Understanding harms associated with gender identity conversion efforts among transgender and nonbinary individuals: The role of preexisting mental well-being

Abstract

Background: Affirmation of gender identity is critical for the mental health and overall well-being of transgender and nonbinary (TNB) persons. Gender identity conversion efforts (GICE), an outlawed practice for licensed professionals in numerous U.S. jurisdictions, have been associated with negative mental health and substance use outcomes. Limited previous literature examining GICE exposure has been criticized for failing to distinguish mental well-being for TNB persons before or after GICE. Our study builds on current literature by examining differences in TNB persons’ psychosocial risk indicators based on their GICE exposure, accounting for pre-GICE mental well-being.Methods: We conducted a secondary data analysis using the 2015 U.S. Transgender Survey (N = 25,810), the largest available national survey aimed at understanding TNB persons’ experiences. Using logistic regression models, we examined how GICE exposure (disaggregated by temporal precedence of initial suicide attempts) is related to health (psychological distress, extra-medical prescription use, healthcare avoidance), socio-structural (public restroom avoidance, housing instability), and interpersonal outcomes (sexual assault, emotional and physical intimate partner violence (IPV)).Results and conclusions: Initiating a suicide attempt post-GICE or in the absence of GICE were the only consistently significant predictors of poor outcomes across all domains compared to participants who never experienced GICE or attempted suicide. Findings suggest, however, that a combination of GICE with a history of suicide attempts (pre- or post-GICE) was indicative of the highest risk across outcomes, highlighting the particularly hazardous nature of combining poor mental well-being and GICE exposures. Our study adds much-needed complexity to our understanding of how GICE exposure’s role in the well-being of TNB persons may differ related to their pre-GICE mental health. Our findings add further credence to previous studies identifying harms associated with GICE exposure, regardless of pre-GICE mental well-being.

Background

The United States (U.S.) is home to ∼1.6 million transgender and nonbinary (TNB) people (Herman et al., 2022). As the visibility of individuals identifying as TNB in the U.S. has increased (Jones, 2022), so have the efforts aimed at pushing TNB persons out of everyday life (MAP, 2023a). Gender identity conversion efforts (GICE) represent an attempt focused on altering the gender identity of TNB persons to align with their assigned sex through the use of psychological interventions (APA, 2021; Mallory et al., 2019). Conversion efforts, regardless of their focus on “converting” gender or sexual minority identities, are outlawed for licensed professionals in many U.S. jurisdictions and opposed by many relevant professional organizations (e.g. American Psychological Association, National Association of Social Workers, American Medical Association) (AMA, 2019; APA, 2021; Foster et al., 2015). Simultaneously, it can be lawfully practiced by mental health professionals in dozens of U.S. jurisdictions as well as by religious counselors (MAP, 2023b). Reporting suggests conversion efforts are also conducted in secrecy by those with or without a license, have been documented to include infliction of physical pain to alter one’s sexual or gender identity, including through exposure to electroshock “therapy” and physical and sexual abuse, and are known to conflate concepts related to gender, sexual orientation, and gender expression among others (APA, 2021; Ashley, 2020, 2021, 2022; Bishop, 2019; Bradshaw et al., 2015; Madrigal-Borloz, 2020; Ta, 2020; The Trevor Project, 2019; Tillewein & Kruse-Diehr, 2023). Despite serious concerns surrounding conversion practices, available data demonstrate that almost 700,000 LGBT adults in the U.S. may have been exposed to conversion efforts at some point in their lives (Mallory et al., 2019). Data on prevalence of conversion efforts targeting gender identity remains limited but a recent national study reported a lifetime GICE prevalence rate of 13.5% among TNB people (Turban et al., 2019).

Previous literature on conversion efforts

To date, much of the literature examining conversion efforts is aimed at understanding the experiences of sexual minorities or combined samples of sexual and gender minorities (SGM; with limited representation from the latter group), finding them to be ineffective and potentially harmful for the well-being of exposed persons (Beckstead & Morrow, 2004; Blosnich et al., 2020; Bradshaw et al., 2015; Forsythe et al., 2022; Meanley et al., 2020; Shidlo & Schroeder, 2002; Turban et al., 2020). Studies related to the efficacy of conversion efforts have found that people with previous exposure to conversion therapy are not likely to experience a significant change in their sexuality (Beckstead & Morrow, 2004; Maccio, 2011; Schaeffer et al., 2000; Shidlo & Schroeder, 2002). To date, only a single study we are aware of has described the efficacy of conversion efforts among TNB persons (Tillewein & Kruse-Diehr, 2023). In their qualitative study of four TNB persons exposed to sexual orientation conversion efforts (SOCE), Tillewein and Kruse-Diehr (2023) found participants pretended that their gender identity was aligned with their assigned sex so that they did not have to endure additional SOCE exposure. Alongside further underscoring the lack of efficacy of conversion efforts and understanding of sexual and gender identities and expressions, the study found these efforts to be largely traumatic for TNB persons (Tillewein & Kruse-Diehr, 2023). Descriptions of such harm are consistent with other work on SOCE and GICE exposure examining well-being among SGM, which have elucidated links between conversion efforts and depression, distress, post-traumatic stress disorder (PTSD), anxiety, guilt, shame, and suicidal ideation (Beckstead & Morrow, 2004; Blosnich et al., 2020; Campbell & Rodgers, 2023; Flentje et al., 2014; Forsythe et al., 2022; Heiden-Rootes et al., 2022; Mammadli et al., 2024; Meanley et al., 2020; Turban et al., 2020). We revisit the discussion of GICE-specific literature later in this article. Stemming from its lack of efficacy as well as the potential for harm, researchers have largely supported policy positions and bans against conversion efforts (Blosnich et al., 2020; Heiden-Rootes et al., 2022; Turban et al., 2020). However, some scholars have attempted to argue for the opposite. In one study of non-efficacious SOCE among 1518 cisgender sexual minorities, Sullins (2022) found significantly higher current minority stress among SOCE-exposed participants, but no differences in substance use, suicidal behaviors, self-harm, and current mental health. Sullins (2022) argued that his findings indicated that banning conversion efforts were not necessary given the lack of harm. It is important to acknowledge, however, that Sullins (2022) methodology and interpretation of the findings contain serious defects, including the misclassification of participants’ history of suicide attempts and overcontrol bias (e.g. controlling for psychological distress), which have cast serious doubt on the validity of the study conclusions (Meyer & Blosnich, 2022).

Gender minority stress and resilience model

The Gender Minority Stress and Resilience Model (GMSRM) (Hendricks & Testa, 2012) can provide critical insights into the harms associated with GICE. The GMSRM, adapted from the Minority Stress Model (Meyer, 2003), contends that in addition to the general stressors experienced by all persons, TNB persons may experience additional minority stressors due to living as gender minorities in a cis- and hetero-normative society. As part of the GMSRM, distal stressors refer to the objective stigmatizing experiences TNB persons may be exposed to including gender identity-related discrimination, victimization, rejection, and non-affirmation. Experiences with distal stressors are theorized to lead to internalized transphobia, nondisclosure of gender identity, and negative expectations, also described as proximal stressors. GMSRM posits that minority stressors (distal and proximal) can have a deleterious impact on the well-being of TNB persons whereas community connectedness and pride can serve as mitigatory factors against minority stressors. Previously, conversion efforts have been conceptualized as a minority stressor for sexual and gender minority persons (Blosnich et al., 2020; Mammadli et al., 2024).

With the goal of “changing” or “converting” one’s gender identity, GICE operates through stigmatization and non-affirmation of the gender identity of TNB persons (Andrade & Campo Redondo, 2022; Cramer et al., 2008). Thus, GICE represents a mechanism of non-affirmation, and, as such, a distal stressor. Minority stressors have been associated with disparities in health and well-being (e.g. mental health, substance use, and healthcare utilization; Bradford et al., 2013; Cicero et al., 2019; Reisner et al., 2015, 2016; Wolfe et al., 2021), socio-structural outcomes (e.g. housing instability, and bathroom avoidance; Begun & Kattari, 2016; Robinson, 2021; Rogers & Rogers, 2021; Weinhardt et al., 2017), and higher risk of interpersonal victimization (e.g. intimate partner violence, sexual assault; Gamarel et al., 2014; Reisner et al., 2014; Sevelius, 2013). Furthermore, a recent study of 254 TNB participants living in the U.S. found GICE exposure as a contributor to future negative expectations (Mammadli et al., 2024).

