A cr(y)tical review of the main cry-it-out studies

This is one for the sleep geeks out there.  I have posted elsewhere about my approach to sleep coaching, why I use the methods I use, and why I do not incorporate cry-it-out or controlled crying into my practice.  I’ve also explained why I don’t judge parents for the methods they use: sleep deprivation is hard to deal with and parents have to do their best to cope in difficult and unique circumstances.

Apply a critical eye to make an informed decision

You might have come across this article because you want to make your own informed decision about the sleep coaching options out there.  Or, maybe you have an instinctive feeling that crying alone methods are not for you and want to understand more about what the available data really tells us.

Several studies have been published which claim to show there is no harm done by leaving a child alone to cry to sleep.  These studies have had lots of press attention over the years and are often cited as persuasive evidence to demonstrate why controlled crying or cry-it-out methods are not harmful.  However, these studies are far from perfect and it would be misleading to present them as so.  It is important to consider them with a critical eye in order to make your own mind up as to what conclusions can be drawn from them.

The main research studies

Let’s look first at a summary of the studies and what they say they found.  Then we’ll consider some of the limitations and flaws in the studies, which will help inform whether we feel we can trust them to inform the way we tackle sleep with our children.

  • The Hiscock study [2008] – this study looked at the long term mental health effects of controlled crying, on 328 mother-infant pairs.  The pairs were randomised to either receive usual care or to undergo a controlled crying programme.  The infants were 7 months old at the start of the study.  The study found fewer mothers in the controlled crying group had a significant depressive illness when their child was two years old (15%) compared to the control group (26%).
  • The Price study [2012] – this was the follow up to the Hiscock study, and aimed to show that controlled crying causes no long term harm.  Some of the infants who took part in the Hiscock study at 7 months old were followed up at 6 years old.  Their cortisol (stress hormone) levels were taken, and the study reported similar cortisol levels between the cry it out group and the control group.
  • The Gradisar study [2012] – this study looked at 43 babies between 6-16 months, seeking to understand whether the type of sleep care/training received impacted on sleep, maternal and infant stress and long term attachment.  Cortisol samples were taken from the babies before and 12 months after the sleep intervention.  Mothers’ stress levels were measured before and after treatment as well.  The Strange Situation test was used after the intervention to test attachment  (the Strange Situation test is a test designed to highlight the types of mother/child attachment).  The study purported to find a “very large decline” in night wakings in the controlled crying group compared with “no change” in the control group.  The study also concluded there was no impact on attachment following controlled crying.

Can we trust what these studies are telling us?

Let’s break this question down into two chunks: first, do these studies demonstrate the efficacy of crying alone as a sleep training method?  And second, can we trust the findings that say controlled crying does no long term harm?

Does crying alone work as a method of sleep training?

Of the three studies above, only Gradisar draws any findings about the efficacy of crying alone methods.  To describe these findings as showing a “very large decline” in night wakings in the controlled crying group, and “no change” in the control group is an exaggeration.  What the study actually showed was a small decline in night waking in both groups, albeit the decline was slightly larger in the controlled crying group (a reduction from 2.7 to 1.4 wake ups per night in the controlled crying group, and a reduction from 2.6 to 2.1 wake ups per night in the control group).  In addition, the very small sample size significantly limits the extent to which these findings can be extrapolated to the population as a whole.  In short, we should be wary of subscribing to the broad brush conclusions reported in this study.

Do these studies prove that crying alone is not harmful?

The short answer is, no.  The samples used in these studies were too narrow and there were too many flaws for us to accept their findings without question.

The small samples used in these studies lacked in racial/ethnic diversity (all three of these studies were Australian) so cannot automatically be extrapolated to the global population as a whole.  Lots of the Hisckock mother-infant pairs had dropped out of the research by the time of the Price follow up which may have had the effect of skewing the data.  In addition, no baseline cortisol sample was taken so there is no way of comparing pre and post-intervention data.

The Gradisar study did compare pre and post-intervention cortisol levels, but was flawed in other ways.  First, the “Strange Situation” attachment test is only considered reliable when used with infants aged between 12 and 20 months.  Some of the Gradisar children would have been 28 months old when they were tested.  Second, the study found no significant difference in attachment type between the controlled crying group and the crying alone group.  However, the data directly contradicts this, showing that 46% of the controlled crying children were insecurely attached, and 39% of the control group were insecurely attached.  Around 30% of the population as a whole is insecurely attached so the incidence of insecure attachment was unusually high in both groups.  This raises questions about how representative the sample was in the first place.

So what can we make of all this?

The cry-it-out debate is age old, passionately fought and unlikely to be definitively resolved any time soon.  Supporters of crying alone methods may argue that it is impossible to prove a negative (i.e. to prove that crying alone is not harmful).  Ultimately, every family needs to decide what is right for them when it comes to solving sleep challenges and this is an intensely personal choice.  To argue otherwise risks minimising the impact of sleep deprivation on every aspect of family life.

Whatever your views on this emotive topic, I would encourage you to apply a critical eye when reading up on the available options.  Decide what is right for you, and if you want to work with a sleep consultant, make sure you find someone who supports rather than challenges your parenting style.

%d bloggers like this: