, , , , , , ,

I was slightly nervous posting about our sleep training. It’s a contentious topic, and I was worried about a backlash. Actually, I did not receive one; knowing that we had actually done it, and were not just considering it, my friends who do think sleep training via our method (some sort of cry-it-out-ish) is not good idea, kindly kept quiet. I appreciate their consideration. Others wrote with encouragement saying “we did that and it was awesome!” which helped me too. But a number of people responded to my comments about cortisol with relief: a ‘thank you for writing that, I had been worried about affecting my child and now I feel better’.

That got me thinking: as Mums (and Dads and Aunts and Uncles and pet cats… OK, not pet cats) we are bombarded with “information” (cough) about how to raise children. And what to eat. And how to have a successful marriage. And when to exercise. It goes on. The internet age has bought stacks of articles, stories and opinions to our fingertips (and many, many pictures of cats).


I am lucky: a large part of my job is to read scientific articles which argue that  they conclude X and to work out quite why they don’t actually quite conclude exactly X, so that I can go and do studies to help us actually conclude X with better data. Essentially, my job is to see through the bullshit and interpret actual data accurately. I spent the last semester teaching my student to do just that and she passed her qualifying exam in flying colors. I am not great at a a lot of this whole Science thing, but I am pretty good at that.

So, I think I have an advantage. I can see through all (well… some… I occasionally call my Mum in a panic) the guilt-trips and the misinformation and the scare stories Yes: scare stories. (check out the daily Mail’s list of things that can give you cancer http://www.anorak.co.uk/288298/scare-stories/the-daily-mails-list-of-things-that-give-you-cancer-from-a-to-z.html/).


Personally, I am never going to stop reading the interwebs (I’d have to do way too much work for a start), so I thought I would share some of my tactics for keeping a sane head.

First, I usually try to source the original scientific article. You can usually find them on PubMed or Google Scholar with some clever searches (can take some practice). You will find a link to the article (which you may need to subscribe to a journal to read), but I do the bulk of my assessment off the abstract. But you don’t need it really. Then I go through the following process, which I will illustrate using the ‘sleep it out causes anxiety disorders in children’ study. I use this study only because it was the subject of my most recent obsessive-research bout. You can do it equally with ‘diet coke causes your leg to fall off’ or whatever the latest ‘thing’ is.

So, back to the study. The abstract can be found here, and an article written about it on the internet can be found here.

Essentially, the study examined a small group of infants who had been left to cry-it-out as sleep training. After 5 days the infants would go to sleep without crying, but their cortisol levels were higher, while the cortisol levels of the mothers were not high. That is all the study says.

Response articles, however, said things like “But there is no doubt that repeated lack of responsiveness to a baby’s cries-even for only five minutes at a time-is potentially damaging to the baby’s mental health. Babies who are left to cry it out alone may fail to develop a basic sense of trust or an understanding of themselves as a causal agent, possibly leading to feelings of powerlessness, low self-esteem, and chronic anxiety later in life”.

I can’t find it now but I did also find an article that said that children left to cry-it-out would, within 5 minutes, “deduce that they are essentially unlovable”.

Holy Moly! Who wants to subject their kids to that?? So, what proces did I use to evaluate these data:

(1) OK, the biggest thing ever: what do the actual Scientists who wrote the study say it means? Not a blogger… not an author… not a media Science writer… not a random Physicist who someone got to comment on the study: what does a Scientist in the same subject say about the data? Preferably, what do the authors of the article say? They say:

“The findings are discussed in relation to understanding the determinants and implications of maternal–infant physiological synchrony in early childhood”

NOTHING about later life. NOTHING about disorders. Not even anything about anxiety. Not even implications for the development of the child. They simply say that mother and child cortisol levels are no longer in sync: they have a ‘physiological asynchrony’.

Already it’s less scary, right? I mean, if I said “do this and your hormones won’t rise and fall in the same rhythm as your child’s” you probably wouldn’t freak out too much.

(2) Then I look at the actual outcome: cortisol. Now let’s look at the claimed outcome: anxiety disorders in later life. Assuming the actual outcome was well measured in this study (and the cortisol was… it’s a good study) can we link the actual outcome with the claimed outcome? NO. Is there any scientific link between infant cortisol at bedtime and later anxiety disorders? Nope. OK, so how about infant cortisol and later anxiety disorders? Errr… nope. OK, we are getting tenuous here, but how about general cortisol (remembering that we have no proof that a temporary nighttime rise in cortisol leads to a general increase in cortisol) and anxiety disorders? Makes sense, right? Well, maybe, but sadly Science does not agree. Some studies support the hypothesis that cortisol (which is linked to acute stressful situations) is associated with anxiety disorders while others,show that within individuals suffering from anxiety “a lower cortisol awakening response… predicted an unfavorable course trajectory”.

So, you may not want your kid to experience cortisol rises. That’s fine. You may think that infant cortisol leads to adult anxiety. That’s also fine. But, know that that is not an evidence-based view.

