ABSTRACT Title of Dissertation: EARLY ADOLESCENT ROMANTIC EXPERIENCES AND PSYCHOSOCIAL FUNCTIONING IN SEXUAL MINORITY YOUTH Samantha Hubachek, Doctor of Philosophy, 2023 Dissertation directed by: Professor Lea Rose Dougherty Department of Psychology Sexual minority youth are at greater risk for negative psychosocial outcomes starting in early adolescence, in part due to stress related to stigma and discrimination. We examined early adolescent romantic involvement as a potential risk or protective factor for the development of psychosocial concerns in sexual minority youth using two independent samples of youth assessed in early adolescence. Study 1 utilized prospective, longitudinal data from the Stony Brook Temperament Study (SBTS) to examine associations between youth romantic experiences, sexual orientation, and psychosocial functioning from early (age 12) to middle (age 15) adolescence (N=392; n=348 heterosexual youth, n=44 sexual minority youth). Study 2 utilized cross- sectional data from the Adolescent Brain Cognitive Development (ABCD) Study, a study with a larger and more diverse sample, to further examine the interaction between romantic experiences and sexual orientation in association with psychiatric symptoms in early adolescence (age 12), as well as whether interactive effects of romantic experiences and sexual orientation on psychiatric symptoms vary based on race/ethnicity or socioeconomic status (N=7184; n=6633 heterosexual youth, n=551 sexual minority youth). Although heterosexual and sexual minority youth were equally likely to have romantic and sexual experience in the SBTS sample, sexual minority youth were more likely to participate in romantic and sexual activity in early adolescence than their heterosexual peers in the ABCD sample. Across both samples, romantic experience in early adolescence was associated with poorer concurrent and subsequent psychosocial outcomes for sexual minority youth. Further, the interactive effects of romantic experience and sexual orientation in association with psychosocial outcomes did not vary based on race/ethnicity or parental education in the ABCD sample. These findings begin to characterize early adolescent romantic experiences in sexual minority youth and suggest that romantic involvement during this period may be linked to psychosocial concerns. This work may inform future clinical interventions targeting mental health in sexual minority youth. EARLY ADOLESCENT ROMANTIC EXPERIENCES AND PSYCHOSOCIAL FUNCTIONING IN SEXUAL MINORITY YOUTH by Samantha Hubachek Dissertation submitted to the Faculty of the Graduate School of the University of Maryland, College Park, in partial fulfillment of the requirements for the degree of Doctor of Philosophy 2023 Advisory Committee: Professor Lea R. Dougherty, Chair Edward Lemay, Ph.D. Ethan Mereish, Ph.D. Kirsty Clark, Ph.D., MPH Devon Payne-Sturges, Ph.D., Dean’s Representative i © Copyright by Samantha Hubachek 2023 ii Acknowledgements Support for this research was provided by funding through NIMH R01 MH069942 (Klein) and NIMH R01 MH122487-01A1 (Wiggins, Dougherty). iii Table of Contents Acknowledgements ....................................................................................................... ii Table of Contents ......................................................................................................... iii List of Tables ............................................................................................................... iv List of Figures ............................................................................................................... v Chapter 1: Background ................................................................................................. 1 Chapter 2: Methods ..................................................................................................... 13 Chapter 3: Results ....................................................................................................... 25 Chapter 4: Discussion ................................................................................................. 38 Appendices .................................................................................................................. 69 Bibliography ............................................................................................................... 85 This Table of Contents is automatically generated by MS Word, linked to the Heading formats used within the Chapter text. iv List of Tables Table 1. Sample characteristics from the Stony Brook Temperament Study (SBTS; Study 1). Table 2. Bivariate correlations among all study variables in the SBTS sample (Study 1). Table 3. Sample characteristics from the Adolescent Brain Cognitive Development Study (ABCD; Study 2). Table 4. Distribution of variables in the ABCD sample by sexual orientation and gender identity (Study 2). Table 5. Concurrent associations between youth romantic experiences and psychiatric symptoms: Moderation by sexual orientation (ABCD sample; Study 2). v List of Figures Figure 1. Interaction between youth sexual orientation and romantic experience at age 12 in association with disruptive behavior disorder symptoms at age 12 (SBTS; Study 1). Figure 2. Interaction between youth sexual orientation and romantic experiences at age 12 in association with peer stress at age 12 (SBTS; Study 1). Figure 3. Interaction between youth sexual orientation and age 12 romantic experiences in association with family stress at age 12 (SBTS; Study 1). Figure 4. Interaction between youth sexual orientation and age 12 romantic experiences in association with KSADS depressive symptoms at age 15 (SBTS; Study 1). Figure 5. Interaction between youth sexual orientation and age 12 romantic experiences in association with CDI depressive symptoms at age 15 (SBTS; Study 1). Figure 6. Interaction between youth sexual orientation and age 12 romantic experiences in association with KSADS disruptive symptoms at age 15 (SBTS; Study 1). Figure 7. Interaction between youth sexual orientation and age 12 romantic experiences in association with family stress at age 15 (SBTS; Study 1). Figure 8. Interaction between youth sexual orientation and age 15 romantic experiences in association with KSADS disruptive behavior disorder symptoms at age 15 (SBTS; Study 1). Figure 9. Interaction between youth sexual orientation and romantic experience in association with internalizing problems (ABCD; Study 2). Figure 10. Interaction between youth sexual orientation and romantic experience in association with externalizing problems (ABCD; Study 2). 1 Chapter 1: Background Sexual minority youth (e.g., those who identify as lesbian, gay, bisexual, or queer or who report same-sex romantic attractions or behavior) are at elevated risk for poorer mental and physical health than their heterosexual peers across the lifespan (Fish et al., 2020). This disparity emerges in early adolescence and is due in part from the fact that sexual minority youth must cope with identity-specific stressors, such as discrimination based on their sexual orientation, in addition to typical developmental stressors (Meyer et al., 2021; Pachankis et al., 2022). Thus, a major focus of research is to identify potential risk and protective factors in the development of psychopathology for sexual minority youth (Clark, 2020; Fish et al., 2020). Limited work has examined adolescent romantic relationships as one such factor and produced mixed results: some studies have suggested that romantic involvement may promote positive outcomes for sexual minority youth (Baams et al., 2014; Bauermeister et al., 2010; Glover et al., 2009; Whitton et al., 2018, 2021), while other studies have suggested that it may be detrimental (Bauermeister et al., 2010; Russell & Consolacion, 2003; Udry & Chantala, 2002; Whitton et al., 2018). Yet no study has examined links between sexual minority romantic involvement and psychosocial functioning in early adolescence, despite it being a formative period for both sexual minority youth identity development and mental health (Fish et al., 2020; Pachankis et al., 2022). Given that research in heterosexual early adolescents suggests that engaging in romantic activity at atypically early ages (prior to age 13) may increase risk for negative outcomes , a critical next step is to examine whether romantic involvement in early adolescence may uniquely incur risk for or be protective against 2 negative psychosocial outcomes for sexual minority youth in early adolescence. This study sought to examine these associations using two independent samples of youth assessed in early adolescence. Sexual minority youth mental health: A minority stress perspective Minority stress theory posits that health disparities for sexual minority youth can be explained in part by chronic stress caused by experiences and expectations of heteronormative stigma (Meyer et al., 2021). Sexual minority youth navigate ‘external’ sources of minority stress at the interpersonal level, including identity- based bullying from peers or discrimination from parents (Meyer et al., 2021; Pachankis et al., 2022), and the structural level, including residing in an area with anti-LGBTQ societal attitudes, institutional practices, or legislation (Goldbach et al., 2023; Hatzenbuehler et al., 2017; Swank et al., 2012). They also navigate ‘internal’ sources of minority stress caused by exposure to such external stressors, including worry about being rejected due to one’s sexual orientation, internalized stigmatizing beliefs about one’s sexual orientation, or the need to conceal one’s sexual orientation (Meyer et al., 2021). These and other sources of minority stress have been linked to poor psychosocial outcomes for sexual minority individuals (Baams et al., 2015; Hatzenbuehler & Pachankis, 2016; Pachankis et al., 2015, 2022). Further, research has suggested that minority stressors differentially impact youth of different sexual identities; For example, bisexual individuals report greater identity uncertainty and concealment, as well as a weaker sense of connection to the LGBT community, placing them at greatest risk for negative outcomes (Chan et al., 2020). In addition to coping with minority stressors, sexual minority youth must manage stress related to 3 universal developmental tasks in adolescence, such as navigating interpersonal relationships with peers, additional responsibilities at home, and increasing demands at school (Clark, 2020; Meyer et al., 2021). Sexual orientation identity development presents both universal (e.g., experiencing sexual attraction, pursuing sexual and romantic involvement) and unique (e.g., questioning and identifying one’s orientation, coming out, coping with stigma related to romantic involvement) developmental stressors for sexual minority youth (Hall et al., 2021). By examining how sexual minority and heterosexual youth may differ or converge in their coping with such developmental tasks, research may aid in identifying whether sexual minority youth have unique needs that should be addressed through distinct interventions to support their navigation of both universal developmental stressors and minority stressors. Developmental timing of adolescent romantic relationships Adolescent romantic involvement is a universal task in youth sexual orientation identity development that provides youth with friendship, support, intimacy, and sexuality, while also fostering independence from one’s parents and closer relationships with peers (Collins et al., 2009; Furman et al., 2009). Romantic involvement in heterosexual adolescents has been linked to positive outcomes (e.g., development of romantic competence and social acceptance from peers) as well as negative outcomes (e.g., depression, anxiety, eating pathology, substance use, and delinquency) in part due to quality of romantic relationships (Beckmeyer et al., 2018; Davila et al., 2016; Furman et al., 2009; Kansky & Allen, 2018; Mirsu-Paun & Oliver, 