Do Fathers Matter?: What Science Is Telling Us About the Parent We've Overlooked (19 page)

Hill told me that her research rests heavily on work by Bruce J. Ellis of the University of Arizona, who helped establish the connection between father absence and adverse outcomes for daughters. Ellis calls himself an evolutionary developmental psychologist. He’s interested in whether Darwin’s theory of natural selection can help explain how children’s environments shape their development—precisely the question that came up in Hill’s study. His work on fathers began in the late 1990s, with efforts to test an interesting theory proposed by other evolution-oriented psychologists. The idea was that early experiences could affect children’s reproductive strategies. Early paternal involvement seems to “set” the reproductive strategy that girls use later in their lives.

Ellis quickly discovered that there was something about fathers that gave them a unique role in regulating their daughters’ development—especially their sexual development—around the time of puberty. In a series of studies beginning in 1999, he found that when girls had a warm relationship with their fathers and spent a lot of time with them in the first five to seven years of their lives, they had a reduced risk of early puberty, early initiation of sex, and teen pregnancy.

Ellis continued this work until the early 2000s, demonstrating the phenomenon in different ways, but he became increasingly frustrated. Clearly, the
association
between fathers’ presence or absence and daughters’ maturation was profound. But he could not determine whether paternal behavior caused the consequences he was seeing in the daughters. The idea that father absence was responsible for early puberty in daughters was compelling, but it wasn’t the only possible explanation. An alternative was that girls who begin puberty early and engage in risky sexual behavior do so because of genes they’ve inherited from their parents—the same genes, perhaps, that led to the truncated parental relationship. Fathers whose genes might be linked to their infidelity might pass those genes on to their daughters, where they could be associated with risky sexual behavior and early puberty. A third guess was that something else in the family’s environment was responsible for the accelerated development of the daughters, not their fathers.

Ellis came up with an innovative way to address the question. He considered families in which divorced parents had two daughters separated by an average of seven years in age. When the parents divorced, the older sister would have had seven more years with her father than the younger sister. To put it another way, the younger sister would be “exposed” to seven more years of father absence than her older sibling. If father absence causes early puberty and risky behavior, then the younger daughter should show more of that behavior than her older sibling. And genes or the family’s environment would not confuse the results because the genes are randomly distributed among the daughters. It was something close to a naturally occurring experiment, Ellis realized. He called it a “quasi-experiment.”

Ellis recruited families with two daughters. Some were families in which the parents divorced; others were intact families, to be used as a control group. He wanted to answer two questions: Was the age at which the girls had their first menstrual period affected by the length of time they spent with a father in the house? And did that age vary depending upon how their fathers behaved? The second question was added because fathers with a history of violence, depression, drug abuse, or imprisonment can affect children’s development, and the researchers wanted to know whether those things might also affect the timing of puberty.

Ellis’s suspicions were confirmed. Younger sisters in divorced families, who spent more time without a father in the house, had their first periods an average of eleven months earlier than their older sisters—but only in homes in which the men had behaved badly as fathers. Ellis told me that while they expected to find that the behavior of the fathers affected their daughters, “we were surprised to get as big an effect as we did.” The conclusion was that growing up with an emotionally or physically distant father in early to middle childhood could be “a key life transition” that alters sexual development.

The next step Ellis took was to look at whether these circumstances could affect the involvement of girls in risky sexual behavior. The correlation between the behavior of fathers and risky behavior by their daughters had already been established. An increased risk of pregnancy or of infection with a sexually transmitted disease was clearly more common in daughters of disrupted families or daughters whose relationships with their fathers were marked by harsh conflict, little warmth or support, and lack of parental supervision. Once again, however, Ellis was interested in sorting out whether these were associated because of external or genetic factors, or whether the emotional distance of the fathers had caused the daughters’ risky sexual behavior.