Health-related factors

As a distal stressor, non-affirmation of one’s gender identity has been linked to poorer mental well-being, risk behaviors (e.g. substance use), and avoidance of healthcare (Johnson et al., 2020). The documented harmful associations of conversion efforts with health-related outcomes may be a product of its impact of stigmatizing and not affirming one’s gender identity which may be plausibly assumed to contribute to the internalization of these negative beliefs and poorer well-being. As described, internalization of transphobia and poorer well-being are associated with GICE exposure. The research on the topic has been primarily concerned with the experiences of sexual minorities, with only five studies examining outcomes specifically among gender minority persons in the U.S. (Campbell & Rodgers, 2023; Heiden-Rootes et al., 2022; Mammadli et al., 2024; Tillewein & Kruse-Diehr, 2023; Turban et al., 2020). These studies have identified higher odds of suicidality, running away, PTSD, disassociation, and psychological distress among GICE exposed TNB participants and concluded that their findings support the policy positions and bans against GICE. The first of these studies, conducted by Turban et al. (2020), found that lifetime GICE exposure was associated with significant increases in odds of lifetime suicidal ideation and attempts, as well as past 30-day severe psychological distress among the sample of TNB adults in the U.S. Although Turban et al. (2020) did not identify differences in substance use based on GICE exposure, Heiden-Rootes et al. (2022) found higher odds of drinking among Black TNB participants and lower odds among White TNB participants exposed to non-religious GICE compared to those who never experienced GICE. It has been suggested that substance use may serve as a coping strategy to deal with minority stressors, including non-affirmation (Felner et al., 2020; Oorthuys et al., 2022; Reisner et al., 2015). Further, literature has indicated that TNB persons who experience stigma during their healthcare utilization may forego future preventative and diagnostic healthcare visits to avoid further stigmatization (Lerner et al., 2020). GICE exposure may be especially concerning as the visit’s purpose would be to stigmatize the person’s marginalized identities. In two recent studies, conversion efforts were linked to fears of being stereotyped in healthcare visits (Mammadli et al., 2024) and cited as a reason for avoidance of mental health and medical professionals (Tillewein & Kruse-Diehr, 2023). Therefore, it is plausible that persons exposed to GICE may be more likely to avoid future healthcare utilization.

Socio-structural factors

Non-affirmation of one’s gender identity and other minority stressors can have deleterious effects on the social and structural outcomes of many TNB persons. Recently, the use of bathrooms congruent with one’s gender identity has become an increasingly pertinent topic of conversation and a target of legislation aimed at limiting the rights of TNB people (MAP, 2023c). Many state and local governments have enacted policies mandating the use of bathrooms that align with one’s sex assigned at birth, casting TNB persons’ attempts at using public bathrooms that align with their gender identity as predatory. Many TNB persons now report avoiding public bathrooms as a manifestation of this legislative and rhetorical transphobia (Lerner, 2021). In addition, the literature suggests that TNB persons who do not feel affirmed in their gender may avoid the use of public restrooms that align with their gender identity due to safety concerns (Weinhardt et al., 2017).

TNB persons’ experiences of socio-structural alienation go beyond the use of public bathrooms. In the U.S., housing instability is disproportionally prevalent among TNB persons (Ecker et al., 2019). At a young age, TNB persons may experience family and social rejection or an unsafe home environment due to the response to their gender identity, which can lead to TNB persons losing access to crucial material and emotional support as well as a home early in their lives. Simultaneously, job opportunities may be limited for many TNB individuals, as they experience widespread employment discrimination due to reactions to their TNB identity that limit their ability to meet basic needs (James et al., 2016). They may face further discrimination when looking for housing and can be alienated from many social and structural institutions, such as homeless shelters where their gender identity may be used against them as an additional barrier (Grant et al., 2011; James et al., 2016; McBride, 2012). TNB persons whose gender affirmation needs are not met may also be more susceptible to socio-structural vulnerabilities, such as housing instability (Gamarel et al., 2021; Sevelius, 2013). Although literature to-date has not examined the link between exposure to a psychological mechanism of non-affirmation, such as GICE and socio-structural outcomes, the existence of such a link is nevertheless worth exploring given the theoretical and empirical links showing the pertinent role minority stressors, including non-affirmation of gender identity, may play in important social and structural outcomes.

Interpersonal victimization

TNB persons experience sexual violence and physical and psychological intimate partner violence (IPV) at disproportionately higher rates than their cisgender counterparts (Peitzmeier et al., 2020). Although we are not aware of literature connecting GICE exposure with sexual victimization or IPV (physical or psychological), GICE as a minority stressor may be connected to physical, psychological, and sexual interpersonal victimization. Given that transphobia and related stigma are linked to IPV vulnerability (Goldenberg et al., 2018; Grossman & D’Augelli, 2006), the stigmatizing and non-affirming role of GICE exposure may act in similar ways. Literature indicates that TNB persons may seek gender affirmation through sex work if their gender affirmation needs are not met, which can further increase sexual and IPV victimization risks (Goldenberg et al., 2018; Sevelius, 2013). As such, disparities in non-affirmation that TNB persons experience resulting from GICE exposure can contribute to the unmet gender affirmation needs of TNB persons, placing them at higher risk of interpersonal victimization.

IPV victimization is also linked to other adverse experiences including housing instability, incarceration, family assault and harassment, discrimination, and recurring gender victimization (Brennan et al., 2012; Goldenberg et al., 2018; James et al., 2016; Peitzmeier et al., 2020; White Hughto et al., 2017). TNB persons experience these adverse events at significantly higher rates. Related to these adverse experiences and minority stressors, TNB persons are also at higher risk for individual and interpersonal factors associated with IPV victimization, including poor mental well-being, substance use, and sexual health vulnerability (Brennan et al., 2012). GICE exposure may be yet another contributing factor.

It has also been suggested that experiences of stigmatization may reduce one’s ability to maintain a healthy balance of power in their relationships (Goldenberg et al., 2018), potentially contributing to IPV victimization. Internalization of minority stressors, including stigma, may be a risk factor that places minorities at higher risk of IPV victimization, as victims may be led to believe that they are culpable for their victimization or that they are hopeless in seeking help due to their gender identity (Kimmes et al., 2019; Stephenson et al., 2010; Stephenson & Finneran, 2016). This is important as conversion efforts have been demonstrated to contribute to the internalization of minority stressors (Balsam & Szymanski, 2005; Kinitz & Salway, 2022).

Criticisms of previous GICE literature and the current study

Noted previously, one of the five studies focused on conversion experiences among TNB persons was conducted by Turban et al. (2020), whose findings have been supportive of bans and policy positions against conversion efforts. However, this study was not absent from criticism. Among others, D’Angelo et al. (2021) argue Turban et al. (2020) claimed causality from associational findings, misappropriated the cutoff on the Kessler-6 psychological distress scale, utilized a biasing question to measure GICE, and did not account for pre-GICE mental well-being. We find many of the arguments made by D’Angelo et al. (2021) to be unsubstantiated and beyond the scope of the current study. It is important, however, to recognize that TNB persons may experience serious mental health concerns before GICE exposure, and not accounting for the temporal precedence of serious mental health concerns with GICE exposure may fail to capture a complete and nuanced understanding of GICE-related harms. To date, we are unaware of any studies that have elucidated the role of GICE exposure on the well-being of TNB persons while accounting for the temporal precedence of GICE exposure and serious mental health issues. This is a limitation that the current study aims to address.

