Beer-guzzling frat boys, catcalling construction workers, and muscle-bound professional wrestlers who prance like peacocks in the ring are often castigated for having too much testosterone. Fathers and husbands who spend hours with their families or spouses are rarely jeered for having too little. Now, research into testosterone’s effect on behavior suggests that married men may experience a unique biological tradeoff that decreases their testosterone levels—thus squelching some of their more flamboyant behaviors and allowing caring and nurturing traits to emerge.

In many animal species, males spend much time locked in testosterone-fueled activities such as fighting and mating. Human males do all this, but also invest in long-term relationships and provide decades of care for their young. This additional dimension puts humans in a special category. “In chimps you don’t find that commitment,” says Peter B. Gray, a graduate student in anthropology. “So the question becomes, what causes men to do this?” Searching for answers, Gray and colleagues compared testosterone levels in married men to levels in their single peers. Their results, published last year in the journal Evolution and Human Behavior, showed married men had significantly lower amounts of the potent hormone.

Gray studied testosterone in saliva collected from 58 men (48 of them Harvard students) between the ages of 20 and 41. Half were married, and of those, 15 were married with children. He took four saliva samples from each man: two in the morning and two in the evening. The subjects also completed questionnaires about their demographic, marital, and parenting backgrounds. Among other things, the questionnaires asked how much time the men spent with their spouses (instead of hanging out with the guys) on their last day off from work, and measured the effort they expended caring for their children. Analysis showed that marriage, fatherhood, and longer periods spent with wives and children were all linked to lower testosterone levels. Fathers in particular had levels significantly lower than those of unmarried men. Researchers also observed that hormone levels in the morning samples were high and relatively even among the men; the differences appeared at night.

"What we have is a correlation that we didn’t know before," says professor of anthropology Peter T. Ellison, Gray’s adviser and a coauthor of the study. "It allows us to look at variation in male hormones in a whole new way." Ellison and Gray posit that major chemical changes occur when men enter relationships or become fathers. Gray explains that the variations between morning and evening saliva samples suggest that something—emotions, experiences, or relationships, perhaps—is driving testosterone down during the day. Both researchers caution that their study is limited in scope, and can’t offer an explanation for what causes testosterone variation. Gray also stresses that his work does not show that fatherhood or marriage causes a hormonal plunge. "I wish I could say that married men experience a drop because of their relationship," he explains. "Our data don’t say either way. Perhaps men with lower testosterone are just more likely to marry."

Still, the study reaches into the dim corners of human evolution and drags out intriguing questions. Have humans engaged in long-term pair bonds for eons, or is this a more recent addition to our behavioral collage? What caused humans to adopt these relationships? What evolutionary advantage stems from nurturing behavior? One theory suggests human males engage in “selfish” reproductive strategies. In other species, males mate as often as possible with a number of females, fighting off rivals and endangering themselves in order to pass on genetic material. Perhaps evolution has imbued human males with another method of securing genetic legacies: instead of pouring their resources into endless rounds of fighting and competing in other ways, men may invest in nurturing relationships and parenting, both reducing their own risk and increasing the likelihood that their offspring will survive.

If testosterone does dip as the result of relationship, Ellison wants to know why. "One of the big remaining questions," he says, "is how does the state of being in a relationship get picked up by your physiology." Together with a group of colleagues, Ellison and Gray are studying testosterone’s role in mating and parenting behavior in a variety of cultural and physical environments. In a series of forthcoming studies, the researchers have examined salival testosterone in Harvard undergraduates and professional-school students—and in Kenyan men, some of whom have two wives. While the suggestion that relationships may douse certain male behaviors is intriguing, Ellison says researchers are ultimately interested in larger questions. "What we’re trying to understand as fully as we can is what our evolutionary pathway has been," he says. "We’ve got a long way to go."

Peter Gray e-mail address: 0dgrayfas [dot] harvard [dot] edu">grayfas [dot] harvard [dot] edu

 Peter Ellison website: www.people.fas.harvard.edu/~pellison/