Abstract:  Extinction methods of sleep training – namely cry-it-out and controlled crying – are assumed to work via teaching a child to “self-soothe”.  It is under this premise that many parents engage in the practice, believing they are “helping” their child.  Unfortunately extinction sleep training neither teaches self-soothing nor helps a child in the sense of social or emotional regulation.  Furthermore, the success or effects of these methods will be highly variable on two variables discussed at length herein: (1) the child’s pre-existing level of emotion regulation, and (2) the degree of distress experienced by the separation from the caregiver.  At the end of the piece, a discussion of what parents can do when they require changes to the situation is presented, as no family is expected to continue on a path that is causing immense distress.


The idea that using extinction sleep training methods will teach your child to self-soothe is a statement that is treated as truth in our society.  Is it true?  Not in the slightest.  It may get your child to fall asleep without signalling to you, but self-soothing it is not.  The main problem with the common rhetoric on extinction sleep training (outside of the glaring one in which parents are told if they don’t use it, they’re setting their child up for a life of failure and they are “martyrs” as parents) is that it ignores very real and important individual differences in both the development of emotion regulation (of which self-soothing is a component) and the degree of distress experienced by children as they are separated from their caregiver.

All humans (children and adults) struggle with emotion regulation if distress is too high.  Children, especially infants, have the added problems that they haven’t had the experiences we have, their bodies don’t physiologically regulate as ours do[1], and they don’t have the self-soothing skills that we have learned over years of practice.  In short, there is more that causes distress to our infants (especially as everything is new) and they have less capacity to deal with it.  When it comes to nighttime, how our children cope and the degree to which they experience distress will be highly variable, resulting in changes and differences in self-settling.

One way to think of this is that for every level of emotion regulation, there is an associated distress level at which point these abilities cannot be implemented.

Self soothing distress chart

So for any given child, whether or not they are able to self-soothe will be dependent upon (1) their self-soothing abilities at that point and (2) how distressing the event is for them.  I want to talk about both of these to further understand the myth of teaching a child to self-soothe using extinction sleep training followed by a discussion of steps parents can take when they need to implement changes.

The Development and Prevalence of Self-Soothing in Infancy

Contrary to what some believe, infants can actually show elements of emotion regulation, like arching away from a negative stimulus or looking to the source of discomfort, but (a) there is a distinction between emotion regulation in terms of avoidance versus self-soothing (remember I said self-soothing is only a component of emotion regulation?) and (b) much of the research isn’t clear that these are effortful self-soothing activities versus behaviours that are in direct response to the stimuli or situation (for a discussion, see [2]).  Although some infants show clear signs of being able to demonstrate early self-soothing techniques (like sucking on their own hand to calm down), many don’t, and none of them use these behaviours exclusively.

The most common form of “self-soothing” in infancy is to seek out a caregiver, for the act of searching for help is a form of emotion regulation (really!).  After all, us adults when we are overwhelmed also seek comfort in other individuals.  Importantly, even in the infants that utilize other forms of self-soothing, none of them use it exclusively.  All of them also use coregulation, and use it far more frequently than any type of self-soothing[3].

Interestingly, few studies have actually looked at the natural developmental trajectory of emotion regulation in infancy (though see [1] for a good discussion of the foundations for social and emotional functioning).  However, there are, luckily for us, some exceptions to this.  First, a study examining emotion regulation in 3- to 13.5-month old infants found that outside of trying to move away from a negative stimulus, the most common form (and often the first sign) of intentional emotion regulation in these infants was to seek out their caregiver (in the case of this study, mothers) to help buffer their negative emotionality[4].

In another study, researchers examined emotion regulation in 5- and 10-month old infants and found that only 15% of 5-month-olds and 22% of 10-month-olds engaged in “self-calming” behaviour when faced with an unpleasant situation (in this case, having their arms held down in a high chair by an experimenter)[5].  The degree to which children displayed these emotion regulatory behaviours was not predicted by either child temperament (maternal rating) or maternal sensitivity or intrusiveness (gauged in a free-play interaction).  At 10 months of age, far later than most sleep training begins, only 22% of infants were able to engage in “self-calming” behaviours when distressed.  (Note that far more were able to engage in avoidance behaviours like trying to arch away, and more were also able to engage in social referencing for help, but neither of these qualify as the type of self-soothing that is advocated for in sleep training.)

