Sexual Identity Development Milestones in Three Generations of Sexual Minority People: A National Probability Sample

Abstract

Using data from the first national probability sample of Black, White, and Latinx sexual minority people in the United States, we examined whether and how sexual identity development timing and pacing differs across demographic subgroups at the intersections of cohort, sex, sexual identity, and race/ethnicity. Among a sample of 1,491 participants aged 18–60 from three distinct birth cohorts, we measured participants’ ages of sexual identity development milestones, including first same-sex attraction, first self-realization of a sexual minority identity, first same-sex sexual behavior, first disclosure to a straight friend, and first disclosure to a family member. Participants from more recent cohorts reported earlier and accelerated pacing of milestones relative to those from older cohorts. Subgroups defined by sex and sexual identity varied in milestone timing and pacing, with gay males reporting earlier onset of some milestones than other subgroups. Those who used newer identity labels (e.g., pansexual, queer) reported younger ages of milestones relative to bisexual participants but similar ages to lesbian and gay participants. Black and Latinx participants reported some milestones at younger ages than White participants. Race-stratified models testing groups at the intersection of cohort, sex, and sexual identity revealed subgroup differences in ages of first disclosure to family, as well as differences in the time between self-realization, same-sex sexual behavior, and disclosure to a straight friend. Results suggest substantial variation in the developmental timing and pacing of milestones across social identities, and the need to further examine how milestone timing is related to identity, stress, and health.

In the past several decades, there have been rapid changes in the cultural, scientific, and public visibility of sexual and gender diversity. Through media, education, and socialization, people across the lifespan engage with new understandings of sexuality and gender beyond heteronormative and binary conceptions. A growing body of research suggests that many people are understanding and claiming a sexual minority identity (i.e., “coming out”) at relatively younger ages than in prior generations (Russell & Fish, 2019). Yet, the processes that characterize sexual identity development for contemporary sexual minority people are understudied, and even less is known about the developmental timing of these processes.

Coming out is often operationalized as disclosing one’s same-sex attraction or sexual minority identity to others; in reality, identity disclosure is only one of several sexual identity development milestones (henceforth referred to as “milestones”), including first same-sex attraction, self-realization of a sexual minority identity, and same-sex sexual behaviors, that mark the processes through which sexual minority people explore, acknowledge, and define their sexual identities. Thus, conceptualizing sexual identity development as a finite “light-switch” process of disclosure obscures the range of developmental processes that may influence sexual minority lives.

Beyond the operationalization of sexual identity development, how does sexual identity formation differ based on cohort, sex, sexual identity, and race/ethnicity? Although extant research has documented milestones across groups defined by these characteristics (e.g., Martos et al., 2015), these studies offer mixed findings and often rely on community samples that may not be representative of sexual minority people more broadly. Research with large probability samples is necessary to better understand the timing and pacing of milestones in the general population, and how these experiences differ for sexual minority people who came of age at distinct sociohistorical moments characterized by divergent cultural understandings of sexual diversity.

The current study utilizes the first national probability sample of Black, White, and Latinx sexual minority people in the U.S. to examine: 1) timing and pacing of milestones; 2) whether and how milestones have shifted over time and across cohorts; and 3) whether and how milestones differ based on sex, sexual identity, and race/ethnicity. We begin with a broad overview of sexual identity development research foundations and the emergence of sexual identity development milestones. Anchored in life course and intersectionality perspectives, we review prior research related to subgroup differences in milestones across cohort, sex, sexual identity, and race/ethnicity. Finally, to extend the current literature, we report findings on the timing and pacing of milestones across demographic subgroups using national probability data designed to address cohort differences among sexual minority people in the United States.

Sexual Identity Development Foundations and the Emergence of Milestones Perspectives

Early clinical and developmental psychologists constructed models defined by linear stages of sexual identity development (e.g., Cass, 1979; Troiden, 1989). These models projected uniform progression from conflict and confusion towards internal awareness and resolution, culminating in public disclosure. Despite their widespread acceptance beginning in the 1980s, important critiques of linear stage models emerged (e.g., Horowitz & Newcomb, 2001). First, they suggested a uniform coming out process, with little acknowledgment of the identity-based, interpersonal, and sociohistorical factors that shape sexual identity development. Second, they proposed unidirectional endpoints, rather than conceptualizing sexual identity development as continuous, iterative, and shaped by sociohistorical, developmental, and interpersonal contexts (D’Augelli, 1994; Diamond, 2006). Third, stage models were primarily cultivated from small, community samples consisting primarily of White, gay, male adults (e.g., Cass, 1979).

Milestones emerged as a new theoretical perspective of sexual identity development that centered heterogeneity and contextual influences that shape developmental processes. Milestones reflect the ages at which sexual minority people report first experiencing pivotal events in the exploration, formation, and integration of their sexual identities. Commonly assessed milestones include the age of first attraction to a same-sex person, self-identification as a sexual minority, same-sex sexual behaviors, and disclosure of a sexual minority identity (e.g., Calzo et al., 2011).

Research on milestones has focused primarily on mean ages of milestone events. However, scholars have also examined the pacing and patterning of milestones as critical measures of sexual identity development. Some studies have demonstrated that the time between, rather than the age, of milestones is associated with outcomes such as problem drinking behavior (Parks & Hughes, 2007). Additionally, research has identified variability in the order of milestones, specifically related to the order of self-identification as a sexual minority, initial same-sex sexual experiences, and identity disclosure. Early conceptual models of sexual identity development (e.g., Troiden, 1989) proposed a “sex-centered” pattern wherein same-sex sexual behavior occurred in tandem with sexual minority self-realization to affirm a sexual minority identity. However, in several recent studies, participants have also reported “identity-centered” ordering (i.e., self-identifying and/or disclosing a sexual minority identity prior to same-sex sexual experience; Calzo et al., 2011; Dubé, 2000; Floyd & Bakeman, 2006). More research is needed to understand the precursors of sex-centered and identity-centered patterning, given that this ordering may be associated with wellbeing (Dubé, 2000).

Life Course and Intersectionality Perspectives

Life course and intersectionality perspectives offer frameworks for centering the diversity that is inherent to sexual identity development processes. A life course approach highlights the historical timing of development as well as the idea that the impact of developmental events is contingent on when they occur in a person’s life (Elder, 1998). From a life course perspective, sexual identity development models should account for the influence of developmental and sociohistorical contexts on the lived experiences of sexual minority people from distinct cohorts (Hammack, et al., 2018). For example, cohort-defining events for people in the United States (e.g., the national legalization of marriage for same-sex couples in 2015) create distinct social contexts for sexual identity development; as such, the timing of milestones likely varies across these sociohistorical contexts and cohorts. To date, there have been only a handful of studies that have considered the impact of developmental stage and historical cohort on sexual identity development (e.g., Calzo et al., 2011; Grov, et al., 2018; Martos et al., 2015), and no studies to our knowledge have explored these differences in a national probability sample of sexual minorities designed to examine cohort differences.

Intersectionality perspectives describe how systems of oppression are driven by identity-based prejudice related to interlocking social identities such as race, class, and gender (Crenshaw, 1991). The contribution of intersectionality perspectives to understanding sexual identity development is threefold: It supports the premise that socialization into specific genders, sexualities, race/ethnicities, and cohorts necessarily shapes the formation of identity and of identity-based discrimination; it calls into question the generalizability of sexual identity development models based only on homogeneous community samples; and it urges us to consider the relations between interlocking social identities and related oppression.

