UK United Kingdom

Safety first: reducing the risks of sleeping with your baby

With so much contradictory information about whether it’s safe for mothers to share a bed with their baby, it’s easy to see why parents are confused. The release yesterday of a paper in BMJ Open added…

Sharing a bed with your baby is not inherently dangerous if the known risk factors for SIDS are eliminated. Image from

With so much contradictory information about whether it’s safe for mothers to share a bed with their baby, it’s easy to see why parents are confused. The release yesterday of a paper in BMJ Open added to this confusion, by warning all parents to simply avoid bed sharing because it increases the risk of SIDS (sudden infant death syndrome) five-fold.

We know that sharing a bed with a baby can be risky in certain circumstances, such as when bedding is soft and the baby can overheat, if there are siblings and pets on the bed, and if you smoke, or are under the influence of drugs or alcohol. Falling asleep with the baby on an armchair or couch can also be dangerous.

But for the many mothers who want to bed share with their baby, risks can be minimised to provide a safe sleeping environment.

Outdated and inconclusive

The BMJ Open study provides weak evidence about the risks of bed sharing for low-risk mothers who breastfeed and provide a safe sleeping environment.

The researchers selectively analysed data from five international studies from 1987 and 1998. It doesn’t include other data sets from more recent studies, which use more comprehensive data collection procedures and definitions.

A further weakness is that 60% of the information on alcohol and illicit drug use was missing, which the authors replaced using complex statistical methods. There was also no data on use of prescribed and over-the-counter medications, which are likely to be common in the first few months after birth, when the SIDS risk is highest.

The authors claim there is an increased risk of SIDS for breastfed babies under three months, even in a safe bed-sharing environment. This claim is confused by the fact that this breastfeeding data was either not available in the original studies or it merely stated whether babies were breastfed at birth.

Another problem is that the number of control babies bed sharing may have been underestimated, as those who may have shared a bed for a portion of the night and didn’t wake up in the parental bed would have been classified as “room sharers”, leading to inflation of the risks associated with bed sharing.

This study should therefore not be used as basis for public health policy for low risk breastfed babies bed sharing in safe sleeping environments.

What does the recent research say?

Sharing a bed with your baby has not been found to be inherently dangerous if the known risk factors for SIDS are eliminated.

A 2009 study from the south-west of England examined 80 SIDS cases over four years and compared them with randomly chosen control families. In their analysis, the researchers excluded known risk factors such as smoking, alcohol and substance use, and falling asleep with baby on a sofa or armchair. When these risks were eliminated, the rate of bed sharing in the SIDS group was actually lower (6%) than the control group where there were no deaths (10%).

A 2009 Alaskan study also found that 99% of SIDS that occurred while sharing a bed involved known risk factors such as maternal smoking or sleeping with someone affected by substances.

Breastfeeding protects against SIDS. Image from

Bed sharing and breastfeeding

For many mums, sharing a bed with their baby is instinctual and a cultural norm, with up to 50% of all babies and 70% to 80% of breastfed babies having shared the parental bed at some time.

We know that bed sharing can help settle infants and it promotes successful breastfeeding. Research has shown that breastfeeding has a direct protective effect against SIDS, reducing the risk by half. This level of protection increases for babies who are breastfed for more than eight weeks.

The protective effect of breastfeeding against SIDS may, in part, relate to the higher prevalence of bed sharing among breastfed babies. Breastfed babies who share a bed with their mother benefit physiologically and have fewer pauses in breathing and more stable temperatures, oxygen saturation and heart rhythms than babies who sleep alone.

Breastfed babies who bed share also naturally tend to lie on their backs with their mothers instinctively cocooning them and protecting them from bed covers.

Current guidelines are counterproductive

It’s important to have public health policies that will reduce the rates of infant deaths associated with sleep. But current state and national public health polices show little insight into the realities of night-time parenting. They say it’s okay to breastfeed babies in bed but they must be placed in their own cot before they fall back to sleep.

With so many parents sharing a bed, such rigid information can be counterproductive. A 2012 Australian study reported health professionals believed that “by insisting on a strict no co-sleeping stance” they “alienate mothers”, which results in them “shutting down” communications with health-care professionals.

This view was confirmed by the mothers interviewed in the study, with one describing the policy as a “scare tactic”. As a result of the policy, some mothers said they had resolved not to disclose their intentions to bed share. This was particularly so for women from cultures where bed sharing was considered normal.

Families need information tailored to minimising the risk to their babies. Image from

Mothers and babies who breastfeed and bed share in a low-risk environment should not be discriminated against – they should have access to information that is tailored to minimising the risk to their babies, by health professionals who are well informed and sensitive to their needs. Only then can the channels of communication remain open and risks of infant death be minimised.

State and national government health authorities urgently need to redraft their policies to enable open discussion of safe sleep and provide options for mothers who intend to bed share.

In the meantime, if you share a bed with your baby or are considering doing so, you can find a detailed discussion of these issues here and evidence-based information on normal and safe infant sleep on the ISIS (Infant Sleep Information Source) website.

