Are babies “stressed out” from sleep training? Part One
Let's talk about cortisol, stress, sleep training, and Middlemiss et al. (2011)
A recent New Yorker article, “The Promise and Peril of a High Priced Sleep Trainer,” describes desperate parents (aren’t we all?!) paying a premium for infant sleep training coaches. Sleep training is a big business, sometimes (though not always) involving evidence-based methods. The journalist speculates about potential harm that sleep training could cause infants and briefly describes two studies on sleep training and infant stress:
In 2011, Wendy Middlemiss, a psychologist at the University of North Texas, led a study of twenty-five babies who underwent a five-day course of extinction sleep training at a clinic in New Zealand. At the start of the course, the levels of cortisol, a stress hormone, in the babies and their mothers were in synch. By the end, cortisol levels had fallen in the mothers but remained “elevated” among the infants, even though they were no longer crying in the night [...] In 2016, Michael Gradisar, an expert in child sleep disorders at Flinders University, in Adelaide, Australia, carried out a similar study on forty-three infants and found that their cortisol levels went down as their sleep improved.
In this Two Part series, I am going to take a deep dive into both studies to figure out why these studies have seemingly opposite results and what they might be telling us about the relationship between infant stress and sleep training. To be clear: I’m not going to discuss the evidence of sleep training in general here, nor am I going to advise for or against it - it’s a unique, personal decision for every parent.
For Part One, I’m going to briefly discuss why cortisol is often implicated in stress and break down the first study mentioned (Middlemiss et al., 2011). Stay tuned for Part Two, where I will discuss the Gradisar et al. (2016) study and summarized the findings from both.
A (brief) introduction to cortisol1
The journalist defines cortisol as “a stress hormone.” This is a personal pet peeve, when people describe any hormones or neurotransmitter as being about just one thing. Hormones are incredibly complicated. Cortisol, released from the adrenal glands of the kidney, has many functions, but one of the main things it does is signal to our bodies that we need to increase our blood sugar. This in turn aids in metabolism and has effects on our immune system, among other things. Cortisol is particularly helpful when we feel stressed - in other words, when we need an extra boost of energy to deal with an obstacle. Cramming for an exam, competing in a triathlon, running from a bear, or getting in a fistfight are all times when cortisol would tell our body to make more blood sugar, thus giving us extra energy to complete those tasks. However, I probably don’t have to tell you that too much stress is a bad thing, and more stress can lead to your body making more cortisol than it needs. Decades of research shows that chronic stress (e.g., living in poverty, experiencing trauma) is related to chronically high levels of cortisol, and high levels of cortisol have multiple physical and mental health consequences.
Back to sleep training: If babies are showing high levels of cortisol after being sleep trained, this might be a sign that sleep training is stressing babies out, even if they aren’t crying or seem like they’re in stress. These high levels of cortisol could lead to other consequences down the line (such as emotional/behavioral problems). To put it in words that would make any parent freak out: Are we stressing out our precious babies without even realizing it so we can selfishly get more sleep?!
Middlemiss et al. (2012)2
Middlemiss and colleagues were interested in physiological synchrony in cortisol response between mothers and infants. In other words, when mom’s cortisol levels are high, are baby’s levels also high (i.e., are they correlated)? Broadly speaking, it’s a healthy and adaptive thing when moms and babies are physiologically "in sync” like this. Does sleep training disrupt mom’s ability to be in sync with their babies?
The researchers recruited 25 infant-mother pairs who were enrolled in a residential, hospital-based sleep training program in New Zealand. The program used extinction, or the “cry it out” method. The researchers measured cortisol in both moms and babies twice on night one (pre- and post-bedtime) and twice on night three (pre- and post-bedtime) of the four night intervention.
It’s important to emphasize that these mothers were recommended to this program by their clinicians or other practitioners because of infant sleep difficulties or concerns with infant feeding/growth. As you can imagine, those difficulties must have been severe enough that moms chose to stay in a hospital for several days to address them. Additionally, the authors note that “many mothers reported a lack of support in caring for their infants at home.” So immediately I am skeptical that any results these researchers find would be applicable to parents and children with different, perhaps less severe sleep difficulties or different home environments.
Additionally, if I was a baby, I might be pretty stressed out in such an unfamiliar setting (a hospital), surrounded by strangers. It’s unfortunate that the researchers didn’t assess some type of baseline cortisol before the sleep training intervention began, because that would help us understand what might be a normative amount of cortisol for these babies (e.g., in their home). Without that comparison information, it will be hard to interpret any cortisol findings because 1) we don’t know what’s average for these babies, and 2) we don’t know whether any potential elevations in cortisol would be due to the sleep training or because of something else (e.g., being in a hospital).
The researchers found that mom and baby cortisol levels were correlated ("in sync") on the first night of the program. On the third night, moms and baby cortisol levels were in sync before bedtime. However, cortisol levels were no longer in sync after the infant fell asleep that night. Rather, mothers’ cortisol levels went down significantly after the infants fell asleep compared to before sleep. This makes sense, because the researchers also found that babies stopped showing behavioral signs of distress (defined as crying for more than 5-10 minutes) on night three; baby is not crying like they were on the first night, so mom’s cortisol levels went down, suggesting they were feeling less stressed. However, infant cortisol levels remained unchanged from pre- to post-bedtime on the third night, despite the fact that they were not crying as much as they were on the first night of the program.
Conclusions
So does this mean that babies were staying stressed out after sleep training (based on their unchanging cortisol levels), even if they weren’t crying anymore? As I said before, the cortisol findings are difficult to interpret because we don’t know whether this lack of change of cortisol on the third night is actually a higher-than-expected amount of cortisol for these babies. Also, the timeline of this study is very short (only three nights), so it doesn’t really provide any evidence that cortisol might stay elevated or remain unchanged over time in sleep trained babies. The authors themselves discuss the possibility that the cortisol response in babies might take more time to even out (and possibly get back in sync with mom) than the duration that they studied. It’s pretty cool, though, that babies stopped crying at bedtime after only three nights of the sleep training intervention! The researchers note that maybe it just takes longer for cortisol levels to “even out” after babies stop showing behavioral signs of stress, and that it’s possible that babies’ cortisol levels will get back in sync with their mothers’ levels over time. They do not suggest that mom and baby will stay out-of-sync indefinitely nor do they suggest that sleep training causes chronic elevations of cortisol (or chronic stress) in children.
Stay tuned for Part Two, where I discuss the Gradisar article and compare the two. Thanks for reading!
I’m not an endocrinologist or stress researcher! This is a very simplified explanation of cortisol and its relationship to stress! There are entire books written on this topic! I don’t think I used enough exclamation marks here!!!!!!!!11111
Middlemiss et al. (2012). Asynchrony of mother-infant hypothalamic-pituitary-adrenal axis activity following extinction of infant crying responses induced during the transition to sleep. Early Human Development, 88, 227-232. https://doi.org/10.1016/j.earlhumdev.2011.08.010