Prevalence of Adolescent Gender Experiences and Gender Expression in Germany

Prevalence of Adolescent Gender Experiences and Gender Expression in Germany

Journal of Adolescent Health xxx (2017) 1e8 www.jahonline.org Original article Prevalence of Adolescent Gender Experiences and Gender Expression in ...

386KB Sizes 0 Downloads 30 Views

Journal of Adolescent Health xxx (2017) 1e8

www.jahonline.org Original article

Prevalence of Adolescent Gender Experiences and Gender Expression in Germany Inga Becker, M.A. a, b, *, Ulrike Ravens-Sieberer, Ph.D. a, Veronika Ottová-Jordan, Ph.D. a, and Michael Schulte-Markwort, M.D. a a b

Department of Child and Adolescent Psychiatry, Psychosomatics, and Psychotherapy, University Medical Center Hamburg-Eppendorf, Hamburg, Germany Institute for Sex Research and Forensic Psychiatry (second affiliation), University Medical Center Hamburg-Eppendorf, Hamburg, Germany

Article history: Received September 12, 2016; Accepted February 1, 2017 Keywords: Gender identity development; Gender dysphoria; Gender variance; Gender nonconformity; Adolescence

A B S T R A C T

Purpose: Adolescence marks a transition point in the development of gender experience and expression. Although there is growing awareness about various gender identities in health research, only limited data on the prevalence of adolescent gender variance in the general population exist. Methods: German female and male adolescents (n ¼ 940) aged 10e16 years participating in the nationally representative “Health Behaviour in School-aged Children” Hamburg survey were asked to report their current gender experience (identification as both feminine and masculine) and gender expression (gender role as a girl or boy). Two overall categories and five subcategories on gender experience and expression were established based on previous research. Results: In total, 4.1% of the adolescents’ responses were rated as variant in gender experience and 3.0% as nonconforming in expression. Both variant experiences and nonconforming expression together were present in only .9% of adolescents. Gender variance was more strongly present in girls and in younger age groups. In detail, 1.6% reported an incongruent, 1.1% an ambivalent, and 1.5% no gender identification. Another 8.0% of the responses could be rated as only somewhat congruent. Conclusions: Fluidity between clearly congruent or incongruent pathways is present in adolescence, including variant as well as possibly still developing (only somewhat clear) gender experiences, whereas clearly incongruent identification and nonconforming expression were less frequent. Understanding adolescent gender development as multidimensional is important to identify the needs of those who do not fit into the current understanding of either female or male. Ó 2017 Society for Adolescent Health and Medicine. All rights reserved.

Adolescence marks an important developmental period and a transition point for girls and boys, especially with regard to the formation of gender identity [1e3]. Gender identity usually * Address correspondence to: Inga Becker, M.A., Department of Child and Adolescent Psychiatry, Psychosomatics, and Psychotherapy, University Medical Center Hamburg-Eppendorf, Martinistraße 52, W29, Hamburg 20246. E-mail address: [email protected] (I. Becker). 1054-139X/Ó 2017 Society for Adolescent Health and Medicine. All rights reserved. http://dx.doi.org/10.1016/j.jadohealth.2017.02.001

IMPLICATIONS AND CONTRIBUTION

Fluidity or variance in gender experiences is common in the normal population during adolescence, whereas clearly gender incongruent experiences are rare. Understanding diverse possibilities for development and variance in gender identity during adolescence may help promote the health of youth with gender dysphoria as well as youth with nonconforming or variant gender experiences.

refers to the experience of oneself as female or male, whereas gender expression refers to role behavior as a girl or boy [3,4]. Normative (congruent or conforming) development thus conceptualizes gender as two dichotomous categories. This binary understanding of gender is increasingly being criticized or questioned [3,5e7]. Umbrella terms like “gender variance” or “transgender” aim to capture one’s experience on a continuum of possibilities other than female and male. Both terms refer to an