In these studies, however, the mother acts as a source of comfort and the distress is caused by a researcher, not the parent.  In sleep training, the parent is the source of the distress by removing themselves and refusing to provide the comfort or regulation the child desires (and as we have seen, the use of coregulation is the go-to for more infants).  How does this affect things?  We don’t know, but as stated in a piece about a group of neuroscientists working to change the way we think about brain development (and who are against the cry-it-out and controlled crying methods of sleep training):

“And questions like who was involved in the event may have more significance than simply the presence of the hormone alone because it indicates which parts of the brain will be involved in processing the stress. In the case of children, the stress initiated by a caregiver may be more significant in terms of brain neuroscience than the stress associated with, say, little Timmy’s school-yard friend Ginny, who knocks him off the swing set from time to time. That stress may cause the boy some difficulty, but the stress associated with an attachment figure leaving him at night to cry alone in his crib may be more significant.

The child’s brain can only process that as an abandonment—it has no other way to make sense of it—and while the results of that abandonment vary considerably in any given household and certainly don’t sentence the child to a lifetime of despondency—or, worse, mediocrity—the child’s brain experiences a lesson it simply cannot order or regulate except by associating care with something other than the parent.” (Source)

Source: Cartoon Stock

Source: Cartoon Stock

Although the author here states that the child’s brain has to associate care with something other than the parent, I would argue the bigger issue is that the brain associates the parent with something other than just care.  Given the role of coregulation in emotional development, I don’t quite know how this can be justified, even if the effects vary widely between individuals due to other circumstances.

In line with this, there is research looking at 6-month-old’s use of self-comforting behaviours in response to several scenarios involving frustration led by their parent[6].  As we saw in the aforementioned research, at 5-months of age, when a researcher causes distress, 15% of infants were able to use self-calming behaviours.  When frustrations were parent-led (e.g., removal of toys or arm restraint by a parent), the attempt of self-comforting behaviours was found in 16% of infants, but they were barely used, with their presence noted for an average of 2% of the time the infants were distressed.  As said by the researchers, “[W]ith few infants displaying the behavior for a short amount of time and most infants not displaying the behaviour.” (p.185)

What can we conclude?  Though some infants are able to show some self-soothing behaviours, they are not regularly used, even (or especially?) when a parent is the source of distress.  Nothing in the literature suggests that infants can be forced to adopt emotion regulation without it being learned over time and in fact, the main form of regulation for infants is coregulation, which is impossible for the child to attain when extinction sleep training is used.

The Role of Distress

The second element to be discussed is the role of distress.  Even for infants who are older and may have some self-soothing skills, the question that remains is how the role of distress influences their ability to use these skills.  There is a bizarre assumption in the sleep training world that all children will experience being alone in a crib the same way (because extinction sleep training is inherently for those who have their children in another sleeping environment; it’s difficult to do CIO when you’re right there, next to your baby, as in a co-sleeping relationship).  That having the door shut is the same to all and that if one child wimpers for 2 minutes, it must mean it isn’t distressing for anyone.  The kids that cry more aren’t truly distressed, just “protesting”.

The problem is that we know there are huge individual differences in the expression and experience of distress.  Much of the research on the development of emotion regulation highlighted above has found as much.  For example, in the study looking at parental frustration for the infant, researchers were able to categorize infants by the degree of frustration they experienced, with some being easily frustrated and others being significantly less so[6].  Furthermore, the research by Stifler and Spinrad, looking at emotion regulation in 5- and 10-month olds, examined the role of prior excessive crying, presumed to be a temperamental trait, and found differences in the experience of distress based on that categorization[5].

Although we shouldn’t need science to tell us there are differences in the experience of distress by an infant, luckily we have it.  What we now need to look at is how this experience of distress influences the use of self-soothing behaviours.  After all, if the idea of sleep training is to get a child to learn to self-soothe, we’d better hope it does in the face of distress, right?

The problem with this is that in all of this research, the relationship between the experience of distress and the use of self-calming or self-soothing behaviours is inversely related.  That is, the greater the distress, the less likely an infant is to be able to use self-soothing behaviours.  In [6], the negative association was very strong and this was not mediated or explained by maternal behaviours when the child was distressed (i.e., sensitive versus intrusive) or by child temperament.  This is in line with older research that found the child’s degree of reactivity has a huge influence on their ability to use any regulatory behaviours (not just self-comforting, but even seeking out assistance or arching away)[7][8].  Indeed, the research on the development of emotion regulation consistently finds that infants are unable to use these behaviours except at low levels of reactivity or distress[8].

