In:: Journal of the American Academy of Child and Adolescent Psychiatry, 2001, 40, 472-481
Follow-up of Adolescents Transsexuals
Correspondence to: Yolanda L.S. Smith, M.Sc.
Mailing address of all authors:
Department of Child and Adolescent Psychiatry
University Medical Center Utrecht
P.O. Box 85500
3508 GA Utrecht, The Netherlands
Tel. no.: (+31) - 30 - 2508401
Fax. no.: (+31) - 30 - 2505444
Statistical expert: Dr. S.H.M. van Goozen
Word count: 5900
Adolescents with Gender Identity Disorder who were Accepted or Rejected for
Sex Reassignment Surgery: A Prospective Follow-up Study
Yolanda L.S. Smith, M.Sc.,
Stephanie H.M. van Goozen, Ph.D.,
Peggy T. Cohen- Kettenis, Ph.D.
Department of Child and Adolescent Psychiatry, University Medical Center
Utrecht, and Rudolph Magnus Institute for Neurosciences.
This work was financially supported by the Stichting Nederlands Gender
Centrum and the Stichting Fondsenwervingsacties.
The authors wish to thank Jos Megens, Mrs. Harmsen, and Dr. Louis Gooren for
their help in the data collection process.
Correspondence to: Yolanda L.S. Smith, M.Sc.
Objective : To conduct a prospective follow- up study with 20 treated
adolescent transsexuals to evaluate early sex reassignment, and with 21 non-
treated and 6 delayed- treated adolescents to evaluate the decisions not to
allow them to start sex reassignm en t at all or at an early age. Method :
Subjects were tested on their psychological, social and sexual functioning.
Follow-up interviews were conducted from 3/95 till 7/99. Treated patients had
undergone surgery 1-4 years before follow- up; non- treated patients were
tested 1-7 years after application. Within the treated and the non- treated
group pre- and post- treatment data were compared. Results between the
groups were also compared. Results : Postoperatively the treated group was
no longer gender dysphoric and psychologically and socially functioning quite
well. Nobody expressed regrets concerning the decision to undergo sex
reassignm ent. Without sex reassignment, the non- treated group also showed
some improvem ent, but they also showed a more dysfunctional psychological
profile. Conclusions : Careful diagnosis and strict criteria are necessary and
sufficient to justify hormone treatm ent in adoles cent transsexuals. Despite the
fact that some of the non- treated patients may actually have Gender Identity
Disorder, the high levels of psychopathology found in this group justify the
decision to not start hormone treatm ent too soon or too easily.
Key words: gender identity disorder, transsexualism, sex reassignment
Despite the early onset of gender identity disorder in many transsexuals, and
in spite of the fact that transsexuals increasingly apply for sex reassignment
surgery (hereafter, SRS) at younger ages, it is common practice not to start
the actual sex reassignme nt (SR) procedure before the age of 18 or 21. One of
the main objections of professionals against a start of the SR procedure before
18 years is the risk of postoperative regrets. One fears that, as a consequence
of the developmental phase itself, adolescents will not be capable of making a
sensible decision about something as drastic as SR. Moreover, medical
interventions in adolescence are expected to lead to unfavorable results
concerning the physical, the psychological, and the social functioning of the
adolescent. Despite these hesitations, transsexual adolescents have been
diagnosed and referred for hormone treatm ent and SRS at the gender clinic of
the Department of Child and Adolescent Psychiatry of University Medical
Center Utrecht (UMCU). In those carefully selected patients that are referred
for hormone treatment, the often- assume d association between
transsexualism and psychopathology has not been found (Cohen et al., 1997;
Cohen- Kettenis and Van Goozen, 1997). This significantly contributed to the
decision to start hormone therapy between the ages of 16 and 18 in two
phases: first, hormones with reversible effects (for male- to- females or MFs
antiandrogens to block further masculinization of the body, for female- to-
males or FMs progestins to suppress menstruation; second, oestrogens to
feminize the MFs and androgens to masculinize the FMs. Though,
postoperative regret or any other unfavorable result is a matter of serious
concern for our clinicians, it is also considered important to avoid life-long
suffering due to postponement of treatment. With early SR two major
negative consequences of late treatm ent may be prevented: (1) irreversible
physical changes (especially a low voice and beard growth in MFs), that may
create life-long traces of the biological sex, and (2) delay or arrest in areas
that are particularly important during adolescence (e.g. peer relationships,
romantic involvements or academic achievemen t) which may, in itself, lead to
additional, yet avoidable problems. Thus, early treatment may be particularly
suitable to prevent unnecessary psychological and emotional problems.