2

I. Becker et al. / Journal of Adolescent Health xxx (2017) 1e8

incongruence between the natal sex and the current gender identity, meaning a not completelyeconforming (and thus nonconforming) gender identity [3,7]. However, clear incongruence is different from a (temporary) fluctuation in the experience of gender, as exploring one’s own sexual and gender identity is an ongoing developmental process when transitioning into adolescence and adulthood [3,8,9]. For some gender nonconforming children or adolescents, variant experiences or expression may be associated with feelings of clinically relevant distress resulting from the experienced incongruence (gender dysphoria [GD] in the DSM-5, American Psychiatric Association [10]). However, gender nonconformity in childhood does not necessarily imply clinical distress and also rarely persists into adolescence (and thus does not require medical interventions; e.g., [11]). Approximately, only 16% of children diagnosed with GD continue to fulfill the criteria into adolescence [11,12]. However, it is likely that there are more developmental pathways in GD adolescents [13] and that there is generally more fluidity in adolescent gender experience and expression than assumed. Increasing public awareness has led to controversial discourses regarding the diagnosis and treatment of GD and the social acceptance of transgender identity and expression [14]. However, our knowledge regarding the spectrum of gender experience and expression in the general population is sparse [7,15e17]. Most studies assessing gender identity or expression focus on clinically referred populations of individuals reporting GD [7,16,17]. The overall meta-analytical prevalence of clinically referred adolescents and adults with a GD diagnosis is increasing (4.6 in 100,000) [18] and shifts toward identification beyond the gender binary [15]. In children, a shift in the proportion of referral rates for those with the female sex at birth compared with the male has been noticed [19]. Although not all children and adolescents with a nonconforming or variant gender identity require clinical attention [3,6,20], many studies have shown that gender nonconformity or being transgender predicts lower levels of psychological well-being risks for developing psychopathology (e.g., [21e24]). Especially prejudice and discrimination may lead to negative mental health outcomes [21,22,25,26]. For youth with a diagnosed GD, medical interventions (cf. the World Professional Association for Transgender Health (WPATH) Standards of Care, Version7, for the health of transsexual, transgender, and gendernonconforming people [14]) are linked to improved psychosocial functioning in the long term [26,27]. Although numbers of GD individuals seeking professional support are increasing, there is no clear consensus about the prevalence of GD and gender variance [14,15,28,29]. Clinical samples might underestimate the number of gender variant individuals in the normal population; not all seek medical assistance or health care, and selection bias exists for clinical samples [16,17,29]. For nonconforming populations who are not presenting at specialized centers, prevalence data are lacking, especially during middle school years [9]. Research addressing the prevalence of gender variant or nonconforming youth is important to better understand their possible developmental multidimensionality [6,8,22e26,30]. Existing prevalence studies report 1.2% and 1.3% transgender identification in middle and high school samples, respectively [9,23]. In a study from New Zealand, another 2.5% of respondents reported being unsure about their gender [23]. Other studies in childhood and adolescence have shown that cross-gender behavior is more frequent than an incongruent identification, that girls present more incongruence and nonconformity than

boys, and that this kind of experience/expression is more frequent at younger ages (e.g., [24,31,32]). However, these studies have mostly used single items assessing the (self-) categorization of gender but not feelings of being feminine and/or masculine (or something else). Thus, parental reports or self-reports on behavioral or cognitive aspects of gender in young people (e.g., [9,31,32]) might overestimate the prevalence of GD or being transgender [16]. Binary concepts such as being female, male, “transgender,” or “of the other sex” leave out the question of whether gender identity in adolescence is fluid. Two recent studies have assessed feelings of gender identification in Belgium and the Netherlands [16,17]. These large population-based studies reported higher numbers of older adolescents and adults with a so-called ambivalent (3.2% of females and 4.6% of males [16]; 1.9% of females and 2.2% of males [17]) than a clearly incongruent identification (around or less than 1% in both studies). The present study thus aimed to replicate the mentioned results of Kuyper and Wijsen [16] and Van Caenegem et al. [17] for adolescents aged 10e16 years by applying similar multidimensional categories for gender experience (including incongruent and ambivalent identification), and expression (gender role behavior). The main research goal was to identify how many adolescents display what we consider a gender variant experience or nonconforming expression, how gender variance is distributed across both gender and age, and what these experiences imply in detail. Methods Study design The present study is representative for high school students from a Northern German region due to the data collection in Hamburg (a Northern German city with approximately 1,800,000 inhabitants). The Hamburg study was part of the German “Health Behaviour in School-aged Children” (HBSC) study, which belongs to a large cross-national survey conducted in collaboration with the World Health Organization Regional Office for Europe [33e36]. The international HBSC study focuses on indicators of adolescent health and of health and behavioral outcomes and includes a set of mandatory items for the collection of a common data set. In addition, each country focuses on specific topics at a national level and may assess regional representative subsamples with a focus on specific additional research questions. In the data collection for the 2013/2014 sample, single descriptive items on gender identification and gender expression were included in the Hamburg HBSC survey. The study was approved by the local ethics committee. Written informed consent was obtained from all study participants according to the institutional guidelines. For all further details regarding the study protocol, please refer to the national and international HBSC overviews [33e36]. Procedure and participants Data collection occurred at local schools in Hamburg in 2013/ 2014, and schools were selected so that the sample was representative of the general population around this age. The data collection occurred within schools, and paper-pencil questionnaires were filled out voluntarily by all participants. All participating classes received a small monetary incentive for their

