Arch Sex Behav (2016) 45:575–585

DOI 10.1007/s10508-015-0614-1

ORIGINAL PAPER

Body Satisfaction and Physical Appearance in Gender Dysphoria

Tim C. van de Grift1,2 • Peggy T. Cohen-Kettenis1 • Thomas D. Steensma1 •
Griet De Cuypere3 • Hertha Richter-Appelt4 • Ira R. H. Haraldsen5 • Rieky E. G. Dikmans2 •
Susanne C. Cerwenka4 • Baudewijntje P. C. Kreukels1

Received: 29 September 2014 / Revised: 25 August 2015 / Accepted: 26 August 2015 / Published online: 16 October 2015
 The Author(s) 2015. This article is published with open access at Springerlink.com

Abstract Gender dysphoria (GD) is often accompanied by MtFs included posture, face, and hair, whereas FtMs were mainly
dissatisfaction with physical appearance and body image prob- dissatisfied with hip and chest regions. Clinicians evaluated the
lems. The aim of this study was to compare body satisfaction with physical appearance to be more congruent with the experienced
perceived appearance by others in various GD subgroups. Data gender in FtMs than in MtFs. Within the MtF group, those with
collection was part of the European Network for the Investigation early onset GD and an androphilic sexual orientation had appear-
of Gender Incongruence. Between 2007 and 2012, 660 adults ances more in line with their gender identity. In conclusion, body
who fulfilled the criteria of the DSM-IV gender identity disorder imageproblemsinGDgobeyondsexcharacteristicsonly.Anincon-
diagnosis (1.31:1 male-to-female [MtF]:female-to-male [FtM] gruent physical appearance may result in more difficult psycho-
ratio) were included into the study. Data were collected before the logical adaptation and in more exposure to discrimination and
start of clinical gender-confirming interventions. Sexual orien- stigmatization.
tation was measured via a semi-structured interview whereas
onset age was based on clinician report. Body satisfaction was Keywords Gender dysphoria
Body image
Appearance

assessed using the Body Image Scale. Congruence of appear- Sexual orientation
ance with the experienced gender was measured by means of a
clinician rating. Overall, FtMs had a more positive body image
than MtFs. Besides genital dissatisfaction, problem areas for Introduction

Gender dysphoria (GD) describes a status in which one expe-

riences an incongruence between assigned and experienced
gender. In line with societal and scientific changes, the devel-
& Tim C. van de Grift
t.vandegrift@vumc.nl
opment of the diagnostic criteria for GD has been subject to
change since it first appeared as a diagnosis in the Diagnos-
1
Department of Medical Psychology, Center of Expertise on tic and Statistical Manual of Mental Disorders (American Psy-
Gender Dysphoria, EMGO Institute for Health and Care chiatric Association, 1980).
Research, VU University Medical Center, P.O. Box 7057,
1007 MB Amsterdam, The Netherlands
Although percentages of severe regret are as low as 1–2 %
2
(Cohen-Kettenis & Gooren, 1999; Landen, Wålinder, Ham-
Department of Plastic, Reconstructive & Hand Surgery, VU
University Medical Center, Amsterdam, The Netherlands
bert, & Lundström, 1998), given the impact of gender affirm-
3
ing interventions, clinicians would like to be assisted in their
Center of Sexology and Gender Problems, Ghent University
Hospital, Ghent, Belgium
treatment recommendations by adequate a priori assessment
4
of factors that predict satisfaction with outcome. Some of the
Department of Sex Research and Forensic Psychiatry,
factors associated with dissatisfaction are physical build,
University Medical Center Hamburg-Eppendorf, Hamburg,
Germany incongruence with the new gender role, poor social support,
5 and severe psychological morbidity (Gijs & Brewaeys, 2007).
Department of Neuropsychiatry and Psychosomatic Medicine,
Division of Clinical Neuroscience, Rikshospitalet, Oslo, Body image and physical appearance are related to psychological
Norway well-being among GD individuals (Vocks, Stahn, Loenser, &

