Swedish twin study: Q & A with J. Michael Bailey

Recently, I posted here and here about a new study from Sweden involving twins and sexuality (Niklas L氓ngstr枚m, Qazi Rahman, Eva Carlstr枚m, Paul Lichtenstein, (2008). Genetic and Environmental Effects on Same-sex Sexual Behaviour: A Population Study of Twins in Sweden. Archives of Sexual Behaviour, June 7, published online.) To get additional perspective on this new report, I asked prominent sexuality researcher, J. Michael Bailey from Northwestern University to comment. My thanks to Dr. Bailey for his time and expertise.

Throckmorton: How are the models used in twin studies able to separate the family influence into genetic and shared environmental effects? Clearly the environmental factors shared by twins do not look influential in this study. Many critics of this kind of work will wonder how the shared family environment cannot be influential. Hence, the question about how we can tell the family influence is genetic and not from the common parenting or social environment.
Bailey: Most studies use twins who have been reared together from birth. Shared environment is the environment shared by siblings reared together, and thus, identical twins and fraternal twins are equally similar in their shared environment (provided they were reared together, they share all of it). Nonshared environment consists of environmental factors not shared even by siblings in the same household. Nonshared environment is the kind of environment that causes differences even between identical twins reared together. The logic of the common twin study (as opposed to the rare study of identical twins reared apart) depends on their being two types of twins that vary in their genetic similarity. The assumption is that this is the only salient way that such twins differ–other differences are presumed not to have an affect on the trait of interest. If this assumption is true, then one looks to see whether identical twins are more similar than fraternal twins. If they are, this is evidence for genetic effects. To the extent that identical and fraternal twins are equally similar to each other (and more similar than random people are to each other), this suggests there are shared environmental factors. To the extent that identical twins reared together are different from each other, this supports the importance of the nonshared environment.
Throckmorton: What are the advantages of this study over your 2000 population based study (Bailey et al, 2000)? What are the limitations in contrast to your work?
Bailey: I think this study is comparable in quality to our 2000 study, although our measures of sexual orientation were more comprehensive.
Throckmorton: In the new twin study, the genetic effects are estimated at 35-39% for men and 18-19% for women. Since we are not assuming one gene codes for sexual orientation, do you have thoughts about what genetic mechanism(s) could be at work? The pedigree studies have had mixed results, have they not?
Bailey: Studies such as this one provide little if any insight into the nature of genetic mechanisms. Rather, they estimate the magnitude of genetic effects, whatever those effects are. Pedigree studies are studies that look at patterns of similarity in families, and there has been some inconsistency, with some studies finding evidence consistent with X-linkage, and others not finding such evidence. But virtually all studies have found higher rates of homosexuality among family members of gay men.
Throckmorton: The estimates of the effect of non-shared environmental factors are very high. The authors indicate that hormones in utero could be an aspect of the non-shared environment. Given that these twins shared the same womb, what kinds of non-shared effects are they referring to?
Bailey: No one knows how this can happen. First of all, nonshared environment is the kind of environment that causes identical twins reared together to be different. This nonshared environment is poorly understood but is clearly important. We know, for example, that if one identical twin has a congenital major brain anomaly such as microcephaly (being born with a very small cortex), the other twin is usually normal. Obviously, this is due to some kind of prenatal environmental factor or event. (It can’t be genetic, because they have the same genes.) We really haven’t a clue what it is.
Differences between identical twins (or fraternal twins for that matter) also reflect measurement error, and the amount of such error depends on how good the measure of sexual orientation (in this case) is. I would say the measure used in the study at hand is okay but not great.
Throckmorton: By the measure of sexual orientation, I assume you are referring to the fact that this study did not assess sexual fantasy or attraction independent of behavior. How do you think a measure of sexual fantasy would impact the results? Do you think we might get lower effects of environment on the inner world?
Bailey: I have no idea, honestly.
Throckmorton: Concerning environment, some critics of pre-natal theories might suggest that twins are not often treated alike and do not really share environments, even though they live in the same house and share the same parents. Is there an empirical answer to this concern?
Bailey: Twin researchers have typically concocted indices of how similarly twins were treated as children. Parents vary in their twin-rearing philosophies, with some favoring “treat ’em alike” and others favoring “treat ’em differently.” Typically, these indices don’t predict how similar twins become.
Throckmorton: Could the non-shared environment also include peer influences, different socialization experiences, trauma, etc.?
Bailey: Yes. Regarding sexual orientation, though, there is both anecdotal and empirical evidence that relevant twin differences arise early. For example, gender nonconformity differences in childhood are common and predict orientation differences in adulthood.

Dr. Bailey’s answer to the last question is an intriguing finding. One twin might display gender nonconformity and the other might not. These differences do arise early and cause me to question that they are often in response to parenting differences. The GNC differences however, will likely lead to very different social environments for each twin.
Going forward, research on non-shared environmental factors would shed light on how environment helps to shape the pre-natal givens.
*Bailey, J.M., Dunne, M.P., & Martin, N.G. (2000). Genetic and Environmental influences on sexual orientation and its correlates in an Australian twin sample. Journal of Personality and Social Psychology, 78, 524-536.