Skip to content

Bed-sharing with Young Infants: Is It Safe After All?

Does bed-sharing with infants increase their risk of SIDS, even without known risk factors such as alcohol use, smoking, and co-sleeping on a couch or chair? A recent study makes what is probably the best attempt to date to answer this question. The study, led by U.K. researcher Peter Blair, was published last week in the journal PLOS ONE and is freely available to the public (yay!).1

Many studies have found that co-sleeping is associated with an increased risk of SIDS, but most of this risk doesn’t come from co-sleeping per se, but rather doing so in particularly hazardous conditions, such as on a couch or with a parent who has been drinking. However, there’s an important, albeit controversial, caveat to this conclusion. Several studies have looked specifically at infants younger than 3 months and still found a significant risk of bed-sharing even in the absence of these other risk factors.2–6 The current study comes to conclusions much more reassuring to bed-sharing parents. In this study, bed-sharing without alcohol, smoking, or couch/chair co-sleeping was not associated with a significant SIDS risk in infants younger than 3 months and even seemed to be protective in older babies. Both of these findings run counter to previous studies and to the sleep recommendations of the AAP, so they deserve a close look.

How was the study conducted?

The study is actually a reanalysis of two previous studies done by the same researchers.7,8 These were case control studies, a type of observational study and the main tool used by researchers to study SIDS risk factors. Combining both data sets together gave the current study a sample size of 400 SIDS infant and 1386 controls.

These studies tracked every sudden unexpected infant death in a defined region (5 regions in the U.K.) for defined periods of time (1993-1996 and 2003-2006). If a death was determined to be SIDS, not due to an identified cause such as a medical condition or suffocation, it was included in the analysis. The families of every SIDS baby were visited within a couple of days of the baby’s death to answer questions about the conditions around the baby’s death, with a follow-up questionnaire scheduled a couple of weeks later. Meanwhile, for each SIDS baby, 4 healthy control babies of the same age and in the same region were recruited, and the control families completed the same interview about how their baby was sleeping at the time the SIDS baby had died. The sleep conditions between the SIDS babies and control babies were then compared to determine significant risk factors for SIDS deaths.

For example, if more of the SIDS babies than the control babies were from households with smokers, then we could conclude that there is an association between smoking and SIDS. In a case control study, this is expressed as an odds ratio (OR), where an OR of 1 indicates no association, OR greater than 1 indicates increased risk, and OR less than 1 indicates decreased risk. Odds ratios are reported with 95% confidence intervals (CI) which tells us how certain we are about this estimate. A very large CI indicates more uncertainty, and if the CI intersects with 1.0 and the P-value is greater than 0.05, then the association is not considered significant.)

It’s always important to keep in mind that a case control study can’t determine that any one factor causes SIDS; it can only establish a correlation. However, for a number of reasons, we can’t do something like a randomized controlled trial for SIDS, so the case control study design is really the best tool we have for looking at SIDS risk factors.

What did the study find?

Like most case control studies of SIDS, this most recent study found that babies that died of SIDS were more likely to be co-sleeping with an adult than to be sleeping alone. However, where the co-sleeping occurred and who was sleeping with baby made a big difference to the risk. For example, co-sleeping on a sofa or chair and bed-sharing with an adult who had more than two alcoholic drinks both increased the odds of dying of SIDS by 18-fold. Bed-sharing with an adult who was a smoker increased the odds by 4-fold. In the absence of these three risk factors, bed-sharing was not associated with increased SIDS risk in this study.

Next, the researchers did an age-based analysis to look at these same risk factors in babies younger than 3 months and older than 3 months. I’ve shown these results in the figure below (created from Table 3 of the paper).

blair et al 2014 dataIn the figure, note that all of the co-sleeping situations were compared to babies that slept alone, which were automatically given an OR of 1. From there, any higher OR indicates greater risk. This is really evident for co-sleeping on a couch or chair, alcohol use, and in young babies, smoking.

In younger infants, there was not a significant risk of bed-sharing in the absence of the other 3 risk factors. However, the OR for bed-sharing in this group was 1.62 (95% CI of 0.96-2.73; p=0.07). Some might call this a marginal risk for bed-sharing, and since at least 5 other studies have found a risk of bed-sharing with young infants, I would be cautious in assuming this is the final word on this question. However, the data are certainly reassuring. Even if there is a small increased risk in young infants, that risk could easily be trumped by the risk of something like driving with severe sleep deprivation if bed-sharing is the only way you get any sleep.

In older infants, bed-sharing in the absence of the other risk factors actually appears to be protective. (This is hard to see in the figure, but the paper reports an OR of 0.08 for bed-sharing in this group. Only one (0.6%) SIDS infant was bed-sharing with not other risks at the time of death compared with 61 (8.5%) of control infants.) This is a tiny sample, so I think we have to be careful about getting too excited about this result, as we’d like to see it replicated in additional studies. But still, it’s interesting and definitely worth more research.

