In Response to an Article About Margot Sunderland's New Book on Co-Sleeping

I recently read this:

http://www.drmomma.org/2009/07/co-sleeping-children-should-sleep-with.html

I’m highly wary of what I read and I feel I have to wade in with some counter arguments and questions.

Where does the upper age limit of 5 come from for co-sleeping? The only reference in the article to age 5 is the seemingly unconnected study of cortisol rise in under 5s going to nursery, which I’ve read about before and it’s a dubious and misleading statistic.

There’s no follow-up study I’m aware of to show whether those same children have a corresponding drop in cortisol when they attend school or not (the first thing I’d look for), or whether 5 year olds who didn’t attend nursery get the same sort of rise in cortisol on their first days at school. I think most of us remember school being scary at first, and why wouldn’t it be? It still is as an adult!
The demonising of cortisol as a ‘stress hormone’ is misleading. Cortisol is released to *counter* stress and produce energy so we can run from danger or fight if cornered. We get a burst of it in the morning (if we’re lucky) to get us out of bed. Without it we struggle to lift the duvet. Producing cortisol while sitting in a car or at a desk is what leads to all the negative problems, but mixing up adult medical issues (occurring in unnaturally sedentary circumstances) with children is incorrect use of data and misses out the fact kids will run all that energy off. There’s too much correlation and not enough cause in these studies.

The main comparison seems to be with babies a few weeks old rather than, say, 2 or 3 years old. Or even 6 months old. Not enough longitudinal info there.

The data for SIDS needs to be cross-referenced with the number of infants squashed or smothered by parents in bed. Ateah and Hamelin’s 2008 survey of bed-sharing mothers in Canada found that 13% of the respondents recalled at least one episode in which someone had rolled onto or part way onto their infants. None were hurt and the sleeper was awakened before any injury occurred, but it can happen and there are high-risk groups, including parents who are very tired (Blair et al 1999) and that’s a group more likely to increase if they’re sharing a bed with a child every night for 5 years. Obviously the bigger ones are a lot safer, but some parents are pretty big too!

The studies on safety seem to indicate that socio-economic status is a major factor with many variables, including overcrowding of the house, condition of the mattress, whether anyone in the house smokes, etc.
http://www.parentingscience.com/bed-sharing.html

The correlation between women not comforted in childhood and having digestive problems as adults would have more impact if there wasn’t a massive change in diet over the last generation and the babies raised on Spock and Ford weren’t the same ones who grew up eating Norman Borlaug’s modified wheat (GM foods aren’t new. Borlaug won the Nobel Prize in 1970 for his wheat, which has now been linked in some studies to IBS).
I’d like to see a study of women who *were* comforted and see how many of them have digestive problems. Why just women in the study I wonder?

“In China, where cosleeping is taken for granted, SIDS is so rare it does not even have a name.” – This is an interesting quote. China, the country where babies are routinely left to die on railway sidings and overpopulation and poverty is such a huge concern. Infanticide has been a problem there for a long time, originating before the ‘one child per family’ policy that was put in place in 1978 and having its roots in Confucianism, widespread poverty and lack of state pensions.

One could argue China hasn’t named SIDS because it’s not a big deal to the health authorities there, but mostly I suspect it’s because the possibility of cultural differences isn’t being taken into account. Maybe the Chinese don’t feel the need to label everything with acronyms and Latin ciphers to make their doctors sound all clever as they dish out meaningless platitudes. What does Sudden Infant Death Syndrome actually mean? Does it require a name? Isn’t it, when you think about it, just a placeholder that describes an observation yet infers no known cause? Why *would* the Chinese have a name for that? Why do *we* have a name for it?
Either way, China is not somewhere I will be taking parenting advice from.

http://io9.com/5948528/the-unintended-consequences-of-chinas-one+child-policy

Incidentally, using data from other countries without strict cross-referencing and allowing for confounding variables will give you some rather interesting ‘facts’, such as the drop in infant mortality rates in China between 1990 and 2008 from 64.6 infant deaths per 1,000 livebirths to just 18.5. The drop corresponds with the increase in the number of women giving birth in hospitals rather than at home, so I could, if I wanted to, say that babies born in hospital have a 71% greater chance of survival and that this has been shown conclusively by the Chinese, but I suspect the advocates of co-sleeping correspond closely with the advocates of home births, so this disinformation won’t be cherry-picked by Western authors as it won't sell many books.

My biggest concern with this article, though, is the blanket assumptions that no unforeseen negative factors will occur as long as we all make huge sacrifices for our children. This is not the world I live in.
Small children in the bed have a habit of preventing restful sleep for the other occupants. Babies do tend to be a bit safer in bed with parents (factoring out oversize, tiredness, secondary smoke factors, drug use, and socio-economic variables) because they can be checked more frequently, but 3yr olds will spend all night grabbing one parent while kicking the other in the face. The resulting lack of restful sleep in both parents can be a cause of huge stress and give rise to cortisol at home for the whole family. Articles like this tend to make parents feel inadequate so they go out and buy more books on parenting. I’m deeply, deeply suspicious of them. It goes beyond suspicion in this case though because it’s bad science and far more research needs to be done before anything claiming to be results can be weighed up properly.

Having looked her up, Margot Sunderland is someone I’m unsure about. This is her advice on handling a major tantrum: 

“Visualise yourself as a lovely warm, calm blanket. Now envelop your child by holding him with his back to you (if he kicks, he will be kicking away from you) and folding your arms over his. If he is a bigger toddler, take an arm in each of your hands and cross his arms. You can also cross your legs over his to contain his legs and prevent kicking. Hold him calmly and use a gentle tone to say soothing words ('It's all right, I am going to hold you until you calm down'), allowing him to release his angry feelings. He won't be in any space to reason with and will, in any case, not be able to activate the reasoning part of his brain while he is distressed. As your tot calms, let him lie in your arms and cuddle until he is over his blow-out. Then offer him reassurance and a different, preferably quiet, activity."

So basically physically restrain them while telling yourself you’re a calm blanket. Have you ever tried to hold a tantrumming toddler ‘calmly’? The whole business is more like restraining an inmate and far from being a soothing comfort.

If a physicist tells me wrapping a magnet in copper wire will do ‘X’ I tend to believe them, because when you know one electron you know them all. When a psychologist tells me children will do ‘X’ I wonder why they’re assuming all children are the same. I’ve met a lot of psychologists and one common factor I’ve found (in many but thankfully by no means all) is a total lack of understanding of basic human behaviour and emotions. Unlike Richard Feynmann and Brian Cox, I believe psychology is a science, however I don’t believe many psychologists are scientists. I’ve seen too many of them just doing it wrong. Cherry-picking data, relying on correlates and omitting causes, failing to allow for confounding variables and patching together studies that need further research before they get used as foundations for new hypotheses is *not* good science.