123
576 Arch Sex Behav (2016) 45:575–585

Legenbauer, 2009). As the experienced incongruence between different developmental pathways (early onset [EO] and late
physique and gender identity/social role is the source of the dys- onset [LO]) among the GD subtypes (Lawrence, 2010). Cate-
phoria, GD has been conceptualized as a body image syndrome gorizing on the basis of onset age, however, is potentially com-
by some (e.g., Money, 1994). However, only a limited number of plicated as well because of differences in puberty onset, biased
studies have specifically focused on body image in this group (Ålgars, recall, and subtype heterogeneity (Lawrence, 2010; Nieder et al.,
Santtila, & Sandnabba, 2010; Bandini et al., 2013; Becker et al., 2011). For example, within the LO MtF group, there are andro-
2015; Bodlund & Armelius, 1994; Fleming, MacGowan, Robin- philic and gynephilic natal men.
son, Spitz, & Salt, 1982; Kraemer, Delsignore, Schnyder, & Recently, Becker et al. (2015) addressed the differences in
Hepp, 2007; Lindgren & Pauly, 1975; Roback, Strassberg, bodily satisfaction between GD individuals and controls. Specific
McKee, & Cunningham, 1977; Vocks et al., 2009; Wolfradt informationonsubtypedifferences,inthiscasewithregardtophys-
& Neumann, 2001). ical appearance and body image, can contribute to better clinical
Body image is thought of as a person’s self-concept resulting care. As both body satisfaction and therapeutic requests may
from more than solely his or her visual self-image. It is concep- be relatedtotheageofonsetandsexualpreferences,knowledgeon
tualized as consisting of attitudes, experiences, and perceptions subtype differences may help to align gender reaffirming interven-
pertaining to one’s physical appearance, based on self-observa- tions to one’s personal situation.
tion and the reactions of others (see Cash & Pruzinsky, 2002). The
degree of body (dis)satisfaction reflects one’s individual self-con- Aims
cept in relation to the social context. One may assume that a posi-
tive body image is a favorable prognostic factor of quality of life This study aimed to use the concepts of body image and physical
after transition (Bodlund & Armelius, 1994) whereas a negative appearance to provide a better understanding of GD, given their
body image may lead to lower quality of life due to lower self- potential value in GD counseling. The main research objectives
esteem, poorer social functioning, and compensatory conditions, were (1) to describe body (dis)satisfaction and physical appear-
such as eating disorders (Ålgars et al., 2010; Bandini et al., 2013; ance with regard to onset age and sexual orientation in natal males
Bodlund & Armelius, 1994; Vocks et al., 2009). We expect that, and females with GD; (2) to examine the relationship between
even in a population of individuals with GD who have serious self-reported body satisfaction and clinician-reported physical
body image problems, there is variation between individuals. In appearance in individuals with GD.
addition, clinically reported data on physical congruence with the
experienced gender may inform us to what extent the source of
body dissatisfaction can be attributed to differences in physical Method
congruence between the individuals.
In the process of the development of the DSM-5 and the prepa- Participants
ration of ICD-11, and as a result of the changing views on the rela-
tionship between GD and psychopathology, the GD diagnosis as Of the 1019 applicants (MtF = 637 and FtM = 382), a total of 660
well as the specification of certain subtypes have been topics of (64.7 %; MtF = 374 and FtM = 286) who received a GID diag-
debate (Zucker et al., 2013). Sexual orientation and onset age nosis (American Psychiatric Association, 2000) and could be clas-
of GD feelings are most frequently used to categorize people sified as EO or LO were included in the study. At the time of assess-
with GD. ment, the DSM-5 was not published yet. Inclusion was based on a
With regard to sexual orientation, Blanchard, Clemmensen, scoring sheet with GID diagnostic and onset age criteria (Nieder
and Steiner (1987) who proposed and studied a distinction etal.,2011;seebelow aswell).Oftheexcludedgroup(n = 359),93
between homosexual and non-homosexual individuals with GD, (9.1 %) did not fulfill all diagnostic criteria for GID, 103 fulfilled
more recently also denoted with (non-)androphilia in male-to- criteriaforGID,butcouldnotbeassignedtoeithertheearlyonsetor
females (MtFs) and (non-)gynephilia in female-to-males (FtMs) late onset category (residual category), and 163 individuals who
(Cerwenka et al., 2014; Lawrence, 2010). In some studies, it was received a GID diagnosis had missing onset age data. There were
found that, compared to non-androphilic MtFs, androphilic MtFs no statistically significant differences in demographic characteris-
presented earlier for sex reassignment and reported more female tics(ageandeducation)betweentheincludedandexcludedgroups.
identification in childhood (Lawrence, 2010; Smith, van Goozen, For 640 applicants, information on both onset age and sexual ori-
Kuiper, & Cohen-Kettenis, 2005a, 2005b). Non-androphilic MtFs, entation was available. MtF applicants (M = 34.1 years, SD = 12.6
ontheotherhand,weremorelikelytohaveahistoryofsexualarousal years) were significantly older compared to FtMs (M = 27.0 years,
by the image of themselves as a woman (Lawrence, 2010). As a SD = 9.6 years, see Table 1). MtFs had significantly higher educa-
result of criticism concerning the self-report bias and fluidity of tion than FtMs, v2(2) = 12.51, p = .002. More than half of all
sexual orientation, a subtyping based on onset age has been included participants were diagnosed in Amsterdam, whereas
described (Lawrence, 2010; Smith et al., 2005b). Proponents of the other clinics included 22.1 % (Ghent), 17.7 % (Hamburg),
onset age-based subtyping argue that it better reflects the and 8.8 % (Oslo) of the subjects. The MtF to FtM ratio of the

123
Arch Sex Behav (2016) 45:575–585 577

whole sample was 1.31:1. Ratios differed per country, which is Measures
in line with earlier research (Kreukels et al., 2012). As shown in
Table 1, FtM applicants were significantly more likely to have Demographic data, information on social transitioning, pre-
(partially) transitioned than MtF applicants in private, v2(2) = vious hormone treatment, and sexual orientation were taken
26.35,p\.001orworklife,v2(2) = 29.82,p\.001.Self-prescribed from a background interview (Kreukels et al., 2012).
hormone use (self-report) prior to admission was significantly The criteria of the formal GID diagnosis were scored on a
more common in MtFs (22.0 %) than in FtMs (7.3 %), v2(1) = self-constructed form, based on the DSM-IV-TR criteria
27.71, p\.001. In MtF applicants, age of admission was signif- (American Psychiatric Association, 2000). Based on a similar
icantly correlated with sexual orientation; younger MtFs were sheet, while using childhood diagnostic criteria, onset age
more likely to report androphilic orientation (r = .32, df = 365, was assessed. These forms were completed by the clinician at
p\.001) whereas this was not found in the FtM group (point the end of the diagnostic phase. If both DSM-IV-TR core
biserial correlations; r = -.01, df = 271). criteria of GID in childhood were fulfilled, individuals were
categorized as EO (pre-pubertal ‘‘strong cross-gender iden-
Procedure tification’’ and ‘‘persistent discomfort about one’s assigned
sex’’). In case of (post-) pubertal onset of the GID (neither‘‘strong
Data collection was part of the European Network for the Inves- cross-gender identification’’nor‘‘persistent discomfort about
tigation of Gender Incongruence (ENIGI) between January 2007 one’s assigned sex’’ before puberty were reported), individ-
and October 2012. Individuals 17 years of age and older applying uals were classified as LO (Nieder et al., 2011). Individuals
for sex reassigning interventions in Amsterdam (the Netherlands), who fulfilled only one of the criteria in childhood were cat-
Ghent (Belgium), Hamburg (Germany), and Oslo (Norway) were egorized in the residual group.
asked to participate. All data were collected during the diagnostic Sexualorientationwasmeasuredbyoneitemfromasemi-struc-
procedure before receiving any clinical gender-confirming med- tured Background Interview (Kreukels et al., 2012) and classified
ical interventions. For information on the ENIGI protocol, see according to the person’s experienced sexual attraction to others.
Kreukels et al. (2012). Rating based on clinician-reported sexual orientation was strongly

