Trajectories of Gender Identity and Depressive Symptoms in Youths

This cohort study explores the frequency of self-reported changes in gender identity among youths and whether trajectories of gender identity were associated with depressive symptoms.


Introduction
Gender identity refers to one's inner sense of being a man, a woman, or something else. 1 Transgender and gender diverse (TGD) people are those for whom gender identity does not align with societal expectations based on their sex assigned at birth.Recent studies 2,3 show that 1.4% to 1.8% of US youths identify as TGD.
For most people, gender identity is a stable aspect of the self 4,5 ; development or change in gender identity over time is less understood.Several studies [6][7][8][9] have explored whether children who met criteria for gender identity disorder (GID; an obsolete diagnostic criteria used in the Diagnostic and Statistical Manual of Mental Disorders [Fourth Edition]) in childhood still meet GID criteria in adolescence or adulthood.Approximately 85% of the children in these studies came to identify as a sexual minority without GID in adolescence or adulthood. 10These studies refer to persistence and desistance of gender incongruence.However, interpretation from these studies are limited because a substantial proportion of study participants were subthreshold for GID diagnosis in childhood, 11 and using such outdated diagnostic criteria is problematic because one could meet GID in childhood criteria without necessarily identifying as TGD. 12 Nonetheless, recent work 13,14 indicates that a large majority of socially transitioned TGD children still identified as TGD 2.5 to 5 years later.Similarly, a UK study 15 found that among TGD youths, 91.7% continuously identified as TGD while being followed up by a gender clinic for minors.
For some, however, understanding of one's gender may vary over time. 16Some scholars refer to this variability as dynamic gender presentations, 16 gender journeys, 17 retransition, or detransition, 18 although there is no consensus on definitions of detransition. 19Of note, Olson et al 13 found a 7.3% retransition rate among socially transitioned TGD children, indicating retransitions are not common in this group.9][20][21] A recent longitudinal study 22 using a national probability sample found that approximately 1% of the US population reported different gender identities over a period of 4 years, a pattern more prevalent in younger than older generations.Importantly, among TGD people who reported past detransitions, external factors such as social rejection are often reported as reasons associated with detransitioning. 18Little is known about gender identity variability among adolescents and young adults.
Recently, a controversial theory of rapid onset gender dysphoria (ROGD) was developed based on a survey of parents of TGD youths who transitioned in adolescence. 23In the ROGD framework, gender dysphoria experienced by adolescents and young adults without prior indicators would be due to social contagion and compromised mental health and would disproportionately affect youths assigned female at birth.These youths would also experience a decline in mental health and functioning after transitioning. 23The original publication has been criticized because of biased sampling and misleading interpretation [24][25][26] ; furthermore, a correction of the original work by Littman et al 23 emphasized that ROGD is not a formal diagnosis. 27Recent cross-sectional studies [28][29][30] do not support the ROGD hypothesis.Longitudinal, community-based studies are needed to investigate whether people who transition to a TGD identity in adolescence report compromised mental health before gender identity transition or worsening mental health after.The only study, 22 to our knowledge, which explores how gender identity variability may affect health found that individuals whose gender identity varied over time engaged in more health-related risk behavior.
The current study assesses trajectories of gender identity in youths from a community-based sample, aged 15 to 21 years, over 4 assessment points, examining variability or changes in gender identity over time.Given that TGD youths often seek hormone therapy to treat gender dysphoria 31,32 and because exposure to lesbian, gay, bisexual, and transgender (LGBT) violence is associated with the mental health of TGD youths, 33,34 our analyses account for hormone and puberty blocker use and exposure to LGBT violence.Analyses investigated whether depressive symptoms varied across distinct trajectories of gender identity among youths, and associations of frequency of gender identity variability with depressive symptoms over time.

Participants and Procedures
This cohort study was approved by the institutional review boards of New York University and the University of Arizona and followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.Data come from a community-based longitudinal study of sexual and/or gender minority (SGM) youths between the ages of 15 and 21 years at baseline ( 4waves of data collection, every 9 months [2012-2015]). 35See the flowchart of participants included in the study across waves in the eFigure in Supplement 1. Community leaders recruited youths who identified as SGM from community-based agencies and college groups for SGM youths.Recruitment also occurred through referrals from other participants.Data were collected in 2 large cities in the US (1 in the Northeast and 1 in the Southwest) (eAppendix 1 in Supplement 1).Parental consent was waived for participants younger than 18 years to assure safety for youths who were not out to their parents; an independent representative was present to ensure youths participants' assent.Those older than 18 years provided written informed consent.Participants received financial compensation.To capture gender identity variability, we focused on youths who participated in at least 3 waves of the study.At each wave, participants were coded as TGD when their gender identity did not match their sex assigned at birth.
A variable indicating the gender identity variability compared with the previous wave was generated for waves 2 to 4. Frequency of gender identity variability was measured as the number of times participants' gender identity changed across the 4 waves of the study (from 0 to 3).

