CDC study does NOT show bed sharing is dangerous for infants
In what can only be described as an incredibly irresponsible publicity stunt, the Centers for Disease Control has sent out press releases claiming that a new study shows than infant bed sharing quadrupled the rate of infant suffocation and strangulation. The study, "US Infant Mortality Trends Attributable to Accidental Suffocation and Strangulation in Bed From 1984 Through 2004: Are Rates Increasing?" shows nothing of the kind. The overall rate of accidental unexplained infant suffocation and strangulation has not changed AT ALL. The only thing that has changed is the method of classifying infant deaths.
This finding is so important that it deserves to be repeated. There is NO evidence that sharing a bed with a parent increases the risk of accidental infant suffocation.
If there is no evidence that bed sharing increases the risk of infant suffocation, why did the CDC send out press releases suggesting that it does? A claim that parents are routinely suffocating their infants is much more likely to get publicity than merely pointing out that new guidelines for classifying unexplained infant death have led to more deaths being classified as unexplained suffocation and fewer deaths being classified as sudden infant death syndrome (SIDS). The total death rate did not change at all; only the classification changed.
This study does have important implications, but not the ones that the press release claimed. The study raises the possibility that the dramatic decline in SIDS attributed to putting babies on their back is not a real decline, but simply an effect of changing the classification of infant deaths.
Let’s look at what the study does not show:
The study does NOT show that there has been an increase in unexplained infant death. The rate of unexplained infant death has barely changed at all from 1998 to 2004.
The study does NOT show that there has been an increase in accidental suffocation or a decrease in SIDS. Both types of death appear the same on autopsy, therefore, there is no objective way to distinguish between the two after the death has occurred. The decision on how to classify the death is subjective, based purely on the decision of the person who filled out the death certificate.
The study does NOT provide a causal link between infant bed sharing and suffocation. There is no evidence in the study that bed sharing causes infant suffocation, because there is no objective evidence in the study of the actual cause of death. Moreover, the category of accidental suffocation is quite broad.
ASSB [accidental suffocation and strangulation in bed] includes suffocation by (1) soft bedding, pillow, or waterbed mattress, (2) overlaying or rolling on top of or against infant while sleeping, or (3) wedging and entrapment of an infant between 2 objects such as a mattress and wall, bed frame, or furniture; and strangulation by asphyxiation, such as when an infant’s head and neck become caught between crib railings.
In other words, most causes of death in this category have nothing to do with bed sharing at all.
So what really happened?
The authors themselves are quite honest about what they actually studied:
…Recent evidence showed that the decline in SIDS, from1998 through 2001, was offset by an increase in ASSB [accidental suffocation and strangulation in bed] and cause unknown deaths, suggesting that there has been a change in the way these [sudden unexplained infant deaths] are classified and reported…
In this study, we explored trends in infant deaths attributed to ASSB since 1984, before and after the release of the 1996 national guidelines, and assessed how the trend indicating fewer SIDS deaths might be explained by the trends showing increases in ASSB and cause unknown deaths…
What did the study actually show?
…[T]he overall [sudden unexplained infant death] rate showed no important increasing or decreasing trend from 1998 through 2004 …The pattern observed for the other [sudden unexplained infant death]-specific causes of death during this time did reveal an increasing trend for [accidental suffocation and strangulation in bed] and cause unknown and a declining trend for SIDS.
In other words, although the overall death rate remained the same, the proportion of deaths attributed to accidental suffocation rose, and the proportion attributed to SIDS declined concomitantly.
As the authors themselves point out:
The impressive decline in SIDS during the 1990s has been credited to national efforts promoting a supine sleep position for infants. Yet since 1998, safe-sleep prevention efforts may have had little effect in reducing total [sudden unexplained infant] deaths overall. Instead, the relatively stagnant total [sudden unexplained infant death] rate together with the observation that declining SIDS rates are being offset by increasing [accidental suffocation and strangulation in bed] and unknown cause rates suggests that the way these deaths are classified has changed. This change in classification and reporting has been observed in recent studies.
The following graph (modified from Figure 2 in the paper) makes this clear. Overall unexplained deaths have not change. The only thing that has changed is that the proportion of deaths attributed to accidental suffocation has increased while the proportion attributed to SIDS has declined by the same amount.