Table 1 Sample characteristics (applicants with formal GID diagnosis)

MtF FtM
n = 374 n = 286

Mean age (in years; SD)a 34.1 (12.6) 27.0 (9.6)

Education
Low (%) 79 (21.5) 66 (23.5)
Intermediate (%) 199 (54.1) 177 (63.0)
High (%) 90 (24.5) 38 (13.5)
Social role at admission (private)
Experienced gender (%) 191 (52.6) 199 (72.6)
Variable (%) 125 (34.4) 54 (19.7)
Natal gender (%) 47 (12.9) 21 (7.7)
Social role at admission (work)
Experienced gender (%) 126 (38.1) 134 (52.8)
Variable (%) 30 (9.1) 42 (16.5)
Natal gender (%) 175 (52.9) 78 (30.8)
Hormone use at admission (%) 81 (22.0) 16 (7.3)
Inclusions per center N % MtF:FtM ratio

Amsterdam 339 51.4 1.63:1

Ghent 146 22.1 2.24:1
Hamburg 117 17.7 1:1.34
Oslo 58 8.8 1:3.46

Due to missing data, variable sums may not add up to the described number of participants
a
t(658) = 7.97, p\.001

123
578 Arch Sex Behav (2016) 45:575–585

correlated to the self-reported measure (phi correlation; ø = .73, onset age was calculated using phi correlations. BIS subscale reli-
df = 616, p\.001). Results were scored on a Kinsey scale ranging abilities were calculated by means of Cronbach’s alphas.
from being exclusively attracted to one’s natal sex to being exclu- All analyses were repeated after excluding participants who
sivelyattractedtotheothersex.Responsecategoriesforbeingattracted wereonhormonal therapypriortoadmission,ashormonesinduce
to transgenders or being asexual were added. Androphilia in MtFs physical changes, and consequently may influence both fem-
and gynephilia in FtMs was defined as being attracted completely ininity/masculinity of body parts and satisfaction with one’s
or primarily to one’s natal sex. Non-androphilic MtFs and non- physique.
gynephilicFtMsincludedallotheroptions(havingabisexualattrac-
tion, an exclusive attraction to the other natal sex, an attraction to
transgenders and asexuality).
Results
Body image was measured by the Body Image Scale (BIS). It
consists of 30 items to determine satisfaction with various body
Distribution of Sexual Orientation and Onset Age
parts, rated on a 5-point scale of satisfaction ranging from very
Among Natal Males and Females
satisfied(1)toverydissatisfied(5)(Lindgren&Pauly,1975).There
are two versions of the scale: one for natal males and one for
Within the MtF subgroup, the androphilic (n = 126)-non-an-
natal females. The BIS includes primary sex characteristics,
drophilic (n = 241) ratio was 1:1.91 and the early (n = 190)-late
secondary sex characteristics, and neutral (non sex-related)
onset (n = 177) ratio was 1.07:1. In the FtM group, the gynephilic
body parts. The BIS contains equivalent sex-specific genital
(n = 219)-non-gynephilic (n = 54) ratio was 4.05:1 and the early
body parts to enable MtF–FtM comparisons. Higher scores
(n = 230)-late onset (n = 43) ratio was 5.34:1 (Table 3). Sexual
represent higher degrees of body dissatisfaction.
orientation and age of onset correlations were ø = .26 (MtFs; df =
Lindgren and Pauly (1975) proposed a subscale analysis of
365, p\.001), and ø = .21 (FtMs; df = 271, p\.001). This indi-
the BIS, using the subscales primary sex characteristics, sec-
cates a higher likelihood of androphilic sexual orientation in early
ondary sex characteristics, and neutral characteristics. How-
onset MtFs and of gynephilic sexual orientation in early onset
ever, these subscales do not allow for comparisons per body
FtMs.
area. Therefore, an alternative clustering based on body areas
within the BIS (see Table 2) was used. Cronbach’s alphas on
the subscales for the sample are shown in Table 2. Differences in Body Image and Physical Appearance
To measure physical appearance, the Physical Appearance Between the Natal Sexes
Scale (PhAS) was used. This scale scores the observer’s
appraisal of the masculinity/femininity of a person’s physical MtFs scored significantly higher than FtMs on the overall scores
appearance, rated on a 5-point scale ranging from most con- of both the BIS (ANOVAs; M = 101.27, SD = 15.66 vs. M =
gruent with the experienced gender (1) to most incongruent 96.27, SD = 14.93; p = .001) and the PhAS (ANOVAs; M =
with the experienced gender (5) (Smith et al., 2005b). The scale 42.28, SD = 9.55 vs. M = 39.18, SD = 7.00; p\.001) scales, indi-
contains 14 items, and scoring differs per natal sex. Higher scores catinglowerbodysatisfactionandalesscongruentphysicalappear
represent a physical appearance that is less congruent with the ance with their experienced gender.
experienced gender. On the BIS items, MtFs reported highest dissatisfaction scores
on the socially related body parts (such as voice), but also on their
Statistical Analysis hair, their face and neck, and posture. FtMs, on the other hand,
reported the highest discomfort with their breasts. Other reported
The degree of masculinity/femininity of the PhAS items was areas of discomfort were the hip region and chest size. Dissatis-
recoded, based on the person’s natal sex. Sexual orientation faction with the genitals was high in both groups, although MtFs
and onset age subgroups were compared with regard to over- tended to score higher on equivalent body parts (e.g., penis versus
all scores, subscale scores, and scores on individual items of clitoris although not significant on all items). After Bonferroni
BIS and PhAS by means of one-way ANOVAs. These tests correction for multiple comparisons, differences between MtFs
were carried out for the total group as well as for MtFs and and FtMs remained significant for most BIS items in ANOVA
FtMs separately. Within the natal sex groups, BIS scores were testing (Table 4).
compared with regard to transition status, using independent Clinicians assessed FtMs’ appearance as more congruent with
t-tests. Bonferroni corrections were used to control for mul- the experienced gender than MtFs’ appearance. MtFs’ appearance
tiple comparisons. Bonferroni corrected p values were .0017 was less congruent with the experienced gender regarding
(.05/30) for BIS comparisons and .0036 (.05/14) for PhAS motor movement, speech and voice, hair, facial features, and
comparisons. Multiple stepwise linear regression analyses of muscularity.AllsignificantdifferencesbetweenMtFsandFtMsin
sexual orientation and onset age predicting PhAS and BIS scores PhAS items, except for Adam’s apple, feet/hands and figure,
were performed. The correlation between sexual orientation and remained significant after Bonferroni correction.

123
Arch Sex Behav (2016) 45:575–585 579

Table 2 Body Image Scale subscales (Lindgren & Pauly, 1975)

Subscale Items Construct analysis

Lindgren and Pauly (1975)

Primary sex characteristics Body hair, breasts, facial hair, penis/clitoris, scrotum/vagina, and a = 0.65
testicles/uterus 6 items
Secondary sex characteristics Appearance, arms, body movement, bottom, chest size, figure, hair, a = 0.84
hips, muscles, thighs, upper arm muscles, voice, waist, and weight 14 items
Neutral characteristics Adam’s apple, chin, eye brows, face, feet, hands, height, legs, nose, a = 0.81
and shoulders 10 items
Body area subscales
Social and hair items Appearance, body hair, body movement, facial hair, hair, and voice a = 0.72
6 items
Head and neck region Adam’s apple, chin, eye brows, face, and nose a = 0.74
5 items
Muscularity and posture Arms, feet, hands, height, legs, muscles, shoulders, upper arm a = 0.79
muscles, and weight 9 items
Hip region Bottom, figure, hips, thighs, and waist a = 0.82
5 items
Chest region Breasts and chest size NA
Genitals Penis/clitoris, scrotum/vagina, and testicles/uterus, and ovaries a = 0.85
3 items

Table 3 Distribution of sexual orientation and onset age subgroups

MtF FtM
Androphilic Non-androphilic Gynephilic Non-gynephilic

Early onset 88 (24.0%) 102 (27.8%) 193 (70.7%) 37 (13.6%)

Late onset 38 (10.4%) 139 (37.9%) 26 (9.5%) 17 (6.2%)
2 2
MtF; v (1) = 25.09, p\.001; FtM; v (1) = 12.55, p\.001

When comparing the outcomes of the BIS and PhAS (sub)- Differences in Body Image and Physical Appearance
scores, clinicians and applicants reports generally showed similar Between Sexual Orientation and Onset Age
patterns. This implies that the physical characteristics related to Subgroups Among MtFs and FtMs
higher dissatisfaction were mostly the ones also considered less
congruent by clinicians. The only item in which self-report con- With regard to overall body satisfaction (i.e., BIS scores), no
flicted with clinician report was the figure (clinicians reported significant differences between sexual orientation and onset
FtMs more congruent although they reported to be more dissat- age subgroups were found in both natal sexes. The only trend
isfied with their figure). was the relatively highly reported body dissatisfaction of LO
No major differences in the overall and subscale scores were FtMs (approaching MtF levels) compared to their EO coun-
observedafterexcludingparticipantsonpriorhormonaltreatment terparts (p = .095; see Fig. 1). Excluding participants who used
(n = 97). Only the BIS Adam’s apple item was no longer signifi- hormonal therapy prior to admission did not change these
cantly different between the MtF and FtM subgroups (ANOVA; findings.
p = .002 after Bonferroni correction). When comparing partici- Concerning the overall congruence of physical appearance
pants with and without prior hormone use, overall physical (i.e., PhAS scores), non-androphilic MtFs were considered
congruence with the experienced gender was significantly lower significantly less feminine than androphilic MtFs (p\.001).
in people who did not receive hormonal treatment (M =41.27) Similarly, LO MtFs scored significantly less feminine than
compared to the ones who did (M = 39.77, p = .006). No sig- EO MtFs (p\.001) in ANOVA testing. In FtMs, no statisti-
nificant difference between overall BIS scores was found (all cally significant subgroup differences were found. Neverthe-
ANOVAs). less, gynephilic and EO applicants tended to score somewhat

123
580 Arch Sex Behav (2016) 45:575–585

Table 4 Body image and physical appearance scores in male-to-females versus female-to-males
Self-reported (BIS) Clinician-reported (PhAS) Test characteristics
MtF FtM MtF FtM BISb PhASb
M (SD) M (SD) M (SD) M (SD) F (df) F (df)