Furthermore, on the basis of numerous follow-up studies it can be
concluded that, in adults, unfavorable postoperative outcome is related to a
late start of the SR procedure rather than an early one (for a review, see
Cohen- Kettenis and Gooren, 1999). Age at assessment also emerged as a
factor differentiating two small groups of adult MF transsexuals with and
without postoperative regrets (Lindemalm et al., 19 87).
Naturally, if a resolution to extreme and life-long cross- gender identity
problems was attainable with less invasive treatment methods, clinicians
should refrain from SRS, in adolescents as well as in older patients. However,
as more extensively discussed by Cohen- Kettenis and Van Goozen (1997), the
literature does not provide convincing evidence that psychotherapy can in fact
alter a fixed cross- gender identity (Cohen- Kettenis and Kuiper, 1984).
Cohen- Kettenis and Van Goozen (1997) conducted a retrospective study
on postoperative functioning of the first 22 consecutive adolescent
transsexual patients who had attended our gender clinic and who had
undergone SRS. They concluded that starting the SR procedure before
adulthood results in positive postoperative functioning, provided that careful
diagnosis takes place in a specialized gender team and that the criteria for
starting the procedure early are strict (see Cohen- Kettenis and Van Goozen
(1997) for details concerning the clinical approach).
To confirm these first results, a prospective follow- up study was
conducted with the next 20 consecutive adolescents who had undergone
SRS. This time, we also investigated what had become of the adolescents
whose application for SR had been rejected or who had refrained from SR in
the first diagnostic phase. Since we expected applicants to be
heterogen eous with respect to gender dysphoria, comorbidity, and perhaps
other factors, this study should be seen as an evaluation of two related but
separate clinical decisions. The first and most important one was whether it
had been a correct decision to allow well-functioning adolesc ent
transsexuals to proceed with the SR procedure after careful screening, given
that they were between 16 and 18. The second one was to find out whether
the decision not to allow other adolescent applicants to proceed with the SR
procedure before 18 had been a justified one.
This study focuses primarily on postoperative gender dysphoria, feelings
of regret, gender role behavior, and an evaluation of the treatment but other
areas, such as psychological, social, and sexual functioning were also
addressed. The same domains were investigated in the treated and non-
The 20 patients (13 FMs, and 7 MFs), who consecutively underwent SRS after
the 22 patients of the first study, were invited to participate at least one year
after their last surgical treatment. All treated (hereafter T-) patients
responded positively and were included. While applying for SR in their first
diagnostic phase, 21 other applicants withdrew their request for SR, were
rejected, or did not show up at appointments. The primary reason for rejection
or withdrawal was that they were not diagnosed transsexuals despite the fact
that some did have gender identity problems. In many of these cases the
psychological or environmental problems were too serious to make an
accurate diagnosis. In other cases, when patients did not show up for
appointments, a diagnosis had not yet been made before we lost contact.
Hence, the rejected/withdrawn cases did not start hormone treatment and the
real-life experience phase (this group we call hereafter the NT-group). Another
six eventually started hormone treatment. Two started after an initial delay
when still in contact with our gender clinic; four were expected to belong to
the NT-group when contacted but appeared to have reapplied at another
gender clinic where adults are treated.
Four individuals of the NT-group were not traceable and could therefore
not be included in the analyses. Some of their data, gathered at the time of
their application, were used descriptively. Three others could not be
interviewed, but some follow- up data were collected through their parents.
Finally, our clinic was contacted by psychiatric institutions requesting
information about two of our (non- traceable) patients, which provided us with
information about their current status. This resulted in a T-group of 7 MFs and
13 FMs, a NT-group of 13 males and 8 females, and a delayed treated (DT-)
group of 4 MFs and 2 FMs. Statistical analyses of the NT-group were based on
9 males and 5 females, who were seen both at the time of application and at
follow-up. All applicants were between 13 and 18 years old at the time of first
assessme nt. Data concerning the DT-group were not included in the statistical
analyses. (See Table 1 for additional information about the DT- and NT-
INSERT TABLE 1 ABOUT HERE
IQ tests. The most recently adapted Dutch versions of the Wechsler scales
were used, which are the WISC-R (van Haasen et al., 1986), and the WAIS
(Stinissen et al., 1970).
