Another cosleeping/bedsharing* study came out today, and I expect to see this one getting lots of press, as it well should. It pulls together data from five previous studies, making it the largest study to date on this issue. It clearly lays out the risks for SIDS for babies who share a bed with parents, and it purports to adequately take other risk factors into account in its analysis.
And it falls short — for the very same reasons that I’ve seen previous studies on SIDS risk and bedsharing fall short: it doesn’t take into account enough of the other possible risk factors for SIDS during bedsharing. Therefore, the study will likely, as others have, cause more harm than good in the form of guilt-ridden parents who are at their wit’s end and have no place to turn for practical, realistic advice. It may even – although I cannot provide evidence for this – continue to be a part of the problem in NOT reducing SIDS deaths due to bedsharing.
First, what did the study find? It found that a breastfed baby of nonsmoking parents was at five times higher risk for SIDS if sharing a bed with parents than if sleeping in his or her own crib/bassinet/cot — but wait.** That “five times” number is actually an adjusted ratio of risk. The real risk is 2.7 times, which is almost meaningless unless you know the absolute risks. (For an excellent analysis of the statistical limitations in this study see this UNICEF commentary on it.) So let’s look at the raw numbers instead: Among the 1,472 babies who died of SIDS, 22.2% were bedsharing with parents; of the 4,679 control babies who did not die of SIDS, 9.6% were sharing a bed with parents the day of the study interview. In absolute risk terms without percentages, 8 of every 100,000 room-sharing babies in their own beds will die of SIDS while 23 of every 100,000 bedsharing babies will die of SIDS. Sounds a little scary, right? Let’s take a closer look at the way the study was done, the confounders they included in their analysis and what the authors are neglecting to address at all.
In any study, confounders are the other factors that might influence the result of an association between two things, usually an exposure and an outcome. Typical confounders include age, sex race/ethnicity, socioeconomic status, education level, occupation, etc. Depending on the study, other factors can be important confounders: in most studies about prenatal issues, for example, the woman’s reproductive history (previous pregnancies, miscarriages, abortions, etc.) and “parity,” or history of childbirth, are potentially important confounding factors. A study that does not adequately account for confounders runs the risk of having bias that influences the findings.
This study, published today in BMJ Open (you can read the full paper yourself), re-analyzes the raw data from five previous case-control studies while accounting for several really important confounders. In a case-control study, researchers identify a group of cases (here, the babies who died of SIDS) and compare them to a control group of babies/families with similar characteristics to find out what could account for differences in outcomes (one group died from SIDS; one did not).
The confounders they took into account are really, really important – many of the same ones that I’ve previously not seen adequately accounted for in previous studies. They include whether the parents smoke (a high risk factor for SIDS in any sleeping arrangement) and whether the baby was breastfed. They also took into account mother’s age, the baby’s birth weight, the position the baby was last left in and other important factors, but they only had the data on the mother’s alcohol and illegal drug usage for two of the five studies — a pretty significant limitation to the overall analysis. They also analyzed the deaths associated with sleeping on couches separately so that those numbers did not bias the results of BEDsharing (as opposed to accidentally falling asleep with a baby on the sofa).
It’s also important to keep in mind that the study is retrospective: the authors gathered data after the SIDS deaths. A prospective study follows participants forward and tends to be less susceptible to recall bias and other forms of bias. Further, suffocation deaths are counted as SIDS deaths because the UK Office of National Statistics “found that many of their characteristics were very similar.” I find this conflation disingenuous and unhelpful, but that’s not the biggest weakness of this study. The biggest problem is that authors left out three of the most important confounders. In fact, from what I can tell in the study, they never even considered them, never even asked parents about them. They are: Did you intend to share a bed with your baby (as opposed to accidentally falling asleep with the baby in the bed)? Do you know ways to reduce the risk of SIDS for a bedsharing baby? Did you use any practices to reduce the risk of SIDS while sharing a bed with your baby?
The last two, to my knowledge, have never been asked in a study about cosleeping/bedsharing. Why not? Perhaps they are nonsensical questions to researchers who, along with the public health community at large, have such a blind spot related to bedsharing that they refuse to acknowledge that there is ANY way to reduce SIDS risk in a baby who sleeps with their parents. The belief seems to be that if they repeat it often enough – “Don’t share a bed with your baby” – and emphasize the risk of SIDS enough, then parents will magically start following their advice. They won’t.