Social and hair items

Appearance 3.37 (1.13) 3.22 (1.07) NA NA 2.76 (1, 606) NA
Body hair 4.33 (0.96) 3.37 (1.05) NA NA 139.45 (1, 606)*** NA
Body movement 2.99 (1.11) 2.58 (0.87) 2.44 (1.00) 2.01 (0.68) 25.19 (1, 613)*** 35.94 (1, 610)***
Facial hair 4.66 (0.77) 3.67 (1.17) 3.10 (1.05) 3.20 (0.82) 154.81 (1, 597)*** 1.61 (1, 612)
Hair 3.14 (1.37) 1.99 (0.77) 2.46 (1.24) 2.11 (0.85) 151.04 (1, 614)*** 15.06 (1, 613)***
Speech NA NA 2.59 (1.04) 2.28 (0.82) NA 15.67 (1, 611)***
Voice 3.99 (1.09) 3.86 (1.12) 3.22 (1.10) 2.92 (0.98) 1.91 (1, 611) 11.77 (1, 614)**
Head and neck region
Adam’s apple 3.62 (3.62) 3.25 (1.04) 3.40 (0.84) 3.24 (0.71) 13.01 (1, 564) *** 5.98 (1, 599)*
Chin 2.98 (1.04) 2.25 (0.90) 3.16 (0.81) 2.88 (0.72) 83.99 (1, 613)*** 19.83 (1, 613)***
Eye brows 2.86 (1.11) 2.18 (0.81) NA NA 69.87 (1, 612)*** NA
Face 3.27 (1.08) 2.68 (0.99) NA NA 47.27 (1, 611)*** NA
Jaw NA NA 3.21 (0.82) 2.89 (0.71) NA 25.08 (1, 611)***
Nose 3.03 (1.21) 2.02 (0.78) 3.24 (0.85) 2.85 (0.68) 142.22 (1, 613)*** 39.59 (1, 612)***
Skin NA NA 2.98 (0.89) 3.06 (0.73) NA 1.73 (1, 613)
Muscularity and posture
Arms 2.64 (0.90) 2.58 (0.98) NA NA \1 (1, 616) NA
a
Feet 3.14 (1.13) 2.52 (0.90) 3.27 (0.95) 3.07 (0.78) 55.24 (1, 616)*** 7.40 (1, 611)**
Handsa 2.93 (1.11) 2.43 (1.00) 3.27 (0.95) 3.07 (0.78) 33.38 (1, 612)*** 7.40 (1, 611)**
Height 2.62 (1.14) 3.02 (1.13) 3.16 (1.03) 3.25 (0.92) 18.66 (1, 610)*** 1.21 (1, 615)
Legs/calves 2.54 (0.97) 2.35 (0.96) NA NA 5.91 (1, 607)* NA
Muscles 3.01 (0.98) 3.12 (1.14) 3.06 (0.80) 2.81 (0.77) 1.55 (1, 610) 14.79 (1, 611)**
Shoulders 2.92 (1.04) 2.35 (1.05) NA NA 45.64 (1, 616)*** NA
Upper arm muscles 3.20 (1.02) 3.20 (1.14) NA NA \1 (1, 597) NA
Weight 2.88 (1.18) 3.03 (1.14) NA NA 2.680 (1, 612) NA
Hip region
Bottom 2.89 (1.08) 3.32 (1.06) NA NA 24.54 (1, 615)*** NA
Figure 3.13 (1.14) 3.56 (1.06) 3.00 (1.06) 2.76 (0.96) 22.75 (1, 610)*** 7.89 (1, 613)**
Hips 3.25 (1.11) 3.62 (1.14) NA NA 16.31 (1, 615)*** NA
Thighs 2.75 (1.03) 3.33 (1.10) NA NA 44.74 (1, 607)*** NA
Waist 3.09 (1.12) 3.63 (1.03) NA NA 37.07 (1, 600)*** NA
Chest region
Breasts 4.20 (1.05) 4.81 (0.60) NA NA 69.45 (1, 601)*** NA
Chest size 3.54 (1.10) 4.02 (1.13) NA NA 26.92 (1, 595)*** NA
Genitals
Penis/clitoris 4.55 (0.82) 4.28 (1.02) NA NA 12.55 (1, 591)*** NA
Scrotum/vagina 4.62 (0.68) 4.54 (0.81) NA NA 1.46 (1, 593) NA
Testicles/ovary 4.64 (0.68) 4.61 (0.77) NA NA \1 (1, 590) NA
Overall 101.27 (15.66) 96.27 (14.93) 42.28 (9.55) 39.18 (7.00) 11.01 (1, 435)** 18.94 (1, 593)***
One-way ANOVA; Bonferroni corrected: PhAS .05/14 = .0036; BIS .05/30 = .0017; NA the item is not applicable to this scale
a
Combined item in PhAS
b
* p\.05, ** p\.01, *** p\.001

123
Arch Sex Behav (2016) 45:575–585 581

120

110

100

BIS
90

80

70

60
MtF - sexual orientation a MtF - onset age b FtM - sexual orientation c FtM - onset age d
androphylic non-androphylic early onset late onset gynephilic non-gynephylic early onset late onset
n= 77 n= 189 n= 129 n = 138 n= 131 n= 33 n= 144 n= 26

Fig. 1 Total Body Image Scale scores in male-to-female and female-to-male sexual orientation and onset age subgroups. One-way ANOVA, absolute
range = 30 (most satisfied)–150 (most dissatisfied). aF\1, df = 1, 264; bF\1, df = 1, 265; cF\1, df = 1, 162; dF\1, df = 1, 168

