On the Science of Changing Sex

The Dutch Connection

Posted in Confirming Two Type Taxonomy by Kay Brown on December 24, 2010

androgynous faceAge of Onset vs. Sexuality

While the TS/TG world rails against even being included in the DSM, a far more interesting scientific debate regarding transsexual typology is being played out in the DSM-V committee and associated scientific circles.  On the one hand is the North American centered contingent who favors continued use of sexual orientation as specifiers/subtypes, and the Netherlands contingent who favors the use of age of onset of “gender incongruent behavior or feelings”.  The Netherlands group summarizes their argument thus:

“Considering the strong resistance against sexuality related specifiers, and the relative difficulty accessing sexual orientation in individuals pursuing hormonal and surgical interventions to change physical sex characteristics, it should be investigated whether other potentially relevant specifiers (e.g., onset age) are more appropriate.”

While the North American position is best summarized by Lawrence:

“Typologies based on sexual orientation, however, employ subtypes that are less ambiguous and better suited to objective confirmation and that offer more concise, comprehensive clinical description.  Typologies based on sexual orientation are also superior in their ability to predict treatment-related outcomes and comorbid psychopathology and to facilitate research. Commonly expressed objections to typologies based on sexual orientation are unpersuasive when examined closely”

The question for me is… what is really motivating the Netherlands position?  Do they believe that sexuality has nothing to do with typology?  Have they themselves ( in the Smith paper) not shown that Blanchard’s typology of Homosexual vs. Non-Homosexual (AGP) transsexuality is valid?  This seeming contradiction begs exploration. The story actually starts decades ago with the first attempts at creating meaningful typologies.  I won’t go into historical details, as that has been done by better historians than I, most especially Lawrence.  It is only important that we focus on the thread of ideas that led to the “Early vs. Late Onset” typology. In the ’60s and ’70s, a number of sexologists proposed typologies based upon a division between “true” transsexuals and “pseudo” transsexuals, or a softened version using the terms, “primary” vs. “secondary”.  The problem was, that different researchers had different ideas as to just what constellation of behaviors indicated which was primary and which secondary.  But one idea did seem to be held in common, the notion that those gender dysphoric individuals who had always been such, and knew it, were different than those who only slowly became gender dysphoric during late adolescence or adulthood.  This would seem to be common-sensical. This then, is the basis of the school of thought that currently predominates in the Netherlands. However, they have replaced the loaded language, softening it still further by replacing “primary” with “early onset” and “secondary” with “late onset”.  It is important to understand that this is in regards to self-reported private awareness, not at what age transition occurs. But… Is it etiologically meaningful?  Or even clinically meaningful?  Is there even a meaningful difference between age of onset and sexuality? Consider the 1994 Netherlands, Doorn study. Our first question, is Early Onset Transsexual (EOT) and Late Onset Transsexual (LOT), merely new names for the same thing as HSTS vs. Non-HSTS?  After all, we know from many studies that most HSTS transition before age 25, with 40% having transitioned as teens:

Type:                                                            EOT    LOT     TV
N=                                                         103       52      31
Imagined partner in adolescence:
Heterosexual male                               63.0   37.3     3.0
Bisexual/changing                                18.0   19.6   15.2
Heterosexual woman                             9.0   17.6   60.6
No image                                                   10.0   25.5   21.2
Non-androphilic                                    37.0   62.7   97.0
Cross-dressing arousing                     26.0   42.2   84.8

Interestingly, as the authors noted, Person and Oversey’s definition of “primary” transsexual of always having had a livelong conviction of being the opposite sex as “asexual” is clearly wrong, and Stoller’s opinion correct; Most transsexuals showing a lifelong conviction of being the opposite sex are homosexual and not asexual.  Thus, we see a high correlation with exclusive androphilia and  being classified as an “early onset” transsexual.  While there is an equally high correlation with being non-exclusively-androphilic (non-homosexual) and being a “late onset” transsexual. So, it would seem that on first blush, that Early Onset Transsexual (EOT) vs. Late Onset Transsexual (LOT) and HSTS vs. Non-HSTS would have very high statistical correlations, but not perfectly.  So, the two competing typologies aren’t that different in practice, only in theory.