Transitioning and detransitioning

Similar to the reinvigoration of the debate surrounding conversion efforts, transitioning and detransitioning experiences of TNB persons are also being increasingly debated (Jorgensen, 2023; MacKinnon et al., 2022; Temple Newhook et al., 2018; Turban et al., 2021; Walls et al., 2024). Although an extensive discussion is beyond the focus of our current work, we find it pertinent to briefly describe the role of de-/transitioning status on the well-being of TNB persons. Recent work suggests detransitioning experiences may impact the well-being of TNB persons. As an example, MacKinnon et al. (2022) describe anticipated stigma and avoidance of healthcare related to their detransitioning among 28 participants who had reversed their initial transitions, the majority of whom did not regret their previous transitions. In some literature positioned against gender-affirming approaches to transgender health, authors have described feelings of isolation, shame, discrimination, and anticipated stigma as well as the need for mental health treatment among detransitioners related to their transitioning and detransitioning experiences (Jorgensen, 2023; Littman, 2021; Marchiano, 2021; Vandenbussche, 2022). It is critical to note that serious methodological and conceptual concerns plague some of the work around detransitioning experiences (Temple Newhook et al., 2018; Walls et al., 2024) and that TNB persons may reverse their gender transition for many reasons including due to societal and interpersonal pressures (Turban et al., 2021; Walls et al., 2024). Similarly, recent work indicates that the decision to reverse gender identity transitions due to societal or interpersonal factors is correlated with an increased risk of outcomes, such as a history of arrests, avoidance of healthcare, housing instability, and IPV (Walls et al., 2024). Given their importance in the overall well-being of TNB persons, we find it necessary to account for de-/transitioning experiences of TNB persons in our study.

In the current study, we examine how exposure to GICE may be associated with health-related, socio-structural, and interpersonal victimization outcomes of TNB persons living in the U.S. Utilizing the U.S. Transgender Survey (USTS) (James et al., 2016) dataset, used by Turban et al. (2020), we build on their work by (a) constructing a GICE exposure variable with categories that account for participants’ suicide attempt history in relation to their GICE exposure and (b) by examining the relationship with temporal GICE and suicide attempt history and health-related (i.e. psychological distress, extra-medical use of prescription substances, healthcare avoidance), socio-structural (i.e. bathroom avoidance and housing instability), and interpersonal outcomes (i.e. sexual assault, emotional and physical IPV). Our study adds complexity to the current limited understanding of conversion efforts and their related harms by examining how GICE exposure may have distinct effects on the lives of TNB persons based on their history of suicide attempts and the temporality of their suicide attempts and GICE exposure. History of suicide attempts and the timing of suicide attempts with GICE exposure were chosen as any history of suicide attempts can be indicative of a serious mental illness that may or may not be related to GICE exposure.

Materials and methods

Data source

The current study was conducted using the 2015 U.S. Transgender Survey (USTS) (James et al., 2016). The USTS is an online cross-sectional survey of 27,715 transgender and nonbinary (TNB) people living across the U.S. A multimethod strategy was employed to recruit respondents for the USTS, including through social media and partnership with hundreds of organizations and groups that serve or are allied with LGBTQ communities (James et al., 2016). Eligible participants included TNB adults (18+) living in the U.S. The survey was available for administration in English and Spanish. Upon completion, interested participants could opt-in to enter a drawing to win a cash prize. The Institutional Review Board (IRB) at the University of Denver reviewed and determined the study to be Not Human Subject Research as the study is a secondary data analysis project of existing, de-identified data. The following describes the USTS variables used in current analyses.

Variables and measures

GICE exposure and suicide attempts

The exposure variable in the current study was constructed using the gender identity conversion therapy efforts (GICE) and suicide attempt variables. GICE exposure was measured by an affirmative response to the following question: “Did any professional (such as a psychologist, counselor, religious advisor) try to make you identify only with your sex assigned at birth (in other words, try to stop you being trans)?” Participants who responded affirmatively were then asked to report their age during their initial GICE experience. Suicide attempts were measured by the following question: “At any time in your life, did you try to kill yourself?” Participants who reported previous suicide attempts were similarly asked to indicate the age of their initial suicide attempt. Using the temporal order of participants’ first GICE exposure and suicide attempt, we constructed a combined GICE and suicide attempt variable. The categories in this newly constructed variable consisted of: “No GICE and no suicide attempt,” “GICE without a suicide attempt,” “No GICE with a history of suicide attempts,” “GICE with pre-GICE initial suicide attempt,” and “GICE with post-GICE initial suicide attempt.” Participants with an unknown order of GICE exposure and suicide attempts were dropped from all analyses (n = 384) as the temporal ordering of the experiences is central to the research questions examined. No GICE exposure and no suicide attempt were the reference category.

Outcomes

Health-related outcomes

There were three health-related outcomes in the current study. Past 30-Day Psychological Distress. The Kessler-6 (K6) Psychological Distress Scale (Kessler et al., 2002) was utilized to measure participants’ psychological distress levels. The K6 consists of six items with a five-point Likert scale ranging from All of the time (1) to None of the time (5). The items measure participants’ past 30-day experiences with nervousness, anxiety, and depression. An example of an item on the K6 includes “During the past 30 days, about how often did you feel restless or fidgety?” A cut-off score of 13 (range 0–24) is indicative of severe psychological distress (Kessler et al., 2003). In this sample, the K6 demonstrated acceptable internal consistency with a Cronbach’s α, of .66. For data analysis, psychological distress was dichotomized around the cutoff score. Past 12-Month Extra-Medical Use of Prescription Substances. Participants who indicated using prescription drugs that were not prescribed to them were asked, “How long has it been since you last used any prescription drugs not as prescribed or not prescribed to you?” Participants who selected “within the past 30 days” or “more than 30 days ago but within the past 12 months” were coded as 1 whereas those who reported the extra-medical use of prescription substances “more than 12 months ago” were coded as 0. Past 12-Month Healthcare Avoidance. Past 12-month healthcare avoidance due to gender identity was measured using the following binary (Yes/No) item: “Was there a time in the past 12 months when you needed to see a doctor but did not because you thought you would be disrespected or mistreated as a trans person?”

Socio-structural outcomes

There were two socio-structural risk factors examined in the current study. Past 12-Month Bathroom Avoidance. Participants reported the past 12-month bathroom avoidance by responding to the following question: “In the past year, did you avoid going to the bathroom because you were afraid of having problems using them? This would include bathrooms in public, at work, or at school.” Participants could respond “I have never avoided them,” “I have sometimes avoided them,” “I have always avoided them,” or “Not listed above (please specify).” For data analysis, participants who reported never having avoided bathrooms were coded as 0 and those who responded using any other options were coded as 1. Past 12-Month Housing Instability. Housing instability was operationalized by the question, “Now just thinking about the past year, have you had any of these housing situations because you are trans? (Please provide an answer in each row.) In the past year… I experienced homelessness.” Participants could select “yes,” “no,” or “does not apply to me.” Participants who responded affirmatively were coded as having had housing instability in the past year.

Interpersonal outcomes

We examined three interpersonal factors as outcomes. Past 12-Month Sexual Assault. Sexual assault was measured by the question, “Now just thinking about the past year, have you experienced unwanted sexual contact (such as oral, genital, or anal contact or penetration, forced fondling, rape)?” IPV items inquired about participants’ experiences with lifetime emotional and physical IPV. Lifetime Emotional IPV. The variable utilized to operationalize emotional IPV consisted of twelve dichotomous items (yes/no). One example of an item on this measure includes “Have any of your romantic or sexual partners ever tried to keep you from seeing or talking to your family or friends?” Participants who responded affirmatively to any of the items were deemed to have previous emotional IPV experience. Lifetime Physical IPV. Similarly, twelve questions were used to operationalize physical IPV with an example being, “Have any of your romantic or sexual partners ever tried to hurt you by choking or suffocating you?” Those who reported experiencing any of the twelve experiences were recoded as experiencing physical IPV.

Covariates

Covariates in the current study included socio-demographic variables and de-/transitioning status. Among sociodemographic covariates, participants’ ages were recorded with the question, “What is your current age?” Age had a curvilinear relationship with interpersonal victimization outcomes in the current study. As such, an age-squared variable was added as a covariate to the three interpersonal outcome models to better account for this curvilinear relationship. Regarding gender identity, participants were asked to choose from one of the 25 gender identity response options (e.g. “agender,” “two-spirit,” “transgender”). Participants who did not identify with any gender identity from the prepopulated list were provided with space to type their gender identity. In data analysis, gender identity categories were “crossdresser,” “nonbinary, assigned male at birth (AMAB),” “nonbinary, assigned female at birth (AFAB),” “transgender woman,” and “transgender man.” Similar to the gender identity question, participants were asked to report their sexual orientation either by choosing from one of the eight prepopulated options (e.g. asexual, gay, same-gender loving, queer) or by specifying their sexual orientation if not listed. In analyses, seven sexual orientation categories were reflected: “heterosexual,” “asexual,” “bisexual,” “lesbian, gay, or same-gender loving,” “pansexual,” “queer,” or “other.” Further, the racial/ethnic identity of the participants was recorded via the question, “Although the choices listed below may not represent your full identity or use the language you prefer, for this survey please select the choice that most accurately describes your racial/ethnic identity. (Please choose only one answer.).” Participants could either select from one of the nine options provided or specify their racial/ethnic identity in their own words. Racial/ethnic identity categories in data analysis consisted of “White,” “American Indian/Native American,” “Asian,” “Biracial/Multiracial,” “Black,” and “Latinx.”

