What self-soothing, sleep training, and attachment research tell us about how babies learn to feel safe.
It is three in the morning. The baby has been fed and changed. You’ve walked the same stretch of hallway so many times you know which floorboard will creak. For a moment, the crying eases. You lower the baby toward the crib, and it starts again.
There is a particular kind of loneliness in that hallway. You’re exhausted. Someone has told you that picking up the baby will create a habit you’ll spend months undoing. Someone else has told you that every minute of crying leaves a mark on the brain. You can hear both voices in your head, and neither one is there to take the next shift.
I want to begin with the baby in your arms. Crying is one of the few ways a young infant can get your attention when something feels wrong. You may discover hunger, a wet diaper, fatigue, or too much noise. Sometimes you will make your way through every possibility and still have a crying baby in your arms. Your response still matters. The child is having a hard moment with someone who is trying to help. American Academy of Pediatrics guidance on responding to cries
We make a mistake when we expect a baby to manage distress they still need help with. We make another when we call every instance of self-settling a trauma. The research draws a much more useful distinction between an emerging skill, a specific approach to sleep, and a life in which care is chronically unavailable. That distinction matters most to the people hearing the crying in real time. Infant sleep observations · Randomized infant sleep trial · Research on severe early deprivation
The two meanings of self-soothing
Watch a baby wake between stretches of sleep. There may be a little stirring, a small sound, a hand that finds its way to the mouth, and then sleep again. Researchers who recorded infants at night observed some babies returning to sleep without a caregiver’s intervention. How often this happened varied with age and from one baby to another. This is one meaning of self-soothing, though self-settling describes more precisely what a camera can see. Goodlin-Jones and colleagues, nighttime observations · Burnham and colleagues, first-year follow-up
The same word can carry a very different instruction. A baby is crying hard. The adult has been told that going in will spoil the lesson, so the adult stays away even as the distress builds. If this becomes part of a broader pattern in which the baby’s distress routinely goes unanswered, the concern is the repeated lack of care. A brief pause to see whether a sleepy whimper passes is a different experience from repeatedly having no one available when you need help. American Academy of Pediatrics guidance on sleep and age · Bucharest Early Intervention Project
As an engineer, I find it helpful to separate what we can observe from what we infer. We can hear the crying stop. From that alone, we have very little information about why it stopped. The baby may be drifting into sleep. They may have found a way to settle. The outward signal and the body’s stress response can also follow different timelines. Quiet is a measurement we can make from the hallway. Feeling safe is a much larger question. Middlemiss and colleagues, infant crying and cortisol
How a baby borrows our calm
The word co-regulation sounds clinical for something families do without naming it. A father lowers his voice as he lifts the baby. A grandmother turns away from the television and walks into a quieter room. An aunt pauses the bouncing when she notices the baby has stiffened. The adult pays attention, offers something, and adjusts to the response. Sometimes the baby needs food, sometimes less stimulation, and sometimes someone to stay close while the crying runs its course. American Academy of Pediatrics guidance on soothing
Researchers have observed a piece of this process in the still-face experiment. A caregiver and baby interact, the caregiver briefly holds a neutral expression and stops responding, and then interaction resumes. In a study of 43 five- and six-month-olds, babies with more responsive caregivers showed better recovery in heart rate and outward distress during the final reunion. The lab task offers a small window into how an infant responds when a familiar person becomes available again. Haley and Stansbury, infant stress and reunion
In yoga, when I speak about holding space, I mean being able to stay present with someone through an experience I cannot have on their behalf. I think of that when a parent says, “I picked her up and she still cried.” You can check what needs checking, ease your grip, slow your voice, and give her a little time with one kind of comfort before trying the next. She may keep crying. There is room in good caregiving for a feeling that takes time to pass.
