Evidence that roomsharing prevents SIDS is extremely weak

Little newborn baby boy sleeping in round crib with canopy

The American Academy of Pediatrics is finally acknowledging that many principles of lactivism are contradicted by the scientific evidence. Sadly, they have replaced old unsupported recommendations with a new unsupported recommendation.

In August, the AAP finally acknowledged that The Baby Friendly Hospital Initiative, a hospital based effort to promote breastfeeding, leads to serious iatrogenic infant injuries and deaths. The relentless promotion of prolonged periods of skin to skin contact and forced rooming in (by closing well baby nurseries) has led to an epidemic of preventable infants deaths from smothering in the mother’s bed and skull fractures caused by falling out of it. Moreover, the lactivist insistence that co-sleeping improves breastfeeding rates has led to infant deaths from SIDS (Sudden Infant Death Syndrome) at home.

Sadly, the new recommendations, while an improvement, continue to promote the prejudices of lactivists that women should be as close to their babies as possible for as many hours of the day as possible in place of the actual scientific evidence. Specifically, the AAP strongly recommends roomsharing despite the fact that the scientific evidence for a protective effect is extraordinarily weak.

According to the new guidelines:

It is recommended that infants sleep in the parents’ room, close to the parents’ bed, but on a separate surface designed for infants, ideally for the first year of life, but at least for the first 6 months.

There is evidence that sleeping in the parents’ room but on a separate surface decreases the risk of SIDS by as much as 50%.

Actually, as far as I can determine, there is little evidence that sleeping in the parents’ room prevents SIDS. There are four case control studies, two from 2005 and 2004 and two from the 1990’s that show an association between infants sleeping in the parents’ room and decreased risk of SIDS. But as basic statistics teaches us, correlation is not causation.

1. The 2005 study is Bedsharing, roomsharing, and sudden infant death syndrome in Scotland: a case-control study that took place in Scotland involving 123 infants who died of SIDS between January 1, 1996 and May 31, 2000, and 263 controls.

The authors found:

Sharing a sleep surface was associated with SIDS (multivariate OR 2.89, 95% CI 1.40, 5.97). The largest risk was associated with couch sharing (OR 66.9, 95% CI 2.8, 1597). Of 46 SIDS infants who bedshared during their last sleep, 40 (87%) were found in the parents’ bed…

In other words, the association between roomsharing and prevention of SIDS was so minor that they didn’t even include it in the abstract.

What did they find when they looked at roomsharing?

Separate room not sharing was not associated with a risk of SIDS on univariate analysis (OR 1.32 95% CI 0.67, 2.60) but became a risk on multivariate analysis (OR 3.26 95% CI 1.03, 10.35)… Further stratified analysis showed that separate room was associated with a significant risk of SIDS only if a parent smoked (OR 12.2 95% CI 2.25, 66.4) and not if parents were nonsmokers (OR 1.25 95% CI 0.16, 10.06)

Roomsharing was protective ONLY when parents were smokers. That’s not evidence that roomsharing is protective against SIDS.

How about the other citations?

2. Sudden unexplained infant death in 20 regions in Europe: case control study was published in 2004.

…Highly significant risks were associated with prone sleeping (OR 13·1 [95% CI 8·51–20·2]) and with turning from the side to the prone position (45·4 [23·4–87·9])… If the mother smoked, significant risks were associated with bed-sharing, especially during the first weeks of life (at 2 weeks 27·0 [13·3–54·9]). This OR was partly attributable to mother’s consumption of alcohol. Mother’s alcohol consumption was significant only when baby bed- shared all night… About 16% of cases were attributable to bed-sharing and roughly 36% to the baby sleeping in a separate room.

Curiously the authors found that a history of roomsharing but not room sharing at the time of death was associated with relative risk of 0.48, while a history of roomsharing and roomsharing at the time of death was associated with a relative risk of 0.32. The authors acknowledge that they have no idea how roomsharing decreases the risk of SIDS death let alone how a history of previous roomsharing could decrease the risk of death. That suggests that confounding variables may lead to a spurious association.

3. Co-sleeping increases the risk of SIDS, but sleeping in the parents’ bedroom lowers it is part of a 1995 Norwegian monograph, Sudden Infant Death Syndrome: New Trends in the Nineties. I have been unable to access the monograph.

4. Babies sleeping with parents: case-control study of factors influencing the risk of the sudden infant death syndrome is a 1999 British case-control study.

The findings are shown in this chart:


The authors acknowledge problems with interpreting their data:

… Certain factors characteristic of infants found in the parental bed were systematically different from those where the infant was found in a cot: bed sharing infants were much younger, few were put down in the prone position, and few were found with their heads covered. Conversely these factors were reversed among infants who slept separately from their parents.

That’s not a definitive result.

So we have 3 studies with widely disparate findings that indicate potential association, but not causation.

The potential protective effect of roomsharing on SIDS fails to meet the majority of the Hill’s Criteria typically used to determine causation.

  •  The strength of the association is unknown.
  • The dose-response relationship is unknown.
  • The potential protective is not consistent. It varies widely among the studies and in one study applies only to parents who are smokers.
  • No one has offered a remotely plausible mechanism by which roomsharing could prevent SIDS.
  •  The possibility of a chance finding orconfounding variables has not been ruled out.
  • There are no experiments or even prospective studies that look at this issue.

In other words, the claim that roomsharing is associated with protection against SIDS is based on data that is weak and conflicting. More importantly, there is no evidence that roomsharing is the cause of any protective association.

Roomsharing is not harmful. No babies will be hurt by following the recommendation.

What’s at risk is the credibility of the AAP and other physicians. When we make recommendations based on weak data, we face the likelihood that the recommendations will be overturned by better research. That has happened repeatedly with recommendations about breastfeeding.

In 2016 we are finally acknowledging that practices beloved of lactivists — exclusive breastfeeding, co-sleeping, pacifier bans and formula restriction — have risks. Thankfully, the AAP is revising its recommendations in response. The last thing they should be doing is making new recommendations that are no better supported by scientific evidence than the old ones.