Do babies die from sleeping on their tummies?

Most people have heard of the phenomenon of babies who scream for whole evenings, until they eventually fall asleep from exhaustion in the middle of the night. The phenomenon is often called "colic", from the assumption that it is all about digestive problems.  

As a mother of five, I'm skeptical of the "colic" label. My experience is that even though it looks like the babies have a severe stomachache, the cure is not a change of diet, but a change of sleep patterns. Among my children, "colicky" babies have been babies who don't sleep enough during the day. The lack of sleep makes the babies wide awake and tormented instead of sleepy in the evening. I'm not alone in this conclusion. I learnt it from this mother of nine at first.

So, enough daytime sleep is the solution. Unfortunately, it is also the problem, because sleeping is dangerous for babies. In Western countries, Sudden Infant Death Syndrome is the most common cause of infant mortality past the neonatal stage. It is the most common cause of death for children between 1 month and 1 year of age. 

It was even worse until the early 1990s, when twice as many children died in their sleep for no apparent reason. By that time, a lot of research was going on, and it concluded that infants who slept on their tummies were more likely to die unexpectedly than those who slept on their backs. Most infants slept on their tummies because Western governments advised parents to put their children to sleep on their tummies. Western governments changed their advice and death rates due to SIDS decreased rapidly.

In theory, this is a success story where science finally overcame prejudice. In practice, this is one of those areas where citizens never become good enough for the state. Governments all over the Western world recommend that babies should sleep on their backs, on a firm surface without any blanket. Like this (minus the pillows): 

According to my experience, that is a recipe for infant sleep deprivation. If I want a baby to wake up because it is not time to sleep, I put him in exactly that position: On his back without a blanket. I find it hard to imagine how parents who actually only let their baby sleep that way go through their child's first year. 

So it seems that parents have a difficult choice to make: Sleep deprive your baby and minimize their risk of death. Or let your baby sleep, but increase their risk of death. When my fifth child refused to follow the advice and sleep on his back, I dove into the question: Do I really risk my baby's life if I let him sleep the way he prefers? Obviously, recommendations for back sleeping saved babies' lives in the 1990's and still do. But does that mean that I double my child's risk of dying if I put him on his tummy? It turns out that nobody seems to know, really.

A numbers problem

The first problem is: When babies die from SIDS, nobody knows what they die from. That is the meaning of SIDS, a diagnosis of exclusion. When grown-ups and older children die it is almost always possible to find a reason. Babies remarkably often die without  a known cause.

Some coroners have started to question the concept of SIDS altogether. There is always a cause, they claim. Mostly suffocation. 

Those disagreements naturally lead to flawed statistics. If it is a matter of taste whether to apply the SIDS label or not, comparisons between times and locations will be more or less useless. This article says that the "decline in SIDS" reported over the last decades might be at least in part a numbers game, because coroners can't agree when to apply the SIDS label. In some locations, babies no longer die from SIDS, but from undetermined causes. 

Still, some of the decrease in SIDS look very real. If I for example look at Swedish statistics over infant deaths, there was something strange going on in the 1980s. From 1950 when the statistics begin until 1990, death rates decreased a bit almost every year in most age brackets, except 1 month to 1 year. In that age bracket deaths instead increased from 1980 to 1990. 

In order to improve the relevance of their statistics, American authorities show deaths from SIDS, Unknown cause and Accidental suffocation and Strangulation in bed together. Then it is possible to see that for example the remarkable differences between different racial groups is probably not a classification issue: Native American babies die unexpectedly at rates at 212 per 100 000 births. Black babies die a lot too at 186 per 100 000. Hispanic babies die unexpectedly at only 54 per 100 000, less than white babies who die unexpectedly at rates of 85 per 100 000. So even though statistics are shoddy, there clearly are important environmental factors behind sudden infant deaths. 

SIDS is mostly a problem for the poor. But not for the poor exclusively: Elon Musk's firstborn son died from SIDS. 

Does tummy-sleeping kill babies? 

No one still knows why babies between 1 month and 1 year die so much when they sleep. Sudden Infant Death Syndrome is related to sleeping in a prone position, to co-sleeping, to overheating, to tobacco smoke. Several different causes of death have been proposed. 

1. Suffocation. Children who die in their sleep are often found with their faces covered. For that reason suffocation is a proposed cause of death. Some researchers think that when air circulation is poor, infants rebreathe the same air again and again. The air gets loaded with more and more carbon dioxide. Most children and adults wake up when they inhale too much carbon dioxide. However, some genetically vulnerable children don't wake up. Instead they pass out and die. 

   Asphyxia could also be due to airway compression in certain sleeping positions. 

2. Over-heating. SIDS occur more during the winter than during the summer and researchers believe that children might die because they get too warm when they sleep.

3. Infections: About half of children who die from SIDS have a seemingly trivial infection.

4. Some genetic illnesses were included under the SIDS label before they were discovered, but now that doctors know they exist the specific illnesses occur on the death certificates instead. This is one of the reasons why SIDS has become a less common diagnosis: Children who before would have been classified as SIDS victims, are now diagnosed with specific diseases. There might be more such clarifications to come.

