Out on a Limb

by Andrew Sullivan

  Does this apply to humans? The evidence certainly suggests that it does, though much of the “proof” is inferred from accidents. Pregnant women who were injected with progesterone (chemically similar to testosterone) in the 1950s to avoid miscarriage had daughters who later reported markedly tomboyish childhoods. Ditto girls born with a disorder that causes their adrenal glands to produce a hormone like testosterone rather than the more common cortisol. The moving story, chronicled in John Colapinto’s book As Nature Made Him, of David Reimer, who as an infant was surgically altered after a botched circumcision to become a girl, suggests how long-lasting the effect of fetal testosterone can be. Despite a ruthless attempt to socialize David as a girl, and to give him the correct hormonal treatment to develop as one, his behavioral and psychological makeup was still ineradicably male. Eventually, with the help of more testosterone, he became a full man again. Female-to-male transsexuals report a similar transformation when injected with testosterone. One, Susan/Drew Seidman, described his experience in The Village Voice last November. “My sex-drive went through the roof,” Seidman recalled. “I felt like I had to have sex once a day or I would die.… I was into porn as a girl, but now I’m really into porn.” For Seidman, becoming a man was not merely physical. Thanks to testosterone, it was also psychological. “I’m not sure I can tell you what makes a man a man,” Seidman averred. “But I know it’s not a penis.”

  The behavioral traits associated with testosterone are largely the cliché-ridden ones you might expect. The Big T correlates with energy, self-confidence, competitiveness, tenacity, strength, and sexual drive. When you talk to men in testosterone therapy, several themes recur. “People talk about extremes,” one man in his late thirties told me. “But that’s not what testosterone does for me. It makes me think more clearly. It makes me think more positively. It’s my Saint John’s wort.” A man in his twenties said: “Usually, I cycle up the hill to my apartment in twelfth gear. In the days after my shot, I ride it easily in sixteenth.” A forty-year-old executive who took testosterone for body-building purposes told me: “I walk into a business meeting now and I just exude self-confidence. I know there are lots of other reasons for this, but my company has just exploded since my treatment. I’m on a roll. I feel capable of almost anything.”

  When you hear comments like these, it’s no big surprise that strutting peacocks with their extravagant tails and bright colors are supercharged with testosterone and that mousy little male sparrows aren’t. “It turned my life around,” another man said. “I felt stronger—and not just in a physical sense. It was a deep sense of being strong, almost spiritually strong.” Testosterone’s antidepressive power is only marginally understood. It doesn’t act in the precise way other antidepressants do, and it probably helps alleviate gloominess primarily by propelling people into greater activity and restlessness, giving them less time to think and reflect. (This may be one reason women tend to suffer more from depression than men.) Like other drugs, T can also lose potency if overused. Men who inject excessive amounts may see their own production collapse and experience shrinkage of their testicles and liver damage.

  Individual effects obviously vary, and a person’s internal makeup is affected by countless other factors—physical, psychological, and external. But in this complex human engine, testosterone is gasoline. It revs you up. A 1997 study took testosterone samples from 125 men and 128 women and selected the 12 with the lowest levels of testosterone and the 15 with the highest. They gave them beepers, asked them to keep diaries and paged them twenty times over a four-day period to check on their actions, feelings, thoughts, and whereabouts. The differences were striking. High-testosterone people “experienced more arousal and tension than those low in testosterone,” according to the study. “They spent more time thinking, especially about concrete problems in the immediate present. They wanted to get things done and felt frustrated when they could not. They mentioned friends more than family or lovers.”

  Unlike Popeye’s spinach, however, testosterone is also, in humans at least, a relatively subtle agent. It is not some kind of on-off switch by which men are constantly turned on and women off. For one thing, we all start out with different baseline levels. Some women may have remarkably high genetic T levels, some men remarkably low, although the male-female differential is so great that no single woman’s T level can exceed any single man’s, unless she, or he, has some kind of significant hormonal imbalance. For another, and this is where the social and political ramifications get complicated, testosterone is highly susceptible to environment. T levels can rise and fall depending on external circumstances—short term and long term. Testosterone is usually elevated in response to confrontational situations—a street fight, a marital spat, a presidential debate—or in highly charged sexual environments, like a strip bar or a pornographic web site. It can also be raised permanently in continuously combative environments, like war, although it can also be suddenly lowered by stress.

  Because testosterone levels can be measured in saliva as well as in blood, researchers like Alan Booth, Allan Mazur, J. Richard Udry, and particularly James M. Dabbs, whose book Heroes, Rogues and Lovers will be out this fall, have compiled quite a database on these variations. A certain amount of caution is advisable in interpreting the results of these studies. There is some doubt about the validity of onetime samples to gauge underlying testosterone levels. And most of the studies of the psychological effects of testosterone take place in culturally saturated environments, so that the difference between cause and effect is often extremely hard to disentangle. Nevertheless, the sheer number and scale of the studies, especially in the last decade or so, and the strong behavioral correlations with high testosterone, suggest some conclusions about the social importance of testosterone that are increasingly hard to gainsay.