He again sought sisters from divorced families. This time he turned to Craigslist and posted announcements in several cities that began, “SISTERS WANTED!” The criteria were very specific: He was looking for families with two sisters at least four years apart in age and currently between the ages of eighteen and thirty-six. He limited his search to families in which the birth parents separated or divorced when the younger sister was under fourteen years of age. Ellis and his colleagues were able to recruit 101 pairs of sisters, some from families in which the parents had divorced, and, using a different ad, some whose parents had not.

This time, the researchers found that risky sexual behavior in daughters wasn’t simply related to how long they’d lived with their fathers but to the combination of how long they lived with their fathers and what the fathers did in the time they spent with their daughters. “Girls that grew up with a high-quality father—and who lived with him for longer periods of time—showed the lowest level of risky sexual behavior,” Ellis said. “Their younger sisters, who spent less time living with him, tended to show the highest level of risky sexual behavior.” In other words, more years of living with a father who was involved with his children protected girls from risky sexual behavior.

The next question, then, was this: Exactly how do fathers exert this effect on their daughters? One possible explanation, as unlikely as it might seem, is that a father’s scent affects his daughters’ behavior. Many animals emit pheromones, chemical messengers that can be picked up by others and can alter their behavior. “There is certainly evidence from animal research, in a number of species, that exposure to the pheromones of unrelated males can accelerate pubertal development, and some evidence that exposure to pheromones of a father can slow it down,” Ellis explained. “If you take a prepubescent female mouse, a mouse that has not gone through sexual maturation yet, and you stick her in a cage where an adult male has been living and the litter in that cage is saturated with his pheromones—that will cause her to go through puberty more quickly. Just living in that cage.”

Similar effects have been seen with other animals. If the same is true of humans, pheromones could help to explain how the presence or absence of fathers affects their daughters—although that remains an untested hypothesis. The evidence for the action of pheromones in humans is fuzzier, but some research suggests that women who sleep with a male partner have more regular menstrual cycles, perhaps because of the presence of the male’s pheromones.

As we finished our conversation, Ellis brought up something I had been wondering about. What effect does father presence or absence have on sons? He told me that we don’t yet know about sons, and that he himself hasn’t studied them yet. His hypothesis is that a father’s involvement could have a different effect on sons, enhancing a competitive urge and spurring sons to achieve more when they grow up and leave the family. His speculation echoes what we learned about imprinted genes in mice: that a male-imprinted gene does different things in daughters and in sons. In females, it affects the way they care for their young, and in males, the way they compete for mates. A son’s inability to succeed in life could be the human counterpart of a male mouse’s deficiencies as a suitor.

*   *   *

One of the ways we know fathers have close and consequential relationships with their children is that men change when they become fathers—and the changes stay with them. We’ve seen briefly how blood levels of hormones such as oxytocin and prolactin change in animals in response to fatherhood, and that those changes can profoundly affect the behavior of fathers and their relationships with their offspring. Now we can look at those changes more closely, and see how they relate to similar changes in us.

One of the best animals for exploring the role of hormones in fathers is the vole. Like mice and rats, voles are close to us in evolutionary terms. As always, we should be careful about assuming that what’s true in voles is true in men. But as we’ve seen, experiments with animals often give us our first glimpse of what we later find out to be true in humans. So voles are worth paying attention to.

Voles, like deer mice, come in a couple of varieties—the prairie vole and the montane vole. The prairie vole is monogamous, and when it mates, the male guards the female. Oxytocin governs much of this behavior. Give a female oxytocin, and she will bond with whatever male is near: Love the one you’re with. If her oxytocin receptors are blocked, she won’t bond to a male, even after mating. But this doesn’t work in montane voles, a closely related promiscuous species. Pump them up with oxytocin and they’re still promiscuous.

Thomas R. Insel, a psychiatrist and the director of the National Institute of Mental Health, studied voles for fifteen years when he ran the Yerkes Regional Primate Research Center in Atlanta. The two species of voles are 99 percent identical, Insel has said, but it’s the remaining 1 percent that makes them interesting—because it leads to very different social behavior.