Utrecht Gender Dysphoria Scale (UGS). Gender dysphoria was measured with
the UGS, consisting of 12 items on which the subject rated his/her agreem ent
on a 5-point- scale. The higher the score, the more gender dysphoria was
indicated (for psychometric data: see Cohen- Kettenis and Van Goozen, 1997).
Body Dissatisfaction/Physical Appearance.
Body Image Scale (BIS). A body image scale (Lind gren and Pauly, 1975) was
used which had been adapted for a Dutch sample (Kuiper, 1991). The higher
the score, the more dissatisfaction was indicated.
Appraisal of Appearance Inventory (AAI). On this 14- item- inventory 3
independent observers (the diagnostician, a nurse of the gender team, and
the researcher) rated their subjective appraisal of the appearance of the
subject on a 5-point- scale of femininity/masculinity with higher scores
indicating more incompatibility of the appearance with the new gender.
Dutch Short MMPI (NVM: Luteyn et al., 1980).
The Dutch version of the Symptom CheckList (SCL-90: Derogatis et al., 1973;
Dutch version: Arrindell and Ettema, 1986).
Treatment Evaluation and Posttreatment Functioning.
Treatment satisfaction. T-patients completed a semi- structured interview
about treat ment outcome, experiences during and after SR, treatment
evaluation, and feelings of regret. NT-patients were asked questions regarding
their way of living and level of functioning after refraining from or delaying SR.
When applicable, questions from the T-patients' interview were adapted.
Social and sexual functioning. In a semi- structured interview questions were
asked about the subjects' current life situation (Doorn et al., 1996). Specific
questions about sexual contact after surgery were omitted for the NT-group.
Results were analyzed per item.
Satisfaction with surgery. T-patients completed a self-developed questionnaire
concerning functionality of the vagina or penis and satisfaction with surgical
results (Cohen- Kettenis and Van Goozen, 1997). Results were analyzed per
Public confrontation questionnaire. A questionnaire assessed reactions of the
social environment and was used to evaluate the subject's experiences of
being able to pass in the new social role (Doorn et al., 1996). Results were
analyzed per item.
Quality of life. The Affect Balance Scale (Bradburn, 1969) was used in the T-
and NT-groups to measure overall psychological well-being. The scale consists
of five positive and five negative items. Only the negative affect scores were
analyzed because in a randomly selected sample Cronbach's alpha for the
positive affect scale was found to be too low (positive affect scale: .59;
negative affect scale: .73; Kempen and Ormel, 1992). The adapted Dutch
version of the scale by the Central Bureau of Statistics was used (1987).
Various questions were analyzed per item because they did not form a scale.
Whenever this was the case we have indicated in the Results section how
many response categories they contained.
IQ was assessed before treatm ent. The UGS, the BIS, the AAI, and the
Personality Questionnaires were administered before and after treatment
because within- subject changes were expected. The remaining instruments
concerned the postoperative situation and were only administered after
treatment. Naturally, instruments pertaining to post- treatment functioning
were not administered to the NT-group.
T-patients came to UMCU or combined a hormone check- up at Free
University Hospital in Amsterdam (FUHA) with the interview and testing. Each
session took two to three hours. In order to avoid socially desirable responses
the subjects were seen by the first author, who had not been involved in their
diagnosis or treatmen t. The Ethics’ Committees of UMCU and FUHA approved
Statistical analys es
Changes over time within the T- and the NT-group were analyzed with
univariate paired t-tests. Differences between groups were analyzed with
univariate independent t-tests. Subjects with a score more than 2.5 standard
deviation above the mean score of their subgroup were considered to be
outliers. Their values were replaced by the mean value of their subgroup.
