Bed sharing increases SIDS risk

The Pediatric Insider

© 2013 Roy Benaroch, MD

The evidence has become quite clear: bed-sharing, or co-sleeping, increases the risk of Sudden Infant Death Syndrome (SIDS).

The latest study to reinforce the risk of bed sharing comes out of the UK (with contributions from New Zealand and Germany). Published in the British Medical Journal in May, 2013, Bed sharing when parents do not smoke: is there a risk of SIDS? An individual level analysis of five major case–control studies combined data from five separate case-control studies on SIDS, creating a data set of 1472 SIDS cases to compare with 4679 healthy babies—the largest data series on SIDS that has ever been collected. The authors were able to separate out the effects of bed sharing along with other SIDS risk and protective factors to determine the risks of SIDS for families who only bed-shared, versus those who combined bed sharing with breastfeeding, smoking, and alcohol use. Other factors like the baby’s age, birth weight, and sleep position were also included. Their results are statistically strong, and show large big effect sizes.

Infants who share a bed with their parents during the first 3 months of life increase their risk of SIDS by five times—even if parents don’t smoke, don’t use alcohol, and exclusively breastfeed. In other words, breastfeeding and other positive SIDS risk factors avoidance does not erase the increased risk of SIDS associated with bed sharing.

In the combined data, 22.2% of babies who died of SIDS versus 9.6% of controls shared beds with their parents. The risk was especially high when other risk factors were present: bed sharing among infants whose parents smoked led to a 65-fold increase in SIDS; if parents consumed alcohol, the risk increased 90-fold. The risk of SIDS was “inestimably large” for bed-sharing if the mother used illegal drugs. But, again, even if none of these other risks were present, there was still a very large increase in SIDS rates. Bed sharing, even among breast-fed babies with no other risk factors, increased the risk of SIDS by a 5-fold compared to babies who slept on their own surface in their parents’ room or in their own rooms.

The AmericanAcademy of Pediatrics has recommended against bed sharing since their 2011 recommendations for the safest sleep environment for babies. Their guidelines are comprehensive and well-referenced, including many specific recommendations:

  • Babies should be put down to sleep on their backs. (That doesn’t mean they must be kept on their backs. Once they can roll, let them roll. Do not use devices that force your baby to stay in one position. Baby sleep positioners kill.)
  • Infants should sleep in a crib or bassinet—on a firm flat surface that’s safety-approved for infant sleeping. Car seats and other devices that hold baby in a sitting or semi-sitting position are not for routine sleep. (Which means that Fisher-Price’s Rock –n- Play Sleeper is specifically contraindicated for sleeping.)
  • Room sharing without bed sharing is recommended.
  • Avoid pillows, quilts, comforters, sheepskins, and other soft surfaces under the infant or in their sleep environment.
  • Avoiding smoking, alcohol, and illicit drug use during and after pregnancy.
  • Breastfeed.
  • Consider offering a pacifier at sleep times.
  • Avoid overheating.
  • Immunize infants according to the established recommendations of the AAP and CDC (that is, don’t use one of the made-up schedules that have no scientific backing.)

Bed sharing is a choice that many families make. Some parents enjoy the closeness of baby, and feel more secure; some nursing moms feel that it makes nursing easier. But parents who choose to bed-share should have honest, well-researched information on both risks and benefits. Bed sharing, even with no other risk factors, dramatically increases the risk that your baby will die of SIDS.

Explore posts in the same categories: In the news, Medical problems

Tags: , ,

You can comment below, or link to this permanent URL from your own site.

14 Comments on “Bed sharing increases SIDS risk”

  1. oldmdgirl Says:

    I don’t disagree with the findings overall… I just wanted to point out that that even in a well designed case control study, missing data can be a huge problem. For instance, 0ver 60% of the data was missing for alcohol and drug use. Bias is possible even without missing data (1) social desirability, 2) people who’ve had a baby die are more likely to be asked about drug use, etc.), and we can see in table 1 that over 60% of the data was missing for these two variables, which most likely was not missing at random further compounding the problem. I’m not saying that EtOH and drugs are not risk factors, but I am saying you probably can’t trust the presented ORs as a result.


