Genetics help determine sexual orientation

Or at least you would think that was the finding when reading the above headline from a UPI story.

Canadian scientists have discovered evidence that genetic factors play a major role in determining whether a man is homosexual or heterosexual. The researchers found an important brain structure -- the posterior region of the corpus callosum -- is larger in homosexual than in heterosexual men. The brain structure consists of a thick band of nerves that connect the left and right hemispheres of the brain.

All right so where is the genetics component? I decided to go look up the article itself. Here is the (truncated methods):

Measures and Procedure

MR Imaging
Each participant underwent an MR imaging research protocol..

Corpus Callosum (CC)
Area measures of the total mid-sagittal surface of the hemisphere, the CC, and four CC subregions, defined to approximate major cortical regions, were made ...

Handedness
Hand-preference scores were based on demonstrated preferences on a nine-item test ...

Cognitive Tests
A small battery of tests was given to each participant to document their cognitive status...

Data Analysis
Independent sample t-tests were used to compare group differences ...

No mention of DNA microarrays, SNP analysis, or anything else remotely resembling genetics. So, right there, the news article should be different. It should say "Neuroanatomy helps determine ...". Wait a minute:

A sample of 12 homosexual men who met the following specific criteria was recruited through advertisements on a university campus and in community-based organizations ... The control group consisted of ten heterosexual men, who were recruited via advertisements on a university campus...

I don't see anything experimental about this setup. It is a purely correlational study. All right, well, at least their data is convincing right?




P=0.06? You don't even have p<0.05? Oh, well you do a one-tailed test and then p=0.03. I guess that's makes everything better. At least they have a fairly large effect size in the isthmus (cohen's d=0.83). Of course, you measured six different ares so you kind of expect one of them to look good.

This wasn't a bad study per se, but the data are just not strong on their own. Since other groups have already shown that such sexual orientation based neuroanatomical differences exist, this study just adds some further evidence. But by far the worst part is seeing the press take research findings and garble them all up for public consumption. It is nothing that a little fact checking wouldn't fix ... before going to press.