I’ve been following a discussion about CIO (crying it out) pros and cons over at iVillage that I found via a trackback to my blog stats. A poster had requested scientific information about the Ferber method/CIO in general. Someone (Thanks! 🙂 ) had posted some of my critiques of the usual garbage that gets passed around the Web as “scientific information abut CIO”, but is nothing of the sort. Of course, one of the other anti-CIO posters (who herself had posted two sources full of said garbage information) was not going to let those links stand. Having skimmed them, she pronounced the studies I referenced to be about “whether CIO works, not whether CIO is safe” (which is patently untrue – see the link referenced here). When her error was pointed out, she still insists we do not know the “dose” of CIO which is sufficient to cause brain damage (as demonstrated by studies which deal with prolonged separations and outright abuse in animals and children) and “If CIO were a drug, there would be clinical studies done to determine the physical effects and potential risks, at what dosage level it causes harm (overdose), and other factors relating to it’s safety before it could be approved for doctors to prescribe it or parents to use it. ” More to the point, she pooh-poohs the idea that parents’ and other observational data could actually be meaningful:
The studies presented are based on parental observation and little else. No medical studies on how an infant’s BRAIN actually responds to CIO have ever been presented. In other words, there is no “evidence of no harm”…. and a great deal of information that points to the possibility of harm. Is there definitive evidence on either side? Not at the moment. Is there reason to doubt the safety of this method? I believe so.
I often come upon AP/NPers (attachment/natural parenting devotees) responding to stuff that I’ve written which is referenced online. Most of it is an embarrassing demonstration of their lack of reading comprehension (and indeed, it’s quite obvious this poster did not initially read the links provided), but this case refers to something slightly different which, I think, could use a bit of elaboration, as it reveals a bias and error of thought common in AP/NP philosophy.
As I have pointed out before, her “great deal of evidence that points to the possibility of harm” leans heavily on untested hypotheses, twisted study results and outright lies and omissions such as this and this. But is it really true that observational studies can’t tell us whether CIO at the “doses” normally used in the context of sleep training are harmful or not?
As I have pointed out before, it’s impossible to conclusively prove a negative. And people who object to a practice they don’t like will often use the precautionary principle in a selective manner, moving the goalposts constantly to keep their pet theory (in this case “CIO causes harm”) alive:
It’s also important to remember that many of those who invoke the precautionary principle do so not out of concern for the masses, but for political purposes. They also tend to invoke it selectively – e.g, PCB’s in baby bottles are ‘proof’ that bottles are bad, but PCB’s in breastmilk are, mysteriously, not a cause for concern. Thimerosal in vaccines, despite the Institute of Medicine’s statement rejecting a connection with autism, is OMG A TOXIN WE’RE INJECTING INTO WIDDLE BABIES!, but the unregulated, non-FDA supervised potion their herbalist gave them? Just fine.
I wonder what that APer would think about the following scenario: A woman who is breastfeeding her newborn baby has an infectious disease which requires her to take penicillin, otherwise her health will deteriorate. Her doctor tells her that she must stop breastfeeding her child immediately, because the long-term effects of penicillin on a newborn baby have not been studied in controlled trials.
If she is like most APers I know, she’ll probably be falling all over herself to denounce this doctor as a moron, and that the woman in question had better get another doctor STAT (and I would even agree). However, the doc is technically correct: there really are no controlled studies which examine the safety of penicillin while nursing, or for that matter, pregnancy. The reason we (and Dr Hale) know that penicillin will cause the baby little to no harm if passed on via breastmilk is…wait for the drumroll…60+ years’ worth of observation.
You see, while observation is not really a good way to infer a positive cause-and-effect relationship, a very long period of observation is quite useful to come to a conclusion regarding the lack of a cause-and-effect relationship. Meaning that if a practice has been common for decades or centuries and side effect X has not been observed after its use, it’s quite reasonable to conclude that X is not a side effect of that practice, or else is a vanishingly rare one.
