Intended for healthcare professionals


Making informed choices on co-sleeping with your baby

BMJ 2015; 350 doi: (Published 02 February 2015) Cite this as: BMJ 2015;350:h563
  1. Peter J Fleming, professor of infant health and developmental physiology ,
  2. Peter S Blair, reader in medical statistics
  1. 1Centre for Child and Adolescent Health, School of Social and Community Medicine, University of Bristol, St Michael’s Hospital, Bristol BS2 8EG, UK.
  1. Correspondence to: P Fleming peter.fleming{at}

Avoid smoking, alcohol, recreational drugs, and co-sleeping on the sofa

For most of human history, and in most parts of the world today, sharing a sleep surface (“co-sleeping”) with the primary caregiver—almost always the mother—has been the normal experience of most human infants. Such close contact has benefits for the baby, including facilitating breast feeding, and even in modern Western societies infant mortality is significantly lower among breast fed infants.1 Routine co-sleeping is common in Western societies and is thought to be increasing with increased breast feeding; studies have shown a close bidirectional relation between breast feeding and co-sleeping.2 On any given night in the UK, around one quarter of infants under 6 months of age spend part or all of the night sharing a sleep surface with a sleeping parent.3

Against this background is the recognition that, since the large fall in the number of sudden infant deaths that followed the “Back to Sleep” campaign, the proportion of such deaths occurring while co-sleeping has risen.4 In some countries, notably the United States, many professionals have accepted that co-sleeping substantially contributes to unexpected infant deaths and should be discouraged.5

For parents who smoke, have drunk alcohol, or have taken recreational drugs co-sleeping is clearly associated with an increased risk of sudden infant death, and co-sleeping on a sofa or armchair even without these factors carries a substantially increased risk ,6 as does co-sleeping with an infant who was born preterm or at low birth weight.7

What is less clear is whether co-sleeping by breastfeeding mothers, in the absence of the other risk factors noted above, is associated with any increased risk of unexpected infant death.

Two recent meta-analyses used patient level data collected in large scale population based case-control studies to investigate whether co-sleeping in the absence of other known risk factors was associated with an increased risk of unexpected infant death.8 9 The first study pooled data from five case-control studies and concluded that, even in the absence of parental smoking, alcohol use, or drug taking, co-sleeping was associated with an increased risk of infant death.8 However, the second study, which used data from two UK studies, concluded that the increased risk of infant death was almost entirely accounted for by the effects of parental smoking, alcohol use, and inappropriate sleep environments (such as co-sleeping on a sofa).9 The first meta-analysis was larger but had limited data on parental alcohol consumption and used imputation to estimate the incidence of alcohol consumption from whole studies and potentially heterogeneous populations where this information was missing. The second study was limited to data in the UK but both contributory case-control studies had detailed information on smoking, alcohol use, and sleep environment.

Because of this uncertainty the UK’s National Institute for Health and Care Excellence (NICE) was asked in 2014 to investigate the evidence linking co-sleeping and the risk of unexpected infant deaths and produce evidence based guidance. The resulting guidance, published in December 2014,10 goes some way to meet the needs of parents but stops short of providing the correct emphasis given the risks involved.

The NICE panel reviewed several large case-control studies as well as the two meta-analyses.8 9 Although the NICE report recognised the extreme difficulty in investigating contributory factors to rare and unexpected events such as sudden infant deaths, it was critical of the lack of information on various potentially contributory factors in the studies reviewed. The panel did not, however, contact study authors to identify whether such information was available.

The guidance notes the reported association between co-sleeping and unexpected infant death but recognises the complex nature of such an association, and emphasises the lack of evidence of a causal link. However, it accepts that “some of the reviewed evidence showed that there is a statistical relationship between SIDS [sudden infant death syndrome] and co-sleeping.”

While recognising that co-sleeping may be intentional or unintentional, the guidance recommends that parents should be informed that “there is an association between co-sleeping and SIDS,” that this association “is likely to be greater when they or their partner smoke,” and that the association “may be greater with parental or carer alcohol consumption, or drug use or low birthweight or premature infants.”

Unfortunately the guidance does not mention the reported increased risk associated with co-sleeping on a sofa or armchair, despite the fact that this relation was found to be highly significant in four studies from the UK and Ireland6 11 12 13; a high prevalence of deaths on sofas has also recently been reported in Wales.14 Bed sharing is relatively common in the general population whereas sofa sharing is not (prevalence <1%); given that recent UK studies suggest a sixth of sudden infant deaths occur in such an environment,4 6 14 this is a rare but lethal infant care practice that warrants far more emphasis, especially if parents are under the misconception that co-sleeping on a sofa is no more dangerous than sharing a bed.

The guidance should be welcomed for making transparent the lack of evidence of a causal relationship between co-sleeping and sudden infant death, while encouraging health professionals to be honest in informing parents of the potential risks arising from co-sleeping—particularly in relation to parental smoking, alcohol, and drug use.

By conflating all forms of co-sleeping and not recognising the much greater risks reported from co-sleeping on a sofa than on a bed, the guidance has missed an opportunity to inform parents and thus potentially to protect infants. Seeking to dissuade parents from this practice carries no identifiable risk and may help reduce the number of avoidable sudden infant deaths.


Cite this as: BMJ 2015;350:h563


  • Competing interests: We have read and understood BMJ policy on declaration of interests and declare PSB is an honorary adviser to Unicef UK on night time infant care practices.

  • Provenance and peer review: Commissioned; not externally peer reviewed.


View Abstract

Log in

Log in through your institution


* For online subscription