Depressive Symptoms
Depressive symptoms were measured utilizing the Beck Depression Inventory for Youth (BDI-Y), which assesses negative thoughts, sadness, and depressive symptoms. 37A sum score was calculated, with higher scores indicating more depressive symptomatology (average internal consistency across waves of α = .94).Sum scores of 13 or less are considered normal; scores of 14 or greater may indicate mild to severe depressive symptoms.

Covariates
Based on prior research, demographic characteristics described below were included in the adjusted models.[40][41][42]

Demographic Characteristics
Time invariant demographic variables were collected at wave 1 and included age at baseline, sex assigned at birth (0 = male; 1 = female), receipt of free lunch in high school (0 = no; 1 = yes; 2 = not reported), and race and ethnicity (0 = non-Latino White; 1 = non-Latino Black; 2 = Latino; 3 = other race or ethnicity or did not report), and recruitment site (0 = Southwest; 1 = Northeast).The other race and ethnicity included Asian, Pacific Islander, American Indian or Alaska Native, and multiracial individuals, and was created due to small sample size of each of the individual categories.Race and ethnicity were included in the study to account for health disparities that may be associated with social determinants and societal marginalization.See Appendix 2 in Supplement 1 for detailed information on the measures utilized for this study.

Explanatory Variables
Given that youths identifying as TGD tend to report more exposure to LGBT violence 43 and often seek puberty blockers and hormone therapy to reduce gender dysphoria 34,44 and that these factors are known to be associated with mental health, 31,33,34 these measures were included in final models as possible explanatory variables between transition and depression.Alternatively, exposure to LGBT violence could be a confounding variable between gender identity changes and depressive symptoms; prior work has suggested that TGD individuals may detransition as a response to stigma, 18 and exposure to LGBT violence is also associated with more depressive symptoms. 33

Hormone Therapy and Puberty Blocker Use
At waves 2 to 4, participants reported history of hormone therapy and puberty blocker use.

Cumulative Exposure to Violence Due to LGBT Identity
Exposure to violence due to LGBT identity was assessed using a 6-item scale in which participants reported how often they had experienced different forms of LGBT violence (0 = never; 3 = at least 3 times). 45At wave 1, participants were asked to consider these experiences in their lifetime.In subsequent waves, participants were asked to consider only the past 9 months.At each wave, a mean score was computed.To obtain a cumulative score, the sum score was calculated by adding the scores from the previous waves to each measure of exposure to LGBT violence across time.

Statistical Analysis
Data were managed and analyzed using Stata 18.0 (StataCorp).First, we conducted analyses of variance to test group differences in frequency of gender identity variability.Bonferroni adjustments were applied to adjust for multiple group comparisons (significance for these analyses were set at P < .008).Hierarchical level modeling (HLM) was used to analyze trajectories of depressive symptoms.To estimate within-and between-person effects, time-varying factors (ie, cumulative exposure to LGBT violence and frequency of gender identity variability) were decomposed into 2 components; between-person (BP) components (level 2) are assessed by the person mean across waves; and within-person (WP) components (level 1) are assessed by the individual deviation from their own mean across waves. 46Thus, the BP component contrasted depressive symptoms of youths who had more gender identity variability with other youths who had less or no gender identity variability, whereas the WP component contrasted depressive symptoms when a participant had more gender identity variability with other periods in which the same participant had less or no gender identity variability.
An empty mean model (ie, without factors included) was tested to estimate the degree to which depressive symptoms variation was associated with BP factors (intraclass correlation [ICC]).
Unadjusted and adjusted models examined whether the trajectory of depressive symptoms varied across gender identity trajectory groups (model 1 and model 2, respectively).Based on model 2, model 3 investigated whether the frequency of gender identity variability was associated with depressive symptoms in both BP and WP levels, while also accounting for hormone and puberty blocker use and cumulative exposure to LGBT violence.Significance testing for HLM analyses were set at a 2-tailed P < .05.Given that missingness in key variables was substantively low (<3%), we addressed missing values with listwise deletion (at the waves participants had missing values) due to the low impact in the sample size (see the eTable in Supplement 1 for sensitivity analysis).HLM uses a mixed-effects model that works with all data available in longitudinal analyses.Data analysis was conducted from September 2022 to June 2023.

