Psychosocial Stressors Explaining the Monosexual–Bisexual Disparity in Mental Health: A Population-Based Study of Sexual-Minority Young Adults

Abstract

Bisexual people report greater mental-health problems (i.e., depression, anxiety, suicidality) compared with their monosexual (i.e., gay or lesbian) peers. Yet existing studies often use nonprobability samples, analyze few psychosocial stressors, and overlook bisexual people’s considerable diversity. We analyzed data from a population-based study of sexual-minority young adults in Sweden (N = 748) that assessed identity-related stressors (e.g., family rejection) and general life stressors (e.g., financial loss). Bisexual respondents reported more mental-health problems and general life stressors, but fewer identity-related stressors, than monosexual respondents. Latent class analysis revealed three distinct bisexual subgroups with varying patterns of gender-based sexual attractions, gender of sexual partners, gender conformity, and sexual-identity centrality that were associated with unique patterns of psychosocial stressors and mental health. Findings show that general life stressors play an important role in bisexual people’s mental health. Future research is needed, especially on the role these stressors play during critical developmental periods such as young adulthood.

The disparity between bisexual (i.e., sexual and/or romantic orientations toward people of more than one gender) and monosexual (i.e., sexual and/or romantic orientations toward people of only one gender) individuals in internalizing mental-health problems (e.g., depression, anxiety, suicidality) is well established within the sexual-minority population (Jorm et al., 2002; Persson & Pfaus, 2015). People with bisexual and other nonmonosexual identities (e.g., pansexual), collectively referred to herein as “bisexual,” represent one of the highest risk populations for internalizing mental-health problems, with these disparities being particularly large in young adulthood. However, reasons for the monosexual–bisexual disparity in internalizing mental-health problems are relatively unclear.

Minority stress theory (Brooks, 1981; Meyer, 2003), which posits that the heightened stress resulting from exposure to stigma-motivated victimization, discrimination, and rejection significantly undermines mental health, is the foremost explanation for the observed greater mental-health burden borne by sexual-minority compared with heterosexual individuals (Bränström et al., 2020). Tenets of minority stress theory have more recently been applied to investigate psychosocial stressors potentially contributing to the observed monosexual–bisexual mental-health disparity among sexual-minority people (Feinstein & Dyar, 2017). This research, conducted among community samples, has found that bisexual individuals, compared with their gay and lesbian counterparts, report higher levels of at least some external minority stressors such as exposure to victimization and assault (Bender & Lauritsen, 2021) and lower LGBTQ-community connection (Balsam & Mohr, 2007) and higher levels of at least some internal minority stressors, including sexual-orientation concealment, internalized stigma, and identity conflict (Balsam & Mohr, 2007; Chan et al., 2020; Persson et al., 2015). Some research suggests that bisexual people also face unique types of bisexual discrimination (e.g., identity invalidation) that negatively influences mental health (Feinstein et al., 2022), which may explain higher levels of discrimination reported among bisexual individuals in some studies (Bränström et al., 2020). However, other studies find that bisexual people report lower levels of discrimination compared with their gay and lesbian peers given their greater tendency to conceal their sexual identities (e.g., Colledge et al., 2015). In addition to experiencing unique patterns of external and internal minority stressors compared with monosexual (i.e., gay or lesbian) individuals, research shows that bisexual individuals experience more exposure to general life stressors known to predict poorer mental health compared with their gay and lesbian peers (Jorm et al., 2002).

Although this existing research has been instrumental in suggesting the possibility that bisexual individuals have more exposure to certain types of psychosocial stressors, including identity-related and general life stressors, compared with both gay and lesbian and heterosexual individuals, several methodological limitations undermine the ability to draw conclusions about the role of these disparities in explaining the monosexual–bisexual disparity in internalizing mental-health problems. Most notably, this research has tended to not examine whether these elevations in certain psychosocial stressors serve as mediators of the monosexual–bisexual disparity in internalizing mental-health problems, and no studies have done so using probability samples. However, one study using nonprobability sampling conducted in Hong Kong with sexual-minority adults demonstrated that the monosexual–bisexual disparity in depression and anxiety was mediated by bisexual individuals’ greater sexual-orientation concealment and lower LGBTQ-community connectedness (Chan et al., 2020). Similarly, another study involving sexual-minority women in Canada revealed that the monosexual–bisexual disparity in depression and anxiety was mediated by bisexual individuals’ greater sexual-orientation concealment and bisexual women’s higher engagement in risky sexual behavior (Persson et al., 2015).

Although these studies conducted in nonprobability samples contribute to an emerging picture of psychosocial explanations for the monosexual–bisexual disparity in internalizing mental-health problems, nonprobability samples of sexual minorities include those who are less concealed, more connected to the LGBTQ community, and more psychologically distressed than those in population-based samples (Hottes et al., 2016). Consequently, these studies likely underrepresent the sizeable proportion of bisexual people in the general population who are concealed and not connected to the LGBTQ community and might overrepresent those who are more psychologically distressed, thereby introducing bias into associations among these factors. Furthermore, existing studies have not examined the role of a comprehensive set of external minority stressors, internal minority stressors, and general life stressors as potential psychosocial contributors to the monosexual–bisexual disparity in internalizing mental-health problems. Additionally, although some existing studies have assessed anxiety and depression, they have not assessed suicidality, despite recent evidence showing that bisexual individuals experience substantially elevated suicide risk compared with their monosexual (e.g., gay and lesbian) peers (Salway et al., 2019).

Previous research with bisexual individuals has also tended to treat this population as a homogeneous population, without considering subgroup differences across the diversity of bisexual manifestations and experiences. However, as the most sizable group of sexual-minority individuals (Bränström et al., 2020), bisexual individuals vary in terms of at least four factors that capture distinct experiences of bisexuality and are associated with mental health. First, bisexual individuals, especially women, vary widely in terms of the gender of the people to whom they experience sexual and romantic attractions (Diamond, 2008; Farr et al., 2014; Galupo et al., 2017). Across sexual orientations, an individual’s gender-based sexual attractions are associated with mental health such that people reporting sexual attractions only to people of the same gender or to people across multiple genders have poorer mental health than those reporting only other-gender sexual and romantic attractions (Teasdale & Bradley-Engen, 2010). Second, the gender of one’s sexual partners has been shown to predict the mental health of bisexual people (Dyar et al., 2014), with bisexual men and women who engage in sexual behavior with both other men and women reporting greater risk than those who engage in sexual behaviors with only people of the other gender (Bostwick et al., 2010; Dyar et al., 2019). Third, some (Li et al., 2016), although not all (Baams et al., 2013), studies have found that bisexual individuals are more gender conforming than gay and lesbian individuals and that gender conformity is associated with better mental health among sexual minorities, at least among men (Roberts et al., 2013). Fourth, sexual-identity centrality, or the prominence of one’s sexual identity to one’s overall self-concept, tends to be lower among bisexual compared with monosexual people (Hinton et al., 2022). Among bisexual people, sexual-identity centrality is associated with mental health, with one longitudinal study showing that those with more central bisexual identities experienced decreasing anxiety over time (Flanders, 2015). Although previous research has highlighted the significance of factors such as gender-based sexual attractions, gender of one’s sexual partners, one’s own gender conformity, and one’s sexual-identity centrality as correlates of the mental health of bisexual individuals, no research has considered how these factors might be related to monosexual–bisexual differences in psychosocial stressors and internalizing mental-health problems.