A theoretical point for the use of Early Onset is that those who experience such, cannot be originally motivated by autogynephilia, and thus must have had a feminine gender identity that later developed into fetishistic use of cross-dressing in some.  The authors make an assumption that autogynephilic arousal cannot exist before puberty, “Cross-dressing, however, in many transvestites starts before age 10, indicating that its original function is not fetishistic.” But, as has been shown, this simply is not true.  Such arousal does occur, and further, obvious penile arousal is not necessary for autogynephilic desire to be present and rewarding.  Thus, the age cut-off for early vs. late onset was based on a false premise.

If Blanchard’s original hypothesis that exclusively androphilic gender dysphoric MTF transsexuals are non-autogynephilic and non-exclusively-androphilic individuals are, then we would expect that there would be a high correlation with the number of individuals who reported autogynephilic arousal to cross-dressing and the number who were not exclusively androphilic.  A casual examination certainly shows that to be the case and a mathematical examination shows that the three point correlation is 0.985403; so close to a perfect 1.0 as to essentially be so, given the rounding errors.  Thus, Blanchard’s hypothesis is confirmed.

This study also suggests something very interesting: the ratio of transvestites, who are predominately exclusively gynephilic, reporting that cross dressing is, or was at one time, sexually arousing is 85%.  This number is very similar to those numbers found in both Blanchard and Nuttbrock for exclusively gynephilic transsexuals.  This is again powerful evidence that we are looking at the same “essential” taxa in these groups, that is to say, that both of them are autogynephilic.  The TS groups, which include higher percentages of bisexual and asexual transsexuals, show lower reported arousal to cross-dressing (but not necessarily lower autogynephilic arousal, as this is not the only form of autogynephilic ideation) show similar reduced percentages as was found in both Blanchard and Nuttbrock.  Thus, this Doorn study shows essentially the same results, and may be added to the list of studies that confirm Blanchard’s hypothesis in nearly all respects. Although Doorn et al. have made a valiant effort to shoehorn the data to fit their hypothesis that early vs. late onset of gender dysphoria is an essential taxanomic distinction, the data actually supports just the opposite conclusion.  Sexual orientation, specifically exclusive androphilia, vs. non-exclusive-androphilia (and by extension, “homosexual” vs. “autogynephilic”, a la Blanchard) are far more salient signifiers, and likely matching etiologies.

Addedum 12/27/2010: I was curious to see if we turn the data around by 90° what would the percentage of each sexuality be:

Sexuality:                      Androphilic      Bisexual      Asexual   Gynephilic
N=                                              85                      29                  23                 18
Early Onset                            77%                  65%              43%             50%
Late Onset                              23%                   35%             57%             50%

So we see from this direction, that again, that report to have been exclusively  or even somewhat androphilic transsexuals are more likely to have been classified as “EOT” than the other sexualities.  That is to say, that androphilia correlates with early onset.

Addendum 12/1/2013:  The question of whether the two typologies are synonymous has been answered by a joint paper in which researchers in the Netherlands and North America use the two interchangeably.

Addendum 6/29/2014:  One of my faithful readers pointed out an interesting paper published in part by the Netherlands (Dr. Cohen-Kettenis) team that touched upon the above issue.  (Thanks.) Looking at just the data on the sexual orientation vs. “EOT” and “LOT” vs. sexual orientation of the MTF research population we see a similar pattern to Doorn’s data.  Interestingly, in this study, they understood that their subjects might not be reporting their sexual orientation accurately, so they include data on their clinician’s assessment of these subjects sexual orientation from their observations over the months and years that the subjects were in transition.  No surprise, these MTF transwomen had misrepresented their sexual orientation in the socially desirable direction for MTF transwomen, toward greater exclusive androphilia.

Type:                                                            EOT               LOT
N=                                                                35                    44
Women (self)                                            15 (43%)         8 (18%)
Women (clinician)                                   14 (41%)        17 (39%)
Men (self)                                                   13 (37%)        23 (52%)
Men (clinician)                                         14 (41%)          4 (9%)
Bisexual (self)                                             2 (6%)          10 (23%)
Bisexual (clinician)                                   5 (15%)         22 (50%)
Asexual (clinician)                                     1 (3%)             4 (9%)

Using the clinician assessment, only 18 (23%) out of the 79 transwomen were exclusively androphilic, with 14 (78%) of them having been categorized as “early onset”, identical to the Doorn study.  Conversely, using clinician assessment, only 21 (34%) out of 61 non-exclusively androphilic transwomen were categorized as “early onset”.  If anything, this shows an even stronger signal than the Doorn study, that non-exclusively androphilic transwoman are likely to be categorized as “late onset”.

