Many sleep training researchers firmly believe the former. “Don’t underestimate the abilities of children to self-regulate,” says Hall, the paediatric sleep researcher who used actigraphy in her study of 235 Canadian families. “Parents can help them learn to self-regulate by giving them opportunities to self-regulate. That’s how you can look at self-soothing – it’s an opportunity to calm themselves down.”
It’s difficult to measure objectively whether babies are truly soothing themselves, or have just given up calling for help.
One way could be to measure cortisol, which is often known as the stress hormone. But cortisol rises and falls in response to factors besides stress, and the studies that have measured it have had mixed results. One found that the babies’ cortisol levels were elevated right after a sleep intervention, but there was no control group of un-trained babies to compare it to. The small study of 43 infants found that cortisol declined, but it didn’t measure cortisol until a week after the intervention. And in an attempt to find out whether sleep training led to elevated stress levels long-term, a third study, Hiscock’s longitudinal study in Australia, took cortisol samples five years later and found no difference between the cohorts.
“I personally have an issue with the cortisol studies,” says Mindell. “Cortisol changes throughout the day. Even sampling cortisol is very difficult. It’s based on many things, including how many hours a person has been awake, how it’s sampled – it’s a complicated thing. People often think ‘oh, if we measure cortisol, we’ll know if the baby’s stressed or not stressed’.”
Even the term “self-soothing” has a confusing history. Coined by sleep researcher Thomas Anders in the 1970s, it’s often used synonymously with the idea that babies can self-regulate. For Anders, however, a self-soothing baby was simply one who put themselves back to sleep without parental intervention – he wasn’t trying to quantify their stress levels.
Of the few studies that have looked at the short- to longer-term outcomes of sleep training, none have found an effect on a baby’s attachment or mental health. Hiscock’s study, for example, the largest and longest longitudinal study done on sleep training, found sleep-trained children were no more likely to be insecurely attached to their caregiver at six years of age than their peers. (Experts like Hiscock say they aren’t aware of any studies that look at potential long-term effects of cold-turkey cry-it-out, just at modified extinction. They also examined healthy babies at least six months old. So these findings aren’t necessarily applicable to infants trained at younger ages, or in other ways.)
Like other longitudinal studies, Hiscock’s lost touch with a number of families when it was time for the final follow-up: 101 of the original 326. That means it is theoretically possible that the sleep training did affect some children in either a negative or positive way long-term, but that their experiences weren’t captured. It’s more likely, though, that any effects of a single intervention simply “washed out” after six years, says Hiscock.
The upsides of responding
Another way to examine the self-regulation question is to consider babies’ developing brains – and their limitations. Human babies are born very neurologically immature compared with other mammals, with brains around one-third of the size of an adult’s. The prefrontal cortex, the “home” of emotional regulation in the brain, is one of the last parts of the brain to mature, not developing fully until one’s mid-20s.
As a result, throughout infancy and toddlerhood, the brain relies on “co-regulation” – the aid of a soothing caregiver – to calm down. In a position adopted by the American Academy of Pediatrics, for example, the National Scientific Council on the Developing Child defines a “positive” stress response as one that results from stress that is brief, “mild to moderate” and which hinges on “the availability of a caring and responsive adult who helps the child cope with the stressor, thereby providing a protective effect that facilitates the return of the stress response systems back to baseline status”.