2017). Normative romantic experiences typically begin with initial romantic 4 attraction and formation of preliminary relationships in early adolescence, followed by exploration and solidification of romantic relationships in middle and late adolescence (Collins, 2003; Collins et al., 2009). As youth progress in their romantic development, they develop appropriate coping skills to manage the emotional ups and downs of navigating romantic relationships (Collins & Laursen, 2004; Connolly & McIsaac, 2009). However, research has supported that youth who deviate from the typical romantic developmental pathway and are involved in intimate relationships at atypically early ages (i.e., prior to age 13) may lack the interpersonal tools necessary to cope with the demands of romantic relationships and the support provided by peers navigating similar experiences, thus placing them at increased risk for negative outcomes (Connolly & McIsaac, 2009; Davila et al., 2009; Hubachek et al., in preparation, Starr et al., 2012). However, it is unclear whether the negative consequences of early adolescent romantic activity persist into later adolescence and adulthood (Connolly et al., 2013; Natsuaki et al., 2009). Research suggests that sexual minority youth experience initial sexual attraction at the same time on average as their heterosexual peers but initiate sexual activity and romantic relationships later on average than their heterosexual peers in later adolescence and early adulthood (Hall et al., 2021). This discrepancy may result from the additional stressors sexual minority youth must navigate as they process feelings of attraction and question their sexual orientation in early adolescence, explore their sexuality and potential identity labels in middle to late adolescence, and cope out and manage related stigma in late adolescence and emerging adulthood (Hall et al., 2021). However, sexual minority youth are also at elevated risk for childhood sexual abuse, which has been linked to 5 earlier sexual initiation (Friedman et al., 2011). For sexual minority youth engaging in early romantic and sexual activity, minority stress experiences may compound the stress of navigating romantic relationships without support, impacting short- and long-term risk for negative outcomes. Thus, it is a critical next step for research to examine early romantic relationships in sexual minority youth and how they relate to risk for negative outcomes. Adolescent romantic relationships and well-being in sexual minority youth A small body of work has examined adolescent romantic involvement in sexual minority youth, and its findings have been mixed (Baams et al., 2014; Bauermeister et al., 2010; Glover et al., 2009; Russell & Consolacion, 2003; Udry & Chantala, 2002; Whitton et al., 2018, 2021). Two studies have found positive associations between romantic relationships with a same-sex partner and youth well- being (Bauermeister et al., 2010; Glover et al., 2009). Same-sex romantic activity was linked to greater relational esteem and less relational depression in sexual minority youth (N=82 youth ages 14-21; Glover et al., 2009), and greater self-esteem for sexual minority boys and less internalized homophobia for sexual minority girls (N=350 youth ages 15-19; Bauermeister et al., 2010). Other studies have examined romantic involvement in sexual minority youth without assessing the sex of their romantic partner (Baams et al., 2015; Whitton et al., 2018, 2021). These studies have similarly linked romantic involvement to fewer depressive symptoms, and less alcohol and cannabis use problems in sexual minority youth, as well as less anxiety in Latinx youth (N= 488 youth assigned female at birth ages 16-31; Whitton et al., 2021), and less psychological distress in Black gay and lesbian youth (N= 248 youth 6 ages 16-20; Whitton et al., 2018). Moreover, this work has demonstrated that adolescent romantic involvement buffers against the negative impacts of minority stressors, such as expected rejection (N=309 youth ages 16-24; Baams et al., 2014) and victimization (N=248 youth ages 16-20; Whitton et al., 2018), on sexual minority psychological wellbeing. While it is unclear whether associations between romantic involvement and psychosocial functioning differ for sexual minority youth with a same- or other-sex partner (Bauermeister et al., 2010; Glover et al., 2009; Russell & Consolacion, 2003; Udry & Chantala, 2002), these findings indicate that a romantic partner may provide sexual minority youth with an additional source of support, in turn, improving psychosocial outcomes for those youth (Whitton et al., 2018). Conversely, adolescent romantic involvement has also been linked to poorer outcomes for sexual minority youth (Bauermeister et al., 2010; Russell & Consolacion, 2003; Whitton et al., 2018). Whitton et al. (2018) observed greater psychological distress in bisexual youth in romantic relationships (N= 248 youth ages 16-20). Further, participating in a romantic relationship with an other-sex partner was associated with greater internalized homophobia in sexual minority boys (N=350 youth ages 15-19; Bauermeister et al., 2010), and participating in both same-sex and other-sex romantic relationships was associated with delinquency and substance use for boys and girls, as well as depression and suicidality for girls (N=18,799 seventh to twelfth grade youth; Udry & Chantala, 2002). It is possible that romantic involvement with an other-sex partner or both same- and other-sex partners cause sexual minority youth to experience stress related to coming to terms with their identity or pressure to conceal their identity (Whitton et al., 2018). However, Russell and Consolacion 7 (2003) observed greater suicidality in sexual minority youth participating in romantic relationships with a same-sex partner (N=5872 tenth and eleventh grade youth) and Udry and Chantala (2002) observed greater emotional problems in boys participating in romantic relationships with a same-sex partner (N=18,799 seventh to twelfth grade youth; Udry & Chantala, 2002). These findings suggest that romantic involvement with a same-sex partner may also incur risk for sexual minority youth. It may be that youth who are romantically involved with a same-sex partner are more likely to have to reveal their sexual orientation to family and peers, thus increasing their likelihood of exposure to identity-based discrimination or victimization (Whitton et al., 2018). In sum, the current literature presents evidence that romantic involvement may have the potential to be protective or to incur risk for sexual minority adolescent youth, and provides a foundation for future research (Baams et al., 2014; Bauermeister et al., 2010; Glover et al., 2009; Russell & Consolacion, 2003; Whitton et al., 2018, 2021). First, this work has supported that associations between romantic activity and psychosocial functioning may vary by youth gender identity, race/ethnicity, and gender of romantic partner (Bauermeister et al., 2010; Glover et al., 2009; Russell & Consolacion, 2003; Whitton et al., 2018, 2021). This highlights the need to examine these and other potential moderators of associations between sexual minority romantic involvement and psychosocial functioning. For example, no study has examined the moderating role of socioeconomic status on associations between youth’s romantic relationships and wellbeing, despite links between socioeconomic status and youth psychopathology (Peverill et al., 2021), as well as stress in romantic relationships (Karney, 2021). Nevertheless, the studies that have 8 examined youth gender identity, race/ethnicity, or gender of romantic partner independently have demonstrated mixed findings (Baams et al., 2014; Glover et al., 2009; Russell & Consolacion, 2003; Whitton et al., 2018, 2021). Therefore, these factors should be examined in combination to assess their joint impact on the complex relations between romantic activity and psychosocial functioning. Further, the majority of studies in this small body of work have employed samples of sexual minority youth with ages ranging from adolescence to young adulthood, neglecting romantic involvement in the early adolescent developmental period and potentially obscuring age-related effects of romantic involvement in middle to late adolescence on psychosocial functioning (Baams et al., 2015; Glover et al., 2009; Whitton et al., 2018, 2021). Future research may build upon this work by examining relevant moderators of associations between sexual minority romantic involvement and psychosocial functioning using a prospective, longitudinal design to assess how romantic involvement may impact psychosocial functioning for sexual minority youth across adolescence. Early adolescence as a critical developmental period Early adolescence is a particularly formative and challenging time for all youth as it is characterized by dramatic biological and social changes related to pubertal development that directly impact the onset of sexual orientation identity and romantic development, as well as overall psychosocial functioning (Connolly & McIsaac, 2009). During early adolescence, youth undergo marked physical and hormonal changes linked to increased physical attraction to peers and desire to initiate romantic experiences (Collins et al., 2009); for sexual minority youth, this period is 9 also characterized by questioning one’s orientation (Hall et al., 2021). Youth also increasingly adopt stereotypically masculine and feminine behaviors that extend to romantic and sexual involvement, and demonstrate the negative attitudes towards sexual minorities (Horn et al., 2016; Mayeux & Kleiser, 2020; Russell & Fish, 2016). In turn, early adolescent sexual minority youth begin to experience greater victimization from peers, during this formative period of sexual orientation identity development (Robinson & Espelage, 2011; van Beusekom et al., 2016). Importantly, research has shown that this increase in peer victimization coincides with sexual minority youth showing poorer mental health than their heterosexual peers (La Roi et al., 2016; Pachankis et al., 2022; Schuster et al., 2015). Given that romantic partners may either be an invaluable source of support in coping with universal and minority stressors alike, or may exacerbate peer victimization and discrimination for sexual minority youth, further research must elucidate the potential impacts of early adolescent romantic involvement for sexual minority youth. It is critical that this be done at the outset of early adolescence, as youth sexual orientation identity and romantic development begin to unfold, to serve as a foundation for future research to assess relations between sexual orientation, romantic involvement, and psychosocial functioning as developmental processes influencing one another over time. Current research The current research sought to begin to characterize early adolescent romantic relationships in sexual minority youth using two independent samples. Study 1 used data from the Stony Brook Temperament Study (SBTS), a community-based prospective, longitudinal study of early temperament and risk for psychopathology 10 (Klein & Finsaas, 2017). Study 1 utilized a well-characterized sample of youth followed from early to middle adolescence, including measures of youth psychiatric symptoms and functioning, youth age, sex assigned at birth, race, and ethnicity, in addition to a comprehensive and developmentally appropriate assessment of youth sexual orientation obtained at age 15 (Badgett, 2009). Further, youth romantic experiences were assessed at both ages 12 and 15. This allowed us to examine concurrent associations between youth sexual orientation and romantic experiences at age 15, as well as to retrospectively link sexual orientation at age 15 to romantic experiences at age 12 to assess the long-term consequences of early adolescent romantic involvement (Pachankis et al., 2022). This study used a multi-method and multi-informant assessment of youth psychosocial functioning, including clinical and life stress interviews with both parents and youth, as well as youth self-report measures, reducing informant biases. In Study 1, we aimed to examine the interaction between youth romantic experiences at ages 12 and youth sexual orientation in association with concurrent psychosocial functioning at age 12 (Aim 1a) and subsequent psychosocial functioning at age 15 (Aim 1b). Next, we aimed to examine the interaction between youth romantic experiences at age 15 and youth sexual orientation in association with concurrent psychosocial functioning at age 15 (Aim 2). Given that research in heterosexual youth has linked early adolescent romantic activity with negative outcomes (Davila et al., 2009; Hubachek et al., in preparation; Starr et al., 2012) and some research in sexual minority youth has linked adolescent romantic activity with negative outcomes (Bauermeister et al., 2010; Russell & Consolacion, 2003; Whitton 11 et al., 2018, 2021), we hypothesized that associations between romantic experiences and negative psychosocial functioning would be strongest for sexual minority youth as compared to heterosexual youth. Despite this study’s strengths, including a well-characterized sample, prospective design, comprehensive and developmentally appropriate assessment of youth sexual orientation, and multi-method and multi-informant approach, it also has limitations. Given the small number of sexual minority youth in the Study 1 sample (n=44), we were underpowered to examine additional moderators in associations between sexual minority youth’s early adolescent romantic experiences and psychosocial functioning. Further, although the assessment of sexual orientation used in Study 1 was comprehensive and developmentally appropriate, it was not administered at age 12, despite evidence that youth can reliably report sexual orientation as early as ages 9-10 (Clark et al., 2020). Given these limitations, we sought to further explore sexual minority youth’s early adolescent romantic relationships using a larger, more diverse sample with information about current youth sexual orientation and gender identity in Study 2. Study 2 used data from the 2-year follow-up assessment of the Adolescent Brain Cognitive Development (ABCD) Study (Potter et al., 2022; Volkow et al., 2018; https://abcdstudy.org). The ABCD Study is a large longitudinal cohort study of early adolescent youth (ages 9-10 at baseline) recruited to reflect the gender, race/ethnicity, and socioeconomic status of the U.S. Population (Compton et al., 2019). As part of this study, youth sexual orientation has been assessed annually (baseline, 1-year follow-up assessment, and 2-year follow-up assessment). At the 2- 12 year follow-up assessment, youth additionally completed a measure of early romantic behavior. Appendix 1 presents a comparison between romantic items assessed in Study 1 (SBTS sample) and Study 2 (ABCD sample). Importantly, this data affords the opportunity to examine associations between youth romantic experiences, youth sexual orientation, and psychological functioning, as well as other potential moderators that prior work suggests may impact relations between romantic experiences, sexual orientation, and psychological functioning, such as race, ethnicity, and socioeconomic status (Karney, 2021; Peverill et al., 2021; Whitton et al., 2018, 2021). In Study 2, we aimed examine the interaction between youth romantic experiences and sexual orientation in association with youth psychiatric symptoms (Aim 3), as well as whether interactive effects of romantic experiences and sexual orientation on psychiatric symptoms varied based on race/ethnicity or socioeconomic status (Exploratory Aim). We hypothesized that associations between romantic experiences and greater psychiatric symptoms would be strongest for sexual minority youth as compared to heterosexual youth. Given that only two studies have examined youth race/ethnicity as a moderator in associations between sexual minority romantic involvement and psychosocial functioning in later adolescence, and this study is the first to examine race/ethnicity and socioeconomic status as moderators of associations between sexual minority early adolescent romantic relationships and wellbeing, we did not formulate a priori hypotheses regarding race/ethnicity and socioeconomic status. In both Study 1 and Study 2, we also reported frequency of early adolescent 13 sexual experiences. However, given the age of youth participating in the current research, the prevalence of sexual experiences was too low to conduct analyses. Chapter 2: Methods Study 1 Participants The sample included 392 (n=317 heterosexual youth and n=39 sexual minority youth with romantic experience data at age 12; n=348 heterosexual youth and n=44 sexual minority youth with romantic experience data at age 15) youth and their primary caregivers recruited from the larger longitudinal Stony Brook Temperament Study (N=609) (Klein & Finsaas, 2017). Families were assessed when the youth was ages 3, 6, 9, 12, and 15. Participants were mostly suburban-dwelling, White and non-Hispanic/Latine, middle class youth living in two-parent homes in the United States with educated parents. Youth completed self-report questionnaires and interviews about their romantic experiences, psychiatric history, and psychosocial stress at ages 12 and 15, as well as a self-report measure of sexual orientation at age 15 only. Parents additionally completed clinical interviews to assess youth stress at age 12 and psychiatric symptoms at ages 12 and 15. All procedures were approved by the Stony Brook University Institutional Review Board. Parents and youth provided informed written assent at both assessments. Measures Sexual orientation. Youth completed the Erotic Response and Orientations Scale (EROS) (Li & Hines, 2016; Storms, 1980) to assess romantic and sexual 14 attractions at age 15. Youth reported on romantic and sexual attractions (e.g., thoughts, feelings, dreams about someone) in the past 6 months on same-sex and other-sex subscales from (0=not at all to 4=almost every day). To prevent misclassification of youth who reported only very occasional attractions to either sex, youth with exclusively other-sex attractions (youth who reported 0 or 1 on the same- sex-attracted subscale and greater than 1 on the other-sex-attracted subscale; n=348, 53.50%) were classified as heterosexual youth and youth with same-sex attractions (youth who reported above 1 on the same-sex-attracted subscale, regardless of other- sex attractions; n=44, 6.8%) were classified as sexual minority youth. Youth who did not report any sexual attractions (youth who reported 0 or 1 on both the other-sex and same-sex subscales; n=25, 3.8%) were classified as non-attracted and removed from analyses. This classification is consistent with our prior work (Pachankis et al., 2022). Scale items on each subscale were normally distributed and internal consistency among the items was high (same-sex-attracted: α = 0.94; other-sex-attracted: α = 0.93). Romantic and sexual experiences. A self-report measure used in prior studies (Starr & Davila, 2009; Steinberg et al., 2006) assessed early adolescent romantic experiences at age 12 and age 15. At age 12, youth rated the lifetime frequency (1=Never, 2=Once or twice, 3=A few times, 4=Many times) with which they had engaged in various romantic experiences that are relatively common and developmentally normative in adolescence including: having ever gone on a date, having flirted with someone, having been romantically attracted to someone, and having kissed a date or romantic partner. Consistently with prior work (Starr & 15 Davila, 2009; Steinberg et al., 2006), the responses were summed to create a total Romantic Experiences scale (α=.89) to assess the frequency with which youth have engaged in romantic experiences. At age 15, adolescents rated the lifetime frequency with which they had ever gone on a date only (1=Never, 2=Once or twice, 3=A few times, 4=Many times). The other romantic experiences assessed at age 12 were not assessed at age 15. Appendices 2 and 3 present complete lists of items administered at age 12 and 15. Youth were also asked about any prior sexual experience at ages 12 and 15. At age 12, youth rated the lifetime frequency (1=Never, 2=Once or twice, 3=A few times, 4=Many times) with which they had engaged in “sexual relations more than kissing but not necessarily intercourse.” Given the low frequency of youth who endorsed this item (see Table 1), this variable was coded as 0=absent and 1=present. At age 15, youth rated the lifetime frequency (0=Never, 2=Once or twice, 3=A few times, 4=Many times) with which they had engaged in “sexual relations (more than kissing) with a significant other or with someone you are not dating”, “sexual relations (more than kissing) while drinking alcohol or doing drugs”, “sexual relations (more than kissing) with someone in another relationship or you didn’t know well,” and/or a “one time hook-up.” Given the low frequency of youth who endorsed these items (see Table 1), youth sexual experiences were coded as 0=absent and 1=present if youth endorsed any of the above items. Appendix 4 presents a complete list of items administered. Youth psychiatric symptoms. The K-SADS Present and Lifetime Version (K- SADS-PL; Kaufman et al., 1997) was used to assess Diagnostic and Statistical 16 Manual of Mental Disorders, Fifth Edition (DSM-5; American Psychiatric Association, 2013) current psychiatric symptoms at ages 12 and 15. Doctoral students in clinical psychology and a Masters level clinician administered the K-SADS separately to the parent and child; summary ratings for each symptom were derived from both reports. All cases with a lifetime diagnosis (46.6% of current sample) were reviewed in a case conference co-led by a child psychiatrist and clinical psychologist. Current DSM-5 psychiatric symptoms were rated on a three-point severity scale and summed to create dimensional scores for each of the following diagnostic categories: any depressive disorder (major depressive and dysthymic disorder) (α=0.98); any anxiety disorder (specific phobia, social phobia, separation anxiety, generalized anxiety, panic disorder, and agoraphobia) (α=0.96), and disruptive behavior disorder (DBD; oppositional defiant and conduct disorder) (α=0.95). A second rater derived ratings from 25 videotaped interviews to assess inter-rater reliability; intraclass correlation coefficients (ICCs) were acceptable (range=0.89–0.97). In addition, youth completed self-report measures of depression and anxiety symptoms and body image at ages 12 and 15 using the Children’s Depression Inventory-2 (CDI-2; Kovacs, 1983; α=0.85 at age 12, α=.83 at age 15), Screen for Child Anxiety Related Disorders (SCARED; Birmaher et al., 1999; α=0.90 at age 12, α=.93 at age 15), and Body Image scale from the Eating Disorders Inventory (EDI; Garner, 1991; α=0.56 at age 12, α=.79 at age 15), respectively. Psychosocial stress. Youth completed the UCLA Life Stress Interview (Rudolph & Hammen, 1999), an approximately 45-minute semi-structured interview for adolescents that assesses episodic/acute and chronic stress, at ages 12 and 15. 17 Parents additionally completed the interview at age 12 only. Chronic stress scores can be interpreted as reflecting social functioning (Harkness & Monroe, 2016). Three domains of functioning were assessed: academic/school, peer, and family stress. Scores from the academic (e.g., performance in classes and degree of difficulty with, and extra help needed for, schoolwork) and school behavior (e.g., problems with teachers, frequency and severity of behavior problems and fights with peers at school) sections were averaged to create the academic and behavior stress score. Scores from the close friends (e.g., number of close friends, as well as degrees of mutual satisfaction, conflict, confiding, and stability of relationships) and social (e.g., degrees of popularity, being a victim of bullying, and engagement in social activities with a wider peer group) sections were averaged to create the peer stress score. The family stress section assesses quality of relationships with family members and degree of family support, tension, and conflict; this information comprised the family stress score. Trained post-baccalaureate research assistant interviewers assessed functioning over the past year with parents and then teens. Ratings derived from both informants were assigned on a scale from 1 to 5 (higher scores indicate higher levels of functional impairment). A post-doctoral fellow independently rated functioning using audio-recordings of interviews; ICCs ranged from 0.57-0.93 (n=32-33) at age 12 and .75-.92 at age 15 (n=30). Pubertal development. Youth completed the Pubertal Development Scale (PDS; Carskadon & Acebo, 1993), a well-validated measure assessing pubertal development (e.g., body hair growth, skin changes, breast development, genitalia development, menstruation, voice deepening), at ages 12 and 15. 18 Sociodemographic variables. Youth and their parents reported on youth age, sex assigned at birth, race/ethnicity, and parent’s highest level of education. Youth sex assigned at birth, race/ethnicity, and parental education were coded as dichotomous variables. Youth sex assigned at birth was coded as 0=male and 1=female. Race/ethnicity was coded as 0=White and non-Hispanic/Latine and 1= Black, Indigenous, and People of Color. Parental education was coded as 0=neither parent graduated with a 4-year college degree and 1=at least one parent graduated with a 4-year college degree. Data analysis plan Statistical analyses were conducted in SPSS Version 27.0. First, we examined youth sexual orientation as a moderator of associations between frequency of romantic experiences at age 12 and psychiatric symptoms (depression, anxiety, disruptive behavior, and body image) and psychosocial stress (academic, peer, and family stress) at age 12 (Aim 1a) and age 15 (Aim 1b) using regression analyses. Next, we examined youth sexual orientation as a moderator of associations between frequency of romantic experiences at age 15 and concurrent psychiatric symptoms (depression, anxiety, disruptive behavior, and body image) and psychosocial stress (academic, peer, and family stress) using regression analyses (Aim 2). All significant interactions were probed using simple slopes tests according to Aiken and West (1991)’s guidelines. All analyses examining age 12 romantic experiences controlled for youth sex assigned at birth, race/ethnicity, parental education, and age and pubertal development at the age 12 visit. All analyses examining age 15 romantic experiences controlled for youth sex assigned at birth, race/ethnicity, parental 19 education, and age and pubertal development at the age 15 visit. Sensitivity analyses were conducted including peer and family stress at age 12 as covariates in associations between romantic experiences at age 12 and psychiatric symptoms at ages 12 and 15, as well as peer and family stress at age 15 as covariates in associations between romantic experiences at age 15 and psychiatric symptoms at age 15. In these analyses, peer and family stress served as proxies for exposure to minority stress from peers and family, respectively, to account for the effect of minority stress exposure on psychiatric symptoms. For analyses examining age 12 romantic experiences as a predictor of age 15 psychiatric symptoms and psychosocial stress, sensitivity analyses were conducted including age 12 psychiatric symptoms and psychosocial stress as a covariate to assess change in symptoms and functioning from age 12 to age 15. Finally, sensitivity analyses were conducted limiting the sample to youth who reported having both same-sex and other-sex attractions (n=36) to assess differences in associations between youth romantic experiences and outcomes for monosexual and bisexual youth. Study 2 Participants Participants were selected from the 2-year follow-up assessment of the ABCD Study (N=7184 youth, Mage=11.98 years, SDage=11.98; n=6633 heterosexual youth, n=551 sexual minority youth). The ABCD study is a longitudinal cohort study of 11,878 youth (ages 9-10 at baseline) recruited from U.S. schools to reflect the gender, race/ethnicity, and socioeconomic status of the U.S. Population (Compton et al., 2019; Potter et al., 2022). This study utilized data collected between 2018-2020 and 20 became available as part of release 4.0 of the ABCD Study. As part of this assessment, youth completed self-report measures of sexual orientation and early romantic behavior. Their parents completed a self-report measure about their child’s mental health. Measures. Sexual orientation. Youth were asked “Are you gay or bisexual?” to assess youth sexual orientation. Youth were provided the following possible response options: “Yes,” “Maybe,” “No,” “I do not understand this question,” or “Decline to answer.” Consistently with prior work, youth who responded “Yes” or “Maybe” were coded as sexual minority youth (n=551), and youth who responded “No” were coded as heterosexual youth (n= 6633; 0=heterosexual, 1=sexual minority; Calzo & Blashill, 2018; Clark et al., 2020). Youth who responded “I do not understand this question” or “Decline to answer” were excluded from analyses. Gender identity. To assess gender identity, youth were asked “Are you transgender?” and completed the ABCD Study Youth Gender Survey that assessed felt-gender (e.g., “how much do you feel like a girl?” and “how much do you feel like a boy?”) using a five-point scale (5=totally, 4=mostly, 3=somewhat, 2=a little, and 1=not at all). Youth gender identity were coded as a 3 level dummy-coded variable (cisgender boy, cisgender girl, and gender minority youth, with cisgender boy serving as the reference group) using youth responses from these three items. Gender minority youth include youth whose gender identity or expression does not align with their assigned sex at birth (Fish et al., 2020). In response to “Are you transgender,” youth were provided the following possible response options: “Yes,” “Maybe,” “No,” 21 “I do not understand this question,” or “Decline to answer.” Youth who responded “no” and whose assigned sex at birth is reported as “male” were coded as cisgender boys (n=3757) and youth who responded “No” whose assigned sex at birth is reported as “female” were coded as cisgender girls (n=3310). Youth who responded “Yes” or “Maybe” were coded as gender minority or whose sex at birth differs from their felt- gender (e.g., whichever gender or genders are rated as totally or mostly) were also coded as gender minority (n= 117; Calzo & Blashill, 2018). Due to the small sample of gender minority youth (117, 1.63%), this group was not further divided into gender minority youth assigned female at birth and gender minority youth assigned male at birth. Romantic and sexual experience. Youth completed the early dating items from the grade 10 Healthy Passages battery (Windle et al., 2004) to assess youth romantic experiences. Youth responded to the following two items reflecting relatively common and developmentally normative romantic experiences in adolescence: “have you ever had a boyfriend or a girlfriend?” and “Have you ever kissed a boy or a girl on the mouth?” Given that this study is the first to assess romantic experiences in early adolescent sexual minority youth and only two items were assessed, Youth Romantic Experience was coded as 0=absent and 1=present if youth endorsed either romantic experience item administered. To assess youth sexual experience, youth responded to the following items: “has a boy or girl has ever put their hands under your clothes?” and “have you ever put your hands under a boy’s or girl’s clothes?” Given that this study is the first to assess sexual experiences in early adolescent sexual minority youth and only two items were assessed, Youth Sexual 22 Experience was coded as 0=absent and 1=present if youth endorsed either sexual experience item administered. Youth psychiatric symptoms. Parents completed the Child Behavior Checklist (CBCL; Achenbach et al., 2001) to assess youth psychiatric symptoms. Parents rated 113 items about their child’s behavior on a 3-point scale (0=not at all true, 1=somewhat true, 2=very true). Parents’ scores for each item were used to create Internalizing Problems and Externalizing Problems summary scores. Pubertal development. Youth completed the Pubertal Development Scale (PDS; Carskadon & Acebo, 1993), a well-validated measure assessing pubertal development (e.g., body hair growth, skin changes, breast development, genitalia development, menstruation, voice deepening) to assess pubertal development. Sociodemographic variables. Youth and their parents reported on youth age, race/ethnicity, and parent’s highest level of education. Race/ethnicity was coded as a 6 level dummy-coded variable (Asian, Black, Hispanic/Latine, Multi-racial/Biracial, Indigenous/Other, and White) with White serving as the reference group given that it was the largest group. Due to small cell sizes, youth who identified as Black, Hispanic/Latine, Asian, Biracial or Multiracial, or Indigenous or Other were included in analyses as Black, Indigenous, and People of Color (BIPOC) and youth who identified as White were included in analyses as White youth (0=BIPOC, 1=White) in analyses examining race/ethnicity as a moderator of associations between the interactive effect of romantic experience and sexual orientation on psychiatric symptoms. Youth parental education was coded as dichotomous variables. Parental education was coded as 0=at least one parent graduated with a 4-year college degree 23 and 1= neither parent graduated with a 4-year college degree and was used as a proxy for socioeconomic status. Data analysis plan Aim 3 was assessed using multilevel mixed effects models using the lme4 package in R version 1.3 (R Core Team, 2020; Bates et al., 2015) to account for hierarchical data and missing data. Fixed effect covariates included youth age, gender identity1 (cisgender girl, cisgender boy, gender minority), race/ethnicity (Asian, Black, Hispanic/Latine, Multi-racial/Biracial, Indigenous/Other White), pubertal development, and parental education (0=at least one parent graduated college, 1=neither parent graduated college). Random effects covariates included ABCD site and family ID, such that family ID was nested within site given their hierarchical relation. This allowed us to account for differences due to geographic location and the fact that siblings from the same family are included in the sample. In addition to these covariates, youth romantic experience (0=no romantic experience, 1=romantic experience), sexual orientation (0=heterosexual, 1=sexual minority), and their interaction were entered as independent variables in the model, and youth concurrent psychiatric symptoms was entered as a dependent variable. A two-way analysis of covariance (ANCOVA) was then used to examine the interaction between youth romantic experiences and sexual orientation in association with youth psychiatric 1 Youth gender identity was examined as a covariate in primary analyses due to small group cell sizes when included as an additional moderator (n=16). The results of the three-way analysis of covariance examining the interaction between youth romantic experiences, sexual orientation, and gender identity in association with psychiatric symptoms can be found in Appendix 5. The interaction was not significantly associated with psychiatric symptoms (p>.05). 24 symptoms. Multiple comparisons were accounted for with the False Discovery Rate (FDR) correction (Benjamini & Hochberg, 1995). Additional analyses were conducted to explore the specificity and robustness of the interactive effects of youth romantic experiences and sexual orientation on psychiatric symptoms. Specifically, multilevel mixed effects models were used to assess whether the interactive effect of youth romantic experiences and sexual orientation on psychiatric symptoms varied based on race/ethnicity or socioeconomic status (as measured by parental education; 0=at least one parent graduated college, 1=neither parent graduated college). First, youth romantic experience, sexual orientation, race/ethnicity (0=BIPOC, 1=White), and their two-way and three-way interaction terms were entered as independent variables in the model, and youth psychiatric symptoms was entered as a dependent variable. Fixed effect covariates included youth age, gender identity, pubertal development, and parental education, and random effects covariates included site and family ID, such that family ID was nested within site. A three-way analysis of covariance was then used to examine the interaction between youth romantic experiences, sexual orientation, and race/ethnicity in association with youth psychiatric symptoms. Multiple comparisons were accounted for with the False Discovery Rate (FDR) correction (Benjamini & Hochberg, 1995). Next, youth romantic experience, sexual orientation, parental education, and their two- and three-way interaction terms were entered as independent variables in the model, and youth psychiatric symptoms was entered as a dependent variable. Fixed effect covariates included youth age, gender identity, pubertal development, and 25 race/ethnicity, and random effects covariates included site and family ID, such that family ID was nested within site. A three-way analysis of covariance was then used to examine the interaction between youth romantic experiences, sexual orientation, and parental education in association with youth psychiatric symptoms. Multiple comparisons were accounted for with the False Discovery Rate (FDR) correction (Benjamini & Hochberg, 1995). Chapter 3: Results Study 1 Preliminary analyses Table 1 presents the means and standard deviations of all study variables and Table 2 presents bivariate correlations among all study variables. Of the 392 participants in the current sample, 348 participants identified as heterosexual youth and 44 participants identified as sexual minority youth at age 15. At age 12, 66 (19%) heterosexual youth and 8 (20.5%) sexual minority youth reported ever having gone on a date, whereas, at age 15, 188 (54.02%) heterosexual youth and 21 (47.72%) sexual minority youth reported having done so. Ten (2.87%) heterosexual youth and 2 (4.55%) sexual minority youth had engaged in at least one sexual experience (defined as “more than kissing but not necessarily intercourse”) at age 12, whereas 94 (27%) heterosexual youth and 11 (25%) sexual minority youth had done so at age 15. There were no significant differences heterosexual youth’s and sexual minority youth’s frequency of romantic or sexual experiences at age 12 or age 15 (ps>.05). 26 Aim 1a. Concurrent associations between age 12 romantic experiences and age 12 psychiatric symptoms and psychosocial stress: Moderation by sexual orientation Youth psychiatric symptoms. Youth sexual orientation significantly moderated the association between frequency of romantic experiences and K-SADS disruptive behavior disorder symptoms at age 12 (b=1.11, SE=.37, pr=.16, p=.003), controlling for youth sex assigned at birth, race/ethnicity, parental education, youth age, and pubertal development at the age 12 assessment (see Figure 1). For sexual minority youth only, greater frequency of romantic experiences at age 12 was significantly associated with greater concurrent disruptive behavior disorder symptoms (b=1.19, SE=.34, pr=.19, p<.001), whereas for heterosexual youth, the link between frequency of romantic experiences and disruptive behavior disorder symptoms at age 12 was not significant (b=.08, SE=.15, pr=.03, p=.599). Associations between age 12 frequency of romantic experiences and age 12 K-SADS depressive symptoms (b=.17, SE=.25, pr=.04, p=.500), K-SADS anxiety symptoms (b=-.45, SE=.85, pr=-.03, p=.600), self-reported CDI depressive symptoms (b=.66, SE=.79, pr=.05, p=.405), and self-reported SCARED anxiety symptoms (b=-2.19, SE=1.56, pr=-.08, p=.163), youth-reported body image (b=-1.49, SE=.82, pr=-.10, p=.068) did not vary based on youth sexual orientation. Youth psychosocial stress. Youth sexual orientation significantly moderated the association between frequency of romantic experiences at age 12 and peer stress at age 12 (b=.29, SE=.14, pr=.12, p=.034), controlling for youth sex assigned at birth, race/ethnicity, parental education, youth age, and pubertal development at the age 12 27 assessment (see Figure 2). For sexual minority youth, the link between frequency of romantic relationships and peer stress at age 12 was not significant (b=.18, SE=.12, pr=.08, p=.160), whereas for heterosexual youth, greater frequency of romantic experiences at age 12 was significantly associated with less peer stress at age 12 (b=- .11, SE=.06, pr=-.11, p=.045). Further, youth sexual orientation significantly moderated the association between frequency of romantic experiences at age 12 and family stress at age 12 (b=.35, SE=.14, pr=.13, p=.015), controlling for youth sex assigned at birth, race/ethnicity, parental education, youth age, and pubertal development at the age 12 assessment (see Figure 3). For sexual minority youth only, greater frequency of romantic experiences at age 12 was significantly associated with greater family stress at age 12 (b=.37, SE= .13, pr=.15, p=.005), whereas for heterosexual youth, the link between frequency of romantic experiences and family stress at age 12 was not significant (b=.02, SE=.06, pr=.02, p=.701). Associations between age 12 frequency of romantic experiences and age 12 academic stress (b=.10, SE=.13, pr=.04, p=.449) did not vary based on youth sexual orientation. However, there was a significant main effect of frequency of romantic experiences at age 12 and academic stress at age 12, such that greater frequency of romantic experiences was associated with greater academic stress (b=.11, SE=.05, pr=.11, p=.040). Aim 1b. Longitudinal associations between age 12 romantic experiences and age 15 psychiatric symptoms and psychosocial stress: Moderation by sexual orientation 28 Youth psychiatric symptoms. Youth sexual orientation significantly moderated the association between frequency of romantic experiences at age 12 and K-SADS depressive symptoms at age 15 (b=1.61, SE=.37, pr=.23, p<001), controlling for youth sex assigned at birth, race/ethnicity, parental education, youth age, and pubertal development at the age 12 visit (see Figure 4). For sexual minority youth only, greater frequency of romantic experiences at age 12 was significantly associated greater K-SADS depressive symptoms at age 15 (b= 1.65, SE= .34, pr=.26, p<.001), whereas for heterosexual youth, the link between frequency of romantic experiences at age 12 and K-SADS depressive symptoms at age 15 was not significant (b=.03, SE=.15, pr=.01, p=.832). Further, youth sexual orientation significantly moderated the association between frequency of romantic experiences at age 12 and self-reported CDI depressive symptoms at age 15 (b=2.04, SE=.79, pr=.14, p=.010), controlling for youth sex assigned at birth, race/ethnicity, parental education, youth age, and pubertal development at the age 12 assessment (see Figure 5). For sexual minority youth only, greater frequency of romantic experiences at age 12 was significantly associated with greater self-reported CDI depressive symptoms at age 15 (b=2.21, SE=.72, pr=.17, p=.002), whereas for heterosexual youth, the link between frequency of romantic experiences at age 12 and self-reported CDI depressive symptoms at age 15 was not significant (b=.17, SE=.33, pr=.03, p=.605). Youth sexual orientation significantly moderated the association between frequency of romantic experiences at age 12 and K-SADS disruptive behavior disorder symptoms at age 15 (b=.86, SE=.32, pr=.15, p=.007), controlling for youth sex assigned at birth, race/ethnicity, parental education, youth age, and pubertal 29 development at the age 12 visit (see Figure 6). For sexual minority youth only, greater frequency of romantic experiences at age 12 was significantly associated with greater K-SADS disruptive behavior symptoms at age 15 (b=1.01, SE=.29, pr=.19 p<.001), whereas for heterosexual youth, the link between frequency of romantic experiences at age 12 and disruptive behavior symptoms at age 15 was not significant (b=.15, SE=.13, pr=.06, p=.258). Finally, youth sexual orientation moderated the association between frequency of romantic experiences at age 12 and youth-reported body image at age 15 (b=1.50, SE=.76, pr=.11, p=.048), controlling for youth sex assigned at birth, race/ethnicity, parental education, youth age, and pubertal development at the age 12 assessment. However, when probed, the link between frequency of romantic experiences at age 12 and youth-reported body image was not significant for same- sex-attracted (b=.38, SE=.31, pr=.07, p=.224) or other-sex-attracted youth (b=-1.11, SE=.69, pr=-.09, p=.109). Associations between age 12 frequency of romantic experiences and age 15 K-SADS anxiety symptoms (b=-.02, SE=.72, pr<-.01, p=.973) and self-reported SCARED anxiety symptoms (b=2.38, SE=1.66, pr=.08, p=.152) did not vary based on youth sexual orientation. Youth psychosocial stress. Youth sexual orientation marginally significantly moderated the association between frequency of romantic experiences at age 12 and family stress at age 15 (b=.14, SE=.07, pr=.11, p=.052), controlling for youth sex assigned at birth, race/ethnicity, parental education, youth age, and pubertal development at the age 12 visit (see Figure 7). For sexual minority youth only, greater frequency of romantic experiences at age 12 was significantly associated with greater family stress at age 15 (b=.18, SE=.07, pr=.15, p=.007), whereas for heterosexual 30 youth, the link between frequency of romantic experiences at age 12 and family stress at age 15 was not significant (b=.04, SE=.03, pr=.07, p=.198). Associations between age 12 frequency of romantic experiences and peer stress (b=.12, SE=.07, pr=.09, p=.085) and academic stress (b=-.04, SE=.07, pr=-.03, p=.595) at age 15 did not vary based on youth sexual orientation. Aim 2. Concurrent associations between age 15 dating experiences and age 15 psychiatric symptoms and psychosocial stress Youth sexual orientation significantly moderated the association between frequency of dating experiences at age 15 and concurrent disruptive behavior disorder symptoms at age 15 (b=-1.08, SE=.37, pr=-.15, p=.003), controlling for youth sex assigned at birth, race/ethnicity, parental education, youth age, and pubertal development at the age 15 visit (see Figure 8). For sexual minority youth, frequency of dating experiences at age 15 was associated with fewer concurrent disruptive behavior disorder symptoms (b=-.73, SE=.34, pr=-.11, p=.033), whereas for heterosexual youth, frequency of dating experiences at age 15 was associated with greater concurrent disruptive behavior disorder symptoms at age 15 (b=.35, SE=.13, pr=.14, p=.008). Associations between age 15 frequency of dating experiences and age 15 K-SADS depressive symptoms (b=.46, SE=.40, pr=.06, p=.249), K-SADS anxiety symptoms (b=-.28, SE=.78, pr=-.03, p=.623), self-reported CDI depressive symptoms (b=-.61, SE=.84, pr=-.04, p=.466), self-reported anxiety SCARED symptoms (b=-.82, SE=1.79, pr=-.02, p=.648), youth-reported body image (b=.34, SE=.82, pr=.02, p=.674), peer stress (b=.03, SE=.07, pr=.02, p=.672), academic stress (b=-.08, SE=.07, pr=-.06, p=.291), and family stress (b=-.08, SE=.08, pr=-.05, 31 p=.325) at age 15 did not vary based on sexual orientation. However, there was a significant main effect of frequency of dating experiences at age 15 on peer stress at age 15, such that greater frequency of dating experience was associated with less peer stress (b=-.11, SE=.03, pr=-.22, p<.001). Sensitivity analyses Sensitivity analyses were conducted including peer and family stress as additional covariates in all analyses examining psychiatric symptoms as the dependent variable. The interactions between sexual orientation and frequency of romantic experience at age 12 in association with KSADS disruptive behavior disorder symptoms at age 12 (b=.99, SE=.37, pr=.15, p=.007), KSADS depressive symptoms at age 15 (b=1.55, SE=.37, pr=.22, p<.001), CDI depressive symptoms at age 15 (b=1.69, SE=.77, pr=.11, p=.038), and KSADS disruptive behavior disorder symptoms at age 15 (b=.76, SE=.32, pr=.13, p=.017) remained significant when controlling for peer stress at age 12. The interactions between sexual orientation and frequency of romantic experience at age 15 in association with KSADS disruptive behavior disorder symptoms at age 15 (b=-1.09, SE=.37, pr=-.15, p=.003) remained significant when controlling for peer stress at age 15. The interactions between sexual orientation and frequency of romantic experience at age 12 in association with KSADS disruptive behavior disorder symptoms at age 12 (b=.90, SE=.36, pr=.14, p=.014), KSADS depressive symptoms at age 15 (b=1.59, SE=.37, pr=.23, p<.001), CDI depressive symptoms (b=1.75, SE=.79, pr=.12, p=.027), KSADS disruptive behavior disorder symptoms at age 15 (b=.67, SE=.31, pr=.12, p=.033), remained significant when controlling for family 32 stress at age 12. The interactions between sexual orientation and frequency of romantic experience at age 15 in association with KSADS disruptive behavior disorder symptoms at age 15 (b=-1.05, SE=.37, pr=-.15, p=.004) remained significant when controlling for family stress at age 15. Further, sensitivity analyses were conducted including psychiatric symptoms and psychosocial stress at age 12 as an additional covariate in all analyses examining psychiatric symptoms and psychosocial stress at age 15 as the dependent variable. By controlling for prior psychiatric symptoms and psychosocial stress, the dependent variable represents residuals, and the effects of frequency of romantic experiences at age 12 on psychiatric symptoms and psychosocial stress at age 15 reflect change in psychiatric symptoms and psychosocial stress from age 12 to age 15. The interactions between sexual orientation and frequency of romantic experience in association with KSADS depressive symptoms (b=1.60, SE=.37, pr=.23, p<.001) and CDI depressive symptoms (b=1.70, SE=.67, pr=.14, p=.012) remained significant. For sexual minority youth, greater romantic experience was associated with greater depressive symptoms on the KSADS (b=1.64, SE=.34, pr=.26, p<.001) and the CDI (b=1.57, SE=.62, pr=.14, p=.012), whereas for heterosexual youth, the links between romantic experience and depressive symptoms on the KSADS (b=.04, SE=.15, pr=.02, p=.280) and CDI (b=-.14, SE=.28, pr=-.03, p=.63) were not significant. The interactions between sexual orientation and frequency of romantic experience in association with disruptive behavior disorder symptoms (b=.30, SE=.26, pr=.06, p=.256), youth- reported body image (b=-.73, SE=.68, pr=-.06, p=.286), and family stress (b=.07, SE=.07, pr=.06, p=.275) were no longer significant. However, an additional 33 interaction emerged. Youth sexual orientation significantly moderated the association between frequency of romantic experiences at age 12 and change in self-reported SCARED anxiety symptom scores from age 12 to age 15 (b=3.62, SE=1.41, pr=.14, p=.011), controlling for youth sex assigned at birth, race/ethnicity, parental education, youth age, and pubertal development at the age 12 assessment. For heterosexual youth, greater frequency of romantic experiences at age 12 was significantly associated with a decrease in self-reported SCARED anxiety symptoms from age 12 to age 15 (b=-1.52, SE=.59, pr=-.14, p=.010), whereas for sexual minority youth, the link between frequency of romantic experiences at age 12 and change in anxiety symptoms from age 12 to age 15 was not significant (b=2.10, SE=.130, pr=.09, p=.107). Finally, sensitivity analyses were conducted limiting the sample of sexual minority youth to youth who reported having both same-sex and other-sex attractions (n=36) to assess differences in associations between youth romantic experiences and outcomes for monosexual and bisexual youth. All significant associations remained significant, and two additional significant associations emerged. First, youth sexual orientation significantly moderated the association between frequency of romantic experiences at age 12 and family stress at age 15 (b=-.15, SE=.07, pr=-.11, p=.040). For sexual minority youth, greater romantic experiences at age 12 was associated with greater family stress at age 15 (b=.19, SE=.07, pr=.15, p=.006), whereas for heterosexual youth, the link between romantic experience and family stress was not significant (b=.04, SE=.03, pr=.07, p=.221). Further, youth sexual orientation significantly moderated the association between frequency of romantic experiences at 34 age 12 and peer stress at age 15 (b=.15, SE=.07, pr=.12, p=.030), controlling for youth sex assigned at birth, race/ethnicity, parental education, youth age, and pubertal development at the age 12 visit. However, when probed, the link between frequency of romantic experiences at age 12 and youth-reported body image was not significant for same-sex-attracted (b=.11, SE=.06, pr=.09, p=.090) or other-sex-attracted youth (b=-.04, SE=.03, pr=-.08, p=.137). Study 2 Preliminary analyses Table 3 presents the means and standard deviations of all study variables and Table 4 presents associations between youth sexual and gender identity and all study variables. In our study sample (N=7184), 6633 of 7184 (92.33%) youth identified as heterosexual and 551 (7.67%) identified as gay or bisexual (i.e., sexual minority youth). 7067 (98.37%) youth identified as cisgender and 117 (1.63%) identified as transgender or as having a felt-gender that differed from their sex assigned at birth (i.e. gender minority youth). 1729 of 6633 (26.07%) of heterosexual youth and 248 of 551 (45.01%) sexual minority youth reported having romantic experience, whereas 59 of 6633 (.89%) heterosexual youth and 14 (2.54%) of 551 sexual minority youth reported having sexual experience. 1924 of 7067 (27.23%) cisgender youth and 53 of 117 (45.30%) gender minority youth reported having romantic experience, whereas 70 of 7067 (.99%) cisgender youth and 3 of 117 (2.56%) gender minority youth reported having sexual experience. Sexual minority youth were significantly more likely to report romantic, 𝜒2(1,7184)=91.52, p<.001, and sexual experience, 𝜒2(1,7164)=13.99, p<.001, than heterosexual youth. Gender minority youth were 35 significantly more likely to report romantic experience than cisgender youth, 𝜒2(1,7184)=18.85, p<.001, but there was no significant difference in likelihood of reporting sexual experience between gender minority and cisgender youth, 𝜒2(1,7164)=2.82, p=.093. Aim 3. Concurrent associations between youth romantic experiences and psychiatric symptoms: Moderation by sexual orientation Internalizing problems. Results of a two-way ANCOVA yielded a significant interaction between youth romantic experiences and sexual orientation in association with youth concurrent internalizing problems, F(1,7027.7)=5.58, p=.017), controlling for fixed effects covariates (age, race/ethnicity, pubertal development, parental education, and gender identity) and random effects covariates (site and family ID; see Table 5, Figure 9). FDR-corrected pairwise comparisons revealed that sexual minority youth with romantic experience (M=52.2, SE=.80, n=248) demonstrated greater internalizing problems than sexual minority youth without romantic experience (M=50.5, SE=.75, n=303; p=.045), heterosexual youth with romantic experience (M=47.7, SE=.59, n=1729; p<.001), and heterosexual youth without romantic experience (M=48.0, SE=.56, n=4904; p<.001). Further, sexual minority youth without romantic experience (M=50.5, SE=.75, n=303) also demonstrated greater internalizing problems than heterosexual youth with (M=47.7, SE=.59, n=1729; p<.001) and without romantic experience (M=48.0, SE=.56, n=4904; p<.001). There were no significant differences in internalizing problems between heterosexual youth with (M=47.7, SE=.59, n=1729) and without romantic experience (M=48.0, SE=.56, n=4904; p=.221). 36 Externalizing problems. Results of a two-way ANCOVA yielded a significant interaction between youth romantic experiences and sexual orientation in association with youth concurrent externalizing problems, F(1, 7012.6)=4.44, p=.035), controlling for fixed effects covariates (age, race/ethnicity, pubertal development, parental education, and gender identity) and random effects covariates (site and family ID; see Table 5, Figure 10). FDR-corrected pairwise comparisons revealed that sexual minority youth with romantic experience (M=48.8, SE=.73, n=248) demonstrated greater externalizing problems than sexual minority youth without romantic experience (M=44.8, SE=.69, n=303, p<.001), heterosexual youth with romantic experience (M=46.8, SE=.54, n=1729; p=.002), and heterosexual youth without romantic experience (M=44.5, SE=.50, n=4904, p<.001). Further, heterosexual youth with romantic experience (M=46.8, SE=.54, n=1729) demonstrated greater externalizing problems than sexual minority youth without romantic experience (M=44.8, SE=.69, n=303, p=.002) and heterosexual youth without romantic experience (M=44.5, SE=.50, n=4904, p<.001). There was no significant difference in externalizing problems between sexual minority youth without romantic experience (M=44.8, SE=.69, n=303) and heterosexual youth without romantic experience (M=44.5, SE=.50, n=4904, p=.589). Exploratory analysis. Concurrent associations between youth romantic experiences and psychiatric symptoms: Moderation by sexual orientation and race/ethnicity2 2 The complete results including two-way interaction and main effects for these analyses can be found in Appendix 6. 37 Given small group cell sizes of sexual minority youth who identified as Black (n=32 with romantic experience, n=24 without romantic experience), Hispanic/Latine (n=41 with romantic experience, 58 without romantic experience), Asian (n=1 with romantic experience, n=5 without romantic experience), Biracial or Multiracial (n=30 with romantic experience, n=36 without romantic experience), or Indigenous or Other (n=3 with romantic experience, n=1 without romantic experience), youth who identified as Black, Hispanic/Latine, Asian, Biracial or Multiracial, or Indigenous or Other were included in analyses as Black, Indigenous, and People of Color (BIPOC). Of sexual minority youth with romantic experience, 107 identified as BIPOC and 139 identified as White. Of sexual minority youth without romantic experience, 124 identified as BIPOC and 178 identified as White. Of heterosexual youth with romantic experience, 900 identified as BIPOC and 820 identified as White. Of heterosexual youth without romantic experience, 1944 identified as BIPOC and 2945 identified as White. Internalizing problems. Results of a three-way ANCOVA indicated no significant three-way interaction between youth romantic experiences, sexual orientation, and race/ethnicity (0=BIPOC, 1=White) in association with youth internalizing problems, F(1,6979.9)=2.63 , p=.105, controlling for fixed effects covariates (age, gender identity, pubertal development, and parental education) and random effects covariates (site and family ID). Externalizing problems. Results of a three-way ANCOVA indicated no significant three-way interaction between youth romantic experiences, sexual orientation, and race/ethnicity (0=BIPOC, 1=White) in association with externalizing 38 problems, F(1,6964.7)=.17, p=.678, controlling for fixed effects covariates (age, gender identity, pubertal development, and parental education) and random effects covariates (site and family ID). Exploratory analysis. Concurrent associations between youth romantic experiences and psychiatric symptoms: Moderation by sexual orientation and parental education3 Internalizing problems. Results of a three-way ANCOVA indicated no significant three-way interaction between youth romantic experiences, sexual orientation, and parental education in association with youth internalizing problems, F(1,6996.1)=.18 , p=.674, controlling for fixed effects covariates (age, gender identity, pubertal development, and race/ethnicity) and random effects covariates (site and family ID). Externalizing problems. Results of a three-way ANCOVA indicated no significant interaction between youth romantic experiences, sexual orientation, and parental education in association with externalizing problems, F(1,6983.6)=.16, p=.687, controlling for fixed effects covariates (age, gender identity, pubertal development, and race/ethnicity) and random effects covariates (site and family ID). Chapter 4: Discussion The current study examined whether associations between romantic experience and psychosocial functioning vary based on sexual orientation using two independent samples of early adolescent youth. Overall, we observed low rates of 3 The complete results including two-way interaction and main effects for these analyses can be found in Appendix 7. 39 sexual experiences for both heterosexual and sexual minority early adolescent youth. However, findings in the ABCD sample (Study 2) suggest that sexual minority youth are more likely to participate in romantic and sexual activity in early adolescence than their heterosexual peers. Furthermore, sexual minority youth with romantic experience in early adolescence demonstrated poorer concurrent and subsequent psychosocial outcomes. Our findings begin to characterize early adolescent romantic experiences for sexual minority youth and suggest that early romantic involvement may be linked to the development of psychopathology for sexual minority youth, bearing important clinical implications for sexual minority youth mental health intervention. Prevalence of early adolescent romantic experience in sexual minority youth Although heterosexual (19.00%) and sexual minority (20.50%) youth both showed low rates of early adolescent romantic experience in the SBTS sample (Study 1), it warrants attention that nearly half of sexual minority youth (45.01%) reported early adolescent romantic experience in the ABCD sample (Study 2), a larger and more diverse sample. Further, sexual minority youth were more likely to engage in initial romantic experiences (e.g., kissing, going on a date) than their heterosexual peers in the ABCD sample. While prior work has suggested that romantic involvement in early adolescence reflects a deviation from the typical trajectory of heterosexual romantic involvement (Connolly & McIsaac, 2009), these findings suggest that early adolescent romantic experience may be more common for sexual minority youth. This disparity may occur in part because romantic involvement with same- and/or other-sex partners allows sexual minority youth to explore their sexual 40 orientation during early adolescence, a critical period of sexual orientation identity development characterized by experiencing sexual orientation and questioning and seeking to identify one’s orientation (Hall et al., 2021). Sexual minority youth also experience greater peer victimization, as well as poorer mental health during early adolescence (La Roi et al., 2016; Pachankis et al., 2022; Robinson & Espelage, 2011; Schuster et al., 2015; van Beusekom et al., 2016), and may thus seek romantic partners as a source of social support. Furthermore, sexual minority youth experience childhood sexual abuse at greater rates than their heterosexual peers, which has been associated with earlier initiation of sexual activity; sexual minority youth who have experienced childhood sexual abuse may thus be more likely to engage in early romantic activity (Friedman et al., 2011). Given that early adolescent romantic experience was common and associated with poorer psychosocial functioning for sexual minority youth, these findings highlight the need to explore what factors motivate sexual minority youth to become romantically involved, as well as to provide education and intervention to support sexual minority youth in navigating romantic involvement. Romantic experience and psychiatric symptoms across early to middle adolescence: Differences between heterosexual and sexual minority youth In both the SBTS and ABCD samples, romantic involvement was associated with poorer psychiatric functioning for sexual minority youth. Results from the SBTS sample (Study 1) revealed that sexual minority youth with greater romantic experience in early adolescence demonstrated greater concurrent (age 12) and subsequent (age 15) disruptive behavior disorder symptoms. Further, sexual minority 41 youth with greater romantic experience in early adolescence demonstrated greater subsequent (age 15) depressive symptoms, even when accounting for prior depressive symptoms (age 12). Similarly, results from the ABCD sample (Study 2) demonstrated that early adolescent sexual minority youth with romantic experience showed the greatest concurrent externalizing and internalizing problems in comparison to their peers. Although research in presumably heterosexual youth has previously linked early romantic involvement with both externalizing and internalizing psychopathology (Connolly et al., 2013; Davila et al., 2016; Friedlander et al., 2007; Hubachek et al., in preparation; Joyner & Udry, 2000), our study is the first to observe this link in early adolescent sexual minority youth. It may be that sexual minority youth who are romantically involved in early adolescence as compared to later adolescence may be ill-equipped with the interpersonal skills, emotion regulation strategies, and support from peers and parents required to competently navigate the emotional highs and lows that accompany romantic experiences, placing them at greater risk for emotional and behavioral dysregulation, as has been posited in the literature examining heterosexual early adolescent romantic involvement (Connolly & McIsaac, 2009; Davila, Steinberg, et al., 2009; Grossman et al., 2018). Considering minority stress theory, these deficits in skills and support may have a greater impact on sexual minority youth than their heterosexual peers given that they must manage stressors associated with romantic involvement experienced by all youth regardless of sexual orientation, as well as minority stressors (Goldbach & Gibbs, 2017; Meyer et al., 2021; Pachankis et al., 2022). For example, sexual minority youth may experience minority stress related to exploring and coming to 42 terms with their sexual identities through romantic involvement, coping with identity- related discrimination and stigma exacerbated by being visibly romantically involved, or living in a region with anti-LGBTQ societal attitudes, institutional practices, or legislation (Goldbach et al., 2023; Goldbach & Gibbs, 2017; Hatzenbuehler et al., 2017; Meyer et al., 2021; Pachankis et al., 2022). Critically, these results suggest that neglecting to assess sexual orientation may have obscured prior findings linking romantic involvement and psychosocial functioning in presumably heterosexual youth. Our findings suggest that the links between poorer emotional and behavioral outcomes and romantic involvement in sexual minority youth persist into later adolescence. Of note, although the prospective associations between greater romantic experience and greater depressive symptoms persisted when controlling for prior symptoms, the prospective association between greater romantic experience and greater disruptive behavior symptoms did not. Thus, romantic involvement may predict increases in depressive symptoms, but not disruptive behavior symptoms between early and middle adolescence; rather, elevated disruptive behavior symptoms in early adolescence remain elevated in middle adolescence. These findings highlight the need to elucidate the mechanisms underlying relations between romantic involvement and various aspects of psychosocial functioning, such as exposure to minority stressors (Goldbach & Gibbs, 2017; Pachankis et al., 2022). Although sensitivity analyses in the SBTS sample showed that associations between early adolescent romantic involvement and psychiatric symptoms in sexual minority youth remained significant when controlling for peer and family stress, future longitudinal 43 research should examine sexual minority identity-specific stressors as mechanisms underlying associations between romantic experience and changes in psychosocial functioning to clarify how romantic involvement relates to sexual minority youth mental health across development. In contrast to associations between romantic experience in early adolescence and psychological functioning, romantic experience in middle adolescence was not associated with poorer psychological functioning for sexual minority youth; in fact, sexual minority youth with greater romantic experience demonstrated fewer concurrent disruptive behavior disorder symptoms than their heterosexual peers. These findings closely align to prior research in adolescent sexual minority youth linking romantic involvement to improved well-being and less psychological distress (Baams et al., 2014; Bauermeister et al., 2010; Glover et al., 2009; Whitton et al., 2018, 2021). Given that engaging in romantic and sexual activity are critical steps in sexual orientation identity development, it may be that romantic involvement supports exploration and affirmation of sexual minority youth’s identity when youth have the appropriate interpersonal skills and emotional maturity to do so (Hall et al., 2021). Moreover, romantic partners may also provide sexual minority youth with support in coping with both minority and universal stressors in adolescence, benefiting their mental wellbeing (Whitton et al., 2018). The current research contributes to this work by specifying a lower limit to the developmental period in which romantic involvement may be protective for sexual minority youth. Our findings suggest that early adolescence may be a sensitive period for sexual minority youth romantic development, in which youth may still be developing the coping skills 44 and relational problem-solving tools to weather the ups and downs of romantic involvement, as well as minority stress experiences. In later adolescence, however, romantic involvement may benefit sexual minority youth mental health and even protect against the negative impact of exposure to minority stress (Baams et al., 2014; Bauermeister et al., 2010; Glover et al., 2009; Whitton et al., 2018, 2021). Given that we were only able to examine romantic involvement in middle adolescence (age 15) as part of the SBTS sample, future longitudinal research should continue to explore associations between romantic involvement and psychosocial functioning in sexual minority youth across adolescence in larger and more diverse samples. Consistently with prior work (Connolly et al., 2013; Friedlander et al., 2007), our results showed that romantic involvement in early to middle adolescence comes with some risk for heterosexual youth, as well. In the ABCD sample, heterosexual youth with romantic experience demonstrated greater externalizing symptoms than heterosexual and sexual minority youth without romantic experience, although lesser externalizing symptoms than sexual minority youth with romantic experience. Furthermore, in the SBTS sample, heterosexual youth with romantic experience in middle adolescence demonstrated greater concurrent disruptive behavior disorder symptoms than their peers. These findings are consistent with prior work highlighting the link between heterosexual romantic involvement and externalizing behaviors across adolescence (Beckmeyer et al., 2018). Surprisingly, we did not observe links between early adolescent romantic experiences and internalizing behaviors in either sample, as have been observed in the literature (Davila et al., 2016). However, much of this work has not assessed youth sexual orientation, potentially obscuring complex 45 relations between early adolescent romantic involvement and internalizing psychopathology that vary based on sexual orientation (Furman & Collibee, 2014). Furthermore, these associations may be present for some heterosexual youth, but not others (Hubachek et al., in preparation). Early adolescence is characterized by dramatic social and biological changes related to pubertal development that directly impact romantic interest and vary for girls and boys (Connolly & McIsaac, 2009; Kaplowitz, 2004; Negriff & Susman, 2011). Indeed, prior work has observed greater associations between early adolescent romantic involvement and psychopathology for girls as compared to boys as well as youth with “off-time” (i.e. early or late) pubertal development (Davila et al., 2016; Hubachek et al., in preparation; Ullsperger & Nikolas, 2017). Future work should assess sexual orientation, gender identity, and pubertal development in combination in association with romantic involvement and psychosocial functioning. Romantic experience and psychosocial stress across early to middle adolescence: Differences between heterosexual and sexual minority youth Results from the SBTS sample showed that greater romantic experience in early adolescence was significantly associated with greater concurrent (age 12) family stress and marginally associated with subsequent (age 15) family stress for sexual minority youth. These findings build on the current literature linking romantic involvement with parent-child stress in presumably heterosexual youth (Davila, Stroud, et al., 2009; Gray & Steinberg, 1999), and suggest that early adolescent romantic involvement may contribute to lasting family stress for sexual minority youth, as well. Although some parents support and accept their child’s sexual 46 minority identity after they come out, many parents experience emotional conflict and distress related to their child’s sexual orientation (Chrisler, 2017; Clark et al., 2021). Further, parents of sexual minority youth who hold heterosexist and homophobic beliefs show poorer psychological adjustment and greater distress related to their relationship with their child (Hubachek et al., 2023). Parents of sexual minority youth may reject their child’s identity by distancing themselves from their child, ignoring their identity, restricting their child’s access to support and resources, attempting to change their child’s identity, or even physically abusing their child; these reactions are associated with poor mental health outcomes for sexual minority youth (Bebes et al., 2015; Savin-Williams & Dubé, 1998). Sexual minority youth who are romantically involved with a same-sex partner may be more likely to have come out to their families, increasing the likelihood of parent-child stress and conflict with parents who are rejecting of their child’s sexual orientation. Exposure to such minority stress experiences may increase risk of poorer psychosocial outcomes for sexual minority youth who are visibly romantically involved. However, it is also possible that youth with unaccepting parents and greater family stress are more likely to seek romantic experiences as a source of social support (Davila, Stroud, et al., 2009). It is notable that sensitivity analyses revealed that greater romantic experience at age 12 was significantly associated with greater family stress at age 15 for youth with both same- and other-sex attractions. Prior work has shown that bisexual individuals are more likely to report experiences of identity uncertainty, identity concealment, and other sexual minority stressors, placing them at even greater risk for negative mental health outcomes than gay and lesbian individuals (Chan et al., 2020); 47 these minority stress experiences may lead to greater family stress for romantically involved bisexual youth. Yet, prospective associations between greater romantic experience and greater family stress did not persist when controlling for prior family stress in either the complete sample of sexual minority youth or in the sample of sexual minority youth with both same- and other-sex attractions only. This suggests that sexual minority youth who are romantically involved show greater family stress in early adolescence and continue to show greater family stress in middle adolescence, as opposed to increases in family stress between early and middle adolescence. Future longitudinal research should examine sexual minority youth’s family, romantic, and psychological functioning, as well as exposure to minority stressors, across adolescence to clarify the directionality and causality of these associations. We also observed that heterosexual youth with greater romantic experience in early adolescence demonstrated less concurrent peer stress than their peers in the SBTS sample. Further, sensitivity analyses revealed that heterosexual youth with greater romantic involvement in early adolescence showed decreased anxiety symptoms between early and middle adolescence in the SBTS sample. Taken together, these findings suggest that heterosexual youth with romantic experience in early adolescence fare better than their sexual minority peers. Prior work in presumably heterosexual youth has evidenced bidirectional associations between peer and romantic relationships, such that youth who are well-liked by their peers and have higher quality friendships are more likely to become romantically involved, which in turn leads to improved peer status and facilitation of further friendships (Connolly et 48 al., 2000; Savickaitė et al., 2020; Zimmer-Gembeck et al., 2004). Moreover, increased social support from peers and romantic partners and greater social status may also improve anxiety symptoms (Scardera et al., 2020). These findings suggest that the benefits of early adolescent romantic experience for peer functioning apply only to heterosexual youth, possibly because heterosexual dating at this time conforms to heteronormative beliefs commonly held in early adolescence (Mayeux & Kleiser, 2020; Moreau et al., 2019). Interestingly, we did not observe increased peer stress in sexual minority youth who are romantically involved in early adolescence, and both heterosexual and sexual minority youth with greater romantic experience later in middle adolescence showed less peer stress. We were not able to assess the gender identity of the youth with whom sexual minority youth are romantically involved in either study; this may have obscured associations between romantic involvement and peer stress in sexual minority youth. For example, some sexual minority youth in our sample may have participated in exclusively other-sex romantic experiences, which could be perceived as consistent with heteronormative attitudes about romantic involvement and thus well-received by peers. Other sexual minority youth in our sample may have participated in exclusively same-sex or both same- and other-sex romantic experiences, violating heteronormative attitudes about romantic involvement and potentially leading to identity-based discrimination from peers and other interpersonal minority stressors. Future work should examine whether associations between sexual minority youth romantic involvement and peer stress vary based on the gender identity of sexual minority youth’s romantic partners to improve our understanding of whom is at greatest risk for poor social functioning. 49 Romantic experience and psychiatric symptoms in early adolescence: Moderation by sexual orientation, race/ethnicity, and socioeconomic status It is notable that associations between romantic experience and psychiatric symptoms did not differ based on race/ethnicity or socioeconomic status for sexual minority or heterosexual youth in the ABCD sample. Minority stress theory posits that health disparities for sexual minority youth can be explained in part by chronic stress caused by experiences and expectations of stigma and discrimination (Meyer et al., 2021). Considering the framework of intersectionality theory, sexual minority youth’s other social identities (e.g., race, ethnicity, socioeconomic status, etc.) intersect with their sexual orientation to additively or multiplicatively impact their experiences of disadvantage and privilege, and, in turn, mental health (Atewologun, 2018; Toomey et al., 2017). Recent research has evidenced that associations between adolescent romantic involvement and mental health vary based on race and ethnicity for sexual minority youth: adolescent romantic involvement was linked to lower levels of psychological distress for Black sexual minority youth and lower levels of anxiety for Latine sexual minority youth assigned female at birth (Whitton et al., 2018, 2021). These findings suggest that romantic involvement may provide Black and Latine sexual minority youth with social support, buffering the influence of intersecting minority stressors on mental health (Whitton et al., 2018, 2021). Due to small group cell sizes, we were unable to examine the moderating effects of multiple racial and ethnic identities on associations between romantic involvement and wellbeing. Collapsing multiple racial and ethnic identities into a single sample of youth identifying as Black, Indigenous, and/or People of Color (BIPOC) incorrectly 50 implies there is homogeneity across BIPOC populations (Buchanan et al., 2021), and may have obscured potential moderating effects. Similarly, parental education was used as a proxy for socioeconomic status and was also collapsed into two groups (i.e., youth with and without a parent who graduated with a 4-year college degree). It may be that both race/ethnicity and socioeconomic status require more nuanced modeling due to complex relations with romantic involvement, sexual orientation, and psychosocial functioning. Strengths and limitations The current research benefits from its use of two independent, well- characterized samples to assess associations between sexual orientation, romantic experience, and psychosocial functioning in early adolescence. Each study presents notable strengths: the SBTS features a prospective design, comprehensive and developmentally appropriate assessment of youth sexual orientation, and multi- method and multi-informant approach, whereas the ABCD study features assessments of both sexual and gender identity, and a large, more diverse sample, affording the opportunity to examine complex interaction models. This work should be interpreted with consideration of several limitations. First, the SBTS and ABCD studies differ in their measurement of sexual orientation, such that the SBTS utilizes a measure of same-sex attraction (Li & Hines, 2016; Storms, 1980), whereas the ABCD study utilizes the sexual orientation identity item from the K-SADS (Kaufman et al., 1997). Sexual attraction and sexual orientation are related, but distinct constructs; initial sexual attraction is an important first step in sexual orientation identity development that precedes identification of one’s sexual 51 orientation (Hall et al., 2021). Second, the SBTS utilized an assessment of sexual orientation at age 15 to retrospectively examine associations between sexual orientation and psychosocial functioning at age 12, despite evidence that youth can reliably report sexual orientation by this age (Clark et al., 2020). Third, neither the SBTS nor the ABCD study afforded the opportunity to assess whether associations between romantic experience and psychosocial functioning vary based on sexual minority identity (e.g., gay, lesbian, bisexual, pansexual, etc.). While the SBTS assessed both same- and other-sex attractions, too few youth reported having exclusively same-sex attractions (n=8) to conduct analyses comparing this group to youth who reported having both same- and other-sex attractions. The sexual orientation item utilized in the ABCD study did not differentiate between monosexual and bisexual orientati