Bisexuality, romantic relationships, and depression: The role of relationship quality

Abstract

This study provides a population-level assessment of relationship quality among partnered bisexual adults and evaluates whether high-quality romantic relationships are linked to fewer depressive symptoms. Using data from the National Longitudinal Study of Adolescent to Adult Health (1998–2008; n = 15,701), we estimated linear regression models to compare relationship quality across sexual identity groups and to examine associations between relationship quality and mental health. Results indicate that bisexual adults report significantly lower relationship quality than heterosexual adults but not significantly lower than lesbian or gay adults, after controlling for relationship type and other covariates. Although high-quality romantic relationships are linked to improved mental health for bisexual individuals, this association is weaker compared to heterosexual individuals. Our findings suggest that bisexual adults experience unique relationship-specific and broader societal stressors that challenge the promotion of mental health through romantic involvement.

1. Introduction

Bisexual adults experience elevated risks for depression in comparison to their heterosexual and gay/lesbian counterparts (Bostwick, 2012; Bostwick et al., 2014; Meyer, 2003). Distinct stigma and discrimination, referred to as binegativity, from both heterosexual and lesbian/gay communities contribute significantly to these mental health disparities (Armstrong and Reissing, 2014; Cox et al., 2013; Dodge et al., 2016; Eliason, 2001; Friedman et al., 2014; Zivony and Lobel, 2014). In romantic relationships, bisexual individuals are often subjected to negative stereotypes, such as being hypersexual, confused, or unfaithful (Dodge et al., 2016; Friedman et al., 2014; Zivony and Lobel, 2014). Bisexual people may also internalize these bi-negative stereotypes and stigma which can, and do, contribute to their mental health burden, leading to higher rates of depression (Bostwick, 2012; Hall, 2018; Pollitt and Roberts, 2021; Velasco et al., 2024).

For most adults, intimate relationships are a prominent protective mechanism against mental health problems (Braithwaite and Holt-Lunstad, 2017; Umberson et al., 2010). Indeed, intimate partners are the primary sources of social support and emotional care, which can alleviate daily stressors and mental health problems (Braithwaite and Holt-Lunstad, 2017; Donnelly et al., 2019; Hsieh, 2014; Hsieh and Liu, 2019; Sarno et al., 2022; Umberson et al., 2010, 2013). Yet, prior research on bisexual people has found no health benefits for those in romantic relationships, suggesting that the protective processes of relationships may differ for bisexual individuals compared to their heterosexual or lesbian/gay counterparts (Feinstein et al., 2016; Feinstein and Dyar, 2017; Hsieh and Liu, 2019; Hsu and Mernitz, 2024).

Scholars have speculated that bisexual people may have lower relationship quality compared to heterosexual, lesbian, and gay populations (Hsieh and Liu, 2019; Smith et al., 2022), which could diminish the potential benefits of having a romantic partner. The mental health benefits of romantic relationships are found in both heterosexual and gay/lesbian populations, particularly when these unions are high-quality (Donnelly et al., 2019; Du Bois et al., 2019; Mernitz et al., 2020). Distressed relationships, by contrast, increase the risk of depression-related issues (Kiecolt-Glaser and Wilson, 2017). To the best of our knowledge, no population-level study has examined relationship quality and depressive symptoms among bisexual adults. We use Wave 1 and 4 of the National Longitudinal Study of Adolescent to Adult Health (Add Health) to determine whether bisexual adults report lower relationship quality than their heterosexual, lesbian, or gay counterparts. We also examine how their relationship quality is linked to depressive symptoms. First, we examine differences in romantic relationship quality between partnered heterosexual, gay/lesbian, and bisexual individuals. Then, we explore whether high-quality relationships are associated with fewer depressive symptoms in bisexual individuals compared to those of other sexual orientations.

1.1. Romantic relationships and mental health among bisexual people

Bisexual people are the most disadvantaged in mental health, especially for depression, compared to all other sexual orientations (Bostwick, 2012; Bostwick et al., 2010; Feinstein and Dyar, 2017; Hsieh, 2014; Institute of Medicine, 2011). When considered separately from other sexual minority populations, scholars found that bisexual people face a ‘double jeopardy’ of negative stereotypes, such as being sexually confused, promiscuous, and unfaithful in romantic relationships, from both heterosexual and lesbian/gay communities (Dodge et al., 2016; Friedman et al., 2014; McLean, 2007; Smith et al., 2022; Zivony and Lobel, 2014). Consequently, discrimination against bisexual individuals by both heterosexual and lesbian/gay communities, can increase their risk of mental health problems (Velasco et al., 2024). Given the lack of social acceptance and community resources, studies have also found that bisexual people are less likely to disclose their sexual identity to friends and family than lesbian/gay people (Barringer et al., 2017; McLean, 2007; Scherrer et al., 2015), and possess fewer social resources—such as emotional and financial support from significant others—than their lesbian/gay counterparts, which are critical to mental health (Hsieh, 2014).

Romantic relationships often protect against poor mental health for young adults (Braithwaite and Holt-Lunstad, 2017; Feinstein et al., 2016; Still, 2021; Whitton et al., 2013). These committed relationships confer health benefits by providing emotional support, encouraging healthy habits, regulating problematic behaviors, and sharing economic resources (Carr and Springer, 2010; Donnelly et al., 2019; Hsieh and Liu, 2019; Waite and Gallagher, 2001). Yet, most prior studies have focused on heterosexual relationships (Braithwaite and Holt-Lunstad, 2017; Carr and Springer, 2010; Liu and Umberson, 2008; Umberson et al., 2013; Whitton et al., 2013) and only more recently have scholars begun to explore whether and how these mental health benefits extend to sexual minority men and women in same-sex relationships (Feinstein et al., 2016; Mernitz et al., 2020; Newcomb, 2020; Reczek, 2020; Reczek et al., 2017; Whitton et al., 2018).

