A Passing Moment…

… Or, Who’s the Fairest of them All?

For years, clinicians, therapists, researchers, and transfolk alike have remarked that “younger transitioners”, transkids, “homosexual transsexuals”, “early onset” (whatever label or demarcator in fashion) MTF transsexuals simply ‘pass’ better than “older transitioners”, autogynephilic transsexuals, “late onset” MTF transsexuals.  For years, I wanted to conduct a study about this.  Well, now we have clinical data to test this observation.

In a study conducted in Europe, by our favorite Netherlands based researchers, they looked at both body image and clinician assessment of gender incongruent physicality.  That is to say, how well or poorly they pass.  The same single clinician rated them all, so while a different clinician may give folks a different absolute score, the relative scores for all subjects is likely to be very accurate.

In deference to the currently debated question and researching the potential validity of which signifier is the accurate basis for a taxonomic typology of transsexuality, the data was presented for both sexual orientation and age of onset.

MtF                                                                 FtM
Androphilic      Non-androphilic           Gynephilic Non-gynephilic
Early onset 88 (70%) 102 (43%)                   193 (88%)   37 (69%)
Late onset  38 (30%)   139 (57%)                   26 (12%)     17 (31%)

The Dutch have long contended that age of onset was the salient signifier, while those in North America contend that it is sexual orientation, specifically “homosexual” vs. “non-homosexual”, which readers of my blog, and those familiar with the literature, know gives a strong signal / correlation with autogynephilia in MTF transsexuals.

In the graphs below, a higher score means more gender incongruent appearance (i.e. ‘readable’), while a lower score means more gender congruent (i.e. ‘passable’).

Now, looking at our earlier observation, do exclusively “homosexual” transsexuals pass better than “non-homosexual”?  For transwomen, the answer is a resounding “YES!”, with a large effect size (d=0.7).  Putting this into everday language, this is to say that the most passible of the non-androphilic transwomen are just barely comparable to the average androphilic transwoman.  Or another way of putting it, nearly half the androphilic transwoman pass better than nearly all non-androphilic.  Or yet another way of putting is that the least passible androphilic is the same as the average non-androphilic transwoman.

However, keep in mind that we know that many “late transitioners” misreport their sexual orientation because of Social Desirability Bias and Autogynephilic Pseudo-Androphilia.  From several studies we know that perhaps 38% report that their sexual orientation ‘changed’ from exclusively gynephilic to androphilic or bisexual. and that estimates of misreported sexual orientation means that from 20-40% of the self-reported androphilic group is in fact, non-androphilic, which would tend to pull the data toward the non-androphilic value.  Even with that possibility, the data still shows that androphilic MTF transwomen pass far better than non-androphilic.  The data also shows a greater range, standard deviation, which we would expect if 20-40% of the self-reported androphilic were in fact a mixture of the two populations.

Dr. Joy Shaffer and Kay Brown at ages 27 & 26.  Joy is non-androphilic, transitioned at age 21/22. Kay is androphilic, transitioned at age 17/18.

So what of our question about early vs. late onset?  Here again, early onset passes better.  But look closely at the data, 46% of the early onset group are androphilic.  If we hypothesize that the salient signifier is sexual orientation and NOT age of onset, then we would expect that the relative score for early onset would be intermediate between androphilic and both non-androphilic and late-onset (which is predominately non-androphilic at 79%).  And that indeed is what we see.  Further, one would predict that since late onset is predominately non-androphilic, that they would have the same level of passability… and indeed that is what we see again.  The data clearly supports the position that sexual orientation and NOT age of onset is the salient signifier, given the stronger signal.  That is to say, variation in the data is explained completely by sexual orientation and that the variation of passability with respect to age of onset is from the correlation between sexual orientation and age of onset.

It is clear that there is a mild correlation with sexual orientation and age of onset, with androphilic MTF transsexual more likely to report early onset (70% vs. 43%).  However, given clinical experiences with each, the meaning of age of onset is quite likely different.  If 43% of non-androphilic transwomen really did have an early onset… why do they all wait so long to socially transition?  The modal age is about 35 years and the average is about 40 years old compared to the 20 years old for androphilic.  I contend that retrospective age of onset is time shifted to an earlier age due to Social Desirability Bias AND to having a different internal meaning to the question.  For transkids (androphilic transwomen), the age of onset is demarcated by extreme somatic and social dysphoria, while for non-androphilic the demarcator is retrospectively found by their strong but vacillating autogynephilic desire for somatic transformation, but with little as yet stable social or somatic dysphoria.  As per Doctor and Prince, it takes considerable time for true gender dysphoria and cross gender identity to develop in non-androphiles / autogynephiles.

