Basic_HPA_Axis“A bit of stress is good for them!”

How many of you have heard this when it comes to sleep training? Day care?  The idea that stress is actually a good thing?  I’ve explained why this is wrong before, but we can add another reason to the list thanks to new research out of Germany: Acute stress reduces cognitive flexibility, even in infants.  Let’s take a look at exactly what this new research can tell us about the effects of stress on the developing brain of infants and why this might matter to what is becoming a common parenting practice in “helping” sleep…

To understand the research, we have to have context and the background here is that we have quite a bit of research on the very negative effects of chronic stress on development.  Children who experience chronic stress are at much greater risk of suffering a host of maladies later on.  The question is what about acute stress?  Acute stress is, by definition, non-chronic, but still resulting in increased cortisol and other physiological changes.  In adults, small amounts of stress can be beneficial as they can increase learning, but can we take work from adults and assume the same holds for young children?  Further, despite some “positives” of minor stress, we know acute stress actually results in more rigid behaviours in adults – they are unable to be flexible in their thinking which diminishes problem solving and can lead to unwanted behaviours.

In other works, I have spoken of the period of hyporesponsivity and why it is important when we discuss stress in infancy.  Briefly, this period is a time when it is difficult (but not impossible) to elicit a cortical response to a stressor.  It begins in early infancy (around 2 months of age) and continues for at least the first year, possibly the first three (this time frame is currently unknown).  The major factor that predicts this hyporesponsivity is secure attachment and responsive caregiving.  Responsive caregivers seem to buffer the child from the effects of stress and in fact the absence of such responsiveness results in increases in cortisol in children.  That is, the removal of responsive caregiving when needed is a major stressor in and of itself.  The fact that we have evolved to have such a period speaks to the need to keep cortisol from negatively influencing neurological development during this time, but also speaks to how the presence of stress informs the child about his or her environment.

Acute stress should therefore never be seen as “minor” for a child who is in the period of hyporesponsivity because the entire point of development and thriving is to avoid such a stress response.  However, evidence for any cognitive change in children is limited simply because the research itself has not been done (not because no effects have been found).  The new research, led by Sabine Seehagen from the Ruhr-Universität Bochum and published in the Proceedings of the National Academy of Sciences

[1], examined the issue of how acute stress affects infant cognition, specifically with respect to rigid versus flexible thinking.

In this study, 26 15-month-old infants’ cognition was tested after either a control condition with no stress (14 infants) or exposure to an acute-stress condition (12 infants).  The stress-condition included a well-known paradigm to induce stress in infants which included three separate minor stressors: exposure to a stranger, exposure to a loud robot, and separation from the caregiver, each of which was 4 minutes long with non-stressful parental interactions in between.  The experience of stress was confirmed both behaviourally and via assessments of cortisol levels compared to pre-stress-condition cortisol and that of the no-stress group.  (However, the current study does not assess the individual contributions of stress and thus it is not clear if one subcondition or a combination led to the stress responses noted.)

After this condition, infants were taught a learning game involving pressing buttons to elicit lights and sounds and then having this go extinct.  The key point to the test here was to see how children will respond when another button is available and the one that previously rewarded them ceases to do so.  Will they remain rigid and stick with the button that worked before or will they be flexible and try something new?  Key points on this methodology:

  • All children learned that both buttons (red and blue) create lights and noise
  • The habit-acquisition phase included pressing only one button until they lost interest
  • During the testing phase, both buttons are available, but pushing the buttons does not elicit the same effect it did during learning and habit-acquisition
  • The buttons were different colours, but exposure to the colours was counterbalanced to ensure a favourite colour wasn’t driving any results

The researchers found that children who had undergone the stress condition got “stuck” on the habit-acquisition button; that is, these children kept pressing that one button without trying another even though they saw that the button was no longer providing the same stimuli it did in the previous phases.  On the other hand, children who had not experienced the stress-condition were less likely to persist in pressing the button that was part of the habit-acquisition phase.  Interestingly, the change was most noticeable over time.  Infants in the stress-condition increased the proportion of time they pressed the habituated button over the 30s of test time whereas the other infants decreased the proportion of time pressing that button (43 to 64% versus 57 to 18%).

The take-home seems to be that acute stress has a similar impact on infants as it does on adults.  Namely, it inhibits our ability to think flexibly which is intricately linked to our problem solving abilities.

Is this study the final word on the issue?

Like any new study, of course not.  This is a preliminary study and should be viewed as such.  An important one, mind you, but still preliminary.  The methodology was appropriate and because of the cost of assessing cortisol, the sample of 26, though it might seem low, is actually in the range of normal for this type of work.  Notably, the findings fit within the context of what we know about the effect of acute stress on adult cognition (which is more robust) which adds more credence to the findings, but it absolutely should be replicated before deciding how much weight should be given to it, though it certainly should not be dismissed because of its preliminary nature.

