Gender Diversity Among Young Adolescents: Are Gender Attitudes,Sexual Orientation,Gender Roles,Gender Identity and Sex Assigned at Birth Associated With Liking to Become a Wo/Man and the Body Image of Young Adolescents in Belgium?

Abstract

This study aims to understand how gender diversity and gender attitudes are linked to the sexual well-being of young adolescents. Data was collected among 561 adolescents in secondary schools in Flanders and binary logistic regressions were performed. Gender diversity was conceptualized as gender identity, gender expression, sex assigned at birth, sexual orientation and personal gender attitudes. Whereas “like to become a wo/man” and body image were used as measures for sexual well-being. Age and migration were added as covariates.The data illustrates that gender diversity is present among young adolescents and that it can be linked to young adolescent’s body image and liking to become a wo/man. In addition, it shows how diverse sets of gender attitudes are differently associated with sexual well-being outcomes. The findings also suggest the internalization of patriarchal and heterosexual societal norms by the respondents. The results indicate an association between on the one hand gender diversity and gender attitudes and on the other hand adolescent sexual well-being. Furthermore, it suggests that gender transformative research and programs that aim to improve adolescent sexual health and well-being should carefully choose their gender focus and invest in improving gender equity beyond the individual level. Additional research is needed to investigate the association between gender diversity and adolescent sexual wellbeing among diverse sexual well-being conceptualizations and specific groups (gender, ethnic, and others) of young adolescents.

Keywords

young adolescents sexual well-being gender diversity gender attitudes global early adolescent study

Introduction

Gender is generally acknowledged as a determinant of health (Levy et al., 2019). It is widely recognized that gender attitudes—that is, views held by individuals regarding the roles men and women should play in society—are linked to adolescent sexual and reproductive health (ASRH) (Ho & Mussap, 2019; Levy et al., 2019; Mark,2013; UNFPA, 2013). Studies have demonstrated that more restrictive attitudes toward gender equality are associated with poorer health outcomes such as violence, unwanted pregnancies and HIV infection (Chandra-Mouli et al., 2018; Greene & Patton, 2020; Kågesten & van Reeuwijk, 2021; Marston & King, 2006; Muralidharan, 2015) and strong correlations between positive attitudes toward gender equality on the one hand and sexual pleasure and good communication between sexual partners on the other have been found (De Meyer et al., 2014; Goicolea et al., 2012; Muralidharan, 2015; Reidy et al., 2015; Tolman et al., 2003; UNFPA, 2013; World Health Organization [WHO], 2013).

However, gender does not only refer to gender attitudes. It is a complex concept that comprises different aspects (Aleshire, 2016; Bracke, 2014; Ho & Mussap, 2019). Although often used interchangeably, biological sex and gender are not synonymous (Aleshire, 2016). To educate people about gender and gender diversity the infographic of the Gender Unicorn is a useful tool (Solotke et al., 2019). The strength of this kind of tool is its history of development and adoption by the trans and gender diverse community (Killermann, 2011). The Gender Unicorn consists of five constructs: gender identity, sex assigned at birth, gender expression, physically attracted to, and emotionally attracted to (Ho & Mussap, 2019; Trans Student Educational Resources, 2015). Gender identity may fluctuate and may be congruent or incongruent with sex assigned at birth (Aleshire, 2016). Gender expression refers to the way in which an individual communicates gender within a cultural context (eg, name, behaviour, hairstyle, and clothing) (Aleshire, 2016). The Gender Unicorn provides the options women, men and other gender(s) when referring to physical and emotional attraction (Ho & Mussap, 2019; Trans Student Educational Resources, 2015).

Studies among specific gender diverse groups indicate how gender can influence health during adolescence. For example, in their review Steensma et al. (2013) conclude that “studies on persons with gender dysphoria—that is, being uncomfortable with the body or with the expected roles of the assigned gender—show that the period of adolescence, with its changing social environment and the onset of physical puberty, seems to be crucial for the development of a non-normative gender identity” (Steensma et al., 2013). Additionally, Gilbey and colleagues indicate that adolescents who are attracted to the same gender are at higher risk for mental health issues than adolescents who feel attracted to other gender(s) (Gilbey et al., 2020).

Moreover, gender norms—adopted in adolescence—reflect different health trajectories for boys and girls. As such, the understanding of the subtle links between the different gender concepts and sexual health and well-being is imperative for a healthy future (Greene & Patton, 2020; Igras et al., 2014;). This investment is still needed since young adolescents are inclined to have stereotypical gender attitudes, ASRH and well-being needs are still prevalent and adolescents with diverse sexual orientations, gender identities and gender roles tend to be marginalized as they are often considered not to comply to the gender norm of (sexually active) adolescents (De Meyer et al., 2017; Halpern, 2010; Heise et al., 2019; Igras et al., 2014; Leonardi et al., 2019; Levy et al., 2019; Liang et al., 2019).

Nevertheless, the different components of gender other than biological sex (male vs. female), are rarely taken into account in research and programmes among young adolescents (Carver et al., 2013). Hence, the specific objective of this research is to add to the understanding of how different components of gender diversity could possibly affect ASH and well-being of young adolescents in Belgium. As such we want to contribute to the understanding of how gender diversity is present among these adolescents and how all different aspects of gender are interlinked and associated with health and well-being.

Materials and Methods

Study Design

This study is conducted within the framework of the Global Early Adolescent Study, the first global study—implemented in 15 countries—which investigates how perceived gender norms are formed in early adolescents and how they influence their sexual health and well-being. The study uses mixed methods. First, an exploratory qualitative phase was done to understand the gendered transitions into adolescence and the role that gender norms play within the key relationships of adolescents (Mmari et al., 2017). The second phase consists of a quantitative longitudinal study. The development and piloting of the survey are described elsewhere (Blum et al., 2019; Moreau et al., 2019; Zimmerman et al., 2019). In each country, the complete survey was piloted and re-piloted among respectively 120 and 75 adolescents between 11 and 14 years old.