The mean age of the T-group was 16.6 years (range 15 to 19) at pretest and
21.0 (range 19 to 23) at follow- up. Ten T-patients had started hormone
treatment between 16 and 18 years of age. The average time elapsed
between the last operation and the time of the follow-up interview was 1.3
years (range 1 to 4). The group's pretreatme nt mean IQ score was 107 (SD =
16; range 85 to 140). At follow-up 9 subjects (48%) were studying (at a school
for business administration or university); 5 had a job and 5 were
unemployed. Ten subjects lived independently or in student dormitories; 1
lived with her partner and 8 were living with (one of) their parents.
The mean age of the NT-group was 17.3 years (range 13.7 to 20.2) at
pretest and 21.6 (range 15.7 to 26.2) at follow-up. The average time elapsed
between SR application and the time of the follow-up interview was 4.2 years
(range 1 to 7). The group's mean IQ score was 104 (SD = 15; range 89 to 130)
at pretest. At follow- up 6 were studying, 5 had a job, and 4 were unemployed.
Six subjects of the NT-group lived independently, 1 with a partner, and 7 lived
with (one of) their parents. Another subject lived in a psychiatric institution.
The T-group reported less gender dysphoria (p < .001) at follow- up than at
pretest. The mean pretest and posttest scores of this group were completely
in the range of those from the previous study (pretest mean = 51.7; SD = 6.3)
(posttest mean = 14.8; SD = 3.2). None of the subjects expressed feelings of
regret about their SR (3 point- scale).
The NT-group was also less gender dysphoric (p = .002) at follow-up
than at the time of application. However, the decrease in gender dysphoria
was much greater within the T- than within the NT-group (p = .002) .
Eleven subjects of the NT-group did not feel any regrets about having
refrained from SR or being rejected (3 response categories). One woman
slightly regretted having refrained from SR because she still had doubts about
her gender identity. Two men also slightly regretted the decision not to start
treatment but in both, the wish for SR was not clearly differentiated from
unrealistic expectations that SR would resolve important non- gender
problems. Finally, one man who strongly regretted not having started SR
wanted only a breast enlargement and no vaginoplasty.
With respect to their overall appearanc e, the majority of the T-group reported
satisfaction: 16 subjects were satisfied or very satisfied, 1 was very
dissatisfied, and 2 were neutral. Satisfaction with primary and secondary
sexual characteristics significantly increased after treatm ent. Also, the group's
mean score on the AAI was lower (p < .001) at posttest, indicating that,
according to observers, their appearance had become more compatible with
the new gender.
Of the NT-group, 6 subjects were (very) dissatisfied with their overall
appearance. Five subjects reported being (very) satisfied, while 3 subjects
expressed a neutral view. Over time the NT-group had become more satisfied
with their primary sex characteristics, although there was no change in
satisfaction with the secondary sex characteristics.
Analysis of the NVM of the T-group showed no significant changes after
treatment. When pre- and post test group means were compared with Dutch
normative data, all scores turned out to be within the average range,
indicating normal functioning. The mean T-group's total score on the SCL-90
(psycho- neuroticism) was not significantly lower at posttest than at pretest,
although compared with Dutch normative data this score fell in the high range
at pretest, and in the above average range at posttest. The mean scores on
the subscales anxiety, depression, and hostility were lower at posttest.
The NVM results of the NT-group showed a more dysfunctional profile. When
compared with a Dutch normative group, the subscale psychopathology was
in the high range at pre- and posttest. The results on the SCL-90 were also
unfavorable. The total score was and remained in the high range, as
compared with normative groups. Here, it is important to keep in mind that
probably the patients who functioned worst had not filled out the tests at
follow-up because they were psychiatrically hospitalized.
Treatment Evaluation and Posttreatment Functioning of the Treated Group.
Satisfaction with surgery. Breast removal is emotionally the most important
type of surgery for adolescent FMs because they are advi sed to postpone
metaidoioplasty (transformation of the hypertrophic clitoris into a micro- penis)
or phalloplasty in view of the fact that the surgical techniques are steadily
improving. Only one FM had undergone phalloplasty and two FMs had a
neoscrotum. For the MFs vaginoplasty (including amputation of the penis) is
the most important surgical intervention.
Eight FMs were satisfied with their breast removal, while 5 were
dissatisfied with the result due to the visibility of the scars. Nevertheless, 9 did
not have any problems baring their chest when swimming.
Three MFs expressed satisfaction with their vaginoplasty: they felt their
vaginas looked natural. Another 3 individuals were not completely satisfied,
mostly because they considered their vagina not deep or feminine enough.