What does this study mean to parents and policy-makers?

In the SIDS research world, the question of bed-sharing safety is hotly debated. Last year, a study led by Bob Carpenter, another U.K. researcher, looked at similar questions and came to very different conclusions.4 That study found that in babies younger than 3 months, bed-sharing increased the odds of SIDS by 5-fold, even in breastfed babies with nonsmoking parents. The study got lots of media attention and was praised by some SIDS researchers but criticized by others. Among its most vocal critics was Peter Blair, author of the current study. Blair was a peer reviewer for the paper and voiced several major objections [PDF], but presumably since the paper had 3 other very positive reviews, it was still published.9 Perhaps his current study was an attempt to fill in the gaps he saw in Carpenter’s study. In our interpretation of this study, we have to remember that it is just ONE study among many, and in science, we have a responsibility to look at ALL the evidence, not just the study that tells us what we want to hear. Still, this new study was careful and well-designed, and it calls into question some of what we’ve been told about the dangers of bed-sharing.

Policy-makers, such as the AAP’s Task Force on SIDS, are faced with trying to translate this contentious area of research to the public. The AAP’s policy is conservative, recommending room-sharing without bed-sharing and stating, “The AAP does not recommend any specific bed-sharing situations as safe.”10 The policy statement goes on to list specific situations where bed-sharing is especially unsafe (such as with babies younger than 3 months, with a smoker, with an adult under the influence of alcohol, drugs, or medications, on a very soft surface with bedding, etc.). However, the overall message is that your baby should sleep alone in a crib, and this is the message that is emphasized in public health campaigns in the U.S.

The problem with the AAP’s policy is that it doesn’t fully convey the vast differences between co-sleeping in hazardous circumstances, such as on a couch or with an impaired adult, and careful and intentional bed-sharing. We’ve also had pretty good evidence for a while that bed-sharing with infants older than 3 months without smoking or alcohol is not dangerous, but this is glossed over in the AAP statement. And now we have some evidence that it may even be safe to bed-share with younger infants, provided we do it carefully. It may be time for the AAP to rethink their approach to safe sleep recommendations.

We also have to recognize that moms have always and will always sleep with their babies, and no official policy or set of guidelines will change that. Sometimes it is the only way anyone gets any sleep. Sometimes it helps moms keep breastfeeding. Sometimes it just feels right. (Not always, but sometimes.) Parents who choose to bed-share shouldn’t feel like they have to hide this fact from their pediatrician. Instead, they should be having an open conversation about how to make their baby’s sleep environment – crib or shared bed – as safe as possible. And judging from the data in this latest study, the most important points to emphasize to parents are these:

  1. Don’t co-sleep on a couch or chair. If you feed your baby on a couch or chair during the night, and you think there is a chance that you’ll fall asleep there, you’ll probably actually be safer – not to mention more comfortable – feeding in bed.
  2. Don’t bed-share if either parent in the bed has had more than 2 drinks of alcohol. (And as a side note, babies should only bed-share with their parents, not kids, pets, or other adults.)
  3. Don’t bed-share if either parent smokes, especially with babies younger than 3 months.

The authors of the study write:

“An important implication of our findings is that to give blanket advice to all parents never to bed-share with their infant does not reflect the evidence. There is a danger that such advice could influence parents to seek alternative, more dangerous sleep surfaces such as a sofa. In our study in 2003-2006 a number of families whose infants died informed us that they had been advised not to bed-share and thus fed the infant (and fell asleep) on a sofa… Of course we should inform the public about risks that can be associated with bed-sharing, but bed-sharing is a widespread socio and cultural norm; giving across the board advice to simply not do it negates the option of highlighting the specific and highly significant risks we have found.”

This is a very important study, and I’m glad to see it published. It asks a question that is important to parents trying to figure out how to get a decent and safe night’s sleep. The findings provide guidance on the most important risk factors for SIDS and reassurance to parents who choose to bed-share that these steps can vastly reduce the risk.

What’s most puzzling to me about this study is that it hasn’t gotten more media attention. It’s been available online for a week now, and I haven’t seen a single news article or blog post about it. I honestly can’t figure that one out.

For more information about how to make your baby’s sleep environment as safe as possible, check out Caring for Your Baby At Night from UNICEF UK. This is both evidence-based and practical and is one of the best resources I have found on the topic.