more congruent with the experienced gender (see Fig. 2). As expected, individuals reported the highest degree of dis-
Repeating the analyses while excluding participants who used satisfaction with their primary and secondary sex characteristics,
hormonal therapy prior to admission resulted in similar find- but body dissatisfaction in GD appeared to go beyond this kind of
ings. sex-anatomically related dysphoria. The findings on primary and
Regression analysis showed that, in the MtF subgroup, satis- secondary sex characteristics were generally in line with earlier
faction with body parts of social relevance and hair was predicted research (Ålgars et al., 2010; Bandini et al., 2013; Bodlund &
by sexual orientation (see Table 5). Furthermore, overall reported Armelius, 1994; Fleming et al., 1982; Kraemer et al., 2007; Vocks
physical congruence was predicted by both sexual orientation etal.,2009;Wolfradt&Neumann,2001).Ourdatashowedhigher
and onset age. In the sample as a whole, sexual orientation and overall scores on both the BIS and the PhAS in MtFs, compared to
onset age were both weak predictors of body image and physical FtMs, indicating less body satisfaction and a physical appearance
appearance, although sexual orientation was a somewhat stronger thatwaslesscongruent withtheexperiencedgender.Asdescribed
predictor than onset age. This suggests that MtFs with early onset earlier, body image is often conceptualized as one’s self-concept
gender dysphoria and androphilic sexual orientation more often of physique in relation to the social context (Cash & Pruzinsky,
have a more satisfactory body image and physical appearance con- 2002). The source of the observed differences in body satis-
gruent with their gender identity. No significant predictors for faction between the groups, therefore, may be found in both
body image and physical appearance scores were found in the physical characteristics and psychosocial characteristics.
FtM applicants. Significantly more FtMs lived (partially) in their experi-
enced gender role, compared to MtFs. Transition in both private
and work life before they entered the clinic corresponded with
Social Transition and Body Image significantly lower reported body dissatisfaction. Therefore, the
difference between the sexes in body (dis)satisfaction may be
In MtFs, social transition in private life at clinical admission related to the difference in social transitioning between the
correspondedwithlowerBISscores(M = 107.97,SD = 13.84vs. groups: FtMs are more satisfied with their body and this may be
M = 96.39, SD = 16.07 for socially transitioned; p\.001), indi- due to more frequent social transition. In society, the masculine
cating lower body dissatisfaction in this group. The same was role of FtMs is generally more accepted than the female role of
found for social transition at work (M = 106.02, SD = 13.53 vs. MtFs, making social transition for the first group easier. Living
M = 94.50, SD = 15.27 for socially transitioned; p\.001). In the in the social role of the experienced gender may contribute to a
FtM group, no such differences were found when performing more positive attitude toward one’s body. On the other hand, the
ANOVAs. ones who already have a more positive body image may also be
the ones that transitioned earlier.
Other factors that differ between the sexes and that may
Discussion explaindifferencesinbody(dis)satisfactionbetweenthesegroups
are prior hormone use and age. We could, however, not confirm
The objective of this study was to assess self-reported body a relation between prior hormone use or age and the degree of
(dis)satisfaction and physical appearance as evaluated by body (dis)satisfaction. Although (self-)administration of hor-
clinicians in relation to various subgroups of individuals with mones is expected to influence the congruence of physical char-
GD before the start of medical treatment. acteristics with the experienced gender, the results may have been

123
582 Arch Sex Behav (2016) 45:575–585

55

50

45

PhAS
40

35

30

25

20
MtF - sexual orientation a MtF - onset age b FtM - sexual orientation c FtM - onset age d
androphylic non-androphylic early onset late onset gynephilic non-gynephylic early onset late onset
n= 117 n= 223 n= 186 n = 160 n= 191 n= 46 n= 213 n= 36

Fig. 2 Total Physical Appearance Scale scores per subgroup in male-to- gender)–70 (least congruent with experienced gender). aF = 39.29, df = 1,
female and female-to-male sexual orientation and onset age subgroups. 338, p\.001; bF = 24.69, df = 1, 344, p\.001; cF = 3.708, df = 1, 235,
One-way ANOVA, absolute range = 14 (most congruent with experienced p = .055; dF\1, df = 1, 247

Table 5 Results of regression analyses for physical appearance and body image with sexual orientation and onset age as predictors
Predictorsa,c bb,d

Whole sample
BIS sum Sexual orientation .10*
BIS social and hair items Sexual orientation .28***
Onset age -.11**
BIS chest region Sexual orientation -.16***
Onset age .09*
PhAS sum Sexual orientation .26***
Onset age .16***
MtFs
BIS social and hair items Sexual orientation .18**
BIS genitals Onset age -.13*
PhAS sum Sexual orientation .27***
Onset age .20***
a
Sexual orientation labels: 1 = androphilic (MtFs) or gynephilic (FtMs), 2 = non-androphilic (MtFs) or non-gynephilic (FtMs); onset age labels:
1 = early onset, 2 = late onset
b
Higher PhAS corresponds with less physical congruence with the experienced gender; higher BIS scores represent higher degree of body
dissatisfaction
c
Phi correlation sexual orientation and onset age; ø = .36, df = 638, p\.001
d
* p\.05, ** p\.01, *** p\.001
BIS Body Image Scale, PhAS Physical Appearance Scale

unsatisfactory, because these individuals applied to a clinic directly, but also influence the social evaluation of a person’s
to receive further gender-confirming treatment. In addition, physical characteristics, and significantly more FtMs lived
hormones may not have been used long enough or in subop- (partially) in their experienced gender role, compared to MtFs.
timal doses to induce the desired physical changes. Finally, sex differences in physical appearance and body satis-
Clinicians judged FtMs as more physically congruent with the faction may be explained by the construction of gender as
experienced gender than MtFs on all listed body items. As MtF described by Kessler and McKenna (1978). The attribution of
applicants were older than FtMs, and age was significantly cor- gender primarily depends on the existence or absence of male traits
related with higher physical incongruence scores, this may (e.g., physical masculinization). Masculine body characteristics
explain some of the difference between the natal sexes. In addi- (e.g., hair growth, facial characteristics) are often more diffi-
tion, social transition may not only favor body satisfaction cult to mask and, therefore, it may be more difficult for MtFs to