Two were disappointed that they could not achieve orgasms. Five MFs had
experienced sexual intercourse without problems. One MF had attempted
intercourse but had a vaginistic response.
Relationships and sexuality. Ten subjects had a stable relationship with one
partner at the time of the interview, 9 had no partner at follow-up or had
never had one. Of the 10 subjects who had a steady sexual partner, 7
expressed satisfacti on with their sex life, 2 expressed a neutral view, and 1 FM
was dissatisfied (5 respons e categories). This FM was unhappy about the fact
that he could not have intercourse with a 'normal' penis with his girlfriend, but
said he was very happy with the relationship itself. He also reported achieving
orgasm every time they had intercourse after his metaidoioplasty.
Several FMs reported that they found it difficult to live without a penis,
especially at moments when they did not know their potential sexual partner
well. Masturbation was not very frequent (5 response categories). MFs
reported a decrease in masturbation frequency after treatm ent (3 response
categories), while FMs reported an increase or no change. Of the 16
individuals who were sexually active, with or without partner, 11 achieved
orgasms regularly (5 response categories).
Eighteen subjects had a compatible sexual orientation (that is MFs
feeling attracted to heterosexual men or homosexual women, and FMs to
heterosexual women or homosexual men)., whereas 1 person was not (yet)
sure about her sexual orientation.
Social life and social contacts. The majority of the T-group (16 subjects) felt
accepted and supported in their new gender role by everyone they knew,
whereas 3 felt accepted and supported by most people (4 response
categories). Sixteen participants had not lost any family memb er or friend or
had lost contact with one person only. Two subjects had lost more than one
friend as a consequence of the SR. Twelve persons indicated that they were
(very) satisfied with their social contacts with the opposite gender as well as
with same gender (13 individuals); 2 subjects were dissatisfied about their
contacts with the opposite gender (5 response categories). Most people (15
subjects) did not feel lonely, 2 felt lonely sometimes, and another 2 felt lonely
quite sometimes (5 response categories).
Superficial contacts such as those with neighbors or shopkeepers were
either non- existent/neutral (6 persons) or positive (13 persons). None of the
subjects had experiences of actually being harassed. One MF subject had a
few times been approached by strangers as a man since the start of her
treatment, while none of the others had ever been approached as someone of
the biological sex. This corresponds with the increase in AAI (p < .001) at
follow-up, indicating a more compatible appearance.
Quality of life. The group reported a reasona ble sense of well-being. Although
not quite comparable, it is worth noting that the negative affect score of this
adolescent group (mean = 4.4; SD = 3.2; range 0 to10) was lower than that of
the randomly selected elderly sample (mean = 6.1; SD = 1.4; range 5 to 10).
Unfortunately no data of a younger comparison group are available.
Self-reported Functioning of the Non- Treated (NT) group at Follow- up.
Relationships and sexuality. Six persons had a stable relationship with a
partner at the time of the interview, as opposed to 9 who had no partner at
follow-up or had never had one. Five subjects reported not knowing what their
sexual orientation was, while another 5 had an incompatible sexual
orientation (meaning that the males felt attracted to homosexual women
and/or heterosexual men, and the females to homosexual men and/or
heterosexual women). Four subjects had a compatible sexual orientation.
Social life and social contacts. Twelve persons reported feeling (very) satisfied
about their social contacts with the opposite gender, whereas 1 person
reported being very dissatisfied (5 response categories). Where same- gender
social contacts were concerned, 9 felt (very) satisfied and 3 felt neutral. Six
(40%) individuals did not feel lonely, another 6 (40%) felt lonely sometimes,
and 3 (20%) felt lonely quite often. (5 response categories).
Quality of life. The NT-group's mean negative affect score (mean = 6.2; SD =
2.6; range 0 to 10) was not different from the mean score of the T-group
(mean = 4.4; SD = 3.2; range 0 to 10) at follow- up. They had, however,
almost exactly the same score as the elderly group (6.1 and 6.2).
Functioning of the Delayed Treated (DT) Group.
The small number of DT-individuals (N=6) and the difference in size compared
to the T-group (N=20) made adequate statistical analyses not possible.