References:

  1. Blair, P. S., Sidebotham, P., Pease, A. & Fleming, P. J. 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, e107799 (2014).
  2. Blair, P. S. et al. Babies sleeping with parents: case-control study of factors influencing the risk of the sudden infant death syndrome. Br. Med. J. 319, 1457–1462 (1999).
  3. Carpenter, R. G. et al. Sudden unexplained infant death in 20 regions in Europe: case control study. Lancet 363, 185–191 (2004).
  4. Carpenter, R. et al. 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, (2013).
  5. McGarvey, C., McDonnell, M., Hamilton, K., O’Regan, M. & Matthews, T. An 8 year study of risk factors for SIDS: bed-sharing versus non-bed-sharing. Arch. Dis. Child. 91, 318–323 (2006).
  6. Tappin, D., Ecob, R. & Brooke, H. Bedsharing, Roomsharing, and Sudden Infant Death Syndrome in Scotland: A Case-control Study. J. Pediatr. 147, 32–37 (2005).
  7. Blair, P. S. et al. Hazardous cosleeping environments and risk factors amenable to change: case-control study of SIDS in south west England. Br. Med. J. 339, b3666–b3666 (2009).
  8. Fleming, P. J. et al. Environment of infants during sleep and risk of the sudden infant death syndrome: results of 1993-5 case-control study for confidential inquiry into stillbirths and deaths in infancy. Br. Med. J. 313, 191–195 (1996).
  9. Horne, R. S., Moon, R. Y., Blair, P. S. & Gilbert, R. Peer Review History for: Bed sharing when parents do not smoke: is there a risk of SIDS? An individual level analysis of five major case–control studies by Carpenter et al. (2013). at <http://bmjopen.bmj.com/content/3/5/e002299&gt;
  10. American Academy of Pediatrics – Task Force on Sudden Infant Death Syndrome. Policy Statement: SIDS and other sleep-related infant deaths: expansion of recommendations for a safe infant sleeping environment. Pediatrics 128, 1030–9 (2011).

Save

48 Comments
  1. mt #

    I have to say, Alice’s writing about this topic (oh, and being a parent!) have really changed my position on bedhsaring. For a variety of reasons from the practical (our bed is in a frame and high off the ground) to the personal (we aren’t committed to a bedsharing lifestyle) we are not a bedsharing family. The AAP position gave me the opportunity to think, smugly, that our family’s decision was the “correct” one.

    However, I know the desperation of sleep deprivation and the agony of sleep training. For families that *need* sleep and have babies too young to train (or don’t want to train), bedsharing can be a solution. I also think that Alice is right that bedsharing is going to happen whatever the AAP says, so no one’s helped by making it something shameful. By having open discussions about bedsharing, the AAP could help families do it as safely as possible.

    One of these discussions could be about the reality that bedsharing can turn from a snap decision into a lifestyle without the parents’ noticing at first. I’ve met many parents that fall into bedsharing because they’re desperate for sleep, but the bed becomes a site of strife later on (which isn’t good for sleep), when the parents are ready for the little one to sleep solo and the child won’t have it. Or worse, when one parent wants to continue bedsharing and the other doesn’t. By removing the taboo, families might feel more free to make this a considered and intentional decision.

    Like

    September 26, 2014
    • My feelings about bedsharing safety have definitely evolved over the last couple of years, too. I think we still don’t really know if there is some small risk associated with bedsharing minus known hazards in young babies, but I also think that is sort of beside the point. We have a lot of data to guide us in identifying the biggest risks, and from there I think we have to trust parents to make the right choices for them. And then, you’re right – parents often need support with sleep no matter which route they choose and if they find that they need to change course, the more support they have in these decisions and transitions, the better.

      Like

      September 26, 2014
      • Ashby #

        I would venture to say that people aren’t obligated to do the 100% most safe option 100% of the time. Other factors come into play. SIDS is a risk, but extreme parental or infant sleep deprivation is also a risk. Sleep dep induced PPD is a risk. These risks need to be weighed against risk of SIDS in any sleep situation, and the AAP standards don’t give a lot of tools for desperate parents that actually work. Most newborns won’t sleep on their backs in a crib, so I find the recommendation to have a the flavor of CYA.

        Like

        September 30, 2014
        • Absolutely… and nothing is 100% safe. In my opinion, careful bed sharing may be a small risk (or not) but it could easily be offset by other risks like sleep deprivation and PPD. It’s a very personal decision.

          Like

          October 1, 2014
        • Using a bedside co-sleeper is consistent with AAP guidelines in that they take no position on it one way or the other. Seems to me that using a bedside co-sleeper is a way to bed-share while conforming to the AAP guidelines. The AAP guidelines say:

          “There are no data regarding the safety of sleepers that attach to the side of an adult bed. However, there are potential safety concerns if the sleeper is not attached properly to the side of the adult bed or if the infant moves into the adult bed. Therefore, the task force cannot make a recommendation for or against the use of bedside sleepers.”

          Like

          October 4, 2014
          • A bedside co-sleeper or bassinet by the bed is also my first choice as a mom and based on the evidence on safe sleep. The AAP is right that there isn’t much research on co-sleeper safety, but some common sense there goes a long way. It lets you stay close to baby for comforting and feeding but gives baby a separate sleep surface with a little physical distance from adult bodies, pillows, and blankets. However, I don’t argue with parents who say that the ONLY way their baby sleeps is touching them and that a co-sleeper just doesn’t work. This is a common reason for bed-sharing, which often isn’t the original plan.

