Abstract
Background/Purpose
Health disparities between sexual minorities and heterosexuals are well documented and have been explained by differential access to healthcare as well as exposure to discrimination. The current research examines the role that healthcare stereotype threat, or the fear of being judged by healthcare providers based on negative group stereotypes, plays in the health of LGB individuals.
Methods
LGB individuals (N = 1507) in three age cohorts were recruited via random digit dialing to participate in a larger study on sexual minority health. Participants completed measures assessing healthcare stereotype threat, lifetime health diagnoses, life satisfaction, and number of bad physical health days and personal distress in the past 30 days.
Results
Healthcare stereotype threat was associated with higher psychological distress and number of reported bad physical health days. Additionally, the Younger and Middle cohorts reported more stereotype threat than the Older cohort, but reported significantly higher levels of life satisfaction in the face of this threat than those in the Older cohort.
Conclusions
Healthcare stereotype threat was related to poorer mental and physical health among LGB individuals; this was true when these outcomes were assessed over the past 30 days but not when they were assessed in general. Cohort differences in healthcare stereotype threat suggest potential important within group variation that needs further investigating. The research broadens the contexts to which stereotype threat is relevant and establishes a stressor related to LGB health.
Keywords: Health disparities, Minority stress, Healthcare stereotype threat, LGB health
Healthcare stereotype threat predicted poorer health among LGB individuals. Older participants reported less threat, but showed worse links between threat and life satisfaction.
Data continually demonstrate health disparities between sexual minorities (e.g., those identifying as gay, lesbian, or bisexual) and heterosexuals. These disparities exist in terms of mental health [1], physical health [2], and in terms of precursors to poor health such as substance abuse [3]. As well, these disparities show up in research using smaller convenience and community-based samples [4] and also samples derived from national probability sampling methods [5]. Finally, these disparities seem to persist, even as public attitudes toward sexual minorities improve [6].
Myriad explanations have been but forth to explain these sexual orientation health differences including differential access to insurance and health services [7]; higher prevalence of HIV and AIDS among sexual minority men as compared to heterosexual men which then leads to a cascade of other health disparities [8]; and even potential problems with sampling and survey methodologies that bias who participates in these types of studies and how they complete the measures [8]. One potential construct in need of empirical attention that may contribute to sexual orientation health disparities is stereotype threat. Stereotype threat theory posits that individuals who belong to groups about which there are negative stereotypes can fear being judged by others based on these stereotypes and can also fear potentially confirming these stereotypes [9, 10]. Such fears can interfere with performance in the immediate situation, perhaps due to anxiety or because of a burden on working memory [11]. Furthermore, stereotype threat theory proposes, and research corroborates, that there are downstream consequences of stereotype threat including disengagement with and disidentification from the domain where the stereotype in question is relevant.
Originally studied in the context of academic performance [10, 12] and other performance domains such as sports [13], stereotype threat theory applies more broadly and is more recently being tested in non-performance domains. Most centrally, researchers are beginning to highlight healthcare stereotype threat [14], or stereotype threat occurring within the realm of health and healthcare. In a laboratory experiment, Abdou and Fingerhut [15] demonstrated the consequences of stereotype threat in the healthcare context for Black versus White women. Priming stereotypes of Black women in a virtual waiting room led strongly identified Black women to report higher anxiety while they ostensibly waited to see a doctor than similarly identified White women. In contrast, when stereotypes were nor present, there were no differences in anxiety between these groups. In a separate survey study, Abdou et al. [14] examined the prevalence of healthcare stereotype threat in a sample of almost 1,500 older adults as part of the Health and Retirement Study. In this sample, approximately 17% of individuals reported healthcare stereotype threat based on at least one social identity. Furthermore, in comparison to those who did not report healthcare stereotype threat, those who did report such threat had higher odds of reporting hypertension, depressive symptoms, and poor self-rated health.
Building on these ideas, Fingerhut and Abdou [16] made the case that healthcare stereotype threat is relevant to the health encounters of sexual minority individuals. Using research on stereotypes about and prejudice toward LGB people, they showed that the necessary precursor to healthcare stereotype threat, namely the existence of negative stereotypes about LGB individuals that are relevant in the context of healthcare, exists. For example, research has suggested that gay men are perceived as immoral [17] and sexually deviant and promiscuous [18]. Lesbian women are similarly perceived as sexually deviant as well as angry and confused [19]. Furthermore, Fingerhut and Abdou [16] showed that LGB individuals continue to report discrimination in healthcare contexts, setting the stage for furthering fears that one will be judged negatively based on sexual orientation in future encounters with healthcare providers. Finally, the authors used data concerning communication with healthcare providers and delays in seeking healthcare to establish some of the theorized downstream consequences of healthcare stereotype threat. Specifically, the authors cited research showing that many LGB people do not disclose their sexual orientation to physicians, a sign of poor communication between patient and provider, as well as research showing that LGB individuals are more likely than their heterosexual counterparts to delay seeking healthcare, a sign of disengagement from the healthcare domain. Providing some support for these ideas in a small convenience sample of LGBT individuals, Ojeida-Leitner and Lewis [20] showed that reports of healthcare stereotype threat were positively associated with self-reported poor mental health, negative affect, and delays in seeking mental health care.