Taken together, life course and intersectionality perspectives allow us to ask new questions about the dynamic interplay between ontogenetic and contextual influences on sexual identity development. These frameworks position individuals and their development in historical, developmental, and social locations, and provide a framework to understand the diversity that characterizes sexual identity formation processes.

Variability in Sexual Identity Development Milestones

Mounting research suggests that sexual minority people experience varied developmental pathways when exploring and forming their identities. In fact, variability in the timing and pacing of milestones may be the rule rather than the exception among diverse sexual minority people living in distinctive sociohistorical contexts (D’Augelli, 1994).

Cohort Differences.

The timing and pacing of milestones may be changing across cohorts (e.g., Floyd & Bakeman, 2006; Grov et al., 2018; Martos et al., 2015). For example, among a non-probability sample of gay men, awareness of attraction and sexual behavior shifted one year earlier in age for every 8–25 calendar years, and disclosure shifted one year earlier for every 2–5 calendar years (Drasin et al., 2008). These differences are thought to coincide with prominent shifts in visibility and attitudes towards sexual minority people over time (Russell & Fish, 2016), as well as the cohort-defining events that exemplify such shifts. Historical events such as the introduction of homosexuality as a mental disorder in the DSM in 1952 and its subsequent removal in 1973, the HIV/AIDS crisis of the 1980s and 1990s, and the national legalization of marriage for same-sex couples in 2015 have had profound effects on cultural notions of same-sex sexuality, with implications for sexual identity development (Hammack et al., 2018).

Sociohistorical shifts towards greater acceptance of sexual minority identity may also condense the time between milestones. Martos and colleagues (2015) found that more recent cohorts of sexual minority people reported shorter times between first feeling attracted to a same-sex person and their first same-sex relationship. Others have found that more recent cohorts of lesbians spend less time between wondering about a same-sex identity and self-realizing their identity, and between self-realizing and disclosing a sexual minority identity than previous cohorts (Parks & Hughes, 2007). Milestone pacing may be directly related to experiences of minority stress (Meyer, 2003) as high stigma environments may increase internalized homophobia and/or delay identity disclosure (Russell & Fish, 2016). Exploring cohort differences in the time between milestones may provide unique insight into sexual identity developmental processes.

Sex and Sexual Identity Differences.

The timing and pacing of milestones may vary by both sex and sexual identity, as well as their intersections. With respect to sex differences, prior studies suggest that males often report same-sex attraction and self-realization milestones earlier than females (e.g., Katz-Wise et al., 2017; Martos et al., 2015). Further, although males are more likely to be younger when they report milestones, they also report taking longer to disclose a sexual minority identity after first self-identifying than females (Martos et al., 2015).

Sexual identity (i.e., whether one identifies as “lesbian,” “gay,” “bisexual,” “queer,” or another identity) may also contribute to milestone timing and pacing. Among some studies examining variability in milestones across sexual identities, those with bisexual identities reported later ages of first same-sex awareness, attraction, and self-identification relative to lesbian and gay individuals (Maguen et al., 2002; Martos et al., 2015). In another study, the time between first same-sex attraction and self-identification was greater for bisexual participants than for lesbian and gay participants, but bisexual participants spent significantly less time between self-identification and disclosure than lesbian/gay participants (Martos et al., 2015). Different-sex attraction may delay bisexual people’s recognition and development of a sexual minority identity that could lead to greater vulnerability to discrimination (Maguen et al., 2002). Among the studies that have examined the intersections of sex and sexual identity, researchers have found that both bisexual and lesbian females tend to report later mean ages of milestones than bisexual and gay males (Katz-Wise, 2017). Representative research exploring this variability could lend insight into the links between gender socialization, biphobia, homophobia and sexual minority identity development.

Taken together, both sex and sexual identity matter for sexual identity development timing. Given that both bisexual and monosexual (i.e., lesbian and gay) people, as well as males and females, consistently vary on outcomes related to stress and health (Russell & Fish, 2016), population-based research examining subgroups of sexual minority people at the intersection of sex and sexual identity could illuminate how these social identities may explain developmental variability among sexual minority populations.

Although most sexual minority people use terms such as “gay,” “lesbian,” and “bisexual” to describe their sexual identities (Russell, et al., 2009), contemporary cohorts increasingly choose alternative sexual identity labels such as “queer” or “pansexual” (Goldberg et al., 2020; Morandini et al., 2017). Prior studies have not systematically examined the development of these newer, increasingly common sexual identities, and it has not been possible to assess whether those with these identities experience distinct milestone timing or pacing. Given the variability of sexual minority identity labels among more recent cohorts, there is a need to integrate newer sexual identities into the study of sexual identity development timing and pacing.

Racial/ethnic Differences.

Prior research is mixed regarding the timing and pacing of milestones across U.S. racial/ethnic groups. Some have theorized that cultural factors such as the primacy of family, conservative religious values, and racism could delay sexual identity development for racial/ethnic minorities (e.g., Greene, 1997). In fact, many studies measuring racial/ethnic differences across milestones observe the opposite: in one study of lesbian women, Black and Latina women reported earlier milestones than White women (Parks et al., 2004). In a study of sexual minority male youth, Latinx youth reported earlier awareness of same-sex attractions than African American and White youth (Dubé & Savin-Williams, 1999). At the same time, a number of studies have found no racial/ethnic differences in milestone timing (Floyd & Stein, 2002; Grov et al., 2006; Martos et al., 2015; Rosario et al., 2004). Given the inconsistency of previous research, and in order to increase understandings of how social locations influence identity development processes that predict risk or resilience, research with probability samples of sexual minority people is necessary to better understand the links between racial/ethnic identity and milestones.

Measurement Gaps

Research exploring milestones across different sociodemographic subgroups suggests context-specific variability in sexual identity development. However, measurement gaps in prior research limit its generalizability to contemporary cohorts of sexual minority people. First, apart from a few recent studies (e.g., Katz-Wise et al., 2017; Martos et al., 2015), and despite expert recommendations (Institute of Medicine, 2011), there has been a lull in research investigating milestones since the early 2000s. Yet there have been critical sociohistorical events (e.g., the legalization of same-sex marriage) that have likely influenced sexual identity development. Second, prior research is dominated by urban community samples that may not generalize to the broader sexual minority population. Last, most studies that examine variability across social identities treat each identity as distinct; we propose that intersectional experiences related to cohort, sex, sexual identity, and race/ethnicity may matter greatly for sexual identity development timing and pacing. Therefore, intersectional approaches to studying milestones are crucial.

The Current Study

The goal of the current study is to examine a diverse national probability sample of sexual minority people aged 18–60 who came of age in three distinct sociohistorical contexts to describe the timing and pacing of five key milestones: the age of first same-sex attraction, self-realization of a sexual minority identity, same-sex sexual experience, disclosure to a straight friend, and disclosure to a family member. We then examine subgroup differences in the timing and pacing of milestones by cohort, sex, sexual identity, and race/ethnicity, as well as by the intersections of these social identities. We hypothesize that:

1. More recent cohorts will report younger ages of milestones than less recent cohorts, and will report less time between milestones.

2. Gay and bisexual males will report younger ages of milestones than lesbian and bisexual females, but will progress through milestones at a slower pace than lesbian and bisexual females.

3. Given the dearth of research exploring the milestones of sexual minority people with newer (e.g., queer, pansexual) identities, exploratory analyses will examine milestone timing and pacing for sexual minority people with these identities.

4. Black and Latinx sexual minority people will report younger ages of milestones than White sexual minority people, but will not differ from White sexual minority people in milestone pacing.