            Like

            October 4, 2014
    • I completely agree with you. I was also wondering about bedsharing prior to this post, and definitely realize that I need to to what is most comfortable for my baby and I, as long as I am making a safe decision.

      Like

      October 3, 2014
  2. Well, I tweeted it for you, so hopefully that gets you a few people reading about it. 🙂

    Like

    September 26, 2014
  3. Reblogged this on Richmond Pediatrics and commented:
    I see many parents co-sleep with their infants and children, usually because of cultural traditions. Science of Mom has been kind enough to review the literature around the risks of co-sleeping and has emphasized these three points for maximum safety:
    Don’t co-sleep on a couch or chair. If you feed your baby on a couch or chair during the night, and you think there is a chance that you’ll fall asleep there, you’ll probably actually be safer – not to mention more comfortable – feeding in bed.
    Don’t bed-share if either parent in the bed has had more than 2 drinks of alcohol. (And as a side note, babies should only bed-share with their parents, not kids, pets, or other adults.)
    Don’t bed-share if either parent smokes, especially with babies younger than 3 months.

    Like

    September 26, 2014
  4. maggie #

    Two completely unrelated to each other comments:
    I understand that the study was specifically focused on SIDS, but the choice to exclude suffication in a study on bed sharing seems to be fraught with the possibility that the results are mis-interpreted. I would argure that the best case, this study is not looking at the risk of mortality with bed-sharing, but instead looking at the impact of bed-sharing on SIDS. Subtle, yet dangerously different when in the hands of main stream media.

    However, a close friend lost a baby to SIDS (in a crib, no blanket,…) For their next child, she went with a co-sleeper in the bed because she couldn’t sleep unless she could roll over and make sure the child was still breathing. It is really good to see the study pointing out the factors that can be controlled to reduce the risk of bed-sharing.

    Like

    September 26, 2014
    • Your first point is really interesting. I think it makes sense to try to separate SIDS from suffocation from a research point of view. The mechanisms are different, so lumping them together could also confuse interpretation of the results. Most of the case control studies in this area do their best to focus on SIDS, and we actually have very little data on suffocation or asphyxiation (which can be seen on an autopsy – SIDS can’t). The research we do have on suffocation doesn’t include a control group so can’t tell us anything about the relative risk of different sleep practices or environments. I think that the assumption is that accidental deaths from something like suffocation usually happen when there are more obvious hazards, and thus, they’re easier to prevent. This is a reasonable assumption, but it would be interesting to see a controlled study on it.

      To your second point – I can’t imagine the anxiety about sleep I would have had I lost a child to SIDS. I would definitely want my baby sleeping close, too.

      Like

      September 27, 2014
      • Your blog says:

        “Both of these findings run counter to previous studies and to the sleep recommendations of the AAP, so they deserve a close look.”

        But the study findings were limited to SIDS, whereas the AAP guidelines are about “SIDS and Other Sleep-Related Infant Deaths” (a quote from the title of the AAP policy statement).

        More quotes from the AAP policy statement:

        “The AAP, therefore, is expanding its recommendations from focusing only on SIDS to focusing on a safe sleep environment that can reduce the risk of all sleep-related infant deaths, including SIDS.”

        “The recommendations outlined herein were developed to reduce the risk of
        SIDS and sleep-related suffocation, asphyxia, and entrapment among infants
        in the general population.”

        Like

        October 4, 2014
        • Right, and I think it makes sense from a policy standpoint to lump SIDS and other SUIDs together for two reasons: 1) they share many risk factors; and 2) it’s often difficult to tell them apart. From a research standpoint, if we’re looking at two causes of deaths with different mechanisms, it makes much more sense to differentiate between the two so that we can understand them better. Case control studies can use a strict criteria to do that, but in the rest of the world, determining whether a sleep-related infant death is truly SIDS or an accidental death is difficult and often varies by jurisdiction (determined by coroners with very little training).

          Like

          October 4, 2014
      • Jenn c #

        Actually there is usually no evidence of suffocation or overlying on autopsy. (Pediatric fellow who recently spent time with the medical examiner.)

        Like

        October 23, 2014
      • gilliantarr #

        I think the prescient point is that conclusions about bed-sharing generally, not strictly SIDS in the context of bed-sharing, are not founded. Saying something like, “The findings provide guidance on the most important risk factors for SIDS and reassurance to parents who choose to bed-share that it may be a safe option if done with care,” ignores that regardless of the association between SIDS and bed-sharing, bed-sharing may still not be safe because of the risk of accidental suffocation and strangulation in bed (ASSB). The more recent declines in SIDS have been largely compensated for by a classification switch to ASSB, so I resist dismissing it so quickly. ASSB hazards are not always obvious, and the studies of bed-sharing in the U.S. show that a large portion of bed-sharing is *not* done safely. Take a look at the review by Ward 2014 for a decent summary of the bed-sharing literature.

