From the Carolina Population Center, University of North Carolina at Chapel Hill, Chapel Hill, NC.
Address correspondence to: J. Richard Udry, Carolina Population Center, University of North Carolina at Chapel Hill, CB# 8120, University Square, 123 West Franklin St., Chapel Hill, NC 27516-2524; email@example.com
Certain principles underlying the papers by Titus-Ernstoff et al. 1 and Sandberg et al. 2 may benefit from further comment. It is widely believed by social scientists that human behavior is shaped exclusively by social conditioning. More specifically, differences in the behavior of the sexes are seen as exclusively a “social construction,” designed uniquely by each society, and with no biological foundation. 3
The authors of these two papers blithely and felicitously mostly ignore this social science theory. 1,2 They operate under a widely accepted biological theory of the origin of sex differences that applies to humans as well as to all other mammals. This theory has stronger theoretical and empirical support than the social construction theory. It goes like this: each species has specific sex differences in behavior that are natural to its repertoire, but fetuses are behaviorally feminine (predisposed to behave in the ways that are species-typical of females) until androgenized (acted upon by testosterone) either prenatally or perinatally, depending on the species. In humans this is a second-trimester event. Because female fetuses do not make any testosterone themselves, they remain feminine. Their only source of testosterone is their mothers’ blood, which contains a small amount of testosterone. Male fetal testes produce lots of testosterone and masculinize the fetus—that is, make it predisposed to behave in ways that are species-typical of males. It does this in a perverse way. When testosterone enters cells that contain the enzyme aromatase, such as brain cells, aromatase transforms testosterone into estrogen. (Estrogen cannot directly enter the brain.) Estrogen then physically masculinizes the male brain permanently. Female fetal brain cells contain aromatase, but because they receive very little testosterone, they are not masculinized (very much). 4 At birth, human infants already have sex-dimorphic brains, and they are predisposed to behave in sex-dimorphic ways characteristic of male and female humans as a species.
“At birth, human infants already have sex-dimorphic brains. . .”
Our authors take for granted the process I have described. Their problem is to discover whether diethylstilbestrol (DES), environmental polychlorinated biphenyls (PCBs) or other “endocrine disrupters” can modify or disrupt the normal process. DES is a synthetic estrogen, widely but unsuccessfully used in the 1950s to prevent miscarriage until its use was prohibited. Subsequent studies were carried out to determine the effects of exposure, if any, on aspects of sex-dimorphic behavior in humans and laboratory animals. These studies variously show feminizing effects, masculinizing effects, both effects, no effects and even masculinizing effects on females and feminizing effects on males. 5–8 PCBs and other androgen disrupters have generally been shown in other animals to be antiandrogens in hormonal transactions, but whether or how they affect sex-dimorphic behaviors in humans is not understood. 9 Fish-eating humans are presumably affected by whatever potpourri of endocrine disrupters is in the fish. These disruptors are variously androgenic, antiandrogenic, estrogenic and antiestrogenic. To sort out the behavioral effects of such a mixture may be hopeless.
There are two theoretical criteria for selecting the behaviors to be examined in studying effects of prenatal biological intruders on gendered behaviors. The first is that the behaviors must be responsive to testosterone. The second is that the behavior must be sex-dimorphic. By the nature of the behavior selection process, the dependent behaviors meet both criteria in the study by Sandberg et al. 2 In the Titus-Ernstoff et al. 1 paper, the dependent behaviors meet neither criterion. There are psychosexual behaviors that are testosterone-dependent but not gender-dimorphic.