How can we relate this to sleep training?

I don’t know anyone who says that extinction sleep training is not at all stressful on the infant.  If they believe that, they’re insane.  The stress of sleep training is, at a minimum, acute stress (it can be long-term as there are families who report using this method for over a month at a time[2]).  Distress during extinction sleep training is likely far higher than what was experienced in this particular study, as in this study infants were crying only around 19% of the time.  And if your child isn’t crying in your “sleep training”, then it’s not extinction sleep training, now is it?

Isn’t a lack of sleep even worse for cognitive development than acute stress?

This is hard to answer because we know that severe sleep deprivation has both short-term and, when chronic, long-term implications; however, much of what parents report as “sleep problems” for their infants and toddlers are not actually problems[3].  They are biologically normal patterns.  Because they are biologically normal, we would not expect this to have any negative effects on cognitive development.  Note that this does not mean that it is not a problem for the parent which is why we have gentle ways to help guide infant sleep when absolutely necessary, but at no point is intentionally stressing a child the appropriate response.

When there is a severe problem with sleep, it is best to look for the underlying cause because sleep in and of itself is not a problem – it’s a basic function, like eating, which is highly variable between people but is not something we need to “teach”.  Our children will eat and eat healthy foods if we present them with those options.  Sometimes the problem is simply that parents have an environment that is not conducive to sleep, such as having screens on before bed or lights that aren’t dimmed, but sometimes it’s a health issue that impedes sleep and medical attention is necessary. (Note that things like nursing to sleep and cuddling and cosleeping are not “bad habits” that need to be avoided or that hurt sleep.  To find out more on that, click here.)  However, again, there is no need to utilize extinction sleep methods that cause acute stress on top of whatever else is causing sleep disruptions.  As I often explain to families, when your child is sick with the flu and not eating, we don’t treat the not eating as the problem and force feed them to “teach them to eat”.  We look for the cause of not eating – the flu – and treat that knowing that the eating will resolve itself when the flu is taken care of.

Thus, the argument that we must prioritize sleep over all else (which is what extinction sleep training does) ignores these key points and when we consider the additional effect of this rigidity in cognition, it should be clear that this is not the way to help when there are sleep problems.  You can find other ways to help your children gently.

Does this say anything about child care?

It is entirely possible that this speaks to why we see externalizing behaviours in children who have spent extended periods in daycare, particularly low-quality daycare.  As I have written on here, the research on stress and daycare is clear that it is an acute stressor for many infants and children, but this is mediated by quality of daycare.  If this is the mechanisms by which we’re seeing negative outcomes associated with daycare, it means we have even more reason to ensure that all children are provided responsive caregiving options.

What questions remain?

The questions that remain, though, are equally important:

  • How long does this need to go on for there to be lasting changes? Remember, this research found short-term cognitive changes, not long-term.  But at what point does this rigidity become a part of a child’s neurological framework?
  • How does this interact with child temperament? As we are learning more about different temperaments and different reactions to parenting[4], we need to be aware that the effects are different for different children and this will be based, at least in part, on their temperament.
  • If this does occur long-term and there does seem to be a fixed pattern emerging, how can the brain return to a flexible state? This is important because we will all make mistakes and hope that we can overcome them, even if it takes longer and a lot of work.
  • What is the evolutionary reason for this? We want to have a plausible idea as to why this type of rigidity comes in the face of stress – what would the advantage be?  Or equally, why would it not harm the child in the long-run?  This is important for any finding.  It has to fit within a theoretical framework and this research is no exception.  Personally I imagine it has to do with the biological stress response and the needs of the body to divert blood away from the brain to the limbs for a fight-or-flight response, but that’s just my initial thoughts.



[1] Seehagen S, Schneider S, Rudolph J, Ernst S, Zmyj N.  Stress impairs cognitive flexibility in infants.  PNAS 2015; 112: 12882-6.

[2] Loutzenhiser L, Hoffman J, Beatch J.  Parental perceptions of the effectiveness of graduated extinction in reducing infant night-wakings.  Journal of Reproductive and Infant Psychology 2014;

[3] Hysing M, Harvey AG, Torgerson L, Ystrom E, Reichborn-Kjennerud T, Sivertsen B.  Trajectories and predictors of nocturnal awakenings and sleep duration in infants.  Journal of Developmental and Behavioral Pediatrics 2014; 35: 309-16.

[4] Belsky J.  The upside of vulnerability.  Scientific American Mind 2015; 26: 40-45.