I. Becker et al. / Journal of Adolescent Health xxx (2017) 1e8

contribution. The Hamburg subsample consisted of 1,250 school children from classes in all public high schools. The response rate was 23% at the school level and 69% at the individual level [34]. Data cleaning for the international sample was conducted according to the international HBSC guidelines. Complete data were available for 1,035 adolescents from Hamburg ranging in age from 10.5 to 16.3 years. Afterward, 95 participants were further excluded due to missing responses on more than one of the three gender scales. Thus, a subset of complete data sets of n ¼ 940 adolescents for gender experience and expression was available for analyses. Missing single items were replaced using the expectation-maximization method. Variables and measures Sociodemographic information was defined as described in previous analyses using the HBSC international survey data [33,36]. All participants were asked at the beginning of the questionnaire to state whether they were “a boy” or “a girl” and how old they were. Age groups were defined according to the common procedure of the HBSC study design: 11 years (with a range between 10.5 and 12.5), 13 years (between 12.6 and 14.5), and 15 years (between 14.6 and 16.3). Information on the family’s nationality was obtained by asking children about their country of birth and their parents’ ethnic background. Socioeconomic status was assessed by the Family Affluence Scale [37]. This scale includes a sum score reflecting socioeconomic status from low to high [37]. In the present study, the item on the number of computers was excluded from the total score (and the total score adjusted, accordingly) because this item does not constitute a differentiating criterion [34]. Gender experience was assessed by two single-item questions asking both girls and boys “how feminine do you experience yourself?” and “how masculine do you experience

3

yourself?” with a rating scale ranging from 1 (“very feminine/ masculine”) to 5 (“not feminine at all/not masculine at all”). Kuyper and Wijsen [16] and Van Caenegem et al. [17] operationalized individuals as “congruent,” “ambivalent,” or “incongruent” in their gender identification when asking them the degree to which they felt “like a woman/like a man.” In the present study, we coded adolescents’ answers on femininity and masculinity based on the established measures from the existing studies [16,17]. Based on the adolescent participants’ indication of femininity/masculinity and their gender, a measure was constructed which included two overall categories of gender experience: (1) “congruent” versus (2) “variant.” We further coded five identification subcategories: three of them (1A) “completely congruent,” (2A) “incongruent,” and (2B) “ambivalent” as suggested by Kuyper and Wijsen and Van Caenegem et al. [16,17]. In addition, the two congruent and variant subcategories (1B) “somewhat congruent” and (2C) “no gender identification” were included in addition to the previously established measures, because we believed that some of the answers represented more differentiating coding possibilities than those applied by previous studies. However, to provide overall prevalence data on possible “congruent” versus “variant” gender experiences, all detailed identification subcategories were merged into these two overall categories (see Table 1). 1.) Congruent gender experience included the following gender identification subcategories: (A) A completely congruent gender identification (stronger identification with one’s own gender than the other gender) is reflected by a lower (and thus stronger identification) score on the scale measuring psychological experience as one’s own gender [16,17]. (B) The somewhat congruent identification subcategory assesses the degree of a somewhat insecure or unstable but nevertheless congruent gender experience. This was reflected by a

Table 1 Coding of gender experience measures and gender identification subcategories Girls

How masculine do you experience yourself?

How feminine do you experience yourself?

Very masculine

Rather masculine

Moderately masculine

Rather not masculine

Not at all masculine

1

2

3

4

5

Ambivalent Incongruent Incongruent Incongruent Incongruent

Completely congruent Ambivalent Incongruent Incongruent Incongruent

Completely congruent Completely congruent Ambivalent Incongruent Incongruent

Completely congruent Completely congruent Somewhat congruenta No genderb No genderb

Completely congruent Completely congruent Somewhat congruenta No genderb No genderb

Not at all masculine

Very feminine Rather feminine Moderately feminine Rather not feminine Not at all feminine

1 2 3 4 5

Boys

How masculine do you experience yourself?

How feminine do you experience yourself?

Very masculine

Rather masculine

Moderately masculine

Rather not masculine

1

2

3

4

5

Ambivalent Completely congruent Completely congruent Completely congruent Completely congruent

Incongruent Ambivalent Completely congruent Completely congruent Completely congruent

Incongruent Incongruent Ambivalent Somewhat congruenta Somewhat congruenta

Incongruent Incongruent Incongruent No genderb No genderb

Incongruent Incongruent Incongruent No genderb No genderb

Very feminine Rather feminine Moderately feminine Rather not feminine Not at all feminine

1 2 3 4 5

Overall gender experience categories: (1) Congruent gender experience ¼ including subcategories of gender identification: (A) completely congruent and (B) somewhat congruent gender identification; (2) Variant gender experience ¼ including subcategories of gender identification: (A) incongruent, (B) ambivalent, and (C) no gender identification. Original (German) translation of items: “Wie weiblich/männlich fühlst du dich?”; Categories adapted from Kuyper and Wijsen, 2014; van Caenegem et al., 2015; original items: “Could you indicate to which degree you psychologically experience yourself as a woman/as a man?” and “I feel like a woman/man” (please note: coding was reversed in the two original studies). a Not included in the two previous studies. b Included in the two previous studies (interpreted as either ambivalent or incongruent).

4

I. Becker et al. / Journal of Adolescent Health xxx (2017) 1e8

moderate identification with the own gender but no opposite identification. In previous studies [16,17], this category was part of the “congruent identification” category. 2.) Variant gender experience included the following gender identification subcategories: (A) An incongruent gender identification reflects a stronger experience as the other gender than on that measuring the natal sex [16,17]. (B) An ambivalent gender identification (equal identification with the own and the other gender) reflects an equal score for the scales capturing experience as feminine and masculine. This includes a strong to moderate identification with both genders [16,17]. (C) No gender identification included equally high scores, and thus equally low identification with both genders. Because one could argue that experiencing oneself as neither feminine nor masculine should be classified as a distinct category [16,17], this subcategory was included to add to the possible gender multidimensionality. Gender expression was obtained through a single item assessing gender role behavior by asking “to what extent do others tell you that you behave like a boy/like a girl?” on a fourpoint scale ranging from “completely like a girl” to “completely like a boy.” Gender expression categories as either “(1) conforming” or “(2) nonconforming” were then divided into the following gender role subcategories: (1A) “Completely conforming gender role” (to sex; score 1 in girls, and 4 in boys). (1B) “Somewhat conforming gender role” (somewhat or predominantly conforming to sex; score 2 in girls, and 3 in boys). (2A) “Nonconforming gender role” (opposite to sex; scores 3 or 4 in girls, and 1 or 2 in boys). Statistical analyses All sample characteristics and gender experience and expression results are reported in valid percentages. For the overall prevalence, confidence intervals (95% CI) are reported. The distribution of variant gender experiences according to gender and age was compared using Fishers’ exact test. To test for

possible group differences, we calculated effect sizes using Cohen’s d and phi coefficients. Power analysis revealed that the present sample was representative for Hamburg students when divided into the two gender groups. All statistical analyses were conducted using SPSS 22.0. Results Sample characteristics Of the 940 adolescents included in the analyses, 52.4% were girls and 47.6% were boys (see Table 2). Girls and boys did not differ significantly with regard to their age at assessment (mean ¼ 13.3 years); however, there were small effects for the distribution of gender across the age groups and school grades. All other sociodemographic characteristics revealed no differences between girls and boys. Most adolescents were born in Germany; one third of the sample had a different ethnic background. The majority came from a family with a medium social background. Prevalence of gender experience and gender expression Most of the adolescents experienced themselves as congruent and conforming (see Table 3). However, 4.1% of the adolescents reported an overall variant gender experience, and 3.0% reported a nonconforming gender expression. Combining gender experience and expression, .9% of the adolescents reported both a variant experience and a nonconforming expression. Another 3.3% reported a variant experience while presenting a conforming expression, and 2.1% reported a nonconforming expression while being congruent in their experience. Distribution of variant gender experiences according to gender and age The highest percentages of variant gender experience could be found among girls at age 11 (43.6%) and girls at age 15 (25.6%;

Table 2 Sample characteristics Gender

Age in years M (SD) Range Age groups % (n) 11 (10.5e12.5) 13 (12.6e14.5) 15 (14.6e16.3) School grade % (n) Year 5 Year 7 Year 9 Nationality % (n) Born in Germany with German parents Born in Germany, parents with other nationality Other country of birth Socioeconomic status (FAS)a % (n) Low (0e2) Medium (3e4) High (5 points)

d/4

Adolescents (N ¼ 940) Girls, n ¼ 493 (52.4%)

Boys, n ¼ 447 (47.6%)

13.35 (1.70) 10.67e16.08

13.34 (1.59) 10.50e16.33

34.9 (172) 29.0 (143) 36.1 (178)

30.9 (138) 38.3 (171) 30.9 (138)

35.1 (173) 28.8 (142) 36.1 (178)

31.1 (139) 38.3 (171) 30.6 (137)

59.4 (293) 32.7 (161) 7.9 (39)

64.4 (288) 28.6 (128) 6.9 (31)

12.5 (56) 55.5 (248) 32.0 (143)

14.0 (39) 55.6 (274) 30.4 (150)

.01 .10

.10

.05

.03

FAS ¼ Family Affluence Scale; M ¼ mean; SD ¼ standard deviation. a FAS based on 3 items (number of cars in the family, own room at home, and number of family vacations in the past 12 months).

I. Becker et al. / Journal of Adolescent Health xxx (2017) 1e8 Table 3 Prevalence of overall gender experience and gender expression categories Adolescent school children (n ¼ 940) Gender experience categories (1) Congruent gender experiencea (2) Variant gender experienceb Gender expression categories (1) Conforming gender expressionc (2) Nonconforming gender expressiond Gender experience and expression categories (1) Congruent and (1) conforming (1) Congruent and (2) nonconforming (2) Variant and (1) conforming (2) Variant and (2) nonconforming

Valid % (n)

95% CI

95.9 (901) 4.1 (39)

94.5e97.1 2.9e5.5

97.0 (912) 3.0 (28)

95.7e98.1 1.9e4.3

93.7 2.1 3.3 .9

92.1e95.2 1.3e3.1 2.2e4.6 .3e1.5

(881) (20) (31) (8)