Does this sound familiar? The idea that just repeating “don’t do it” will somehow make people stop doing it? Teen abstinence comes to mind: if we just tell teens not to have sex at all instead of teaching them safe ways to avoid pregnancy and sexually transmitted infections, then teen pregnancy and STI rates among teens will plummet, right?
Except that strategy hasn’t worked very well, as the public health community knows. They get sex education right. They have found that abstinence-only education does not have different long-term outcomes than more comprehensive sex ed and that comprehensive programs teaching safe sex and abstinence together are effective. In fact, abstinence-only education might even increase sexual activity, not to mention unprotected sex.
And that is why I have serious problems with this study and similar ones. The researchers are contributing valuable information to the research literature, and they claim to provide information “to enable an informed choice to be made by parents as to whether the risks associated with bed sharing outweigh the postulated benefits.” But parents cannot make an informed choice if the researchers are refusing to study or even acknowledge bedsharing practices that can reduce the risk of SIDS.
By denying parents information on best practices about ways they can reduce risks while bedsharing in addition to encouraging them not to bedshare, the public health community potentially shares responsibility for the continued SIDS cases in bedsharing. Their short-sighted, dogmatic, counterproductive, bang-their-heads-against-the-wall-expecting-different-results thinking might actually contribute to SIDS deaths among bedsharing babies. The study authors wrote, “Our analysis estimates that 88% of bed sharing deaths are attributable to bed sharing, that is, would not have occurred had the baby not been bed sharing.” My question is, how many of those deaths are attributable to parents’ not being adequately informed about the best ways to reduce the risk of SIDS if they were choosing to sleep with their baby?
Parents need sleep. There are some babies who absolutely, positively will not sleep on their own, period. My son was one of them. From day one in the hospital, he WOULD. NOT. SLEEP. by himself. I had two options: my son could sleep with me (which I had sworn I wouldn’t do before he was born), or I could go sleep-deprived for weeks on end. The latter was untenable — and much less safe. If I had continued to rock my newborn to sleep, lay him down in the bassinet and then walk away for the ten minutes it took before he woke up crying, hour after hour after hour, night after night, then I would have been a nervous wreck and a worthless parent. I would have placed my son at greater risk during the day for carelessness resulting from sleep deprivation. I wouldn’t have been able to drive – except that some days I had to, which would have been unsafe. I get dizzy with sleep deprivation; could I have dropped him? Sleeping with my son enabled me to get sleep and was tremendously more efficient for breastfeeding through the night.
I realize I may be promoting a paleofantasy fallacy, but I do think it’s safe to say that mothers, like other mammals, have been sleeping with their babies for thousands of years. I doubt Ms. Cavewoman placed her baby on a bed of leaves in a separate corner away from her at night, and much of the non-Western world sleeps with their babies. (In fact, even “sleep training” is a pretty modern and Western practice because most of the world, for most of human existence, has not had to return to work after six weeks of unpaid leave to conform to a biologically artificial 9-5 schedule.)
Parents need solutions, not shaming. These kinds of studies pile on the guilt for a mother who desperately needs sleep and whose only option might be to sleep with her child, but the public health community refuses to do evidence-based studies into best practices to give her practical advice she can actually use. And I’m leaving aside the fact that millions of mothers want to sleep with their babies, regardless of whether they feel they have to or not, which is a completely valid choice. There are definitely benefits to cosleeping/bedsharing, and there are excellent resources where parents can get information on safe cosleeping/bedsharing guidelines.
To be sure, I am not suggesting there is no increased risk of SIDS among parents whose babies sleep with them. I don’t believe adequate studies have been done to truly find out by taking into account ALL the necessary confounders, but I would not be flabbergasted to find there is an increased risk. However, we take risks all the time, every day. The point of public health is to teach the public how to reduce those risks as much as is possible. Even if we take this study’s findings at face value, 15 additional babies out of every 100,000 will die of SIDS because of bedsharing. Frankly, that is not a high risk, and my risk of endangering my baby through sleep deprivation is likely higher. The risks of dying in a car accident hover somewhere around 1 in 17,000 – without considering drowsiness or distraction by cell phones, etc. If not sharing a bed at all significantly reduces a child’s risk of SIDS and a parent can do that and chooses to do that, that’s great. For the rest of us, the public health community is neglecting its duty in teaching safe, effective ways to cosleep/bedshare while reducing the risk of SIDS or suffocation.