Finally, we should remember that adolescent and adult sex-dimorphic behavior is responsive to circulating adolescent and adult testosterone as well as to prenatal hormone structuring of the brain. The adult effects are activational, that is, dependent on current levels of testosterone. According to my research on women, sensitivity to the activational effect is inversely dependent on the amount of prenatal androgenization. 10 If the process works the same way for men as well, one implication of this inverse interaction is that females are more sensitive to activational effects, whereas males are more sensitive to prenatal effects. The other implication is that prediction of adult gendered behavior from prenatal events omits the perhaps equally important (and interacting) simultaneous effects of current hormone exposure on the same behaviors. One has only part of the picture, perhaps an entirely misleading part. Sandberg et al. 2 will not need to worry so much because their subjects are prepubertal, but for the Titus-Ernstoff et al.1 sample of adults this omission is a big concern.
Meanwhile, what about the social scientists who do not believe the biological theory in the first place? They must find these papers futile. But not all is lost. No one has said that their theory is all wrong. I have shown elsewhere 10 that the biological and the social science theories work simultaneously. Women with low prenatal androgen exposure showed very strong childhood gender socialization effects on behavior. Parental efforts to make them more feminine really worked. But women who had higher prenatal androgen exposure showed progressively attenuated effects from parental efforts at feminine socialization. The most androgen-exposed women were more masculine, and seemed to be immunized against the parents’ substantial efforts to feminize them.
Researchers into the possible prenatal effects of endocrine disrupters and other biological intruders should take note. Parents are not passive observers of their children’s gendered behaviors. Parents have ideas about how they want their children to behave. They work hard to achieve their goals. They often succeed. They may succeed in wiping out the very effects the hormone-based theory tells us to expect.
Neither of these papers shows credible effects on gendered behaviors of prenatal exposures to synthetic hormones or environmental endocrine disrupters. Are these results any surprise? Are the effects really not there? We really do not know. The natural process is complicated beyond the ability of a single research study to comprehend. What with hormones changing into other hormones, prenatal plus adult effects that inversely interact, hormonal effects interacting with socialization effects—combined with parents’ trying to steer the boat—it is truly amazing that we have made as much progress as we have.
About the Author
J. RICHARD UDRY is Kenan Professor in the Departments of Sociology and Maternal/Child Health at the University of North Carolina, Chapel Hill. His scientific passion is integrating biological and social science models of behavior.
1. Titus-Ernstoff L, Perez K, Hatch EE, et al.
Psychosocial characteristics of men and women exposed prenatally to Diethylstilbestrol. Epidemiology 2003; 14: 155–160.
2. Sandberg DE, Vena JE, Weiner J, Beehler GP, Swanson M, Meyer-Bahlburg HFL. Hormonally active agents in the environment and children’s behavior: Assessing influences on gender-dimorphic outcomes. Epidemiology 2003; 14: 148–154.
3. Browne KR. Biology at Work: Rethinking Sexual Equality. New Brunswick, NJ: Rutgers University Press, 2002.
4. Panksepp J. Affective Neuroscience: The Foundations of Human and Animal Emotions. New York: Oxford University Press, 1998.
5. Berenbaum SA. Effects of early androgens on sex-typed activities and interests in adolescents with congenital adrenal hyperplasia. Horm Behav 1999; 35: 102–110.
6. Ehrhardt AA, Meyer-Bahlburg HF, Rosen LR, et al
. The development of gender-related behavior in females following prenatal exposure to diethylstilbestrol (DES). Horm Behav 1989; 25: 526–541.
7. Lisk JD, Ehrhardt AA, Meyer-Bahlburg HFL, Rosen LK, Gruen RS, Veridiano N. Gender-related behavior development in females exposed to diethylstilbestrol (DES) in utero: an attempted replication. J Am Acad Child Adolesc Psychiatry 1991; 30: 29–37.
8. Reinisch J, Sanders SA. Effects of prenatal exposure to diethylstilbestrol (DES) on hemispheric laterality and spatial ability in human males. Horm Behav 1992; 26: 62–75.
9. National Research Council. Exposures: sources and dynamics of hormonally active agents in the environment. In: Hormonally Active Agents in the Environment. Washington DC: National Academy Press, 1999; 52–76.
10. Udry JR. Biological limits of gender construction. Am Sociol Rev 2000; 65: 443–457.