What the brain scans actually studied
Some of the most important imaging research on early deprivation comes from the Bucharest Early Intervention Project in Romania. Young children who had been living in institutions were randomly assigned either to a foster care program or to the care available at the time. The institutions exposed children to deprivation across many parts of daily life, including limited social, emotional, and language stimulation. At around age eight, MRI scans showed differences in the microstructure of some white matter pathways, including pathways involved in emotion and cognition. In several pathways, the children assigned to foster care showed patterns closer to those of children raised in families. Bick and colleagues, randomized foster care and white matter study
The amygdala, a brain structure involved in processing emotionally significant information, has been studied in this context too. One study found larger adjusted amygdala volumes among children adopted later out of institutions. A separate study of previously institutionalized adolescents did not find larger amygdala volumes. These are studies of early deprivation across a child’s daily life, with different samples and outcomes. They cannot isolate the effect of unanswered crying or give us a minute-by-minute threshold for harm. Tottenham and colleagues, amygdala study · Hodel and colleagues, adolescent MRI study
The finding I keep returning to in the Bucharest project is what happened when care changed. The foster care program altered much more than who came when a child cried. It gave children a family setting in which they could build relationships with caregivers. Researchers found gains in attachment security as well as differences in some brain measures. The study gives us a reason to take early care seriously and to make dependable relationships available to children who have had too few of them. Smyke and colleagues, foster care and attachment · Bick and colleagues, white matter development
What cortisol can tell us about a crying baby
Cortisol is a hormone involved in the body’s stress response. What matters depends on what researchers measure. They might look at the change after a challenge, the time it takes to come back down, or the way levels rise and fall over a day. These are different questions. In a follow-up from the Bucharest project, children assigned to institutional care showed a smaller cortisol response to a laboratory stressor years later than children assigned to foster care. The simple story that early deprivation always keeps cortisol high misses how differently the stress system can respond over time. McLaughlin and colleagues, early caregiving and stress reactivity
A small study often cited in sleep debates followed 25 infants, ages four to ten months, during a residential sleep program. By the third day, the researchers observed less outward distress around sleep onset, while the infants’ cortisol after settling had not declined in parallel with the mothers’. The study had no comparison group, and its first measurements were taken on the first day of the program. It leaves open how closely visible distress tracks physiology in this setting. Its design could not establish how the program changed cortisol relative to ordinary nights at home, or whether there was any lasting harm. Middlemiss and colleagues, residential sleep study
A baby’s cry is worth attending to without a hormone test to certify the distress. And when researchers do take a hormone measurement, it deserves to be read on its own terms. A sample from one evening and a stress response measured years after severe deprivation each tell us something about the people and circumstances studied.
What the sleep-training trials found
In a 2016 randomized trial, researchers studied 43 infants ages six to sixteen months. They compared sleep education with two approaches. One used gradually spaced caregiver check-ins. The other, called bedtime fading, temporarily set bedtime closer to when the infant naturally fell asleep and then moved it earlier. Both approaches shortened the time it took to fall asleep. At the twelve-month follow-up, researchers detected no group differences in the attachment or emotional and behavioral outcomes they measured. With 43 babies, the trial had limited power to pick up small effects. Its cortisol samples were taken in the morning and afternoon, so they could not tell us what happened physiologically during a particular episode of bedtime crying. Gradisar and colleagues, randomized sleep trial
A larger Australian trial followed families offered behavioral sleep support when their infants were around eight to ten months old. At age six, the researchers detected no differences on measured outcomes such as child emotional and conduct problems or parent-child closeness. About 69 percent of the families eligible for the follow-up participated. Taken together, these trials give some reassurance about the approaches studied in older infants. A newborn whose feeding and sleep are still taking shape calls for a different approach. The American Academy of Pediatrics encourages prompt responses to crying in the first few months. Price and colleagues, five-year follow-up · American Academy of Pediatrics guidance for the first months
One family may choose to rock their baby to sleep. Another may change the bedtime routine because everyone is running out of sleep. The choice deserves attention to the baby’s age, feeding needs, and health, as well as what the adults can sustain. A change at bedtime tells us very little about a relationship without knowing how that baby is cared for through the rest of the night and day.
