Debated; McKenna argues near-baseline for specific conditions
Bed-sharing with parental smoking	High	Smoking is a major independent risk factor
Bed-sharing after parental alcohol consumption	Very high	Alcohol dramatically increases SIDS risk
Sofa or chair sleeping with infant	Extremely high	Soft surface, entrapment risk, no arousal capacity
Bed-sharing on soft mattress with heavy bedding	High	Soft surface, airway obstruction risk

The table makes visible something the simple “never bed-share” recommendation obscures: risk is not binary but varies enormously based on specific conditions, and the safest version of bed-sharing carries very different risk than the most hazardous version.

The Sofa Problem: Where the Real Risk Is

Why Sofas and Chairs Are the Most Dangerous Surface

If there is one safety message that the research supports most strongly, it is not “never bed-share” but “never sleep with an infant on a sofa, armchair, recliner, or other soft, enveloping surface.” The entrapment and airway obstruction risks on soft surfaces are dramatically higher than on a firm flat mattress. Infants can become wedged between cushions, between a parent’s body and cushion seams, or can sink into a soft surface in ways that obstruct breathing.

Research on sleep-related infant death consistently finds that sofas and chairs are dramatically over-represented relative to their use as sleep surfaces. A 2018 CDC analysis of infant sleep-related deaths found that accidental suffocation and strangulation in bed — many of which involve soft surface sleep — accounted for a significant and growing share of post-neonatal deaths. The sofa is the highest-risk location in the home for infant sleep and the one most urgently targeted by safety messaging.

Critically, sofa sleeping is often the unplanned consequence of a policy that is too rigid: a parent who is exhausted, who has been told never to bring an infant to the parental bed, falls asleep on the sofa with the infant while nursing. The research suggests that the rigid “never in bed” message may, in some cases, push parents toward the more dangerous sofa scenario by making the firm parental mattress feel equally forbidden. This is one reason some researchers, including McKenna, argue for guidance that explicitly distinguishes between surface types rather than treating all parent-infant co-sleep as equivalent.

Developmental Outcomes of Co-Sleeping

What the Research Shows Beyond Safety

Safety appropriately dominates the co-sleeping conversation, but developmental researchers have also examined whether parent-infant sleep proximity produces different developmental outcomes. The evidence here is more mixed and more difficult to interpret causally than the safety evidence, because families who bed-share differ in many ways from those who don’t, making it difficult to isolate the effect of sleep proximity itself.

Research has found correlations between bed-sharing and longer breastfeeding duration — consistent with McKenna’s finding that breastfeeding is facilitated by nighttime proximity. Extended breastfeeding is associated with improved immune outcomes and some cognitive outcomes. Whether bed-sharing causes better developmental outcomes (beyond supporting breastfeeding) or is simply correlated with attachment parenting practices that have their own effects is not established.

Maternal sleep quality is also relevant. Research on maternal sleep during the infant period finds that bed-sharing mothers show fragmented sleep but also faster return to sleep after nighttime feedings, while mothers using separate sleep spaces show fewer nighttime awakenings but longer arousals when they do occur. Total sleep time may be equivalent; subjective sleep quality often favors whatever arrangement the family chose intentionally.

Transitioning Out of Co-Sleeping

A common parental concern is whether bed-sharing creates difficulty transitioning to independent sleep. Research on this question is limited but does not support the strong version of the concern: children who bed-shared in infancy do not show systematically higher rates of sleep difficulties in childhood compared to those who slept independently, provided the transition is handled gradually and responsively. The transition, when it becomes developmentally appropriate for the family, is easier when framed as a positive development rather than a removal of something familiar.

For parents navigating decisions about kids’ sleep at older ages and the mental health implications of sleep quality, the research on sleep and kids’ mental health addresses the developmental sleep science beyond infancy.

What to Watch for Over the Next 3 Months

If your family is currently bed-sharing or considering it, the next three months are most practically important for identifying and eliminating specific risk factors — not for resolving the policy debate. Work through this specifically: Is there any smoking in the household? (This is the single most important modifiable risk factor for SIDS and should be addressed regardless of sleep arrangement.) Does your sleep surface have soft elements — a pillow-top mattress, loose bedding that could cover an infant’s face, a gap between the mattress and headboard or wall? Is alcohol consumption a factor in any caregiver who sleeps with the infant? Are either parent on sedating medications?

These factors, not the bed-sharing itself, carry the highest burden of risk. Families that address all of them are in a meaningfully different risk category than the population statistics suggest.

