Cover Story

It's the flip side of postpartum depression: The excitement, bonding and utter contentment of new motherhood, bolstered by a flood of "cuddle hormones" that are particularly potent in nursing mothers. At least that's what mothering magazines regularly imply.

But it's probably not that simple. Researchers have had a tough time pinning hormones to postpartum depression, let alone to more subtle behaviors such as bonding and caregiving. In fact, it's only been in the last decade or so that researchers suspected any role of hormones in postpartum mood and parenting behavior. In large part, people assumed women learned to love and nurture babies through experience, not through chemistry.

"In female primates, there's an interest in infants from the start," says Dario Maestripieri, PhD, of the University of Chicago, who looks at the influence of estrogen and progesterone on parenting behavior in monkeys. "It didn't look as if puberty or pregnancy made a big change in willingness to care for infants, and that suggested that there's no need for hormonal changes for females to become maternal. With that knowledge, no one did any research for a long time."

Studies in rodents such as rats and prairie voles began to change that. Researchers found that hormones and peptides such as estrogen, progesterone, oxytocin and vasopressin played a big role in how well these animals cared for their young. Indeed, female rats have little interest in pups until just before delivering their first litter. Researchers can change that by injecting key hormones directly into their brains. And they can make poor parents out of rats and other rodents by restricting those same hormones. It turns out that all the hormones linked to parenting in rodents are also present and active in primates, with spikes near the end of pregnancy and during nursing.

That got researchers thinking and investigating. There's now enough evidence that hormones play at least some role in postpartum mood and parenting behaviors in nonhuman as well as human primates that people are flocking to the field, says Maestripieri. That said, no one expects a clear-cut causal picture to emerge.

"The bottom line in mammals is that hormones are not necessary for individuals to be interested in babies and to be parental," he says. "However, changes in certain hormones might result in changes in parental responsiveness."

Heightened attraction via cortisol

Just how hormones may influence human parents' caregiving is much more difficult to measure than the all-or-nothing experiments researchers can do with rodents. Researchers can't manipulate hormone concentrations in humans. Even just measuring hormone levels can be tricky: the levels of interest are in the brain, which can't be accessed, and blood levels, though helpful, are often not sensitive enough to pick up subtle changes. To top it off, the studies themselves are complicated by ethical restrictions.

These difficulties make the work challenging but not impossible, say researchers.

Probably the best evidence to date that postpartum hormones influence how women interact with their babies comes from the lab of psychologist Alison Fleming, PhD, at the University of Toronto. Her strongest data revolve around the hormone cortisol, typically associated with the body's stress reaction. She and her colleagues find that women with higher levels of cortisol postpartum are more attracted to their babies' scents, and more attentive to them. These behaviors, in turn, have been associated with a stronger mother-infant bond.

"Cortisol postpartum is associated with heightened attraction to infant odors, more sympathy to infant cries and more active maternal-approach behavior in interactions with the infant," says Fleming. And it's not just mothers who show hormonal effects. Studies in rodents, and even a few in humans, show changes in hormones in fathers. In particular, Fleming and her colleagues find that fathers with lower testosterone levels show greater sympathy and need to respond to infant cries. There's also anecdotal evidence that the excitement and stress felt by new adoptive parents may trigger some of the same hormone changes seen in new parents--particularly elevated cortisol levels.

"It may be that we have a physiology that encourages social attachment in all kinds of people," says University of Illinois at Chicago neurobiologist Sue Carter, PhD, who studies the effect of hormones on behavior postpartum in human and nonhuman animals.

That said, Fleming emphasizes that her research finds that correlations between hormones and mothers' behavior are usually mediated by environment or the parents' experience with babies and caregiving.

Buffering the stress response

Other intriguing data have emerged on oxytocin, known as the "cuddle hormone." Its effect in rodents is dramatic--encouraging parenting behavior in virgin females as well as males and making both genders more social in general. Its effect in humans is not as clear cut. Because it's difficult to study in humans, researchers have turned to monitoring an indirect measure--lactation. Women who nurse produce oxytocin in large amounts and those who don't have less.

The main finding to date is that breast-feeding appears to buffer women's stress response. The strongest evidence for this stress buffer comes from years of work in rats. When these animals are lactating, they show a reduced stress reaction in response to experiences such as surgical trauma and electric shock, as measured by several different stress-related hormones.

Encouraged by such strong findings, Carter and her colleagues began to look at how lactating and nonlactating women responded to the stress of strenuous exercise. In a study led by Margaret Altemus, MD, they found that release of three stress-related hormones--adrenocorticotropin hormone, cortisol and vasopressin--in lactating women was cut in half compared with women who bottle-feed their infants.

"Lactating women produced much lower levels of stress hormones in response to exercise," says Carter.

Two new studies--one by Carter and her colleagues and one by a German team led by Marcus Heinrichs, PhD--indicate that breast-feeding can also buffer women from social stress, but the effect is most apparent just after nursing.

"Overall, I think the research in humans tells us--and the animal research supports this--that the lactating female is better able to handle all sorts of stressors," says Carter.

Another take-home message from these studies of lactation, she adds, is that any research looking for a link between hormones, behavior and postpartum mood needs to take into account whether, when and how often women are breast-feeding. Unfortunately, much of the research to date has not accounted for these potentially important differences, says Carter.

Messing with the balance

Researchers are just beginning to think about how modern medicine might influence hormones' effects on behavior and mood--including possible side effects. Among those researchers is Binghamton University psychologist Diane Witt, PhD, who started this fall as program director for behavioral neuroscience at the National Science Foundation. She's concerned about how alopathic medicine may be messing with women's natural hormonal balance through cultural norms and medical interventions. For example, as many as 73 percent of American women who give birth get pitocin--a synthetic form of oxytocin--to induce or speed along labor. Doctors even give the drug to women who have a Caesarean section to promote contraction of the uterus.

"This is a hormone that we know from animal research is implicated in all kinds of behaviors including learning and memory, sex, social behavior and feeding behavior," says Witt. "We may be changing the mother's or baby's brain when we give oxytocin."

Doctors have long contended that the drug, given intravenously, does not pass across the blood-brain barrier into the mother's brain. But no one knows for sure, says Witt. And until we do, we should be more cautious.

Indeed, Swedish researcher Kerstin Uvnas-Moberg, PhD, finds that exposing animals to extra oxytocin during early life permanently reduces their stress reactivity and lowers their baseline blood pressure. And the University of Illinois's Carter finds that male prairie voles exposed to extra oxytocin on the first day of life tend to form pair bonds more quickly in adulthood. In contrast, males exposed to a single shot with an agent that blocks the effect of oxytocin--something used to prevent premature labor in women--were less parental than normal as adults and, in fact, were aggressive toward pups.

Some of these could be good changes, says Witt, but, she adds, "more is not always better. For a chemical to work, it needs receptors and we know you can downregulate receptors by overloading them. So, if, for example, oxytocin is important for the initiation of maternal behavior, and we blast away that system at the time of labor, we could really be affecting subsequent expression of maternal behavior."

So far, this is all theory. But there's enough suggestive data now that people are starting to take notice. And, say researchers, hopefully that means more funding will follow.

Beth Azar is a writer in Portland, Ore.