Sleep training works short-term, and the follow-up trials find no lasting harm

Behavioural sleep training for infants — methods such as graduated extinction ("controlled crying") and bedtime fading — improves sleep in the short term, and the trials that followed children for years afterward found no evidence that it harms their emotional development, mental health, or attachment to their parents. The same trials also found no lasting benefit once the children grew older: the effects, good and feared, largely wash out with time [s1] [s2] [s3]. That is a more reassuring and more modest picture than either side of the sleep-training debate usually presents.

The trial that measured the feared harms directly

The most pointed study is a randomised controlled trial that set out to test the specific fears attached to sleep training: that letting a baby cry raises stress hormones and damages the parent-child bond. Researchers randomly assigned 43 infants aged 6 to 16 months to graduated extinction, bedtime fading, or a sleep-education control group [s1]. Crucially, they did not just ask parents how things went — they measured infants' stress directly through morning and afternoon salivary cortisol, and, twelve months later, assessed the mother-child bond using the "strange situation," the standard laboratory procedure for classifying attachment security [s1].

On sleep, the interventions worked. There were significant group differences in sleep latency, the time taken to fall asleep (P<.05); in the number of night awakenings (P<.0001); and in wakefulness after sleep onset (P=.01), with large improvements for the graduated-extinction group in particular [s1]. On the harms, the trial found the opposite of what critics predict: salivary cortisol showed small-to-moderate declines in the sleep-training groups compared with controls, not increases [s1]. And at the twelve-month follow-up, there were no significant differences between groups in children's emotional and behavioural problems, and no significant differences in secure-versus-insecure attachment [s1]. The authors' conclusion was that graduated extinction and bedtime fading deliver real sleep benefits "yet convey no adverse stress responses or long-term effects on parent-child attachment or child emotions and behavior" [s1].

The trial is small — 43 infants — which limits how finely it can detect a rare harm. But it is the study that measured the mechanisms the fears are actually about, and it did not find them.

What happens years later

The longer-term question was answered by a five-year follow-up of a much larger population trial. Children with parent-reported sleep problems at seven months had been randomised, at 8 to 10 months, to a brief behavioural sleep programme or usual care; the follow-up reassessed 225 of those families (69%) when the children were six years old [s2].

The result was a broad null. There was no evidence of differences between the sleep-training and control families on any outcome measured: children's emotional scores (P=.8) and conduct scores (P=.6), sleep problems (9% vs 7%, P=.2), chronic stress (29% vs 22%, P=.4), child-parent closeness (P=.1) and conflict (P=.4), disinhibited attachment (P=.3), maternal depression, anxiety and stress (P=.9), and authoritative parenting (63% vs 59%, P=.5) [s2]. The authors concluded that behavioural sleep techniques "have no marked long-lasting effects (positive or negative)" and that parents and professionals can use them to reduce the short-to-medium-term burden of infant sleep problems and maternal depression [s2].

This cuts both ways, and honesty requires saying so. The dreaded long-term harms did not appear — but neither did lasting benefits for the child. By school age, the trained and untrained children looked the same.

The benefit that did persist: mothers

Where a durable effect did show up was in mothers, not children. A separate cluster-randomised trial followed families to the child's second birthday and found that mothers in the intervention group were less likely to report clinical depressive symptoms than control mothers: 15.4% versus 26.4% on a community threshold, and 4.2% versus 13.2% on a stricter clinical threshold [s3]. Parenting style and child mental health did not differ between the groups, and the proportion of children with a sleep problem was similar (27.3% intervention vs 32.6% control) [s3]. The trial concluded that the infant-sleep intervention produced "sustained positive effects on maternal depression symptoms" with "no evidence of longer-term adverse effects" on parenting or child mental health [s3].

This reframes what sleep training is for. In these trials, its clearest lasting payoff was to maternal mental health during a vulnerable period, rather than to the child's later development in either direction.

What the evidence supports

Taken together, the trials support a careful statement: behavioural sleep interventions improve infant sleep in the short term, reduce maternal depressive symptoms, and — in the studies that looked hardest, including direct cortisol and attachment measures — show no evidence of harm to children's stress physiology, attachment, or later mental health, while also showing no lasting developmental benefit [s1] [s2] [s3]. The strongest limitation is sample size for detecting uncommon effects, and the trials studied structured programmes delivered with support, not every version of "cry it out."

What the evidence does not do is tell any particular family whether to sleep-train or how. It describes what happened, on average, to children and mothers in these trials. This article is informational and is not medical advice; concerns about an infant's sleep or a parent's mood are worth raising with a clinician.