Longitudinal Analyses
The ICC results showed 58.5% of the variance of depressive symptoms was at the BP level, and 41.5% was at the WP level (ICC coefficient, 0.585; 95% CI, 0.536 to 0.633).HLM unconditional models indicated that depressive symptoms had a linear trajectory accounting for random effects.HLM analyses assessing trajectories of depressive symptoms show that at baseline (intercepts) in the unadjusted model (model 1), participants from the TGD and the cisgender to TGD groups reported LGBT, lesbian, gay, bisexual, and transgender; NA, not applicable; TGD, transgender and gender diverse.
a Slope of depressive symptoms for the cisgender group.
indicated that the cisgender to TGD group also reported more depressive symptoms at baseline when compared with the TGD to cisgender group, but this finding was not significant (Β = 6.02;SE = 2.30; P = .05).There were no gender identity trajectory group differences in the rate of change of depressive symptoms over time (slopes), and there were no differences based on post hoc group comparisons.Importantly, the baseline difference in depressive symptoms between cisgender to TGD group and cisgender group was not significant after accounting for exposure to LGBT violence (model 3, Β = 3.31; SE = 2.36; P = .16).
Lastly, tests of the association of gender identity variability with depressive symptoms over time (model 3) showed that, at the BP level, patterns of depressive symptoms did not differ for youths who reported more variability in gender identity compared with their counterparts who reported less or no gender identity variability (Β = 2.43; SE = 2.51; P = .33).Furthermore, WP estimates indicated that youths did not report more depressive symptoms at times when they reported more gender identity variability compared with other periods when they reported less or no gender identity variability (Β = 0.23; SE = 0.74; P = .75).

Discussion
Gender identity is complex, and typically characterized as stable, or as a movement from cisgender to TGD (and for some, back to cisgender).Even longitudinal studies typically ask participants about their gender identity only once, overlooking ways that understanding of gender identity may vary or change for youths over time.Although a few longitudinal studies 13,14 have examined whether gender identity varies among socially transitioned TGD children, trajectories of gender identity in samples of youths are still not well understood. 47,48In this community-based longitudinal cohort study, 1 in 5 (18.3%) youths reported a different gender identity over a period of approximately 3.5 years (the majority reported the same gender identity over time).Almost one-third of the youths who reported a different gender identity did so more than twice.These findings differ from clinical samples where the majority of TGD youths consistently identified as TGD, 15 but complement recent longitudinal work 48 revealing that while gender identity is stable for the majority of youths, shifts in gender identity are not uncommon and should not be considered abnormal.9][50][51] Importantly, while changes in gender identities can be driven by developmental gender identity exploration, 52 prior work has indicated that it can be driven by a social adaptation to stigma. 18Much scientific and public attention has focused on mental health for TGD youths; our findings show that youths who reported the most change in gender identities during the study period (TGD to cisgender) were in one of the groups with the lowest levels of depression across all waves.It is possible that positive mental health can help youths feel comfortable exploring gender identity, despite societal stigma.Notably, depressive symptoms among participants in this group were stable over time.This stability might be a result of floor effects.Also, for this group, identity variability was typically between nonbinary and cisgender identities; they may have had less nonconforming gender expressions, which perhaps relates to their lower exposure to LGBT violence relative to other TGD groups.Furthermore, gender identity variability was not associated with more depressive symptoms, either between participants (BP) or for individuals over time (WP).Thus, youths who reported more changes in their gender identities were no different in terms of their mental health compared with those with fewer changes, and longitudinally, youths were not more depressed after a shift in their gender identities.These findings are more consistent with an explanation of adolescent gender identity exploration and development, rather than arguments that gender identity changes would be associated with compromised or worsening in mental health. 11r examination of groups based on gender identity trajectories (consistently cisgender or TGD, or reported changes to or from cisgender and TGD) showed that youths who reported the most change in gender identities over time (the cisgender to TGD and TGD to cisgender groups) tended to often identify as genderqueer or nonbinary, consistent with prior studies that show that youths more often identify with nonbinary gender identities. 29,48,53It may be that youths who identify as genderqueer or nonbinary may be more comfortable with gender identity exploration.