In Belgium, the research was implemented in 23 secondary schools in three provinces in Flanders (Antwerp, Eastern Flanders, and Flemish-Brabant), using a multistage sampling design. The three provinces were selected based on feasibility (language and distance). The first stage of the sampling design was on the level of the schools for which we used a combination of exhaustive and convenient sampling. Secondary schools in the three provinces were ranked based on the educational underprivileged indicator (“OKI-index”). The index is a digit between 0 (min) and 4 (max), with higher scores referring to a higher proportion of disadvantaged youth. Exhaustive sampling was used for every secondary school (with a full educational program for students from twelve until 18 years old) in all cities in the three provinces and with an OKI index between 1.2 and 4 (max OKI index). The headmasters of all the selected schools were contacted by email and telephone. Due to low participation rates, we subsequently shifted to convenience sampling. Using personal contacts, we contacted headmasters of schools with an OKI index lower than 1.2 but above the Flemish mean (0.91). Recruitment started in September 2018 and ended in June 2019. Twenty-three secondary schools in seven cities (Antwerp, Ghent, Mechelen, Oostende, Aalst, Harelbeke, Lokeren) participated. Once the school accepted to participate, they decided which classes were allowed to participate in the study. In these classes, all students were eligible to participate. After parents consented (through a digital form or on paper), the students were asked to complete an assent form on paper.

The survey was self-administered on tablets using Survey CTO software. Responding to the survey was done in groups of maximum 40 students. In each group session one or more researchers—depending on the size of the group—were present to help the students where necessary.

Ethical approval was obtained from the Johns Hopkins Bloomberg School of Public Health Institutional Review Board, the World Health Organization Ethical Review Board, and by the Bioethics Committee of Ghent University Hospital in Belgium.

Description of the Analysis, the Model and the Sample Size

The concept of the Gender Unicorn shows the differences between gender identity, gender expression, sex assigned at birth, and physical and emotional attraction (Trans Student Educational Resources, 2015). Besides these personal gender traits, the model was complemented by two gender attitude scales: the Gender Expression Scale (GEX) and the Freedom of Sexual Expression Scale (FSEX), as research has indicated how diverse gender attitudes can influence ASH. Since we assumed that the analysis could be influenced by age and migration background, we added these variables as covariates. To operationalize adolescent sexual well-being, which we consider to be an essential part of ASRH, two variables from the Global Early Adolescent Study survey were selected as outcome variables: “like to become a wo/man” and “body image”. These two variables are also referred to in the conceptual framework of adolescent sexual well-being, developed by Kågesten and Van Reeuwijk (2021). Binary logistic regressions were performed using SPSS 28.0 as the outcome variables did not fulfill to the requirements to conduct a linear regression and we wanted to investigate the relationship between gender diversity concepts and adolescent sexual well-being The sample size was calculated based on power of 80%, confidence level of 0.05, and relative risk of 2.5 between 2 subpopulations (e.g., boys and girls). We chose a rare event for the calculation (10%) and allowed the comparison to be made among a subsample of the population. Under these conditions, a sample of 450 girls and 450 boys was required.

Description of the Variables

To operationalize gender diversity we used the concept of the Gender Unicorn. The following measures were included: gender identity, gender expression, sex assigned at birth, sexual orientation (both romantic and sexual physical attraction) and personal gender attitudes. Gender identity was measured by asking boys “Some boys also feel a bit like a girl, and there are also girls that feel a bit like a boy. To what extent do you think of yourself as a real boy?” A Likert scale was offered to answer the question from 0% “a real boy” to 100% “a real boy.” The opposite question was asked for girls. Afterwards both answers were merged and recoded into a dummy variable: less or more than 50% overlap between gender identity and sex assigned at birth. Due to the lack of an existing scale to measure gender expression among adolescents at the time of data collection, the gender expression of the participants was captured by using the items of the short BEM Sex Role Inventory (BSRI-12) (Carver et al., 2013). In this scale 12 items (gentle, warm, dominant, tender, sympathetic, affectionate, with leadership abilities, possessing strong personality, act as a leader, defend own beliefs, make decisions easily and sensitive to others’ needs) are presented and adolescents need to respond on a 5-point Likert scale from 1 “disagree a lot” to 5 “agree a lot”. To have meaningful variables, we conducted a Principal Component Analysis (PCA). Sexual orientation was measured by two separate questions about romantic and physical attraction. To know to whom the respondents ever felt romantically attracted we asked the following question “Have you ever felt that you were in love with a boy or a girl?”. The respondents could indicate that they did not or did with a girl, a boy or both. Afterwards the question was recoded into the dummy variable “Ever been in love with someone of the same sex” (yes/no). Regarding physical attraction, we asked the respondents if they have ever felt attracted to boys and/or girls. The answer options ran from “100% attracted to boys” (1) to “100% attracted to girl” (5). The middle option (3) referred to “equal attraction to girls and boys”. Afterwards the answers were recoded to more (1) or less or equal (0) than 50% attracted to the opposite sex. Sex assigned at birth was measured by asking the question “Are you a boy or a girl?”. As in Belgium the option “other” did not legally exist at the time of data collection, this was not offered. For all variables, the respondents also received the answer options that they didn’t know or refused to answer.

The gender attitudes of the adolescents were represented by two gender scales that were part of the Belgian Global Early Adolescent Study survey: the GEX and the FSEX. The fist scale consisted of the following items: It is ok that boys polish their nails, It is normal that girls play soccer, Boys who wear pink are ridiculous, It is normal that boys do rope skipping, Boys should pay a lot of attention to their hair, Girls should wear skirts and dresses, Boys should not wear earrings, Boys who like dancing are strange, Boys who don’t like sports are not real boys, Girls should be interested in make-up, Girls should avoid raising their voice to be lady like, Boys should always defend themselves even if it means fighting, Girls are expected to be humble, Boys should be able to show their feelings without fear of being teased, Boys who behave like girls are considered weak, It is important for boys to show they are tough even if they are nervous inside, It is okay to tease a girl who acts like a boy, It is okay to tease a boy who acts like a girl. This scale had a high Cronbach Alpha of .905.

The FSEX consists of the following items: Boys are supposed to be attracted to girls, Girls are supposed to be attracted to boys, Boys who are attracted to other boys should be treated the same as everyone else, Girls who are attracted to other girls should be treated the same as everyone else, Girls who are attracted to other girls should not be teased, Boys who are attracted to other boys should not be teased, Parents should treat their daughter the same whether she loves a boy or a girl, Parents should treat their son the same whether he loves a boy or a girl, It is ok for a boy to have sexual intercourse with another boy, It is ok for a girl to have sexual intercourse with another girl. Also this scale had a high Cronbach Alpha of 932.