Method

Data Source and Sample

Data come from the first wave of the Generations Study, a national probability study designed to examine identity, stress, and health across three cohorts of lesbian, gay, and bisexual people in the United States (Meyer et al., 2020). The study was explicitly designed to capture cohorts of sexual minority people who came of age in different historical contexts marked by distinct visibility and acceptance of sexual diversity. Age cohorts were defined based on an analysis of historical events and consideration of the age of respondents at critical periods of development (for further details, see Frost et al., 2019). The inclusion (youngest) cohort were 18–25 at recruitment and came of age at a time when sodomy laws were ruled unconstitutional and public discourse was shifting to equity and inclusion. The visibility (middle) cohort were 34–41 years old at recruitment and were adolescents when the HIV/AIDS epidemic was emerging as a public health crisis, leading to greater visibility of sexual minority identity. The pride (oldest) cohort were 52–59 years old at recruitment, and came of age as the first post-Stonewall generation, when homosexuality was considered a mental illness, when sodomy was illegal, and when pride festivals were first occurring. The cohort design allows for assessment of whether the timing and pacing of sexual identity milestones have changed alongside these historic shifts.

Participants were recruited through the Gallup Daily Tracking Survey (GDTS), a telephone interview conducted with a national probability sample of over 366,000 participants in the United States (Meyer et al., 2020). LGBT participants in the GDTS were identified through random digit dialing using the question, “Do you, personally, identify as lesbian, gay, bisexual, or transgender?” Those who responded affirmatively were further screened for eligibility. Those who identified as transgender were recruited for a complementary study (Krueger et al., 2020a) that focused on issues specific to the transgender community, and thus were excluded from the present sample. Respondents were eligible for participation in the Generations Study if they identified as LGB and met each of the following criteria: (a) identified as any identity except heterosexual; (b) were ages 18–25, 34–41, or 52–59 at the time of recruitment; (c) had, at minimum, a 5th grade education; (d) had conducted the GDTS phone interview in English, and (e) identified their race/ethnicity as Black, Latinx, or White (or multiple racial/ethnic categories including at least one of these). The eligibility limitation to three race/ethnic groups was based on projections of sufficient numbers in the GDTS of sexual minority participants in each race/ethnic group to permit robust statistical estimation of subgroup differences (Meyer et al., 2020).

Respondents who met the eligibility criteria were invited to participate in the Generations Study and were either emailed or mailed a self-administered survey questionnaire. Respondents received a $25 gift certificate for participation. The study protocol was reviewed and approved by the Institutional Review Board of the University of California, Los Angeles as IRB of record and relying IRBs in collaborating institutions under the project name Identity Stress and Health in Three Cohorts of LGB individuals (IRB# 14–000500). Of those screened through the GDTS, 3.5% identified as LGBT; 27.5% of the LGB participants were eligible for the study based on the aforementioned criteria. Of those eligible, 81% agreed to participate in the survey and among those, 48% completed the survey between March, 2016 and March, 2017 for a total cooperation rate (proportion of completed out of all who were eligible) of 39%. The study was designed to oversample Black and Latinx respondents, and the recruitment period for these groups was extended through March, 2018. Sampling weights account for the oversampling of Black and Latinx respondents and provide population-representative estimates. The final baseline sample included 1,518 respondents. The analytic sample was restricted to participants who provided their sexual identity and a valid response to at least one milestone age (n=1,491). The analytic sample characteristics are reported in Table 1.

Table 1.

Demographics of study sample

	Attraction	Self-Realization	Same-sex Sex	Disclosure (Friend)	Disclosure (Family)
	n (%) X2(df), p	n (%) X2(df), p	n (%) X2(df), p	n (%) X2(df), p	n (%) X2(df), p
Cohort
Inclusion (ages 18–26)	633 (96.22)	643 (96.35)	502 (74.26)	620 (94.39)	509 (77.85)
Visibility (ages 32–43)	354 (96.78)	352 (95.87)	350 (96.83)	343 (92.96)	320 (88.05)
Pride (ages 50–60)	453 (96.53)	445 (95.31)	457 (98.12)	428 (90.68)	407 (86.63)
X2(df), p	.23(2), .91	.63(2), .80	133.60(2), <.001	4.70(2), .17	21.39(2), <.001
Sex
Female	766 (95.84)	777 (96.38)	669 (78.04)	746 (93.83)	648 (79.18)
Male	674 (97.20)	663 (95.62)	640 (90.38)	645 (92.91)	588 (84.85)
X2 (df), p	1.90(1), .26	.56(1), .58	38.75(1), <.001	.50(1), .56	7.62(1), .02
Sexual Identity
Lesbian/gay	804 (97.60)	792 (96.18)	791 (94.58)	783 (95.57)	760 (93.44)
Bisexual	477 (97.77)	475 (95.43)	377 (72.67)	443 (90.55)	329 (67.78)
Newer	159 (87.39)	173 (97.76)	141 (73.59)	165 (95.08)	147 (81.40)
X2 (df), p	49.64(2), <.001	2.09(2), .53	124.05(2), <.001	14.33(2), <.01	141.48(2), <.001
Race/Ethnicity
White	933 (96.04)	933 (95.27)	859 (83.15)	901 (92.68)	798 (80.69)
Black	220 (96.89)	221 (97.57)	204 (88.28)	207 (91.13)	188 (80.50)
Latinx	287 (97.03)	286 (97.28)	246 (78.65)	283 (97.50)	250 (84.47)
X2 (df), p	.88(2), .73	4.23(2), .25	9.05(2), .07	11.55(2), .01	2.42(2), .46
Analytic Sample	1440	1440	1309	1391	1236

Note. Cell sizes are unweighted and percentages are weighted. % is the percentage of respondents in each demographic subgroup who reported a milestone age. Rao Scott X² is weighted and indicates chi square differences in % who reported a milestone between subgroups. Race= race/ethnicity. Bold values indicate p< .05.

Measures

Sexual identity development milestone age.

Five items were used to assess the age at which participants reported milestones (Martin & Dean, 1987): 1) “At what age were you first sexually attracted to someone of the same sex as you?”; 2) “At what age did you first realize that you were LGB?”; 3) “At what age were you the first time you had sex with someone of the same sex?”; 4) “At what age did you first tell a straight friend that you were LGB?”; and 5) “At what age did you first tell a family member that you were LGB?” Importantly, prior to the milestone questions, the survey text read, “Remember, by LGB we mean a sexual minority identity that you identify with.” Response options included providing an age in years at which the milestone occurred, selecting “never,” or selecting “don’t know.”

Sexual identity development milestone pacing.

Four variables were generated to assess the time between milestones that have been theorized to follow a temporal order (Martos et. al, 2015; Troiden, 1989). These include the time between: 1) age of first same-sex attraction and first self-realization of a sexual minority identity; 2) age of first self-realization and first same-sex sexual behavior; 3) age of first same-sex sexual behavior and first disclosure to a straight friend, and; 4) age of first disclosure to a straight friend and first disclosure to a family member. Given that prior research has found considerable variability in the ordering of milestones (e.g., Calzo et al., 2011; Floyd & Stein, 2006), negative values were retained in order to account for those who did not follow the proposed chronological order.

Cohort.

A cohort variable was constructed from participants’ answer to the question, “What year were you born?” Participants’ age was calculated by subtracting this answer from the year that the respondent completed the survey. Respondents were then assigned to one of three cohorts: the inclusion (youngest) cohort, aged 18–25; the visibility (middle) cohort, aged 34–41; and the pride (oldest) cohort, aged 52–59. A small number of participants aged out of the cohort between the time their age was originally collected from the GDTS and the dissemination of the survey, or there were slight discrepancies between dates reported on the screener and on the survey. We retained participants within two years of the age criteria, resulting in cohorts that included age ranges of 18–26, 32–43, 50–60 years.