The author’s conclusions are quite modest:
Infant mortality rates attributable to ASSB have quadrupled in the last 2 decades, with the most dramatic increase in rates occurring after 1996. The reason for the increase is unknown …
It would be more accurate to add that it is also unknown if this represents a real increase or merely reflects a change in the classification. One thing is quite clear, however; this study does NOT show that infant bed sharing leads to an increased risk of accidental strangulation and it is irresponsible to suggest or imply that it does.
TIP:
Enter the amount, and click "Tip" to submit!
aliquot
Roger Fallihee
madcelt
Norwonk
John Moore
Esther AR
Salon.com
Comments
In other words, if the baby died on its face, it was reported as accidental strangulation, and if it died on its back, it was reported as SIDS?
I do hope that they find that the "back to sleep" business isn't necessary. We have a national epidemic of flat headed babies!
"In other words, if the baby died on its face, it was reported as accidental strangulation, and if it died on its back, it was reported as SIDS?"
What appears to have happened is that the criteria for SIDS have been made much stricter. Therefore, deaths that were attributed to SIDS in the past are now classified as "accidental suffocation or strangulation in bed."
In addition, ASSB does not even mean that the baby was sleeping in a bed, or that the infant was smothered. For example, ASSB includes babies getting their heads caught between the bars of a crib.
around the world, bed sharing is the norm in many places--I think we need to be more sensitive and flexible. In my own practice, I never preach against bed sharing, I just make sure parents are being as safe as possible when they do it.
"Where did you get the press release?"
The main press release was entitle RATES OF INFANT SUFFOCATION, STRANGULATION IN BED INCREASE.
Given all that, I wonder where this paper is coming from, since there is plenty of other solid science behind the sleeping position association. As I said, this is odd - but then the CDC often does odd things, usually with politics behind it.
http://tinyurl.com/aan472
We raised our children in the 80s. At the time breast feeding was not the norm. My wife breast feed both the children and we let them sleep in our bed until they were around two. They both had their own bed, and most nights started out in their own bed, but since my wife breast feed it was easier for me to get up, get the crying child, change them if needed, and bring them back to bed and let my wife breast feed them back to sleep. We got more sleep as a couple and the children slept most the night. I use to joke it was natural child spacing.
How is it that you claim these have NOTHNG to do with bedsharing. How likely is my infant to be sleeping in a water bed if said infant is not sleeping with me? How likely is my infant to get wedged between the mattress and the headboard in a crib with a tight fitting mattress, compared to sleeping in an adult bed with a headboard and no warning about the gap between the board and the mattress?
While I agree that the CDC can be alarmist, I do think that many of these causes of death are more likely to be associated with cosleeping than with sleeping in a safe crib (one with slats only a small distance apart and one without bumpers, pillows, etc). I find your claim that rate changes are due to redifinition to be perfectly credible, but c'mon--admit that many of the causes of ASSB are more likely to be present with co-sleeping.
(And, if you are wondering, I shared a bed with all 3 of my kids for periods of time while they were infants, and they all survived just fine, so I am not anti-bed sharing).
"How is it that you claim these have NOTHNG to do with bedsharing. How likely is my infant to be sleeping in a water bed if said infant is not sleeping with me?"
I don't know about water beds, but many people can afford only one bed and nothing else. Therefore, the infant will sleep in the bed with the parent at night, and by himself during nap times.
The key point is that the CDC implied that PARENTS are accidentally suffocating their children by sharing a bed with them, when there is no evidence of any kind to indicate that is what is happening. First, the overall death rate from sudden unexplained infant death showed no change at all. Second, the purported decrease in SIDS deaths exactly matches the purported increase in accidental suffocation, making it more likely that the only thing that has happened is that deaths are being reclassified.
I want to emphasize that I am not claiming that it is impossible for infants to suffocate because of co-sleeping, just that this study showed nothing of the kind. Essentially, it showed only that a new classification of infant deaths led to a new distribution of possible causes of death. Had the same classification system been in use before 1994, it is likely that there would have been no demonstrable change over the lifetime of the study.
Now that we got that problem out of the way, how about mine ... how do you get the kids out of your bed? Eighteen month old twin boys, restless sleepers ... ridiculously cold winter and our bedroom being the only reliably warm room in the house. No one is getting enough sleep.
I disagree. If accidental suffocation occurs in situations in which infants are likely to be sleeping with parents, then I do think that there is an indirect link between parental behaviors and infant deaths. As I would diagram it with my Research Methods students: co-sleeping --> sleeping on saoft bedding/with headboard/on waterbed (etc)--accidental suffocation.
