Prenatal exposure to drugs of abuse: methodological considerations and effects on sexual differentiation.

The pattern of results from the studies reviewed above indicates that alcohol, morphine, nicotine, marijuana, and possibly cocaine can influence reproductive aspects of the neurobehavioral sexual differentiation process to varying degrees. However, with the exception of alcohol, little is currently known regarding the effects of these drugs on nonreproductive sex-related behaviors. Future studies are needed to define the extent of perinatal disruption induced by each drug on the nonreproductive aspect of the sexual differentiation process. It is increasingly clear that the neurobehavioral development of reproductive and nonreproductive behaviors is not influenced to the same degree by alterations in the perinatal hormonal or monoaminergic environment, probably reflecting a fundamental underlying difference in the relative contributions of different brain areas to each behavior (Meaney and McEwen 1986). This fact points to the necessity of greater inclusion of sex-related behaviors in animal models used to assess the teratogenic potential of a given drug on the sexual differentiation process. In light of recent demonstrations of regional structural sex differences in the human CNS (Allen et al. 1989, 1991; deLacoste-Utamsing and Holloway 1982; Hofman et al. 1988; Swaab and Hofman 1988) as well as reports of structural differences in male homosexuals (LeVay 1992; Swaab and Hofman 1990), there is an increasing interest in the contribution of prenatal drug exposure to homosexuality in humans. These findings appear to have led some investigators to interpret behavioral results from animal studies of prenatal drug exposure as being relevant to understanding the causes of homosexuality in humans (Dahlgren et al. 1991; Hard et al. 1984). However, while data from the animal models reviewed above can provide invaluable preclinical evidence to help understand the effects of perinatal drug exposure on brain development and the process of sexual differentiation, the authors believe that the results of these studies provide minimal useful information with respect the prenatal influence of these drugs on homosexual behavior in humans. Animal models of homosexuality are inherently inadequate for several reasons. No adequate model exists for homosexual behavior in the rodent in the absence of pharmacological administration of steroids. Normal male rats that show low levels of masculine sex behavior in the presence of estrous females do not exhibit increased tendencies to mount other males nor to lordosis when mounted by another male. In fact, male preference behavior for an estrous female rat does not appear to be influenced by perinatal androgen exposure (Merkx 1984). A second issue that cannot be addressed in an animal model is the fact that sexual orientation in human is determined by an interaction between hormonal, environmental, and cultural factors (Money 1987). This problem, and others, with a developmental animal model of human homosexuality have been considered by Sachs and Meisel (1988), to which the reader is referred for a more extensive discussion. Finally, in humans there is also the issue of gender identity, which refers to traits or conditions of maleness or femaleness. The degree to which gender identity in humans is causally linked to cultural or biological influences is an area of current debate (Gentile 1993; Unger and Crawford 1993), but such identity is clearly beyond the scope of animal modeling. Therefore, issues related to sexual orientation of humans and prenatal drug exposure likely await data from future human studies for further resolution.