It’s interesting to note that while this study acknowledged the well-recognized fact that “late onset” transsexuals tend to misrepresent their sexual orientations, they failed to recognize that self-report of early childhood gender atypicality is also suspect.  Further, accurate self-report of early cross-dressing and gender dysphoria does not equate to early gender atypicality, as it can and often does result from childhood expression of autogynephilia.  Lawrence found in her 2005 survey that 76% of self-reported non-exclusively androphilic transwomen will describe themselves as having been feminine (gender atypical) as children, but only 25% self-reported that others would have thought so.  In future studies, I would like to see reliable data of parents or other caregivers reports on gendered behavior in addition to self report.  I strongly suspect that we will see an even stronger signal with respect to clinician assessed sexual orientation (especially before transition) and parental reports of childhood gender atypicality.

Addendum 8/17/2014

While reviewing data for another post, I realized that I had in an older post, data that supported my contention that exclusive androphilic MTF transkids reported being “early onset” at the same rates as the above referenced papers.  In the Tsoi study of transwomen in Singapore, which has been shown to have zero percentage of autogynephilic transsexuals, the following data was gathered:

Singapore HSTS Life Arcs
(Cumulative Percentages at Age Indicated)

Age:                                                             6    12   18     24 Total %
Start to feel like a girl                       27    74  99  100 100

Note the number 74% of this population that reported having recognized that they “feel like girls” by age 12.  Puberty occurs around age 12-14 in males… so this study shows that 74% exclusively androphilic MTF transsexuals report being “early onset” in this study.  But note that ALL of them feel that way by age 18.

So, three studies, one from the Netherlands, one from greater Europe, and one from Asia, showing 77%, 78%, and now 74% of exclusively androphilic transsexuals would be classified as “early onset”.  This is yet more data that shows that the concept of “early” vs. “late” is nearly synonomous with Blanchard’s “Homosexual” vs. “Autogynephilic” and that sexuality, exclusive androphilia vs autogynephilia is the actual taxonomic distinction.  Had the null hypothisis, that “early onset” did NOT correlate with exclusive androphilia… we would have seen widely diverging data.  Further, given that this last study shows that looking at a country’s transsexual population and testing their sexuality by ‘early’ vs. ‘late’ would have shows widely diverging data, as the Singapore study as zero “early onset” gynephiles… of course, they also have zero “late onset” gynephiles as well, so that’s not totally fair… but still, had “early” vs. “late” been a valid taxonomy, one would not have had this Singaporean result… as it is sexuality, not age of onset that varies from country to country in correlation with their Individuality vs. Collectivity.

Addendum 12/1/2015:

I found a Korean study that also lists the ages of becoming gender dysphoric.  Of 43 MTF transsexuals, all but one was exclusively androphilic and that one bisexual:

Gender dysphoric by age:
Preschool    Middleschool    Adult (18)
30%                83%                      100%

The median age was 10.9 +/- 4.1 for gender dysphoria.  The mean age for social transition was 19.7 +/- 1.2 years old, typical for transkids from other places in the world, including the United States.

Addendum 1/22/2019:

Yet more data published in 2019; this time from Germany where they too prefer the “early” vs. “late” distinction.  However, their cut line was at the age of majority of 18 years old.  Note that the very same pattern, nearly identical percentages were found.

N=40     Androphilic     Non-Androphilic
Early      10 (77%)             9 (33%)
Late         3 (23%)            18 (67%)

Addendum 1/24/2019:

In reviewing my blog entries today, I realized I had yet another study from 2014 that I had overlooked earlier, the Cerwenka paper:

N=201    Androphilic     Non-Androphilic
Early      54 (73%)             53 (40%)
Late        20 (27%)             78 (60%)

Oh look… it shows the same trend.  Imagine that.



Peggy T. Cohen-Kettenis and Friedemann Pfäfflin, “The DSM Diagnostic Criteria for Gender Identity Disorder in Adolescents and Adults” http://www.springerlink.com/content/c54551hj463111j1/

Anne A. Lawrence, “Sexual Orientation versus Age of Onset as Bases for Typologies (Subtypes) for Gender Identity Disorder in Adolescents and Adults” http://www.springerlink.com/content/7135712p31525871/

C. D. Doorn, J. Poortinga and A. M. Verschoor, “Cross-gender identity in transvestites and male transsexuals” http://www.springerlink.com/content/u63p723776v57m11/