CI ¼ confidence interval. a Including subcategories of congruent and somewhat congruent gender identification. b Including subcategories of incongruent, ambivalent, and no gender identification. c Including subcategories of completely conforming and somewhat conforming gender role. d Including subcategory of nonconforming gender role.

see Table 4). In total, 74.4% of the variant gender experiences were found among girls and 25.6% among boys. Boys reported overall lower frequencies of variant gender experiences, with the highest percentage found for those aged 11 years (12.8%). Comparisons of the distribution revealed no significant effects; however, effect sizes were moderate. Prevalence of detailed gender experience and expression categories Within the variant gender experiences, 1.6% of the adolescents reported an incongruent, 1.1% an ambivalent, and 1.5% no gender identification (see Table 5). In total, 8.0% of the adolescents reported only somewhat congruent gender identification. Within the conforming gender role category, 58.3% of the adolescents reported a completely conforming gender role, whereas 38.7% reported only somewhat conforming gender role. Boys generally presented greater congruent gender experiences than girls. Discussion The results of the present study show that variant gender experience is diversely present in the adolescent population aged 10e16 years. In total, 4.1% of the adolescents’ overall responses could be classified as gender variantdor as something other than a clear congruent gender experience (e.g., incongruent, ambivalent, or no gender identification). This finding is comparable to the previously reported variant experiences from

5

the two studies assessing similar categories [16,17]. However, our present results are also in line with previous findings reporting on adolescent cross-gender or transgender identification in the general population [9,23,32]. The proportion of adolescents reporting either an incongruent identification (1.6%) or a variant gender experience while also assuming a nonconforming gender expression (.9%) was quite small (comparable to numbers reporting approximately 1% transgender prevalence). Furthermore, a variant gender experience did not always also present with a nonconforming gender expression, or vice versa, as reflected in the two different theoretical concepts [3,4], which should be considered in research. Second, girls appeared to report less gender typical experiences and more variance in the distribution of gender expression categories than boys. Approximately, two thirds of the variant experiences were present in girls, and only half of the girls reported behaving completely like the own gender. This is in line with the findings of higher nonconformity prevalence among female participants in Belgium [17] and in nonclinical child and adolescent samples (e.g. [24,31,32]) as well as comparable to numbers from clinical GD samples reporting higher referral numbers for natal males in the past and currently a shift toward the female sex [19]. However, it remains difficult to say whether such differences are related to greater societal tolerance or really a result of higher prevalence [24]. More than half of the variant experiences were reported within the youngest group (aged 11 years), which might support previous findings from studies reporting higher nonconformity or variance in younger age groups in children (e.g. [24]) or hypotheses regarding a possible relationship of gender identity development with gender, age, and pubertal onset [19]. Third, our results illustrate the diversity in gender experience and expression during adolescence. Looking more closely at the adapted categories from previous research [16,17] reveals how varied the gender experience responses may be in younger adolescent samples. Apart from incongruent, gender variant experiences could be divided into ambivalent (1.1%; equally strong identification with both genders) or no gender (1.5%; equally low identification with both genders) identification. Whereas Clark et al. [23] reported 2.5% of their sample reporting being not sure about the own gender, we categorized another 8% of the responses as only somewhat congruent, implying that they only moderately felt like their own sex but not like the opposite sex. This might imply a rather unstable or possibly still developing gender identification instead of clear congruence (or incongruence). The results of the present study add to the current lack of knowledge with regard to adolescent gender identity development, in general, and also in the debate around the diagnosis and treatment of GD [3,14,28]. In the words of Goldner [38] (p. 163), “trans may be uncommon, but gender variance itself

Table 4 Distribution of variant gender experience category according to gender and age Age categories

Age 11 Age 13 Age 15

Gender variance (n ¼ 39)

Valid % (n)

Adolescents

Group comparison

Girls (n ¼ 29)

Boys (n ¼ 10)

Test

p

4

43.6 (17) 5.1 (2) 25.6 (10)

12.8 (5) 7.7 (3) 5.1 (2)

Fisher’s exact ¼ 3.316

.163

.308

(2) Variant gender identification ¼ including subcategories of incongruent, ambivalent, and no gender identification. Valid % in reference to total gender variant experiences.