        Also, the association between ASSB and bed-sharing is being hypothesized by the proportion of cases exposed compared to the prevalence of the exposure in the population. It is crude and leaves much to be desired, but it is where observational research often starts. The fact that >50% ASSB deaths occur while bed-sharing but that only 13.5% of infants usually bed-share suggests that there is a problem. Studies vary pretty greatly in these numbers, so I hope better studies are on their way to test this association more rigorously, but in the interim we cannot ignore it.

        My take-away is that even if we were to hypothesize that there is no excess risk of injury in the case of:

        -a term infant >=4mo dressed lightly and using a pacifier,
        -in bed only with her well-rested, never-smoker, non-intoxicated mother,
        -sleeping on a firm mattress on the floor with no nearby walls, objects, or cords, and
        -with no bedding or pillows,

        that is not how bed-sharing in the U.S. occurs. It is easy to call for guidelines for safe bed-sharing, but what should those be? Establishing safety (as opposed to risk) is very difficult in this context, and there simply isn’t the evidence in the realm of sleep-related infant deaths to tell people what is safe. We know that bed-sharing as it is done is a risk. It is only a hypothesis that bed-sharing in the absence of risk factors x, y, and z is safe, and it would be irresponsible of the AAP or any other agency to assert it as anything more. Really, the UNICEF pamphlet does what the AAP guidelines do (although the UNICEF one is less complete), they just rephrase known risk factors in terms of safety and use the phrase, “If you decide to share a bed with a baby.” The real difference is the user-friendliness of the presentation. I agree that culturally-sensitive risk reduction messages need to get out there, but I think we need to be *very* careful to not convey that bed-sharing is safe, because no study has shown that.

        Finally, in reference to this study, note the comparison group in the regression. It is all infants who did not co-sleep. To infer what is being claimed from this study, the comparison would need to be against infants sleeping alone whose parents did not smoke and were not intoxicated. There is some evidence that those factors are associated with SUID even in the absence of bed-sharing, so I would want to see the analysis with that bias removed.

        Like

        December 2, 2014
  5. Amy #

    I agree with the above poster Maggie – leaving out suffocation seems to be leaving out the big risk of bedsharing. I’ve read the James McKenna writing, which indicate that bedsharing is protective against SIDS, but I always suspected that the bigger risk of bedsharing is suffocation from soft bedding, blankets or pillows over baby’s head, getting caught between the bed and the wall, or being rolled on. People often conflate SIDS and suffocation, which I think confuses everyone.

    I wound up bed-sharing for a year – we were on the fence about it, but my daughter wouldn’t sleep more than 20 minutes at a stretch anywhere else in the first weeks – even in a co-sleeper with my hand on her – and we were clearly at risk of falling asleep wherever I was holding feeding her at night, so the bed was the safest option for us. We read a lot and got advice from several people on how to make make our bed as safe as it could be – we removed the pillowtop from our otherwise firm mattress, used separate thin & breathable blankets for me & my husband, had only one pillow apiece, and if my husband had more than one drink, he slept elsewhere. I worried about it all the time in the beginning, but I couldn’t see any safer nighttime option for us. I think that the lack of widely-available safe bed-sharing guidelines in the US is really a recipe for disaster. Parents should have those guidelines at the ready right when they go home from the hospital, in case they fall into bedsharing; they shouldn’t be expected to research them on-the-fly in their overwhelmed and sleep-deprived state. Maybe the nurses at the hospital should even include a review of them upon discharge – there were several other items the nurses’ checklist for discussion.

    Like

    September 26, 2014
    • The SIDS/suffocation distinction is confusing, and as I said above, it would be nice to see more research on the suffocation side too. However, it’s also true that the two causes of death seem to share many risk factors, so if you take steps to prevent suffocation, you’ll reduce the risk of SIDS too. And both outcomes are devastating and worth trying our best to avoid.

      I think it is a great idea to provide safe sleep guidelines for both crib and bed-sharing sleep environments on discharge from the hospital. The UNICEF pamphlet I linked at the end of the post is a great model for how to go about that. Guidelines don’t have to say that bedsharing is necessarily safe, but they can acknowledge that many parents bedshare and how to go about making it as safe as possible.

      Like

      September 27, 2014
  6. catherine #

    I’m a little confused by the sentence that begins “Only one (0.6%) SIDS infant…” I went to the article and it’s not clear to me there, either. Maybe it’ll click later, but are “control infants” here referring to those infants recruited from the community? Or to infants (older than 3 months) who died of SIDS who weren’t co-sleeping?