The role of temperament was also made clear in both [5] and [6] in that infants that were easily distressed showed lower levels of emotion regulation regardless of the immediate distress.  Furthermore, babies that were rated as being easily frustrated showed greater physiological reactivity to distress and in turn, less physiological regulation to distress.  The cluster of behaviours seen in the research (which also includes less attentiveness and distress to novelty) has the researchers suggesting a specific temperament pattern in which the development of emotion regulation may be slower and require more assistance.

Source: Unknown

Source: Unknown

If we take these findings and relate them to more modern research on sleep training per se, we may get a clearer picture.  First, we should look at these issues of self-soothing and distress with respect to the research by Middlemiss and colleagues that looked at cortisol reactivity and synchrony during cry-it-out sleep training[9].  These researchers found that during the extinction program, infants cries were associated with a very high spike of cortisol during the separation before sleep.  Notably, by day 3, infants were no longer crying, but showed the same cortisol spike that they did on day 1.

One of the biggest concerns on this research, however, has been the high standard deviations in the group, which is indicative of high variability between infants.  I hope the section here on distress helps to clarify this in that distress is highly variable.  Some infants become very distressed to certain situations that others find only mildly distressing, if distressing at all.  This speaks to the influence of temperament in the use of sleep training and any potential long-term effects.  Indeed, research on parental care and stress responsivity more generally shows that infants who are more reactive show greater elevations in cortisol under conditions of less-than-optimal care (defined as sensitive and responsive care)(see [10] for a review).

The second element of this study worth discussing is the effect on synchrony.  In the Middlemiss and colleagues research, the physiological synchrony between mother and baby that was high and statistically significant on day 1 was dampened to the point of non-significance on day 3.  Synchrony is important to development as it influences a host of outcomes pertaining to well-being (see here for a discussion), and is supposed to be present in caregiver-infant dyads.  The loss of this synchrony was in fact the main finding of note in this particular piece of research because it highlights how disruptive extinction sleep training is.

If we go back to role of coregulation in emotion regulation, we can clearly see the possible negative effect of a loss of synchrony.  Coregulation works because the infant is able to be calmed physiologically (even if not emotionally) by the caregiver; however, there needs to be a relationship that allows for this.  A total lack of synchrony would mean the parent is unable to calm the distressed child at all.  In cases where synchrony has been dampened, the ability of parents to physiologically soothe their infant will be diminished, but not absent.  Returning to the quote above on how the child’s brain makes sense of cry-it-out in terms of parental abandonment, the findings on synchrony are in line with the view that it may have a significant effect.  (The question now becomes, for how long?)

The second bit of recent research worth mentioning is that which examined the efficacy of sleep training done at home.  Whereas most of the research on how effective extinction programs (i.e., cry-it-out and controlled crying) have been done in labs where there are very strict procedures in place, how this looks at home may be very different, and indeed that’s exactly what researchers found[11].  Instead of the quick fix that it is advertised to be, extinction programs at home were largely unsuccessful, with half of the people who tried it reporting having to try it 4 or more times and 40% of families reporting needing to use it for longer than a week before either success or giving up (nearly 13% used it for more than a month straight).  On top of all this, 40% of families said it did not improve their infant’s sleep at all.  Notably, the parents who reported that the technique was highly stressful for their infant were also those who found it least effective; knowing the link between distress, self-soothing, and coregulation, is this any surprise?

Putting It All Together

What does all of this mean?  I can only tell you how I take it, you are free to come to your own conclusions.  For me, this tells us two big things: (a) the idea that extinction sleep training methods will work for all children ignores a key variable, namely the degree of distress experienced by an infant during these methods, and (b) that this distress is a key indicator as to how the process of self-settling will impact the child, the parent, and the later emotion regulation.

Extinction works by taking away responsiveness until the child gives up.  At its worst – straight up crying-it-out – it is a method that tells parents to ignore their child’s distress at all costs.  No matter how securely attached their child is prior to this method, the lack of a parental presence means the child will experience a cortisol spike, the duration of which is unknown.  How this affects their developing brain will be subject to a host of other factors – how long this goes on for, the temperament of the child, the degree of attachment after sleep training, and so on –so of course we can’t make definitive conclusions.

We can, however, say that self-soothing isn’t what’s going on during extinction sleep training.  Whatever it is – self-settling or giving up – isn’t in the realm of emotion regulation as we know it.  Given the link between distress and ability to regulation one’s emotions, causing a child immense distress is not a means to elicit or teach self-soothing.