For all items in both scales, adolescents were asked to reply on a five point Likert scale from “disagree a lot” (1) to “agree a lot” (5), also the option “refuse to answer” was given. Some of the items were recoded so a higher score would mean being more inclined toward gender inequitable attitudes. The answer option “refuse to answer” was recoded as a missing value. Consequently, to obtain the mean value on the scale, a sum was made of all the items and divided by the amount of items which the scale consists of. If respondents had more than one third of missing answers, they were not taken into account in the analysis. The remaining missing answers (maximum one third) were replaced by the mean of the other items within the scale. The last predictors in the model are age and migrant background. Age was checked by the simple question “How old are you?”. To assess the migrant background of the adolescents, we asked whether the respondent’s mother was born in Belgium or not. The most common answer possibilities in Flanders were presented to the adolescents, as well as the option “other”. We subsequently recoded the variable to a binary outcome variable (yes vs. no). For the variables age and migration background, the response options “I don’t know” and “refuse to answer” were recoded as missing.

Finally, we used two measures to assess the sexual well-being of the young adolescents. The first one asked in a single question whether or not the respondent liked to become a wo/man. The second measure was a scale for body image that consisted of the following items: On the whole, I am satisfied with my body, I worry about the way my body looks, I like the way I look, I often wish my body were different, I am worried that my body is not developing normally (Cronbach Alpha = .816). For both outcomes, the respondents needed to answer on a Likert scale from 1 “disagree a lot” to 5 “agree a lot” and could also choose to refuse to answer the question. For the second variable, they could additionally choose: “I don’t know” or “I don’t understand the question”. The latter were considered as a missing answer. The variable “like to become a wo/man” was recoded into a dummy variable referring to not liking to become a wo/man or being neutral versus liking to become a wo/man. To develop the body image scale, the items were recoded so a higher score meant that the adolescents were more satisfied with their body. Subsequently, the scores of all items were summed and divided by five. For each respondent with at least four non-missing values, the missing value was replaced by the mean of the respondent’s scores on the other items in the scale. If respondents had more than 1 missing value, they were excluded from the analysis. The dummy variable refers to having a score above or equal to and below the sample median of the body image scale.

Results

Background Characteristics

In total, 1,008 adolescents participated in the study. Table 1 presents the correlations between the variables of the model.

Table 1.

Correlation Table Variables of the Model.

Variables		Sex assigned at birth	Age (mean)	Mother born in Belgium *	More than 50% overlap between gender identity and sex assigned at birth	More than 50% physically attracted to the opposite sex	Ever been in love with someone of the same sex	GEX (median)	FSEX (median)	Gender expression—feminine (mean)*	Gender expression—leader (mean)	Gender expression—being assertive (mean)
Sex assigned at birth	Pearson correlation	1	−.096*	.062	−.180**	.096*	.168**	−.360**	−.286**	.228**	−.066	−.072
	Sign (2-tailed)		.024	.145	<.001	.022	<.001	<.001	<.001	<.001	.117	.090
Age (mean)	Pearson correlation	−.096*	1	−.038	.043	.078	−.042	.119**	.100*	−.058	.003	−.006
	Sign (2-tailed)	.024		.374	.306	.063	.325	.005	.018	.173	.946	.880
Mother born in Belgium *	Pearson correlation	.062	−.038	1	−.045	.130**	.107	−.269**	−.338**	.146**	−.127**	−.084*
	Sign (2-tailed)	.145	.374		.290	.002	.012	<.001	<.001	<.001	.003	.046
More than 50% overlap between gender identity and sex assigned at birth	Pearson correlation	−.180**	.043	−.045	1	.040	−.194**	.164**	.137**	−.071	.127**	.026
	Sign (2-tailed)	<.001	.306	.290		.348	<.001	<.001	.001	.092	.003	.533
More than 50% physically attracted to the opposite sex	Pearson correlation	.096*	.078	.130**	.561	1	−.237**	−.195**	−.110**	.053	−.008	−.005
	Sign (2-tailed)	.022	.063	.002	.040		<.001	<.001	.009	.207	.842	.901
Ever been in love with someone of the same sex	Pearson correlation	.168**	−.042	.107*	−.194**	−.237**	1	−.122**	−.156**	.028	−.063	−.035
	Sign (2-tailed)	<.001	.325	.012	<.001	<.001		.004	<.001	.505	.135	.410
GEX (median)	Pearson correlation	−.360**	.119**	−.269**	.164**	−.195**	−.122**	1	.712**	−.329**	.092	.039
	Sign (2-tailed)	<.001	.005	<.001	<.001	<.001	.004		<.001	<.001	.029	.355
FSEX (median)	Pearson correlation	−.286**	.100*	−.338**	.137**	−.110**	−.156**	.712**	1	−.280**	.035	.098
	Sign (2-tailed)	<.001	.018	<.001	.001	.009	<.001	<.001		<.001	.414	.021
Gender expression —feminine (mean)*	Pearson correlation	.228**	−.058	.146**	−.071	.053	.028	−.329**	−.280**	1	−.020	.003
	Sign (2-tailed)	<.001	.173	<.001	.092	.207	.505	<.001	<.001		.642	.943
Gender expression – leader (mean)	Pearson correlation	−.066	.003	−.127**	.127**	−.008	.561	.092*	.035	−.020	1	−.019
	Sign (2-tailed)	.117	.946	.003	.003	.842	−.063	.029	.414	.642		.654
Gender expression –being assertive (mean)	Pearson correlation	−.072	−.006	−.084*	.026	−.005	−.035	.039	.98*	.003	−.019	1
	Sign (2-tailed)	.090	.880	.046	.533	.901	.410	.355	.021	.943	.654

p < .05. **p < .01.

Of all the participants, 561 had no missing values for none of the variables included in the model. The remaining 447 participants who had one or more missing values, were excluded from the analyses. Table 2 presents the socio-demographic characteristics of the respondents that were included and excluded from the analyses.

Table 2.

Description of Respondents on Gender Diversity Covariates and Outcome Variables.