6

I. Becker et al. / Journal of Adolescent Health xxx (2017) 1e8

Table 5 Gender experience and expression categories Gender experience/identification categories

(1) Congruent gender experience category (A) Completely congruent gender identification (B) Somewhat congruent gender identification (2) Variant gender experience category (A) Incongruent gender identification (B) Ambivalent gender identification (C) No gender identification Gender expression/role categories 1) Conforming expression category (A) Completely conforming gender rolea (B) Somewhat conforming gender rolea 2) Nonconforming expression category (A) Nonconforming gender rolea

Girls (n ¼ 493)

Boys (n ¼ 447)

Valid % (n)

95% CI

Valid % (n)

95% CI

83.9 (413) 10.3 (51)

80.7e86.8 7.7e13.1

92.4 (413) 5.4 (24)

89.8e94.7 3.3e7.5

2.4 (12) 1.4 (7) 2.0 (10)

1.2e4.0 .4e2.5 1.0e3.3

.7 (3) .7 (3) .9 (4)

.0e1.5 .0e1.5 .0e1.8

Valid % (n)

95% CI

Valid % (n)

95% CI

45.2 (223) 50.7 (250)

40.7e50.0 46.1e54.9

72.7 (325) 25.5 (114)

68.7e76.6 21.7e29.7

4.1 (20)

2.4e6.1

1.8 (8)

.7e3.2

CI ¼ confidence interval. a According to sex assigned at birth.

is not rare.” [7]. Given the findings by Kuyper and Wijsen [16], who found a much lower number of possible GD treatment applicants, and the results on transgender identification [9,23], one could argue that an approximate estimation of 1% is valid for gender incongruence or transgender prevalence during adolescence. Thus, one could argue in favor of the recommendations of Cohen-Kettenis and Pfäfflin [20], that the diagnosis in adolescence requires a multidimensional approach and that variant gender identities do not necessarily need medical intervention [16]. However, the recent studies on adolescent transgender identification have also shown that estimates of approximately 1% at the middle and high school level [9,23] are higher than previous estimates for adults seeking transition-related medical interventions [18,26]. Given the growing presence of clinical referrals for nonbinary or gender queer identities [15], estimating the actual prevalence with regard to the further development of adolescents and their treatment wishes in the general population poses a challenge for research and needs further exploration. The present results support two hypotheses: on the one hand, one could argue that gender identity development is fluid instead of clear binary; and on other hand, the present results highlight the developmental process of exploring personal experiences and roles [3,7,8,30]. Therefore, it seems important to keep the developmental perspective in mind when carefully considering treatment options. At the same time, due to the increased vulnerability in gender variant youth, understanding their experiences as more complex, fluid, and multiplied may help improve overall health outcomes [8,22e26,30]. Because transgender or gender nonconforming youth are at risk of developing minority stress (a model introduced by Meyer [39]), specific programs designed to the needs of gender variant adolescents are important. The stress transgender adolescents as well as sexual and gender minorities in general experience are correlated with the development of psychopathology and often thought to result from experiences of victimization and discrimination [21,22,25,26,40,41], whereas resilience is negatively correlated with psychological distress in transgender adults [41,42]. Resilience factors such as coping and social support may buffer the effect of potential stressors and thus reduce negative outcomes [41]. Strategies for developing individual- and

group-level interventions [42] focusing on resilience or minority coping [40] may support gender variant youth on their path to adulthood while preventing further negative health outcomes. In addition or alternative to such programs and medical services for transgender youth, the WPATH Standards of Care also highlight the importance of social support through peers or families with regard to overall well-being and alleviation of GD [14]. Recent studies have shown that parental support for transgender children and adolescents is related to fewer psychopathologies [43,44]. Thus, interventions and clinical approaches that promote resilience via parental or general social support (e.g. in school or through peers and communities) may be beneficial to overall mental well-being of all adolescents that identify beyond the gender binary. However, more research is necessary to better understand the links between sexual and gender identity development and experiences of stress among sexual and gender minorities, e.g. [25]. One could argue that the number of adolescents questioning or displaying only somewhat congruence about their gender experience in the present study is also due to gender categorizations’ lack of diversity [7,30]. The lack of studies on the population size of gender variance is one methodological challenge [9,29], and the high number of only somewhat congruent gender experiences in our study might also reflect this methodological problem. However, by applying differentiating categories, the present study captured other possible experiences apart from the binary gender understanding. To better understand adolescents’ individual experiences of gender, it is critical to develop instruments that accurately, culturally sensitively, and dimensionally measure gender (experiences) [9,26,29,30], as suggested by Joel et al. [7] or the Williams Institute [45]. The present Hamburg subsample is representative for German high school students who live in larger suburban areas and thus might be limited with regard to adolescents with different urbanicity or educational backgrounds. A few other methodological limitations existed with regard to the study’s design: first, participants were asked whether they were a boy or girl. Therefore, this study lacks a more sensitive two-step measure (cf. the recommendations by the Williams Institute [45]). Questions also remain regarding the possible clinical needs