To date, research on these topics has not extensively examined bisexual populations. Studies that distinguish between bisexual people and other sexual orientations often find negligible health benefits from marriage (Hsieh and Liu, 2019) as well as other types of romantic relationships (Hsu and Mernitz, 2024). For instance, in a national probability sample, Hsieh and Liu (2019) failed to find an association between marital status and better physical health. Instead, they found that married bisexual people report worse health outcomes than unmarried ones after controlling socioeconomic resources and health behaviors. They also noted that only same-gender relationships confer health benefits for bisexual people, but most bisexual people are in different-gender relationships (Hsieh and Liu, 2019). For young adult populations in particular, being involved in a romantic relationship is detrimental for bisexual young adults’ mental health (Feinstein et al., 2016; Whitton et al., 2018). In a community sample of young adults, Whitton et al. (2018) found that bisexual people experienced greater psychological distress when they were in a romantic relationship than when they were single or unpartnered. Evidence so far suggests that bisexual adults in relationships may be at a greater risk for mental health problems than those who are single or unpartnered.

1.2. The role of relationship quality

Romantic relationships alone may not necessarily protect mental health; rather, relationship quality plays a more significant role in influencing health outcomes (Dush and Amato, 2005; Kiecolt-Glaser and Wilson, 2017; Proulx et al., 2007; Still, 2021; Thomeer et al., 2013; Umberson et al., 2006; Williams, 2003). High-quality relationships confer the most mental health benefits compared to low-quality relationships, a finding that holds true for those in different-sex and same-sex relationships (Dush and Amato, 2005; Mernitz et al., 2020; Still, 2021). Specifically, higher-quality relationships are negatively associated with depressive symptoms for married same-sex couples (Mernitz et al., 2020). By contrast, low-quality relationships can be more detrimental to health than relationship dissolutions among married heterosexual individuals (Williams, 2003), and may accelerate health declines with age (Umberson et al., 2006).

Thus, scholars have speculated that relationship quality plays a critical role in shaping the health benefits of romantic relationships for bisexual individuals (Hsieh and Liu, 2019; Whitton et al., 2018). It is possible that bisexual individuals either face challenges in forming high-quality committed relationships or report lower relationship quality regardless of relationship type. On the one hand, negative stereotypes against bisexuality contribute to insecurity regarding commitment in relationships with bisexual partners, particularly among women (Armstrong and Reissing, 2014). Relationship quality varies by relationship type, with those in dating relationships often reporting lower relationship quality than those who are married or cohabiting (Brown and Bulanda, 2008; Giordano et al., 2012; Waite and Joyner, 2001). Thus, bisexual people may report lower relationship quality on average if they are more likely to be in dating rather than cohabiting or marital relationships.

On the other hand, stressors specific to bisexuality within romantic relationships can lower relationship quality, and acceptance from significant others is crucial to relationship quality (Fuller and Hovland, 2022) regardless of relationship type. In two studies based on community samples of bisexual men, Smith et al. (2022) found that a partner’s binegativity was associated with lower relationship satisfaction, and the lack of identity affirmation in a relationship was linked to unprotected sexual behaviors (Feinstein et al., 2021). Binegativity from a romantic partner can manifest as bisexual identity erasure (Feinstein et al., 2019; Mernitz et al., 2022), such as dismissing bisexuality as a “phase” (McLean, 2007; Scherrer et al., 2015) or requiring bisexual individuals—particularly women—to “prove” their bisexuality (Mernitz et al., 2022).

Partner binegativity may also contribute to partners feeling insecure in committed relationships and becoming hypervigilant about potential attractions to others. Chronic distrust and worries about infidelity (Israel and Mohr, 2004; Tabatabai, 2016; Zivony and Lobel, 2014), rooted in bi-specific misconceptions that frame bisexual people as incapable of satisfaction with one partner, can further strain relationship dynamics (Anderson and McCormack, 2016). These dynamics may also discourage bisexual adults from disclosing their identity to partners, leading to concealment that generates additional daily stress that is associated with worse mental health (Anderson and McCormack, 2016; Barringer et al., 2017; Feinstein et al., 2020; Pallotta-Chiarolli, 2016). While lesbian and gay individuals also experience relationship-related stressors, these challenges tend to come from external sources rather than from their romantic partners (LeBlanc and Frost, 2020). Consequently, bisexual individuals may be at a higher risk of experiencing low-quality relationships, regardless of relationship type, as they face greater challenges once in a romantic partnership.

Thus far, prior studies demonstrate that same-sex and different-sex committed relationships are often similar in quality (Joyner et al., 2019), and that relationship quality plays a crucial role in mitigating depression-related issues among gender and sexual minority populations (Mernitz et al., 2020; Sarno et al., 2022). However, existing studies are limited in that they do not provide population-level evidence on relationship quality and depression-related issues among bisexual individuals or how these experiences compare to those of heterosexual, lesbian, and gay individuals. Barriers to committed relationships and bi-specific stressors from partners, family, and other social relationships, such as binegativity, identity concealment, partner invalidation (Anderson and McCormack, 2016; Barringer et al., 2017; Feinstein et al., 2020; Mernitz et al., 2022; Pallotta-Chiarolli, 2016; Scherrer et al., 2015; Smith et al., 2022), and/or higher degree of internalized bi-negativity (Frost and Meyer, 2009; Pollitt and Roberts, 2021; Velasco et al., 2024) can dampen any benefits bisexual adults derive from intimate partnerships.

