Saturday, June 21, 2008

Jumping Into "The Fray" on Cerebral Asymmetry and Sexual Orientation

Fig. S1 (Savic & Lindström, 2008). Part of the left cerebral hemisphere VOI in a male heterosexual subject. Subject’s right side is to the right in the image.

At The Fray, a reader discussion forum at Slate Magazine, a knowledgeable commenter named shusaku provided his own critique of the Savic and Lindström article that purportedly demonstrated the Symmetry Of Homosexual Brain Resembles That Of Opposite Sex.
Slate --> The Fray --> Human Nature

a scientific criticism of the article presented
by shusaku

Since PNAS is not, to my knowledge, a true peer-reviewed journal. I've decided to post a little review of the article addressing its validity.
[NOTE: that is not entirely true. Read the footnote1 for an explanation.]

Shusaku continues with this biting critique:
1) Structural MRI measruements - the regions of interest (ROIs) demarcated in this study included the ventricles, which are not gray or white matter. This indicates that the measurements of structural asymmetry can easily be confounded by differences in ventricular size. In which case, it is equally likely gray or white matter asymmetry cannot be determined by the results presented. There result of differences in structural asymmetry are further undermined by the fact that the cerebellum demarcations only contain gray/white matter and show no differences in asymmetry between subject groups.
Ouch. Why the authors chose to use that particular metric (as opposed to, say, regional brain volumes or cortical thickness) isn't at all clear from the Introduction. In the Discussion, Savic and Lindström stated:
Among studies explicitly comparing the entire hemispheres most, although not all, suggest that the right hemisphere is larger in men (32, 34, 38, 39).
By contrast, in a study on sex differences in cortical thickness, Sowell et al. (2007) opined:
Given the apparent specificity of differences in male and female cognitive advantages, and regional specificity of brain–behavior relationships, global differences in brain size between the sexes that have been readily observed with relatively gross methods might not be the most relevant structural dimorphism when investigating neural substrates of sex differences in cognition.
This is a very important point, because a major result in their study was that gray matter thickness in the right inferior parietal and posterior temporal cortices was 0.45 mm thicker in women than in men [all of unknown sexual orientation], as illustrated below.

Fig. 1 (Sowell et al., 2007). (A) Maps of differences between the sexes in thickness of gray matter (males coded 1, females coded 0 for all maps displayed) for the entire group of 176 subjects showing differences in gray matter (in millimeters) between the male and female subjects according to the color bar on the right. Warmer colors (less than 0 on the color bar) are regions where gray matter thickness is greater in the female than in the male subjects, and cooler colors (greater than 0) are regions where the males have greater gray matter thickness than the female subjects. Note the approximately 0.45 mm increase in cortical thickness in females in the right posterior temporal lobe.

And that's only by way of illustrating Shusaku 's point #1 contra Savic and Lindström. He also criticized the amygdala functional connectivity analysis (a whole 'nother topic), the assumption that learning and environment cannot account for the results, and the exclusion of bisexuals (although all studies seem to exclude them, so that one's not really fair).

But all's fair in science and revulsion:
There are many more technical problems with the analytical techniques used in this study. I'm not going to bother going through all of them. Needless to say, this study proves absolutely nothing, and is just another example of bad science. As an MRI researcher, this article offends me personally, for it gives MRI and PET research a bad rap; I'm going to go throw up now.


1 There are multiple routes to publishing in PNAS. One track is direct submission to the editors, who send it out for peer review. Another track is for a member of the National Academy of Sciences to either sponsor ("communicate") a paper by other authors or to submit ("contribute") his/her own paper (both without peer review from the PNAS Editorial Board). Savic and Lindström submitted to the Editorial Board, so the paper did go out for peer review. However, I noticed something that's not usually allowed, i.e., the acting Editor for the manuscript is affiliated with the same institution as the authors (Karolinska Institutet).


Savic I, Lindström P. (2008). PET and MRI show differences in cerebral asymmetry and functional connectivity between homo- and heterosexual subjects. Proc Natl Acad Sci. June 16, 2008, 10.1073/pnas.0801566105.

Sowell ER, Peterson BS, Kan E, Woods RP, Yoshii J, Bansal R, Xu D, Zhu H, Thompson PM, Toga AW. (2007). Sex Differences in Cortical Thickness Mapped in 176 Healthy Individuals between 7 and 87 Years of Age. Cerebral Cortex, 17(7), 1550-1560. DOI: 10.1093/cercor/bhl066.

Subscribe to Post Comments [Atom]


At June 30, 2008 10:07 PM, Anonymous Anonymous said...

The Neurocritic writes:

However, I noticed something that's not usually allowed, i.e., the acting Editor for the manuscript is affiliated with the same institution as the authors (Karolinska Institutet).

PNAS seems to be taking a more relaxed attitude towards perceived conflicts of interest these days. So the recent Jaeggi et al. (2008) working memory training article that caught a lot of press attention recently was edited by Edward E. Smith even though John Jonides was a co-author. (For those of you who can't spot the possible issue here, Smith and Jonides have co-authored like 30 papers over the years.) To be honest, I don't think it's productive to try to avoid every hint of a perceived COI because the world does not neatly divide into people who are and are not in perceived conflict, and if it did, the two parts would not always be equally qualified to review your stuff. But some things would seem to be best avoided.

At June 30, 2008 11:18 PM, Blogger The Neurocritic said...

Oh! I did not notice that the Jaeggi et al. (2008) article was edited by Edward E. Smith...

It seems that PNAS should have been able to find highly qualified editors (without perceived COI) for both these papers.


At July 01, 2008 10:34 AM, Blogger kopalinski said...

Hello, I have one question to you. Do you have original article of Savic and Lindström (PET and MRI show differences in cerebral asymmetry and functional connectivity between homo- and heterosexual subjects)?

At July 01, 2008 10:41 AM, Blogger The Neurocritic said...

Yes, I have the original article.

At July 01, 2008 10:45 AM, Blogger kopalinski said...

Great. Could you send me it on my email? If it's not a problem for you, this is my address

I would be grateful

At July 01, 2008 10:50 AM, Blogger The Neurocritic said...

Sure, no problem.

At July 01, 2008 10:51 AM, Blogger kopalinski said...

thank you.

Greetings from Poland


At May 19, 2009 8:23 AM, Anonymous Anonymous said...

Hi Neurocritic!
I'd like to get in touch with you for an article I'm writing on the Savic Lindström research for a Swedish magazine, would love to get in touch with you!

Would be very grateful if you could send me an e-mail at karin.lenke [at]

no need to publish this comment, just could't find an e-mail address to you!



Post a Comment

<< Home

eXTReMe Tracker