Co-sleeping: what the research actually shows in 2026

This is the most heated topic in the parenting world and I am going to be unpopular with everyone for a paragraph or two. I am an evidence-led consultant. The evidence on bedsharing is more nuanced than either side of the debate usually admits. Here is what the research actually shows, with citations.

The biological framework

Two decades of anthropological and biomechanical research on infant sleep make the case that humans evolved for some form of close mother-infant nighttime contact. Solitary infant sleep is a modern Western convention, not a biological default.

That is not a moral argument. It is an observational one. The conclusion researchers draw is more measured than the headlines suggest: bedsharing is risky in some configurations and considerably less risky in others. The job of public health messaging is to give parents the tools to reduce risk, not to issue a blanket "never do this" that gets ignored by 40% of US families anyway.

The Safe Sleep 7

The harm reduction framework, developed by lactation consultants and refined through La Leche League, sets out the conditions under which bedsharing risk is lowest. The Academy of Breastfeeding Medicine has published similar guidance.

Lower risk bedsharing requires all of these:

• Baby is full-term and not low birth weight.
• Baby is healthy (no respiratory infection at the time).
• Baby is breastfed (alters mother's sleep architecture so she stays more responsive).
• Mother is sober (no alcohol, no sedating medication, no recreational drugs).
• Mother is a non-smoker (and the bedroom is smoke-free).
• Sleep surface is firm, flat, and clear of pillows and duvets around baby.
• Baby is on their back, in the gap between mother and the edge of the bed (not between two adults).

Remove any one of those and risk goes up materially. Remove two or three and you are in significantly elevated risk territory.

The non-negotiable risk factors

Sofa sleeping or armchair sleeping with a baby. Never. Multiple studies put the relative risk of SIDS or fatal accident here at 10 to 20 times higher than a crib. If you are exhausted and about to nod off feeding on the sofa, put the baby in the crib first. Always.

Bedsharing with smoking parents. Carpenter et al. and multiple US epidemiological studies put the bedsharing risk substantially higher when either parent smokes, even if they do not smoke in the bedroom. The pathway is not fully understood, but the data is consistent across studies.

Bedsharing with alcohol or sedatives. Greatly increases the risk of overlay and reduces the parent's arousal threshold. This is the most preventable risk factor in modern bedsharing data.

Bedsharing under 4 months for any baby who was premature or low birth weight. The thoracic and arousal vulnerabilities are higher.

What the AAP still says

The AAP does not recommend bedsharing. The 2022 update softened the language to recognize that families bedshare for many reasons and provided risk-reduction guidance for those families. This is the closest the AAP has come to harm reduction.

For US parents, the AAP guidance is the operating standard. Some hospital systems and lactation organizations within the US take a more explicit harm-reduction stance closer to the Safe Sleep 7 framework.

The "should I sleep train if we bedshare" question

You can absolutely sleep train without ending the bedsharing arrangement, if that is what you want. The skills being taught are independent falling-asleep skills, which can be taught at nap time in the cot first, then carried into night sleep. The two are not the same conversation.

I have clients who continue bedsharing for the second half of the night and who have taught their babies excellent independent sleep skills for the first half. Whichever you choose, choose it deliberately, and run it consistently. The babies do not care what you choose. They care that you do not flip every night.