Prairie voles are social creatures. Males and females form long-lasting monogamous bonds, and both parents share the care of their young. Both male and female montane voles abandon their young shortly after birth and do not form long-lasting bonds with one another. In the laboratory, Insel observed that prairie voles spent more than 50 percent of their time huddling with their mates. When a prairie vole dies, its mate will usually live alone rather than choose a new partner. Insel and his colleagues wanted to know what it was in the scant bit of DNA that differed between these two species that made them so different with regard to mating and parenting. The answer, he believed, could shed light on the mating characteristics of humans as well.

It soon emerged that one of the key factors underlying vole behavior was indeed oxytocin, which also functions as a neurotransmitter—that is, it is involved in signaling within the brain. Oxytocin is associated with social behavior, childbirth, nursing, sex, and maternal bonding. It causes the uterus to contract during labor, and a synthetic form of it, Pitocin, is routinely used to stimulate or induce labor in pregnant women.

In 1992, Insel looked in the brains of prairie and montane voles for oxytocin receptors—the molecular entryways where oxytocin can attach and allow neurons to respond. He found oxytocin receptors in half a dozen places in the brains of monogamous prairie voles—including some in an area called the nucleus accumbens, the brain’s reward center, which is where cocaine exerts its pleasurable effects. In the polygamous montane voles, in contrast, he found little evidence of oxytocin receptors in those places. To check the findings, he looked at the brains of two other kinds of voles—one monogamous, and one not—and found similar discrepancies. When he looked for such differences related to other hormones, he couldn’t find them. And there was one other intriguing finding. When montane vole females gave birth, the distribution of the oxytocin receptors shifted in certain regions in their brains just as the mothers began to demonstrate maternal behavior.

Males also produce oxytocin, but for them another chemical very similar to oxytocin—known as vasopressin—seems to be the key substance responsible for bonding. In voles, it helps create the bond between males and females. As is the case with oxytocin, the monogamous males have vasopressin receptors in many places where the polygamous voles do not. In 2004, Larry J. Young of Emory University—a collaborator of Insel’s in some of those early experiments—conducted an experiment related to vasopressin that he describes as “mind-blowing.” He and his colleagues took the gene for the vasopressin receptor from a monogamous vole and injected it into the brain of a nonmonogamous vole known as a meadow vole. They injected it into the part of the brain that’s related to reward and addiction. “When we put these transformed meadow voles with a female and allowed them to mate and then we tested them, these animals had formed a bond with that partner.” In every other regard, they behaved exactly as before. “This shows that you can transform behavior, even complex social behavior like bonding, just by changing the expression of a single gene in a single brain area.”

The researchers were also interested in whether the boost in vasopressin receptors would alter the meadow voles’ normally indifferent paternal behavior. They didn’t find differences in such behaviors as how much time males spent licking or huddling with their offspring. But the fathers approached their offspring more quickly and spent more time with the pups than males that didn’t get the boost in vasopressin receptors. Various experiments showed that some manipulations of the hormone would block bonding but not fathering behavior, or vice versa. The conclusion was that bonding and fathering are both related to vasopressin, but that each involves a distinct circuit in the brain—and that those circuits differ from one species of vole to another.

C. Sue Carter, a neuroscientist at the University of Illinois at Chicago and the codirector of the university’s Brain-Body Center, did the original research on male prairie voles—specifically, on what happens when a male vole meets his baby. Among the social, bonding, monogamous prairie voles, males are more likely to engage in spontaneous parenting behavior than virgin females. “A female vole who hasn’t had babies is not parental when exposed to a baby,” she told me. Among males, on the other hand, about 80 percent spontaneously respond to babies as if they already know how to be fathers. “It’s so mysterious,” she says. “There’s no hormonal priming, they haven’t been pregnant, they haven’t seen a baby since they were babies themselves.” Females begin to behave as parents once they’ve given birth. The idea is that exposure to the babies probably leads to hormonal changes. But, Carter said, when the male hasn’t been exposed, he’s still parental. “Why do they become instantly parental, and why does the female not? The baby is like some magic potion, and for some reason males are more sensitive than females.”