Sex.

Respondents’ sex was coded as male or female according to their reported sex assigned at birth. Eighteen respondents with missing data were assigned a value from the sex they reported on the GDTS. Given that participants who identified as transgender at the screening process took part in a different study, hereafter, we use the term “male” to refer to those assigned male at birth and “female” to refer to those assigned female at birth.

Sexual identity.

Participants were asked, “Which of the following best describes your current sexual orientation?” Possible response options were “Lesbian,” “Gay,” “Bisexual,” “Queer,” “Same-gender loving,” and “Other (write in).” Three sexual minority categories were retained for analysis: lesbian/gay, bisexual, and newer sexual minority identity (e.g., queer, pansexual, asexual spectrum, anti-label, other).

Race/ethnicity.

During the screener process, race/ethnicity was measured by asking “Are you of Hispanic, Latino, or Spanish origin—such as Mexican, Puerto Rican, Cuban or other Spanish origin?” and “Which of the following describes your race: White, Black or African American, Asian, American Indian or Alaska Native, or Native Hawaiian or Pacific Islander?” Participants were eligible for inclusion if they reported being Black or African American, White, or Latinx during the GDTS, including those with multiple racial/ethnic identities who listed Black or African American, White, or Latinx as one of their identities. The following method (Krueger, et al., 2020b) was used to classify groups: All participants who reported that they were Hispanic/Latino were categorized as Latinx regardless of any other answer; then, participants who indicated that they were Black or African American were categorized as Black regardless of other races selected (with the exception of Latinx); finally, anyone who indicated that they were White including any other race, except Latinx and Black, was categorized as White. A three-category race variable was retained for analysis, including three mutually-exclusive categories: Black, White, and Latinx.

Analytic Approach

Data were analyzed using Stata 15 (StataCorp, 2017). The survey “svy” estimation command was used to apply weights for population-based estimates. There are no stratification or cluster weights in the sample. First, Rao Scott chi-square analyses were performed to test differences between participants who did and did not report an age of first same-sex attraction, self-realization as a sexual minority, same-sex sexual behavior, disclosure to a straight friend, and disclosure to family across groups defined by cohort, sex, sexual identity, and race/ethnicity. Next, among those who reported a milestone age, a series of analyses of covariance (ANCOVA) were performed to test differences between milestone age and pacing by cohort, sex, sexual identity, and race/ethnicity. Additional race-stratified ANCOVAs were used to test milestone differences across groups defined by the interaction between cohort, sex, and sexual identity. Bonferroni post-estimation and contrast procedures were used to test mean differences between groups.

Prior studies have documented wide ranges of responses to milestone ages, especially those related to same-sex attraction and self-awareness. For example, in D’Augelli and Hershberger’s (1993) study of milestones among LGB youth in urban community centers, participants reported ages of self-awareness of an LGB identity between 1 and 19. Herdt and McClintock (2000) assert that attraction before adrenarche is fundamentally distinct from sexual attraction. Therefore, the meaning of reports of same-sex attraction and awareness before the developmental emergence of sexual attraction is ambiguous. In light of this, we conducted a series of sensitivity analyses with uncensored milestone ages (see supplementary Tables S1 and S2) and milestone ages censored at 8. Given minimal differences across models and Herdt and McClintock’s theoretical argument that pre-adrenarche milestones are fundamentally distinct from sexual identity development milestones, milestone age was censored at age 8.

Results

Differences in Reported Milestones

Rao Scott chi-square analyses testing differences between participants who did and did not report milestone ages by sociodemographic characteristics are presented in Table 1. There were cohort differences in the percentage of respondents who reported an age of first same-sex sexual behavior (χ2(2,N =1,491) = 136.60, p<.001) and an age of disclosing a sexual identity to a family member (χ2(2,N =1,491) = 21.39, p<.001). Roughly 97% of participants from the visibility (middle) cohort and 98% of those from the pride (older) cohort reported an age of first same-sex sexual behavior, relative to 74% of those from the inclusion (youngest) cohort. About 88% of those from the visibility cohort and 87% of those from the pride cohort reported disclosing their sexual minority identity to a family member, compared to 78% of those from the inclusion cohort. There were no generational differences in the proportion of reported ages of first same-sex attraction, self-realization of a sexual minority identity, or disclosure to a straight friend.

Sex differences were also present. More males reported an age of first same-sex sexual experience (χ2(1,N =1,491) = 38.75, p<.001) and disclosure to a family member (χ2(1,N =1,491) = 7.62, p=.02). No significant differences emerged with respect to same-sex attraction, self-realization of a sexual minority identity, or disclosure to a straight friend.

Reports of milestone age also varied by sexual identity, including same-sex attraction (χ2(1,N =1,491) = 49.64, p<.001), same-sex sexual behavior (χ2(2,N =1,491) = 124.05, p<.001), disclosure to a straight friend (χ2(1,N =1,491) = 14.33, p<.01), and disclosure to a family member (χ2(1,N =1,491) = 141.48, p<.001). Almost all lesbian/gay (98%) and bisexual (98%) respondents reported same-sex attraction while 87% of those with newer identities reported this milestone. Whereas 95% of lesbian/gay respondents reported an age of first same-sex sexual behavior, 73% of bisexual participants and 74% of those with newer identities reported this milestone. More than 95% of those with gay, lesbian, and newer identities reported an age of disclosure to friends, while 91% of bisexuals disclosed to a straight friend. Overwhelmingly, lesbian/gay respondents reported disclosure to family (93%), whereas 68% of bisexual people and 81% of those with newer identities did. There were no differences in the percentage of those reporting an age of sexual minority self-realization.

There were few differences in reporting milestone ages by race/ethnicity. There were no racial/ethnic differences in the proportion of those reporting same-sex attraction, self-realization of a sexual minority status, same-sex sexual behavior, or disclosure to a family member. Whereas Latinx people most commonly reported an age of disclosure to a straight friend (98%), Black participants (91%) and White participants (93%) reported less frequent disclosure to a straight friend (χ2(2,N =1,491) = 11.54, p=.01).

Timing of Sexual Identity Development Milestones

Table 2 presents estimated marginal means from ANCOVA analyses testing sociodemographic differences in the age of each milestone, adjusted for cohort, sexual identity, sex, and race/ethnicity. Cohort differences were prominent across all milestones. Those from the inclusion (youngest) cohort reported younger ages of first same-sex attraction (M=11.75, SE=.14) than those from the pride (oldest) cohort (M= 13.08, SE=.38, p=.005) but did not differ from the visibility (middle) cohort. Participants from each cohort statistically differed in their mean age of first realizing their sexual minority status. There was an approximately 2.5 year lag between reported ages of those from the inclusion cohort (M=13.86, SE=.15), the visibility cohort (M=16.32, SE=.38), and the pride cohort (M=18.81, SE=.48). The inclusion cohort reported younger ages of first same-sex sexual behavior (M=16.40, SE=.16) relative to the visibility (M=18.46, SE=.38, p<.001) and pride cohorts (M=19.30, SE=.51, p<.001), who did not statistically differ. Same-sex sexual behavior and the age of first disclosure to a straight friend occurred on average nearly simultaneously in the inclusion cohort, whereas those in the visibility and pride cohorts reported younger mean ages of same-sex sexual behavior than disclosure. Across cohort, mean ages of first disclosure to a straight friend occurred about 4 years apart and the inclusion (M=15.91, SE=.13), visibility (M=20.58, SE=.38), and pride (M=24.40, SE=.51) cohorts all significantly differed from one another (p<.001). The ages at which participants first disclosed to family also differed across cohorts (M=16.88, SE=.14 vs. M=22.21, SE=.40 vs. M=26.48, SE=.56, respectively, p<.001). Across cohorts and milestones, only one subgroup did not differ from any other group: those from the visibility cohort reported similar mean ages of first same-sex attraction relative to those from the inclusion and pride cohorts, respectively.