  Testosterone is clearly correlated in both men and women with psychological dominance, confident physicality, and high self-esteem. In most combative, competitive environments, especially physical ones, the person with the most T wins. Put any two men in a room together and the one with more testosterone will tend to dominate the interaction. Working women have higher levels of testosterone than women who stay at home, and the daughters of working women have higher levels of testosterone than the daughters of housewives. A 1996 study found that in lesbian couples in which one partner assumes the male, or “butch,” role and another assumes the female, or “femme,” role, the “butch” woman has higher levels of testosterone than the “femme” woman. In naval medical tests, midshipmen have been shown to have higher average levels of testosterone than plebes. Actors tend to have more testosterone than ministers, according to a 1990 study. Among seven hundred male prison inmates in a 1995 study, those with the highest T levels tended to be those most likely to be in trouble with the prison authorities and to engage in unprovoked violence. This is true among women as well as among men, according to a 1997 study of eighty-seven female inmates in a maximum-security prison. Although high testosterone levels often correlate with dominance in interpersonal relationships, it does not guarantee more social power. Testosterone levels are higher among blue-collar workers, for example, than among white-collar workers, according to a study of more than four thousand former military personnel conducted in 1992. A 1998 study found that trial lawyers—with their habituation to combat, conflict, and swagger—have higher levels of T than other lawyers.

  The salient question, of course, is, How much of this difference in aggression and dominance is related to environment? Are trial lawyers naturally more testosteroned, and does that lead them into their profession? Or does the experience of the courtroom raise their levels? Do working women have naturally higher T levels, or does the prestige of work and power elevate their testosterone? Because of the limits of researching such a question, it is hard to tell beyond a reasonable doubt. But the social context clearly matters. It is even possible to tell who has won a tennis match not by watching the game, but by monitoring testosterone-fille
d saliva samples throughout. Testosterone levels rise for both players before the match. The winner of any single game sees his T production rise; the loser sees it fall. The ultimate winner experiences a postgame testosterone surge, while the loser sees a collapse. This is true even for people watching sports matches. A 1998 study found that fans backing the winning side in a college basketball game and a World Cup soccer match saw their testosterone levels rise; fans rooting for the losing teams in both games saw their own T levels fall. There is, it seems, such a thing as vicarious testosterone.

  One theory to explain this sensitivity to environment is that testosterone was originally favored in human evolution to enable successful hunting and combat. It kicks in, like adrenaline, in anticipation of combat, mental or physical, and helps you prevail. But a testosterone crash can be a killer too. Toward the end of my two-week cycle, I can almost feel my spirits dragging. In the event of a just-lost battle, as Matt Ridley points out in his book The Red Queen, there’s a good reason for this to occur. If you lose a contest with prey or a rival, it makes sense not to pick another fight immediately. So your body wisely prompts you to withdraw, filling your brain with depression and self-doubt. But if you have made a successful kill or defeated a treacherous enemy, your hormones goad you into further conquest. And people wonder why professional football players get into postgame sexual escapades and violence. Or why successful businessmen and politicians often push their sexual luck.

  Similarly, testosterone levels may respond to more long-term stimuli. Studies have shown that inner-city youths, often exposed to danger in high-crime neighborhoods, may generate higher testosterone levels than unthreatened, secluded suburbanites. And so high T levels may not merely be responses to a violent environment; they may subsequently add to it in what becomes an increasingly violent, sexualized cycle. (It may be no accident that testosterone-soaked ghettos foster both high levels of crime and high levels of illegitimacy.) In the same way, declines in violence and crime may allow T levels to drop among young inner-city males, generating a virtuous trend of further reductions in crime and birth rates. This may help to explain why crime can decline precipitously, rather than drift down slowly, over time. Studies have also shown that men in long-term marriages see their testosterone levels progressively fall and their sex drives subsequently decline. It is as if their wives successfully tame them, reducing their sexual energy to a level where it is more unlikely to seek extramarital outlets. A 1993 study showed that single men tended to have higher levels of testosterone than married men and that men with high levels of testosterone turned out to be more likely to have had a failed marriage. Of course, if you start out with higher T levels, you may be more likely to fail at marriage, stay in the sexual marketplace, see your testosterone increase in response to this, and so on.

  None of this means, as the scientists always caution, that testosterone is directly linked to romantic failure or violence. No study has found a simple correlation, for example, between testosterone levels and crime. But there may be a complex correlation. The male-prisoner study, for example, found no general above-normal testosterone levels among inmates. But murderers and armed robbers had higher testosterone levels than mere car thieves and burglars. Why is this not surprising? One of the most remarkable, but least commented on, social statistics available is the sex differential in crime. For decades, arrest rates have shown that an overwhelmingly disproportionate number of arrestees are male. Although the sex differential has narrowed since the chivalrous 1930s, when the male-female arrest ratio was twelve to one, it remains almost four to one, a close echo of the testosterone differential between men and women. In violent crime, men make up an even bigger proportion. In 1998, 89 percent of murders in the United States, for example, were committed by men. Of course, there’s a nature-nurture issue here as well, and the fact that the sex differential in crime has decreased over this century suggests that environment has played a part. Yet despite the enormous social changes of the last century, the differential is still four to one, which suggests that underlying attributes may also have a great deal to do with it.