In sum, the limited existing research on the monosexual–bisexual disparity in internalizing mental-health problems has relied on nonprobability samples, investigated a restricted set of psychosocial stressors, and typically categorized bisexual people as a monolith despite their notable diversity. Overcoming these limitations requires a population-based data set that includes a sizeable sample of bisexual and monosexual sexual-minority individuals, measures of both identity-specific and general psychosocial stressors and a comprehensive assessment of the diversity of the bisexual experience across gender-based sexual attractions, gender of sexual partners, gender conformity, and sexual-identity centrality. The Pathways to Longitudinally Understanding Stress (PLUS) cohort study represents a rare combination of these methodological strengths and allows pursuing the aims of the current study, which were to (a) establish the monosexual–bisexual disparity in internalizing mental-health problems in a population-based sample; (b) empirically distinguish between unique subgroups of bisexual respondents across dimensions of gender-based sexual attractions, gender of sexual partners, one’s own gender conformity, and one’s sexual-identity centrality; (c) assess internalizing mental-health problems and a comprehensive set of identity-specific and general psychosocial stressors across the bisexual subgroups uncovered above; and (d) examine the role of these psychosocial stressors in explaining monosexual–bisexual disparities in internalizing mental-health problems, including depression, anxiety, and suicidality.

Transparency and Openness

We report how we determined our sample size, all data exclusions, all manipulations, and all measures in the study. All data were produced under the Swedish Statistics Act and the European Union Data Protection Regulation, according to which privacy concerns restrict the availability of personal data for research. Aggregated data can be made available by the authors, subject to ethical vetting. Data syntax for all analyses can be found at https://osf.io/53hyv/. This study was not preregistered. All procedures involving human subjects were approved by the Stockholm Regional Ethical Review Board.

Method

Participants and data

This study utilized data from the Wave 1 survey assessment of the PLUS cohort study. PLUS is a population-based cohort study of young adults in Sweden designed to prospectively investigate the mechanisms contributing to the increased risk of common mental-health problems among sexual-minority young adults (Bränström et al., 2023; Pachankis et al., 2024). Respondents for the PLUS cohort study were recruited from the 2015, 2016, and 2018 Swedish National Public Health Survey (SNPHS), a nationally representative health survey conducted by the Public Health Agency of Sweden. To assemble the PLUS cohort, all 2,973 respondents between the ages of 17 and 34 years old who indicated a nonheterosexual sexual orientation in the 2015, 2016, and 2018 SNPHS were invited to participate along with a randomly selected comparison group of 2,973 heterosexual respondents of the same age range and from the same survey years. Invitations to participate in the PLUS cohort study were sent via mail to the respondents’ home addresses, guiding them to an online informed consent form available in both Swedish and English. A total of 2,222 respondents provided consent and completed the Wave 1 survey assessment in October 2019.

To investigate mental-health disparities between bisexual and monosexual individuals and uncover distinct subgroups among bisexual respondents, we made several analytic choices to define the study sample. First, we restricted the sample to 823 respondents who identified as either monosexual (i.e., gay or lesbian) or bisexual (i.e., bisexual or pansexual). Then, to facilitate a clear delineation of gender-based categorization of sexual attraction and gender of partners (e.g., same-gender, other-gender) and minimize potential problematic assumptions (Matsuno & Budge, 2017), we dropped the 9.1% (n = 75) of sexual-minority respondents who reported a “genderqueer,” “nonbinary,” or “other” gender identity or who reported primary sexual attraction to genderqueer/nonbinary individuals or a history of genderqueer/nonbinary sexual partners. We acknowledge the nuanced and complex nature of genderqueer/nonbinary identities among bisexual people that warrants investigation beyond the scope of this study. Therefore, the final analytic sample included 748 young adults comprising 184 monosexuals and 564 bisexuals.

Measures

Sexual orientation.

We assessed respondents’ sexual orientation with the following question: “Which of the following best represents how you think of yourself?” Respondents’ options were “lesbian or gay”; “straight, that is, not lesbian or gay”; “bisexual”; or “something else.” Those who responded “something else” were then presented with a second question: “What do you mean by something else?” Response options for this question were “queer,” “pansexual,” “asexual,” or “demisexual.” As reported earlier, the current study included sexual-minority respondents who reported a monosexual (i.e., gay or lesbian) or bisexual (i.e., bisexual or pansexual) sexual orientation. Respondents identifying as queer were excluded from the study because of the broad nature of the identity, which encompasses individuals with both monosexual and multigender attractions within the sexual-minority community (Worthen, 2023). Consequently, we were unable to categorize individuals as monosexual or bisexual on the basis of queer self-identification.

Gender.

Gender was measured by the following question: “What is your current gender identity?” Response options were “male/man,” “female/woman,” “trans male/trans man,” or “trans female/trans woman” (The GenIUSS Group, 2014). On the basis of these self-reported options for gender, nine transgender men and four transgender women were represented in the current study. To facilitate gender-based classifications of bisexual identity, as described below, we created a binary gender category denoting men (inclusive of transgender men) and women (inclusive of transgender women).

Indicators of bisexual experience.

The following four indicators of bisexual experience were assessed: gender-based sexual attraction, gender of sexual partners, one’s own gender conformity, and one’s sexual-identity centrality.

Gender-based sexual attraction.

Gender-based sexual attraction was measured by the question “In general, are you mostly sexually attracted to . . .” with the following mutually exclusive response options: “males (men),” “females (women),” “transgender men,” “transgender women,” or “both males (men) and females (women)” (The GenIUSS Group, 2014). To ensure inclusivity and consistency with classification of respondents’ own gender, we adopted an approach grouping gender-based sexual attraction to men and transgender men as one category (referred to as “men”) and women and transgender women as another category (referred to as “women”). We then created a three-level categorical indicator variable classifying respondents according to their reported gender-based sexual attractions: attraction to the same gender, attraction to the other gender, or attraction to both genders.