Where you are spot on is in criticizing the claim of a trend. You are quite persuasive in arguing that the numbers reflect not a change in behavior of parents nor infant deaths, but instead a reclassification of cause of death.
I just finished reading the article mentioned in full-text (ordered it from the library last night along with the thimerosal study and the breastfeeding one). The increase in SUID cases attributable to suffocation is clearly due to diagnostic shift from the SIDS group to the suffocation (AASB) and unknown groups, much in the way the apparent rise in autism cases corresponded to the drop in other psychiatric diagnoses such as mental retardation. But that just means that those relatively high infant suffocation rates were there all along - they were just mistakenly attributed to SIDS rather than the actual cause, and perhaps once more stringent death-scene investigations were required to arrive at a SIDS diagnosis, the coroners were finding evidence that babies who would have formerly been classified as SIDS deaths were actually suffocation victims. And this by no means exonerates accidental suffocation as a by-product of bedsharing - on the contrary.
While we can't know exactly how much of the increase in the suffocation category was due to bedsharing because of the large proportion of missing data on the 2003-04 death certificates, the fact that the largest known cause of suffocation (table 3) is overlaying (exclusive to non-crib sleep surfaces), that the largest known category of deaths occurred in adult beds (27.5%), and that over 50% of the deaths occurred while bedsharing is suggestive there is a correlation. Add to this that the CPSC also came up with increased infant suffocation death in adult beds (Drago and Dannenberg 1999), and that the peak months of age in which AASB occurs correspond almost exactly to the interval in which bedsharing babies, even in nonsmoking conditions, are at increased risk of dying from SIDS/SUDI as compared to non-bedsharing infants (0-3 months)...it's very suggestive that bedsharing in those early months does predispose infants to suffocation death.
You can possibly claim that some or all of these deaths were due to "unsafe" bedsharing (though you'd be hard pressed to prove it with such incomplete info except in cases where sofas, waterbeds or soft bedding were involved), but given the big picture, it is, IMO, highly likely that bedsharing played a role in these deaths.
Just another point - you say that "deaths that were attributed to SIDS in the past are now classified as "accidental suffocation or strangulation in bed." " ...chances are that if there were no clear indications at the scene of death that it was a suffocation death, the infant death would be classified not as AASB, but as "unknown" - another category which increased more than 3-fold percentage-wise (and more than two-fold in actual rates per 100,000 population).
"But that just means that those relatively high infant suffocation rates were there all along - they were just mistakenly attributed to SIDS rather than the actual cause, and perhaps once more stringent death-scene investigations were required to arrive at a SIDS diagnosis, the coroners were finding evidence that babies who would have formerly been classified as SIDS deaths were actually suffocation victims. And this by no means exonerates accidental suffocation as a by-product of bedsharing - on the contrary."
It seems to me that so much data is missing that it is impossible to draw any conclusions about co-sleeping. I'd be interested in your perspective as a pediatrician on the following issues.
1. There's no objective way to tell the difference between SIDS and accidental suffocation (unless the baby is found trapped by sheets or between crib rails). Therefore, the cause of death reflects the subjective impression of the individual filling out the death certificate.
2. The paper does not tell us (because the authors do not know) what proportion of babies were sleeping WITH a parent, what proportion were sleeping in a bed WITHOUT a parent, and what proportion were trapped in sheets or crib rails. It makes a huge difference.
3. The paper does not tell us (because the authors do not know) why parents were sleeping with their babies. Is there a difference in death rates for co-sleeping for philosophical reasons, and co-sleeping because of chaotic life conditions such as the entire family sleeping in one bed because they cannot afford a second bed or crib?
4. The paper does not tell us whether parents who were sleeping with the baby at the time the baby died were impaired by drugs or alcohol. Are unimpaired parents accidentally smothering their babies or does smothering only occur in the setting of parent impairment.
Obviously, we don't want even a single baby to die, and clearly there are some things that should be avoided whether the baby is sleeping in a bed or crib, such as pillows, comforters and blankets. Nonetheless, it seems to me to be irresponsible to suggest or imply that unimpaired parents who co-sleep for philosophical reasons risk accidentally suffocating their babies.
I never understood the big "debate" around bed sharing. It has risks, just as using a crib does.