    Like

    September 26, 2014
    • catherine #

      Oh – and thanks for this. I still have a lot of fear about co-sleeping – the traditional/AAP message is pretty ingrained in my mind – but I’d love to feel differently about it. And research helps. 🙂

      Like

      September 26, 2014
    • I know, it took a while for me to wrap my head around that finding too, and it IS pretty surprising compared with previous studies. It helps to see the actual numbers in Table 3 of the paper. Among older infants (>3 months) who died of SIDS, only one (0.6%) was bed-sharing without smoking or alcohol. Most of this older SIDS group was actually sleeping alone, and the rest were in more obviously dangerous co-sleep environments. Meanwhile, 61 control infants (8.5%) were bed-sharing without smoking or alcohol at the time. The control infants were the healthy infants recruited from the community. Bed-sharing was thus much more common in healthy infants than in those that died of SIDS, making it look like bed-sharing was protective. Does that make sense?

      Like

      September 27, 2014
  7. Thank you, thank you for again weeding through the research and publishing something a layperson can understand. That is why I love your blogging so much!
    Well written, timely, and clear. I’ve been waiting for a bed sharing article from you and I’m so happy to see it.

    Like

    September 27, 2014
  8. Michelle Tseng #

    Hi – thanks for your post. I saw the link to your post on my friend’s facebook post. I skimmed the article, as well as the two original articles from which the data for PLoS ONE article obtained. A few questions – What did you think about the authors combining the two studies that were done at two different times and with slightly different methods? Table 3 of the PLoS ONE paper – both the SIDS group and the control group had 10.4% death when co-sleeping with parents without hazards (did I interpret that correctly?). Why is the OR >1 if the two % are the same? Should it not be 1 exactly? I’ve not done OR analysis before so perhaps it has something to do with the raw data? Also I see nothing about corrections for multiple tests – is that not a worry in this type of study? Thanks, Michelle

    Like

    September 28, 2014
    • Hi Michelle,
      I think it is reasonable to be concerned about combining the two studies, but in the scheme of things, I don’t think combining the two datasets is inappropriate for a few reasons. First, the two studies were done by the same group and in the same regions of the U.K., and the methodology was similar (but as you say, not exactly the same). However, the Carpenter study released last year combined 5 studies that differed from one another MUCH more, so I’d say this latest study is a big improvement in that regard (although the Carpenter study had a much higher sample size). Second, it is fortunately getting harder and harder to conduct these studies since the SIDS rate has declined. The 2003-2006 data set is the latest case control data we have, and it was a relatively small study – too small to analyze the data by age as in the current study. Still, you’re right to question it. Even in the decade between these two studies, there were some notable changes in the epidemiology of SIDS, including a greater proportion of cases on sofas and an average younger age of death in co-sleeping infants.

      As to your second question, I’m not stats expert and I’ve never run stats on a case control study, but I’m guessing that the OR is >1 because of the adjustment for age and time of last sleep, which were handled differently in the two studies. It does seem strange, though, that you’d end up with a p-value of 0.07 for a comparison between 10.4% and 10.4%! I’m not sure about your question on multiple comparisons, but I know I don’t usually see mention of this in case control studies. Maybe someone who is more familiar with these statistical methods can weigh in?

      Like

      September 28, 2014
      • maggie #

        Is the SIDS rate declining because of better or more prevalent autopsy techniques that are now categorizing those deaths as suffocation or other known causes? Im my area, there was a famous case of a woman who lost nine children to SIDS, only to find over a decade later that they were all deliberately suffocated….

        Like

        September 29, 2014
        • The biggest drop in SIDS coincided with the Back to Sleep campaign, which started in the mid-90’s and was associated with a very noticeable (and pretty immediate) drop in cases. After that, I think we’ve definitely seen some diagnostic shift. For a long time, SIDS was used as a cause of death much more liberally, and diagnosis really varied between jurisdiction. For example, the local coroner might call a death SIDS even though it was more likely suffocation, in order to spare the parents from blaming themselves too much. Now the CDC has a much stricter criteria, and we’re seeing fewer deaths classified as SIDS and more as accidental injuries like suffocation, plus a great deal classified as “undetermined,” used in instances like where there may be some evidence in the environment for suffocation but it isn’t clear from the autopsy. Between the 2 U.K. studies included in this most recent analysis, the authors also observed a drop in smoking around babies, side sleeping, and heavy bedding association with SIDS deaths. The epidemiology is obviously really complex, which I think does limit the usefulness of historic data in informing current practices.

          Like

          September 29, 2014
  9. I have co-slept with my girl since she was born and it has worked very well. I do think that it is an individualized decision–for example, I’m not a restless sleeper and so I felt comfortable putting her next to me as a newborn.

    Like

    September 29, 2014
  10. Dr Rob #

    Well done! A complicated study synthesized for a lay readership! That’s hard to do!

    Like

    September 29, 2014
  11. Joana #

    Do You know if they tested for artificial feeding as a variable?