How a relationship handles the missed moments
Attachment takes shape across many exchanges. A large meta-analysis found that children tended to show more secure attachment when caregivers were more sensitive to their cues. The link appeared for mothers and fathers and was modest. A single caregiving habit can tell us only so much about a relationship. And in a randomized trial with families at risk for neglect, a program that coached caregivers in how to respond more nurturingly led to higher rates of secure attachment than the comparison program. The way adults respond can change, and children can benefit from that change. Madigan and colleagues, caregiver sensitivity meta-analysis · Bernard and colleagues, randomized caregiving intervention
Sensitive caregiving also shows up in research on how children learn. A 2026 synthesis found modest associations with thinking and language skills and a much smaller association with fewer social and emotional difficulties. These studies examined caregiver sensitivity broadly. Think of the person who follows a baby’s gaze, names the dog walking past, or pauses so the baby has a turn to babble. Everyday responsiveness includes those exchanges too. Nivison and colleagues, review of developmental outcomes
Even face-to-face play contains plenty of missed signals. An infant looks away and the adult keeps trying to make them smile. A few seconds later, the adult notices and gives them space. The infant looks back, and the exchange begins again. Researchers Edward Tronick and Jeffrey Cohn observed repeated shifts between coordination and miscoordination in ordinary play. I find that comforting because it gives us something we can do after a hard moment. We can notice what we missed and try again. Tronick and Cohn, face-to-face coordination
Picture a parent who has spent ten minutes trying to rock a baby who wanted to feed. At last the parent recognizes the cue and changes course. There is no speech to make about those ten minutes. The next thing is the feed. Repair at this age is often that ordinary.
What to do on a hard night
When a young baby cries, start with what the cry could be asking for. Check feeding, the diaper, temperature, and signs of discomfort. Try a steady hold, a quiet voice, or gentle rocking. An older baby who briefly stirs and is already settling may need a little space. When the crying escalates or sounds unusual for your baby, go closer and see what is happening. If crying is unusual or persistent, or the baby seems ill, contact their pediatric clinician. American Academy of Pediatrics guidance on crying · American Academy of Pediatrics guidance on infant sleep
If you feel close to losing control, put your baby on their back in an empty crib or bassinet with a firm, flat sleep surface. Step away briefly, check back, and ask another adult to take over if someone is available. Never shake a baby. There are nights when the most caring thing one adult can say to another is, “Give me the baby. Go lie down.” American Academy of Pediatrics guidance for overwhelmed caregivers · American Academy of Pediatrics safe sleep policy
I think we misunderstand independence when we hurry it along in someone who has only just arrived. There will be a day when this child reaches for your hand and then lets go to cross the room. Tonight, perhaps they need you to sit with them in the dark. Let that be enough. And if there is a second pair of hands in the house, let those hands carry some of the night too.
Frequently Asked Questions
Will I spoil my baby by picking them up every time they cry?
This is one of the most persistent myths, and the attachment research points the other way. For an infant, crying is communication, not manipulation, and responding to it does not create a bad habit, it teaches the baby that the world is reliable and that they are safe. You cannot spoil a baby by meeting their need to be comforted; you are building the foundation they will stand on later.
What does secure attachment actually mean?
It is the deep sense a child develops that when they are distressed, someone comes, consistently and reliably. That repeated experience of being answered wires in a basic trust: that they matter, that the world responds, that they can reach out and be met. It is not about being a perfect parent; it is about being a dependable one often enough that the pattern holds.
What if I cannot respond every single time?
You do not have to be flawless, and the research is reassuring here. Secure attachment grows from good-enough responsiveness over time, not perfection in every middle-of-the-night moment. There will be nights you are exhausted and slow to the crib; what matters is the overall pattern of a baby who learns, again and again, that someone comes. Be gentle with yourself as well as with the baby.