A child who does not experience distress, however, is likely to be less affected by the process.  Does this mean it should be endorsed for these children?  I still say no.  First, extinction methods aren’t often used if a child is going to sleep without distress on one’s own.  What point would there be?  So with this, it becomes difficult to argue that the method would even be appropriate to begin with.  (If people are using it to stop night wakings, then I would really caution against that – if you have a child going to sleep independently and waking in distress, than I’d hazard a guess to say something else is going on because there are no cues you’ve been using that the child might be looking for.)

The question many parents will have now, then, is what the heck are they supposed to do?  Am I suggesting that you sit and struggle with a situation that isn’t working for you?


In a nutshell, here is what I would recommend:

  1. If you’re at your wit’s end and need immediate help and you have some cash, hire a night doula for a couple nights. If you don’t have cash, see if a trusted family member or friend can get you through a couple nights.  Get yourself caught up on sleep and know that you have someone loving there to care for your child at night (although if you’re exclusively breastfeeding and not pumping, you may need to be woken for a couple feeds).
  2. If you’re not at your wit’s end (yet) or you’ve gotten caught up and don’t feel the immediate need to make a drastic change, check out these reasons why an infant’s sleep patterns may not be “normal” and if you feel your child’s patterns fit this, start looking into the causes behind it. Often it’s feeding or health issues that result in increased wakings and distress and once they are resolved, infant sleep improves dramatically.  Of note, if your child was premature or had health problems at birth, you may experience more wakings than normal as your child adapts and that’s okay.
  3. If your child’s sleep qualifies as “normal” but still poses problems for you and your family, then you should look to some of the gentle resources listed here. These are not quick fixes and will take time and patience to implement.  Like all changes, not all will work with all kids (just as extinction methods have been found to be less than optimal in terms of success when done at home, as mentioned above).  You will need to start making changes with the knowledge that it may take some time before you see a noticeable improvement.
  4. If at any time you hit your wit’s end, return to point 1 for a couple nights until you feel yourself able to cope again. At no point is it good for you to be at risk of harming your baby or yourself; however, what this stage is will vary person-to-person.  I know people that struggled but were never at their wit’s end with hourly wakings for months, but I know people who have felt like they were going to lose their mind with a baby who was waking 3-4 times a night.
  5. At all times remember that your child isn’t doing this to make you hate them or resent them. They really aren’t.  They are trying – albeit sometimes unsuccessfully – to communicate with you.  They may protest a change, but when done with responsiveness, you can help them transition as well without causing undo stress.  Just remember you should be working together and finding compromise that works for all.


[1] Porges SW.  Physiological regulation in high-risk infants: a model for assessment and potential intervention.  Development and Psychopathology 1996; 8: 43-58 (Note: This is not just an article on high-risk infants, but rather elaborates a model for all infants with a focus on how to identify high-risk ones.)

[2] Eisenberg N, Spinrad TL.  Emotion-related regulation: Sharpening the definition.  Child Development 2004; 75: 334-9.

[3] Campos JT, Frankel CB, Camras L.  On the nature of emotion regulation.  Child Development 2004; 75: 377-94.

[4] Rothbart MK, Ziaie H, & O’Boyle CG. Self-regulation and emotion in infancy. In N. Eisenberg & R.A. Fabes (Eds.) Emotion and its regulation in early development: New directions for child development, No. 55: The Jossey-Bass education series (1992) (pp. 7-23). San Francisco: Jossey-Bass Publishers.

[5] Stifter CA, Spinrad TL.  The effect of excessive crying on the development of emotion regulation.  Infancy 2002; 3: 133-52.

[6] Calkins SD, Dedmon SE, Gill KL, Lomax LE, Johnson LM.  Frustration in infancy: implications for emotion regulation, physiological processes, and temperament.  Infancy 2002; 3: 175-97.

[7] Kopp C.  Regulation of distress and negative emotions: a developmental perspective.  Developmental Psychology 1989; 25: 343-54.

[8] Stifler CA, Braungart JM.  The regulation of negative reactivity: Function and development.  Developmental Psychology 1995; 38: 448-55.

[9] Middlemiss W, Granger DA, Goldberg WA, Nathans L.  Asynchrony of mother-infant hypothalamic-pituitary-adrenal axis activity following extinction of infant crying responses induced during the transition to sleep.  Early Human Development 2012; 88: 227-32.

[10] Gunnar MR, Donzella B.  Social regulation of the cortisol levels in early human development.  Psychoneuroendocrinology 2002; 27: 199-220.

[11] Loutzenhiser L, Hoffman J, Beatch J.  Parental perceptions of the effectiveness of gradual extinction in reducing infant night-wakings.  Journal of Reproductive and Infant Psychology 2014; dx.doi.org/10.1080/02646838.2014.910864.