Yet, given that bisexual individuals experience heightened risks of depression (Bostwick, 2012), the psychosocial resources embedded within high-quality relationships may provide more mental health benefits than they would for other groups. For example, Hsieh (2014) found that the health benefits of emotional support and friendships vary across sexual orientations, with bisexual individuals benefiting more from having sufficient confidantes than heterosexual individuals. It is therefore plausible that being in a high-quality relationship, characterized by emotional support and acceptance from one’s partner, may confer greater mental health benefits for bisexual individuals, given their elevated levels of psychological distress, compared to other groups.

1.3. The current study

Building on current evidence, this study has two main objectives: (1) to examine whether bisexual adults report different levels of relationship quality compared to other sexual orientation groups and to determine whether this difference is driven by their concentration in less committed relationships (i.e., dating rather than marriage or cohabitation), and (2) to assess how relationship quality is linked to depressive symptoms, as well as how this association may vary across sexual orientation groups. We also address whether bisexual adults benefit more or less from high-quality relationships. We propose the following hypotheses:

• Hypothesis 1. Bisexual adults will report significantly lower relationship quality compared to heterosexual and lesbian/gay adults.

• Hypothesis 2. The lower relationship quality for bisexual adults is partially explained by relationship types.

• Hypothesis 3. Relationship quality is associated with fewer depressive symptoms among all sexual orientation groups.

• Hypothesis 4 (exploratory). The association between relationship quality and depressive symptoms varies across sexual orientation groups.

2. Methods

2.1. Data

This study draws from Waves 1 and 4 of the National Longitudinal Study of Adolescent to Adult Health (Add Health) (Harris et al., 2019). Add Health is a nationally representative study that first sampled 80 high schools and 52 middle schools in the US. Wave 1 was collected during the 1994–1995 school year (n = 20,745) and Wave 4 was collected in 2008 when participants were between the ages 24 to 34 (n = 15,701). We choose relationship data from Wave 4 because the romantic relationship quality scale was measured using 7 items for a current relationship. Out of the initial sample of 15,701 adolescents who completed Wave 4, we include all individuals who report a current relationship (regardless of relationship type; n = 11,478). We further restrict our sample to those without missing data (3.6 % of sample has missing data, thus, we use listwise deletion), leading to an analytic sample of 11,062 people.

2.2. Measures

Romantic Relationship Quality.

At Wave 4, respondents who are in a current relationship are asked to describe their relationships in detail. Seven questions on a 5-point Likert scale from strongly agree to strongly disagree are used for this study. The questions include: “We (enjoy/enjoyed) doing even ordinary day-to-day things together,” “I (am/was) satisfied with the way we handle our problems and disagreements,” “I (am/was) satisfied with the way we handle family finances,” “My partner (listens/listened) to me when I need someone to talk to,” “My partner (expresses/expressed) love and affection to me,” “I (am/was) satisfied with our sex life,” “I (trust/trusted) my partner to be faithful to me.” Respondents with missing value(s) on any of the items are excluded from the analytic sample. The relationship quality score ranges from 7 to 35 with higher values indicating better quality (Cronbach’s α = 0.89).

Depressive Symptoms.

At Wave 4, depressive symptoms are self-reported by the respondent. We measure depressive symptoms from the abridged 10-item assessment from the Center for Epidemiologic Studies Depression Scale (CESD). Respondents answer how often they had experienced negative and positive emotions in the past 7 days (e.g., “I felt depressed”). The answers indicate the frequency of emotions and were coded from 0 (never or rarely) to 3 (most of the time and all the time). Answers to positive affect questions (e.g., “I enjoyed life”) were reverse coded. The sum of values from 10 items is the indicator of symptoms, where higher scores indicate more depressive symptoms (Cronbach’s α = 0.83).

Sexual Identity.

Sexual identity is measured by the question at Wave 4: “Please choose the description that best fits how you think about yourself.” The answers include 100 % heterosexual (straight), mostly heterosexual (straight) but sometimes attracted to people of your own sex, bisexual, mostly homosexual but sometimes attracted to people of the opposite sex, 100 % homosexual (gay), not sexually attracted to either males or females, refused, and don’t know. Those who reported 100 % heterosexual or homosexual are coded as heterosexual or lesbian- and gay-identified respectively (n = 9,631 for heterosexual, n = 119 for lesbian/gay). Answers indicate asexuality or questioning (“don’t know” or “refused”) are considered missing.

To ensure our results are robust, we present both the main and sensitivity findings. First, in this study, we collapse the mostly heterosexual (n = 1,082), bisexual (n = 157), and mostly homosexual (n = 73) identity categories into a bisexual sample (n = 1,312). In the results presented below, we include those who respond “mostly heterosexual” and “mostly homosexual” in our bisexual sample given that other studies have shown that both are distinct from “100 % heterosexual” and “100 % homosexual” (Savin-Williams and Vrangalova, 2013; Vrangalova and Savin-Williams, 2012). These groups clearly identify away from mono-sexuality, and both sexual identity and relationship status are measured at Wave 4. Thus, these individuals are in current relationships and report a non-monosexual identity (Barringer et al., 2017; Feinstein et al., 2020; Vrangalova and Savin-Williams, 2012). Because the “mostly heterosexual” group is large and could disproportionately influence results, we reclassified them as heterosexual in an additional analysis, leaving 230 respondents in the bisexual sample. Then, we also conducted a sensitivity analysis considering only self-identified bisexual individuals (n = 157) in the bisexual sample. Across all specifications, our results remain substantively unchanged regardless of how bisexual adults are classified.

Relationship Type.

We identify respondents’ current relationship type using romantic relationship history data from Add Health (Harris et al., 2019). Respondents are asked “type of relationship with partner” and whether the relationship was current for each partner. Answers include marriage, cohabitation, pregnancy, current dating, and most recent. We exclude past relationships and pregnancy-only relationships and classify relationships into three categories: current marital relationships, cohabiting relationships, and dating relationships.

Covariates.

We control for the following relationship characteristics that are associated with relationship quality based on prior literature (Frost et al., 2022; Joyner et al., 2019; Still, 2021): the age difference between partners, whether respondents are in an inter-racial relationship (1 = yes), the duration of the relationship (measured in months). We also control for whether the relationship was with a same-sex partner (1 = yes), as it is also linked to health-promoting effects for sexual minority population (Hsieh and Liu, 2019; Hsu and Mernitz, 2024). Other controls include respondent’s sex assigned at birth at Wave 1, racial/ethnic category constructed at Wave 1 (Udry et al., 2003), age at Wave 4, region of residence at Wave 4, college education at Wave 4 (1 = yes), and whether or not the respondent was a parent (1 = yes). We also control for depressive symptoms at Wave 1 (10 items, α = 0.81) when modeling mental health measures given their association with both relationship history and later mental health in adulthood (Mernitz and Kamp Dush, 2014).

2.3. Analytic plan

We test first and second hypotheses that bisexual adults will report lower relationship quality than heterosexual and lesbian/gay individuals (Hypothesis 1) and that relationship type may explain this difference (Hypothesis 2) using linear regression models to independently predict romantic relationship quality by sexual identity group (bisexual, gay/lesbian, and heterosexual). We assess whether the association between relationship quality and bisexual identity changes after accounting for relationship type. We then perform pairwise comparisons across sexual identity groups to explore differences.

To test our third and fourth hypotheses, we use linear regression to predict depressive symptoms. We include sexual identity, relationship quality, and other covariates to examine whether relationship quality is independently associated with each mental health outcome (Hypothesis 3). We then test an interaction between sexual identity and relationship quality to investigate whether these associations vary across sexual orientation groups (Hypothesis 4). Interaction results are also visualized to facilitate interpretation. All analyses are weighted and conducted in R.

3. Results

3.1. Descriptive results

Table 1 presents descriptive statistics by sexual identity, along with results from bivariate regressions on key measures. Bisexual adults report significantly lower relationship quality (28.09) than heterosexual adults (29.42), but only marginally lower than lesbian/gay individuals (29.44, p < 0.1). Although the mean difference between bisexual and heterosexual adults is statistically significant (28.09 vs. 29.42), the magnitude of this difference is modest. Given that the overall average relationship quality score is 29.2 on a 7–35 scale, most respondents reported relatively high-quality relationships. Nonetheless, bisexual adults consistently scored below this average, suggesting that even small differences may reflect meaningful disparities when considered alongside their elevated depressive symptoms and other disadvantages. Bisexual adults also report the highest levels of depressive symptoms (7.6 at Wave 4), whereas heterosexual, lesbian, and gay adults report depressive symptom scores below 7. The majority of bisexual adults in the sample are women (83 %), and most (94 %) are in different-gender relationships. Being in a different-gender relationship may lessen exposure to stigma targeting same-sex couples, but it does not shield bisexual adults from stigma specific to their identity. Experiences such as identity erasure and invalidation remain common and can arise irrespective of partner gender.

Table 1.

Descriptive statistics by sexual identity.

Variables	Heterosexual Pop, N = 9,631d	Bisexual Pop, N = 1,312d	Lesbian/Gay Pop, N = 119d
Relationship Quality W4	29.42 (5.26)b	28.09 (5.70)a	29.44 (5.95)
Depressive Symptoms W4	5.58 (4.46)b	7.57 (5.04)ac	5.65 (4.47)b
Depressive Symptoms W1	6.37 (4.58)	7.90 (5.22)	6.96 (5.39)
Relationship Types
Married	56 %	46 %	13 %
Cohabitating	24 %	36 %	51 %
Dating	19 %	18 %	36 %
Age W4	28.43 (1.86)	28.03 (1.75)	28.56 (1.82)
Female	47 %	83 %	39 %
Race
White	70 %	75 %	61 %
Black	14 %	7.6 %	16 %
Latino	11 %	11 %	17 %
Asian	3.1 %	2.4 %	2.0 %
Other	1.8 %	3.6 %	4.3 %
College Degree W4	31 %	32 %	39 %
Relationship Duration (Year)	5.54 (3.91)	5.12 (3.88)	2.70 (2.65)
Age Difference	3.24 (3.37)	3.83 (3.92)	4.39 (4.28)
Interracial Relationship	17 %	24 %	31 %
Have Children	56 %	51 %	11 %
Same-Sex Relationship	0.3 %	6.1 %	95 %
Region of Residence W4
Northeast	12 %	16 %	16 %
Midwest	29 %	30 %	23 %
South	43 %	33 %	40 %
West	16 %	21 %	21 %

Data from Add Health Wave 1&4.

^asignificantly different from heterosexual sample (p < 0.05).

^bsignificantly different from bisexual sample (p < 0.05).

^csignificantly different from lesbian/gay sample (p < 0.05).

^dUnweighted N to show sample sizes.

Regarding relationship type, while 56 % of heterosexual adults are married, only 46 % of bisexual individuals and 13 % of lesbian/gay individuals are married. The low marriage rate among lesbian/gay individuals likely reflects the fact that data collection occurred before 2015, when same-sex marriage was nationally legalized.

3.2. Regression results

Table 2 summarizes the regression results on relationship quality. Bisexual adults report significantly lower relationship quality than heterosexual adults but not lower than lesbian and gay adults. Before accounting for relationship type, partnered bisexual adults report significantly lower relationship quality than partnered heterosexual adults (b = −1.49, p < 0.05, 1 in Table 2). However, there is no significant difference in relationship quality between partnered bisexual and gay/lesbian adults. Notably, these findings for partnered bisexual adults remain robust even after controlling for sociodemographic and relationship characteristics.

Table 2.

Regression results on relationship quality.

	1	2	3a	4a	5b	6b
Bisexual (Ref = Het)	−1.49***	−1.49***	−1.65*	−1.65*	−1.53*	−1.54*
	(0.25)	(0.25)	(0.68)	(0.69)	(0.63)	(0.63)
Lesbian/Gay	−0.46	−0.50	−0.49	−0.53	−1.40	−1.43
	(1.00)	(1.00)	(1.15)	(1.16)	(1.72)	(1.73)
Cohabitating (Ref = Mar.)		−0.36+		−0.41*		−0.41*
		(0.20)		(0.20)		(0.20)
Dating		−1.17***		−1.18***		−1.18***
		(0.24)		(0.25)		(0.25)
Male (Ref = F.)	−0.11	−0.10	0.09	0.09	0.09	0.10
	(0.15)	(0.15)	(0.15)	(0.15)	(0.15)	(0.15)
Age	−0.11*	−0.11*	−0.10*	−0.10*	−0.10*	−0.10*
	(0.04)	(0.04)	(0.04)	(0.04)	(0.04)	(0.04)
Black (Ref = White)	−1.75***	−1.58***	−1.68***	−1.50***	−1.67***	−1.50***
	(0.19)	(0.18)	(0.20)	(0.19)	(0.20)	(0.19)
Latino	−0.62*	−0.54*	−0.59*	−0.51*	−0.59*	−0.51*
	(0.24)	(0.24)	(0.24)	(0.24)	(0.24)	(0.24)
Asian	−0.04	−0.00	0.03	0.07	0.03	0.07
	(0.53)	(0.52)	(0.53)	(0.53)	(0.53)	(0.53)
Other	−0.35	−0.29	−0.39	−0.33	−0.40	−0.33
	(0.50)	(0.50)	(0.49)	(0.49)	(0.49)	(0.49)
College Degree	1.03***	1.00***	1.04***	1.01***	1.05***	1.02***
	(0.14)	(0.14)	(0.14)	(0.14)	(0.14)	(0.14)
Have Children	−0.82***	−0.97***	−0.80***	−0.95***	−0.80***	−0.95***
	(0.18)	(0.18)	(0.18)	(0.18)	(0.18)	(0.18)
Interracial Rel.	−0.34	−0.33	−0.38+	−0.36+	−0.38+	−0.37+
	(0.20)	(0.21)	(0.21)	(0.21)	(0.21)	(0.21)
Age Diff.	−0.03	−0.03	−0.03	−0.03	−0.03	−0.03
	(0.02)	(0.02)	(0.02)	(0.02)	(0.02)	(0.02)
Rel. Duration (Year)	0.03+	−0.02	0.04+	−0.02	0.04+	−0.02
	(0.02)	(0.02)	(0.02)	(0.02)	(0.02)	(0.02)
Same-Sex Rel.	0.34	0.43	0.52	0.63	0.89	0.99
	(0.79)	(0.79)	(0.98)	(0.98)	(1.58)	(1.59)
Midwest (Ref = NE)	0.34+	0.31	0.37+	0.34+	0.37+	0.34+
	(0.19)	(0.19)	(0.19)	(0.19)	(0.19)	(0.19)
South	0.56**	0.50**	0.61***	0.55**	0.61***	0.55**
	(0.18)	(0.18)	(0.18)	(0.18)	(0.18)	(0.18)
West	0.48*	0.45*	0.48*	0.44*	0.48*	0.44*
	(0.20)	(0.21)	(0.21)	(0.21)	(0.21)	(0.21)
N	11,062	11,062	11,062	11,062	11,062	11,062
R-squared	0.04	0.05	0.04	0.04	0.04	0.04

⁺p < 0.1,

^*p < 0.05,

^**p < 0.01,

^***p < 0.001.

Data from Add Health.

^a“Mostly heterosexual” identity is recategorized as heterosexual sample.

^b“Mostly heterosexual” identity is recategorized as heterosexual sample, and “Mostly Homosexual” identity is recategorized as Lesbian/Gay sample.

To test Hypothesis 2, we then include relationship type in the model. We find that the association between relationship quality and bisexual identity remains unchanged (b = −1.49, p < 0.05, 2 in Table 2), even though relationship type is significantly associated with relationship quality. As expected, being in a dating relationship is associated with lower relationship quality compared to marriage. This finding suggests that the lower relationship quality reported by bisexual adults cannot be explained by the types of relationships in which they are concentrated. Descriptive results reveal that similar percentages of bisexual and heterosexual adults are in dating relationships, with 46 % of bisexual adults and 56 % of heterosexual adults being married.

Therefore, we find partial support for our first hypothesis: bisexual adults report lower relationship quality than heterosexual adults but not significantly lower relationship quality than lesbian/gay adults. Given the small sample size for lesbian/gay adults, this comparison should be interpreted with caution. Furthermore, Hypothesis 2 is not supported, as relationship type does not explain the lower relationship quality reported by bisexual adults.

We also conducted sensitivity analyses to assure robustness. We recategorize individuals who identify as “mostly heterosexual” as heterosexual, leaving a much smaller bisexual sample (bisexual sample n = 230). With this adjustment, our findings on bisexual adults’ lower relationship quality, regardless of relationship type, remain robust (b = −1.65, p < 0.05, 3 and 4 in Table 2). Additionally, we further recategorized “mostly homosexual” into lesbian/gay sample, yielding only 157 bisexual individuals. Our findings remain substantively the same (b = −1.54 p < 0.05, 6 in Table 2)

To test Hypothesis 3, we examine the associations between relationship quality and depressive symptoms. Table 3 presents the linear regression results for mental health outcomes. First, bisexual adults report significantly more depressive symptoms (b = 1.02, p < 0.05, DS 2 in Table 3) compared to heterosexual adults. Relationship quality is associated with fewer depressive symptoms (b = −0.22, p < 0.05, DS 2 in Table 3). These associations remain significant after adjusting for demographic, socioeconomic, and relationship covariates.

Table 3.

Regression results on depressive symptoms.

	1	2	3	4a	5a	6b	7b
Bisexual (Ref = Het)	1.64***	1.02***	−1.37	1.42*	−2.40	1.58**	−0.03
	(0.19)	(0.20)	(0.96)	(0.56)	(2.57)	(0.58)	(2.45)
Lesbian/Gay	0.07	0.08	−1.64	0.27	−1.23	0.15	−4.36
	(0.51)	(0.80)	(2.96)	(0.86)	(2.95)	(1.11)	(2.76)
Relationship Quality	−0.26***	−0.22***	−0.23***	−0.23***	−0.23***	−0.23***	−0.23***
	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)
Bisexual × RQ			0.08*		0.14		0.06
			(0.03)		(0.09)		(0.09)
Lesbian/Gay × RQ			0.06		0.06		0.16
			(0.10)		(0.10)		(0.09)
Cohabitating (Ref = Mar.)		0.25	0.24	0.28*	0.28+	0.28*	0.28+
		(0.14)	(0.14)	(0.14)	(0.14)	(0.14)	(0.14)
Dating		0.21	0.19	0.21	0.20	0.21	0.20
		(0.18)	(0.19)	(0.19)	(0.19)	(0.18)	(0.18)
Dep. Symptoms W1		0.26***	0.26***	0.26***	0.26***	0.26***	0.26***
		(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)
Male (Ref = F.)		−0.62***	−0.62***	−0.74***	−0.74***	−0.74***	−0.74***
		(0.11)	(0.11)	(0.11)	(0.11)	(0.11)	(0.11)
Age		−0.03	−0.03	−0.04	−0.04	−0.04	−0.04
		(0.03)	(0.03)	(0.03)	(0.03)	(0.03)	(0.03)
Black (Ref = White)		0.70***	0.70***	0.64**	0.63**	0.64**	0.63**
		(0.20)	(0.20)	(0.19)	(0.19)	(0.19)	(0.19)
Latino		0.21	0.21	0.19	0.20	0.19	0.20
		(0.21)	(0.21)	(0.20)	(0.20)	(0.20)	(0.20)
Asian		0.54+	0.53	0.49	0.49	0.49	0.48
		(0.32)	(0.32)	(0.32)	(0.32)	(0.32)	(0.32)
Other		0.28	0.25	0.30	0.30	0.29	0.30
		(0.42)	(0.42)	(0.43)	(0.42)	(0.43)	(0.43)
College Degree		−0.85***	−0.85***	−0.85***	−0.85***	−0.85***	−0.85***
		(0.11)	(0.11)	(0.11)	(0.11)	(0.11)	(0.11)
Have Children		−0.17	−0.17	−0.19	−0.19	−0.19	−0.19
		(0.11)	(0.11)	(0.11)	(0.11)	(0.11)	(0.11)
Interracial Rel.		0.15	0.16	0.17	0.18	0.17	0.17
		(0.15)	(0.15)	(0.15)	(0.15)	(0.15)	(0.15)
Age Diff.		0.05**	0.05**	0.05**	0.05**	0.05**	0.05**
		(0.02)	(0.02)	(0.02)	(0.02)	(0.02)	(0.02)
Rel. Duration (Year)		−0.01	−0.01	−0.01	−0.01	−0.01	−0.01
		(0.02)	(0.02)	(0.02)	(0.02)	(0.02)	(0.02)
Same-Sex Rel.		−0.38	−0.40	−0.69	−0.91	−0.30	−0.56
		(0.71)	(0.69)	(0.78)	(0.75)	(1.07)	(1.01)
Midwest (Ref = NE)		−0.03	−0.02	−0.04	−0.04	−0.04	−0.04
		(0.19)	(0.19)	(0.18)	(0.18)	(0.18)	(0.19)
South		−0.28+	−0.27+	−0.31*	−0.30*	−0.31*	−0.30*
		(0.15)	(0.15)	(0.15)	(0.15)	(0.15)	(0.15)
West		−0.02	−0.02	−0.01	−0.01	−0.01	−0.01
		(0.18)	(0.18)	(0.18)	(0.18)	(0.18)	(0.18)
N	11,062	11,062	11,062	11,062	11,062	11,062	11,062
R-squared	0.11	0.21	0.21	0.21	0.21	0.21	0.21

⁺p < 0.1,

^*p < 0.05,

^**p < 0.01,

^***p < 0.001.

Data from Add Health.

^a“Mostly heterosexual” identity is recategorized as heterosexual sample.

^b“Mostly heterosexual” identity is recategorized as heterosexual sample, and “Mostly Homosexual” identity is recategorized as Lesbian/Gay sample.

Next, we test an interaction between sexual orientation and relationship quality to determine whether these associations vary by sexual orientation. The interaction term is significant in our regression model predicting depressive symptoms (b = 0.08, p < 0.05, DS 3 in Table 3). To facilitate interpretation, the results are visualized in Fig. 1. Fig. 1 illustrates the predicted depressive symptoms based on relationship quality across sexual orientation groups. Bisexual adults do not derive greater mental health benefits from high-quality relationships; instead, they experience fewer benefits compared to heterosexual adults. While all three groups exhibit similar depressive symptoms when relationship quality is low, heterosexual adults receive the greatest mental health benefits from high-quality relationships, reporting significantly fewer depressive symptoms. Bisexual adults in higher quality relationships are associated with fewer depressive symptoms than bisexual adults reporting lower relationship quality (Δ_bisexual = −0.15, p < 0.05). However, when examining only those with above-average relationship quality (${\overline{x}}_{\mathit{\text{relationship}}\mathit{\text{quality}}}=29.2$), bisexual adults report significantly more depressive symptoms than heterosexual adults in relationships of similar quality.

Fig. 1.

Predicted depressive symptoms by romantic relationship quality and sexual orientations (Estimates based on 3 in Table 3).

3.3. Sensitivity check

Similarly, we conduct additional regression analyses on depressive symptoms using a series of alternative samples of bisexual adults in relationships. First, we exclude those who identify as “mostly heterosexual” from the bisexual category. In this sample, we find an even weaker association between relationship quality and depressive symptoms for bisexual adults, and the association is not statistically significant (Δ_bisexual = −0.09, p >. 1). In Fig. 2, bisexual adults in higher-quality relationships (relationship quality score > 29.2) still report significantly more depressive symptoms than heterosexual adults in relationships of the same quality. We then replicate the results by moving “mostly homosexual” to lesbian/gay sample, and found a slightly stronger and significant association between relationship quality and depressive symptoms for bisexual adults (Δ_bisexual = −0.17, p < 0.05), and the interaction term is no longer statistically significant, likely driven by small sample size in this categorization. When focusing on respondents in higher-quality relationships (relationship quality score >29.2), bisexual adults report significantly more depressive symptoms than heterosexual adults in relationships of comparable quality (see Fig. 3).

Fig. 2.

Predicted depressive symptoms by romantic relationship quality and sexual orientations (Estimates based on 5 in Table 3)

Note: Same model as Fig. 1, but “Mostly Heterosexual” sample is re-classified as “Heterosexual” sample.

Fig. 3.

Predicted depressive symptoms by romantic relationship quality and sexual orientations (Estimates based on 7 in Table 3)

Note: Same model as Fig. 1, but “Mostly Heterosexual” sample is re-classified as “Heterosexual” sample and “Mostly Homosexual” is re-classified as “Lesbian/Gay” sample.

4. Discussion

This study demonstrates that relationship quality is linked to mental health outcomes among bisexual adults. Consistent with prior speculations (Hsieh and Liu, 2019) and our first hypothesis, partnered bisexual adults reported significantly lower relationship quality than heterosexual adults. Moreover, they report lower relationship quality even after controlling for relationship type, which contradicts our second hypothesis. This suggests that bisexual adults tend to perceive lower relationship quality across different relationship types. Binegativity is distinct from the minority stressors faced by same-sex couples because it can come directly from one’s romantic partner, and is associated with lower relationship quality for bisexual individuals (Smith et al., 2022). Additionally, bisexual adults disclose their identity to significant others less frequently than lesbian/gay adults (Barringer et al., 2017). They may conceal their bisexual identity in relationships due to potential binegativity from their partners, which is linked to lower relationship quality. The lack of difference in relationship quality between lesbian/gay adults and bisexual/heterosexual adults may reflect that lesbian/gay adults also face discrimination against their same-sex relationships, which may contribute to partners’ worse mental health and increase relationship conflict (LeBlanc and Frost, 2020; Neilands et al., 2020).

Consistent with our third hypothesis, we find that high-quality relationships are associated with fewer depressive symptoms. It is likely that bisexual adults in high-quality relationships have more accepting partners who hold less binegative attitudes (Smith et al., 2022) and they feel more comfortable sharing their bisexual identity and past experiences with supportive partners (Feinstein et al., 2021; McLean, 2007). Scholars suggest that a partner’s bipositivity—acceptance and understanding of bisexuality—is crucial for bisexual adults to feel accepted in intimate relationships (Fuller and Hovland, 2022). Therefore, a high-quality relationship likely reflects the presence of a supportive romantic partner, which may buffer against bi-specific discrimination experienced outside of the relationship.

For our exploratory fourth hypothesis, we find that bisexual adults benefit less—not more—from high-quality relationships compared to heterosexual adults. Our findings consistently demonstrate that bisexual adults in above-average-quality relationships report significantly more depressive symptoms than heterosexual adults in relationships of similar quality. High-quality romantic relationships alone may be insufficient to fully mitigate the bi-specific stressors present in the broader social environment, which are associated with the elevated mental health challenges faced by bisexual adults (Bostwick, 2012; Institute of Medicine, 2011; Meyer, 2003). Specifically, even in high-quality relationships, bisexual adults may experience internalized binegativity by absorbing societal stigma against bisexuality, which can erode their mental health (Meyer, 2003; Pollitt and Roberts, 2021; Velasco et al., 2024). Similarly, partner invalidation of bisexual identity and microaggressions, such as heightened surveillance of attractions to other genders, can undermine well-being despite otherwise positive relationship qualities (Armstrong and Reissing, 2014; Feinstein et al., 2019, 2021; Smith et al., 2022). To preserve trust and relationship stability (Anderson and McCormack, 2016; Feinstein et al., 2020; Pallotta--Chiarolli, 2016), some bisexual adults may conceal their identity within relationships, a strategy that may bolster perceived quality but may also generate daily psychological stress (Feinstein et al., 2020). Taken together, all these factors may explain why bisexual adults in high-quality relationships show greater depressive symptoms than their heterosexual counterparts.

Furthermore, bisexual adults overwhelmingly form different-sex relationships, which may confer fewer health benefits than same-sex relationships, as prior research has found that same-sex relationships are associated with better health outcomes for bisexual individuals (Hsieh and Liu, 2019). Same-sex partners are more likely to invest time in health regulation and emotional work that support mental well-being than different-sex partners (Reczek and Umberson, 2012; Umberson et al., 2017; Wang and Umberson, 2022). Prior studies also show associations suggesting that different-gender partners of bisexual adults may be more likely to express interpersonal hostility (Smith et al., 2022), whereas same-gender partners are more likely to endorse tolerant beliefs, as lesbian and gay individuals on average report lower levels of binegative attitudes than heterosexual adults (Dodge et al., 2016; Friedman et al., 2014). Therefore, bisexual adults in different-gender relationships may be less likely to have partners who affirm their sexual identity, which can help explain their lower reported relationship quality and weaker mental health benefits. The overwhelming number of bisexual adults in different-sex relationships may limit the health benefits they receive, reflecting broader societal discrimination and stigma against same-sex relationships at the time the data were collected in 2008.

Interestingly, our bisexual sample consisted of predominantly women (83 %), echoing existing research that women are more likely to endorse a bisexual identity than men (England et al., 2016) and are more sexually fluid in their attraction, behavior, and identity than men (Katz-Wise, 2015). Because women consistently report higher levels of depressive symptoms than men in population-based studies (Salk et al., 2017), part of the elevated depressive symptoms observed among bisexual adults may be linked to gender differences rather than sexual identity alone. Further, pre-existing depressive symptoms may influence how respondents self-report their relationship quality because those with more symptoms may view their relationships more negatively. In our study, bisexual adults had higher baseline depressive symptoms which may shape how they interpret and report their lower relationship quality relative to heterosexual adults. Future studies with larger bisexual samples and measures that distinguish gender from sexual orientation could help clarify whether individuals with elevated depressive symptoms tend to perceive their relationships less positively and, in turn, refine our understanding of how high-quality relationships are linked to mental health among bisexual adults.

This study has several limitations. While we control for several partner characteristics, partner’s sexual identity is not measured in Add Health. Therefore, we cannot assess the congruence between respondents’ sexual identity and that of their partners, which may be relevant to relationship quality. For example, bisexual adults in relationships with other bisexual individuals may report higher relationship quality than those in mixed-orientation relationships, as they may experience less binegativity (Kays et al., 2014; Vencill and Wiljamaa, 2016). Similarly, we cannot determine whether respondents have disclosed their bisexual identity to their partners. Those who conceal their sexual identity within their relationship may experience lower relationship quality and poorer mental health. It is also important to directly measure partner binegativity and bipositivity among individuals who have disclosed their bisexual identity, as these factors may shape relationship experiences (Fuller and Hovland, 2022; Smith et al., 2022). Future research would also benefit from disaggregating bi + identities, such as pansexual, heteroflexible, and homoflexible identities, that diverge from monosexual orientations (Barringer et al., 2017; Savin--Williams and Vrangalova, 2013; Vrangalova and Savin-Williams, 2012), in order to examine whether and how their relationship quality and mental health outcomes differ from one another.

While we control for depressive symptoms at Wave 1, selection into romantic relationships based on mental health status may still influence our results. For instance, those with poorer mental health may have difficulty entering into, or sustaining high-quality relationships, regardless of sexuality. Depressive symptoms are associated with entering into those relationships – and progressing within them (Sandberg-Thoma and Kamp Dush, 2014), suggesting that those in high-quality relationship may also be those with better mental health. This bidirectionality may help explain our finding that bisexual adults in high-quality relationships reported more depressive symptoms than their heterosexual counterparts in high-quality relationships, given that bisexual adults had elevated baseline depressive symptoms compared to heterosexual adults. Longitudinal research could better disentangle the bidirectionality within these association and provide a more comprehensive understanding of relationship quality and mental health among bisexual adults.

Finally, it is important to note that the Add Health data were collected in 2008, before the national legalization of same-sex marriage. Although contemporary nationally representative longitudinal data with bisexual populations is limited, our findings may not fully reflect the experiences of bisexual adults in more recent contexts. Future research collecting and using more recent population-level data will be crucial to assess whether the patterns observed here persist in recent cohorts and social contexts.

This study is among the first population-level analyses to examine the associations between relationship quality and mental health with a focus on partnered bisexual adults. Using a nationally representative sample, our findings align with prior research suggesting that bisexual adults tend to report lower-quality relationships than other sexual identity groups, particularly heterosexual adults. Furthermore, while high-quality relationships are associated with better mental health for bisexual individuals, the mental health benefits appear smaller than those observed among heterosexual counterparts. These results suggest that bisexual adults’ experiences with relationship-specific stressors, such as identity erasure and concealment, may be linked to both their mental health and relationship quality (Barringer et al., 2017; Feinstein et al., 2019; Mernitz et al., 2022; Smith et al., 2022). Moreover, the broader social context—beyond intimate relationships—plays a crucial role in shaping bisexual adults’ mental health outcomes. Addressing bi-specific stigma and discrimination both within relationships and in the wider social environment is essential for improving relationship quality and alleviating the mental health burdens faced by bisexual adults.

Data availability

The National Longitudinal Study of Adolescent to Adult Health (Add Health) is available here:https://addhealth.cpc.unc.edu/