123
Arch Sex Behav (2016) 45:575–585 583

present themselves in a feminine way than vice versa. A similar physically ‘‘pass’’ as the experienced gender when transi-
argument for social transition and body image may be followed tioning earlier in life, resulting in higher chance of finding a
here for social transition and physical appearance. Social partner from the preferred gender role and developing a more
transition may be easier for FtMs, because their appearance is satisfactory self-image.
more easily aligned to their experienced gender. In addition, Inrelationtophysicalappearance,one’ssexualorientationmay
people who have already transitioned might be more easily per- also be informative on possible membership of a certain subcul-
ceived as their experienced gender than those who have not. ture. The male homosexual subculture is known to have high stan-
The observation that FtMs were perceived as more congruent dards on physical appearance whereas the lesbian subculture is
with their experienced gender, even without (hormonal) treat- more tolerant toward diversity in appearance (Ålgars et al., 2010;
ment, may also be indicative of the social attitudes toward bodily Morrison, Morrison, & Sager, 2004; Vocks et al., 2009). Applying
masculinity and femininity. For MtFs, pronounced features, such high bodily standards to oneself may increase the likelihood of
as jaw line or facial hair growth, may impede their feminine appear- body dissatisfaction. Sexual orientation and relational functioning
ance. These body attributes, which are most difficult to hide, are may also influence one’streatment preferences (Cerwenka et al.,
theoneswiththe highest dissatisfaction scores, whencompared to 2014); sexuality could be a decisive factor in choosing for a phal-
the other sex. The different areas of dissatisfaction for the natal loplasty or characteristics of the neovagina (such as depth).
sexes could also be the result of a difference of importance which The relationship between sexual orientation and onset age
is attributed to this item in personal and societal standards (e.g., remains a topic of debate (Lawrence, 2010). Recently, the sub
masculine mesomorphic standard), how this body item impacts workgroup on the DSM-5 classification concluded that clinical
social interaction, and if it can be influenced via modifying tech- decisionsare currently no longer based onthe sexual orientation
niques (e.g., such as make-up, clothing, surgery, or weight loss). classification (Zucker et al., 2013). Onset age and sexual orienta-
With regard to reported total body (dis)satisfaction, no sub- tion correlate as low as ø = .26 (MtFs) and .21 (FtMs), a find-
type differences were found within the MtF and FtM groups. ing in line with earlier research (Lawrence, 2010). Therefore,
Clinicians, however, viewed the physical appearance of appli- one cannot be substituted for the other. Although both variables
cants with a sexual preference for theirnatal sex (i.e., androphilic appeared to be weak predictors of body (dis)satisfaction and of
MtFs and gynephilic FtMs) and with an early onset more con- clinician-viewedphysicalcongruence,sexualorientationappeared
gruent with their experienced gender. Their sexual preference to be a stronger predictor of physical appearance, and (aspects of)
and relational experiences may have steered androphilic MtFs body image than onset age. Therefore, information on sexual ori-
and gynephilic FtMs toward presenting their appearance in a entation, acknowledging the shortcomings of this concept, may
more congruent way. In contrast, non-androphilic MtFs and contribute to a more focused counseling in some individuals when
non-gynephilic FtMs may have had ‘‘heterosexual’’ relation- it comes to body changing interventions. Gender role in previous
ships prior to admission, in which they may have been more relationships and the assumed impact of medical interventions
likely to present their physique in a way that corresponded with should be subject of counseling. Also, sexual behavior should be
social norms of the natal sex, rather than of the experienced considered when choosing gender affirming interventions (e.g.,
gender (Cerwenka et al., 2014). In case of LO gender dysphoric possibility of vaginal penetration).
people, this may be related to the fact that cross-gender identi-
fication and presentation became more present at a later age. Limitations
Findings on physical differences between sexual orientation sub-
types (Blanchard et al., 1987), such as lower body weight of The current study was limited by the self-report character of
androphilic MtFs, have not been replicated (Smith et al., 2005b). sexual orientation, onset age, and BIS. As data were collected
Moreover, sexual orientation is described to be fluid over time during the diagnostic phase, individuals might have respon-
(Cohen-Kettenis & Pfäfflin, 2010). Therefore, an explanation for ded in a socially desirable way to receive a diagnosis and,
the perceived subtype differences in physical congruence may be therefore, treatment. Furthermore, the assessment of physical
more likely found in differences in relational role and the use of appearance was done by only one clinician. However, earlier
body part modifying techniques between the sexual orientation data published on this scale found inter-observer correlation
subgroups. Individuals who are enabled to live in the social role coefficients ranging from .68 to .79 for the individual items
of their experienced gender within their relationships may feel (Smith et al., 2005a).
more empowered to express this role socially through clothing, Data on the BIS and PhAS scales were collected at different
hairstyle, make-up, and physical behavior. moments of the diagnostic phase; data on body image were col-
An explanation for the more congruent physical appear- lected at the beginning of the diagnostic procedure, whereas data
ance of the EO versus the LO applicants may be found in their on physical appearance were collected later on in the diagnostic
younger age. As mentioned before, younger age was found to process. Finally, as mentioned earlier, the concept of sexual ori-
be associated with a more congruent physical appearance entation has its limitations. In the GD population, sexual orien-
with the experienced gender. In addition, it may be easier to tation may be subject to change over the course of transition,