Gender of sexual partners.

Information on the gender of one’s sexual partners was assessed by the question “In your lifetime, have your sexual partners been (select that all that apply) . . .” with the following possible response options: “males (men),” “females (women),” “both males (men) and females (women),” “transgender men,” “transgender women,” or “I have never had sex” (The GenIUSS Group, 2014). Again, we grouped men as one category and women as another category, inclusive of transgender men and women, respectively. This approach yielded a four-level categorical variable classifying respondents on the basis of whether they exclusively engaged in sexual activity with the same gender, the other gender, both genders, or had never engaged in sexual activity.

Gender conformity.

We assessed gender conformity with two questions measuring appearance and mannerisms, respectively: “On average, how do you think people would describe your appearance, style, or dress?” and “On average, how do you think people would describe your mannerisms?” (The GenIUSS Group, 2014). Both questions were assessed on a scale of 1 (very feminine) to 7 (very masculine), and responses to these two questions were averaged (α = .88). From this average, we created a binary indicator variable such that women (men) who scored 4 (equally feminine and masculine) or higher (lower) were assigned 1 (gender nonconforming) or 0 (gender conforming).

Sexual-identity centrality.

To assess sexual-identity centrality, we used the identity-centrality subscale of the Lesbian, Gay, and Bisexual Identity Scale (Mohr & Kendra, 2011), which consists of four items asking about the degree to which one’s sexual-minority identity is central to their sense of self (e.g., “To understand who I am as a person you have to know that I’m LGB”) on a scale of 1 (disagree strongly) to 6 (agree strongly; α = .82). Item responses were averaged to create an overall score and then trichotomized to create the following three-level categorical indicator variable of sexual-identity centrality: respondents who reported less central sexual identity (scored < 3), respondents who reported moderate sexual-identity centrality (scored from 3 to 4), and respondents who reported more central sexual identity (scored > 4).

Sociodemographic factors.

Respondents reported their age, nativity status (Swedish-born or immigrant), and level of education.

Psychosocial stressors.

External minority stressors. Respondents completed measures of the following four external minority stressors: family rejection, discrimination, sexual-orientation victimization, and LGBTQ-community (non)connectedness. Past-year family rejection was assessed by the family-reaction subscale of the Gay-Related Stress Scale (Lewis et al., 2002), a nine-item measure of experiences of family rejection in the past 12 months (e.g., “Distance between me and my family due to my sexual orientation”). For each item, respondents selected “yes” or “no,” and an average score was generated, with higher scores denoting more exposure to past-year family rejection (α = .70). Past-year discrimination was assessed by the Everyday Discrimination Scale (Williams et al., 1997), a 10-item scale asking respondents to report how frequently they had experienced interpersonal stigma, prejudice, and discrimination over the past 12 months (e.g., “People acted as if they are better than you”) on a scale from 1 (never) to 4 (often). Responses were averaged to provide an overall score, with higher scores reflecting more experiences of discrimination (α = .89). This measure was used to capture general experiences of discrimination rather than those specific only to sexual orientation, consistent with previous research involving sexual-minority individuals (Frost et al., 2015). Past-year sexual-orientation victimization was assessed by the Sexual Orientation Victimization Questionnaire (D’Augelli & Grossman, 2001), a seven-item scale asking respondents to report how frequently they had experienced victimizing events in the past year because of their sexual orientation (e.g., “threats of violence” or “verbal insults”) on a scale from 0 (never) to 3 (three or more times). Responses were averaged to provide an overall score, with higher scores reflecting greater frequency of sexual-orientation victimization experiences (α = .64). LGBTQ-community connectedness was measured by the Community Connectedness Scale (Frost & Meyer, 2012), a three-item scale asking respondents to what extent they feel connected to their “town or city’s LGBTQ+ community” on a scale from 1 (disagree strongly) to 4 (agree strongly). Responses were averaged for an overall score, with higher values reflecting a stronger connection to the LGBTQ community (α = .81).

Internal minority stressors.

Respondents completed measures assessing the following key internal minority stressors: sexual-minority rejection sensitivity, sexual-orientation concealment, and internalized stigma. Rejection sensitivity was assessed with an adapted version of the Gay-Related Rejection Sensitivity Scale (Pachankis et al., 2008), in which, to reduce participant burden, we selected only six of the original 14 items. For each item, respondents were presented with an ambiguous situation of sexual-minority-related rejection (e.g., “Your colleagues are celebrating a coworker’s birthday at a restaurant. You are not invited.”) Respondents then rated (a) how concerned or anxious they would be that the situation happened because of their sexual orientation on a scale of 1 (very unconcerned) to 6 (very concerned) and (b) how likely they believe the situation occurred because of their sexual orientation on a scale of 1 (very unlikely) to 6 (very likely). To calculate an overall score, the concern and likelihood scores for each scenario were multiplied, and then the product scores were averaged (concern α = .82, likelihood α = .83). Sexual-orientation outness was measured by asking respondents to rate how much they were “out of the closet” to six different groups of close people in their lives (i.e., family members, straight friends, LGBTQ+ friends, health-care providers, coworkers, neighbors) on a scale of 0 (out to none) to 3 (out to all; Meyer et al., 2002). Responses were averaged to obtain an overall score, with higher values reflecting a greater level of outness (i.e., lower level of concealment). Internalized stigma was assessed with the internalized homonegativity subscale of the Lesbian, Gay, and Bisexual Identity Scale (Mohr & Kendra, 2011), which consists of three items asking the respondents about negative self-evaluations of their sexual orientation (e.g., “If it were possible, I would choose to be straight”) on a scale of 1 (disagree strongly) to 6 (agree strongly). Responses were averaged, with higher scores reflecting a greater degree of internalized stigma (α = .84).

General life stressors.

General life stressors were evaluated using the Life Events List (Cohen et al., 1991), which assesses the occurrence and nature of major life events within the preceding 12 months. Participants were prompted to indicate whether each event had affected them personally or someone else (e.g., their spouse or partner). To derive a tally of general life stressors, items addressing personal events typically perceived as negative and stressful were aggregated (e.g., “Have you been hospitalized for a serious, i.e., life-threatening, illness during the past 12 months?”) For events with uncertain significance (e.g., “Have there been significant changes in your personal finances during the past 12 months?”), subsequent questions determined whether the change was positive or negative (e.g., “Has the change been for the better or worse?”), and those rated as negative were included in the count of general life stressors. A total of 21 negative and stressful events (e.g., job loss, separation or divorce, death of a loved one, financial loss, illness, relationship problems) were summed to compute an overall score, with higher scores indicating a greater frequency of general life stressors.