Experts believe SIDS occurs due to a combination of an infant being vulnerable and environmental stressors. Premature babies and babies with low birth weight are over-represented among SIDS victims. The Triple-Risk Model was launched in 1994. It says that SIDS occurs when three factors are present simultaneously: An underlying vulnerability in the infant, a critical developmental period and an exogenous stressor, like overheating or asphyxia. According to the Triple-Risk Model, SIDS does not cause death in normal infants but, rather, only in vulnerable infants with an underlying abnormality. SIDS is uncommon in babies younger than a month, peaks around three months and becomes uncommon again after six months. There seem to be a critical window when some babies are sensitive to environmental factors. If all babies are raised without exogenous stressors like prone sleeping position during those months, also the vulnerable children will survive in most cases, according to this theory. 

The question is why tummy sleeping, which many of us who are now middle-aged survived very well as infants, is a stressor at all. Researchers say it could be something with suffocation. Many unexpectedly dead children are found in circumstances that suggest that they died from suffocation. That is a reason to suspect that the excess deaths among tummy-sleeping children is due to asphyxia: When children sleep in a prone position they are more likely to get their noses into the surface underneath them. Prone position could also obstruct their airways (but curiously, prone position improved the condition of hospitalized covid-19 victims). 

A study of children in Tasmania found that children who slept in a prone position on a kind of soft and fluffy mattress filled with a certain type of tree-bark ran a 20 times higher risk of dying from SIDS compared to the average rate. Children who slept prone on an ordinary foam mattress only ran three times higher risk of SIDS. Several studies show that the association between SIDS and prone sleeping is stronger in the day than in the night. As a parent I immediately speculate that it is because babies fall asleep everywhere during the day, but mostly sleep in a bed or a cradle during the night. That way, more children would lie prone on unsuitable fluffy surfaces during the day than during the night. I can't find any study that specifies whether the excess daytime deaths in tummy-sleeping babies were linked to fluffy surfaces. 

So what is then my child's risk of dying if I put him on his tummy but always avoid fluffy surfaces (and pillows and stuffed animals)? I really don't know. No one else seems to know either. No one has investigated what happens when people break the number one rule but try hard to stick to most of the other rules. 

I also wonder, am I alone in discovering that many children prefer sleeping on their tummies? Do other people actually follow the advice and put their babies to sleep on their backs? Only two thirds do, according to a study. Alarming, according to a scientist behind the study. Relieving for someone in my position: If a lot of people actually do like I do and disregard the advice and rates of SIDS are still low compared to the early 1990s, that means that a lot of parents break the rules and get away with it. As it is, I think nobody knows how many babies actually sleep on their tummies, because as soon as an issue becomes tabooed, people will be less willing to tell the truth about it. Western governments recommend that babies sleep on their backs, in their own bed without a blanket. We can be reasonably sure that most babies do not sleep like that. We just don't know in what way people act against those recommendations.

One recommendation that people act against openly is the anti-co-sleeping recommendation. Co-sleeping with other people is more or less as much of an additional risk factor for SIDS as a prone sleeping position. In this study, bed sharing was no better than a prone sleeping position. This article summarizes studies. Curiously, there seems to be no taboo against bed sharing. People I know admit openly that they co-sleep with their infants without the smallest trace of shame. Probably it is because bed-sharing is seen as beneficiary in itself, while people who place their babies in the "wrong" position are assumed to be uneducated. I guess that the back-sleeping recommendations are partially responsible for the current trend to co-sleep with babies: If babies are not allowed to sleep as they like, they will wake up more often. Then parents will try to find other ways to calm them in the night, like sleeping in the same bed. 

Can baby breathing monitors mitigate the risk? 

SIDS almost always occur when an infant sleeps. Prone position makes many babies sleep better. Could it be that prone position is more dangerous because sleeping well is in itself dangerous for a certain minority of babies? 

The current proposed solution is to keep all babies in a state closer to awakenness in order to prevent them from dying in their sleep. It seem to work at least partially, but to a cost: Not sleeping enough causes babies great distress. It also causes parents and siblings distress and loss of productivity. How many accidents are caused by sleep-deprived parents? How much less attention is saved for siblings when parents need to handle a sleep-deprived baby? How are relations in families affected when a sleep-deprived baby needs to be consoled an important part of the time? 

It would be great if there was a method that could enable babies to sleep well without increasing their risk of dying. Can baby monitors tell caregivers when a baby sleeps too well to be safe? 

Again, nobody knows. Here is a good article that summarizes the knowledge about breathing monitors that makes exactly this conclusion. Governments all over the Western world recommend against breathing monitors. Why do governments have to issue recommendations about things they don't know? Until the 1990s, governments recommended parents to put babies on their tummies and that seems to have killed children. Now governments recommend against breathing monitors instead. If breathing monitors actually turn out to decrease infant deaths, how many babies will have died needlessly because of the anti-monitor recommendations? 