I. Becker et al. / Journal of Adolescent Health xxx (2017) 1e8

of gender variant or nonbinary youth, since we did not assess distress related to incongruence nor self-identification as transgender or non-binary. An overall rate of approximately 11% complete missing responses to the gender items might include a response bias. Other studies also included options about understanding the questions correctly (e.g., [23]). Our study also only captured those aged 10 years and older, and thus does not reflect any prevalence data for younger children. We also cannot make any assumption about how stable the reported numbers would be over the long term. The high prevalence of gender variant responses in the youngest age category might be understood as a developmental factor of gender identity, but it is also possible that the younger participants were simply not able to answer such gender questions. Thus, the relationship between gender identification and age in research should be further explored. The categories that we used based on previous research were conducted for older adolescent and adult gender identification assessment [16,17] and might thus lack of validity for our sample. The fluidity of gender identity development during adolescence creates a challenge for estimating the size of gender experience and expression populations of this age. Although the numbers from the present study might be equally malleable with regard to the long-term development of gender identity, they demonstrate that variant gender experiences and expression are common during adolescence. Future research should conduct longitudinal studies including sensitive dimensional measures to asses any changes in self-reported gender experiences to determine the permanency of early adolescent gender fluidity as well as related factors that may contribute to overall wellbeing. Understanding gender in more multiplied ways can lead to a better understanding of possible diverse developments during adolescence and may thus help improve overall health outcomes. Acknowledgments HBSC is an international survey conducted in collaboration with WHO/EURO. A complete list of the participating researchers can be found on the HBSC website (www.HBSC.org). The authors would like to thank all the schoolchildren, teachers and scientific co-workers who participated in the HBSC survey. The authors also thank Dr. Claus Barkmann and Dr. Timo Nieder for their helpful comments. Funding Sources The data collection for each HBSC survey is funded at the national level. Funding for the Hamburg study was provided by the Werner-Otto-Stiftung and Hamburg macht Kinder gesund e.V. The funding source had no involvement in the process of the study’s design, data collection, analysis, or publication. References [1] Erikson EH. Identity, youth, and crisis. New York, NY: W.W. Norton Inc.; 1968. [2] Marcia JE, Waterman AS, Matteson DR, et al., eds. Ego Identity: A Handbook for Psychosocial Research. New York, NY: Springer Verlag; 1993. [3] Steensma TD, Kreukels BP, de Vries AL, Cohen-Kettenis PT. Gender identity development in adolescence. Horm Behav 2013;64:288e97. [4] Money J. The concept of gender identity disorder in childhood and adolescence after 39 years. J Sex Marital Ther 1994;20:163e77.

7

[5] Fausto-Sterling A. Sexing the body: Gender politics and the construction of sexuality. New York, NY: Basic Books; 2000. [6] Diamond LM, Butterworth M. Questioning gender and sexual identity: Dynamic links over time. Sex Roles 2008;59:365e76. [7] Joel D, Tarrasch R, Berman Z, et al. Queering gender: Studying gender identity in ‘normative’ individuals. Psychol Sex 2014;5:291e321. [8] Institute of Medicine (IOM) Committee on Lesbian, Gay, Bisexual, and Transgender Health Issues and Research Gaps and Opportunities. The Health of Lesbian, Gay, Bisexual, and transgender People: Building a Foundation for better understanding. Washington (DC): National Academies Press; 2011. [9] Shields JP, Cohen R, Glassman JR, et al. Estimating population size and demographic characteristics of lesbian, gay, bisexual, and transgender youth in middle school. J Adolesc Health 2013;52:248e50. [10] APA (American Psychiatric Association). Diagnostic and statistical Manual of mental Disorders. 5th edition. Washington, DC: American Psychiatric Association; 2013. [11] Steensma TD, McGuire JK, Kreukels BP, et al. Factors associated with desistence and persistence of childhood gender dysphoria: A quantitative follow-up study. J Am Acad Child Adolesc Psychiatry 2013;52:582e90. [12] Ristori J, Steensma TD. Gender dysphoria in childhood. Int Rev Psychiatry 2016;28:13e20. [13] Steensma TD, Cohen-Kettenis PT. More than two developmental pathways in children with gender dysphoria? J Am Acad Child Adolesc Psychiatry 2015;54:147e8. [14] Coleman E, Bockting W, Botzer M, et al. Standards of care for the health of transsexual, transgender, and gender-nonconforming people, version 7. Int J Transgend 2012;13:165e232. [15] Richards C, Bouman WP, Seal L, et al. Non-binary or genderqueer genders. Int Rev Psychiatry 2016;28:95e102. [16] Kuyper L, Wijsen C. Gender identities and gender dysphoria in The Netherlands. Arch Sex Behav 2014;43:377e85. [17] Van Caenegem E, Wierckx K, Elaut E, et al. Prevalence of gender nonconformity in Flanders, Belgium. Arch Sex Behav 2015;44:1281e7. [18] Arcelus J, Bouman WP, Van Den Noortgate W, et al. Systematic review and meta-analysis of prevalence studies in transsexualism. Eur Psychiatry 2015;30:807e15. [19] Aitken M, Steensma TD, Blanchard R, et al. Evidence for an altered sex ratio in clinic-referred adolescents with gender dysphoria. J Sex Med 2015;12: 756e63. [20] Cohen-Kettenis PT, Pfäfflin F. The DSM diagnostic criteria for gender identity disorder in adolescents and adults. Arch Sex Behav 2010;39: 499e513. [21] Arcelus J, Claes L, Witcomb GL, et al. Risk factors for non-suicidal self-injury among trans youth. J Sex Med 2016;13:402e12. [22] Baams L, Beek T, Hille H, et al. Gender nonconformity, perceived stigmatization, and psychological well-being in Dutch sexual minority youth and young adults: A mediation analysis. Arch Sex Behav 2013;42: 765e73. [23] Clark TC, Lucassen MF, Bullen P, et al. The health and well-being of transgender high school students: Results from the New Zealand adolescent health survey (youth’12). J Adolesc Health 2014;55:93e9. [24] van Beijsterveldt CEM, Hudziak JJ, Boomsma DI. Genetic and environmental influences on cross-gender behavior and relation to behavior problems: A study of Dutch twins at ages 7 and 10 years. Arch Sex Behav 2016;35:647e58. [25] Katz-Wise SL, Rosario M, Calzo JP, et al. Endorsement and timing of sexual orientation developmental milestones among sexual minority young adults in the Growing Up Today Study. J Sex Res 2017;54:172e85. [26] Connolly MD, Zervos MJ, Barone CJ, et al. The mental health of transgender youth: Advances in understanding. J Adolesc Health 2016;59:489e95. [27] de Vries ALC, McGuire JK, Steensma TD, et al. Young adult psychological outcome after puberty suppression and gender reassignment. Pediatrics 2014;134:696e704. [28] Drescher J, Cohen-Kettenis PT, Reed GM. Gender incongruence of childhood in the ICD-11: Controversies, proposal, and rationale. Lancet Psychiatry 2016;3:297e304. [29] Olson-Kennedy J, Cohen-Kettenis PT, Kreukels BP, et al. Research priorities for gender nonconforming/transgender youth: Gender identity development and biopsychosocial outcomes. Curr Opin Endocrinol Diabetes Obes 2016;23:172e9. [30] Frohard-Dourlent H, Dobson S, Clark BA, et al. “I would have preferred more options”: Accounting for non-binary youth in health research. Nurs Inq 2017;24. [31] Achenbach TM, Edelbrock CS. Behavioral problems and competencies reported by parents of normal and disturbed children aged four through sixteen. Monogr Soc Res Child Dev 1981;46:1e82. [32] Cohen-Kettnis PT, Pfaefflin F. Transgenderism and intersexuality in childhood and adolescence making choices. Thousand Oaks, CA: Sage Publishing House; 2003.

