I’ve been aware of this poster for a while now. Macall Gordon, an ‘academic’ (at the time of the poster’s making she was studying for her Master’s degree in Antioch University, Seattle) presented it at the 2007 Biennial Meeting of the Society for research in Child Development, and chaired a roundtable discussion symposium on the subject of “Crying Before Sleep: What Do We Know About Behavioral Sleep Training in Infancy?” there.
The poster’s claims, supposedly, lay a foundation for claiming there is no evidence in the research for using CIO methods in children under 12 months of age, because we don’t know what the psychological effects will be.
Gordon claims there is no basis for the recommendations of sleep experts that parents can implement CIO techniques starting at 3-6 months, and that animal studies suggest that extinction (not necessarily in the context of sleep, mind) is stressful. She doesn’t quite go as far as to claim that the stress will bake babies’ brains in a cortisol bath, turning them into mush, in the manner of Dr. Sears and other hysterics…but it’s implied.
Gordon has reiterated her ideas in an interview with the Talaris Institute, and on her own website and blog. She, to be fair, stops short at claiming outright that CIO should never be done, but it’s clear she’d like to undermine parents’ confidence in the method.
In this post, we’ll have a look at Macall Gordon’s claims and see if there is any merit to them. But before that, you might want to know that Gordon is/was an executive editor at Attachment Parenting International and an advocate of cosleeping – and I might add that she doesn’t seem to have looked at the quality of research in support of cosleeping quite as closely or critically as she has done with CIO (Note the inclusion of two non-peer reviewed master’s theses making claims that cosleeping makes for psychologically superior children, and the usual baseless “cosleeping may prevent SIDS” claim). Not that this disqualifies her from analyzing the literature regarding CIO, but knowing her bias is always helpful.
Claim #1: There is no basis for recommending parents start CIO methods between 3-6 months, as suggested by many popular parenting books.
Actually, there is very good physiological evidence that by age 3-4 months, and certainly by 6, a baby is ready for CIO. That doesn’t mean you must CIO the instant baby turns the requisite age, of course, but this is the minimal age for using these techniques. It’s not exactly a secret, as Ferber (at least) does partially explain the rationale in his book. Reasons of psychological safety have never really come into the equasion because, well, one would think the generations-long experience with the practice should speak for itself.
*Babies are born with sleep cycles that are different than those of adults:
Normal Sleep Development
Neonates cycle through rapid–eye-movement (REM) and non-REM (NREM) sleep differently than older children or adults. In the first few months of life, infants have not yet become fully adjusted to a day-night cycle, and more of the control of sleep is internal. Full-term infants can spend as much as 50% of their total sleep time in REM sleep. At this age, the term used is active sleep. The infant often enters the sleep cycle in REM. By about 4 months of age, as sleep cycling approaches the adult mode, sleep-state progression matures and NREM sleep typically precedes REM sleep.
*Newborns have very labile emotional and hormonal responses to stressful events – even a diaper change can be interpreted as a stressor. By 4-6 months, this behavioral and hormonal response is much attenuated, and is similar to that of older babies.
*Smaller babies (under 3-4 months) often need to wake up in the night to feed, and may become malnourished if prevented from nursing at night.
(Yes, I know about Babywise and the like. I don’t endorse such methods, for all the reasons detailed above).
Claim #2: Most of the studies done regarding CIO don’t examine babies under 1 year of age, and those that do, don’t do so exclusively and we don’t know how many babies were looked at. Very few look at the psychological effects of CIO.
Gordon lists a whole bunch of studies which really do deal with older children, the majority of which relate mostly to whether the method is effective, not the effect on the child’s psychological wellbeing. So in a sense, she’s right. But that doesn’t change the fact that there were two (and now are 3) studies which examine this question in populations of infants, and all of them have found no negative, and even some positive, psychological effects. They are France et al and Eckerberg et al. While the former does not state the proportion of children in each age group, the latter states clearly that over 2/3 of the children studied (67/95) were between 6-17 months of age, and the results were also stratified by age (6-18 months and >18 months, respectively). The last study, which has since been published and which I’ve looked at here, deals exclusively with infants who were 6-12 months old at the time of the sleep intervention, with, yet again, no evidence of psychological harm to them. (Interestingly enough, Sarah V., who addressed the issue long before I did, managed to come up with exactly the same two studies that were around at the time. Imagine that – both of us managed to bring up only the relevant literature…)
So despite the fact there are not oodles of supportive research dealing with our specific issue – when one considers there are exactly ZERO studies that have found the opposite, and adding 90 years’ worth of observational data with no negative effects shown…I dunno, but it seems that right now the scales are tipped rather firmly towards no evidence of harm from CIO, and it’s up to the anti-CIO crowd to prove (not make theoretical claims, prove) the contrary if they want to claim otherwise.
Claim #3: The psychological test used doesn’t actually measure infant psychological wellbeing, but rather describes behaviors more related to regulatory capacities, temperament, and developmental level (e.g. “Can accept sudden advances of a stranger.” “Enjoys a crowd.” “Can recover easily when upset.”) Nor does it prove that CIO doesn’t cause attachment problems.
The first two studies mentioned (France and Eckerberg) used a modified questionnaire based upon the Flint Infant Security Scale. Neither of them are solely reliant on that measure, however: Eckerberg et al correlates the Flint scores with Visual Analogue Scales (VAS) the parents used to assess their own and their baby’s mood, and maternal emotional status (which might have affected the scores the mother gave) was tested with the Swedish Parenthood Stress Questionnaire. France et al augmented the results of the Flint Test with the Child Behavior Characteristics scale (according to Borgatta & Fanshel), another test that, like Flint, has been used for decades to assess infant emotional status. (The third study by Hiscock et al assessed infant mental health only at the end of the followup using tests appropriate for older babies and children, not the Flint test, and compared the results to the population at large.).
While it’s true the items on the Flint scale (and most of the other scales designed for use in this age) mostly assess ‘regulatory capacities, temperament and developmental stages’…what exactly would Gordon have them assess? It’s not like you can ask babies how they feel.
But when a baby that age shows evidence of emotional trauma, it’s usually in the form of emotional dysregulation, irritablility and being difficult to soothe, social withdrawal and regression in developmental stages. The Flint and other scales assess exactly these elements. The fact that the researchers were using at least two different tests in each study, all of which spotted no evidence of trauma, is suggestive that CIO is not emotionally traumatic (and may even be beneficial). But hey, if Gordon has some other test she’d like applied to babies undergoing CIO which she feels would better assess this…she was welcome to suggest it. Yet she didn’t.
But what about attachment? Well, the Hiscock study, having assessed the children at an older age and compared the to the general public, could theoretically have done so, but being that you can’t do a Strange Situation test in babies younger than 9-10 months or so, there would have been no way to assess the babies’ attachment beforehand and compare. Still, it might be worth doing and comparing the results to the control group.
There are other, more subtle signs in the Hiscock study which suggest attachments were not ruptured (and perhaps improved) as a result of CIO, though; more on that in Part II, in which I will continue wading through Gordon’s pseudo-academic claims of possible harm from CIO.