Internalizing mental-health outcomes.

The following three common mental-health problems were assessed: depression, anxiety, and suicidality. Depression was assessed by the 20-item Center for Epidemiologic Studies Depression Scale (Radloff, 1977), which measures depressive symptomatology experienced in the past week (e.g., “I felt sad” or “I felt lonely”) on a scale from 0 (rarely or none of the time) to 3 (most or all of the time). Items are summed to create a total overall score, with higher scores reflecting greater depression (α = .92). Anxiety was assessed with the Generalized Anxiety Disorder–7 (Spitzer et al., 2006), a seven-item scale assessing anxiety symptoms over the past 2 weeks (e.g., “Not being able to stop or control worrying”) on a scale from 0 (not at all) to 3 (nearly every day). Items are summed to create an overall score, with higher scores reflecting greater anxiety (α = .88). Suicidality was measured with the five-item Suicidal Ideation Attributes Scale (van Spijker et al., 2014). The first item asks respondents to report the past-month frequency of their suicidal thoughts from 0 (never) to 10 (always); those who report > 0 then receive four additional items assessing their level of control over and distress related to these suicidal thoughts. Items are summed to create a total score, with higher scores reflecting greater severity suicidal ideation (α = .90).

Statistical analysis

Analyses proceeded in five sequential steps. First, we computed means, standard deviations, and Pearson correlations for all study variables. Second, we examined which psychosocial stressors would explain the monosexual–bisexual disparity in internalizing mental-health problems (i.e., depression, anxiety, and suicidality) through hierarchical regression models run separately for each outcome. Across four steps for each mental-health outcome, we introduced covariates (Model 1), then external minority stressors (Model 2), then internal minority stressors (Model 3), and finally general life stressors (Model 4). This approach was utilized to ascertain the incremental impact of various psychosocial stressors on the relationship between bisexual versus monosexual sexual orientation and mental-health outcomes. Third, latent class analysis (LCA) was conducted to uncover subgroups based on the four indicators of diversity within the bisexual experience: gender-based sexual attraction, gender of sexual partners, one’s own gender conformity, and one’s own sexual-identity centrality.

Briefly, LCA is a person-centered statistical technique used to identify unobservable, or latent, subgroups within a population on the basis of patterns of responses to observed categorical indicator variables (Weller et al., 2020). LCA has been feasibly used in previous studies to uncover latent subgroups of sexual-minority people based on indicators of sexual identity, attraction, and behavior (Dawson et al., 2024). In the current study, we applied LCA to construct models with between two and six classes. To facilitate subsequent analyses comparing bisexual classes to the group of monosexual respondents, we used an LCA approach in which we assigned all monosexual respondents an arbitrary value (i.e., 999) on all indicator variables, effectively constraining them to a single class. This allowed us to observe respondents with a monosexual sexual-orientation identity separately from latent classes of bisexual respondents determined on the basis of their patterns of responses to the observed categorical indicators. A four-class solution (i.e., three bisexual latent classes and the constrained monosexual class) was selected by evaluating fit statistics, including the Akaike information criterion (AIC), the Bayesian information criterion (BIC), the bootstrap likelihood ratio test, and entropy, as well as considering class size, stability, and interpretability (Weller et al., 2020).

Next, we used Mplus (Version 8.11; Muthén & Muthén, 2017) to implement the modified Bolck-Croon-Hagenaars (BCH) approach to conducting LCA with distal outcomes to estimate means, standard errors, and mean difference scores on all mental-health outcomes and psychosocial stressors between respondents with a monosexual sexual-orientation identity and the three latent subgroups of bisexual individuals (Asparouhov & Muthén, 2014; Bakk and Vermunt, 2016). Briefly, the BCH approach is the currently recommended method for modeling associations between latent classes and distal variables (e.g., mental-health outcomes) because it uses a weighted multiple group analysis that attenuates bias introduced as a result of measurement error in latent class assignment (Asparouhov & Muthén, 2014). Last, we conducted a hierarchical regression analysis similar to the one in the second step. In this iteration, we used a four-level independent variable representing the LCA solution that separates respondents into three bisexual subgroups and one monosexual subgroup, the latter of which was set as the reference group in regression analyses. To account for potential measurement error introduced through LCA, we performed this hierarchical regression in Mplus using the modified BCH method (Asparouhov & Muthén, 2014).

Results

Sample descriptive information and Pearson correlations

Table 1 depicts descriptive information and Pearson correlations among all study variables. Of the 748 young adult respondents, 75% identified their sexual-orientation identity as bisexual and 25% identified their sexual-orientation identity as monosexual. Pearson correlations revealed statistically significant associations between reporting a bisexual sexual orientation and internalizing mental-health outcomes. Specifically, reporting a bisexual (vs. monosexual) sexual orientation was positively correlated with symptoms of depression (r = .16), anxiety (r = .14), and suicidality (r = .12); reporting a bisexual sexual orientation was also positively associated with general life stressors (r = .12). Conversely, a negative correlation was observed between reporting a bisexual sexual orientation and family rejection (r = −.13), sexual-orientation victimization (r = −.12), LGBTQ-community connectedness (r = −.21), sexual-orientation outness (i.e., reflecting higher sexual-orientation concealment; r = −.24), and internalized stigma (r = −.21).

Table 1.

Descriptive Statistics and Correlations for Study Variables

Variable	n	M	SD	Min.	Max.	1	2	3	4	5	6	7	8	9	10	11
Bisexual identitya	748	0.75	0.43	0.00	1.00	—
Depression	748	21.02	12.31	1.00	57.00	.16*	—
Anxiety	748	8.42	5.44	0.00	21.00	.14*	.80*	—
Suicidality	743	6.03	10.24	0.00	50.00	.12*	.68*	.55*	—
Family rejection	745	0.12	0.17	0.00	0.78	−.13*	.18*	.15*	.10*	—
Discrimination	746	1.55	0.54	1.00	3.70	.07	.44*	.40*	.34*	.26*	—
Sexual-orientation victimization	745	0.08	0.19	0.00	1.43	−.12*	.17*	.12*	.16*	.32*	.36*	—
LGBTQ-community connectedness	715	2.02	0.78	1.00	4.00	−.21*	−.07*	−.03	−.08*	.10*	−.05	.10*	—
Sexual-minority rejection sensitivity	743	8.92	6.65	1.00	36.00	−.02	.25*	.25*	.19*	.20*	.27*	.17*	.15*	—
Sexual-orientation outness	741	1.68	0.91	0.00	3.00	−.24*	−.19*	−.16*	−.09*	−.01	−.11*	.02	.30*	−.11*	—
Internalized stigma	743	1.62	1.04	1.00	6.00	−.21*	.06	.03	.05	.22*	.04	.13*	−.06	.04	−.17*	—
General life stressors	748	2.01	1.85	0.00	11.00	.12*	.42*	.39*	.33*	.22*	.35*	.17*	−.07	.11*	.02	−.02

Note: Min. = minimum; Max. = maximum.