123
584 Arch Sex Behav (2016) 45:575–585

perhaps even more than in non-GD populations (Cohen-Kettenis American Psychiatric Association. (2000). Diagnostic and statistical
& Pfäfflin, 2010). Therefore, the conclusions of this study regard- manual of mental disorders (4th ed.). Arlington, VA: Author.
Bandini, E., Fisher, A. D., Castellini, G., Lo Sauro, C., Lelli L.,
ing this concept merely apply to the phase before clinical inter- Meriggioloa, M. C., … Ricca, V. (2013). Gender identity disorder
ventions and the diagnostic considerations made at that point. and eating disorders: Similarities and differences in terms of body
Also, in this study sexual orientation was coded as a dichotomous uneasiness. Journal of Sexual Medicine, 10, 1012–1023
measure whereas actual sexual orientation may be viewed on a Becker, I., Nieder, T. O., Cerwenka, S., Briken, P., Kreukels, B. P.,
Cohen-Kettenis, P. T., … Richter-Appelt, H. (2015). Body image in
continuum. The classification based on the onset of gender dys- young gender dysphoric adults: A European multi-center study.
phoria is limited by the fact that some people could not be cate- Archives of Sexual Behavior. doi:10.1007/s10508-015-0527-z.
gorized as early or late onset (i.e., the residual group). Blanchard, R., Clemmensen, L. H., & Steiner, B. W. (1987). Heterosexual
and homosexual gender dysphoria. Archives of Sexual Behavior, 16,
139–152.
Conclusion Bodlund, O., & Armelius, K. (1994). Self-image and personality traits in
gender identity disorders: An empirical study. Journal of Sex and
Body image problems in GD go beyond sex characteristics and Marital Therapy, 20, 303–317.
congruence of physical appearance only. As body dissatisfac- Cash, T. F., & Pruzinsky, T. (2002). Body image: A handbook of theory,
research, and clinical practice. New York: Guilford Press.
tion may be indicative of one’s ability to adapt to one’s body and Cerwenka, S., Nieder, T. O., Briken, P., Cohen-Kettenis, P. T., Cuypere,
of (hidden) clinical expectations, it seems a valuable target of G. D., Haraldsen, I. R., … Richter-Appelt, H. (2014). Intimate
counseling at admission. Particularly, individuals with low body partnerships and sexual health in gender-dysphoric individuals
satisfaction extending beyond sex characteristics only should before the start of medical treatment. International Journal of Sex-
ual Health, 26, 52–65.
receive special attention. Information on sexual orientation may Cohen-Kettenis, P. T., & Gooren, L. J. G. (1999). Transsexualism: A
be informative as it may have an impact on preferences for body- review of etiology, diagnosis and treatment. Journal of Psychoso-
related interventions. Ultimately, more effective counseling matic Research, 46, 315–333.
should make individuals more resilient during transition and Cohen-Kettenis, P. T., & Pfäfflin, F. (2010). The DSM diagnostic criteria
for gender identity disorder in adolescents and adults. Archives of
medical interventions. As slightly different patterns of clinician- Sexual Behavior, 39, 499–513.
reported PhAS and self-reported BIS scores were observed, one Fleming, M. Z., MacGowan, B. R., Robinson, L., Spitz, J., & Salt, P.
should be aware of the potential bias between external and (1982). The body image of the postoperative female-to-male trans-
internal interpretation of physique. A congruent appearance sexual. Journal of Consulting and Clinical Psychology, 50, 461–462.
Gijs, L., & Brewaeys, A. (2007). Surgical treatment of gender dysphoria
does not necessarily imply a positive body image. Therefore, in adults and adolescents: Recent developments, effectiveness, and
clinicians should remain sensitive to potential body image issues challenges. Annual Review of Sex Research, 18, 178–224.
in all applicants, not specifically in the group who they expect it Kessler, S. J., & McKenna, W. (1978). Gender: An ethnomethodological
to have. approach. Chicago: University of Chicago Press.
Kraemer, B., Delsignore, A., Schnyder, U., & Hepp, U. (2007). Body
image and transsexualism. Psychopathology, 41, 96–100.
Acknowledgments The authors would like to thank all clinicians and Kreukels, B. P. C., Haraldsen, I. R., De Cuypere, G., Richter-Appelt, H.,
participants, Ellis van der Putten-Bierman for her contribution in an Gijs, L., & Cohen-Kettenis, P. T. (2012). A European network for
earlier phase of the study, and Jos Megens for his permanent support for the investigation of gender incongruence: The ENIGI initiative.
the ENIGI initiative. European Psychiatry, 27, 445–450.
Landen, M., Wålinder, J., Hambert, G., & Lundström, B. (1998). Factors
Compliance with Ethical Standards predictive of regret in sex reassignment. Acta Psychiatrica Scan-
dinavica, 97, 284–289.
Conflict of interest None. Lawrence, A. A. (2010). Sexual orientation versus onset age as bases for
typologies (subtypes) for gender identity disorder in adolescents
Open Access This article is distributed under the terms of the Creative and adults. Archives of Sexual Behavior, 39, 514–545.
Commons Attribution 4.0 International License (http://creativecomm Lindgren, T. W., & Pauly, I. B. (1975). A body image scale for evaluating
ons.org/licenses/by/4.0/), which permits unrestricted use, distribution, transsexuals. Archives of Sexual Behavior, 4, 639–656.
and reproduction in any medium, provided you give appropriate credit to Money, J. (1994). Body-image syndromes in sexology: Phenomenology and
the original author(s) and the source, provide a link to the Creative classification. Journal of Psychology & Human Sexuality, 6, 31–48.
Commons license, and indicate if changes were made. Morrison, M. A., Morrison, T. G., & Sager, C. L. (2004). Does body
satisfaction differ between gay men and lesbian women and hetero-
sexual men and women?: A meta-analytic review. Body Image, 1,
127–138.
Nieder, T. O., Herff, M., Cerwenka, S., Preuss, W. F., Cohen-Kettenis, P.
References T., De Cuypere, G., … Richter-Appelt, H. (2011). Age of onset and
sexual orientation in transsexual males and females. Journal of Sexual
Ålgars, M., Santtila, P., & Sandnabba, N. K. (2010). Conflicted gender Medicine, 8, 783–791.
identity, body dissatisfaction, and disordered eating in adult men Roback, H. B., Strassberg, D. S., McKee, E., & Cunningham, J. (1977).
and women. Sex Roles, 63, 118–125. Self-concept and psychological adjustment differences between
American Psychiatric Association. (1980). Diagnostic and statistical self-identified male transsexuals and male homosexuals. Journal of
manual of mental disorders (3rd ed.). Washington, DC: Author. Homosexuality, 3, 15–20.

123
Arch Sex Behav (2016) 45:575–585 585

Smith, Y. L., van Goozen, S. H., Kuiper, A. J., & Cohen-Kettenis, P. T. Wolfradt, U., & Neumann, K. (2001). Depersonalization, self-esteem
(2005a). Sex reassignment: Outcomes and predictors of treatment and body image in male-to-female transsexuals compared to male
for adolescent and adult transsexuals. Psychological Medicine, 35, and female controls. Archives of Sexual Behavior, 30, 301–310.
89–99. Zucker, K. J., Cohen-Kettenis, P. T., Drescher, J., Meyer-Bahlburg, H. F.,
Smith, Y. L., van Goozen, S. H., Kuiper, A. J., & Cohen-Kettenis, P. T. Pfäfflin, F., & Womack, W. M. (2013). Memo outlining evidence for
(2005b). Transsexual subtypes: Clinical and theoretical signifi- change for gender identity disorder in the DSM-5. Archives of Sexual
cance. Psychiatry Research, 137, 151–160. Behavior, 42, 901–914.
Vocks, S., Stahn, C., Loenser, K., & Legenbauer, T. (2009). Eating and
body image disturbances in male-to-female and female-to-male
transsexuals. Archives of Sexual Behavior, 38, 364–377.

123