Hence, if many breastfeeding women have been taking penicillin since the 1940s and there have been no reported cases of severe infant adverse reactions as a result, it’s fairly safe to conclude that penicillin in breastmilk has no severe adverse effects, and allow its continued use in breastfeeding mothers.
Consider another observational project, this time regarding vaccine reactions: VAERS. This project serves as a portal for reporting adverse events which happened after administration of various vaccines. As VAERS has been around for several decades, if a certain adverse event is not listed as a purported result of a specific vaccination, it’s pretty safe to assume that that that particular adverse event is not a side effect of that vaccine, and there is no obvious reason to study it. For example, if nobody has ever reported that the DTaP vaccine caused their child to grow elephant ears, it’s fairly certain that DTaP probably does not cause this. (Mind you, that doesn’t mean that if a group of parents suddenly did report that their children developed elephant ears after DTaP, then DTaP would necessarily be the cause – just that it would have to be investigated).
What does all this have to do with CIO? Well, CIO in the context of sleep training has been around for a long while. Parents have been exhorted by various experts to let their babies cry themselves to sleep and during the day, and I might add in a much less controlled (i.e., higher “dosage”) manner than usually promoted today, for at least 80-90 years now. Despite this, one will be hard-pressed to find even one credible case of a child who was sleep-trained by CIO (Weissbluth, Ferber or otherwise), at the appropriate age and in the context of a loving, responsive parental relationship, who was harmed by the practice. For example, I suspect Dr. Sears’ recounting of the scary-sounding “Shutdown Syndrome” is highly embellished, but even there, he is talking about a young (4 months old) child whose parents changed several of their parenting practices all at once, including CIO. And I would think that the example in my previous post is one nobody wants to touch with a 10-foot pole…
So given the rather long period of observation – whole generations raised on this practice – and the lack of credible evidence of harm coming about solely by CIO in the “doses” used today (and even in higher “doses” as used in the past), it seems fairly reasonable to conclude that use of CIO in these “doses” does not even come close to resembling the crying and stress brought about by genuine child abuse and neglect (which has plenty of observable negative results), and is almost certainly safe. Something, mind, which cannot be said about certain forms of cosleeping in the early months…but that fact rarely penetrates the minds of the AP ideologues.
But what about the claim that we can’t know about the harmful effects of CIO because “No medical studies on how an infant’s BRAIN actually responds to CIO have ever been presented”, or because we don’t know how babies’ hormones react to the stress of CIO? Even then, we would need observable phenomena to check if the hypothesis that “CIO damages baby brains” is valid. Because how would we know if a certain hormonal pattern or a given pattern on the fMRI is associated with an adverse psychological outcome, if we don’t correlate the patterns we see to scores on psychological tests or other directly observable phenomena? A given level of cortisol or a lower-functioning area of the brain in an fMRI test is not, in and of itself, a pathology – it could also signal a healthy, adaptive response to a challenge. Never mind that even then, we would still have to prove that the pattern is caused by CIO and is causative of the psychological effects. In any case, I would submit that the current state of knowledge regarding stress hormones in particular is still very, very far away from enabling us to reach such conclusions (and will elaborate on that further in a future post), even if we had demonstrated a sustained difference in stress hormone levels and response in CIO’ed babies – which has not yet been done, to remind you.
But we can see that even when an attempt is made to assess the hormonal state of babies undergoing CIO, the anti-CIO mamas still get the vapors and run to sign petitions claiming such an experiment is unethical. In fact, one of the references the anti-CIO poster herself brought as evidence claims so. Mind you, I have no idea if the study itself is well-designed or will actually prove anything worthwhile. But the anti-CIO contingent is indulging in a circular argument here: We just know CIO is harmful even though it’s never been studied and there is no empirical evidence of such (and some empirical evidence to the contrary), therefore it’s unethical to do an experiment to test whether CIO is harmful or not.
One might even suspect the anti-CIO contingent is afraid of finding out the answer…