^a1 = bisexual (i.e., bisexual or pansexual); 0 = monosexual (i.e., gay or lesbian).

^*p < .05.

Results from LCA documenting bisexual subgroups

For the LCA, an inspection of fit statistics for models with two to six classes found that a four-class solution was supported given that it yielded the lowest BIC and highest entropy (.86) of any model; lower AIC than models with two, five, or six classes; and adequate class size with good interpretability. As mentioned above, respondents who reported a monosexual sexual-orientation identity were artificially constrained into a single class to allow for subsequent analyses; thus, here we describe the other three latent classes comprising bisexual respondents. Bisexual Class 1 included respondents with the highest likelihood of same-gender sexual attractions and same-gender sexual partners, lowest gender conformity, and moderate-to-low sexual-identity centrality; Bisexual Class 2 included respondents with the highest likelihood of both-gender sexual attractions and both-gender sexual partners, moderate gender conformity, and highest sexual-identity centrality; and Bisexual Class 3 included respondents with the highest likelihood of other-gender sexual attractions and other-gender sexual partners, highest gender conformity, and lowest sexual-identity centrality. Figure 1 depicts predicted class probabilities for the 564 bisexual respondents included in the LCA, showing that 30 (5.3%) were in Bisexual Class 1, 133 (23.6%) were in Bisexual Class 2, and 401 (71.1%) were in Bisexual Class 3.

Fig. 1.

Predicted class probabilities for uncovered bisexual latent classes. Bisexual Class 1 includes respondents with the highest likelihood of same-gender sexual attractions and same-gender sexual partners, lowest gender conformity, and moderate-to-low sexual-identity centrality; Bisexual Class 2 includes respondents with the highest likelihood of both-gender sexual attractions and both-gender sexual partners, moderate gender conformity, and highest sexual-identity centrality; and Bisexual Class 3 includes respondents with the highest likelihood of other-gender sexual attractions and other-gender sexual partners, highest gender conformity, and lowest sexual-identity centrality.

Sociodemographic characteristics across sexual-orientation groups

Table 2 depicts sociodemographic characteristics across monosexual and bisexual respondents stratified by latent class. Monosexual respondents were older, and a higher percentage were immigrants, men, and university-educated compared with all bisexual latent classes. Bisexual Class 2 respondents were the youngest and reported the lowest educational attainment (66.2% high school or less). Bisexual Class 2 also included the highest percentage of women (90.2%), whereas Bisexual Class 1 included the highest percentage of men (43.3%). Bisexual Class 3 was primarily composed of women (82.3%).

Table 2.

Sample Characteristics (N = 748)

	Monosexual (i.e., gay or lesbian; n = 184)		Bisexual Class 1 (predominantly same-gender-attracted; n = 30)		Bisexual Class 2 (both-gender-attracted; n = 133)		Bisexual Class 3 (predominantly other-gender-attracted; n = 401)
Variable	n	%	n	%	n	%	n	%
Nativity
Immigrant	24	13.0	3	10.0	12	9.0	29	7.2
Swedish-born	160	87.0	27	90.0	121	91.0	372	92.8
Education
High school or less	94	51.1	18	60.0	88	66.2	233	58.1
Some college or technical school	31	16.9	4	13.3	21	15.8	64	16.0
University or more	59	32.1	8	26.7	24	18.1	104	25.9
Gender
Man	94	51.1	13	43.3	11	8.3	69	17.2
Woman	82	44.6	16	53.3	120	90.2	330	82.3
Transgender man	6	3.3	0	0.0	1	0.8	2	0.5
Transgender woman	2	1.1	1	3.3	1	0.8	0	0.0
Sexual orientation
Gay or lesbian	184	100.00	0	0.00	0	0.00	0	0.00
Bisexual	0	0.00	27	90.0	129	97.0	389	97.0
Pansexual	0	0.00	3	10.0	4	3.0	12	3.0

Note: Monosexual and Bisexual Class 1–3 respondents were on average 26.1 (SD = 4.9), 25.2 (SD = 5.1), 23.9 (SD = 5.0), and 25.2 (SD = 5.0) years old, respectively.

Differences in internalizing mental-health outcomes and psychosocial stressors across sexual-orientation groups

Table 3 depicts mean differences in internalizing mental-health problems, external minority stressors, internal minority stressors, and general life stressors across respondents reporting a monosexual sexual-orientation identity and the three bisexual latent classes. Monosexual respondents reported lower depression, anxiety, and suicidality symptoms compared with Bisexual Class 2 and Bisexual Class 3, but not Bisexual Class 1, respondents. In fact, Bisexual Class 1 respondents reported significantly lower anxiety symptoms than did monosexual respondents.

Table 3.

Mean Differences Between Monosexuals and Latent Classes of Bisexual Young Adults (N = 748)

Variable	Monosexual (i.e., gay or lesbian; n = 184)	Bisexual Class 1 (predominantly same-gender-attracted; n = 30)	Bisexual Class 2 (both-gender-attracted; n = 133)	Bisexual Class 3 (predominantly other-gender-attracted; n = 401)
Depression	17.61 (0.83)a	17.06 (2.30)ac	25.91 (1.57)b	21.53 (21.62)c
Anxiety	7.07 (0.37)a	5.25 (0.84)b	10.31 (0.67)c	8.84 (0.30)c
Suicidality	3.89 (0.58)a	5.06 (2.07)ab	8.85 (1.47)b	6.32 (0.58)b
Family rejection	0.16 (0.01)a	0.18 (0.04)a	0.20 (0.03)a	0.08 (0.01)b
Discrimination	1.49 (0.04)a	1.39 (0.10)a	1.76 (0.07)b	1.54 (0.03)a
Sexual-orientation victimization	0.12 (0.02)ab	0.06 (0.04)ac	0.17 (0.03)b	0.03 (0.01)c
LGBTQ-community connectedness	2.30 (0.06)ab	2.05 (0.17)a	2.48 (0.10)b	1.76 (0.04)c
Sexual-minority rejection sensitivity	9.19 (0.50)a	7.41 (1.17)a	11.72 (0.89)b	8.18 (0.35)a
Sexual-orientation outness	2.07 (0.06)a	1.75 (0.19)abc	1.74 (0.10)b	1.48 (0.05)c
Internalized stigma	2.00 (0.10)a	2.22 (0.31)a	1.39 (0.12)b	1.45 (0.05)b
General life stressors	1.62 (0.11)a	1.75 (0.38)ab	2.26 (0.26)b	2.15 (0.10)b

Note: Mean differences were estimated using the Bolck-Croon-Hagenaars method implemented in Mplus (Version 8.11; Muthén & Muthén, 2017). Standard errors are presented in parentheses. Within each row, values with different subscripts are significantly different, p < .05.

