Hopp til innhold

Systematic review

Bed sharing, pacifier, breastfeeding and cot-death - is there a connection?

Key message

The rate of cot-death for Norway in 2004 was 0,4 per 1000 live births.  Even with such a low rate approximately 20 children die inexplicably every year.

Co-sleeping (baby and mother in same bed) has been cited as a risk factor for cot death.

Some have claimed that co-sleeping during the whole or parts of the night can promote successful breastfeeding.  Studies have shown that the use of pacifier might prevent cot-death but there is concern that the use of a pacifier impedes successful breastfeeding.  It has also been argued that breastfeeding is a protective factor against cot-death.

A multi-professional group conducted a systematic review as part of this Health Technology Assessment.


We systematically searched for studies of experimental and observational design in the following databases: Cochrane Library, MIDIRS, Cinahl, Embase, Medline and Swemed, June 2004.  The search was updated in February 2005.

We assessed studies that had evaluated newborn and infants and the effect of sleeping alone or in parent’s bed, the use of pacifier and/or breastfeeding with the outcomes breastfeeding rates and cot-death.  The studies were critically appraised and included or excluded according to inclusion criteria for study design and methodological quality.


The search identified more than 1000 references.  40 studies were included according to the inclusion criteria.  Only a few studies compared groups that were randomly allocated to one intervention or another.  It is therefore difficult to determine causation.

Researchers have already shown that there is an increased risk for cot-death with co-sleeping if the mother has smoked during pregnancy or if parents smoke after birth, if they have drunk alcohol or taken sedatives.

Five studies assessed the effect of co-sleeping on breastfeeding.  They all showed that the infants that sleep in their parents’ bed were breast-fed more than those who slept alone.  However, it is not clear if co-sleeping encourages breastfeeding or if it is the mothers who breastfeed who choose to have the babies in bed with them.

Research has not shown that co-sleeping is a risk factor for cot-death for infants older than eight weeks.  It is unclear if co-sleeping gives a higher risk for infants less than eight weeks old.  The results from the eleven included studies vary, but eight do not show an association.

Three randomised studies showed that the use of pacifiers by infants probably does not influence overall breastfeeding.  It is however, unclear if an early introduction to pacifiers interferes with breastfeeding duration.  Another unanswered question is whether early introduction to pacifier influences the rates of full and partial breastfeeding.  We do not know if the use of pacifier is a marker for breastfeeding problems or if early use of pacifier leads to less breastfeeding.

All included studies found that the use of pacifier is associated with a reduction in the risk of cot-death.  However, because of limitations in the study design it is impossible to determine if there is causation.

When looking at the effect of breastfeeding on cot-death we included six studies that found no significant association for reduced risk for breastfed infants.  Seven studies showed that breastfeeding could protect against cot-death.  It is unclear whether breastfeeding protects against cot-death.


Because of limitations in the study designs we can not draw any firm conclusions about the factors we investigated.  There is an increased risk for cot-death with co-sleeping if the mother has smoked during pregnancy.  It is uncertain whether there is a reduced risk of cot-death with breastfeeding or with the use of pacifier, if co-sleeping encourages breastfeeding, if co-sleeping is associated with increased risk of cot-death for infants older than eight weeks or if the use of pacifiers influences overall breastfeeding.  It is unclear whether co-sleeping with newborns less than eight weeks of age is harmful.

About this publication

  • Year: 2005
  • By: Norwegian Knowledge Centre for the Health Services
  • Authors Alquist, Ragnhild; Grøgaard, Jens; Hellenes, Kaja; Hofmann, Bjørn; Myr, Rachel; Reinar, Liv Merete; Rognum, Torleiv Ole; Nilsen, Anne Britt Vika; Nilsen, Ellen M; Saugstad, Ola Didrik; Aalen, Odd..
  • ISSN (digital): 1503-9544
  • ISBN (digital): 82-8121-049-4