	Description of included respondents (n = 561)				Description of excluded respondents (n = 447)
Variables		Mean (SD) or median (percentile 25, percentile 75)	Frequency	Valid percentage	Mean or median	Frequency	Valid percentage
Sex assigned at birth	Boy		307	54.7		254	56.8%
	Girl		254	45.3		193	43.2%
Age (mean)		13 (1)			13 (1)
Mother born in Belgium *	No		273	48.7		233	55.1
	Yes		288	51.3		190	44.9
More than 50% overlap between gender identity and sex assigned at birth	No		35	6.2		31	7.6
	Yes		526	93.8		378	92.4
More than 50% physically attracted to the opposite sex	No		68	12.1		39	16.8
	Yes		493	87.9		193	83.2
Ever been in love with someone of the same sex	No		536	95.5		357	97.0
	Yes		25	4.5		11	3.0
GEX (median)		2.27 (1.78–2.72)			2.33 (1.89–2.88)
FSEX (median)		2.40 (1.50–3.10)			2.70 (1.80–3.20)
Gender expression–feminine (mean)*		−0.06 (1.02)			0.11 (0.95)*
Gender expression – leader (mean)		0.01 (0.99)			−0.03 (1.02)
Gender expression –being assertive (mean)		0.02 (0.97)			−0.03 (1.05)
Body image_scale (median)		3.80 (3.00–4.40)			4.00 (3.00–4.40)
Body image median	Below or equal to the median		263	46.9
	Above the median		298	53.1
Like to become a woman	Not like to become a wo/man or being neutral		181	32.3		108	33.6
	Like to become a wo/man		380	67.7		213	66.4

p < .05. **p < .01.

Most missing values were related to missings in the three components of the PCA of the BSRI-12 (128 each) and in the question about physical attraction. In the later, 215 respondents indicated either not knowing (176) or not being willing (39) to answer the question. The mean age of participants that were included and excluded in the analyses was 13 years old. Slightly more boys than girls participated in the study and were also included in the analyses (54.7% vs. 45.3%). For half of the respondents, their mother was not born in Belgium. There were significantly more adolescents whose mother was born in Belgium in the group that was included in the analysis compared to the excluded group. In both groups, more than 90% of the adolescents felt their gender identity and their sex assigned at birth largely overlap. The same can be said about the respondents who were or had ever been in love with someone of the opposite sex. 87.9% of all respondents included in the analyses consider themselves predominantly attracted to the opposite sex. The median scores (max 5) for the included and excluded respondents were higher on the FSEX (respectively 2.70 and 2.42) compared to the GEX (2.33 and 2.28). In the total sample and among the included and excluded participants, boys scored significantly higher than girls on the GEX and the FSEX.

After conducting a PCA on the short BEM Sex Role Inventory (BSRI-12), the twelve items were grouped in three gender expression components: feminine (a), leader (b) and being assertive (c). The mean of the “feminine” component was significantly different between the included group of participants and the excluded group. For the outcome variable “body image”, the median for the included respondents was 3.80 (SD = 3.00). The dummy scale for the included cases (below or equal vs. above the median) contained more respondents in the latter category (298 vs. 263). In both groups, more or less 67% of the adolescents liked to become a wo/man. Tables 3 and 4 describe in more detail results related to the gender scales and the PCA of the BSRI-12.

Table 3.

Differences Between Boys and Girls on the Gender Scales.

Variables		Girls	Boys
GEX (mean)	Total sample	2.07	2.54
	Included participants	2.03	2.49
	Excluded participants	2.11	2.59
FSEX (mean)	Total sample	2.19	2.73
	Included participants	2.12	2.67
	Excluded participants	2.32	2.83

Note. In both groups (included and excluded), the score of the boys is significantly different form the score of the girls (p-values < 0.01.)

Table 4.

Description of Principal Component Analysis (n = 1,008).

		Component
Items of the short BEM sex role inventory (BSRI-12)		One feminine	Two leader	Three being assertive
Gentle		0.708	0.160	−0.133
Warm		0.656	0.301	−0.094
Dominant		−0.408	0.511	−0.374
Tender		0.614	0.290	−0.087
Sympathetic		0.485	0.397	0.092
Affectionate		0.580	0.383	−0.057
With leadership abilities		−0.328	0.685	−0.392
Possessing a strong personality		−0.263	0.570	0.470
Act as a leader		−0.414	0.684	−0.339
Defend your own beliefs		−0.226	0.473	0.585
Make decisions easily		−0.060	0.404	0.438
Sensitive to others’ needs		0.418	0.225	0.136
Initial eigenvalues	Total	2.625	2.464	1.228
	% of variance	21.876	20.534	10.230
	Cumulative %	21.876	42.410	52.640

The Logistic Analyses

After a descriptive analysis of the variables, the multicollinearity between the variables of the two models was tested. No indications for multicollinearity were found which means that the location parameters were the same across the response categories. In a next phase, two logistic regressions were performed to discover which gender variables are associated with the outcome variables “like to become a woman” and “body image”.

In the first model with the outcome variable “like to become a wo/man,” three coefficients were significant (p < .05) and eight were not. Continuous predictors were entered linearly into the model. The Hosmer and Lemeshow Test did not indicate any violation of the linearity assumption. The results of the final model are described in Table 5.

Table 5.

Logistic Regression Models of “like to become a wo/man” and “body image.”

Independent Variables	Dependent variables (n = 561)
	Like to become a wo/man	Body image
	Exp (B)	Exp (B)
Sex assigned at birth^a	.254**	.411**
Gender identity^b	1.116	2.247
Romantic attraction^c	.679	.277*
Physical attraction^d	.274	.624
Gender expression–feminine	1.211	.981
Gender expression–leader	1.424**	1.258**
Gender expression–assertive	1.228*	1.477**
GEX	1.422	.640*
FSEX	.998	1.382*
Age	1.054	.831
Migrant context^e	.770	1.155
Nagelkerke R square	.219	.175
Hosmer and Lemeshow test	p = .989	p = .590

Note.*p < .05. **p < .01.

Boys versus girls (ref.).

Less versus more (ref.) than 50% overlap between gender identity and sex assigned at birth.

Never been in love with someone of the same sex or having been in love (ref.).

Feeling less or equal than 50% attracted to the opposite sex versus feeling more than 50% attracted to the opposite sex (ref.).

The respondent’s mother was not born in Belgium versus the respondent’s mother was born in Belgium (ref.).

The regression analysis showed how sex assigned at birth (Exp B = 0.25; 0.16–0.39) and two factors of the gender expression scale—being a leader (Exp B = 1.42; 1.17–1.74) and being assertive (Exp B = 1.23; 1.01–1.50)—were significantly associated with “liking to become a wo/man.” If we consider sex assigned at birth, the results indicated that the odds of girls liking to become a wo/man were 75% lower than for boys, taking into account all the other variables in the model. In addition, in the same model, “acting as a leader” and “being assertive,” increased the odds of liking to become a wo/man with respectively, 42% and 23%. The remaining variables in the model—“age,”“feminine gender role,”“physical attraction,”“romantic attraction,”“gender identity,”“gender attitudes” (GEX and FSEX) and “migrant background”—were not significantly associated with the dependent variable.