Regarding external minority stressors, Bisexual Class 3 respondents reported lower family rejection and lower LGBTQ-community connectedness compared with all other groups and lower sexual-orientation victimization compared with monosexual and Bisexual Class 2 respondents. Bisexual Class 2 respondents reported higher discrimination compared with all other groups.

Regarding internal minority stressors, Bisexual Class 2 respondents reported higher sexual-minority rejection sensitivity than all other groups. Monosexual respondents reported lower concealment (i.e., higher outness) compared with Bisexual Class 2 and Bisexual Class 3 respondents. Monosexual and Bisexual Class 1 respondents reported higher internalized stigma than Bisexual Class 2 and Bisexual Class 3 respondents.

Regarding general life stressors, monosexual respondents reported lower general life stressors compared with Bisexual Class 2 and Bisexual Class 3, but not Bisexual Class 1, respondents.

Role of psychosocial stressors in explaining monosexual–bisexual disparities in internalizing mental-health outcomes

In the hierarchical regression models depicted in Supplemental Table 1 in the Supplemental Material available online, we first investigated the role of psychosocial stressors on the observed differences in internalizing mental-health outcomes between respondents reporting a monosexual sexual orientation and respondents reporting a bisexual sexual orientation overall before conducting more detailed analyses involving the subgroups of bisexual individuals uncovered in the LCA, as described next. The models demonstrate the incremental impact of sequentially introducing various psychosocial explanatory factors (external minority stressors, internal minority stressors, general life stressors) on monosexual–bisexual disparities in depression, anxiety, and suicidality, respectively. Across internalizing mental-health outcomes, bisexual individuals consistently reported statistically significantly higher symptom levels compared with monosexual individuals, even after accounting for covariates (Model 1), external minority stressors (Model 2), and internal minority stressors (Model 3). Of note, the monosexual–bisexual disparity in anxiety symptoms was marginally significant in Model 3 (p = .052). However, after including general life stressors (Model 4), the previously observed statistically significant disparity in internalizing mental-health problems between monosexual and bisexual respondents ceased to hold across the three outcomes of depression, anxiety, and suicidality. These findings indicate that general life stressors explained the observed monosexual–bisexual disparity in internalizing mental-health problems, over and above the role of internal and external minority stressors.

Table 4 arrays results from hierarchical linear regression models with a four-level independent variable comparing monosexual respondents to the three latent classes of bisexual respondents. As depicted across all three mental-health outcomes, compared with monosexual respondents, Bisexual Class 2 and Bisexual Class 3—but not Bisexual Class 1—respondents evidenced significantly higher mental-health symptoms when controlling for sociodemographic covariates (Model 1). In fact, regarding anxiety symptoms, Bisexual Class 1 respondents evidenced significantly lower anxiety symptoms than monosexual respondents.

Table 4.

Hierarchical Regression Models Documenting Associations Among Sexual-Orientation (Bisexual Latent Classes vs. Monosexual), Mental Health, and Psychosocial Stressors

	Model 1 (covariates only—age, education, nativity, gender)			Model 2 (+ external minority stressors—family rejection, discrimination, sexual-orientation victimization, LGBT Q-community connectedness)			Model 3 (+ internal minority stressors—sexual-minority rejection sensitivity, sexual-orientation outness, internalized stigma)			Model 4 (+ general life stressors)
Variable	Est.	SE	P	Est.	SE	P	Est.	SE	P	Est.	SE	P
Depressive symptoms (CES-D)
Bisexual Class 1	−0.92	2.37	.696	−0.49	2.22	.824	−0.59	2.25	.792	−1.01	2.11	.633
Bisexual Class 2	6.62	2.01	.001	4.31	1.94	.026	4.48	1.93	.021	4.01	1.79	.025
Bisexual Class 3	3.17	1.13	.005	2.57	1.16	.027	2.35	1.19	.049	1.08	1.14	.345
Monosexual	ref	ref	ref	ref	ref	ref	ref	ref	ref	ref	ref	ref
Anxiety symptoms (GAD-7)
Bisexual Class 1	−1.97	0.89	.027	−1.68	0.84	.045	−1.65	0.83	.048	−1.83	0.76	.017
Bisexual Class 2	2.47	0.85	.004	1.72	0.83	.038	1.79	0.83	.031	1.59	0.80	.045
Bisexual Class 3	1.37	0.50	.006	1.31	0.51	.010	1.31	0.53	.013	0.77	0.51	.131
Monosexual	ref	ref	ref	ref	ref	ref	ref	ref	ref	ref	ref	ref
Suicidality (SIDAS)
Bisexual Class 1	1.01	2.06	.627	1.41	1.94	.469	1.63	1.98	.411	1.34	1.90	.482
Bisexual Class 2	4.73	1.79	.008	3.48	1.81	.054	4.00	1.83	.029	3.68	1.70	.030
Bisexual Class 3	2.51	0.92	.007	1.75	1.01	.084	2.17	1.07	.043	1.29	1.04	.215
Monosexual	ref	ref	ref	ref	ref	ref	ref	ref	ref	ref	ref	ref

Note: Bisexual Class 1 includes respondents with the highest likelihood of same-gender sexual attractions and same-gender sexual partners, lowest gender conformity, and moderate-to-low sexual-identity centrality; Bisexual Class 2 includes respondents with the highest likelihood of both-gender sexual attractions and both-gender sexual partners, moderate gender conformity, and highest sexual-identity centrality; and Bisexual Class 3 includes respondents with the highest likelihood of other-gender sexual attractions and other-gender sexual partners, highest gender conformity, and lowest sexual-identity centrality. Models were estimated using the Bolck-Croon-Hagenaars method implemented in Mplus (Version 8.11; Muthén & Muthén, 2017). Unstandardized coefficients are presented. Statistical significance at p < .05 is denoted in bold. Est. = estimate; CES-D = Center for Epidemiological Studies Depression Scale; GAD-7 = General Anxiety Disorder–7; SIDAS = Suicidal Ideation Attributes Scale; ref = reference.