We are still left with an open question.  Why do androphilic transwomen pass so much better than non-androphilic?  Three possible hypotheses exist, 1) Having a truly earlier age of onset and social transition age, they experience less masculinization from endogenous androgens. 2) Self selection for passibility as they are motivated to fit into society better, being both physically and behaviorally extremely gender atypical (and not autogynephilically motivated). 3) Actually being, as a group, intrinsically more physically gender atypical.  (That is to say, that the etiological cause for their behavioral gender atypicality causes physical atypicality as well.)

Its also quite possible that any or all of these may be operating.  In fact, I strongly believe that all three are, in fact, operating.  MTF transkids do transition and obtain HRT at an earlier age.  They (we) do care and want to pass to better our lives.  And, from research into gender atypical children, it has been noted that gender atypical and dysphoric male children are considered more attractive than their gender typical male peers.  This ‘attractiveness’ is caused by hypomasculinity (masculine faces aren’t “pretty”).

Looking at the data for FtM transmen, there seems to be a small signal.  I would really like to see a study with more subjects, as this didn’t seem to be as statistically significant as we would like.  Even if real, the effect size is small.  However, we do have collaborating data from earlier studies that show that FtM transkids are judged to be more physically masculine than non-trans-girls.

(Addendum 9/10/2017:

There have been hints from a number of studies that there is a correlation between sexual orientation and subtle gender atypical facial physiognomy.  A new study just how strongly supports this observation.  Using a deep-layer neural net AI trained to categorize faces as heterosexual or homosexual, can differentiate between two faces, one of a heterosexual and one of a homosexual, of the same natal sex at 91% accuracy for males and 83% for females.  That is, the researchers in essence found that there is a sexual orientation dimorphism with an effect size of d=1.9 for males and d=1.4 for females.  This is an amazingly high effect size for both populations.  In fact, this is higher than the effects sizes found for passability.  Of course, I believe this is because many of the subjects have been incorrectly categorized, as I mentioned above.  But it could also be due to this study being restricted to the face and the passability study looking at one’s entire physique.  Note that the effect sizes for the sexes both agree in proportion, males being larger than females.  This adds more strength to the hypothesis that at least some of the cause of the difference in passability between the two types is native gender atypicality of the “homosexual” transsexual taxons.)

(Addendum 2/25/2021:

We have yet more data regarding the effect of perinatal testosterone exposure and adult sexually dimorphic facial features in a paper I only recently chanced upon.  This time DIRECT measurement of infant blood levels of testosterone and the degree of masculinization of the face in both sexes at age 20, right at the end of adolescence.  The correlation is r=0.55 for males and r=0.48 for females.  The interesting thing about the study is that it found NO correlation with adult testosterone levels.  This indicates that the level of masculinization of the face is an early organizational effect of testosterone, not a simple activational effect of testosterone at puberty.  This explains why even HSTS who were not able to be on puberty blockers or HRT early still have notably feminine features.  This agrees with other studies that suggest that perinatal testosterone levels correlate with both later gendered behavior and sexual orientation.  Thus, for “homosexual transsexuals” we now have a fairly strong set of converging data that supports an early organizational effect of testosterone (low in MTF, high in FtM), while at the same time, this strongly supports the hypothesis that “non-homosexual transsexuals” do not. )

Further Reading:

Differences between androphilic vs. non-androphilic transwomen, passing, and transition decision making.

References:

Tim C. van de Grift, Peggy T. Cohen-Kettenis, Thomas D. Steensma, Griet De Cuypere, Hertha Richter-Appelt, Ira R. H. Haraldsen, Rieky E. G. Dikmans, Susanne C. Cerwenka, , Baudewijntje P. C. Kreukels, “Body Satisfaction and Physical Appearance in Gender Dysphoria” Archives of Sexual Behavior
DOI: 10.1007/s10508-015-0614-1

Zucker KJ\, Wild J, Bradley SJ, Lowry CB., “Physical attractiveness of boys with gender identity disorder.” Archives of Sexual Behavior. 1993 Feb;22(1):23-36.
http://link.springer.com/article/10.1007/BF01552910

Stephanie A. Mcdermid, Kenneth J. Zucker, Susan J. Bradley, Dianne M. Maing, “Effects of Physical Appearance on Masculine Trait Ratings of Boys and Girls with Gender Identity Disorder” Archives of Sexual Behavior
http://link.springer.com/article/10.1023/A%3A1018650401386

Sari R. Fridell, Kenneth J. Zucker, Susan J. Bradley, Dianne M. Maing, “Physical attractiveness of girls with gender identity disorder” Archives of Sexual Behavior
http://link.springer.com/article/10.1007/BF02437905

Michal Kosinski, Yilun Wang, “Deep neural networks are more accurate than humans at detecting sexual orientation from facial images.”
https://osf.io/zn79k/

Whitehouse, et al., “Prenatal testosterone exposure is related to sexually dimorphic facial morphology in adults” Preceedings of the Royal Society https://doi.org/10.1098/rspb.2015.1351

Advertisement