The second model, with “body image” as a dependent variable, is also described in Table 5. Similar to the previous model, the Hosmer and Lemeshow Test did not indicate any violation of the linearity assumption. The results illustrated how romantic attraction (Exp B = 0.28; 0.09–0.85), sex assigned at birth (Exp B = 0.41; 0.28–0.61) and GEX (Exp B = 0.64; 0.41–0.99) were significantly negatively associated with body image. The other significant associations were positive: FSEX (Exp B = 1.38; 1.05–1.82) “acting as a leader” (Exp B = 1.26; 1.05–1.51) and “being assertive” (Exp B = 1.48; 1.21–1.79). The negative associations showed that adolescents who had ever been in love with someone of the same sex were less likely to have a positive body image compared to those who were never in love with someone of the same sex. In comparison to boys, the odds of girls reporting a positive body image was 60% lower. Additionally, adolescents with higher scores on the GEX (i.e., more stereotypical attitudes) were less likely to report a positive body image compared to the ones who had lower scores on this gender attitude scale. A score above the median on the body image scale was more likely for adolescents who act assertive and as a leader. In addition, the odds of having a positive body image were 38% higher for adolescents who scored higher on the FSEX. The remaining variables in the model—“age,”“gender identity,”“physical attraction,”“feminine gender expression” and “migrant background”—were not significantly associated with the outcome variable.

Discussion

To the best of our knowledge this is one of the first studies to investigate the association between, on the one hand, young adolescents’ individual traits related to gender diversity and gender attitudes and, on the other hand, their sexual health and well-being. We can draw five important conclusions based on these results, which will be discussed in the subsequent paragraphs.

First, this research illustrates that gender diversity is present among young adolescents. For example, not all respondents indicate a 100% overlap between their sex assigned at birth and their gender identity. Neither are all respondents heterosexual or do they share similar attitudes toward gender equality.

Second, the results underline the importance of taking into account different aspects of gender diversity to understand and improve adolescent sexual well-being: gender expression, sexual orientation and sex assigned at birth. Most factors were correlated to the two outcome variables. In our analyses, only gender identity and physical attraction were not significantly associated with any of the outcomes which could possibly be related to the younger age of our respondents. These results are in line with former research. A study among 569 adolescents (median age of 14 years) in Australia found that 13.2% of the respondents was unsure about their gender identity (Blacklock et al., 2021). In addition, research in the US on the role of physical attractiveness in adolescent romantic relationships’ indicated that physical attractiveness was unrelated to general relationship satisfaction or to any positive relationship experiences (Furr, 2009).

Thirdly, the results suggest the internalization of patriarchal and heterosexual societal norms by the respondents. These are reflected in the diverse components of gender diversity and are related to the well-being of the respondents. If we consider sex assigned at birth, our analyses depict that being a boy is favorable for both sexual well-being outcomes. Acting as a stereotypical boy—that is, as a leader and assertive—is also positively associated with liking to become a wo/man and body image. However, the first component of gender expression “being feminine” is not significantly associated with any of the outcomes. An additional illustration that the respondents are aware of gender stereotypical norms is related to the level of romance and the individual gender attitudes. Indeed, the results indicate a negative association between ever been in love with someone of the same sex and body image. Concerning the gender attitudes, more stereotypical gender attitudes on the FSEX lead to higher scores on body image. Both results are in line with the norm to be romantically and sexually attracted toward the opposite sex. We can assume that the respondents applied the binary division between heterosexuality and non-heterosexuality while responding to the questions. Being heterosexual or agreeing that heterosexuality should be the norm could therefore be associated with a better body image, since the respondents might feel closer to the norm compared to sexual minority groups. The presence of the patriarchal and heterosexual norms in the Flemish society has as well been documented by other researchers (Al-Attar et al., 2017; D’Haese et al., 2016; Dewaele et al., 2013; Mmari et al., 2018; Yu et al., 2017). Additionally, challenging these norms has found to be difficult and with possible negative consequences such as verbal, physical, and/or psychological retribution (D’Haese et al., 2016; Yu et al., 2017). Our research suggests similar negative associations with adolescent sexual well-being. This finding is in line with the conceptual framework for early adolescents of Blum et al. that states that national forces shape adolescent health directly and indirectly and that systematic gender inequality is an example of this (Blum et al., 2014).

Fourth, our analyses demonstrate that there is not such a thing as “thé gender attitudes of an adolescent.”The two different sets of gender attitudes (gender expression and freedom of sexual expression) are differently associated with the sexual well-being outcomes. As illustrated, gender is a broad concept which refers to diverse components such as gender identity and gender expression (Heise et al., 2019). When we measure adolescents’ gender attitudes, they can be related to all of these diverse components and the attitude toward them does not necessarily need to be the same. Our research focusses on attitudes toward issues that are relevant in the Flemish society: gender expression and freedom of sexual expression. Although both are expressions of personal identity, their association with sexual well-being is different. A higher score on GEX is correlated with a less positive body image, whereas a higher score on FSEX is correlated with a higher body image. The difference could be related to the fact that the items of GEX are grounded in personal experiences and are situated on the general behaviour level, whereas the FSEX items refer only to sexuality and to (sexual) minority groups. Former research from the Global Early Adolescent Study shows that adolescent boys and girls are aware that puberty marks a point at which girls’ and boys’ lives begin to separate and take different (health) trajectories (Mmari et al., 2018). We can hypothesize that the adolescents with higher scores on the GEX agree with societal norms and existing stereotypical expectations for women and men and are therefore less content if their body is not how it is expected to be. The FSEX items however, as mentioned before relate to sexual diversity. This difference in results between the scales suggests that it is important to carefully select the gender attitude(s scales) and the outcomes in gender transforming intervention programs that aim to improve ASRH and well-being. Moreover, our data subscribes the usefulness of these interventions as our models indicate that gender attitudes are associated with ASRH and well-being. Both of these conclusions are also reflected in the systematic review of Muralidharan et al which shows how gender-integrated programs improve health and gender outcomes across the range of health areas (Muralidharan, 2015). They recommend to “Specify the causal pathway by which gender can benefit health, and develop well-defined, gender-specific qualitative and quantitative measures to study how and to what extent gender changes health outcomes.” The same reflections are made by other researchers that focus on SRH of adolescents between 10 and 14 years. These researchers illustrate how gender transformative programs can improve a broad range of outcomes such as ASRH. To be able to do so, they stress the need for gender indicators that reflect the day-to-day realities of these young adolescents (Igras et al., 2014). As mentioned, this could also be relevant to explain the different associations between GEX and FSEX with body image.