Nevertheless, most mean DT-group scores on the tests measuring
psychological functioning were higher than the mean T-group scores. At
pretest the DT-group scored higher than the T-group on 4 of the 5 NVM scales
(negativism: mean = 29.8, SD = 6.0; somatization: mean = 14.3, SD = 10.3;
shyness: mean = 23.3, SD = 3.2; psychopathology: mean = 9.3, SD = 4.0),
and on 6 (anxiety, agoraphobia, depression, inadequacy, sensitivity, hostility)
of the 8 SCL-90 subscales, creating a higher total mean score on psycho-
neuroticism (mean = 212, SD = 75.5). At posttest the mean DT-scores on the
UGS (mean = 39.8, SD = 6.0), the primary BIS (mean = 19.6, SD = .55), the
NVM shyness (mean = 17.4, SD = 10.2) and NVM psychopathology scale
(mean = 9.0, SD = 8.0), and the same 6 SCL-90 subscales and its total score
(mean = 182, SD = 88.6) were higher than the mean T-group scores. In sum,
the DT-group was psychologically functioning poorer than the T-group at pre-
and posttest. At posttest the DT-group also showed more gender dysphoria
and more body dissatisfaction. However, it is worth remembering that 4 of the
6 persons were approached before they had completed the SRS, because they
were expected to belong to the NT- group.
INSERT TABLE 2 ABOUT HERE
As for the T-sample, our results were very similar to the results of the earlier
retrospective study (Cohen- Kettenis and Van Goozen, 1997). The groups were
comparable with respect to various background variables and in both groups
gender dysphoria had disappeared after treatment. This, of course, is the
main goal of SR. Postoperatively the adolescents were also more satisfied with
their primary and secondary sexual characteristics than at pretest and they
functioned socially and psychologically quite well. Just as in the first study, the
adolescents scored in the normal range with respect to psychological
functioning. Above all, no one expressed feelings of regret concerning the
decision to undergo SRS. Thus, one to five years after surgery, SR does seem
to have been therapeutic and beneficial. Compared with 141 adult Dutch
transsexuals the adolescents seemed to fare better (Kuiper and Cohen-
Kettenis, 1988). The findings of the adult transsexuals are most likely to be
caused by their late treatment because they belonged to the first ever treated
group in The Netherlands, and many of them had to wait until they were in
their fifties, or even longer, before SRS became available. As a consequence
they had more social and psychological problems and they received much less
support from their environment than the adolescents did (Cohen- Kettenis and
Kuiper, 1988). Another reason for their less favorable outcome is that criteria
for (overall functioning of) adolescent applicants are stricter than they are for
adults (see below).
Because in many adolescent transsexuals hormone treatment had
started before they reached the last pubertal phases, they rather easily
passed in their new role. This may also partly explain why they function better
than the adults did. The fact that three observers independe ntly evaluat ed the
adolescents' appearance in accordance with their new gender role supports
the impression of the first study and corroborates the adolescents' satisfaction
with their appearance. Another aspect of this positive outcome may be
attributable to the strict criteria for treatment eligibility. Additional criteria for
treatment eligibility as compared to adults are applied to adolescent starters
before the age of 18. As a consequence, those patients selected for early
treatment belong to the best functioning transsexuals. Finally, most of the
transsexuals in our study were FMs and we know from other studies that FMs
fare in many respects better than MFs postoperatively (Pfäfflin and Junge,
1992). Thus we can conclude that careful diagnosis and strict criteria are
necessary and sufficient to justify hormone treatment (and thereby a start of
SR) in well-functioning adolescents, even if they are younger than 18.
A second aim of the study was to examine whether the decision not to
proceed with the SR procedure for some applicants had been a sensible one.
