Three studies are routinely cited as proof that controlled crying is safe and effective. Read closely, none of them shows that. There is no good evidence that it works, no adequate evidence that it does no harm, and one of the three points the other way on attachment. What we can say honestly is that the research is not good enough to settle the question.
You will be asked about this, often by well-informed parents who have read something in the national press and want to know where you stand. Being able to answer with reason, authority and confidence is what makes you a credible practitioner. Answering with a slogan, in either direction, does not.
Watch: is sleep training always harmful for a baby, answered through research.
Table of contents
What we are actually talking about
Leaving infants to cry divides opinion sharply, and it travels under a lot of names. You will hear it called controlled comforting, spaced soothing, rapid return, controlled timed crying, crying down, cry it out and graduated extinction. They are variations on the same underlying approach.
Worth knowing before the conversation starts: parents may never have heard that there is any alternative. You may be the first person to suggest a gentler strategy exists.
The arguments on both sides
| For crying to sleep | Against crying to sleep |
|---|---|
| Often quite quick. | Many child development and mental health experts believe it damages the trust between parent and child. |
| The parent or maternity nurse is more in control of when sleep happens. | The child has not actually learned to self-soothe. The correct term is learned helplessness. |
| The baby will give up crying eventually and either sleep or lie quietly. | It is very distressing to listen to a baby crying hard for any length of time. |
| It fits schedule-based parenting philosophies. | Some babies vomit when very distressed. Some become dangerously overheated. |
| Easier to plan a day around strictly timed naps and bedtimes. | For some babies it is not quick at all, and the crying persists far longer than expected. |
| Requires very little input from the carer. | After illness, developmental change or travel, the process often has to be repeated. |
| Straightforward, with few elements beyond being consistent. | You have no guarantee the child is not ill, scared, hungry or in danger if you are ignoring their cries. |
The Hiscock study
Hiscock et al, 2008 looked at long-term mental health effects on mothers and infants after graduated extinction. 328 mother-baby pairs in Australia, randomised at 7 months to usual care or a graduated extinction programme.
Fewer mothers in the extinction group had a significant depressive illness when their child was two. But what the study did not find was any statistically significant improvement in sleep outcome, in either group. And mothers in the intervention group still reported high depression at 15%, against 26% in the control group.
So: a study frequently cited as proving sleep training works did not demonstrate a sleep benefit.
The Price study
Price et al, 2012 followed the Hiscock cohort up at six years, taking cortisol samples, and reported similar levels between groups. It is widely treated as proof of no long-term harm.
The problems are substantial:
- 46% of the original sample refused to take part in the follow-up
- It reuses a sample where there was no significant sleep difference in the first place, so the original errors carry through
- No baseline cortisol was taken before the intervention, so there is nothing to compare against
- No cortisol was taken during or immediately after the extinction, so the size and duration of the stress response is unknown
- Black and minority ethnic groups were under-represented and non-English speakers excluded
- No actual cortisol data is provided. The researchers say levels were "similar" without defining it
Given those flaws, the claim that graduated extinction has no long-term negative effects is not supportable. The study was not designed well enough to make it.
The Gradisar study
Gradisar et al, 2016 recruited 43 babies aged 6 to 16 months to compare graduated extinction, fading and supportive care, measuring cortisol before and 12 months after, and using the Strange Situation Test for attachment.
On the sleep results. Reported as a "very large decline" in waking for the extinction group against "no change" in controls. The actual numbers:
- Graduated extinction: 2.7 wake-ups down to 1.4
- Control: 2.6 wake-ups down to 2.1
A difference of roughly one waking. That is not what "very large decline versus no change" implies.
On attachment, the more troubling finding. 46% of children in the graduated extinction group were insecurely attached, against 39% of controls. Population prevalence is around 30%. Both groups are high, and the extinction group highest, which sits badly with the conclusion that attachment was unaffected.
The Strange Situation Test is also validated for children aged 12 to 20 months. The oldest children here were 16 months at the start and therefore 28 months at follow-up, well outside that range. And the study is under-powered: the researchers acknowledge each group needed 21 cases and had 14 or 15.
What we can honestly conclude
There is no credible research that conclusively settles how perceived sleep problems should be handled. Specifically:
- No evidence that controlled crying works.
- Not enough evidence that it has lasting positive benefits.
- Limited evidence that it reduces maternal stress and depression any more than education and support do.
- No evidence at all that it does not damage attachment, and the Gradisar data points the other way.
- Research into infant mental health indicates that leaving babies alone to cry induces high stress and produces learned helplessness (Ronald de Kloet 2004; Waters et al 2000; Middlemiss et al 2012).
- Little is known about the dose response of crying. How much crying is too much is an open question.
Ultimately it is a parent's choice how their baby falls asleep. But it is worth being clear with yourself about your own position, because you will be asked.
What this means for your practice
- Have the conversation before the baby arrives where you can. It is far easier to discuss objectively in advance than at 2am on night three.
- You may take over from a practitioner who used controlled crying. You do not have to persist with it if it is not working.
- You do not have to take jobs where you are fundamentally not on the same wavelength as the parents. Screening kindly and early saves everyone.
- Know the studies. A parent who quotes Price or Gradisar at you is not being difficult; they have read the coverage. Being able to explain what those studies actually found is what separates a credible practitioner from an opinionated one.
This article is for guidance and education. It is not a substitute for medical advice. Always follow current safer sleep guidance, and encourage families to speak to their GP, midwife or health visitor about a baby you have concerns about.
This is one lesson from the Gentle Sleep course
Accredited at Levels 3 and 4, covering normal infant sleep, circadian rhythms, safe sleep and SIDS, sleep associations, and the strategies that work without leaving a baby to cry.