Further to this, it is important to point out that young adolescents are often ignored in research and even more in research that focusses on sexual health and well-being and not on risks (Igras et al., 2014). A report, published in 2016 by the Population Council on “Reviewing the evidence and charting a course of research and action for very young adolescents” mentions a growing interest in the very young adolescent population. However, they also state that more data on ASRH is available for adolescents aged 15 to 19 than for (unmarried) girls and boys aged 10 to 14 (McCarthy, 2016). The research presented indicates that research on this topic among young adolescents is possible and meaningful.

Finally, concerning the socio-demographic variables, our research indicates that both sexual well-being outcomes are not significantly associated with “migrant background” nor “age.” Other researchers have found a significant decrease in body image as adolescents get older (Markland & Ingledew, 2007). Although the results in our study were not significant (p = .064) they suggest the same direction. In their research, Golan et al. (2013) mention that a sample of adolescents with mean age 13.5 (ours is 13) years old, is very young to investigate changes in body image. The irrelevance of the migrant status in both analyses could suggest that the physical key transitions in development where the young adolescents went through (e.g., pubertal maturation and brain development) at the time of the study were more important than their background.

Our sample size was relatively small to capture the broad variety of all levels of gender diversity. Conducting this research with specific groups of gender nonconforming adolescents can shed a different light on how gender attitudes contribute to sexual health trajectories of these adolescents who are in general considered to be more at risk than the overall population (Blondeel et al., 2016; Leonardi et al., 2019). Associated with this, we cannot consider our sample as a representation of the overall vulnerable urban adolescents in Flanders, as our respondents were partially recruited through convenient sampling. Concerning the sample, we also need to add that the final models were built on a smaller number of cases than on our overall sample of 1,008 respondents. This was mainly due to the missing values in the three components of the PCA and the lack of knowledge among the respondents about whom they are physically attracted to. This could possibly be related to the young age of the respondents and their lack of experience with it.

We used scales that we developed and validated or that have been applied and validated in earlier research, though mostly among older age groups. During the process of data collection we noticed that a significant amount of our respondents needed help to interpret certain questions. This could mean that some of the respondents (who did not dare to ask for help) have misunderstood some questions and as such our data could be biased. Besides the fact that the questions should be adapted to age we could also argue that a cultural adaptation may have been necessary. Following de Graaf, we support the idea of cross-cultural variations in the conceptualizations of gender, gender expression and gender diversity. Ethnicity, culture and religion, and their overlapping relationship with gender need further exploration (de Graaf & Carmichael, 2019). Therefore, we plea for the development of reliable scales that can measure gender diversity in a non-clinical context among a variety of (young) adolescents. Due to the lack of existence of these scales, we used in our research the BSRI-12 to measure gender expression. However, we are aware that gender expression is a broader concept than gender roles. In future research, the entire spectrum of gender expressions should be taken into account.

Additionally, we only used two variables to measure adolescent sexual well-being, both at the individual level. In other research that focusses on sexual well-being of adolescents, additional variables at individual and interpersonal level are used or suggested such as communication between sexual partners, focusing on sexual satisfaction of oneself and the partner and freedom from pain (Anderson, 2013; De Meyer et al., 2014; Harden, 2014; Michielsen et al., 2016). Testing our models with these factors, could provide additional information for ASRH programs. The same comment could be made for using additional scales that measure a variety of gender attitudes. And finally, inherent to cross-sectional studies, our research did not allow to formulate causal relationships between the predictors and the two outcomes of adolescent sexual well-being.

Conclusion

Our research indicates that gender diversity is present among young adolescents and that this diversity and gender attitudes of young adolescents are associated with ASRH and well-being. Additionally, it suggests that gender transformative research and programs that aim to improve ASRH and well-being should carefully choose their gender focus and outcomes. They should take into account gender diversity and invest in improving gender equity beyond the individual level. More research is needed to see if this also counts for diverse conceptualizations of sexual well-being and health and how this trajectory works for specific groups of diverse (gender, ethnic, and others) young adolescents.

Footnotes

Acknowledgements

This work was undertaken as part of the Global Early Adolescent Study, a study lead by Johns Hopkins Bloomberg School of Public Health in collaboration with the WHO and UNFPA and 15 global institutions.

Author Contribution

All authors discussed the objectives and set-up of the article. S.D.M. was the field coordinator in charge of data collection, conducted the final analysis and elaborated a first and final version of the article. Both co-authors contributed to the analysis and writing process. All the manuscript’s authors have seen and approved the final submitted version.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: It was funded by the David and Lucile Packard Foundation, the Bill and Melinda Gates Foundation, the United States Agency for International Development (USAID), the World Health Organization, the Fund for scientific research Flanders and the Flemish Ministry of Innovation, Public Investment, Media and Poverty Reduction.

Ethical Approval

Ethical approval was obtained from the Johns Hopkins Bloomberg School of Public Health Institutional Review Board, approval number 00005684 and by the Bioethics Committee of Ghent University Hospital in Belgium, approval number 2015/0473.

ORCID iD

Sara De Meyer

Data Availability Statement

Data sharing not applicable to this article as no datasets were generated or analyzed during the current study.