Sign in to Favourite

Join the conversation

13 Comments sorted by

  1. Chris O'Neill

    Retired Way Before 70

    I heard in the media that they were blaming the risk on bacteria. Would that mean you shouldn't sleep with anyone?

  2. Andrei Cleland

    Digital Video Production

    Wonderful, wonderful journalism.
    Our child sleeps with us and would always, no matter what doctors or the mainstream media said about it, because to us, it is the obvious, most lovely experience. (It also happens to be the easiest!) However it's nice to get some evidence that we're not as unscientific or backwards or foolhardy as some think. It's not rocket science that if smokers and drinkers co-sleep the risks will be much, much higher. How can these considerations be side-stepped and muddy the statistics? and then, how can these stats be used to justify expensive government funded campaigns to breath fear into parents so that they no longer trust their instincts?
    The 6% LESS risky is where my money is on. However does that factor in how much more likely a bed-sharing family is to be breast feeding longer, and strengthening an infant to decrease the likelihood of SIDS even further?

  3. Lydia Isokangas


    I concur, bed sharing with a breastfed baby is a very safe and natural thing to do. I didn't do this with my first baby until he was almost 6 months old because I listened to all the rubbish information out there. As soon as he was in the bed and feeding as he needed during the night my husband and I could actually sleep and get some rest! My next two children didn't even bother with the cot - it became a soft toy storage bucket and a change table. Consequently my husband and I felt far more rested and patient with all the troubles that infants can give you.

    How many deaths and nasty accidents happen with parents who are too tired because they're not bed-sharing?

  4. Richard Hockey

    logged in via Twitter

    One would have thought that the excellent article by Carpenter et al would be the last word on bed sharing along with the recent paper by Vennemann et al (J Pediatr 2012;160:44-8) presented compelling evidence of the risk of bedsharing even in the absence of alcohol and tobacco use. But yet again we see TC being used as a platform to promote a particular viewpoint by way of selective use of evidence. This article deliberately sets out to muddy the waters regarding the important public health message which was the subject of the Carpenter paper. It is high time that TC subjects its articles to some form of peer review to weed out misuse such as this.

    1. Catherine Fetherston

      Associate Professor, School of Health Professions at Murdoch University

      In reply to Richard Hockey

      I think it is very important to note in response to this that even Venneman et al., in their meta-analysis, concluded from their findings that "It is not clear whether public health advice should advise against bedsharing in general or just particular hazardous circumstances in which bedsharing occurs" (Vennemann et al J Pediatr 2012;160, p47).

    2. Nonie Jekabsons

      Tree Spotter at -

      In reply to Richard Hockey

      it's "natural" to breastfeed and share a bed. It is instinctive. Our (now adult) children are healthy. If studies into this are to be taken seriously then consider other factors such as breast vs bottle feeding, (including the relative 'carrying capacity' of breast sizes - let's not ignore that one), feeding regimes, parent lifestyle re sleep patterns, type of bed, room ventilation/fresh air... long term psychological effects... the list goes on. Humans are animals, not statistics.

  5. Angela Conway

    Practice Manager

    As far as I understand the evidence, falling asleep with baby in a chair or on a sofa is pretty risky.Should she walk the floor, while feeding or comforting? Perhaps she may fall with baby or drop baby when fatigue overwhelms her? So what should exhausted mum do if she is trying to feed her baby ( a most necessary activity especially in first weeks and months!)? I cannot see how retreating to bed, if other risk factors such as drug taking, smoking, soft bedding etc. are eliminated, could not be the sanest, safest, healthiest, most sensible thing to do. There mum can drop off to sleep without the risks posed elsewhere and baby can get fed and comforted. We need to be clear about real relative risks. And clear about the real, practical realities of caring well for small babies. So please, lets have this debate.

  6. Sarah Vaughan

    General practitioner

    A quick question while I hopefully get my brain together for a longer comment at some point: Could you clarify where you got the 6% and 10% figures from the 2009 Blair et al study? I've read through the study but can't find anything that breaks the figures down by place of sleeping in low-risk situations. This may well be because I'm trying to read it after midnight and should be in bed!

    1. Catherine Fetherston

      Associate Professor, School of Health Professions at Murdoch University

      In reply to Sarah Vaughan

      Thanks Sarah, you will find the figure you want in the paragraph prior to the discussion in the linked paper. You may also be interested in Peter Blair's follow up discussion paper: Perspectives on Bedsharing at:
      and keep an eye out for responses in letters to the editor of BMJ Open. UNICEF have also posted a good response at

    2. Sarah Vaughan

      General practitioner

      In reply to Catherine Fetherston

      Thanks, Cathy! The authors didn't test that figure for statistical significance, but I just ran it on an on-line calculator and the difference comes out as not statistically significant. Did you calculate it differently?

    1. Catherine Fetherston

      Associate Professor, School of Health Professions at Murdoch University

      In reply to Sarah Vaughan

      thanks Sarah -the blog looks great and certainly supports what is being said by so many of us. The stats you refer to in the Blair paper were calculated within a multivariate logistic regressions model which when all known risk factors were eliminated resulted in very low numbers in each group (5 or 6% compared to 9 or 10%) and is why perhaps significance was not commented on.