Table 2.

ANCOVA post-hoc pairwise comparison of cohort, sex, sexual identity, and race/ethnicity differences in timing of sexual identity development milestones.

	First Same-Sex Attraction			First self-realization			First Same-Sex Sex			First Disclosure (Friends)			First Disclosure (Family)
	M	SE	F	M	SE	F	M	SE	F	M	SE	F	M	SE	F
Total	12.08	.12		15.22	.15		17.48	.17		18.29	.17		19.83	.20
Age Range	8–45			8–56			8–50			8–55			8–57
Cohort			5.25**			57.14***			22.72***			166.24***			193.91***
Inclusion (18–26)	11.75a	0.14		13.86ab	0.15		16.40ab	0.16		15.91ab	0.13		16.88ab	0.14
Visibility (32–43)	12.25	0.30		16.32ac	0.38		18.46a	0.38		20.58ac	0.38		22.21ac	0.40
Pride (50–60)	13.08a	0.38		18.81bc	0.48		19.30b	0.51		24.41bc	0.51		26.48bc	0.56
Sex			4.24*			6.11*			13.13***			.49			.23
Female	12.31a	0.19		15.53a	0.21		18.06a	0.24		18.37	0.20		19.76	0.21
Male	11.74a	0.18		14.75a	0.21		16.74a	0.25		18.16	0.20		19.92	0.25
Sexual Identity			12.36***			15.55***			.33			8.90***			5.35**
Lesbian/Gay	11.46a	0.19		14.34a	0.22		17.34	0.22		17.84a	0.20		19.35a	0.21
Bisexual	12.93ab	0.23		16.31ab	0.26		17.63	0.33		19.03ab	0.26		20.58a	0.29
Newer	11.58b	0.27		14.92b	0.31		17.68	0.41		17.66b	0.31		19.84	0.38
Sex x Sexual Identity			8.41***			3.62*			6.19**			1.02			6.09**
Lesbian	12.28a	0.33		15.04a	0.37		18.46ab	0.37		17.86	0.34		19.08a	0.34
Bisexual Female	12.88bc	0.27		16.37bc	0.30		17.73c	0.36		19.10ab	0.30		20.47ab	0.31
Newer Female	11.58bd	0.28		15.32d	0.34		18.48d	0.47		17.91	0.36		20.28c	0.42
Gay	10.74ace	0.17		13.63abde	0.23		16.27acde	0.24		17.75a	0.23		19.57	0.27
Bisexual Male	13.52de	0.40		16.70ef	0.44		18.35e	0.63		18.98c	0.43		20.82d	0.57
Newer Male	12.12	0.57		14.03cf	0.66		16.09b	0.70		16.76bc	0.54		17.74bcd	0.65
Race			6.78**			4.78**			13.12***			2.57			2.53
White	12.38a	0.16		15.56a	0.18		17.94a	0.21		18.43	0.18		20.01	0.20
Black	11.32a	0.25		14.61	0.35		16.01ab	0.32		17.61	0.33		19.03	0.39
Latinx	11.79	0.25		14.70a	0.28		17.31b	0.32		18.37	0.26		19.88	0.30

Note. The same superscripts indicate significant differences at the p<.05 level. Bonferroni post-hoc comparisons were used to minimize Type 1 error. Estimated marginal means adjusted for the effects of cohort, sexual identity, sex, and race/ethnicity are reported. Data are censored at age 8. See supplemental Table S1 for uncensored results. Race=Race/ethnicity.

^*p <.05

^**p<.01

^***p<.001.

Milestone differences at the intersection of sex and sexual identity were prominent. Gay males reported the earliest age of first same-sex attraction (M=10.74, SE= .17) and differed from lesbians (M=12.28, SE= .33, p=.001), bisexual females (M=12.88, SE= .27, p<.001), and bisexual males (M= 13.52, SE= .40, p<.001), but not from those with newer identities. Bisexual males reported the oldest mean age of first same-sex attraction (M= 13.52, SE= .40) and were significantly older than females with newer identity labels (M=11.58, SE= .28, p=.001) and gay males (M=10.74, SE= .17, p<.001) when they first experienced same-sex attraction, but not from lesbian or bisexual females nor from males with newer identities. Bisexual females reported significantly older ages of first same-sex attraction (M=12.88, SE= .27) than females with newer identities (M=11.58, SE= .28, p=.008).

Gay males also realized their sexual minority identities at the youngest age (M=13.63, SE= .23); statistically younger than lesbians (M=15.04, SE= .37, p=.02), bisexual females (M=16.37, SE= .30, p<.001), females with newer identities (M=15.32, SE= .34, p=.001), and bisexual males (M=16.70, SE= .44, p<.001), but not males with newer identities. Bisexual males (M=16.70, SE= .44, p<.001) reported the oldest mean age of first realization and in addition to gay males, differed from males with newer labels (M= 14.03, SE= .66, p=.01); bisexual males did not differ from lesbians, bisexual females, or females with newer identity labels.

Gay males reported first same-sex sexual behavior significantly younger than all other subgroups (M=16.27, SE= .24), with the exception of males with newer identities. Lesbians reported significantly older mean ages of same-sex behavior (M=18.46, SE= .37) relative to males with newer identity labels (M=16.09, SE= .70, p=.04). No other same-sex behavior differences were observed.

Males with newer identities reported the youngest mean age of disclosure to a straight friend (M=16.76, SE= .54) and differed from bisexual females (M=19.10, SE= .30, p=.002) and bisexual males (M=18.98, SE= .43, p=.02), but not from gay males, lesbian females, or females with newer identities. Bisexual females reported the oldest mean age of disclosure to a straight friend (M=19.10, SE= .30) and significantly differed from gay males (M=17.75, SE= .23, p=.007) and males with newer labels (M=16.76, SE= .54, p=.002), but not from bisexual males, lesbians, or females with newer labels. Bisexual males reported the oldest mean ages of disclosure to family (M=20.82, SE= .57) and were older than males with newer labels (M=17.74, SE= .65, p=.006), who reported the youngest mean age of disclosure to family. Males with newer labels reported significantly younger mean ages of disclosure to family than bisexual females (M=20.47, SE= .31, p=.002), females with newer labels (M=20.28, SE= .42, p=.015), and bisexual males (M=20.82, SE= .57, p=.006).

With respect to milestone age, White participants reported later ages of same-sex attraction (M= 12.38, SE= .16) than Black participants (M= 11.32, SE= .25, p=.001), and later mean ages of self-realization (M= 15.56, SE= .18) than Latinx participants (M= 14.70, SE= .28, p=.03). Additionally, Black participants reported significantly younger ages of first same-sex sexual behavior (M= 16.01, SE= .32) than White (M= 17.94, SE= .21, p<.001) or Latinx (M= 17.31, SE= .32, p=.009) participants, who did not differ from one another. No racial/ethnic differences were observed in age of first disclosure to a straight friend or family member.