For the outcome of depression, the introduction of external minority stressors (Model 2) attenuated the magnitude of the association between monosexual and Bisexual Class 2 respondents, but the association remained statistically significant. The addition of internal minority stressors (Model 3) had little effect on the association. Introducing general life stressors (Model 4) slightly reduced the effect size, but the disparity in depressive symptoms between monosexual and Bisexual Class 2 respondents remained statistically significant. For the disparity in depression between monosexual and Bisexual Class 3 respondents, introducing external minority stressors (Model 2) and internal minority stressors (Model 3) only slightly attenuated the magnitude and significance of the association. Adding general life stressors (Model 4) fully attenuated the disparity in depression between monosexual and Bisexual Class 3 respondents.

A similar pattern of results was evidenced for the outcome of anxiety. There was a significant disparity in anxiety symptoms between monosexual and Bisexual Class 2 and Bisexual Class 3 respondents, respectively, across the first three sets of models (adding sociodemographic covariates, then external minority stressors, then internal minority stressors). When general life stressors were added to Model 4, the anxiety disparity between monosexual and Bisexual Class 3 respondents was fully attenuated; the anxiety disparity between monosexual and Bisexual Class 2 respondents was partially attenuated but remained statistically significant. Across all sets of models, Bisexual Class 1 respondents had statistically significantly lower anxiety symptoms than monosexual respondents.

For the outcome of suicidality, Bisexual Class 2 and Bisexual Class 3 respondents reported significantly higher suicidality in Model 1 (when adjusting for sociodemographic covariates) compared with monosexual respondents. When external minority stressors were added (Model 2), the suicidality disparities were attenuated but reappeared when internal minority stressors were added (Model 3), perhaps reflecting a suppression effect (Beckstead, 2012). In Model 4, when adjusting for covariates, external minority stressors, internal minority stressors, and general life stressors, the disparity in suicidality remained between monosexual and Bisexual Class 2 respondents, whereas it was attenuated among Bisexual Class 3 respondents.

Supplemental Table 2 in the Supplemental Material depicts results from the four-step hierarchical regression models in which the outcome was specified as a latent construct representing overall internalizing mental-health problems derived from the observed measures of depression, anxiety, and suicidality. The scale of the latent outcome variable was set to be equivalent to the scale of the depression measure (i.e., Center for Epidemiologic Studies Depression Scale). These supplemental analyses confirm that Bisexual Class 1 respondents did not demonstrate higher internalizing mental-health symptoms than respondents reporting a monosexual sexual-orientation identity. For Bisexual Class 2 respondents compared with monosexual respondents, a statistically significant elevation in internalizing mental-health problems persisted across all four models. For Bisexual Class 3 respondents, the disparity in internalizing mental-health problems between respondents in this class and monosexuals was statistically significant when adjusting for covariates, external stressors, and internal stressors but was fully attenuated when general life stressors were added to the final model.

Discussion

In this population-based study of sexual-minority young adults, we sought to elucidate the monosexual–bisexual disparity in internalizing mental-health problems. Our findings align with previous studies (Bostwick et al., 2010; Jorm et al., 2002), revealing that, on average, young adults reporting a bisexual sexual-orientation identity experience higher levels of depression, anxiety, and suicidality compared with their monosexual peers. However, by measuring the diversity of the bisexual experience (i.e., in terms of gender-based sexual attractions, gender of sexual partners, one’s own gender conformity, and one’s own sexual-identity centrality), our analyses add nuance to this overarching pattern. Notably, a small percentage of bisexual individuals, particularly those resembling monosexual individuals in terms of same-gender attractions, did not exhibit significantly worse mental-health outcomes than their monosexual counterparts and, in fact, had lower anxiety symptoms. Moreover, for the large percentage of bisexual individuals with predominantly other-gender attractions (approximately 71%), their disproportionate exposure to general life stressors (e.g., job loss, financial loss, relationship problems) fully explains their higher levels of internalizing mental-health problems. For bisexual individuals reporting both-gender attractions, their higher level of depression, anxiety, and suicidality compared with their monosexual peers remains unexplained. Together, these findings suggest that, although bisexual individuals are at significantly greater risk of internalizing mental-health problems compared with other sexual-orientation groups, this risk is not borne equally across the bisexual population and is instead distributed across important dimensions of the bisexual experience. These findings also suggest that general life stressors beyond sexual identity-related social stress might underlie a large proportion of bisexual individuals’ greater risk compared with monosexual people, with important implications for advancing theory, research, and intervention.

This study represents the first to investigate dimensions of bisexual experience in a population-based sample recruited by random selection into a national health survey, therefore effectively eliminating potential selection biases inherent to prior investigations and permitting an examination of the full diversity of the bisexual experience. Our findings reveal a spectrum of experiences within the bisexual community across dimensions of gender-based sexual attraction, sexual-partner gender, gender conformity, and sexual-identity centrality, with significant implications for mental health. Notably, more than two thirds of the bisexual individuals in our study belonged to a latent subgroup characterized by a higher likelihood of other-gender sexual attractions and sexual partners, high gender conformity, and low sexual-identity centrality. These individuals may have been underrepresented in previous nonprobability research focusing on LGBTQ populations. Further, results from the current study show that the small proportion of bisexual individuals whose experiences aligned most closely with patterns of monosexual behaviors and experiences did not exhibit poorer mental health compared with their gay and lesbian counterparts, although the small sample size of this group (n = 30) limited power to detect significant differences. Together, these findings suggest that prior research that treats all bisexual individuals as a homogeneous group has overlooked important subgroup experiences that predict internalizing mental-health problems.

This study highlights the significant contribution of general life stressors to the monosexual–bisexual mental-health disparity, in some cases over and above minority stressors. Indeed, reporting a bisexual (vs. monosexual) identity was positively correlated with greater exposure to general life stressors, and general life stressors attenuated the monosexual–bisexual disparity in depression, anxiety, and suicidality, more so than did internal or external minority stressors, for the largest group of bisexual respondents. Bisexual people’s elevated exposure to general life stressors compared with their monosexual peers aligns with previous research showing that bisexual people have lower educational attainment and income compared with monosexual people (Fredriksen-Goldsen et al., 2024; Mittleman, 2022). The underlying reasons for this economic and educational disparity are not fully understood. However, it has been argued that the disparity may stem from bisexual people’s vulnerability to discrimination at the intersections of biphobia and misogyny, particularly given that a majority of bisexual individuals are women, a pattern consistent with findings in the current study. These observed socioeconomic inequities can then increase the likelihood of experiencing financial precarity, leaving bisexual individuals more vulnerable to general life stressors such as job loss, financial strain, and relationship stressors. In the current study, bisexual respondents reported lower educational attainment than their gay and lesbian peers, corroborating this hypothesis.

