by Paul Tough
10. Attachment
Meaney and other neuroscientists have found intriguing evidence that something like the LG effect takes place in humans as well. In collaborations with geneticists over the past decade, Meaney and his researchers have been able to demonstrate that a dam’s licking and grooming doesn’t affect her pups just on the level of their hormones and brain chemicals. It goes much deeper than that, all the way down to the control of gene expression. Licking and grooming a rat pup in its earliest days of life affects the way certain chemicals are affixed to certain sequences on the pup’s DNA, a process known as methylation. Using gene-sequencing technology, Meaney’s team was able to establish which part of a pup’s genome got “switched on” by licking and grooming, and it turned out to be the precise segment that controlled the way the rat’s hippocampus would process stress hormones in adulthood.
That discovery alone caused a sensation in the world of neuroscience. It showed that, in rats at least, subtle parental behaviors had predictable and long-lasting DNA-related effects that could actually be traced and observed. What made the discovery relevant beyond the rodent world was an experiment that Meaney’s team then went on to do using the brain tissue of human suicides —some from suicides who had been maltreated and abused in childhood, and some from suicides who had not. The researchers sliced into the brain tissue and examined the sites on the DNA that are related to the stress response in the hippocampus—the human equivalent of the rat DNA sequence that was switched on by early parental behavior, profoundly affecting the rat’s future reactions. They discovered that the suicides who had been maltreated and abused in childhood had experienced methylation effects on that exact segment of their DNA, though the abuse had the opposite effect of licking and grooming: it had switched off the healthy stress-response function that licking and grooming had switched on in the rat pups.
The suicide study is definitely intriguing, but on its own, it does not amount to conclusive evidence about the effect of parenting on stress function in humans. But more solid evidence is beginning to emerge, thanks to some innovative studies that build on Meaney’s research. Clancy Blair, a researcher in psychology at NYU, has been conducting a large-scale experiment in which he has followed almost from birth a group of more than twelve hundred infants. Every year or so, beginning when the infants were just seven months old, Blair measured the way their cortisol levels spiked in reaction to stressful situations—a simple way to evaluate how well a child is handling stress, kind of a bare-bones index of allostatic load. Blair found that environmental risks, like family turmoil and chaos and crowding, did have a big effect on children’s cortisol levels—but only when their mothers were inattentive or unresponsive. When mothers scored high on measures of responsiveness , the impact of those environmental factors on their children seemed almost to disappear. High-quality mothering, in other words, can act as a powerful buffer against the damage that adversity inflicts on a child’s stress-response system, much as the dams’ licking and grooming seemed to protect their pups.
Gary Evans, the Cornell scientist who tested the Simon-playing ability of the cohort of children in upstate New York he has been studying for almost two decades, conducted a similar experiment as Blair’s, though his subjects were in middle school. He collected three different kinds of data for each child: a cumulative-risk score that took into account everything from the ambient noise in a child’s home to the results of a questionnaire about family friction; an allostatic-load measure that incorporated blood pressure, the level of stress hormones in urine, and body mass index; and a rating of maternal responsiveness, which combined the child’s answers to a series of questions about his or her mother with a researcher’s observations of the mother and child playing Jenga together (another Hasbro game!). Evans found mostly what you’d expect: the higher the environmental-risk score, the higher the allostatic-load score—unless a child’s mother was particularly responsive to her child. If that was the case, the effect of all of those environmental stressors, from overcrowding to poverty to family turmoil, was almost entirely eliminated. If your mom was particularly sensitive to your emotional state during a game of Jenga, in other words, all the bad stuff you faced in life had little to no effect on your allostatic load.
When we consider the impact of parenting on children, we tend to think that the dramatic effects are going to appear at one end or the other of the parenting-quality spectrum. A child who is physically abused is going to fare far worse, we assume, than a child who is simply ignored or discouraged. And the child of a supermom who gets lots of extra tutoring and one-on-one support is going to do way better than an average well-loved child. But what Blair’s and Evans’s research suggests is that regular good parenting—being helpful and attentive during a game of Jenga—can make a profound difference for a child’s future prospects.
Some psychologists believe that the closest parallel to licking and grooming in humans can be found in a phenomenon called attachment. Attachment theory was developed in the 1950s and 1960s by a British psychoanalyst named John Bowlby and a researcher from the University of Toronto named Mary Ainsworth. At the time, the field of child development was dominated by behaviorists, who believed that children developed in a mechanical way, adapting their behavior according to the positive and negative reinforcement they experienced. Children’s emotional lives were not very deep, behaviorists believed; an infant’s apparent yearning for his mother was nothing more than an indication of his biological needs for nourishment and physical comfort. The dominant advice to parents in the 1950s, based on behavioral theory, was to avoid “spoiling” infants by picking them up or otherwise comforting them when they cried.