References

Al-Attar

De Meyer

El-Gibaly

Michielsen

Animosa

L. H.

Mmari

(2017). “A Boy would be friends with boys … and a girl … With girls”: Gender norms in early adolescent friendships in Egypt and Belgium. Journal of Adolescent Health, 61(4s), S30–S34. https://doi.org/10.1016/j.jadohealth.2017.03.013

Aleshire

M. E.

(2016). Sexual orientation, gender identity, and gender expression: What are they? JNP-Journal for Nurse Practitioners, 12(7), E329–E330. https://doi.org/10.1016/j.nurpra.2016.03.016

Anderson

R. M.

(2013). Positive sexuality and its impact on overall well-being. Bundesgesundheitsblatt, Gesundheitsforschung, Gesundheitsschutz, 56(2), 208–214. https://doi.org/10.1007/s00103-012-1607-z

Blacklock

McGlasson

Chew

Murfitt

Hoq

(2021). Challenges in measuring gender identity among transgender, gender diverse, and non-binary young people. Public Health, 200, E4–E5. https://doi.org/10.1016/j.puhe.2021.09.012

Blondeel

Say

Chou

Toskin

Khosla

Scolaro

Temmerman

(2016). Evidence and knowledge gaps on the disease burden in sexual and gender minorities: A review of systematic reviews. International Journal for Equity in Health, 15, 16. https://doi.org/10.1186/s12939-016-0304-1

Blum

R. W.

Astone

N. M.

Decker

M. R.

Mouli

V. C.

(2014). A conceptual framework for early adolescence: A platform for research. International Journal of Adolescent Medicine and Health, 26(3), 321–331. https://doi.org/10.1515/ijamh-2013-0327

Blum

R. W.

Sheehy

Basu

El Gibaly

Kayembe

Zuo

Ortiz

Chan

K. S.

Moreau

(2019). Measuring young adolescent perceptions of relationships: A vignette-based approach to exploring gender equality. PLOS One, 14(6), e0218863. https://doi.org/10.1371/journal.pone.0218863

Bracke

(2014). The unbearable lightness of ‘gender and diversity’. DiGeSt. Journal of Diversity and Gender Studies, 1(1), 41–50. https://doi.org/10.11116/jdivegendstud.1.1.0041

Carver

L. F.

Vafaei

Guerra

Freire

Phillips

S. P.

(2013). Gender differences: examination of the 12-item bem sex role inventory (BSRI-12) in an older Brazilian population. Plos One, 8(10), e76356. https://doi.org/10.1371/journal.pone.0076356

10.

Chandra-Mouli

Plesons

Amin

(2018). Addressing harmful and unequal gender norms in early adolescence [Editorial Material]. Nature Human Behaviour, 2(4), 239–240. https://doi.org/10.1038/s41562-018-0318-3

11.

D’Haese

Dewaele

Van Houtte

(2016). The relationship between childhood gender nonconformity and experiencing diverse types of homophobic violence. Journal of Interpersonal Violence, 31(9), 1634–1660. https://doi.org/10.1177/0886260515569063

12.

de Graaf

N. M.

Carmichael

(2019). Reflections on emerging trends in clinical work with gender diverse children and adolescents. Clinical Child Psychology and Psychiatry, 24(2), 353–364. https://doi.org/10.1177/1359104518812924

13.

De Meyer

Jaruseviciene

Zaborskis

Decat

Vega

Cordova

Temmerman

Degomme

Michielsen

(2014). A cross-sectional study on attitudes toward gender equality, sexual behavior, positive sexual experiences, and communication about sex among sexually active and non-sexually active adolescents in Bolivia and Ecuador. Global Health Action, 7, 24089. https://doi.org/10.3402/gha.v7.24089

14.

De Meyer

Kågesten

Mmari

McEachran

Chilet-Rosell

Kabiru

C. W.

Maina

Jerves

E. M.

Currie

Michielsen

(2017). “Boys should have the courage to ask a girl out”: Gender norms in early adolescent romantic relationships. Journal of Adolescent Health, 61(4s), S42–S47. https://doi.org/10.1016/j.jadohealth.2017.03.007

15.

Dewaele

Van Houtte

Cox

Vincke

(2013). From coming out to visibility management—A new perspective on coping with minority stressors in LGB youth in Flanders. Journal of Homosexuality, 60(5), 685–710. https://doi.org/10.1080/00918369.2013.773818

16.

Furr

W. R.

(2009). The role of physical attractiveness in adolescent romantic relationships [Unpublished doctoral dissertation]. University of Tennessee.

17.

Gilbey

Mahfouda

Ohan

Lin

A. S.

Perry

(2020). Trajectories of mental health difficulties in young people who are attracted to the same gender: A systematic review. Adolescent Research Review, 5(3), 281–293. https://doi.org/10.1007/s40894-019-00128-8

18.

Goicolea

I. T. M.

Edin

Ohman

(2012). When sex is hardly about mutual pleasure: Dominant and resistant discourses on sexuality and its consequences for young people’s sexual health. International Journal of Sexual Health, 24(4), 303–317.

19.

Golan

Hagay

Tamir

(2013). The effect of “in favor of myself’’: Preventive program to enhance positive self and body image among adolescents. PLoS One, 8(11), 9. https://doi.org/10.1371/journal.pone.0078223

20.

Greene

M. E.

Patton

(2020). Adolescence and gender equality in health [Editorial Material]. Journal of Adolescent Health, 66(1), S1–S2. https://doi.org/10.1016/j.jadohealth.2019.10.012

21.

Halpern

C. T.

(2010). Reframing research on adolescent sexuality: Healthy sexual development as part of the life course. Perspectives on Sexual and Reproductive Health, 42(1), 6–7. https://doi.org/10.1363/4200610

22.

Harden

K. P.

(2014). A sex-positive framework for research on adolescent sexuality. Perspectives on Psychological Science A Journal of the Association for Psychological Science, 9(5), 455–469. https://doi.org/10.1177/1745691614535934

23.

Heise

Greene

M. E.

Opper

Stavropoulou

Harper

Nascimento

Zewdie

(2019). Gender inequality and restrictive gender norms: Framing the challenges to health. Lancet, 393(10189), 2440–2454. https://doi.org/10.1016/s0140-6736(19)30652-x

24.

Mussap

A. J.

(2019). The gender identity scale: Adapting the gender unicorn to measure gender identity. Psychology of Sexual Orientation and Gender Diversity, 6(2), 217–231. https://doi.org/10.1037/sgd0000322

25.

Igras

S. M.

Macieira

Murphy

Lundgren

(2014). Investing in very young adolescents’ sexual and reproductive health. Global Public Health, 9(5), 555–569. https://doi.org/10.1080/17441692.2014.908230

26.

Kågesten

van Reeuwijk

(2021). Adolescent sexual wellbeing: A conceptual framework. SocArXiv. https://doi:10.31235/osf.io/9as6e

27.

Kar

S. K.

Choudhury

Singh

A. P.

(2015). Understanding normal development of adolescent sexuality: A bumpy ride. Journal of Human Reproductive Sciences, 8(2), 70–74. https://doi.org/10.4103/0974-1208.158594

28.

Killermann

(2011). The gingerbread person. https://www.itspronouncedmetrosexual.com/2011/11/breaking-through-the-binary-gender-explained-using-continuums/

29.

