Assessing Bisexual Stigma and Mental Health Status: A Brief Report

Abstract

Bisexual women often report higher rates of depression and mental health problems than their heterosexual and lesbian counterparts. These disparities likely occur, in part, as a result of the unique stigma that bisexual women face and experience. Such stigma can in turn operate as a stressor, thereby contributing to poor mental health status. The current pilot study tested a new measure of bisexual stigma and its association with mental health. Results suggest a moderate positive correlation between the two, and point to areas for future consideration when measuring bisexual stigma.

Over the past three decades, investigations into the relationship between sexual orientation and health have increased exponentially. The burgeoning field of lesbian, gay, bisexual and transgender (LGBT) health has highlighted a number of critical health inequities among these groups when compared to heterosexual or cisgender counterparts. Often, however, bisexual persons have been overlooked in this research, subsumed into “lesbian/bisexual” or “gay/bisexual” categories -- or, in those studies that rely on behavioral measures of sexual orientation, grouped into an “any homosexual experience” category. Although such maneuvers are often driven by a need to increase statistical power (given how small sexual minority sub-samples can be), the price of such analytic choices is a continued lack of understanding and knowledge about health needs and experiences of bisexual persons, as well as a lack of knowledge about if and how they may differ from other groups.

As more and more studies have begun to consider bisexual groups as distinct and separate from lesbian or gay groups, alarming differences have emerged within sexual minority groups, in addition to those differences seen between sexual minorities and heterosexual groups (Bostwick, Boyd, Hughes, & McCabe, 2010; Diamant, Wold, Spritzer, & Gelberg, 2002; Jorm et al., 2002; McCabe, Hughes, Bostwick, Boyd, & West, 2009; Steele, Ross, Dobinson, Veldhuizen, & Tinmouth, 2009). In a number of these studies, bisexual groups, and more specifically bisexual women, have had poorer health outcomes than heterosexual or lesbian women. Differences related to substance use and mental health were particularly acute.

For example, in a population-based Australian study by Jorm and colleagues (Jorm et al., 2002) they found that as a group, bisexuals had the worst mental health outcomes compared to the homosexual and heterosexual groups. Measures included depressive symptoms, anxiety symptoms and negative affect. The only mental health finding that was not significantly higher or worse among bisexuals was the suicidality measure. The authors also found that as a group, bisexuals had less positive social support from family, more negative support from friends, and were more likely to report both childhood and current adverse events (Jorm et al., 2002).

In a more recent paper using data from a national population-based study in the United States (Bostwick, Boyd, Hughes, & McCabe, 2010), my colleagues and I found very high rates of DSM-IV mood and anxiety disorders among bisexual and lesbian women, though bisexual women had the highest rates on almost all disorders. For example, whereas 30.5% of heterosexual women reported depression in their lifetime, 44% of lesbians and nearly 59% of bisexual women reported this (p < .01). Across nine different mood and anxiety disorders, bisexual women had the highest prevalence rates for seven out of the nine disorders (Bostwick et al., 2010).

In order to understand the health behaviors and outcomes of bisexual women—or any group for that matter—it is crucial to acknowledge the socio-cultural context in which they are located, which is to say one in which bisexuality is still highly stigmatized. Stigma associated with bisexuality is evidenced through negative attitudes toward bisexuality (Herek, 2002), the proliferation of harmful stereotypes about bisexuality as “not real” or the province of persons who are deeply confused or lying (Carey, 2005), and the rejection (from both heterosexual and gay and lesbian communities) that bisexual persons (often women; see Rust, 1995), must contend with (Israel & Mohr, 2004; Lehavot, Balsam, & Ibrhim-Wells, 2009; Ochs, 1996). Stigma experiences, in turn, likely operate as stressors among bisexual groups (Brooks 1981); Meyer, 2003), accounting in part for the health disparities we see.

In order to partially test the putative relationship between bisexual stigma and mental health among bisexual women, I created a brief measure of bisexual stigma as part of a pilot study of bisexual women’s health, the Women’s Health and Identity Study (WHIS). Aims for the current paper are to:

1. describe the stigma measure, including results from the sample,

2. report on feedback from participants on the measure, and

3. test the association between bisexual stigma and mental health.

Methods

The Women’s Health and Identity Study (WHIS) was a mixed method pilot study, whose aim was to explore bisexual women’s experiences of sexual identity stigma and discrimination and how such experiences may affect their health. The study included a new measure of bisexual stigma. The WHIS consisted of a short, self-administered survey and semi-structured qualitative interviews. All participants completed the survey. One-quarter of the desired target sample (desired sample=60; actual sample=47) was randomized a priori into the qualitative interview component. The survey took approximately 20 minutes to complete, and qualitative interviews ranged from 20 minutes to an hour. All participants were given a $10 gift card as an incentive. Data were collected from December 2006 through June 2008 in Ann Arbor, Michigan and Chicago, Illinois. The final N was 47, with 13 women also completing qualitative interviews.

Convenience and snowball sampling were used to recruit participants. Information about the study was distributed through personal networks and listservs; in community publications, both LGBT-focused and general population; and via flyers posted in coffee shops, churches, bookstores, libraries and LGBT organizations. In addition, all respondents were asked to pass along study information to other bisexual women who might be interested in participating.

Potential participants telephoned the study office to go through a brief screening, to ensure that they met inclusion criteria: participants had to be at least 25 years of age, identify as bisexual, and identify as a woman. The author met each participant at a place of their choosing and convenience, including their homes, the author’s office, or a coffee shop. However, qualitative interviews were only conducted in private locations.

Upon meeting, the consent procedures were outlined. Each participant was given a copy of the form for their records. Once participants signed the consent form, they were given the survey to complete. The author remained nearby to provide clarification should any questions be unclear or confusing. Upon completion of the survey, every participant was asked three open-ended questions: Do you have any additional comments or anything else to add? Did the questions make sense? Were questions related to stigma and discrimination clear; was there anything missing? Only responses to the latter question are reported on here.

The study was approved by the University of Michigan and Adler School of Psychology Institutional Review Boards. In addition, a Certificate of Confidentiality was obtained from the National Institutes of Health to assure additional protection of the participants’ data.

Measures

Demographics

Sexual identity was assessed during the phone screening noted above, with the following question: Recognizing that sexual identity is only part of your identity, how do you define yourself? Those women who said bisexual were included. In addition, based on further discussion with some participants, women who chose the label “queer/bisexual” were also included. One unlabeled woman was also included, based on lengthy discussion with the researcher. Race/ethnicity, age, relationship status, and education level were also recorded.

Stigma

The stigma measure for this study drew upon components of Pinel’s Stigma Consciousness Scale (1999) and the rejection dimension of Fife and Wright’s (2000) multi-dimensional measure of stigma. In addition, there was a single question to assess internalized biphobia and a single item meant to assess the larger cultural delegitimization and contestation of bisexuality as a “real” identity (created by the author). Response options ranged from strongly agree (5) to strongly disagree (1). The final scale consisted of eleven questions (see Table 1). Cronbach’s alpha was .83, demonstrating good internal reliability.

Table 1.

Item and Subscale Frequencies and Means of Bisexual Stigma Measure (n=47)

Scale Items	Strongly Disagree/Disagree % (n)	Neither agree nor disagree % (n)	Strongly Agree/Agree % (n)	Mean (sd)
Stigma Consciousness
I worry that my behaviors will be viewed as stereotypically bisexual	36.2 (17)	19.1(9)	44.7 (21)	3.06 (1.39)
Stereotypes about bisexuals affect me	12.8 (6)	4.3 (2)	83.0 (39)	3.98 (1.17)
Most lesbians/gays have a problem with bisexuals	10.6 (5)	40.4 (19)	48.9 (23)	3.45 (1.02)
Most heterosexuals have problem with bisexuals	25.5 (12)	44.7 (21)	29.8 (14)	3.02 (0.94)
Subscale mean	-	-	-	3.38 (0.85)
Rejection
Fear that lesbians will reject me	8.5 (4)	8.5 (4)	83.0 (39)	4.09 (0.95)
Fear that gay men will reject me	55.3 (26)	25.5 (12)	19.1 (9)	2.53 (1.12)
Fear that heterosexual men will reject me	57.4 (27)	12.8 (6)	29.8 (14)	2.66 (1.37)
Fear that heterosexual women will reject me	27.7 (13)	27.7 (13)	44.7 (21)	3.21(1.10)
Treated with less respect because I am bisexual	14.9 (7)	19.1(9)	66.0 (31)	3.64 (1.09)
Subscale mean	-	-	-	3.23 (0.74)
Contestation Others view my bisexual identity as “untrue” or not real	10.6 (5)	17.0 (8)	72.3 (34)	3.94 (1.03)
Internalized Biphobia Sometimes I wish I weren’t bisexual	66.0 (31)	4.3 (2)	29.8 (14)	2.28 (1.48)
Total stigma score	-	-	-	3.26 (0.71)

Mental Health

Depressive symptomology was assessed using the Community Epidemiological Survey of Depression (CES-D), 20-item version (Radloff, 1977). The CES-D is not a diagnostic tool but rather is meant to screen for symptoms of depression. It has shown excellent reliability across a variety of groups. Those participants with scores of 16 or above are considered to meet the cut-off for depression, with higher scores indicating more depressive symptomology. Cronbach’s alpha was .81 in the current sample.

Analysis

Univariate statistics were run, including frequencies and means. In addition, two-tailed Pearson correlation statistics were calculated in order to test the associations between the overall stigma “score”, the four stigma sub-scales, and CES-D scores. Due to the small sample size and exploratory nature of this pilot study, a less conservative p-value of p <.10 was used to determine statistical significance.

Results

Of the participants, 83% identified as white, 75% had a Bachelor’s degree or higher, and 65% were in a relationship (married, living with partner, or in a committed relationship but not living with a partner) at the time of study. The age range of the sample was 25 to 66, with an average age of 33.5 (sd=9.2). The mean score on the CES-D was 18.9 (sd=11.7), reflecting the fact that a majority of the sample was experiencing depressive symptoms at the time of the interview.

Table 1 shows the means and frequencies for each of the single items, as well as the subscales of the bisexual stigma measure. Those items that respondents most agreed with (answered either “strongly agree” or “agree”) were “stereotypes about bisexuality affect me personally” (83%), “I fear that lesbians will reject me because of my bisexuality” (83%), and “ I feel that others views my bisexual identity as ‘untrue’ or not a real identity” (72%). Respondents most disagreed with the statement “Sometimes I wish I weren’t bisexual”, with 66.0% strongly disagreeing or disagreeing. Overall, the mean stigma score for the sample was 3.26, suggesting a tendency toward agreement with the items.

There was a modest positive correlation between CES-D scores and stigma (r=.26, p<.10), suggesting that more stigma was associated with higher depressive symptomology (Table 2). This association appeared to be driven by the stigma consciousness subscale, which was the only subscale that was significantly associated with the CES-D score (r=.28, p <.10). In terms of the correlations of the stigma subscales and the overall measure itself, stigma consciousness, rejection, and contestation were significantly and highly correlated with one another, and with the overall scale. The single item assessing internalized biphobia was not significantly associated with either the stigma consciousness subscale or the contestation item.

Table 2.

Correlations Between Stigma Variables and CES-D

		1	2	3	4	5
1.	Stigma	1
2.	Stigma Consciousness	.873**	1
3.	Rejection	.926**	.702**	1
4.	Contestation	.734**	.602**	.622**	1
5.	Internalized Biphobia	.444**	.139	.341*	.239	1
6.	CES-D	.257†	.284†	.196	.045	.140

^†p < .10,

^*p < .05,

^**p < .01 level

Participant comments about the stigma measure suggested either changes in wording or additional areas that should be considered in future studies. Two participants specifically noted that the word fear may not be appropriate or the “right word” to use in the scale (it is used in the stem of the questions related to fear of rejection). One participant suggested assessing expectations of rejection instead.

Other constructs or areas to possibly incorporate into the bisexual stigma measure included disappointment from others about the person’s bisexuality, not being taken seriously, not being accepted by others, and other people’s perceptions of the bisexual person. Additionally, some participants suggested the need to capture aspects of stigma within relationships, as well as from other bisexual persons.

There were also comments related to the overall structure of the measure. The issue of whether or not ‘disagree’ or ‘agree’ statements were adequately suited to capture the full extent of people’s experiences was noted. Rather, the suggestion was to assess how frequently people had experienced different enactments of stigma. Finally, one woman suggested that questions should distinguish between “larger community perceptions” versus interactions with individuals.

Discussion

The Women’s Health and Identity Study tested a new measure of bisexual stigma, as well as the relationship between stigma and mental health status. Preliminary tests showed that the measure has high internal reliability. Correlations between sub-scales of the measure were generally high, suggesting that they are measuring a common construct.

There was a modest relationship between bisexual stigma and participants’ mental health status, wherein stronger endorsement of stigma items was positively associated with more depressive symptomology. These findings provide some support to the hypothesis that mental health disparities among bisexual women may in part be associated with the unique stigma that bisexual women face. In order to more rigorously test this putative relationship using the bisexual stigma measure described here, a number of things should be taken into consideration in future studies.

The current study had a very small sample, which was fairly homogeneous in terms of both race/ethnicity and education. Further, the bisexual stigma measure was only pilot tested among women. The continued invisibility of bisexual men’s experiences, across multiple domains of inquiry, needs to be addressed and corrected (Steinman, 2001). Additional testing of this instrument is needed among larger and more diverse samples, to ensure its reliability and validity across different populations. In addition, larger samples will allow for the use of data reduction techniques, such as factor analysis, to further assess relationships between the variables. Conducting exploratory and confirmatory factor analyses can help to identify the underlying factor structure of the measure, such that sub-scales items are verified or if need be, reassessed or dropped all together.

Further, additional items or concepts should be included in future iterations of the measure in order to better capture the breadth and depth of bisexual women’s experiences with stigma. In particular, lack of acceptance, disappointment from others, and not being taken seriously due to a bisexual identity should be incorporated. Such items may more adequately capture social rejection, and in fact are similar to items found in Fife and Wright’s multi-dimensional measure of stigma, which the current measure was in part based upon (Fife &Wright, 2000). These questions should also specify the sources of rejection, e.g., family, friends, the “LGBT” community, and/or romantic partners. Finally, the measure would be strengthened by adding items related to exclusion, e.g., “I have been excluded from lesbian and gay events due to my bisexual identity”, given the exclusionary experiences that many bisexual women continue to report (Bower, Gurevich, & Mathieson, 2002; Esterberg, 1997; Heath & Mulligan, 2008; Lehavot, Balsam, & Ibrhim-Wells, 2009).

Conclusion

This brief report details results from the Women’s Health and Identity Study, specifically, the use of a new measure to assess bisexual stigma and determine its association with mental health, especially depression. This represents one of the first attempts to quantitatively measure stigma related to bisexuality. These preliminary results are instructive for how to improve and strengthen future iterations of a bisexual stigma measure.