Relationship status, another demographic covariate, was documented with the following question “What is your current relationship status?” Response options consisted of “partnered, living together,” “partnered, not living together,” and “single.” Participants could also specify their relationship status in their own words. This variable was dichotomized as “partnered” or “single” for the purposes of data analysis. Additionally, participants reported their educational attainment by choosing one of the 13 choices provided [e.g. GED, some college, no degree (including currently in college), professional degree (e.g. MD, JD)]. The following categories comprised the education variable in data analysis: “less than high school,” “high school degree,” “some college,” and “college graduate.” The other two demographic covariates included in all analyses were employment status and household income. Employment status was captured via the question “What is your current employment status?” with 10 options, including “student,” “retired,” and “work for pay from sex work, selling drugs, or other work that is currently considered illegal.” For the purposes of analyses, employment categories were recoded as “employed,” “unemployed,” “out of the labor force,” and “unspecified.” Lastly, participants were asked to choose from 18 prepopulated income brackets to report their pretax household income level in 2014. These options were recoded as the following in data analysis: “no income,” “<$10,000,” $10,000–$24,999,” “$25,000–$49,999,” “$50,000–$99,999,” “$100K+,” and “unknown.”

De-/transition status was included as an important theoretical covariate in all analyses (Author, under review). Participants were asked “Have you ever de-transitioned? In other words, have you ever gone back to living as your sex assigned at birth, at least for a while?” with the response options “I have never transitioned,” “no,” and “yes.” These categories were preserved during data analysis.

Missing data

The degree of missingness varied depending on the variable, ranging from 0% to 2.09%. Missingness levels higher than 5–10% are likely to lead to biased analysis, with 5% deemed as a more appropriate threshold for large datasets (Dong & Peng, 2013; Madley-Dowd et al., 2019). Due to the low percentage of missingness in the current study, all observations with missingness in the relevant variables (n = 1895; 6.8%) were dropped from analyses. The total analytical sample in the current study consisted of 25,436 TNB participants.

Data analysis

All analyses were conducted using Stata 16.1 (StataCorp, 2019). Initially, we conducted univariate statistics to obtain a basic view of participant characteristics. Next, we conducted multivariate logistic regression models to examine the relationship between combined GICE and suicide attempt experiences and the outcome variables while controlling for demographic variables and de-/transition status. Data analysis was conducted using the svy command to account for the complex survey design of the USTS. The USTS data were collected using convenience sampling. Accordingly, to allow for comparison of findings with the general U.S. population, weights were available to account for race/ethnicity, age, and educational attainment related differences. These sampling weights were derived from the population estimates of the 2014 American Community Survey (James et al., 2016). Sample demographics are presented without weighted estimates, whereas all multivariate analyses were conducted using weighted samples. Eight separate models were estimated in the current study. Bonferroni correction for multiple comparisons was conducted to reduce the risk of Type I error (a = .01).

Results

Participant characteristics

Participant characteristics of the analytic sample can be found in Table 1. Briefly, the average participant was 31 years old with most identifying as White (82.4%). Most commonly, participants identified as transgender women (32.8%) and queer (20.7%) with a majority reporting having a partner (51.9%) and never detransitioning (53.7%). Most had some college education (46.4%) or a college degree (37.9%) as their highest level of education and were employed (65.1%).

Table 1.

Demographic characteristics of the participants (N = 25,436).

	Prevalence % (n)	Mean (Std. dev.)
Gender identity
Transgender woman	32.77% (8335)
Transgender man	28.97% (7481)
Nonbinary (AFAB)	28.67% (7292)
Nonbinary (AMAB)	6.94% (1654)
Crossdresser	2.65% (674)
Race
White	82.38% (20,954)
Black	2.81% (714)
Latinx	5.26% (1337)
Biracial/multiracial	5.57% (1417)
Asian American	2.87% (731)
American Indian/Native American	1.11% (283)
Age	–	30.97 (13.30)
Sexual orientation
Heterosexual	12.05% (3063)
Asexual	10.93% (2780)
Bisexual	14.88% (3786)
Lesbian/Gay/SGL	16.50% (4198)
Pansexual	18.32% (4660)
Queer	20.68% (5259)
Other	6.64% (1690)
Relationship status
Single	48.11% (12,237)
Partnered	51.89% (13,199)
Educational attainment
Less than high school	3.22% (818)
High school	12.42% (3159)
Some college	46.42% (11,808)
College degree	37.94% (9651)
Employment status
Unemployed	12.93% (3287)
Out of the labor force	21.60% (5491)
Employed	65.05% (16,769)
Unspecified	.42% (106)
Household income
No income	3.64% (926)
<$10,000	11.18% (2845)
$10,000–$24,999	18.10% (4604)
$25,000–$49,999	20.64% (5250)
$50,000–$99,999	22.69% (5772)
$100,000+	15.3% (3895)
Unknown	8.43% (2144)
De-/transition status
Never transitioned	38.48% (9786)
Never detransitioned	53.66% (13,650)
Detransitioned	7.86% (2000)
Primary independent variable
GICE and suicide attempt history
No GICE no suicide attempt	55.98% (14,240)
No suicide attempt w/GICE	5.52% (1403)
Suicide attempt No GICE	31.59% (8035)
Suicide attempt pre-GICE	2.04% (520)
Suicide attempt post-GICE	4.87% (1238)
Outcome prevalence
Health-related outcomes
Psychological distress
No evidence of SMI (K6 Score <13)	61.3% (15,592)
Evidence of SMI (K6 Score 13+)	38.7% (9844)
Past year extra-medical use of prescription substances	15.99% (4068)
Healthcare avoidance due to anticipated provider disrespect	22.01% (5599)
Socio-structural outcomes
Housing instability	7.94% (2019)
Avoidance of public bathroom	61.88% (15,740)
Interpersonal outcomes
Emotional IPV	43.34% (11,025)
Physical IPV	40.84% (10,389)
Sexual assault	9.44% (2400)

SGL: same-gender loving; SMI: serious mental illness; IPV: intimate partner violence.

Health-related outcomes

Table 2 provides findings from logistic regression models examining the relationship between temporal GICE and suicide attempt experiences and health-related outcomes while controlling for demographic covariates and de-/transition status.

Table 2.

Health-related outcomes.