Timo O. Nieder, Melanie Herff, Susanne Cerwenka, Wilhelm F. Preuss, Peggy T. Cohen-Kettenis, Griet DeCuypere, Ira R. Hebold Haraldsen, and Hertha Richter-Appelt, “Age of Onset vs. Sexual Orientation in Male and Female Transsexuals” (2011)

TaeSuk Kim, et al., “Psychological Burdens are Associated with Young Male Transsexuals” (2006) DOI: 10.1111/j.1440-1819.2006.01525.x

Zavlin, D. et al., “Age-Related Differences for Male-to-Female Transgender Patients Undergoing Gender-Affirming Surgery”, Journal of Pediatric Surgery (2019)

Cerwenka, S. et al., “Sexual Behavior of Gender Dysphoric Individuals Before Gender-Confirming Interventions: A European Multicenter Study” (2014)

Comments Off on The Dutch Connection

And Things That Go Bump In The Night…

Posted in Brain Sex by Kay Brown on December 10, 2010

Phantom Pains… or just a pain in the neck?

shrinking brainAnyone who has had surgery knows that pain is common companion for a while afterward.  But, when a significant part of the body is removed, we no longer receive any sort of message from that part… or do we?

In the brain there is a map, a neural representation of the body.  In fact, there are several such maps.  Each map may have different function than the others.  For example, one map is for motor (muscle) control, where-in each muscle is related to a part of the body it effects.  Another map is for the sense of body part location, and is usually closely associated with the motor control map; after all, if one is going to control the muscles, one needs to know what needs to be moved to where.  Yet another map is about other sensations, including pain.  There is even a map of the body which is largely informed by visual information, that is to say, that is kept updated by literally seeing where those parts are in relation to each other.

These maps, if not fed real information, sometimes just make stuff up!  When that happens, an individual may “feel” the presence of missing limb or organ.

One of the leading researchers in the field of phantom sensation is renown neurologist and psychophysicist,  Vilayanur Subramanian Ramachandran, M.D., Ph.D.  He is a professor in the Psychology Department at the University of California, San Diego.  This man is no hack.  In fact, some of my own work in my chosen professional field is influenced by his work in the visual sciences.

So, it was a surprise that he should couple his work in phantom limb sensation with trying to understand the nature of transsexuality. Dr. Ramachandran hypothesised that transsexuals have a sex atypical somatosensory map of their genitalia.  That is to say, that at least some, if not all of the pre-wired neural maps in the neocortex are sexually dimorphic in the region that maps the genitalia.  This is not an unreasonable hypothesis on first examination.  After all, male and female human beings do have differently shaped genitals, are sexually dimorphic, so why wouldn’t the somatosensory maps be sexually dimorphic to match?

However, there are some theoretical objections to this idea.  First and foremost of which is that the genitalia are not really all that different in quality… and only superficially different in quantity.  Nearly each feature of the external and even of some of the internal structures are homologous.  That is, for each feature found in a male, there is a feature that matches it in the female, which is only different in degree, not in kind.  The most obvious example is the glans of the penis is homologous with the glans of the clitoris.  Inside of the penis, and down into its root inside of the body, is spongy tissue that expands when blood pressure fills it with blood.  Inside of the clitoris and down into its root inside of the body, is spongy tissue that expands when blood pressure fills it with blood.  Quite literally, a penis is a very large clitoris; And a clitoris is a very small penis.  Oh there are differences in how the urethra is routed, but even there, they start in the same place.  In men there are two glands called the Cowper’s, which produce a clear fluid that aids in lubrication during sex.  In women there are two glands called the Bartholin’s which produce a clear fluid that aids in lubrication during sex.  Why are they called two different names?  Finally, the scrotal sac is the same tissue as the labia majora, but have fused together.  Thus, the two sexes, which seem so different to a naive observer, are really very nearly the same to a student of anatomy.  So, given that the two are really very nearly the same, shouldn’t the neural maps be the same?

We might then object and say, yes, but they are differently sized and configured.  But then, one might argue back, that one man’s body isn’t identical to another’s, surely the neural wiring must be flexible enough to learn of these differences and make accommodations?

Finally, it is just these accommodations and plasticity of the neocortex that Ramachandran calls into play in his other work on the genesis of phantom sensation, as the no longer required regions of the original map are recruited and repurposed to processing sensations from nearby functions in the brain.  (On a side note, this does not always mean nearby in the sense of external somatic topology… for example, the face and hands are “nearby” in the somatosensory maps, but quite distant externally.)