  2. Tara Says:

    There are multiple problems with this study (that can contribute to bias, as the other commenter noted) and with the assumption that putting a baby to sleep in a bassinet or crib instead of co-sleeping is even an option for many families. At the end of the day, even if the numbers in this study are taken at face value, the increased risk from bedsharing is approx 15 additional SIDS deaths out of 100,000 babies. I would imagine the number of babies who die in car accident by drowsy-driving parents is higher. As I posit in my analysis of this study, an unwillingness of pediatricians and the public health community to discuss with parents responsible ways to reduce the risks of SIDS and suffocation while cosleeping contributes to these deaths as well. You can read that post here: In addition, UNICEF offers an excellent critique of some of the statistical problems with the study here:


  3. oldmdgirl Says:

    You know, that blog post Tara references raises some interesting points. The first one is: Are people who bedshare doing it because they are exhausted and desperate, or because they intend to bedshare. If it’s the former, interventions need to focus on coping strategies for exhausted and desperate parents (and also, is SIDS just generally higher among exhausted and desperate parents whether they bedshare or not). If it’s the latter, interventions need to focus on reducing intent to bedshare.

    Anyway, just some more thoughts.


  4. Nayenette Says:

    Sigh. Just from the confusion of the terms bedsharing and co-sleeping means no one should take this blog post seriously…


  5. Dr. Roy Says:

    Nayenette, many sites equate the terms cosleeping and bedsharing. For instance, the “Redwineandapplesauce” blog post about this same study, linked in the comment by Tara, includes this clarification: “Cosleeping is sometimes taken to mean sleeping in the same room but in separate beds and is other time taken to mean sharing a bed.”

    However, the AAP doesn’t use the terms interchangeably. From their 2011 safe sleep guideline: “The terms “bed-sharing” and “cosleeping” are often used interchangeably, but they are not synonymous. Cosleeping is when parent and infant sleep in close proximity (on the same surface or different surfaces) so as to be able to see, hear, and/or touch each other.139,140 Cosleeping arrangements can include bed-sharing or sleeping in the same room in close proximity.140,141 Bed-sharing refers to a specific type of cosleeping when the infant is sleeping on the same surface with another person.140 Because the term cosleeping can be misconstrued and does not precisely describe sleep arrangements, the AAP recommends use of the terms “room-sharing” and “bed-sharing.””

    Thanks for pointing out the opportunity to clarify the verbiage.


  6. Dr. Roy Says:

    Fair point on the exact ORs re: alcohol and drug use; but would this missing data spuriously inflate the association of bedsharing (independent of those two risk factors) with SIDS? Both the cases and controls had the same questions asked, so it wasn’t a matter of the SIDS families being asked more drug-related questions.


  7. Dr. Roy Says:

    Thank you Tara for your thoughtful reply, and also for your interesting blog post.

    SIDS is quite uncommon– in the last year with good data, 2006, there were 2327 SIDS deaths in the USA, which works out to about 50-60 per 100,000 live births per year. If, as you say, bedsharing accounts for 15 per 100,000, that’s a large percentage of those deaths.

    You’d like to see pediatricians more willing to discuss ways to make bedsharing safer. Clearly the referenced study shows that the combination of bedsharing + other risks is very, very much more dangerous that bedsharing without other risks. But there is still a substantial increased risk from bedsharing alone, when done as safely as possible, with no other risk factors. This has been seen in multiple studies (that’s why bedsharing was discouraged in the 2011 AAP guideline, published before the paper discussed here.)


  8. Rhoda W. Fulton Says:

    Try telling that to the loving, nurturing, caring GRIEVING parents who thought they were bed sharing safely. What they wouldn’t give to have a “do-over” and not take the risk. It’s not worth it . . .


  9. Tara Says:

    I fear I am being misunderstood in subtle but very important ways. Bedsharing plus the other risks (substance use, etc.) is clearly dangerous. I do not propose that pediatricians encourage bedsharing. However, it is possible to *reduce the risks* of bedsharing, even aside from the other significant risk factors already noted, if parents are going to bedshare.

    Is it possible to reduce those risks to the point that SIDS/suffocation-related deaths in bedsharing instances equal those in non-bedsharing (preferably room-sharing) scenarios? Likely not, though we do not have the data to actually say so because no studies have been done that take into account more specific practices that reduce bedsharing risks. (To be fair, it would be a difficult study to conduct — a case control study with so many detailed confounders that it may well end up more qualitative than quantitative and difficult to analyze.)

