The Kids

Cry, Baby, Cry

A journal jumps on the Dr. Sears bandwagon to say sleep training is dangerous. Science says otherwise.

A baby crying it out.
Cry it out

Photo by Oleg Kozlov/Thinkstock

Last week, one of my good friends, who has a 6-week-old baby, went to a breast-feeding support group run by her local hospital. The nurse overseeing the group passed around a copy of the June issue of the journal Clinical Lactation, which was devoted entirely to infant sleep. “You’re not going to like what it says,” she warned my friend and the other exhausted moms in attendance. What it said, quite adamantly, was that “crying-it-out,” the popular term for the sleep training method that involves letting your (not hungry or in-need-of-a-diaper-change) baby cry himself to sleep, is dangerous.

Dangerous how? An editorial in the journal suggests that our common use of crying-it-out may be responsible, at least in part, for the fact that Americans “have a longstanding pattern of poorer health” compared to other high-income nations, where crying-it-out isn’t as popular. Crying-it-out is so stressful to babies that it  “can rearrange certain systems in the brain,” another piece in the issue argues. A third article compares our society’s poor treatment of babies to racism, noting that the editors of the journal Pediatrics—the official journal of the American Academy of Pediatrics—show “great ignorance and disdain for babies” because in 2012 they published the results of a randomized trial suggesting that sleep training does not cause lasting psychological harm. “By allowing this irresponsible and unethical conclusion, the editors are encouraging parents to do great harm to their children and our fellow citizens,” the author, Darcia Narvaez, a psychologist at the University of Notre Dame, proclaims.

My first instinct upon seeing the issue was to burn my copy of Healthy Sleep Habits, Happy Child, the popular sleep-training book that endorses crying-it-out. But I tend to read a lot of medical journals, and I was struck by the brash, sweeping language Narvaez and the other researchers were using. They didn’t sound like scientists who had carefully analyzed the research on crying-it-out and then concluded that it might be harmful. They sounded more like people who had decided that crying-it-out was a bad idea and then looked for science to back up their belief. The same goes for the claims against crying-it-out made by the controversial but highly popular Sears family doctors—William, Robert, James, and Peter. The language they use is scary and vague—more dogmatic than scientific.

Despite the fevered debate around sleep training, I have never had strong opinions about crying-it-out, in part because I never needed to consider it—my son was a good sleeper. (I know, I know: You hate me.) But with this journal issue in hand, I decided it was time for me to figure it out. I’ve learned that many of the claims made in Clinical Lactation and by the Sears family are, at best, overstatements of the science; at worst, they completely misconstrue it. Worse, they don’t mention many of the studies that suggest sleep training to be safe, nor do they seem to consider the important fact that sleep training works and therefore gives babies and their parents much-needed shut-eye, which is crucial for healthy development (not to mention good parenting).

The first argument made by the journal (and the Sears pediatricians) is that crying-it-out is stressful for babies, flooding their sweet little brains with hormones such as cortisol that interfere with healthy brain development. Yes: Chronic, toxic stress is bad for young brains. But several nights of crying during an otherwise happy infancy does not constitute chronic stress. The types of stress that have been shown to cause developmental problems in children include sexual and physical abuse and serious neglect, described as “the absence of sufficient amounts of essential experiences,” which some children raised in state-run institutions in China and Romania suffered, for instance. The Sears family also cites a study showing that babies who cry for prolonged periods have lower IQs than kids who don’t, but the study looked at kids who happened to cry a lot more than usual, not those who were left alone to cry. (The Sears family confused the direction of the causal arrow in the study, too, because the researchers concluded that “underlying neurological problems may be the cause of prolonged crying,” not the other way around.) When Time magazine senior editor Jeffrey Kluger contacted the authors of some of the studies cited by William Sears for an article he wrote in May 2012, the scientists said it was unfair that Sears had used their work as evidence against sleep training. “Our paper is not referring to routine, brief stressful experiences, but to abuse and neglect,” explained Yale psychologist Joan Kaufman. “It is a mis-citation of our work to support a non-scientifically justified idea.”

There is, however, one study cited multiple times in Clinical Lactation that specifically assesses the effects of crying-it-out on infant cortisol levels. In this 2012 study, University of North Texas educational psychologist Wendy Middlemiss and her colleagues tracked the behavior and cortisol levels of 25 infants, ages 4 to 10 months, as they went through a five-day sleep training program à la crying-it-out in a sleep lab in New Zealand. The researchers measured the blood cortisol levels of the infants and their mothers before and after the babies were put to sleep on the first and the third nights. On the first night, the babies cried, but their cortisol levels did not go up upon being left to cry themselves to sleep. On the third night, none of the babies cried much at all, and their cortisol levels stayed constant. In other words, the sleep training worked.