So, like good scientists, we conduct experiments… Ramachandran reasoned that if there was a significant difference between the putatively sexually dimorphic somatosensory maps of the genitalia between control men and MTF transsexuals, then there may be a difference in their responses to penectomy.

Lacking the ability to recruit men at random to remove their external genitalia, we find groups that have need to do so for other reasons.  There are many MTF transsexuals who willingly have their penises removed during Sex Reassignment Surgery (SRS).  Dr. Ramachandran reasoned that men whose penises were partially removed because of cancer, but otherwise randomly selected by Dame Fortune, would serve as the controls.

On theoretical grounds, this too has some problems.  First, modern SRS is not a simple amputation.  In fact, a great deal of effort is taken to maintain as much of the epidermis and their enervation, intact, inverted and invaginated into the neovagina.  Only a portion of the external corpus cavernosa, the spongy erectile tissue, is excised.  The scrotal tissue is divided and repurposed as the labia majora, essentially reversing the prenatal developmental process.  Thus, some of the tissue is in the same place as before, and is in the place expected for a female.  And the penile skin, though now inverted, is topologically still in much the same place.  This is not the best one-to-one match to the controls.  (Ironically, Dr. Ramachandran would have found a better match back in his country of birth, as the traditional surgery for Hijra women, which are essentially the same as the transsexuals of the West, did in fact involve a simple amputation of the penis.)

Thus, there are many reasons to doubt whether there should be, a priori, a sexually dimorphic difference in the somatosensory maps of males and females.  There is also the little matter of the fact that there are two kinds of transsexuals, with different etiologies, which may confound the experimental data.  And finally, the dramatically different results of penile amputation and modern SRS would lead one to suspect that any differences in response, would be attributable to the differences in surgical result.  But… hypothesis are meant to be tested…

So, using data from a 1951 paper on phantom sensations of amputated penises in 12 men and his own research into phantom sensations in 20 MTF transsexuals, he found that six of the 20 MTF transsexuals had phantom sensations while seven of the 12 controls did.  Wow, big difference!!!

Wait… not so fast… actually… well…  that’s actually a small difference in a small sample size.  Let’s imagine that we have two coins, and we have no idea about the nature of statistics and chance.  I flip one coin twenty times and it comes up head six times out of twenty, while you flip the other coin twelve times and it comes up heads seven times.  So, do we say that my coin is more likely to come up tails the next time, while yours is more likely to come up heads?  Ummm…. no.

Random chance alone could give us the same numbers flipping coins as Ramachandran got in his statistics.  In fact, Anne Lawrence (gotta love her, she keeps turning up just like that proverbial penny) showed that properly analyzed using not one, but three different and widely accepted statistical tests showed that this small of a sample with this small of a difference in data has no statistical significance.  It might be random noise… it might be real… we simply don’t know.

We simply don’t know.

How could we know?  First, remove the theoretical problem of comparing apples and oranges (partial penile amputation compared to modern SRS) by collecting data on Hijra’s in India and Pakistan, and carefully note their sexual history and sexual orientation.  Second, get a lot more samples of both control and MTF populations.  Until then; we simply don’t know.

We simply don’t know.

(Addendum 10/7/2014:  Dr. Ramachandran likely chose to examine this topic because he is primarily interested in phantom sensations, but it occurs to me that to explore the concept of sexually dimorphic brain maps, we could, and perhaps should, use pre-operative transsexuals.  One could attempt to stimulate various points of genital anatomy and determine if their is any psychophysical response differences between transsexuals and controls.  My prediction is that there will be absolutely none, in neither transkids nor autogynephilic transwomen.  I also predict that we would also find no difference between FtMs and control females.)

For more essays on trans-brains see Brain Sex.


Lawrence, A. A. (2010). “Transsexuals and nontranssexuals do not differ in prevalence of post-penectomy phantoms: Comment on Ramachandran and McGeoch” (2008), Journal of Consciousness Studies, 17(1/2), 195-196.

Ramachandran, V.S. & McGeoch, P.D. (2007), “Occurrence of phantom genitalia
after gender reassignment surgery”, Medical Hypotheses, 69, pp. 1001–3.

Ramachandran, V.S. & McGeoch, P.D. (2008), “Phantom penises in transsexuals:
Evidence of an innate gender-specific body image in the brain”, Journal of Consciousness Studies, 15(1), pp. 5–16.


Tagged with: ,

Comments Off on And Things That Go Bump In The Night…