For the large majority of the NT-patients SRS was contraindicated. Most of the
NT-subjects who were still in contact with the team when this decision came
about agreed. The data clearly support the decision not to allow the NT-group
to start SR. Our questionnaire data, which may be biased to the positive side
(see above), showed that the NT-subjects functioned worse than the T-
subjects. More than half of the NT-group had been given a psy chiatric
diagnosis at application and/or follow- up. One subject had such a difficult time
coping with his psychiatric problems that he had been hospitalized a number
of times and had committed suicide around the time of the follow-up
interview. Not to allow these patients to start medical treatm ent seems to
have been good decision. It is of interest that the majority of the NT-group had
found other ways for dealing with their gender problem to the point that they
actually reported having less gender dysphoria. We can think of two reasons
for this decrease. One is that these applicants had received treatm ent for
major non- gender problems. It could well be that they had overestimated their
gender dysphoria at the time of their application and that they were able to
appraise the intensity of their gender problems in a more realistic way at the
time of follow-up. Another reason is that the intensity and perhaps also the
quality of the gender dysphoria had been different in the NT-group. Indeed, at
the time of application the variation in scores on the gender dysphoria scale
(as reflected by the standard deviations) was much larger in the NT-group
than in the T-group and some clinical reports suggest that there was far more
gender confusion and uncertainty about SRS in the NT-group. Some very
gender dysphoric but unstable applicants pursued SRS again when they were
older (the DT-group). Their general level of functioning was still worse than
that of the T-group despite the fact that some had received additional
Limitations of the stud y
The study also had some limitations, which have to be addressed. One is that
despite our replication of the results of our first study, the total number of
subjects involved remains small. Moreover, our minimum follow-up period
was one year. It goes without saying that longer- term follow-ups are needed
to assess the ultimate outcom e.
A second limitation concerns the possible selection bias of our sample.
Adolescents who apply at a young age probably are in relatively favorable
circumstances, because their parents, though not happy that their children
are transsexual, are usually supportive of treatment. It is also likely that
adolescents with less extreme or more fluctuating cross- gender identities do
not pursue SR so early in life. Because we do not know how many adolescent
transsexuals do not apply for SR, we do not claim that our patients are
represent ative for all Dutch adolescent transsexuals. Our conclusions are
therefore limited to the group that does apply for SR before adulthood.
Third, of those adolescents who apply only the best functioning are
selected for early medical treatment, the majority being FMs. In contrast to
data on sex ratios of our pre- pubertal GID children (boys to girls 5:1; N=120)
and of adult populations (MFs to FMs 3:1, Bakker et al., 1993), the sex ratio
of adolescent applicants for SRS in our clinic approaches 1:1 (N=125). As
mentioned before, after the first diagnostic phase more FMs than MFs were
allowed to start treatme nt, because they better fulfilled the additional
diagnostic criteria for adolescents, namely being psychologically stable and
having a supportive background.
Fourth, although the post- treatm ent interviews were not conducted by
the clinicians who had been involved in the treatment, the patients may still
have emphasized the positive effects because of their belief that the
examiner had a stake in the outcome by virtue of being associated with the
Fifth, a different research design in which, on a random basis, half of the
applicants who are eligible for treatm ent would be treated and the other half
not, would have been better in methodological respect. Clearly, however, such
a study is for ethical reasons impossible to conduct.
Finally, subject retention was better in the T-group than in the NT-group.
The relatively well-functioning NT-patients came to be interviewed and tested,
but many of the non- participants lived in psychiatric or other institutions. So
the data as presented here probably present an optimistic picture and are
most likely not representative of the entire group of NT-applicants.
Clinical implication s
On the basis of the findings of the previous and current study, it seems
reasonable to conclude that early hormone treatment does not necessarily
lead to worse postoperative functioning than later hormone treatment. It
appears to be possible to prevent false positives when following careful
diagnostic procedures. However, careful diagnosis and decision- making in
adolescence does not preclude that rejected or withdrawing applicants will
seek SRS later in life.
It is important to keep in mind that many applicants are no good
candidates for SR and will probably never be. Applicants vary greatly in terms
of family background, education, psychopathology, motivation to explore
gender issues, and outcome. Although psychopathology may be the result
rather than the underlying problem of the GID, SR may also be sought as a
solution to non- gender problems. Starting hormone treatment before
adulthood should not be considered when too many adverse factors operate
simultaneously, despite the possibility that applicants may actually be
transsexual. This is because it is more complicated to make an accurate
diagnosis in problematic adolescents than in well-functioning adolescents,
even for experienced multidisciplinary teams.
