Research · Sleep training
One of the most common questions about infant sleep is whether sleep training that allows the child to cry can be harmful. Some methods involve a period of protest, which has raised concerns about possible effects on stress, attachment and mental health.
This page compiles scientific studies that have examined the effects of sleep training on children's sleep, behaviour and relationship with their parents. The most studied methods are extinction (cry it out) and graduated extinction (the Ferber method), but more gradual methods have also appeared in studies.
26 empirical studies and 1 commentary (plus 6 debate entries), published 1992–2023. Sorted by evidence strength.
Scoping review of 120 studies: sleep training is effective for children with sleep difficulties, with positive effects on both children's sleep and parental wellbeing. More research needed for older children and diverse populations.
Review of 12 studies: sleep training improves parent-reported sleep problems and mothers' health. The longest follow-up (5 years) found no negative effects on the children according to the study's measures of behaviour, stress and attachment.
Review of 11 studies on behavioural sleep interventions for infants under 6 months: 8 of 11 studies showed improved sleep following parent education. Long-term follow-ups found no adverse effects on the children.
Systematic review: existing evidence does not support that sleep training for infants younger than 6 months improves outcomes for mothers or infants.
Background research: parental behaviour at bedtime (especially active soothing) is the factor most strongly linked to children's sleep problems. Examines associations – not an evaluation of sleep training.
Systematic review of 52 studies: 94% of the studies reported that sleep training was effective, with lasting improvements for up to 6 months.
Meta-analysis of 13 RCTs: sleep training improved nocturnal sleep and reduced maternal depression.
Meta-analysis of 10 studies: sleep interventions reduced child sleep problems and improved maternal sleep quality. No clear effect on maternal depression.
Meta-analysis of 32 RCTs: sleep training increased nocturnal sleep by an average of 9 minutes. No effect on daytime sleep.
Meta-analysis of 9 studies: children slept longer overall, but number of wakings did not decrease. Small effect on mothers' mood.
Meta-analysis of 28 studies: sleep training (behavioural methods) improves sleep onset and wakings in young children.
Ferber produced faster sleep onset and longer sleep. No differences in cortisol (at waking) or attachment (at 12 months) between groups.
Improved parental wellbeing. Small effect on children's actual sleep.
Improved sleep in children and reduced depression in mothers. Greatest effect for mothers with higher depression symptoms.
RCT: sleep training using controlled crying (graduated extinction) reduced sleep problems at 2 months. Mothers with high depression scores improved the most.
Swedish RCT: written information was as effective as therapist support for implementing graduated extinction.
No harm to attachment or behaviour. Children cried less at 18 months.
No differences between sleep-trained and control groups at age 6 in the measured outcomes: behaviour, cortisol and indiscriminate friendliness (a simplified attachment measure, not a full attachment assessment).
Background research: mothers' sleep-related beliefs during pregnancy predicted infant sleep patterns during the first year. Examines associations – not an evaluation of sleep training.
Follow-up of the 2007 study: reduced depression in mothers at 2 years. No negative effects on children's behaviour or mental health.
Swedish study: sleep problems, daytime behaviour and family wellbeing improved within two weeks.
2.5-year follow-up: treated infants slept as well as healthy controls. Improvements were stable without ongoing therapist support.
No difference in sleep time or cortisol. Responsive methods were perceived as less stressful.
Children stopped crying but still had elevated cortisol. Small study (n=25).
Everyday crying response patterns in early infancy did not affect attachment. NB: Not about sleep training.
No harmful effects. Children were assessed as having more secure attachment and lower anxiety after sleep training.
Critical review: questions the necessity of behaviour-based sleep methods from biological and cultural perspectives.
In the studies that exist, no clear negative long-term effects have been found on the outcomes measured (behaviour, mental health, stress, the parent–child relationship). Many studies show that parents experience fewer sleep problems and improved wellbeing. The effect on the child's actual sleep duration is, however, small on average. Results primarily apply to children from around 6 months of age.
Read more about sleep training methods or about crying and stress.