Pacing of Sexual Identity Development Milestones

Table 3 presents ANCOVA results testing mean differences in the pacing between milestones across social identities; estimated marginal means adjusted for the effects of cohort, sexual identity, sex, and race/ethnicity are reported. Similar to the timing of milestones, the time between milestones differed by cohort (see Figure 1). The inclusion cohort spent less time between the age of first same-sex attraction and self-realization of a sexual minority identity (M= 2.07, SE= .13, p<.001) when compared to the visibility (M= 3.96, SE= .32, p<.001) and pride (M= 5.72, SE= .38, p=.001) cohorts, who also differed from one another. Pacing between self-realization and same-sex sexual behavior diverged across cohorts: the pride cohort experienced less time between self-realization and same-sex sexual behavior (M= 0.49, SE= .48) than the inclusion (M= 2.88, SE= .17, p<.001) and visibility cohorts (M=2.16, SE= .35, p=.014), who did not differ. All three cohorts differed in the time between same-sex sexual behavior and disclosure to friends: in the inclusion cohort (M= −0.66, SE= .16), disclosure occurred prior to same-sex sexual behavior on average, whereas in the visibility (M= 1.93, SE= .39) and pride cohorts (M= 5.23, SE= .39), disclosure followed same-sex sexual behavior. Finally, the inclusion cohort reported less time between disclosing to a friend and disclosing to a family member (M= 1.16, SE= .09) than the visibility (M=1.89, SE= .25, p=.022) and pride cohorts (M=2.27, SE= .30, p=.002), who did not differ from one another.

Table 3.

ANCOVA post-hoc pairwise comparison of cohort, sex, sexual identity, and race/ethnicity differences in pacing of sexual identity development milestones.

	Attraction - Realization			Realization - Sex			Sex - Disclosure (Friend)			Disclosure (Friend) - Disclosure (Family)
	M	SE	F	M	SE	F	M	SE	F	M	SE	F
Total	3.09	.13		2.23	.16		1.11	.16		1.52	.10
Range	−10–43			−41–34			−18 – 43			−20–33
Cohort			46.01***			10.81***			99.16***			8.30***
Inclusion (18–26)	2.07ab	0.13		2.88a	0.17		−0.66ab	0.16		1.16ab	0.09
Visibility (32–43)	3.96ac	0.32		2.16b	0.35		1.93ac	0.39		1.89a	0.25
Pride (50–60)	5.72bc	0.38		.49ab	0.48		5.23bc	0.39		2.27b	0.30
Sex			1.37			2.46			16.81***			2.63
Female	3.22	0.17		2.47	0.21		0.54a	0.19		1.38	0.13
Male	2.90	0.19		1.91	0.26		1.86a	0.25		1.71	0.15
Sexual Identity			1.36			8.60***			5.53**			3.83*
Lesbian/Gay	2.85	0.19		2.76a	0.22		0.84a	0.19		1.28a	0.13
Bisexual	3.34	0.21		1.29ab	0.29		1.77ab	0.30		1.66	0.17
Newer	3.19	0.35		2.62b	0.45		0.37b	0.37		2.12a	0.30
Sex x Sexual Identity			1.66			.36			6.02**			4.83**
Lesbian	2.81	0.31		3.12a	0.34		−0.28ab	0.27		1.00a	0.19
Bisexual Female	3.50	0.24		1.42a	0.28		1.55ac	0.31		1.53	0.17
Newer Female	3.56	0.39		2.94	0.49		−0.04cd	0.42		2.27a	0.35
Gay Male	2.77	0.22		2.38	0.27		1.93bd	0.26		1.56	0.17
Bisexual Male	3.06	0.36		1.31	0.62		1.31	0.62		1.85	0.35
Newer Male	2.01	0.64		1.99	1.00		0.40	0.65		1.08	0.37
Race			.15			4.22*			3.96*			.43
White	3.13	0.15		2.30	0.18		0.81a	0.18		1.53	0.11
Black	3.11	0.33		1.31a	0.40		1.98a	0.39		1.31	0.27
Latinx	2.98	0.24		2.79a	0.35		1.33	0.30		1.63	0.21

Note. The same superscripts indicate significant differences at the p<.05 level. Bonferroni post-hoc comparisons were used to minimize Type 1 error. Estimated marginal means adjusted for the effects of cohort, sexual identity, sex, and race/ethnicity are reported. Data are censored at age 8. See supplemental Table S2 for uncensored results. Race=Race/ethnicity.

^*p <.05

^**p<.01

^***p<.001.

Figure 1.

Cohort Differences in Milestone Timing

Note. Estimated marginal means adjusted for the effects of sexual identity, sex, and race/ethnicity are reported. Dotted lines represent 95% confidence intervals. Bonferroni post-hoc comparisons were used to minimize Type 1 error.

With respect to milestone pacing at the intersection of sex and sexual identity, no differences were observed in the time between same-sex attraction and self-realization of a sexual minority identity. Lesbians spent significantly more time between realizing a sexual minority identity and engaging in same-sex sexual behavior (M= 3.12, SE= .34) than did bisexual females (M= 1.42, SE= .28, p=.002). Lesbians (M= −0.28, SE= .27) and females with newer identities (M= −0.04, SE= .42) engaged in same-sex sexual behavior and disclosure to a straight friend nearly concurrently, with significantly less time between these milestones relative to gay males (M=1.93, SE= .26, p=.020) and bisexual females (M=1.55, SE= .31, p=.027), for whom same-sex sexual behavior preceded disclosure to a friend on average. Finally, females with newer identities spent more time between disclosing to a straight friend and disclosing to a family member (M= 2.27, SE= .35) than did lesbians (M=1.00, SE= .19, p=.025).

Few racial/ethnic differences were observed in milestones pacing. Latinx participants spent more time on average between realizing their sexual minority identity and engaging in same-sex sexual behavior (M= 2.79, SE= .35) than did Black participants (M= 1.31, SE= .40, p=.013). Additionally, Black participants spent more time between engaging in sexual behavior and disclosing a sexual minority identity to a friend (M= 1.98, SE= .39) than did White participants (M= .81, SE= .18, p=.022).

Intersectional Models

Table 4 presents ANCOVA results of race-stratified models testing interactions between sex, sexual identity, and cohort in the timing and pacing of milestones. Due to sample size limitations and given the relatively few differences between racial/ethnic groups in main effects models, we combined Black and Latinx racial/ethnic groups. Results showed no significant differences in milestone timing and pacing among White participants. Alternatively, among participants of color (i.e., Black and Latinx participants), three-way interactions were significant for age of disclosure to family members, pacing between first self-realization and first same-sex sexual behavior, and between first same-sex sexual behavior and disclosure to a straight friend.

Table 4.

ANCOVA F Values Testing Intersectional Models

	Attraction		Self-Realization		Same-sex Sex		Disclosure Friend		Disclosure Family
Milestone Timing	F	p	F	p	F	p	F	p	F	p
Cohort x Sex Id x Sex (White)	0.43	0.79	0.44	0.78	0.50	0.74	1.88	0.11	0.81	0.52
Cohort x Sex Id x Sex (POC)	0.72	0.58	2.23	0.06	0.71	0.58	1.56	0.18	2.80	0.03*

	Attraction- Self-Realization		Self-Realization-Sex		Sex - Disclosure (Friend)		Disclosure (Friend)-Disclosure (Family)
Milestone Pacing	F	p	F	p	F	p	F	p
Cohort x Sex Id x Sex (White)	1.05	0.37	0.81	0.52	0.85	0.49	0.98	0.42
Cohort x Sex Id x Sex (POC)	1.27	0.28	3.91	<0.01**	2.57	0.04*	1.21	0.31

Note. Data are censored at age 8. Sex Id= sexual identity. POC= People of color, i.e., Black and Latinx participants in our sample.