In a series of studies in the 1960s and early 1970s, Ainsworth showed that the effect of early nurturance was exactly the opposite of what the behaviorists expected. Babies whose parents responded readily and fully to their cries in the first months of life were, at one year, more independent and intrepid than babies whose parents had ignored their cries. In preschool, the pattern continued—the children whose parents had responded most sensitively to their emotional needs as infants were the most self-reliant. Warm, sensitive parental care, Ainsworth and Bowlby contended, created a “secure base” from which a child could explore the world.
Although psychologists in the 1960s had at their disposal many tests to evaluate the cognitive abilities of infants and children, they had no reliable way to measure a child’s emotional capacities. So Ainsworth invented a method to do just that, an unusual procedure called the Strange Situation. At Johns Hopkins University, in Baltimore, where Ainsworth was a professor, a mother would bring her twelve-month-old child into a lab set up as a playroom. After playing with her infant for a while, the mother left the room, sometimes leaving the child with a stranger, sometimes leaving him or her alone. After a brief interval, she would return. Ainsworth and her researchers observed the whole procedure through one-way mirrors, and then categorized the children’s reactions.
Most children greeted the returning mother happily, running to her and reconnecting with her, sometimes tearfully, sometimes with joy. These children Ainsworth labeled securely attached, and in subsequent experiments over the past few decades, psychologists have come to believe that they make up about 60 percent of American children. Children who did not have a warm reunion—pretending to ignore the mother when she returned; lashing out at her; falling to the floor in a heap—were labeled anxiously attached. Ainsworth found that a child’s reaction in the Strange Situation was directly related to his parents’ degree of responsiveness in that first year of life. Parents who were attuned to their child’s mood and responsive to his cues produced securely attached children; parenting that was detached or conflicted or hostile produced anxiously attached children. And early attachment, Ainsworth said, created psychological effects that could last a lifetime.
11. Minnesota
But Ainsworth’s contention that early attachment had long-term consequences was, at that po
int, just a theory. No one had figured out a reliable way to test it. And then in 1972, one of Ainsworth’s research assistants, Everett Waters, graduated from Johns Hopkins and entered the PhD program in child development at the University of Minnesota. There he met Alan Sroufe, a rising young star at the university’s Institute of Child Development. Sroufe was intrigued by what Waters told him about Ainsworth’s work, and he quickly embraced her ideas and her methods, setting up a lab with Waters where they could perform the Strange Situation test with mothers and children. Before long, the institute had become a leading center of attachment research.
Sroufe joined forces with Byron Egeland, a psychologist at the university who had received a grant from the federal government to conduct a long-term study on low-income mothers and their children. From the local public-health clinic in Minneapolis, they recruited 267 pregnant women, all about to become first-time mothers, all with incomes below the poverty line. Eighty percent of the mothers were white, two-thirds were unmarried, and half were teenagers. Egeland and Sroufe began tracking this group of children at birth, and they have been studying them ever since. (The subjects are now in their late thirties; both Egeland and Sroufe recently retired.) The evidence the study produced, which Egeland and Sroufe and two coauthors summarized most completely in their 2005 book The Development of the Person, stands as the fullest evaluation to date of the long-lasting effects of early parental relationships on a child’s development.
Attachment classification, the Minnesota researchers found, was not absolute destiny—sometimes attachment relationships changed in the course of childhood, and some children with anxious attachments went on to thrive. But for most children, attachment status at one year of age, as measured by the Strange Situation and other tests, was highly predictive of a wide range of outcomes later in life. Children with secure attachment early on were more socially competent throughout their lives: better able to engage with preschool peers, better able to form close friendships in middle childhood, better able to negotiate the complex dynamics of adolescent social networks.
In preschool, two-thirds of children in the Minnesota study who had been securely attached in infancy were categorized by their teachers as “effective” in terms of their behavior, meaning they were attentive and engaged and rarely acted out in class. Among children who had been observed to be anxiously attached a few years earlier, only one in eight was placed in the effective category; the large majority of those children were classified by their teachers as having one or more behavior problems. (The teachers didn’t know how the kids had done on the Strange Situation.) Children whose parents had been judged disengaged or emotionally unavailable in early assessments of their parenting style did the worst in preschool, and teachers recommended special education or grade retention for two-thirds of them. When teachers ranked students on indicators of dependency, 90 percent of the children with an anxious-attachment history fell in the more dependent half of the class, compared with just 12 percent of children with secure histories. When teachers and other children were surveyed, the anxiously attached children were more often labeled mean, antisocial, and immature.
When the children in the study were ten, researchers invited a randomly selected group of forty-eight students to four-week-long sessions at a summer camp, where they were closely observed and discreetly studied. Counselors (again, unaware of the students’ attachment classifications at one year) rated campers who had had secure attachment in infancy as more self-confident, more curious, and better able to deal with setbacks. The ones with anxious-attachment histories spent less time with peers, more time with the counselors, and more time alone.