    Like

    October 1, 2014
    • Great question, and I’m glad you asked. Breastfeeding was included in the analysis, but NOT breastfeeding did not turn out to be a significant risk associated with bed sharing and SIDS. In other words, formula-feeding and bed sharing in the absence of the big 3 risky behaviors (couch/chair sleeping, alcohol, smoking) did not carry an added risk. Breastfeeding was more common among babies that coslept than those that did not. Among those who coslept, 44% of SIDS infants and 46% of controls were breastfeeding. Among those who did not, 9% of SIDS infants and 25% of controls were breastfeeding. That means that breastfeeding appears to be protective, but only in babies that were sleeping alone. The authors speculate that this could mean different mechanisms for SIDS deaths in bed sharing and solo sleeping infants. Other studies have found breastfeeding to be protective overall, but I’m not sure if anyone else has really looked at the data in this way. For what it is worth, although many bed sharing advocates discourage bed sharing in formula-feeding dyads, I’m not aware of a study that actually shows bed sharing to be more dangerous for formula-fed babies vs. breastfed babies, and this one suggests that it isn’t.

      Like

      October 1, 2014
  12. This blog mentions anthropologist/bed-sharing advocate James McKenna, but it fails to mention that his guidelines are much stricter than the “most important points to emphasize” mentioned in the blog. Seems to me that some poster are treating these “most important points to emphasize” as if they were a comprehensive set of guidelines, but I am not sure if that was what Alice intended.

    Mckenna’s guidelines include: no medications, no drugs, no one who is excessively unable to arouse, no long hair, no excessive obesity.

    Unfortunately, Mckennna fails to to define the criteria for “excessively unable to arose” and “excessive obesity”. Not sure how one is suppose to use vague guidelines, particularly given that Mckenna says it’s important that you and your spouse have as sufficiently strong belief in your bed-sharing practices such that you would be immune of self-blame or blaming each other should your baby die of SIDS. Of course, you can tell if you are far from excessively obese, but I am not sure how you tell if you are excessively unable to arouse on any given night. I am not a sleep expert, but I imagine that this is a function of your level of sleep deprivation and your level of exercise on any given day.

    Here’s a link to Mckenna’s guidelines:

    http://cosleeping.nd.edu/safe-co-sleeping-guidelines/

    Like

    October 4, 2014
    • Many, but not all, of McKenna’s recommendations are factors that have shown up in other case-control studies but weren’t examined directly in the study described in this post. In the current study, the big three factors mentioned were enough to describe the majority of the risk of bed-sharing. My guess is that many of the other risk factors were also co-occurring in cases with co-sleeping on couches/chairs, alcohol, or smoking (i.e. most cases involving drug use also involved one of those other risk factors). “Excessive tiredness” does show up in some studies, as well as the AAP’s guidelines, and I agree that this is really hard to define, especially for new parents!

      I like McKenna’s guidelines because they really emphasize being intentional about making the sleep environment as safe as possible, so I’m glad that they’re widely shared. I like the UNICEF UK info that I posted at the bottom of the post even better.

      Like

      October 4, 2014
      • Another issue is that McKenna’s guidelines seem to me to require flexibility that bed-sharing parents don’t seem to have. It’s my impression that bed-sharing parents must bed-share every night if they follow the UNICEF UK guidelines. If you can’t even use well-researched sleep-training procedures then you certainly can’t stick a baby in a crib for one night with no training or acclimation. That means you can never use medications and you can never be excessively tired.

        It seems unrealistic to expect any parent to follow these bed-sharing guidelines.

        The book “Parenting with Reason” recommends that bed-sharing parents learn how to use sleep training procedures (Ferber) because they may need flexibility, but this is anathema to the UNICEF UK world view.

        Like

        October 4, 2014
      • Another of McKenna rule is that the mattress should be on the floor: “It is important to be aware that adult beds were not designed to assure infants safety!”

        http://cosleeping.nd.edu/safe-co-sleeping-guidelines/

        It’s easy to overlook this rule since it’s not in his bulleted list of rules.

        If one adopts McKenna’s viewpoint, then one should probably stop calling it “bed-sharing” since no typical adult bed should be shared.

        Like

        December 15, 2015
  13. The Carpenter study had 1472 SIDs cases and it was a breastfeeding only study. If I am reading it right, the Blair study had 400 SIDs cases and including breastfed and non-breastfed and using statistical methods to factor out the breastfeeding effect. One would expect the Blair study to have less power. Perhaps one could even show with a power analysis that the Blair study is consistent with the Carpenter study, but I am not sure of that.

    Like

    October 5, 2014
    • I made a mistake Carpenter had only 500 or so breastfed. Only about 25% of his overall sample were non-smokers. The paper tends to focus on results for a slice of his dataset.

      Like

      October 5, 2014
  14. climafuturo #

    UNICEF UK also state in bold that “The safest place for your baby to sleep is in a cot by the side of your bed“, which to my understanding means that bed-sharing is more dangerous (less safe) than a bedside cot. I think this is not usually underlined in the “co-sleeping is safe” propaganda.