8

I. Becker et al. / Journal of Adolescent Health xxx (2017) 1e8

[33] Currie C, Inchley J, Molcho M, et al., eds. Health Behavior in School-Aged Children (HBSC) study protocol: background, methodology and mandatory items for the 2013/14 survey. St Andrews: Child and Adolescent Health Research Unit, University of St Andrews; 2014. Available at: http:// www.hbsc.org/news/index.aspx?ni¼2418. Accessed December 15, 2016. [34] Ottová-Jordan V, Bletsch A, Ravens-Sieberer U, eds. Gesundheit und Wohlbefinden von Kindern und Jugendlichen in Hamburg. [Health and Well-Being of Children and Adolescents in Hamburg]. Lengerich: Pabst Science Publishers; 2015. [35] Ottova V, Hillebrandt D, Kolip P, et al. [The HBSC Study in Germanyestudy design and methodology]. Gesundheitswesen 2012;74:8e14. [36] Roberts C, Freeman J, Samdal O, et al. The Health Behaviour in School-Aged Children (HBSC) study: methodological developments and current tensions. Int J Public Health 2009;54(Suppl. 2):140e50. [37] Boyce W, Torsheim T, Currie C, Zambon A. The family affluence scale as a measure of national wealth: validation of an adolescent self-report measure. Soc Indic Res 2006;78:473e87. [38] Goldner V. Trans: Gender in free fall. Psychoanal Dial 2011;21:159e71.

[39] Meyer IH. Minority stress and mental health in gay men. J Health Soc Behav 1995;36:38e56. [40] Meyer IH. Prejudice, social stress, and mental health in lesbian, gay, and bisexual populations: conceptual issues and research evidence. Psychol Bull 2003;129:674e97. [41] Meyer IH. Resilience in the study of minority stress and health of sexual and gender minorities. PSOGD 2015;2:209e13. [42] Breslow AS, Brewster ME, Velez BL, et al. Resilience and collective action: Exploring buffers against minority stress for transgender individuals. PSOGD 2015;2:253e65. [43] Simons L, Schrager SM, Clark LF, et al. Parental support and mental health among transgender adolescents. J Adolesc Health 2013;53:791e3. [44] Olson KR, Durwood L, DeMeules M, McLaughlin KA. Mental health of transgender children who are supported in their identities. Pediatrics 2016;137:e20153223. [45] Herman JL, ed. Best practices for asking questions to identify transgender and other gender minority respondents on population-based surveys. Los Angeles, CA: The Williams Institute; 2014.