The baby monitor issue is a typical case of absence of evidence rather than evidence of absence. Studies to detect the effect of breathing monitors in the general population would need to be very big and very expensive: A breathing monitor costs about 100 dollars. If the research program is going to buy monitors to the about 10 000 participants needed to find any effect, it will cost a lot of money. Recommending people to buy one themselves before anyone has been able to prove they are efficient is probably "unethical" in some way. So we have the classical vicious circle: The effect of the monitors can't be assessed because people can't be told to buy them because their effect hasn't been assessed. 

According to the article above, monitor manufacturers have a lot of data, but haven't shared it with researchers: Expensive monitors connect to people's smartphones and send all data they collect about things like oxygen levels, breathing, back or tummy position. Hundreds of thousands of such monitors have been sold but the data has not been analyzed. Manufacturers themselves claim to have received no reports of children who died when their monitors were used.

A team of researchers published a study in 2002, where parents of high-risk infants in Georgia had been given breathing monitors. Although the children were selected for being high risk, they died from SIDS at a rate lower than average in Georgia at the time. 

“If a kid’s not breathing for more than 30 seconds, I would like to know, if that was my kid", Pediatrician Gary Freed, one of the authors of the study, is quoted saying. 

That last quote summarizes my opinion about breathing monitors: If my child stops breathing, I don't know if I would be able to do anything about it. But even if I could do nothing, I would like to know about it.

I also wonder: If SIDS at least in part is caused by suffocation, how could breathing monitors not help? Suffocation is a gradual process. Wouldn't children stop breathing and moving before they die from lack of oxygen? The only way monitors could not help is if babies continue moving until they are brain-dead or otherwise impossible to resuscitate. That could be the case, of course, but we don't know that it is. Pregnant women are advised to be observant of decreased fetal movements. Apparently, fetal deaths are often preceded by fetal distress. Why wouldn't then infant deaths be preceded by infant distress? The fact that researchers don't know how that distress manifests itself doesn't mean it doesn't exist. 

In Justine Musk's account, SIDS was gradual:

"By the time eBay bought PayPal in 2002, we had moved to Los Angeles and had our first child, a boy named Nevada Alexander. The sale of PayPal vaulted Elon's net worth to well over $100 million. The same week, Nevada went down for a nap, placed on his back as always, and stopped breathing. He was 10 weeks old, the age when male infants are most susceptible to SIDS (Sudden Infant Death Syndrome). By the time the paramedics resuscitated him, he had been deprived of oxygen for so long that he was brain-dead. He spent three days on life support in a hospital in Orange County before we made the decision to take him off it. I held him in my arms when he died."

Nevada Alexander apparently died from a lack of oxygen. Could his (wealthy) parents have done anything to resuscitate him sooner if they had known about it 30 seconds after he stopped breathing?

One concern is that monitors are simply bad. This story says so: The advanced devices which promise to show the child's heart rate and oxygen saturation rate of the blood are cumbersome to use and often give false alarms. For several reasons, no one wants a monitor that gives false alarms. And why should I know my child's heart rate and oxygen levels, if I also don't have a very clear guideline about when to perform CPR or call an ambulance? 

On the other hand, I want to know whether my baby breathes, because I know that babies are supposed to breathe. So I have always opted for the non-intrusive breathing-only monitors. They consist of a plate to be placed under the mattress and a unit that registers movement. These simple devices have never given us any false alarms. In the world of gadgets it seems to be a general rule that the fewer things a gadget tries to achieve, the better it will achieve them. That rule also seems to apply to breathing alarms. We use this battery-powered one, which can also be used outdoors since it doesn't need grid electricity (good if your baby likes to sleep in a pram).

Personally, I use the breathing alarm as a baby-is-alive machine. With my first child, who accepted to sleep in a supine position, I didn't have a breathing monitor. Instead I frequently held my hand under his nose until I felt a stream of air, just to get some peace of mind. With the breathing monitor, I don't need to disturb my baby that way. I can see the green light blinking from several meters away. I don't know exactly what it means, except that my child is moving. At any moment, when it tells me my child is alive, it might really tell me about my child's death-struggle. But at least it says that now, at this moment, my baby is not yet dead. If that is the only thing I can know, I prefer to know that rather than nothing at all.

In real life

Sleeping is dangerous for babies. Sleeping less intensely has appeared to relieve, although not entirely take away the danger. Governments and researchers, who don't have to meet the sleep-deprived babies in person, overall think that is a reasonable solution. As a parent, I think it is not. The SIDS research from the last few decades made great progress in identifying the problem. Now more research is needed to identify a solution. 

I wrote this post more than half a year ago, when I had a rather new baby. I didn't publish it because I was a little afraid of scaring readers away with my parental nerdiness. However, it totally aligns with Wood from Eden's new goal to become the most successful of unsuccessful blogs. Since this is a text I would have liked to find on the internet myself, I put it on the internet, plain and simple.