The results of our studies point to the desirability of early rather than
late medical interventions. Thus far, the patients in our studies were not
younger than 16 when hormone treatment was started. This raises the
question of the lower limits of such interventions. While a lowering of the age
of cross- sex hormone treat ment is not yet indicated, other medical
interventions offer new possibilities. Recently we described a case of an FM
who had attended the gender clinic at age 16 and had SRS at 18 (Cohen-
Kettenis and Van Goozen, 1998). At application and since the age of 13, she
had already been in treatment by a paediatric endocrinologist with a
luteinizing hormone- releasing hormone (LHRH) agonist, depot triptorelin. This
substance binds so strongly to the pituitary that endogenous LHRH can no
longer exert its effects. Consequently, the pituitary secretion of LH and FSH
and therefore the gonadal production of sex steroids stop. As a result, and
when administered before puberty, puberty will not occur. Given after the
start of puberty, pubertal development will not proceed. An advantage of
pubertal delay over cross- sex hormone treatment is that no irreversible steps
are taken. Moreover, the therapist and the transsexual have time to explore
any problems underlying the cross- gender identity or clarify gender confusion.
Thus far we have successfully used puberty delaying hormones in a few
additional cases. This 'diagnostic aid' could become the next phase in the
management of adolescent transsexualism.
Arrindell WA, Ettema JHM (1986), SCL-90: Handleiding bij een
Multidimensionele Psychopathologie- indicator. Lisse: Swets & Zeitlinger
Bakker A, van Kesteren P, Goo ren LJG, Bezemer PD (1993), The preva lence
of trans sexualism in the Netherlands. Acta Psychiat Scand 87:237- 238
Bradburn N (1969), The Structure of Psychological Well-being . Chicago, IL:
Aldine Publishing Company
Central Bureau of Statistics (1987), De Leefsituatie van de Nederlandse
's Gravenhage: Staatsuitgeverij/CBS-publikaties
Cohen, L. de Ruiter, C. Ringelberg, H. Cohen- Kettenis, PT (1997),
Psychological functioning of adolescent transs exu als: Persona lity and
psy chopathology. J Clin Psychol 53:187- 196
Cohen- Kettenis PT, Gooren LJG (1999), Transsexualism, a review of etiology,
diagnosis and treatment. J of Psychosom Res 46:315- 333
Cohen- Kettenis PT, Kuiper AJ (1984), Transseksualiteit en psychotherapie.
Tijdschr Psychoth 3:153- 166
Cohen- Kettenis PT, Kuiper AJ (1988), Social aspects of sex reassignme nt
surgery. In: Sexology, Eicher W, Kockott G, eds. Berlin: Springer- Verlag,
pp 187- 191
Cohen- Kettenis PT, van Goozen SHM (1997), Sex reassignment of adolescent
transsexuals: A follow- up study. J Am Acad Child Adolesc Psychiatry
Cohen- Kettenis, PT van Goozen SHM (1998), Pubertal delay as an aid in
diag nosis and treat ment of a transsex ual adolescen t. Eur J Child Adol
Psychiat 7:246- 248
Derogatis LR, Lipman RS, Covi L (1973), SCL-90: an outpatient psychiatric
rating scale - preliminary report. Psychopharmacology Bulletin 9:13- 27
Doorn CD, Kuiper AJ, Verschoor AM, Cohen- Kettenis PT (1996), Het Verloop
van de Geslachtsaanpassing: een 5-jarige Prospectieve Studie . Rapport
voor de Nederlandse Ziekenfondsraa d
Kempen GIJM, Ormel J (1992), Het meten van psychologisch welbevinden bij
ouderen. Tijdschr Gerontol Geriatr 23:225- 235
Kuiper AJ (1991), Transsexualiteit: Evaluatie van de Geslachtsaanpassende
Behandeling . Amsterdam: Free University Press
Kuiper AJ, Cohen- Kettenis PT (1988), Sex reassignment surgery: A study of
141 Dutch transsexuals. Arch Sex Behav 17:439- 457
Lindemalm G, Körlin D, Uddenberg N (1987), Prognostic factors vs. outcome in
male- to-female transsexualism: A follow up of 13 cases. Acta Psychiatr
Scand 74:268- 274
Lindgren T, Pauly I (1975), A body image scale for evaluating transsexuals.
Arch Sex Behav 4:639- 656
Luteyn F, Kok AR, van der Ploeg FAE (1980), NVM. Nederlandse Verkorte
MMPI, Handleiding . Lisse: Swet