Possibly this study doesn't (will take some time to read it more thoroughly later) but international experience such as these articles relating to New Zealand Coroner's cases on the subject (with four out of seven cases being heard sharing a bed and the other three were placed unsafe positions) I don't think it very safe to claim it doesn't pose a risk or it's simply a change in classification. Even though there might be a degree of subjectivity, I don't think it would extend as far as completely missing a genuine case of SIDS where there is no evidence of ASSB or a death with clear evidence of ASSB. The New Zealand report of a coroners hearing below shows cases, but judging from the comments in the second, related article they are likely to be classed as SIDS rather than the true cause which was that ASSB did occur and caused the death of the infants, with 50% of SIDS cases being associated with bed sharing. It's possible that misclassification is attributing ASSB cases as SIDS, with a significant under-reporting of ASSB in the past due to limitations in categorisation etc.
“Mums sob as joint inquest held: Mothers’ sobs filled courtroom 7 at Wellington District Court as inquests were held on how their babies died in bed, possibly because of “unsafe sleeping environments”, a pathologist said. In four of the seven cases the babies were sharing a bed with others. Three were in bassinets, but face-down, which probably caused them to suffocate. Wellington coroner Garry Evans, who has condemned bed-sharing as a deadly practice in at least 15 other cases during the past eight years, reserved his findings. He told the families yesterday that the inquests were not a witch-hunt. “There is no suggestion these babies received anything except tender loving care. We are simply here to see what contributed to their deaths.” Judy-Anne Tito, who accidentally smothered her son Nephi, said she knew it was risky sleeping with her tiny twins “but it was the only way they would settle…..”. http://www.stuff.co.nz/4787740a23918.html
“Dangers of sharing bed with babies not clear: More consistent messages about the dangers of bedsharing with babies are needed, child health experts say. Auckland University professor of child health research Evan Mitchell said bedsharing absolutely increased risk of babies dying from sudden infant death syndrome (Sids). “About 50 percent of the deaths are occurring in a bedsharing situations,” he told Radio New Zealand….” http://www.stuff.co.nz/4788593a11.html
Second, all arguments regarding what part co-sleeping plays in infant mortality rates within the US are moot. Why? Because there have been no comprehensive studies performed, factoring in all variables, such as bedding, stress levels, alcohol & drug consumption, etc., to determine what effect co-sleeping has on overall infant mortality rates. Data can be extrapolated from various statistics, however, without having a controlled, well documented study, the likely hood of anyone ever being able to make an informed, comprehensive argument affirming or opposing co-sleeping is next to nothing. Bare bones statistics are easy to manipulate, for any interested party. Controlled, detailed case studies, while they can be skewed, are less open to interpretation.
Sober, educated (at least self educated), and attentive parents choose to sleep with their babies, for various reasons. Given that we don't force parents to use cloth diapers (sposies have nasty chemicals that can cause a whole host of health problems), or mandate mothers breast-feed for x number of days to prevent their baby from contracting a potentially life threatening case of some nasty bug, or jail mothers with children who have an unknown allergy that becomes fatal, I think making parents feel guilty for a (relatively) safe form of care and bonding with their child is ridiculous and unnecessary.
I'm a family physician, not a pediatrician, but I'll do my best. Since there's a lot more to this than can be stuffed into a single comment, you can also go to my blog ( http://mainstreamparenting.wordpress.com) and look at the posts in the category of "infant sleep".
Amy: "1. There's no objective way to tell the difference between SIDS and accidental suffocation (unless the baby is found trapped by sheets or between crib rails). Therefore, the cause of death reflects the subjective impression of the individual filling out the death certificate."
Unless you think all coroners in the US are horribly biased or incompetent, there's no reason to believe the results are distorted towards the "accidental suffocation" pole. On the contrary, when coroners started working according to more rigorous and objective criteria to examine death scenes, they found the bias formerly went the other way: a subjective assessment was more likely to misattribute actual cases of suffocation to SIDS rather than the reverse.
Amy:"2. The paper does not tell us (because the authors do not know) what proportion of babies were sleeping WITH a parent, what proportion were sleeping in a bed WITHOUT a parent, and what proportion were trapped in sheets or crib rails. It makes a huge difference."
According to the death certificate info in table 3, at least 51% of the deaths were in the context of bedsharing. I doubt many parents put their tiny babies to bedshare with siblings only, so it seems reasonable that those 51% were sleeping with at least one parent.