	Past 30-day	Past 12-month	Past 12-month
	Psychological distress	Extra-medical use of prescription substances	Healthcare avoidance
	aOR (SE)	aOR (SE)	aOR (SE)
Age	.95 (.0038)***	.97 (.0039)	.97 (.0038)***
Gender identity
Nonbinary (AMAB)	Ref	Ref	Ref
Crossdresser	.68 (.1840)	.56 (.1552)*	1.08 (.4542)
Transgender woman	1.02 (.1897)	.87 (.1284)	2.06 (.3355)***
Transgender man	.83 (.1440)	.92 (.1440)	2.27 (.3837)***
Nonbinary (AFAB)	1.08 (.1728)	.99 (.1486)	1.65 (.2655)**
Sexual orientation
Heterosexual	Ref	Ref	Ref
Asexual	1.58 (.2871)	.83 (.1830)	.77 (.1670)
Bisexual	1.14 (.1651)	1.05 (.1744)	.85 (.1630)
Lesbian/Gay/SGL	1.32 (.2159)	.92 (.1491)	.68 (.1141)*
Pansexual	1.51 (.2156)**	1.15 (.1968)	.94 (.1426)
Queer	1.34 (.1978)	1.65 (.2664)**	1.38 (.1937)
Other	1.67 (.3251)**	1.03 (.2134)	1.00 (.1963)
Race
White	Ref	Ref	Ref
American Indian/Native American	.90 (.2201)	.83 (.2353)	1.39 (.4421)
Asian American	.90 (.2270)	.67 (.1120)*	1.57 (.4160)
Biracial/multiracial	1.52 (.4038)	1.20 (.1932)	1.38 (.2186)*
Black	.81 (.1366)	.72 (.1328)	.98 (.1789)
Latinx	1.06 (.1477)	1.09 (.1541)	1.14 (.2001)
Relationship status
Single	Ref	Ref	Ref
Partnered	.94 (.0750)	1.23 (.1101)*	1.05 (.1010)
Educational attainment
Less than high school	1.37 (.2797)	1.02 (.2267)	1.49 (.3247)
High school	1.66 (.1665)***	1.20 (.1443)	.80 (.0932)
Some college	1.27 (.0770)***	1.54 (.0868)	.80 (.0560)***
College degree	Ref	Ref	Ref
Employment status
Employed	Ref	Ref	Ref
Unemployed	1.62 (.1946)***	1.15 (.1584)	1.06 (.1545)
Out of the labor force	1.69 (.1840)***	.84 (.0952)	1.12 (.1317)
Unspecified	2.13 (.8742)	1.11 (.3934)	.68 (.3844)
Household income
No income	1.23 (.3683)	1.29 (.3948)	2.01 (.7163)
<$10,000	1.41 (.2567)	.85 (.1419)	1.72 (.3397)**
$10,000–$24,999	1.25 (.1704)	.92 (.1465)	1.55 (.2395)**
$25,000–$49,999	1.06 (.1329)	.90 (.1288)	1.56 (.2194)***
$50,000–$99,999	1.13 (.1621)	.86 (.1191)	1.27 (.1970)
$100K+	Ref	Ref	Ref
Unknown	1.16 (.1961)	.94 (.2033)	1.56 (.3250)*
De-/transition status
Never transitioned	Ref	Ref	Ref
Never detransitioned	.61 (.0531)***	.97 (.1037)	1.85 (.1872)***
Detransitioned	.89 (.1107)	1.52 (.2293)**	2.98 (.5015)***
GICE and suicide attempt history
No GICE no suicide attempt	Ref	Ref	Ref
No suicide attempt w/GICE	1.66 (.2507)***	1.32 (.2706)	2.80 (.5720)***
Suicide attempt No GICE	2.77 (.2438)***	1.71 (.1590)***	1.66 (.1740)***
Suicide attempt pre-GICE	5.30 (1.2727)***	1.77 (.3984)*	4.17 (.9328)***
Suicide attempt post-GICE	3.40 (.6127)***	2.35 (.4327)***	2.98 (.5190)***
_cons	1.09 (.2715)	.33 (.0878)***	.11 (.0327)

SGL: same-gender loving.

Note. Linearized standard errors reported. *p ≤ .05, **p ≤ .01, ***p ≤ .001.

Past 30-day psychological distress

Serious psychological distress was prevalent among 38.7% of the participants. Among demographics, higher age was associated with lower odds of serious psychological distress (aOR = .95; 95% CI [.94–.96]). Pansexual participants (aOR = 1.51; 95% CI [1.39–1.99]) and those with “other” sexual orientations (aOR = 1.67; 95% CI [1.14–2.45]) were more likely to have serious psychological distress relative to their heterosexual counterparts. Educational attainment and employment status were also significant demographic factors associated with the odds of serious psychological distress. More specifically, relative to college graduates, participants with high school education were 1.7 times (95% CI [1.36–2.02]) more likely to experience severe psychological distress whereas those with some college education were 1.3 times (95% CI [1.13–1.43]) more likely to experience severe psychological distress. Higher severe psychological distress odds were also observed among participants who were unemployed (aOR = 1.62; 95% CI [1.28–2.05]) or out of the labor force (aOR = 1.68; 95% CI [1.36–2.09]) relative to employed participants. Participants who transitioned but never detransitioned had a lower likelihood of experiencing severe psychological distress relative to those who never transitioned (aOR = .61; 95% CI [.52–.72]).

Regarding GICE and suicide attempt experiences, relative to participants without any GICE exposure or suicide attempt, significantly higher odds of serious psychological distress were observed for participants in all categories. Participants who experienced GICE without a suicide attempt history were 1.7 times (95% CI [1.24–2.24]) more likely to report severe psychological distress in comparison with the no GICE no suicide group whereas participants with a previous suicide attempt in the absence of GICE were 2.8 times (95% CI [2.33–3.29]) more likely to report severe psychological distress. Relative to the no GICE no suicide group, the highest odds of severe psychological distress were observed among participants who reported their initial suicide attempt pre-GICE (aOR = 5.30 (95% CI [3.31–8.49]) followed by post-GICE (aOR = 3.40; 95% CI [2.38–4.84]).

Past 12-month extra-medical prescription substance use

Overall, 16% of the participants reported extra-medical use of prescription substances over the past year. Among demographic factors, sexual orientation was the only significant correlate of extra-medical use of prescription substances. Particularly, queer TNB participants were 1.7 times (95% CI [1.20–2.27]) more likely to report extra-medical prescription substance use over the past 12-months relative to heterosexual TNB participants. Participants who detransitioned were also 1.5 times (aOR = 1.52; 95% CI [1.13–2.04]) more likely to report extra-medical prescription substance use in comparison with those who never transitioned. Compared to the no GICE no suicide group, participants who attempted suicide without GICE exposure were 1.7 times (95% CI [1.42–2.05]) and those who attempted suicide for the first time post-GICE were 2.4 times (95% CI [1.64–3.38]) more likely to report extra-medical use of prescriptions.

Past 12-month healthcare avoidance

Approximately 22% reported healthcare avoidance due to anticipated provider disrespect or mistreatment related to participants’ gender identity. Age was a significant factor, with older participants reporting lower odds of healthcare avoidance (aOR = .97; 95% CI [.97–.98]). Over the past year, transgender women (aOR = 2.06; 95% CI [1.50–2.83]), transgender men (aOR = 2.27; 95% CI [1.63–3.16]), and nonbinary AFAB participants (aOR = 1.65; 95% CI [1.20–2.26]) had higher odds of avoiding healthcare related to anticipated disrespect or mistreatment compared to nonbinary AMAB participants. Further, participants with some college education were less likely to avoid healthcare (aOR = .80; 95% CI [.70–.92]) relative to those with at least a college degree. Household income was also a significant correlate of healthcare avoidance. Participants with household incomes <$10,000 (aOR = 1.72; 95% CI [1.17–2.54]), between $25,000 and $49,999 (aOR = 1.55; 95% CI [1.15–2.10]), and between $50,000 and $99,999 (aOR = 1.56; 95% CI [1.19–2.06]) had higher odds of healthcare avoidance relative to those with household income higher than $100,000. Further, participants who detransitioned (aOR = 1.85; 95% CI [1.52–2.26]) as well as participants who transitioned but never detransitioned (aOR = 2.98; 95% CI [2.14–4.15]) were both at a higher risk of reporting healthcare avoidance relative to participants who never transitioned.

GICE and suicide attempt history were significant correlates of healthcare avoidance among participants. TNB participants who never attempted suicide but were exposed to GICE were 2.8 times more likely to avoid healthcare (95% CI [1.87–4.17]) over the past year relative to the no GICE no suicide attempt group, whereas participants who attempted suicide without GICE exposure were 1.7 times more likely to avoid healthcare over the past year due to anticipated stigma (95% CI [1.35–2.03]). Further, participants who attempted suicide initially before GICE exposure had 4.2 times higher healthcare avoidance risk in comparison with the no GICE no suicide group (95% CI [2.69–6.47]). Similarly, participants who attempted suicide for the first time post-GICE were 3 times more likely to avoid healthcare (95% CI [2.12–4.19]) relative to the no GICE no suicide group (Figure 1).

Figure 1.

Odds of experiencing health vulnerabilities based on GICE exposure and suicide attempt history.

Figure 2.

Odds of experiencing socio-structural vulnerabilities based on GICE exposure and suicide attempt history.

Socio-structural outcomes

Table 3 displays results from the models examining the relationship between temporal GICE exposure and suicide attempt history and socio-structural outcomes while controlling for demographic covariates and de-/transition status.

Table 3.

Socio-structural outcomes.