^*p <.05

^**p<.01

^***p<.001.

Figure 2 depicts race-stratified marginal mean ages of first disclosure to family at the intersection of cohort, sex, and sexual identity (See supplemental Table S3 for numeric values). Few significant differences emerged within the inclusion cohort: lesbians were younger (M= 15.22, SE= .41) when they first disclosed to family when compared to bisexual females (M= 17.31, SE= .30, p<.001). In contrast, subgroup differences emerged across cohorts. Lesbians from the inclusion cohort reported significantly younger ages of disclosure (M= 15.22, SE= .41) than lesbians from visibility (M= 19.83, SE= 1.01, p=.004) or pride cohorts (M= 24.63, SE= 2.36, p=.015), who did not differ from one another. Similarly, gay males from the inclusion cohort reported significantly younger ages of disclosure to family (M= 16.79, SE= .33) than gay males from the visibility (M= 22.57, SE= 1.07, p<.001) and pride (M= 25.31, SE= 1.83, p=.001) cohorts, who also did not differ from one another.

Figure 2.

ANCOVA analyses of Milestone Timing of Disclosure to Family at the Intersection of Cohort, Sex, and Sexual Identity

Note. Models are race-stratified. Estimated marginal means are reported. Bonferroni post-hoc comparisons were used to minimize Type 1 error. Error bars represent 95% confidence intervals. Youngest= inclusion cohort, ages 18–26. Middle= Visibility cohort, ages 32–43, Oldest= Pride cohort, ages 50–60. FPOC= Female, person of color. FW= Female, White. MPOC= Male, person of color. MW= Male, White. Bi= bisexual.

Figure 3 depicts the pacing between self-realization of a sexual minority identity and same-sex sexual behavior. Among the inclusion cohort, all subgroups reported positive mean values, meaning that on average, each subgroup realized that they were a sexual minority prior to engaging in same-sex sexual behavior. In contrast, some subgroups from the visibility and pride cohorts had negative means, indicating that on average, same-sex sexual behavior preceded self-realization. Although bisexual male respondents from the inclusion cohort reported almost a 3- year gap between first self-realization of a sexual minority identity and same-sex sexual behavior (M= 2.92, SE= .75), bisexual male respondents from the pride cohort reported nearly 4.5 years between first engaging in same-sex sexual behavior, and self-identifying with a sexual minority identity (M= −4.31, SE=.94, p <.001).

Figure 3.

ANCOVA analyses of Milestone Pacing of Self-Realization to Same-Sex Sexual Behavior at the Intersection of Cohort, Sex, and Sexual Identity

Note. Models are race-stratified. Estimated marginal means are reported. Bonferroni post-hoc comparisons were used to minimize Type 1 error. Error bars represent 95% confidence intervals. Youngest= inclusion cohort, ages 18–26. Middle= Visibility cohort, ages 32–43, Oldest= Pride cohort, ages 50–60. FPOC= Female, person of color. FW= Female, White. MPOC= Male, person of color. MW= Male, White. Bi= bisexual.

Finally, Figure 4 represents the time between same-sex sexual behavior and disclosure of a sexual minority identity to a straight friend. There were no significant differences in pacing between subgroups in the inclusion cohort. In contrast, there were generational differences across those who held similar identities across sex and sexual identity. For example, gay males in the inclusion cohort reported disclosing to a friend almost a year prior to same-sex sexual behavior (M= −0.83, SE= .39); this differed from those in the visibility cohort (M= 4.59, SE= 1.09, p=.001), who engaged in same-sex behavior 4.5 years prior to disclosing and those in the pride cohort (M= 7.28, SE= 1.16, p<.001), who engaged in same-sex behavior about 7 years before disclosing. The visibility and pride cohorts did not significantly differ from one another.

Figure 4.

ANCOVA analyses of Milestone Pacing of Same-Sex Sexual Behavior to Disclosure at the Intersection of Cohort, Sex, and Sexual Identity

Note. Models are race-stratified. Estimated marginal means are reported. Bonferroni post-hoc comparisons were used to minimize Type 1 error. Error bars represent 95% confidence intervals. Youngest= inclusion cohort, ages 18–26. Middle= Visibility cohort, ages 32–43, Oldest= Pride cohort, ages 50–60. FPOC= Female, person of color. FW= Female, White. MPOC= Male, person of color. MW= Male, White. Bi= bisexual.

Discussion

Using data from the first national probability sample of Black, Latinx, and White sexual minority people in the United States, the current study illuminated the degree to which sexual identity development timing and pacing varies across demographic subgroups occupying distinct social locations across cohort, sex, sexual identity, and race/ethnicity. Overall, our results suggest substantial variation in the developmental timing and pacing of milestones among sexual minority people, revealing the value of life course and intersectional approaches to the developmental science of sexual identity.

Consistent with our first hypothesis, milestone timing and pacing varied considerably by cohort. Those from the inclusion (youngest) cohort reported same-sex attraction on average two years earlier, self-realization five years earlier, and disclosure to friends and family a full decade earlier than those from the pride (oldest) cohort. Those from the inclusion cohort also progressed through milestones nearly twice as fast as those from the pride cohort. Our findings underscore the importance of a life course framework and its attention to historical context for understanding sexual identity development (Hammack et al., 2018; Russell & Fish, 2019). As sexual diversity becomes increasingly visible in the United States, contemporary cohorts of sexual minority people are coming out at younger ages (Russell & Fish, 2016). Future research must consider the distinct developmental contexts that follow from this dramatic cultural shift.

Our study found distinct milestone pacing across cohorts. The pride cohort reported nearly concurrent timing of self-realization and same-sex sexual behavior, with disclosure following 5 years later on average. The visibility cohort spent approximately two years between first realizing a sexual minority identity, engaging in same-sex sexual behavior, and disclosing a sexual minority identity. In the inclusion cohort, both self-realization and disclosure to a straight friend preceded same-sex sexual behavior on average. These findings are consistent with prior studies that have identified distinct trajectories of sexual identity development characterized by “sex-centered” and “identity-centered” patterns (Calzo et al., 2011; Dubé, 2000; Floyd & Bakeman, 2006). Whereas early models of sexual identity development (e.g., Troiden, 1989) theorized “sex-centered” processes whereby sexual activity was a central impetus for exploring sexual minority identities, our study aligns with more recent research (Calzo et al., 2011; Floyd & Stein, 2002; Savin-Williams & Diamond, 2000) which suggests that many sexual minority people report “identity-centered” development. Further, identity-centered patterns are more common among recent cohorts of sexual minority people (Floyd & Bakeman, 2006), reflecting the impact of recent historical changes on sexual identity development processes.

Our second and third hypotheses predicted differences in timing and pacing of milestones across subgroups defined by sex and sexual identity. Monosexual participants reported earlier mean ages of milestones than bisexual people, and this difference was predominantly driven by gay males. This finding is in line with our hypotheses and with prior evidence that males and monosexuals report relatively earlier milestones than their female and bisexual peers (e.g., Martos et al., 2015). Bisexual people have historically been subjected to biphobia (Wandrey et al., 2015) from both heterosexual and sexual minority communities. Such stigma could result in a delay in sexual minority identity development due to bisexual invisibility and prejudice (Wandrey et al., 2015). Continued research at the intersection of sex and sexual identity will help to advance understandings of the means by which gender and sexual identity intersect and are linked to sexual identity development timing.

Although gay males reported earlier milestones than other groups at the intersection of sex and sexual identity, they did not accelerate through milestones more quickly than others. Specifically, they disclosed their sexual identities about two years after their first same-sex sexual experience on average, whereas lesbians and females with newer identities reported near-concurrent ages of same-sex sexual behavior and disclosure. Additional analyses demonstrated that this gap was driven largely by gay males from the visibility and pride cohorts. These males may have spent more time between milestones because they came of age during the HIV/AIDS pandemic, a period in which gay identity was culturally associated with contamination, death, and disease (Hammack et al., 2018). Gay young men at that time may have therefore delayed sexual identity disclosure to avoid stigma and prejudice.

This study was among the first to compare the milestone timing and pacing for people with newer and increasingly common sexual identities such as pansexual and queer compared to monosexual and bisexual identities. Those with newer identities reported younger mean ages of first same-sex attraction, self-realization, and disclosure than those with bisexual identities and largely did not differ from those with traditional monosexual identities. Given that newer labels such as queer and pansexual are often thought to indicate more inclusive attraction across sexuality and gender spectra (Morandini et al., 2017), it was surprising that the timing of milestones more closely aligned with monosexual people than bisexual people. It may be that identification with newer identities signals specific political alignment, gender ideology, education, or socialization (Goldberg et al., 2020) rather than distinct sexual identity developmental trajectories. More research is needed to understand the motivations of those who endorse newer labels and their sexual identity development processes.

Although there were few differences across racial/ethnic subgroups, where there were differences, Black and Latinx participants reported younger ages of same-sex attraction and self-realization, respectively, than White participants. Additionally, Black participants reported earlier ages of first same-sex sexual behavior than White and Latinx participants. This finding is consistent with prior literature supporting earlier emergence of some milestones for racial/ethnic minorities (Dubé & Savin-Williams, 1999; Parks et al., 2004). Given that they already hold a minoritized racial/ethnic identity, Black and Latinx sexual minority people may integrate a marginalized sexual identity earlier (Parks et al., 2004). In other words, socialization into a racial/ethnic minority identity may better equip racial/ethnic minority people to understand and integrate sexual minority attractions and experiences. More research is needed to test these ideas. Specifically, qualitative studies may more closely interrogate the meaning and relative salience of sexual identity development processes for racial and ethnic minorities (Bowleg, 2008).

Race-stratified intersectional models revealed that among Black and Latinx participants in our sample, there were differences in disclosure to family among groups defined by sex, sexual identity, and cohort. Those in the inclusion cohort did not differ from one another, with the exception of lesbians and bisexual females. In contrast, many subgroup differences emerged across cohorts. For example, lesbians in the inclusion cohort came out 5 years earlier on average (about 15 years old) than lesbians in the visibility cohort (about 20 years old), who themselves disclosed their identity 5 years before lesbians in the pride cohort (about 25 years old). Taken together, these results suggest that above and beyond the intersection of important social identities, sociohistorical contexts shape the developmental unfolding of sexual identity.

We also observed differences in the time between self-realization of a sexual minority identity and same-sex sexual behavior, and between same-sex sexual behavior and disclosure to a straight friend among participants of color in our sample. Cohort differences were prominent, affirming that identity-centered (relative to sex-centered) patterning was increasingly common in more recent cohorts. There has been increasing visibility in the last two decades with regard to sexual diversity and language to express sexual minority identities. In recent cohorts, people less commonly “confirm” their sexual minority identities by first engaging in same-sex sexual behavior, whereas in older generations, patterns were more consistent with either engaging in sexual behavior but not thinking of oneself as a sexual minority, or confirming one’s sexual minority identity by engaging in same-sex sexual behavior. The pattern for younger cohorts—understanding one’s sexual attractions before becoming sexually active—is more consistent with what is considered typical adolescent sexual development (Tolman & McClelland, 2011).

Overall, our study provides a descriptive understanding of differences in milestone timing and pacing across subgroups of sexual minority people. Future research should focus on linking these sexual identity developmental processes with individual adjustment and health. For example, insight into which sexual identity development processes will continue to occur earlier for future cohorts could assist in creating age-appropriate interventions aimed at coping with minority stress. Further, researchers could activate this study’s findings to examine associations between milestone timing and mental health, substance use, and academic achievement (e.g., Fish & Pasley, 2015). Last, this information can be used to address a perennial challenge in the field of sexual identity development: understanding sexual minority health both before and after sexual identity disclosure. With accurate measures of milestone timing across subgroups, scholars can design longitudinal investigations of how health vulnerabilities such as suicidality or substance use change before versus after developmental milestones.

Limitations

The cohort design of the study raises three methodological areas for future inquiry. First, mean ages for milestones in some groups exceeded the age range of those from the inclusion (youngest) cohort. It may be that some younger participants had not yet experienced some milestones. Future research should follow participants prospectively to track milestones that may occur in later years, as well as fluidity of sexuality (Diamond, 2006). Second, recall bias may have differentially influenced participant responses such that bias may be greater among older cohorts, for whom more time would have passed since their first sexual identity developmental experiences. Although there is evidence that sexual minority youth’s self-reports of sexual behavior and orientation are reliable across time (Scrimshaw et al., 2006), future cognitive testing of milestones measures should explicitly assess the potential recall bias inherent in such measures. Third, the age of puberty may be declining over time (Euling et al., 2008), which may be linked to age differences in milestones observed across cohorts in our sample. Subsequent studies should account for pubertal timing when examining milestones.

Prior studies of the timing and pacing of milestones have typically allowed participants to report any age for milestones. We censored responses of participants who reported very young ages of milestones that are inconsistent with developmental understandings of the emergence of sexuality (Herdt & McClintock, 2000). Decisions regarding censoring were theoretically supported and empirically tested with sensitivity analyses. Future studies should utilize cognitive testing, qualitative studies, and sensitivity analyses to contribute to deeper understanding of best practices for analyzing data about sexual identity development.

The current study reflects a sample that was recruited to represent sexual minority people from three racial/ethnic groups in the United States. Although this design was required for developing a national probability sample of sexual minority adults for which robust race/ethnic group comparisons could be made, sexual minority people of other races who did not also identify as White, Latinx, or Black were excluded (Meyer et al., 2020). Moreover, although our study included participants who identified with newer labels (e.g., queer, pansexual), the recruitment strategy focused on people who identified as LGB; thus our findings regarding those with newer identities may not represent the full range of people with such identities, many of whom may not have elected to participate in a study of LGB people.

Last, milestones do not fully represent how sexual identity development unfolds within people and over time. In fact, prior studies suggest that for many, sexual identity development is fluid and nonlinear (e.g., Diamond, 2006). Longitudinal and/or qualitative studies may better capture that sexual identity development cannot be understood through a series of singular events that are thought to “conclude” identity development processes. Rather, our study contributes a better understanding of the “beginnings” of these processes, and more specifically, when and how these beginnings may be variable rather than monolithic.

Conclusion

Sexual minority identity development varies across social identities and contexts. This variability requires intersectional and life course approaches to theory and research. Shedding light on the factors that contribute to differences in the timing and pacing of milestones may provide critical insight into the developmental and social drivers of sexual minority people’s health trends, and is crucial for tailoring prevention and interventions efforts aimed at reducing disparities in health and well-being.

Given earlier recognition of sexual minority identity in recent cohorts, deeper understandings of sexual identity development during adolescence are needed. Whereas sexual minority identity was once fundamentally defined by one’s sexual partners, sexual minority identity processes are now occurring in early adolescence and may center the development of an identity related to, yet independent of, sexual behavior. As sexual minority identity developmental processes begin earlier and take a faster developmental pace, intersectional and life course theories should integrate models of adolescent development in order to understand crucial precursors and contexts of sexual identity development among diverse populations.