Implications for Current Controversies
Concerns about youths who identify as TGD have been raised in the past decade, particularly due to reports of greater numbers seeking treatment, 54 in particular by youths assigned female at birth. 47,54,55Results from this study offer insight into several debates.
ROGD proponents suggest that the rising numbers of TGD-identifying youths are due to compromised mental health and social contagion. 23Results of recent studies 28,29 are not consistent with these claims.Psychological well-being and demographic characteristics of youths referred to transgender clinics have been mostly consistent for more than a decade (except for sex ratio). 55rthermore, in a cross-sectional study, 28 later transgender identity acknowledgment was not associated with more compromised mental health among TGD youths.Findings from our study are also inconsistent with the ROGD hypothesis in at least 2 ways.First, although youths whose identities changed from cisgender to TGD reported higher levels of depressive symptoms when compared with consistently cisgender youths, these differences disappeared when we accounted for exposure to LGBT violence.Of note, they also experienced more exposure to LGBT violence than youths who identified consistently as cisgender, possibly due to higher gender nonconformity. 56Second, our study indicates that youths who transitioned to TGD during the study (cisgender to TGD) experienced stable levels of depressive symptoms over time, a marked contrast with the argument that youths who identify as TGD in adolescence and early adulthood will experience worsening mental health. 23,24Yet, despite being stable, youths from the cisgender to TGD group reported sustained high levels of depressive symptoms over time.In addition to the accumulation of exposure to LGBT violence among this group (Table 2), transitioning to a more stigmatized identity may be mentally taxing because of exposure to new types of violence 57 (eg, gender-based violence), expectation of rejection from family and others, 1,45 and loss of support.

Limitations
Despite substantial strengths of this study, including its community-based, longitudinal design, there are several limitations.While solely relying on self-reported gender identity, we were unable to examine, for example, which participants met the criteria for gender dysphoria.However, scholars have pointed out the importance of having youths report their own gender, 12 including providing not only binary options. 58Furthermore, youths were not asked to explain variation in their gender identities.While transitions are part of developmental gender identity exploration for many youths, 52 prior work has shown that external pressures are often associated with detransitioning from a TGD identity. 18,59We also do not know what proportion of these TGD youths went through social transitions (ie, changed pronouns or gender presentation to align with experienced gender); TGD youths may be more targeted for exposure to LGBT violence when they present themselves as TGD or in gender-nonconforming ways.Our findings may not be representative of all youths because our sample was recruited in part from SGM-focused community organizations.It could be that youths involved in these groups have a more positive sexual and gender identity development and receive more support in general.Additionally, the data from this study are more than a decade old, and much has changed in prevalence and visibility, care and affirmation, and political debates regarding TGD youths. 60New prospective, community-based studies are needed to understand gender identity development and change, and associations with mental health for cisgender and TGD youths.

Conclusion
This cohort study documented a diversity of gender identity trajectories in adolescence and early adulthood.Changes in gender identity were not associated with depressive symptoms.Furthermore, the group of youths who reported the most change in gender identities were among those with the lowest level of depressive symptoms.These findings suggest a pattern in which gender identity exploration is a normal part of adolescent development for some youths. 48,52Acknowledgment of this by health care clinicians 22 may help alleviate anxiety related to treating TGD youths whose gender identity may change across time.Although youths who changed from a cisgender to a TGD identity reported higher levels of depressive symptoms at baseline compared with consistently cisgender youths, these differences were explained by higher exposure to LGBT violence.Health care clinicians should pay particular attention to youths transitioning to TGD identities; additional support in this process can help mitigate the adverse effects of exposure to LGBT violence from peers or family rejection.Moreover, delays in providing care can result in more stress for these youths. 61timately, more longitudinal studies are needed to understand gender identity trajectories and health.

Figure 2 .CIS
Figure 2. Trajectories of Depressive Symptoms Over Time Among Youths With Different Gender Identity Trajectories19

Table 1 .
37mographic Characteristics of the Baseline Among Different Gender Identity Trajectories Membership[32 participants]).Overall, 1 in 5 participants (18.3%) reported a different gender identity over the study period.While 20 of 32 participants (62.5%) in the TGD group reported hormone use, only 6 of 28 participants (21.4%) in the cisgender to TGD group, and 1 of 32 participants (3.1%) in the TGD to cisgender group had used hormones.Use of puberty blockers was reported by 12 of the 92 participants who identified as noncisgender (ie, binary transgender or genderqueer and nonbinary) during the study; the majority were from the TGD group.Table2displays cumulative exposure toLGBT violence scores for the 4 gender identity trajectory groups.Participants presented mild levels of depression across all waves (mean [SD], 13.61[11.02]).37Figure1displays the proportion of gender identities for each trajectory group across waves.For participants from the cisgender to TGD and TGD to cisgender groups (ie, the groups defined by gender identity change), when they identified as TGD, they most often identified as genderqueer or a Percentages by column.bPercentages by row.c Other or not reported included those who did not report any race or ethnicity and participants who were American Indian or Alaska Native, Asian, Pacific Islander, or multiracial.These participants were grouped in this category due to small numbers.cisgender