    On the other hand, the AAP is just putting in practice the precautionary principle in the face of conflicting evidence. Translating science into policy is not a scientific issue but a political one. What if I now say that prone sleep position is not the cause of increased risk of SIDS, but rather a poor higiene of clothes and sheets? (see e.g. the comment to Carpenter et al 2013 by Goldwater & Bettelheim published 23 May 2013: http://bmjopen.bmj.com/content/3/5/e002299.full/reply) Would it be ethical to check that and criticize the “back-to-sleep” campaign for being conservative?

    Like

    October 6, 2014
    • I absolutely agree that the vast majority of the evidence we have so far supports the statement that “The safest place for your baby to sleep is in a cot by the side of your bed.“ (This recent study sort of doesn’t – for infants older than 3 months – but only in the absence of other hazards). Both the UNICEF and AAP statements emphasize this point. The difference is in how they go on to offer practical advice. I actually think the AAP statement is evidence-based and carefully worded to emphasize situations in which bed-sharing is particularly dangerous and should be avoided. I can appreciate that their role is provide evidence-based advice, not to tell us what we want to hear (and I think that bed-sharing advocates are guilty of the second, which makes their message more popular with many parents). I do think that the AAP could do a better job of recognizing that moms and babies will continue to sleep together no matter what they recommend, and it would be helpful if they also provided evidence-based guidelines for how to make bed-sharing as safe as possible if that is the choice of the family. That’s really what UNICEF UK does, and I like that approach (even while continuing to emphasize that the safest place for baby to sleep is in a crib next to your bed). It’s really about effective communication of risk, and we’re talking about a HUGE range of risk here, depending on bed-sharing conditions and about keeping lines of communication open between families and pediatricians. In practice, I think many pediatricians appreciate these points and are open to these discussions with families, but the AAP could do a better job of providing them with guidance.
      For example, see this page of advice from the AAP to parents. It doesn’t ever mention ways to make bed-sharing more safe; it just says don’t do it: http://www.healthychildren.org/English/ages-stages/baby/sleep/Pages/Preventing-SIDS.aspx
      Also, here’s my Q and A with a pediatrician about this topic: https://scienceofmom.com/2013/05/23/sids-and-bedsharing-a-pediatricians-perspective/

      Like

      October 7, 2014
      • climafuturo #

        I think I would rather be in between the UNICEF and the AAP. It’s true that the AAP don’t properly address the real-life situations in which families bedshare (though it gives some advice on what to avoid when they “emphasize situations in which bed-sharing is particularly dangerous”), but I see the UNICEF leaflet as a bit of promoting bed-sharing, in the sense that the statement that a cot is safer is somewhat hidden in the middle of the text, instead of at the very beginning as a starting point. Besides, maybe because I’m not a native English speaker, it’s not clear to me whether by “a cot by the side of your bed” they mean a bedside co-sleeper or an independent cot placed close to your bed.

        By the way, thank you for your great article and your comments. As a father of a poor sleeper toddler who bed-share with her mother, I really appreciate your perspective on these subjects. Also very interesting your summary of the new paradigm-shifter research on gluten introduction and celiac disease. Thank you!

        Jesús R.

        Like

        October 8, 2014
        • maggie #

          To a native English speaker, “a cot by the side of your bed” usually means an independent cot placed close to your bed. Since the UNICEF recommendation seems to come from the UK, I would add that “cot” in Britian is called “crib” or “bassinet” in America.

          Like

          October 8, 2014
          • climafuturo #

            Thank you for the clarifications, maggie!

            Like

            October 8, 2014
  15. Reblogged this on kelle938 and commented:
    I can’t not snuggle with my boobaa

    Like

    November 3, 2014
  16. I’ve nominated you for the One Lovely Blog Award 🙂

    Like

    November 4, 2014
  17. krs #

    This post is so interesting. Are there any theories why cosleeping on a sofa or chair leads to such a dramatic increase in risk of SIDS? I can certainly see why it would be a suffocation hazard, but I am curious about SIDS. Different sleeping positions maybe?

    Like

    April 25, 2015
  18. Really good read. I am bed sharing with my two and a half year old. He had a seizure about a year ago and after the long hospital stay I couldn’t get him to sleep alone anymore. After a while I started taking him in my bed and now he has moved in. I think at the beginning I was more concerned what others would think than about the safety of my child. You automatically wake up if you get too close and he will definitely let you know when he doesn’t have enough space. However, everyone and even doctors were saying he should sleep in his own bed. Well tell that a stubborn little kid. So I think the stigma of your child needs to sleep alone stressed me more out than his safety.

    Like

    October 28, 2015

Trackbacks & Pingbacks

  1. What are the real hazards with bed-sharing? New research for you. | The Sleep Store Blog
  2. Should Your Baby Sleep in Your Room? For How Long? Balancing Sleep, Safety, and Sanity | Science of Mom

Comments are closed.

%d bloggers like this: