My physician-gaming-group-friends are being swamped by concerned parents because of a new study in Pediatrics that has led to many news articles declaring that swaddling – wrapping an infant tightly in a cloth or blanket – may increase the risk of Sudden Infant Death Syndrome, or SIDS.
But a careful read of the study should help everyone to take a deep breath and calm down.
SIDS is still the leading cause of death in babies age 1 month to one year old. We still aren’t sure what causes it. Risk factors include smoking in parents, prematurity, and being put to sleep on the stomach. This is the reason that back-to-sleep campaigns advocate for babies to be laid down to sleep in a supine, not prone, position.
Babies like to sleep on their stomachs, though, and they can be fussy on their backs. So some parents like to swaddle them, believing that a tightly wrapped infant is a more comforted infant. Others, though, argue that this is bad for babies, reduces spontaneous arousals, and can increase the risk of SIDS.
This new study was a meta-analysis of four case-control studies that looked at how swaddling was related to SIDS deaths. A case control study pairs infants who died of SIDS (cases) with similar infants who did not (controls) and sees how they differ. In this specific case, they were interested in whether infants who died of SIDS were more likely to have been swaddled than infants who did not.
They found that the overall risk of swaddling was not statistically significant. However, the most recent study was found to be heterogeneous, and when they removed it, the relationship became significant, with an odds ratio of 1.38.
That’s the top-line result leading to concerning headlines. But there are a number of factors to consider before anyone panics.
First of all, case-control studies are notoriously susceptible to “recall bias”. This occurs when someone is more likely to remember a fact (like swaddling) after a bad outcome (like SIDS) than after a good outcome (healthy baby).
These studies were also very, very different. Two were from the UK, one was from Tasmania, and one was from Chicago, all over a two decade period. The prevalence of swaddling was also very different, ranging from 35.7% of infants in Tasmania in the 1980’s to 5.7% in England two decades later. It’s likely that other things were very different in those studies as well.
None of the studies really described what constituted “swaddling”, and the ways in which the researchers asked parents about “swaddling” differed greatly from study to study. Analyses controlled for very few other factors, apparently just age in some analyses.
The biggest caveat, however, has to do with positioning. The greatest risk of SIDS seemed to come from being swaddled and then placed prone to sleep. That goes against back-to-sleep practices, and was also rare. Being swaddled and placed on the side to sleep also carried a higher risk. But babies who were swaddled and placed on their backs to sleep had a statistically significant, but very small risk. For instance, 5.9% of SIDS cases were supine and swaddled, compared to 5.2% of controls. SIDS risk with swaddling also increased with age.
But a key message is that of all the babies who died of SIDS, more than 85% weren’t swaddled at all. Less than 6% were put sleep on their backs while swaddled. We’re talking about 38 babies. That’s a very small number to change practice over.
This is also is just an association, not proof of causation. Without controlling for other factors, we can’t know if there’s something else (ie race, socio-economic status, smoking, education, etc.) that’s confounding the results.
For those of you who have never swaddled a baby before, it’s not easy. Keeping them still while you wrap them up tight requires practice. The wrapping needs to be secure as well, because you don’t want loose cloths or blankets in a crib – that’s a hazard in and of itself.
As the authors note, guidance about swaddling advises that it stop either after babies are four months old, or after babies start to rollover on their own. That makes sense. The authors further argue that swaddled babies should be put on their backs to sleep, which we already do.
But there’s nothing in this article that really screams that swaddling – in and of itself – is dangerous. The evidence really isn’t that compelling. If you’re swaddling your newborn well, and placing them on their back to sleep, there’s not much here to worry about.