Also watch for the sofa drift: if you find yourself repeatedly falling asleep with an infant on a sofa or chair — not by choice, but because you’re too exhausted to get to the bed — that is both a safety signal and a parental exhaustion signal. The sofa-sleeping risk is the most urgent practical safety concern regardless of your position on the broader bed-sharing debate.

Frequently Asked Questions

The AAP says never to bed-share. Should I follow that guidance?

The AAP’s guidance reflects real SIDS risk data and is appropriate to take seriously. The limitation of the guidance is that it doesn’t distinguish between very different risk scenarios. If you choose to bed-share, applying the evidence-based risk-reduction framework (firm surface, sober non-smoking parents, breastfeeding, no soft bedding) addresses the specific factors most strongly associated with increased risk. If you’re uncertain, room-sharing without bed-sharing is well-supported and significantly reduces SIDS risk.

Is it true that countries with high bed-sharing rates have lower SIDS rates?

Yes, and the most supported explanation is that countries like Japan and Hong Kong with high bed-sharing rates also have low rates of the specific risk factors — particularly maternal smoking, alcohol use, and soft sleep surfaces — that dramatically increase SIDS risk when combined with bed-sharing. The cross-cultural data does not prove that bed-sharing is safe; it supports the conclusion that risk factors matter enormously and should not be averaged away in the data.

At what age does co-sleeping become less of a SIDS risk?

SIDS is defined as unexplained infant death under one year of age, with peak incidence between one and four months. The risk falls significantly after six months and is very low after one year. After infancy, the safety calculus around parent-child sleep proximity shifts primarily to sleep quality questions (for both parent and child) rather than SIDS. The AAP’s recommendation for separate sleep surface specifically addresses the first 12 months.

Will bed-sharing make it hard for my child to sleep independently later?

Research does not support the concern that bed-sharing in infancy creates lasting sleep independence problems. Children who bed-share in infancy do not show systematically higher rates of sleep difficulties in childhood. Transition to independent sleep, when age-appropriate, is easier with a gradual, positive approach than with abrupt changes — and the research suggests it is achievable regardless of the infant sleep arrangement.

What should I do if I’m too exhausted to get to the bed safely with my baby?

This is the sofa-risk scenario, and it is the highest-urgency safety situation in the co-sleeping literature. The practical answer: make the parental bed as prepared as possible for safe infant sleep before you go to sleep, so that if you fall asleep while nursing or holding the baby, you are on a firm flat surface rather than a sofa. Have a plan with your co-parent for who takes the baby if you fall asleep while holding them. Parental exhaustion is a real factor in infant death, and the answer to it is not more rigidity about sleep location but safer default environments.

---

About the author

Ricky Flores is the founder of HiWave Makers and an electrical engineer with 15+ years of experience building consumer technology at Apple, Samsung, and Texas Instruments. He writes about how kids learn to build, think, and create in a tech-saturated world. Read more at hiwavemakers.com.

Sources

1. American Academy of Pediatrics. (2022). Safe sleep: Recommendations. Pediatrics, 150(1), e2022057990.
2. McKenna, J. J., & Gettler, L. T. (2016). There is no such thing as infant sleep, there is no such thing as breastfeeding, there is only breastsleeping. Acta Paediatrica, 105(1), 17–21.
3. Blair, P. S., Sidebotham, P., Pease, A., & Fleming, P. J. (2014). Bed-sharing in the absence of hazardous circumstances: Is there a risk of sudden infant death syndrome? An analysis from two case-control studies conducted in the UK. PLOS ONE, 9(9), e107799.
4. Carpenter, R., McGarvey, C., Mitchell, E. A., Tappin, D. M., Vennemann, M. M., Smuk, M., & Carpenter, J. R. (2013). Bed sharing when parents do not smoke: Is there a risk of SIDS? An individual level analysis of five major case-control studies. BMJ Open, 3(5), e002299.
5. McKenna, J. J., Ball, H. L., & Gettler, L. T. (2007). Mother-infant cosleeping, breastfeeding and sudden infant death syndrome: What biological anthropology has discovered about normal infant sleep and pediatric sleep medicine. American Journal of Physical Anthropology, 134(S45), 133–161.
6. Centers for Disease Control and Prevention. (2018). Infant sleep practices and sudden unexpected infant death. CDC.gov.
7. Mindell, J. A., Sadeh, A., Wiegand, B., How, T. H., & Goh, D. Y. T. (2010). Cross-cultural differences in infant and toddler sleep. Sleep Medicine, 11(3), 274–280.
8. Task Force on Sudden Infant Death Syndrome, Moon, R. Y., & AAP. (2022). Evidence base for 2022 updated recommendations for a safe infant sleeping environment to reduce the risk of sleep-related infant deaths. Pediatrics, 150(1), e2022057991.