	Past 12-month	Past 12-month
	Bathroom avoidance	Housing instability
	aOR (SE)	aOR (SE)
Age	.97 (.0034)***	.98 (.0061)**
Gender identity
Nonbinary (AMAB)	Ref	Ref
Crossdresser	1.37 (.2955)	.86 (.4712)
Transgender woman	1.42 (.2131)*	.80 (.2032)
Transgender man	2.92 (.4462)***	.56 (.1390)*
Nonbinary (AFAB)	.87 (.1166)	.45 (.1135)**
Sexual orientation
Heterosexual	Ref	Ref
Asexual	1.61 (.2329)***	1.08 (.3062)
Bisexual	1.23 (.1629)	1.51 (.3513)
Lesbian/Gay/SGL	1.16 (.1460)	1.43 (.3765)
Pansexual	1.43 (.1763)**	1.85 (.4342)**
Queer	2.28 (.2770)***	1.36 (.3201)
Other	1.50 (.2583)*	1.88 (.5185)*
Race
White	Ref	Ref
American Indian/Native American	.92 (.2167)	1.24 (.4379)
Asian American	1.35 (.3280)	1.60 (.6494)
Biracial/multiracial	1.32 (.3395)	1.71 (.5170)
Black	.53 (.0768)***	1.64 (.3572)*
Latinx	.99 (.1343)	1.13 (.2714)
Relationship status
Single	Ref	Ref
Partner	1.00 (.0714)	.73 (.0916)*
Educational attainment
Less than high school	.73 (.1335)	1.65 (.3943)*
High school	1.09 (.1052)	1.66 (.2960)**
Some college	1.06 (.0615)	1.34 (.1443)**
College degree	Ref	Ref
Employment status
Employed	Ref	Ref
Unemployed	1.06 (.1324)	1.90 (.3045)***
Out of labor force	.86 (.0798)	.95 (.1638)
Unspecified	.52 (.1517)*	.56 (.3398)
Household income
No income	1.13 (.3660)	10.30 (5.2428)***
<$10,000	1.10 (.1688)	6.84 (3.2422)***
$10,000–$24,999	1.27 (.1499)*	4.23 (1.9353)**
$25,000–$49,999	1.12 (.1181)	2.07 (.9424)
$50,000–$99,999	1.00 (.0985)	1.77 (.8697)
$100K+	Ref	Ref
Unknown	1.20 (.1876)	5.78 (2.8270)***
De-/transition status
Never transitioned	Ref	Ref
Never detransitioned	1.39 (.1160)***	1.61 (.2303)***
Detransitioned	1.67 (.2165)***	2.20 (.4097)***
GICE and suicide attempt history
No GICE no suicide attempt	Ref	Ref
No suicide attempt w/GICE	1.28 (.2310)	1.61 (.3967)
Suicide attempt No GICE	1.33 (.1057)***	2.28 (.3483)***
Suicide attempt pre-GICE	1.82 (.4554)*	3.17 (.8631)***
Suicide attempt post-GICE	1.92 (.3515)***	3.08 (.7172)***
_cons	1.34 (.2822)	.02 (.0101)***

SGL: same-gender loving.

Note. Linearized standard errors reported. *p ≤ .05, **p ≤ .01, ***p ≤ .001.

Past 12-month bathroom avoidance

Avoidance of bathrooms was highly prevalent, with 61.9% reporting avoiding bathrooms over the past year due to fears. Older TNB participants were less likely than their younger counterparts to avoid public bathrooms (aOR = .97; 95% CI [.96–.98]). Transgender men were 2.9 times more likely to report avoiding bathrooms compared to their AMAB nonbinary counterparts (95% CI [2.17–3.94]). Higher risk was also observed among asexual (aOR = 1.61; 95% CI [1.21–2.14]), pansexual (aOR = 1.43; 95% CI [1.12–1.82]), and queer (aOR = 2.28; 95% CI [1.79–2.89]) TNB participants compared with heterosexual TNB participants. Regarding race, Black participants were significantly less likely to report avoiding bathrooms than White participants (aOR = .53; 95% CI [.40–.70]). Additionally, participants who never detransitioned (aOR = 1.39, 95% CI [1.18–1.63]) and those who reported detransitioning (aOR = 1.67, 95% CI [1.30–2.15]) experienced higher bathroom avoidance odds vs. those who never transitioned.

TNB participants who attempted suicide, but never experienced GICE were 1.3 times more likely to avoid bathrooms relative to participants who experienced neither (95% CI [1.14–1.55]), whereas those who attempted suicide initially post-GICE (95% CI [1.34–2.75]) had 1.9 times higher odds of avoiding bathrooms due to fears over the past year. Participants who reported having a pre-GICE initial suicide attempt and those who attempted suicide in the absence of GICE were not significantly different than those who experienced neither GICE nor suicide attempts.

Past 12-month housing instability

Over the past year, ∼7.9% of the participants experienced housing instability, with younger participants reporting higher odds of past year housing instability (aOR = .98; 95% CI [.97–.99]) relative to their older counterparts. Further, nonbinary AFAB participants had lower odds of housing instability (aOR = .45; 95% CI [.27–.74]) relative to nonbinary AMAB participants. However, pansexual participants had 1.9 times higher odds of housing instability compared to heterosexual TNB participants (95% CI [1.17–2.93]). Employment and educational attainment were also significant predictors of housing instability. Specifically, unemployed participants (aOR = 1.9, 95% CI [1.39–2.60]) reported higher odds of past year housing instability compared to their employed counterparts. High school graduates (aOR = 1.66; 95% CI [1.17–2.35]) and participants with some college education (aOR = 1.34; 95% CI [1.08–1.65]) had higher past year housing instability odds than participants with at least a college degree. Unsurprisingly, household income was also a salient correlate of housing instability. Participants with unknown (aOR = 5.78, 95% CI [2.21–15.08]) or no household income (aOR = 10.30, 95% CI [3.80–27.93]), as well as those with household incomes <$10,000 (aOR = 6.84, 95% CI [2.70–17.32]) and between $10,000 and $24,999 (aOR = 4.23, 95% CI [1.72–10.37]) had higher housing instability odds over the past year relative to participants with household incomes over $100,000.

Certain de-/transition-related experiences were correlated with higher odds of housing instability. Particularly, participants who detransitioned (aOR = 2.20, 95% CI [1.53–3.17]) and those who never detransitioned (aOR = 1.61, 95% CI [1.22–2.12] were both at a higher risk of experiencing housing instability compared to participants who never transitioned. Furthermore, participants who attempted suicide without GICE exposure (aOR = 2.28, 95% CI [1.69–3.07]), had initial suicide attempt before GICE (aOR = 3.17, 95% CI [1.86–5.41]), or had an initial suicide attempt post-GICE (aOR = 3.08, 95% CI [1.95–4.86]) were all more likely to experience housing instability over the past year than participants who never attempted suicide and were never exposed to GICE.

Interpersonal outcomes

Past 12-month sexual assault

Almost one in ten participants reported experiencing sexual assault (9.4%) over the past year. There was a significant curvilinear relationship between age and past year sexual assault whereby the odds of sexual assault decreased until ∼64-years, at which point it gradually increased. Transgender men had lower odds of past year sexual assault (aOR = .45, 95% CI [.30–.67]) relative to nonbinary AMAB participants whereas biracial participants experienced higher odds of past year sexual assault relative to their White counterparts (aOR = 2.02; 95% CI [1.36–3.00]). In addition, participants making <$10,000 (aOR = 1.96, 95% CI [1.28–3.00]) and between $10,000 and $24,999 (aOR = 1.60, 95% CI [1.14–2.23]) had higher odds of experiencing past year sexual assault vs. those who make more than $100,000. Regarding GICE and suicide attempts, participants who had GICE exposure but no suicide attempt history (aOR = 2.20; 95% CI [1.69–2.85]), attempted suicide initially pre-GICE (aOR = 2.91; 95% CI [1.75–4.84]), and attempted suicide initially post-GICE (aOR = 3.68; 95% CI [2.36–5.72]) all experienced a higher risk of past year sexual assault relative to participants who did not attempt suicide or experience GICE (Table 4).

Table 4.

Interpersonal outcomes.

	Past 12-month Sexual assault	Lifetime Emotional IPV	Lifetime Physical IPV
	aOR (SE)	aOR (SE)	aOR (SE)
Age	.88 (.0232)***	1.09 (.0221)***	1.09 (.0219)***
Age2	1.00 (.0003)***	1.00 (.0002)***	1.00 (.0002)***
Gender identity
Nonbinary (AMAB)	Ref	Ref	Ref
Crossdresser	.55 (.2010)	1.10 (.2255)	.84 (.1759)
Transgender woman	1.11 (.2301)	1.17 (.1636)	.82 (.1147)
Transgender man	.45 (.0908)***	1.28 (.1835)	1.28 (.1816)
Nonbinary (AFAB)	.65 (.1101)*	1.19 (.1624)	1.19 (.1611)
Sexual orientation
Heterosexual	Ref	Ref	Ref
Asexual	.75 (.2096)	.83 (.1356)	.78 (.1167)
Bisexual	.95 (.2197)	1.06 (.1300)	.96 (.1156)
Lesbian/Gay/SGL	.88 (.2263)	1.15 (.1427)	1.02 (.1253)
Pansexual	1.13 (.2627)	1.57 (.1862)***	1.42 (.1664)**
Queer	1.26 (.2563)	1.40 (.1649)**	1.41 (.1641)**
Other	1.26 (.3153)	1.60 (.2658)**	1.53 (.2465)**
Race
White	Ref	Ref	Ref
American Indian/Native American	2.27 (.7055)**	2.06 (.5120)**	2.24 (.5268)***
Asian American	1.68 (.7039)	1.02 (.2374)	.64 (.1101)**
Biracial/multiracial	2.02 (.4075)***	1.55 (.3333)*	1.70 (.3982)*
Black	1.16 (.2572)	.97 (.1489)	.95 (.1418)
Latinx	1.04 (.1971)	.93 (.1241)	1.00 (.1245)
Relationship status
Single	Ref	Ref	Ref
Partner	1.05 (.1265)	1.37 (.1017)***	1.73 (.1255)***
Educational attainment
Less than high school	1.43 (.3703)	1.07 (.2139)	1.07 (.2053)
High school	.90 (.1350)	.95 (.0902)	1.03 (.0980)
Some college	1.06 (.0980)	1.05 (.0584)	1.16 (.0665)**
College degree	Ref	Ref	Ref
Employment status
Employed	Ref	Ref	Ref
Unemployed	1.31 (.2280)	1.21 (.1456)	1.16 (.1338)
Out of labor force	.78 (.1276)	.93 (.0960)	.86 (.0905)
Unspecified	1.66 (.9666)	.70 (.2499)	1.04 (.2881)
Household income
No income	1.39 (.3862)	.92 (.2618)	.94 (.2314)
<$10,000	1.96 (.4251)**	1.25 (.2091)	.99 (.1555)
$10,000–$24,999	1.60 (.2744)**	1.29 (.1475)*	1.56 (.1810)***
$25,000–$49,999	1.50 (.2784)*	1.24 (.1296)*	1.37 (.1446)**
$50,000–$99,999	1.11 (.1869)	1.17 (.1220)	1.12 (.1196)
$100K+	Ref	Ref	Ref
Unknown	1.50 (.3809)	1.10 (.1636)	1.28 (.2003)
De-/transition status
Never transitioned	Ref	Ref	Ref
Never detransitioned	1.16 (.2053)	1.10 (.0950)	1.14 (.1018)
Detransitioned	1.62 (.3433)*	1.65 (.2292)***	1.73 (.2428)***
GICE and suicide attempt history
No GICE no suicide attempt	Ref	Ref	Ref
No suicide attempt w/GICE	1.91 (.4980)*	1.78 (.2746)***	1.36 (.1823)*
Suicide attempt No GICE	2.20 (.2928)***	2.14 (.1793)***	2.19 (.1773)***
Suicide attempt pre-GICE	2.91 (.7537)***	4.98 (1.0677)***	4.02 (.8185)***
Suicide attempt post-GICE	3.68 (.8290)***	4.30 (.8671)***	4.52 (.8886)***
_cons	.77 (.4321)	.04 (.0186)***	.05 (.0196)***

SGL: same-gender loving; IPV: intimate partner violence.

Note. Linearized standard errors reported. *p ≤ .05, **p ≤ .01, ***p ≤ .001.

Lifetime emotional IPV

Another common outcome for TNB participants was emotional IPV, with 43.3% indicating experiencing emotional IPV. Demonstrating a significant curvilinear relationship, the odds of emotional IPV increased until the age of 46, after which the odds of lifetime emotional IPV decreased. Further, participants with pansexual (aOR = 1.57, 95% CI [1.24–1.98]), queer (aOR = 1.40, 95% CI [1.12–1.77]), and “other” (aOR = 1.59, 95% CI [1.15–2.21]) sexual orientations had higher odds of lifetime emotional IPV compared to their heterosexual counterparts. Native American/American Indian participants had 2.1 times higher odds of lifetime emotional IPV relative to White participants (95% CI [1.26–3.35]). Relationship status was also a significant correlate of emotional IPV whereby partnered participants experienced higher odds relative to single participants (aOR = 1.37; 95% CI [1.18–1.58]). Participants who detransitioned were at a higher risk of experiencing emotional IPV over the past year compared to those who never transitioned (aOR = 1.65, 95% CI [1.26–2.17]).

In comparison with the no GICE no suicide group, higher odds were observed among GICE exposed participants who never attempted suicide (aOR = 1.78, 95% CI [1.32–2.41]), attempted suicide without GICE exposure (aOR = 2.14, 95% CI [1.82–2.52]), attempted suicide initially before GICE (aOR = 4.98, 95% CI [3.27–7.58]), and attempted suicide initially post-GICE (aOR = 4.30, 95% CI [2.89–6.38].

Lifetime physical IPV

The prevalence of physical IPV was slightly lower than that of emotional IPV, with 40.8% of the participants experiencing this outcome. Similar to emotional IPV, physical IPV odds had a significant curvilinear relationship with the age of the participants, increasing until the age of 46 and proceeding to decrease after. Higher physical IPV odds were observed among participants with pansexual (aOR = 1.42, 95% CI [1.13–1.79]), queer (aOR = 1.41, 95% CI [1.13–1.77]), and “other” (aOR = 1.53, 95% CI [1.12–2.10]) sexual orientations relative to their heterosexual counterparts. Native American/American Indian participants were 2.2 times (95% CI [1.41–3.55]) as likely as their White counterparts to report experiencing physical IPV over the past year whereas Asian TNB participants had lower odds (aOR = .64, 95% CI [.45–.89]). Furthermore, higher physical IPV odds were also observed among partnered participants compared to single participants (aOR = 1.73, 95% CI [1.50–1.99]) and among those with a household income between $10,000 and $24,999 (aOR = 1.56, 95% CI [1.24–1.96]) and between $25,000 and $49,999 (aOR = 1.37, 95% CI [1.11–1.69]) relative to participants who reported a household income higher than $100,000. Likewise, participants who detransitioned were 1.7 times more likely to experience physical IPV in comparison with participants who never transitioned (95% CI [1.31–2.28]).

Participants who experienced GICE without a suicide attempt history were not at a higher risk of experiencing physical IPV relative to participants who never experienced either (aOR = 1.36, 95% CI [1.05–1.77]). On the contrary, participants with a suicide attempt history in the absence of GICE (aOR = 2.19, 95% CI [1.87–2.57]) and those with a pre-GICE initial suicide attempt history (aOR = 4.02, 95% CI [2.70–5.99] had higher odds of experiencing physical IPV compared to those who did not experience either. The highest observed risk relative to the no GICE no suicide group was among participants who attempted suicide initially post-GICE (aOR = 4.52, 95% CI [3.08–6.65]) (Figure 3).

Figure 3.

Odds of experiencing interpersonal vulnerabilities based on GICE exposure and suicide attempt history.

Discussion

Our study is among the first to examine how GICE efforts may be associated with the health and well-being of TNB persons living in the U.S. while accounting for the timing of their potential serious mental health concerns in relation to their GICE exposure. Although a nuanced relationship was identified among TNB persons related to their experiences with GICE and initial suicide attempts, a clear pattern emerged across domains of outcomes. Overall, compared to participants who had neither experienced GICE nor reported ever attempting suicide, participants who experienced their initial suicide attempt post-GICE as well as those who reported attempting suicide in the absence of GICE exposure were the only groups that consistently demonstrated a higher risk for all outcomes across the three domains. Crucially, the odds for all outcomes were higher among the former group. Perhaps, this may be indicative of a particularly unfavorable risk profile associated with experiencing serious mental health concerns, such as suicide attempts only after GICE. Although we do not have the necessary data to link GICE exposure and the subsequent suicide attempts of the participants, GICE may potentially play an insidious role in patterning harmful outcome trajectories for exposed participants even when contrasting with the outcomes of those with serious mental health concerns in the absence of GICE. Participants who experienced GICE after their initial suicide attempt were also at a higher risk for most of the outcomes, often at rates similar to or higher than participants who reported initial suicide attempts post-GICE. Despite some variation in odds across the domains based on temporal order, these overall patterns point to particularly hazardous implications of combining GICE and poor mental well-being.

Health-related findings

Across all health variables, participants who reported both GICE exposure and a previous suicide attempt had higher odds of negative health outcomes than participants who reported either GICE or a past suicide attempt, as well as those who reported neither GICE nor a previous suicide attempt. This adds additional support to positions held by many professional organizations (e.g. AMA, 2019; APA, 2021) and researchers (e.g. Turban et al., 2020) regarding the harms of GICE.

Furthermore, the order that participants experienced GICE in reference to their initial suicide attempt significantly impacted the odds of negative health outcomes. We found that attempting suicide initially pre-GICE was associated with the highest odds of both healthcare avoidance and psychological distress, suggesting that experiencing additional minority stress in the form of GICE non-affirmation may be particularly harmful to those with a previous history of suicidality. Although further research is needed to understand the mechanisms of risk among those who experienced GICE after a suicide attempt, gender non-affirmation has been associated with both healthcare avoidance (e.g. Boyer et al., 2022) and psychological distress (e.g. Lett et al., 2022) in previous literature, and it is likely that experiencing GICE after a suicide attempt further exacerbates an already high baseline of distress. This pattern was reversed for extra-medical use of prescription substances, with individuals who first attempted suicide after GICE having the highest odds. Previous work has identified family rejection as a predictor for both GICE (Veale et al., 2022) and substance abuse (Klein & Golub, 2016), so it is possible that participants in this group experienced elevated levels of rejection, though additional research is needed to explore these possible relationships.

Socio-structural findings

The order of one’s suicide attempts and their exposure to GICE was associated with different trajectories in socio-structural outcomes. Higher levels of bathroom avoidance were observed among persons who attempted suicide without GICE exposure and among those who initiated suicide attempts post-GICE. Previously, psychological distress and suicidality have been linked to bathroom avoidance among TNB persons (Lerner, 2021). Our findings indicate that attempting suicide following GICE exposure is linked to higher levels of bathroom avoidance than attempting suicide alone, highlighting that although bathroom avoidance may still be experienced in the absence of GICE among TNB persons, the combination of the two is certainly not predictive of a more favorable outcome. This may be due to the non-affirming and stigmatizing nature of GICE, which has been linked to bathroom avoidance (Andrade & Campo Redondo, 2022; Cramer et al., 2008; Lerner, 2021). To our surprise, we did not identify similar trends among TNB participants who reported an initial suicide attempt before their GICE experience. It is not entirely clear why experiencing GICE following an initial suicide attempt is not also linked to increased bathroom avoidance risk. It may potentially be related to the lower sample size of this particular group, but more work is necessary to obtain a more complete understanding of how GICE exposure among those with preexisting poor mental well-being may contribute to the avoidance of public bathrooms.

Further, suicidality alone and in combination with GICE exposure was indicative of higher housing instability risk, with those who experienced both GICE and suicidality reporting higher odds regardless of the order. The relationship between GICE exposure and housing instability can perhaps be explained by GICE exposure’s role in the running away experiences of TNB persons (Campbell & Rodgers, 2023). Considering that TNB persons report housing instability due to running away from home when faced with unsupportive family environments (DeChants et al., 2021) and the previously documented contribution of GICE exposure to the risk of running away, it is plausible that one pathway from GICE exposure to housing instability goes through running away. This may be especially applicable to TNB persons who have a history of suicide attempts before or after their GICE exposure.

Interpersonal findings

The highest rates of negative interpersonal outcomes—past-year sexual assault, lifetime emotional IPV, and lifetime physical IPV—were among respondents who had experienced both GICE and suicidality with the results suggesting that the trajectory of experiencing GICE first, followed by suicide attempts is associated with particularly high interpersonal vulnerability. Not surprisingly, experiencing either GICE or suicidality singularly was also associated with elevated risks compared to experiencing neither, but these risks were attenuated compared to experiencing both.

Given that GICE predicts higher rates of internalized transphobia (Veale et al., 2022), and that transphobia is linked to IPV vulnerability (Goldenberg et al., 2018; Grossman & D’Augelli, 2006), it seems feasible that one potential mechanism underlying the relationship between GICE and its association with negative interpersonal outcomes may very well be internalized transphobia. This mediational pattern would be in line with recent findings demonstrating internalized transphobia’s mediating effect between GICE and psychological distress (Veale et al., 2022) and the association between GICE and increased self-shame as well as decreased self-esteem (Dehlin et al., 2015). One additional possibility explaining the complex relationship between GICE, suicidality, and IPV is that they are mutually reinforcing, similar to the relationship between IPV and other indicators of social marginalization that Poteat et al. (2016) suggested.

Similar to The Trevor Project’s (2019) findings that young TNB people who experienced GICE during their adolescence were more likely to report suicidality as young adults than their counterparts who had not experienced GICE, our results suggest that the order of these experiences matter with GICE preceding suicidality being associated with highest likelihood of negative interpersonal relational outcomes for two of the three outcomes. However, our findings also suggest that there are likely multiple trajectories associated with increased risks, underscoring the importance of banning GICE and ensuring access to suicide prevention services for TNB people.

Limitations

The study’s findings should be considered within the context of its limitations. Data utilized in the current study is cross-sectional. As such, causality cannot be inferred from the findings. Further, participants for whom the temporal precedence of age at the time of initial GICE exposure and initial suicide attempt could not be determined were dropped from analyses. It is plausible that patterns in outcomes may have been different if the data on temporal precedence were available for all participants who previously experienced GICE and attempted suicide. Contrary to some previous work (Heiden-Rootes et al., 2022; Turban et al., 2020), we did not disaggregate GICE exposure based on the setting in which it was administered (i.e. religious vs. non-religious). Future studies should examine whether the administration of GICE in religious vs. non-religious settings is associated with differing outcomes. It is also important to mention that the category of never transitioning is comprised of a mixture of TNB participants, including some who had not transitioned because of choice as well as others who had not transitioned due to inability to do so. Relatedly, given the way the response options were provided, it is unclear whether respondents may have interpreted the “never transitioned” category to specifically mean medical transitioning or if they also considered legal (e.g. name/gender change on documents), and social (e.g. name/pronoun change socially) transitioning. Lastly, given the age of the dataset and the changes in the context of TNB social and health policy since 2015 in the U.S., it would be valuable to examine how the patterns in outcomes would differ if the underlying data were more recent.

Conclusion

This study builds on previous work, including Turban et al. (2020), to better understand the relationships between GICE, suicide attempt history, and negative health outcomes. To our knowledge, this study is the first to explore the temporal relationship between GICE and suicide attempt history, finding that participants who reported both experiencing GICE and a suicide attempt had higher odds of negative outcomes across all variables than those who did not. This pattern persisted regardless of whether GICE exposure occurred before or after the initial suicide attempt, adding further evidence regarding the harms of GICE. More specifically, those who attempted suicide pre-GICE had the highest odds of psychological distress, healthcare avoidance, emotional IPV, and housing instability, while those who first attempted suicide post-GICE had the highest odds of extra-medical substance use, bathroom avoidance, sexual assault, and physical IPV. Further research is needed to better understand the mechanisms behind GICE harm and negative health outcomes, but our findings—along with those of previous researchers—underscore the importance of banning GICE and supporting TNB-affirming policies.

Statement on human rights and on the welfare of animals

This article does not contain any studies of human participants or animals performed by any of the authors.

Funding Statement

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Disclosure statement

No potential conflict of interest was reported by the author(s).