These findings also may align with prior research indicating that bisexual individuals encounter more stressful life events over their life span compared with gay and lesbian individuals, including early adverse experiences (McLaughlin et al., 2012; Persson & Pfaus, 2015). Recent longitudinal evidence from a birth cohort indicates that early childhood adversity correlates with increased stressful life events into adulthood (Brennan et al., 2024), both of which are strongly correlated with internalizing mental-health problems (Phillips et al., 2015; Raposa et al., 2014). Thus, although the current study did not directly measure early life stressors, it may be one factor in explaining the observed greater exposure to general life stressors and poorer mental health among bisexual young adults compared with their monosexual peers in young adulthood. Future research measuring dimensions of bisexual identities and experiences and general life stressors across the life course represents an important future research direction capable of providing a more comprehensive understanding of directional paths toward internalizing mental-health problems.

This study emphasizes the need for mental-health support tailored to the observed mental-health disparities affecting bisexual young adults. Although the field of sexual-minority mental health has made strides with evidence-based interventions such as LGBTQ-affirming cognitive behavioral therapy (Pachankis et al., 2022), designed to alleviate the higher burden of internalizing mental-health problems in the sexual-minority population, the current study suggests that the psychosocial contributors for bisexual young adults may not align fully with traditional minority stress pathways commonly targeted by such interventions, such as internalized stigma. Consequently, bisexual individuals may benefit from more general trauma-informed mental-health support (for a review, see Oral et al., 2016) to cope with exposure to stressful life events, irrespective of their direct relevance to sexual orientation.

Results from the current study further highlight that even when accounting for psychosocial stressors, including minority stressors and general life stressors, bisexual individuals with a high likelihood of both-gender attractions still experience substantially higher depression, anxiety, and suicidality than their gay and lesbian peers. At the same time, this group reported the highest sexual-identity centrality and significantly higher LGBTQ-community connectedness than their bisexual peers who reported same-gender or predominantly other-gender attractions. These findings suggest that interventions that leverage both-gender-attracted bisexual young adults’ existing identity centrality and LGBTQ-community connectedness may hold particular promise for reducing negative mental-health outcomes in this population. Indeed, research indicates that loneliness may play an important role in the elevated risk for suicidality among bisexual individuals (Mereish et al., 2017), suggesting that interventions focusing on building community and support could be particularly beneficial. The relatively strong identity centrality and LGBTQ-community connectedness experienced by both-gender-attracted bisexual individuals also implies that group-psychotherapy modalities may be especially effective given the group’s existing connections to the community and their shared identity-relevant struggles and strengths. Interventions can perhaps further strengthen these assets among this population in a way that ensures healthy identity development and community connection. A recent pilot group-therapy intervention intended to build healthy community and coping skills among Black and Latino gay and bisexual men revealed that the intervention helped participants feel less alone in their shared minoritized identities and was associated with reductions in internalizing mental-health outcomes (Jackson et al., 2022). Findings from the current study suggest that developing and testing a similar group-based intervention targeting both-gender-attracted bisexual individuals may warrant consideration.

Findings from this research should be interpreted in light of six limitations. First, the study’s cross-sectional design precluded the establishment of causality among the variables measured. Future research using longitudinal methods is essential to elucidate the interplay between bisexual identity measured at one time point, psychosocial stressors assessed at a subsequent time point, and mental-health outcomes at a still later time point. Second, as mentioned above, we did not measure exposure to general life stressors outside of the previous year, including exposure to earlier life stressors, a notable limitation given the importance of understanding stressful experiences across the life course in shaping internalizing psychopathology in young adulthood. Future research, especially with bisexual populations, should seek to assess exposure to general life stressors across the life course to identify relevant timings of exposures to stressful experiences in relation to monosexual–bisexual disparities in internalizing psychopathology. Third, participants in this study sample were recruited from Sweden, a country with high levels of sexual-minority social acceptance, and findings may not generalize to more stigmatizing geographic locations, warranting future research. Fourth, although we benefited from population-based sampling, mitigating selection bias based on sexual orientation, we faced limitations in recruiting a substantial number of transgender individuals because of their lower population prevalence. Moreover, to enhance classification of gender of sexual attractions and sexual partners in the LCA and to mitigate problematic assumptions, we did not include individuals with nonbinary/genderqueer identities, sexual attractions, and sexual partners. Whether any of the associations reported here may differ across the full spectrum of gender identities is unknown, representing an important avenue for future research with sufficient power. Fifth, there was insufficient sample size to stratify by gender, and we were therefore unable to capture different patterns of psychosocial stress exposure and internalizing mental-health problems between men and women, another important avenue for future research with a larger sample size. Last, this study did not assess bisexual-specific minority stressors—such as binegativity and identity invalidation (Feinstein & Dyar, 2017; Katz-Wise et al., 2017)—that may play an important role in the mental health of bisexual individuals. This limitation underscores the importance of measuring minority stressors relevant to the experiences of bisexual people in future population-based research.

Conclusion

Taking advantage of a rare combination of methodological strengths (e.g., population-based sample of sexual-minority young adults, comprehensive measurement of psychosocial stressors), this study advances understanding of the sizeable monosexual–bisexual disparity in internalizing mental-health problems. By delineating three distinct subgroups of the bisexual experience on the basis of gender-based sexual attraction, gender of sexual partners, one’s own gender conformity, and one’s own identity centrality, this study both illuminates the considerable diversity within the bisexual young adult population and provides valuable insights into heterogeneous psychosocial and mental-health risks patterned by that diversity. Notably, the association of external and internal minority stressors with the observed monosexual–bisexual mental-health disparities was relatively small. Thus, results of this study suggest some limitations in the applicability of psychosocial stressors outlined in the minority stress model (Brooks, 1981; Meyer, 2003) when applied to understanding mental-health disparities between monosexual and bisexual individuals. Instead, our findings suggest the need for future theory-driven research to explore the role of general life stressors in shaping the mental health of bisexual individuals compared with their monosexual peers during sensitive developmental periods, including young adulthood. Future research should aim to elucidate how targeted interventions can effectively reduce internalizing mental-health problems that disproportionately affect bisexual young adults, considering both sexual-orientation-specific and general pathways.