  This, then, is what it comes down to: testosterone is a facilitator of risk—physical, criminal, personal. Without the influence of testosterone, the cost of these risks might seem to far outweigh the benefits. But with testosterone charging through the brain, caution is thrown to the wind. The influence of testosterone may not always lead to raw physical confrontation. In men with many options it may influence the decision to invest money in a dubious enterprise, jump into an ill-advised sexual affair, or tell an egregiously big whopper. At the time, all these decisions may make some sort of testosteroned sense. The White House, anyone?

  The effects of testosterone are not secret; neither is the fact that men have far more of it than women. But why? As we have seen, testosterone is not synonymous with gender; in some species, it is the female who has most of it. The relatively new science of evolutionary psychology offers perhaps the best explanation for why that’s not the case in humans. For neo-Darwinians, the aggressive and sexual aspects of testosterone are related to the division of labor among hunter-gatherers in our ancient but formative evolutionary past. This division—men in general hunted; women in general gathered—favored differing levels of testosterone. Women need some testosterone—for self-defense, occasional risk-taking, strength—but not as much as men. Men use it to increase their potential to defeat rivals, respond to physical threats in strange environments, maximize their physical attractiveness, prompt them to spread their genes as widely as possible, and defend their home if necessary.

  But the picture, as most good evolutionary psychologists point out, is more complex than this. Men who are excessively testosteroned are not that attractive to most women. Although they have the genes that turn women on—strong jaws and pronounced cheekbones, for example, are correlated with high testosterone—they can also be precisely the unstable, highly sexed creatures that childbearing, stability-seeking women want to avoid. There are two ways, evolutionary psychologists hazard, that women have successfully squared this particular circle. One is to marry the sweet class nerd and have an affair with the college quarterback: that way you get the good genes, the good sex, and the stable home. The other is to find a man with variable T levels, who can be both stable and nurturing when you want him to be and yet become a muscle-bound, bristly gladiator when the need arises. The latter strategy, as Emma Bovary realized, is sadly more easily said than done.

  So over millennia, men with high but variable levels of testosterone were the ones most favored by women and therefore most likely to produce offspring, and eventually us. Most men today are highly testosteroned, but not rigidly so. We don’t have to live at all times with the T levels required to face down a woolly mammoth or bed half the village’s young women. We can adjust so that our testosterone levels make us more suitable for co-parenting or for simply sticking around our mates when the sexual spark has dimmed. Indeed, one researcher, John Wingfield, has found a suggestive correlation in bird species between adjustable testosterone levels and males that have an active role to play in rearing their young. Male birds with consistently high testosterone levels tend to be worse fathers; males with variable levels are better dads. So there’s hope for the new man yet.

  From the point of view of men, after all, constantly high testosterone is a real problem, as any fifteen-year-old boy trying to concentrate on his homework will tell you. I missed one deadline on this article because it came three days after a testosterone shot and I couldn’t bring myself to sit still long enough. And from a purely genetic point of view, men don’t merely have an interest in impregnating as many women as possible; they also have an interest in seeing that their offspring are brought up successfully and their genes perpetuated. So for the male, the conflict between sex and love is resolved, as it is for the female, by a compromise between the short-term thrill of promiscuity and the long-term rewards of nurturing children. Just as the female does, he optim
izes his genetic outcome by a stable marriage and occasional extramarital affairs. He is just more likely to have these affairs than a woman. Testosterone is both cause and effect of this difference.

  And the difference is a real one. This is so obvious a point that we sometimes miss it. But without that difference, it would be hard to justify separate sports leagues for men and women, just as it would be hard not to suspect judicial bias behind the fact that of the ninety-eight people executed last year in the United States, 100 percent came from a group that composes a little less than 50 percent of the population; that is, men. When the discrepancy is racial, we wring our hands. That it is sexual raises no red flags. Similarly, it is not surprising that 55 percent of everyone arrested in 1998 was under the age of twenty-five—the years when male testosterone levels are at their natural peak.

  It is also controversial yet undeniable that elevating testosterone levels can be extremely beneficial for physical and mental performance. It depends, of course, on what you’re performing in. If your job is to whack home runs, capture criminals, or play the market, then testosterone is a huge advantage. If you’re a professional conciliator, office manager, or teacher, it is probably a handicap. Major League Baseball was embarrassed that Mark McGwire’s 1998 season home-run record might have been influenced by his use of androstenedione, a legal supplement that helps increase the body’s own production of testosterone. But its own study into andro’s effects concluded that regular use of it clearly raises T levels and so improves muscle mass and physical strength, without serious side effects. Testosterone also accelerates the rate of recovery from physical injury. Does this help make sense of McGwire’s achievement? More testosterone obviously didn’t give him the skill to hit seventy home runs, but it almost certainly contributed to the physical and mental endurance that helped him do so.