Amy:"3. The paper does not tell us (because the authors do not know) why parents were sleeping with their babies. Is there a difference in death rates for co-sleeping for philosophical reasons, and co-sleeping because of chaotic life conditions such as the entire family sleeping in one bed because they cannot afford a second bed or crib?"
The issue isn't why they share a bed, but whether the bed is shared safely or not. I don't think you can just assume that because parents have decided to share a bed with their infant on philosophical grounds, the parents will necessarily provide a safe cosleeping environment on a consistent basis. Some of them will not because they don't know all the 'rules' (and some of the most popular websites don't provide all of them; for example, Dr. Sears doesn't mention parents shouldn't smoke, and he also recommends placing the baby between the mother and the wall - even though wedging in adult beds between mattress and wall is a common cause of infant suffocation.
Some sources convince parents that bedsharing protects against SIDS (despite the complete lack of evidence for this) and that any suggestion to the contrary is a conspiracy by the crib manufaturers. Besides, one's "mommy instinct" will kick in if something happens during the night.
It's pretty amazing how many cosleepers on philosophical grounds think that their good intentions alone will keep their baby safe - a quick search for images of the family bed will show people cosleeping with baby near pillows, near heavy duvets, with obese mothers (also a no-no), or with siblings. Baddock et al studied a group of 40 white, highly-educated women who coslept with their small infants; 3 (7.5%) were smokers, and 4(10%) also slept with other children. (see http://mainstreamparenting.wordpress.com/2008/07/17/the-big-sids-coverup/).
Amy:"4. The paper does not tell us whether parents who were sleeping with the baby at the time the baby died were impaired by drugs or alcohol. Are unimpaired parents accidentally smothering their babies or does smothering only occur in the setting of parent impairment."
We don't know the answer to either of those questions. We DO know from various other studies that alcohol and drug use are not the only risk factors in cosleeping situations (and some - like smoking or the baby's head being covered during sleep - are of greater magnitude). I think it'd be a big stretch to assume that all the cases of suffocation deaths, either in this study or generally, occur in the context of alcohol or drug use. Three's also the category of "excessively exhausted" parents...which would be pretty common, I should think, in the baby's early weeks, and difficult for parents to self-assess.
"Unless you think all coroners in the US are horribly biased or incompetent, there's no reason to believe the results are distorted towards the "accidental suffocation" pole."
It's not a question of being biased. It's simply that the terms have been redefined. Nothing has actually changed.
"According to the death certificate info in table 3, at least 51% of the deaths were in the context of bedsharing. I doubt many parents put their tiny babies to bedshare with siblings only, so it seems reasonable that those 51% were sleeping with at least one parent."
First, approximately half were not in the context of bed sharing, which means that at least half of the deaths had nothing to do with bed sharing at all. Second, it is not seem reasonable to me to assume that those infants who were bed sharing were sleeping with a parent. It could have been other children in the family, grandparents, mother's boyfriend, etc. It is difficult for us to imagine a situation in which an entire family has only one bed, but those situations do exist.
"I don't think you can just assume that because parents have decided to share a bed with their infant on philosophical grounds, the parents will necessarily provide a safe cosleeping environment on a consistent basis."
We shouldn't assume anything, including assuming that it is co-sleeping itself that led to the deaths. Sudden accidental suffocation is not really a cause of death, it is a term of exclusion, and it certainly does not tell us what actually killed the child. Not only do we lack a basis for assuming that co-sleeping itself causes an increase in accidental suffocation, we don't have a mechanism of action. Certainly this particular paper provided neither.
I want to make it very clear that I am not claiming that co-sleeping can't cause accidental suffocation. My point is that it was irresponsible for the publicists to suggest 1. that the incidence of accidental suffocation has increased, because the paper shows nothing of the kind, and 2. that co-sleeping causes an increased risk in suffocation, because the paper doesn't show that, either.
State Medical Examiners' offices funding varies across the states, death scene investigations by responding police vary incredibly, existence and quality of mortality review committees are not standardized, etc.
For SIDS, the best autopsy results are with pathologists who have received special training in pediatric autopsies. But not all towns/cities/states have people with this expertise.
As Amy points out, as new classifications are devised, people start switching diagnostic categories, but not always based on good evidence.
http://pediatrics.aappublications.org/cgi/eletters/123/2/533
I've practiced safe bed sharing with both of my kids and I believe it to be safe with proper precautions.