What Middlemiss considers a cause for concern is that the babies’ cortisol levels never dropped during the course of the study (the mothers’ cortisol levels did, on the third night). “What’s dangerous about the situation,” Middlemiss explained to me, “is that the mother has no idea—because the behavioral indication [i.e. the crying] is absent—that the infant had remained stressed.” But how do we know the infants were stressed to begin with if their cortisol levels never rose? Middlemiss says that the babies in her study must have been stressed as soon as they arrived at the sleep lab—it’s a foreign environment—so it didn’t matter that the training itself didn’t incite a further increase. But there were no comparison groups included in the study to validate such a conclusion—no infants lounging around at home with much lower cortisol levels—so it’s unfair to assume that the sleep-trained infants were stressed. (Middlemiss told me that she once took cortisol measurements of babies while they were at home, and that their cortisol levels were lower than the babies tested in the sleep lab, but she didn’t include this information in her study.) Even if the babies at the sleep lab were highly stressed, the obvious take-home is that parents shouldn’t bring their babies to sleep labs—not that they shouldn’t cry it out.

Another word that gets thrown around a lot when people talk about sleep training is “attachment.” Attachment is an extremely misunderstood concept; basically, it describes a child’s relationship with his mother or father as it develops over the course of the first year of life. A child who is securely attached to his mother is confident that she is there for him, because she has been repeatedly and appropriately responsive to his cues and needs. A child who is not securely attached is not so sure mom can be counted on, because she has been unpredictable in her responsiveness or perhaps even abusive. Securely attached children go on to have stronger relationships throughout life and are more confidant, cooperative, caring and emotionally stable than those who aren’t securely attached to their caregivers. Attachment is powerful—no question—and important.

But attachment isn’t extremely fragile, nor is there a formula parents need to follow to ensure that it develops. “Don’t get me wrong: I think nursing is great for lots of reasons, co-sleeping is fine, and carrying a baby in a sling is great, but you can do all of those things and not be a sensitive parent,” says Alan Sroufe, a professor emeritus at the University of Minnesota Institute of Child Development, who has studied attachment for decades. And no, you won’t threaten a secure attachment with your baby if you let her cry at night a few times, either: Sroufe let his own daughter cry it out for a few days when she was about 8 months old. (It worked.) ”Did I think she would be traumatized by this? No. This business of being sensitive and responsive—it’s about being sensitive and responsive the vast majority of the time,” he says. (Sroufe believes, however, that crying-it-out is inappropriate for younger babies; some researchers have drawn a “safety” line at 6 months of age because that’s when infants develop object permanence, the ability to understand that mom and dad still exist when they’re not visible.)

But what if it takes weeks of intense crying, night after night, to sleep train your child? Here’s the thing: When crying-it-out is done properly, the experts say, it doesn’t take weeks. It takes days. In Middlemiss’ study, the babies stopped crying by the third night. A 1988 trial also reported significant improvements in infant sleep within three days using the method. Yet we’ve all heard horror stories about parents who have had to endure weeks upon weeks of all-night screamfests. I asked Marc Weissbluth, a pediatrician at Northwestern University and the author of the best-selling Healthy Sleep Habits, Happy Child, about the discrepancy, and he says that crying-it-out can take a long time, but typically only if the parents “have the child’s bedtime too late, or they’re not napping the child, or they’re doing intermittent reinforcement,”—i.e. they’re going back in to soothe the child instead of truly letting them cry it out. (Extremely overtired babies resist sleep training, and parents who soothe their babies during training reward the crying, giving them reason to do it again and again.) Fix these problems, Weissbluth says, and crying-it-out should work in three days.

It’s not just Weissbluth saying this. When the American Academy of Sleep Medicine reviewed the literature on infant and child sleep training, it reported that in 17 out of 19 published studies, unmodified extinction—the clinical term for crying-it-out—effectively reduced bedtime resistance and the frequency of nighttime wakings, concluding that it “has a strong record of accomplishment.” The two published clinical trials on graduated extinction, the technique popularized by sleep researcher Richard Ferber, which involves leaving your baby to cry for increasing periods of time (but not necessarily all night), was deemed successful, too, but it takes longer. None of the studies found side effects associated with sleep training. In a 2012 randomized clinical trial, Australian researchers followed up with 173 6-year-olds who had been sleep trained as babies, some of whom with graduated extinction, and found that they were no different than non-sleep-trained 6-year-olds with regards to emotional development, psychological health, parent-child closeness, and parental attachment.

And it’s not just that a short stint of sleep training isn’t harmful—it could actually be beneficial. People tend to underestimate the importance of sleep, but it’s absolutely crucial for healthy development, and longitudinal studies have suggested that sleep-deprived infants often go on to become sleep-deprived children. (Here’s some info on how much sleep children of different ages need.) Sleep deprivation affects parts of the developing brain involved in regulating emotions and thinking logically. Kids who don’t sleep well are also more likely to injure themselves. Sleep-deprived adults, as in parents who are up at all hours of the night tending to sleepless babies, are much worse at deciphering emotional cues and being emotionally expressive themselves—problems that could potentially threaten the parent-child attachment bond. Several trials have also found that sleep training reduces the risk of maternal depression—by as much as a factor of three—which is (actually!) chronically stressful for kids; a 2009 study reported that infants of depressed moms were more likely to be anxious and socially disengaged than were infants of healthy moms.

Crying-it-out is not for every parent, I know. But desperate parents—or parents who just want to be done with the 2 a.m. wake up— should feel fine trying the method. It’s not just that there’s no evidence of harm in crying-it-out—there is some solid evidence of no harm. When sleep training works, and research suggests it often does, it can provide long-term benefits for the entire family—giving babies the sleep they need to develop into healthy toddlers and giving parents the rest they need to be sensitive, confident, and happy caregivers.