Sexual Identity Continues to Shift for an Unexpectedly Long Time

Data on 12,000 adults shows why labels like "straight," "gay," and "bi" don't suffice.

holding hands

Decades of research on sexual preference and orientation have pointed to one simple fact: No label can fully capture the range of human sexuality. Traditional terms like gay, straight, and bisexual are just a few that can describe who a person is attracted to or has had relationships with in the past. But identity can change over time, and any one label can end up being too limiting. A new study in The Journal of Sex Research shows that sexuality changes well past the point we might assume our identity becomes cemented.

In the recent paper, Christine Kaestle, M.P.H, Ph.D., presents data from 12,000 adults showing how much sexuality shifts during the teenage years into a person’s 20s. Kaestle tracked people from their late teenage years into their late 20s and early 30s, accounting for various aspects of sexuality by tracking not only how people self-identified but also who they reported being attracted to and who they had sexual relationships with.

“Longitudinal patterns show substantial change through the early and late 20s,” writes Kaestle, an associate professor of human development and family science at Virginia Tech. This substantial change, she writes, indicates that developing one’s sexual orientation is not just something a person can take for granted from the get-go but a journey that spans young adulthood.

lesbian couple, marriage
These survey results indicate that different dimensions of sexuality can change dramatically from adolescence to adulthood.

This doesn’t mean that being attracted to someone of the same sex is just a phase of adolescence, though. What Kaestle emphasizes is that this survey shows just how complicated sexuality is, because it clearly can’t be encapsulated with any single measure.

“Sexual orientation involves many aspects of life, such as who we feel attracted to, who we have sex with, and how we self-identify,” Kaestle said.

In the paper, she argues that past surveys that attempted to track and assess sexual orientation and attraction have missed some of these subtle nuances because they only asked about a single measure of sexuality, whether that was sexual attraction, sexual relationship behavior, or sexual identity. In this way, much of the research on sexuality has missed the mark for a lot of people since these factors can differ over time. Not only that, but people can report that they are heterosexual, for instance, while also engaging in relationships with same-sex partners.

In addition to revealing a temporal aspect to sexual identity, the paper also revealed differences in the way sexual identity differs between men and women, which led Kaestle to create nine categories of sexuality between the sexes.

Based on her data on self-reported feelings and behaviors, she classified men into four categories: straight, minimal sexual expression, mostly straight/bi, and emerging gay. For women, she identified all four of these categories but also added one more: mostly straight discontinuous. (The category that is most likely to have only same-sex relationships and attractions is labeled “emerging” because Kaestle’s study was more about mapping the journey of sexuality than the final destination.)

Sexuality isn't just about sex, it also involves attraction and relationships.
Sexuality isn't just about identity; it also involves attraction and relationships.

Even within the straight category, Kaestle reports a significant number of survey respondents reported attraction to people of the same sex at different points in their lives, especially during the teenage years, even if they never acted on it. To be clear, this group — which represented the majority of both men and women — showed the fewest changes over time.

Individuals in the mostly straight/bi groups reported largely opposite-sex attraction and relationships but also increasing levels of same-sex attraction and relationships, especially in their early 20s. The mostly straight discontinuous category described women who reported attraction to both men and women in their early 20s but reported almost exclusively same-sex attraction in their late 20s.

Kaestle highlights how important it is to consider various dimensions of behavior and attraction when talking about sexuality, emphasizing that it’s an especially dynamic process for people in their teens or 20s.

“Examining multiple dimensions of identity, romantic experiences, and sexual behavior longitudinally, and in combination rather than individually, provides new insight into the developmental experiences of different feelings, the exploration of behaviors, and the sometimes evolving nature of sexual orientation identity in adolescence into adulthood, with unique results for males and females,” she writes.

Kaestle’s study is an important advancement in our understanding of the dynamism of human sexuality, but even she admits it has room to be more inclusive. Since it treats gender as a binary, it doesn’t give a clear analysis of the changing identities of transgender or gender-nonconforming young people. And due to her limited sample size, she points out there could be categories that exist that have not been identified yet, suggesting there are infinite identities in between, all of which contribute to the richness and complexity of human sexuality.

Abstract: Defining sexual minority status longitudinally over critical developmental periods is essential for understanding the roots of health disparities. Theory supports multidimensional continuums, but current research often examines single measures of sexual activity, sexual attractions, or self-labeled identity separately. Here, a new typology of longitudinal latent classes describes dynamic multidimensional processes continuing from late adolescence (ages 16 to 18) through the late 20s. Using Add Health data (N = 6,864), longitudinal latent class analysis (LLCA), a person-centered approach, showed significant differences between the orientation experiences of males and females (invariance tests led to stratification by sex). The male LLCA model predicted four classes: straight males (87.4%), minimal sexual expression males (6.5%), mostly straight and bi males (3.8%), and emerging gay males (2.4%). The female LLCA model predicted five classes: straight females (73.8%), minimal sexual expression females (7%), mostly straight discontinuous females (10.2%), emerging bi females (7.5%), and emerging lesbian females (1.5%). Some classes represent generally consistent indicators across dimensions over time, while other classes describe more emerging or discontinuous trajectories. Substantial changes were common not only from late adolescence to the early 20s but also from the early 20s to the late 20s, indicating that sexual orientation development continues throughout emerging adulthood.