(3) I check whether the outcome (here: cortisol) has been associated with ONLY the status variable (here cry-it-out training). Well… no. Cortisol is associated with any infant distress. We put our kids through distress all the time (last time I used the example of Sam howling in the car seat when I was on the interstate and couldn’t comfort him). So, cortisol at night is a question of adding something the kid already experiences, not bringing a whole new qualitiative experience.

More importantly: there is no evidence about whether cortisol is experienced by non-cry-it-out training. There is no control group. It is not only possible, but plausible to me, that if you have a kid like Sam who seems to hate the very concept of sleep, any form of sleep training is going to stress him out (this was highlighted to me when he would go ballistic at cuddling on the boppy pillow because it was used at night).

(4) I also look at the general methodology of the study: who was the data collected on? If it is not humans I treat conclusions pertaining to humans extremely skeptically. Is it on a large sample? This was on 25 infants – that is not a lot at all. It is probably what we would call a ‘pilot study’. But, that does not invalidate the conclusions of the study, it just means that we need to see it repeated. It is very possible that with such a small sample size there was some particular characteristic about either the mothers or the children that don’t pertain to you. Has the study been repeated by an independent group? In this case – not yet.

(5) I also consider the point that in most studies where we are looking at potential harm people are not randomized to a group. That is, people choose to follow a certain path / method and perhaps WHY they chose that actually explains the outcome. In the case of the sleep study, we have infants who could not sleep through the night, and mothers who responded to an advert for such (were they at the end of their tether?) and agreed to take part in the study. Would I follow a training plan for my kid at the direction of a Scientist to earn some bucks? No. Would I sleep in a sleep center with my kiddo when he was that young for 5 days? No. So maybe there is something about these mothers who consented to do such that leads to the rise in cortisol.

The clearer example is on breast vs. formula feeding. There are great breast and great formula feeding mothers (I was both, so clearly I am doubly great… that’s Science). But, if you ARE a negligent mother OR you suffer PPD or whatever, you are less likely to go through the hassle of breastfeeding. So, there will be more ‘struggling’ mothers in the formula feeding group. Or, even if you are super fabbie wonderful faultless mother (like moi, naturally) then there may be a reason you formula feed. I didn’t produce enough milk. Possibly because I was separated from my baby so much early on, or because I went back to work so quickly, or because I didn’t have a great post partum diet. Maybe THESE reasons – which aggregate more among formula feeders – and not the formula cause group differences between formula fed and breast milk fed babies (where they exist).

(6) I am wary of ‘evolution’ arguments. Yes: our environment is changing faster than our DNA can evolve (although note: much of who we are is not what the code of our DNA says, but how much it is turned on and off [i.e. are the genes doing anything in our body, or just lying dormant] and that can be changed in as little as 12 hours…). But regardless, our environment has changed, and we have to help our kids fit into it at some point. When and how is up to you.

(7) Sift out sensationalist language. Young Scientists, especially those in behavioral Science, learn early on not to say “we have shown that such and such is the case” but rather “our data support the idea that..” or “we have added to a body of evidence arguing that…”. Internet articles are sneaky-sneaky using a mix of both:

“But there is no doubt (oh no!) that repeated lack of responsiveness to a baby’s cries-even for only five minutes at a time-is potentially (oh… wait..) damaging to the baby’s mental health. Babies who are left to cry it out alone may (only may?) fail to develop” yada yada yada. Don’t focus on the first bit: focus on the uncertainty.

(8) You will KNOW if something is – for sure – bad for your kid. There are no Mummy-wars about whether one should smoke in pregnancy to stay relaxed, or teach your child to skip meals to keep body fat low and prevent child-onset obesity. Or if mothers should drink a butt load of delicious delicious wine to blank out the sound of their infants crying. We know some stuff. The rest? It’s not evidence based. It’s instinct and guesswork.

But yeah, I have my random opinions about my kid – some rational, some not. I am obsessed with the idea that Sam has homemade baby food only until he is 2 – even the organic-in-a-pouch stuff is not “good enough” (I am sure I will reassess when I see the work… but hey, Wes is a SAHD, he’s got to earn his keep 😉 ). I don’t really know why, although if I had to, I could bring out a heap load of scare stories showing you why he MUST have that if he is to be AT ALL healthy and functional later on in life. I would not have had him vaccinated had I been in the UK – even though Science really does not back that stance up. However, I don’t care too much about baby bath products (poison apparently), crying-it-out and disposable diapers. Go figure.

All vaccinated up

All vaccinated up

But look – you gotta read the stuff to know what you think about it. I certainly like my friends sending me articles that go contrary to my opinion so I can reevaluate my opinion. David (one of my UAB mentors) once said to me “Be who you are, but be it knowingly”. I say: “Raise your kid in your own way. but do it in an informed manner”. And do it without guilt.

My birthing coach Johanna put it best when I was fretting about sleep training: she asked if I really thought one thing could make up for all the days and nights of unconditional love.

Anyway, if Sam is anything like me, he’ll f- himself up more than I ever could anyway 🙂

Image credits