First, a few basic facts about genetics and behavior:
- Genes are sections of DNA in our chromosomes, located in each cell nucleus, that contain the instructions for our bodies to develop and function. Genes are the units of heredity passed down from parents to their offspring. (2)
- Genes contain the blueprint for building our bodies, but they do not determine our actions. There is no human behavior that is genetically dictated. As Dr. Neil Whitehead notes, “No genetically determined human behavior has yet been found.” (3)
- Human behavior develops from a complex interaction of biological, chemical, familial, societal and spiritual influences. Our behavior also includes elements of volition – choice.
- Any attempt to say that a behavior is caused by a gene is deterministic and mechanistic – it turns humans into automatons and eliminates our will. Our ability to make choices is an essential part of what makes us human.
So what about twin studies in general?
- Identical twins develop when one fertilized egg separates into two individuals. As such, identical twins have nearly identical DNA. Researchers study the concordance of a trait in identical twins to see whether or not a trait is genetically-linked. In a concordant pair of twins, both show the same characteristic. On the other hand, in a discordant pair only one twin has the characteristic. Higher concordance levels may mean a higher possibility of genetic causation.
- Researchers also study similarities and differences between pairs of identical twins and pairs of fraternal twins, who develop from two separate fertilized eggs. By comparing the two groups, researchers hope to find clues about the whether or not a characteristic is genetically caused. While twin studies are often helpful in the study of traits that have a direct physical expression, they are less reliable when it comes to behavioral traits.
- Twin studies are legitimately criticized because of the possibility of errors – including environmental-genetic interaction, shared intrauterine influences of identical twins, and probability of greater similarity of treatment of identical twins. As such, “these pitfalls and others imply that data from human twin studies should be interpreted with caution and reservation.” (4)
- Even if twin studies examining behavior were considered to be valid and reliable, the twin studies that have been conducted to date actually demonstrate that homosexuality is not genetic. As Dr. Neil Whitehead notes, “In a nutshell, if you take pairs of identical twins in which one twin is homosexual, the identical co-twin is usually not homosexual. That means, given that identical twins are always genetically identical, homosexuality cannot be genetically dictated. No-one is born gay (emphasis his).” (5)
- In general, twin studies show that when one male identical twin is gay-identified, the other will be gay-identified one time in nine, or only 11% of the time. The percentages are similar for female homosexuality. If homosexual behavior, attractions and identity were completely genetic, the similarity would be closer to 100%. What this tells us is that same-sex attracted twins are not born that way. Thus, within a pair of identical twins – where one twin has same-sex attractions – the other will have opposite-sex attractions eight out of nine times. (6)
So what about this 2008 Swedish twin study?
About the study:
- Sweden keeps a twin registry, and researchers invited twins between the ages of 20 and 47 to participate in a web-based survey or telephone interview (Swedish Twin Study of Adults: Genes and Environments – STAGE). (7)
- Among the other data gathered, twins who participated were asked for the lifetime number of opposite-sex and same-sex partners.
- From these two questions, the authors of the study constructed two variables – whether the individual twins had ever had a same-sex partner and the total number of same-sex partners.
About the results:
The study of twins and homosexuality was touted in a press release as “the largest in the world so far.”(8) The authors said that “more than 7,600 Swedish twins….responded to a 2005-2006 survey of health, behaviour, and sexuality.”(9) This is true, but the press release failed to state that in their analysis:
- Only 7 pairs of male identical twins were found where both had any same-sex partner in their lifetime;
- Only 26 pairs of female identical twins were found where both had any same-sex partner in their lifetime;
- Only 64 pairs of identical male twins were found where only 1 of the pair had any same-sex partners;
- And only 188 pairs of identical female twins were found where only 1 of the pair had any same-sex partners. (10)
Significantly, their data show a very low number of identical twin pairs who had both engaged in same-sex behaviors. The data also show a very low concordance rate (where both twins show the same trait) – less then 10% for the males and slightly over 12% for the females. Again, the study points to what we already know: there is a low concordance rate for homosexuality among identical twins, and the main factors in homosexuality are not genetic.
The study also examined the incidence of same-sex behavior for fraternal twins:
- 3 male pairs of fraternal twins were found where both had same-sex partners;
- 13 female pairs of fraternal twins were found where both had same-sex partners;
- And 50 male pairs were found where only one of the pair had any same-sex partners;
- And 127 female pairs were found where only one of the pair had any same-sex partners.
Here, the analysts used a statistical formula to compare the fraternal and identical pairs. Again, the validity of this type of statistical analysis is problematic because of the assumptions it makes and is challenged by many researchers.
From this formulaic comparison, this is what the study authors reported for male identical twins:
- The genetic effects explained 39% of the variance in having a same-sex partner;
- Shared environment accounted for 0% variance;
- And the unique environment of each twin accounted for 61% of the variance.
Using this formula, the proportions of the genetic and shared environmental effects are calculated based on differences in concordance rates among the identical and fraternal pairs, and the unique environment effect is what is leftover – that which cannot be adequately explained by either of the other components.
Wide confidence intervals:
The study had extremely wide confidence intervals for all the results. The confidence interval is similar to a margin of error – it shows how reliable the results are. A wide confidence interval shows a very unreliable result. For the male identical twins, the confidence intervals were unacceptably wide:
- The genetic effects could be as low as 0% or as high as 59%;
- Likewise, the shared environment could have been as low as 0% or as high as 46%;
- And the unique environmental effect could be as low as 41% or as high as 85%. (11)
Here, it should be noted that the confidence intervals were very large for all the other groups in the analysis. In short, the study’s analysis of genetic vs. environmental influence is almost meaningless because the confidence intervals are so large. And, notably, the large margin of error reveals that the influence of genetics on choosing same-sex partners could very well be zero. That is, that genetics do not contribute to same-sex sexual behavior at all. In fact, even the authors say that “wide confidence intervals suggest cautious interpretation,”(12) but neglect to mention this in their press release.
In addition to analyzing whether both twins in a pair had ever had same-sex partners, the researchers also looked at the total number of lifetime same-sex partners. They only found two male and twelve female pairs of identical twins with the same number. This demonstrated overwhelmingly low concordance rates among male and female identical twins – less than 3% and 6%, respectively. Once again, this section of the study showed very wide confidence intervals – error boundaries. These three factors – the small number of pairs, the low concordance rate and the wide confidence intervals – give evidence to the unreliability of this study. Even if their analyses were accurate, the results showed the much greater influence of environment over genetics.
Our Conclusions:
The press release for the study tells us that “society's attitudes have little impact on choice of sexual partner.” In reality, the authors studied nothing about society’s attitudes or about how people choose sexual partners.
Here’s what one of the researchers said in the press release, “The results show, that familial and public attitudes might be less important for our sexual behaviour than previously suggested. Instead, genetic factors and the individual’s unique biological and social environments play the biggest role. Studies like this are needed to improve our basic understanding of sexuality and to inform the public debate.”(13)
Because the study did not study familial and public attitudes and did not give us meaningful data with accurate measures, we are left to conclude that the authors were motivated to shape public attitudes. They did not find that homosexuality was genetically determined. A better title for their press release would be: “Study of small sample of twin pairs – with wide error boundaries – shows high influence of environment on homosexual behavior.”