Leonardi

Frecker

Scheim

A. I.

Kives

(2019). Reproductive health considerations in sexual and/or gender minority adolescents. Journal of Pediatric and Adolescent Gynecology, 32(1), 15–20. https://doi.org/10.1016/j.jpag.2018.09.010

30.

Levy

J. K.

Darmstadt

G. L.

Ashby

Quandt

Halsey

Nagar

Greene

M. E.

(2019). Characteristics of successful programmes targeting gender inequality and restrictive gender norms for the health and wellbeing of children, adolescents, and young adults: A systematic review. The Lancet Global Health, 8, e225–e236. https://doi.org/10.1016/s2214-109x(19)30495-4

31.

Liang

Simelane

Fortuny Fillo

Chalasani

Weny

Salazar Canelos

Jenkins

Moller

A. B.

Chandra-Mouli

Say

Michielsen

Engel

D. M. C.

Snow

(2019). The state of adolescent sexual and reproductive health. Journal of Adolescent Health, 65(6s), S3–S15. https://doi.org/10.1016/j.jadohealth.2019.09.015

32.

Mark

(2013). Adolescence: A second chance to tackle inequities. Lancet, 382(9904), 1535–1535. https://doi.org/10.1016/S0140-6736(13)62308-9

33.

Markland

Ingledew

D. K.

(2007). The relationships between body mass and body image and relative autonomy for exercise among adolescent males and females. Psychology of Sport and Exercise, 8(5), 836–853. https://doi.org/10.1016/j.psychsport.2006.11.002

34.

Marston

King

(2006). Factors that shape young people’s sexual behaviour: A systematic review. Lancet, 368(9547), 1581–1586. https://doi.org/10.1016/s0140-6736(06)69662-1

35.

McCarthy

Brady

Hallman

(2016). Investing when it counts: Reviewing the evidence and charting a course of research and action for very young adolescents. Population Council. https://www.popcouncil.org/uploads/pdfs/2016PGY_InvestingWhenItCounts.pdf

36.

Michielsen

De Meyer

Ivanova

Anderson

Decat

Herbiet

Kabiru

C. W.

Ketting

Lees

Moreau

Tolman

D. L.

Vanwesenbeeck

Vega

Verhetsel

Chandra-Mouli

(2016). Reorienting adolescent sexual and reproductive health research: Reflections from an international conference. Reproductive Health, 13, 3. https://doi.org/10.1186/s12978-016-0117-0

37.

Mmari

Blum

R. W.

Atnafou

Chilet

de Meyer

El-Gibaly

Basu

Bello

Maina

Zuo

(2017). Exploration of gender norms and socialization among early adolescents: The use of qualitative methods for the global early adolescent study. Journal of Adolescent Health, 61(4s), S12–S18. https://doi.org/10.1016/j.jadohealth.2017.07.006

38.

Mmari

Moreau

Gibbs

S. E.

De Meyer

Michielsen

Kabiru

C. W.

Bello

Fatusi

Lou

Zuo

Al-Attar

G. S. T.

El-Gibaly

(2018). ‘Yeah, I’ve grown; I can’t go out anymore’: Differences in perceived risks between girls and boys entering adolescence. Culture, Health & Sexuality, 20(7), 787–798. https://doi.org/10.1080/13691058.2017.1382718

39.

Moreau

De Meyer

Vu Manh

Guiella

Acharya

Bello

Maina

Mmari

(2019). Measuring gender norms about relationships in early adolescence: Results from the global early adolescent study. SSM-Population Health, 7, 014. https://doi.org/10.1016/j.ssmph.2018.10.014

40.

Muralidharan

Fehringer

Pappa

Rottach

Das

Mandal

(2015). Transforming gender norms, roles, and power dynamics for better health: Evidence from a systematic review of gender-integrated health programs in low- and middle-income countries. Futures Group, Health Policy Project. United States Agency for International Development USAID. https://www.healthpolicyproject.com/pubs/381_GPMIndiaSummaryReport.pdf

41.

Reidy

D. E.

Smith-Darden

J. P.

Cortina

K. S.

Kernsmith

R. M.

Kernsmith

P. D.

(2015). Masculine discrepancy stress, teen dating violence, and sexual violence perpetration among adolescent boys. Journal of Adolescent Health, 56(6), 619–624. https://doi.org/10.1016/j.jadohealth.2015.02.009

42.

Solotke

Sitkin

N. A.

Schwartz

M. L.

Encandela

J. A.

(2019). Twelve tips for incorporating and teaching sexual and gender minority health in medical school curricula. Medical Teacher, 41(2), 141–146. https://doi.org/10.1080/0142159x.2017.1407867

43.

Steensma

T. D.

Kreukels

B. P.

de Vries

A. L.

Cohen-Kettenis

P. T.

(2013). Gender identity development in adolescence. Hormones and Behavior, 64(2), 288–297. https://doi.org/10.1016/j.yhbeh.2013.02.020

44.

Tolman

D. L.

Striepe

M. I.

Harmon

(2003). Gender matters: Constructing a model of adolescent sexual health. Journal of Sex Research, 40(1), 4–12. https://doi.org/10.1080/00224490309552162

45.

Trans Student Educational Resources. (2015). The gender unicorn. Trans Student Educational Resources. http://www.transstudent.org/gender

46.

United Nations Population Fund (UNFPA). (2013). The state of world population 2013: Motherhood in childhood: Facing the challenge of adolescent pregnancy. UNFPA. https://www.unfpa.org/publications/state-world-population-2013

47.

World Health Organization (WHO). (2013). HIV and adolescents: Guidance for HIV testing and counselling and care for adolescents living with HIV: Recommendations for a public health approach and considerations for policy-makers and managers. World Health Organization. https://apps.who.int/iris/handle/10665/94334

48.

Zuo

Blum

R. W.

Tolman

D. L.

Kagesten

Mmari

De Meyer

Michielsen

Basu

Acharya

Lian

Lou

(2017). Marching to a different drummer: A cross-cultural comparison of young adolescents who challenge gender norms. Journal of Adolescent Health, 61(4s), S48–S54. https://doi.org/10.1016/j.jadohealth.2017.07.005

49.

Zimmerman

L. A.

Moreau

Wilopo

Blum

(2019). Measuring agency as a dimension of empowerment among young adolescents globally; findings from the Global Early Adolescent Study. SSM—Population Health, 8, 100454. https://doi.org/10.1016/j.ssmph.2019.100454