    However, just as it is foolhardy and risky for public health officials to teach abstinence only to teens, with the expectation that this strategy will succeed in reducing STI and pregnancy rates, it is foolhardy and risky for doctors and the public health community to continue believing that telling parents never to bedshare will succeed in reducing these deaths. Argue until the cows come home about why the parents do it, but many will choose intentionally, purposefully to do it and should be able to have a frank conversation with their pediatrician in which they learn the ways to *reduce* (not eliminate, which is impossible) the risks as much as possible if they are absolutely not going to not bedshare. Having partners sleep in separate beds, removing all covers, having a firm mattress (of course no smoking, alcohol, etc) — heck, having mom sleep on the floor. No, these are not perfect solutions. No, they are most likely not as safe as room-sharing with a separate bed for baby (again the missing data caveat). But they are a heck of a lot “safer” (I find the terms “safe”/”safer” misleading and inappropriate) than telling a parent outright not to bedshare and then having that parent go home without any guidance, bedsharing, doing so in ways that further increase the risk of SIDS/suffocation, and then becoming a statistic.

    (There are many families for whom bedsharing was not initially a choice but it became the only option because of child sleeping problems. Some may claim those parents simply need guidance in getting their child to sleep and stay asleep, but that is a naive position, as evidenced by the pediatricians who contacted me both on that blog post and privately to thank me because they became conflicted themselves *once they had children of their own* and discovered that all those best practices do not always pan out like the books and experts say they will.) (On a side note, I must add in that I’ve a firmly evidence-based science reporter who promotes the CDC vaccination schedule, etc. I only add this so I am not mistaken as a representative of a particular parenting “school” of thought that has a particular agenda.)

    To Rhoda, I cannot imagine the grief of any parent who loses a baby. Yet there are parents who lose babies in sofas when they fall asleep by accident, sitting down for just a moment, because they are so utterly exhausted and trying so desperately not bedshare and rather to “do things right.” There are others who make dangerous careless errors induced by sleep-deprivation, including while driving, during that day that have cost lives. Forgive my beating a dead horse here, but it is time for public health officials to acknowledge the realities of this issue and address it more appropriately, sensitively and realistically.


  10. oldmdgirl Says:

    It could spuriously inflate the risk of bedsharing if missingness was associated with either the outcome or exposure of interest.


  11. Jess Says:

    Quick question that’s only moderately related: One of the sleep recommendations you (well, and the AAP) make is to put babies to sleep on their backs. Absolutely, I’m with you. And then it says when they roll, you can let them stay there. Again, totally with you. But why is it recommended that you have to put babies to sleep on their backs once they can roll over. My little guy rolls to his tummy every.single.time. So when he wakes up to feed in the middle of the night, falls asleep feeding, and then I put him back down on his back, he wakes up, rolls over, and fusses for awhile. If I put him down straight on his tummy, he goes to sleep much more quickly. Is that still terribly, horribly bad? Or is it a “better safe than sorry” recommendation made without evidence? Do you know of any studies that look at tummy sleeping after babies can roll themselves?


  12. Dr. Roy Says:

    Babies roll at different ages, and even “rolling” isn’t a single skill that everyone can always agree on. I think the AAP committee that wrote the 2011 guideline was looking for a straightforward statement that applies to all babies.

    Obviously, if your baby is rolling all over the place easily the moment you put him down, it probably doesn’t matter one bit what position he’s first put down in.


  13. Keely Says:

    I don’t remember where I read this so perhaps I am missing some details but I have heard that SIDS is non existent in Japan, a country where bed sharing is the norm. Wouldn’t that suggest that there are other factors at play here?


  14. Dr. Roy Says:

    Keely, it is a myth that SIDS doesn’t occur in Japan. The 1st cite, below, from Japanese authors at Tokyo Women’s Medical University, refers to SIDS as “one of major contributing factors to infant mortality in Japan.”

    It is problematic to compare SIDS rates between countries, as definitions vary. Some countries have considered “overlying” deaths (suffocation) in SIDS counts, while others have not; and the requirement for a formal autopsy and scene investigation isn’t always fulfilled.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: