Adolescence is a time characterized by a greater reliance on peers, the onset of romantic relationships, and a strengthening sense of sexuality (Nickerson and Nagle, 2005; Brown and Bakken, 2011; Tolman and McClelland, 2011; Connolly et al., 2014). Adolescence is also a period when sexual identity begins to develop; though, experiences and processes of sexual identity development often differ among adolescents (Perrin, 2002; Savin-Williams and Cohen, 2004). For example, some adolescents may have a consistent sexual identity over time, whereas others may report shift in their sexual identity, referred as sexual identity fluidity (Ott et al., 2011; Katz-Wise and Hyde, 2015; Diamond et al., 2017). Sexual identity fluidity is more common among sexual minority adolescents (e.g., gay, lesbian, bisexual, queer), with recent estimates noting sexual identity changes among sexual minority youth ranging from 28% to 67% (Diamond, 2008; Morgan et al., 2018; Silva, 2018; Cohen et al., 2020; Srivastava et al., 2022).

Sexual identification can be stressful for youth, stemming from: social norms to maintain a consistent sexual identity over time; pressure to choose identity labels that may approximate heterosexuality (e.g., mostly heterosexual, bisexual, pansexual) due to the privileging of heterosexuality; and expectations to choose a label that may not fit with one’s dynamic experiences of sexual/romantic attractions, behaviors, and relationships (Hall, 2019). Some research suggests that sexual identity fluidity is associated with worse mental health, including depression (Everett, 2015; Feinstein et al., 2019).

There are several theories that may help explain the association between sexual identity fluidity and worsening mental health, including identity control theory and minority stress theory. Identity control theory posits that one’s identity must align with their internal assessment on how the identity and associated behaviors within the context of social structures where the identities are embedded (Burke, 2006). With respect to sexual identities, it would mean that different aspects of sexuality (i.e., attraction, behavior and identity) must have coherence and concordance and one is able to manage these within their social environments (Igartua et al., 2009; Rosario et al., 2011). However, discordance among these aspects of sexuality may result in one or more shift in sexual identity over time (Rosario et al., 2011). Applying the minority stress framework, the internal process of understanding, internalizing, and confusion in regard to one’s sexual identity may represent an additional sexual minority stressor, referred to as identity management stress (Goldbach et al., 2017). For example, discordance among aspects of sexuality or developing identity in a homonegative climate could lead to identity dissatisfaction, and stress (Igartua et al., 2009; Ott et al., 2011; Page et al., 2013; Horley and Clarke, 2016).

Decades of theoretical and empirical evidence have shown that chronic stress can lead to negative mental health outcomes, such as depression (Pearlin, 1999; Marin et al., 2011). Meyer (2003) argues that sexual minority adolescents are exposed to heightened stress related to a variety of stigma and discrimination-related experiences based on their non-heterosexual status, often referred to as sexual minority stress (Meyer, 2003; Marshal et al., 2011; Goldbach et al., 2014). Research suggests that adolescents experiencing stress associated with identity management (e.g., addressing disagreements between sexual behavior or attraction and sexual identification) may report more symptoms of depression (Clark et al., 2015; Everett, 2015; Caplan, 2017). Among sexual minority youth, depression occurs at a greater rate during adolescence, compared to their cisgender heterosexual peers (Cole et al., 2002). Unfortunately, both identity-related stressors and depression are also known to have harmful effects on multiple developmental outcomes (e.g., later mental health, educational and economic outcomes) through their late adolescence and early adulthood (Dekker et al., 2007; Fergusson et al., 2007; Bulhões et al., 2020).

In addition, these processes of sexual identity development may differ by sex assigned at birth. Some evidence suggests that cisgender males typically describe their sexuality often as unchanging; however, cisgender females often describe it as more fluid, evolving, and contextual, leading to different pathways of identity development (Kinnish et al., 2005; Diamond, 2008; Radtke, 2013). This difference in identity development by sex assigned at birth could be extended to differential rates of fluidity in sexuality identity between cisgender males and females. For example, research suggests that cisgender females may be more likely to report sexual identity fluidity compared to males over time (e.g., Oi and Wilkinson, 2018 [19.3 vs. 13.3%]; Fricke and Sironi, 2020 [18.0 vs. 6.2%], Savin-Williams et al., 2012 [17.8 vs. 6.2%], and Stewart et al., 2019 [26.0 vs. 11.0%]).

Evidence suggests the importance of examining stressful experiences associated with identity management and depressive symptomology among sexual minority adolescents. However, these associations have not been studied longitudinally. In addition, it is unknown if these dynamic processes (identity management stress and depression) vary between those who report sexual identity fluidity versus those who did not. Moreover, given empirical evidence that sexual identity fluidity rates differ by sex assigned at birth, it is also important to understand if the association between identity management stress, depression, and sexual identity fluidity differs by sex assigned at birth.

The present study utilizes a large (N = 1,077) longitudinal sample of sexual minority adolescents. The goals of the present study are to assess time-sequential associations between two processes (identity management stress and depression) over time. In line with identity control theory, the present study sought to understand longitudinal associations between identity management stress and depression and how these processes may differ for those who report sexual identity fluidity versus those who do not.

Bivariate analyses were used to examine the differences in reported sexual identity fluidity (ever versus no change) by demographic variables at baseline. T-test for difference in sample means were conducted for the continuous variable (age), while chi-square tests were conducted for categorical variables (sex assigned at birth, race/ethnicity and sexual identity).

We examined if there were differences in time-sequential associations between depression and identity management stress, by sexual identity change status (no change versus change) using a multigroup autoregressive cross-lagged (ARCL) model. The cross-lagged design comprises two or more variables at two or more time points. It yields three types of effects: synchronous associations (correlations between different variables measured at the same time), stability effects (correlations between the same variable measured at different times), and cross-lagged effects. The cross-lagged effects refer to the prediction of one or more variables by other (temporarily preceding) variables, controlling for the baseline level of the predicted variable. A multi-step process was employed to time-sequential associations between depression and identity management stress, by identity change status (Figure 1).

In our first model, each variable was allowed to predict subsequent follow-up assessment of itself, measuring the stability of individual differences in the construct from one occasion to the next. Cross-lagged effects were estimated, controlling for the previous level of the construct being predicted. Thus, when depression at 6-month follow-up was predicted by identity management stress at baseline, depression at baseline was controlled to rule out the possibility that the cross-lagged effect is simply due to correlations between depression and stress at baseline (Anyan et al., 2020). In our second model, we used a multiple group analysis framework where sexual identity change and sex assigned at birth variables were used to permit direct comparisons of the association between change in depressive symptoms and stress between groups. Model fit was determined using the following indices: χ2 goodness-of-fit statistic, the comparative fit index (CFI ≥ 0.90), the root-mean-square error of approximation (RMSEA ≤ 0.08), and the standardized root-mean-square residual (SRMR ≤ 0.08) (Browne and Cudeck, 1993; Hu and Bentler, 1999). Analyses were carried out in the structural equation modeling (SEM) framework using Mplus 8.0; We used the full information maximum likelihood estimator for all analyses, which assumes data are missing at random and uses all data available for each participant (Muthén and Muthén, 2009).

The current study advances our understanding of sexual identity fluidity, and its association with depression and identity management stress over time. The study examined difference in time-sequential associations between depression and identity management stress, by fluidity status using a multigroup autoregressive cross-lagged model. In the sample, 40% of sexual minority adolescents reported at least one change in sexual identity over 4 time-points over 18-month period. This finding is consistent with the literature, where 28-67% of sexual minority adolescents and youth have reported fluidity in sexual identity orientation over time (Diamond, 2008; Morgan et al., 2018; Silva, 2018; Cohen et al., 2020).

With regard to our research question (i.e., if temporal relationship between identity management stress and depression over time differs by sexual identity fluidity status), results show that depression predicted greater identity management stress, irrespective of sexual identity fluidity status. However, in our second model (grouping by fluidity in sexual identity and birth sex), results indicated that the association between depression and identity management stress over time differed by sexual identity fluidity for cisgender females, but not for cisgender males. Among cisgender females who reported fluidity in sexual identity, we found a reciprocal effect of identity management stress and depression over time (depression predicted subsequent identity management stress, and identity management stress predicted subsequent depression). However, among those females who did not report fluidity in sexual identity, we did not find any cross-lagged effects. Among males, we found that depression predicted subsequent identity management stress, irrespective of sexual identity fluidity status. This result draws support from limited evidence where adolescents with discordant or fluid sexual identity labels are more likely to report higher rates of depression (Everett, 2015; Caplan, 2017). For example, Caplan (2017) reported that in their school sample, those with a concordant sexual orientation (agreement between sexual behavior or attraction, and sexual identification) report significantly lower depressive symptoms scores than do those with a discordant sexual orientation. Our results suggest that among female adolescents, sexual identity fluidity is associated with cognitive and emotional disruptions as they reconfigure their identities and navigate social support networks, leading to heightened stress and depressive symptomology. Additionally, prior research examining impact of sexual identity fluidity on mental health risks have looked at change from the point of directionality (a change towards same-sex identities or a change towards heterosexual identities (e.g., Everett, 2015). However, in our study, we examined fluidity in sexual identity labels, and not in sexual orientation on a Likert scale. Hence, we are not able to assess directionality of change, as the understanding of authentic sexual identity and a shift toward or away from it would differ from individual to individual.

Our results indicate that sexual identity fluidity process differs between cisgender females and males. Among males, the association between these two processes (identity management stress and depression) did not differ by sexual identity fluidity status. The suggested difference in that sexual identity change process between cisgender females and males, could be attributed to difference in sexual identity development and related experiences by birth sex (Kinnish et al., 2005; Diamond, 2008; Radtke, 2013). A recent systematic review of changes in sexual identities by Srivastava et al. (2022) found that sexual identity fluidity was more common among cisgender female compared to their male counterparts (often by more than 10 percentage points); however, it remains difficult to draw reliable conclusions about the extent and cause of these differences by sex assigned at birth. We believe one of the reasons for higher rates of fluidity among females in our sample compared to males could be attributed to sexual identification at baseline. For example, in our sample, those who identified with less common sexual identities (ex., pansexual, asexual, queer, and questioning) at baseline were more likely to report a change in sexual identity over 18-month period, compared to those who identified as gay, lesbian, and bisexual. Post hoc analyses suggests that 85% of cisgender males identified with more common sexual identities (gay or bisexual) at baseline, compared to only 66% of cisgender females. Further, those identifying as more common identities (gay, lesbian, bisexual) were least likely to report a change at the subsequent time point (14.1% vs. 49.3%; χ² = 132.1, p < 0.001) compared to participants identifying with less common identities (ex., pansexual, asexual, queer, and questioning). Given the variations in reported changes and reasons for the change, more research is needed to understand the causes of such differences by assigned sex at birth. Our finding also finds some support in the extant literature, for example, Glover et al. (2009) reported that in their study males used more traditional label (e.g., gay), whereas females presented greater variability in attraction and self-labels of sexual orientation, and were more likely to experience non-exclusive attractions, thus making sexual orientation and identification a much more fluid process compared to males. Some of these differences can be explained by dominant notions of masculinity for males and more acceptability of fluidity in sexuality for females (Kilianski, 2003; Breen and Karpinski, 2013; Diamond, 2016); however, the nuances associated with sexual identity development and processes of change as they differ between cisgender females and males need further investigation.

Current research and practice with sexual minority adolescents has suggested role of enabling environment in healthy identity development among sexual minority adolescents (Perrin, 2002; Savin-Williams and Cohen, 2004; Legate et al., 2019). Studies suggest that adolescents with unsupportive or homophobic familial environment, delay their coming out processes or even try to meet the heterosexual expectations of their predominately heterosexual families (Waldner and Magrader, 1999; Cox et al., 2010; Legate et al., 2019). For example, a study with 502 adolescents reported less acceptance and greater difficulty with disclosure were associated with higher scores on internalized homophobia (negative self-perceptions and self-devaluation for being non-heterosexual) (Cox et al., 2010). Additionally, confusion about one’s sexual identity can also delay endorsement and integration of sexual identity label (Igartua et al., 2009; Rosario et al., 2011). In short, lack of enabling environment may result in delaying of identity integration, resulting in development of less-authentic identities likely to change later. We believe some adolescents are more likely to report sexual identity fluidity than others, which may be because of underlying factors like access to community and support, relationship status, presence of a role model, etc. Moreover, these processes of fluidity could also be an additional stressor among sexual minority adolescents. The results add to the limited knowledge on the complex relationship between minority stress and depression as it relates to sexual identity fluidity.

This study had several limitations. This study focused on cisgender youth, it does not discuss the disparities experienced by transgender and gender non-binary adolescents (Price-Feeney et al., 2020; Srivastava et al., 2020). Sexual identity fluidity were measured as a difference in response to sexual identity question between two time-points, however, this may not be a conscious change on adolescent’s part. Internet survey research has distinct advantages, especially for reaching marginalized, geographically dispersed minority populations (Stern et al., 2020). We recruited a large sample of diverse sexual minority adolescents from both urban and rural areas of United States. However, internet-based recruitment and data collection also have limitations and challenges. In terms of generalizability, our findings are limited to adolescents who have access to the internet and online spaces. Although there is now significant evidence that the demographic and behavioral characteristics of those recruited online are similar to those recruited through more traditional, in-person venues. Additionally, internet survey research also has validity concerns (e.g., duplicate participations). However, our study protocols addressed these concerns through rigorous validity checks. All data were self-reported; however, anonymity was ensured by not collecting any identifying information, which minimized response bias.

Despite these limitations, to our knowledge, this paper is the first to examine the difference in relationship between depression and identity management stress by sexual identity fluidity status in a nationwide sample of sexual minority adolescents. Given pervasive homonegative social and political climates in many areas, sexual minority adolescents will continue to experience sexual identity development in less supportive environments, adding to the minority stress experiences. Among females, the evidence of relationship between identity management stress and depression as it associates with sexual identity fluidity is important to support female adolescents through their healthy identity development processes. However, the underlying and understudied variations by sex and gender need further examination.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by University of Southern California Institutional Review Board. Written informed consent for participation was not provided by the participants’ legal guardians/next of kin because: All participants provided assent to participate, and longitudinal participants provided informed consent at the first follow-up survey after reaching age 18. All study procedures were reviewed and approved by the University of Southern California Social–Behavioral Institutional Review Board, including a waiver of parental permission given the potential for harm due to unintentional “outing” to a parent during the consent process.

AS contributed to the conceptualization, formal analysis, and writing, reviewing, and editing the manuscript. WJH and EAK contributed to the interpretation of results and editing the manuscript. JTG is the PI on the grant, and contributed to the conceptualization, and interpretation of results. All authors contributed to the manuscript revision, read, and approved the submitted version.