Finally, the researchers followed the children through high school, where they found that early parental care predicted which students would graduate even more reliably than IQ or achievement-test scores. Using measures of early parenting only and ignoring the students’ own characteristics and abilities, the researchers found they could have predicted with 77 percent accuracy, when the children were not yet four years old, which ones would later drop out of high school.
It is easy to see parallels between what Michael Meaney’s researchers found in their rat pups in Montreal and what Alan Sroufe and Byron Egeland found in the children they studied in Minnesota. In both cases, certain mothers performed certain specific idiosyncratic parenting behaviors in the earliest days of their children’s lives. And those behaviors—licking and grooming in the rats, responding sensitively to infants’ cues in the humans—seem to have had a powerful and long-lasting effect on the children’s outcomes in a variety of similar ways: the human and rat babies who received the extra dose of early care were, later on, more curious, more self-reliant, calmer, and better able to deal with obstacles. The early nurturing attention from their mothers had fostered in them a resilience that acted as a protective buffer against stress. When the regular challenges of life emerged, even years later—an open-field test, a disagreement among strong-willed kindergartners—they were able, rats and humans alike, to assert themselves, draw on reserves of self-confidence, and make their way forward.
12. Parenting Interventions
There is a direct link between Mary Ainsworth’s research on attachment and Nadine Burke Harris’s clinic in Bayview−Hunters Point, and that link is a San Francisco psychologist named Alicia Lieberman. In the mid-1970s, Lieberman studied with Ainsworth at Johns Hopkins University in Baltimore. It was the era when Ainsworth was conducting her first big study of parenting and attachment, and under Ainsworth’s direction, Lieberman, then a graduate student, spent long hours watching and coding videotape of new mothers interacting with their babies, looking for the small, specific examples of sensitive and responsive maternal behavior that promoted secure attachment for the infants. Today, Lieberman runs the Child Trauma Research Program at the University of California at San Francisco, where she has become, in recent years, a close collaborator with Nadine Burke Harris.
Lieberman told me that while she admires the study that Sroufe and Egeland did in Minnesota, she feels there are two important ideas missing from their analysis. The first is an explicit recognition of how plainly difficult it is for many parents in neighborhoods like Bayview−Hunters Point to form secure attachments with their children. “Often, the circumstances of a mother’s life overwhelm her natural coping capacity,” Lieberman told me when I visited one of the clinics where she works in San Francisco. “When you are bombarded by poverty, uncertainty, and fear, it takes a superhuman quality to provide the conditions for a secure attachment.” In addition, a mother’s own attachment history can make her parenting challenge even greater: research from the Minnesota study and elsewhere shows that if a new mother experienced insecure attachment with her parents as a child (no matter what her class background), then it will be exponentially more difficult for her to provide a secure, nurturing environment for her own children.
The other thing that is underemphasized in the Minnesota study, Lieberman said, is the fact that parents can overcome histories of trauma and poor attachment; that they can change their approach to their children from one that produces anxious attachment to one that promotes secure attachment and healthy functioning. Some parents can accomplish this transformation on their own, Lieberman said, but most need help. And that is what she has spent most of her career doing: figuring out how best to provide that help. In the years after she left Johns Hopkins, she developed a treatment, called child-parent psychotherapy, that combines Ainsworth’s theories on attachment with more recent research on traumatic stress. In child-parent psychotherapy, therapists work with at-risk parents and their infant children simultaneously to improve attachment relationships and protect both parents and children from the effects of trauma. Two therapists in Lieberman’s program now work on site at Burke Harris’s clinic, providing the treatment to dozens of patients.
Lieberman’s treatment is relatively intensive, administered in weekly sessions that can continue for as long as a year. But the principle
behind it—improving children’s outcomes by promoting stronger relationships between children and their parents—is increasingly in use across the country in a wide variety of interventions. And the results, when these interventions are evaluated, are often powerful.
In one study, Dante Cicchetti, a psychologist at the University of Minnesota , tracked a group of 137 families with documented histories of previous child maltreatment; these were families, in other words, where children were at very high risk. Each family had a one-year-old child who was the focus of the intervention. At the beginning of the study, all the infants were evaluated in the Strange Situation procedure, and the results were predictably awful: just one of the 137 infants demonstrated secure attachment, and 90 percent of them were classified as having disorganized attachment, the most problematic type of anxious attachment. Then the families were divided randomly into a treatment group and a control group. The treatment group was given a year of Lieberman’s child-parent psychotherapy, and the control group received the standard community services provided to families reported for maltreatment. When the children were two years old, 61 percent of the ones in the treatment group had formed a secure attachment with their mothers, while in the control group, only 2 percent of the children were securely attached. Attachment-promoting parenting, Cicchetti had shown, can be nurtured in even the most troubled parents, and the benefit to both them and their children can be profound.