The central purpose for the study presented here was to provide further empirical evidence to support strong theoretical reasoning and tangential sources of data linking stereotype threat and health outcomes for LGB people. Based on a longstanding literature on minority stress [21], or the unique stress experiences that LGB individuals encounter on account of their minority sexual identity, and on a more recent, smaller body of research on healthcare stereotype threat in non-LGB individuals, it is predicted that higher levels of reported healthcare stereotype threat related to sexual orientation will be associated with poorer health outcomes for sexual minority individuals.
Though the central aim of this particular project was to empirically examine links between healthcare stereotype threat and various health-related outcomes among LGB people, the data were collected as part of a larger project examining potential generational differences in stress, health, and well-being among LGB adults. Specifically, the data were part of the Generations Study (http://www.generations-study.com), which is a 5-year, multi-wave and multi-method study examining health and well-being as well as healthcare utilization and experiences with minority stress among three distinct cohorts of LGB individuals. As such, the data allow and call out for examinations across cohorts of sexual minorities in terms of exposure to healthcare stereotype threat as well as links between healthcare stereotype threat and health outcomes.
Hammack and colleagues [22] suggested that researchers interested in sexuality and sexual minority populations should incorporate a paradigm that “accommodates the significant social and historical change[s]” related to LGB experiences as these changes likely affect the way LGB individuals experience their identity, their place in society, and ultimately their health. Referring specifically to gay men, they wrote the following: “Rather than treating gay men as a unified social category, a life course paradigm seeks to interrogate the historical variability that characterizes gay men’s evolving subjectivities and practices in matters of health and identity development.” As an example, these researchers go on to describe how the life experiences of gay people coming out at the height of the AIDS crisis, a time where there were very few legal protections for LGB people and where societal acceptance of homosexuality was limited, may look very different from those coming out at the height of the marriage equality movement, where both the highest court in the USA and the majority of American people favored extending marriage rights to same-sex couples. Nevertheless, it is not entirely clear the ways in which these changes affect the everyday experiences and health and well-being of individuals. Meyer [23] speaks to this in questioning whether increased societal acceptance leads to less exposure to minority stressors (e.g., stigma and discrimination) while in the same breath wondering if new stressors and/or new coping mechanisms (or the reduction in coping altogether) appear in the face of newfound legal and social acceptance. The analyses presented here will explore differences in reports of healthcare stereotype threat and links between healthcare stereotype threat and health outcomes among three distinct cohorts of LGB individuals which will be further described below in the Methods. Given the nascent stage of research on cohort differences within the LGB community, specific predictions regarding cohort differences were not defined a priori.
Method
Participants and Procedure
Participants were recruited via Gallup, Inc., using random digit dialing of both landline and cell phones. To be invited to participate in the study, individuals had to: self-identify as LGB (and not transgender); have at least a sixth grade education; be proficient enough in English to complete the measures; and identify as Black, LatinX, or White, or multiple categories including at least one of these three. Participation was limited to these three groups as Gallup estimated that there would not be sufficient numbers of individuals from other racial and ethnic groups and who also met the other criteria to provide a large enough sample to allow for subgroup analyses. Again, given power issues, multiracial participants who indicated that they were Hispanic/Latino/a were categorized as such regardless of any other racial/ethnic identities. From there, multiracial participants who included Black as one of their identities were categorized as Black. Finally, multiracial participants who indicated that they were White and did not include Hispanic/Latino/a or Black as racial/ethnic identities were categorized as White. For more information on how race/ethnicity was categorized, see Krueger et al. [24]. Participants who identified as transgender, even if they met the other criteria, were included in a separate study focused on the experiences of transgender individuals (www.transpop.org).
Additionally, participants had to be of an age that was within one of three designated cohorts. Using a life course and social ecological approach to understanding the experiences of individuals [25], the Generations researchers identified key historical events related to the LGB community that likely interacted with life course development for sexual minority individuals and thus shaped subsequent life experiences. Specifically, the researchers identified the Stonewall Inn riots in 1969, the beginning of the ACT UP organization in the fight against AIDS in 1987, and the ruling by the Supreme Court in Massachusetts opening marriage equality to same-sex couples in 2003 as pivotal historical events that define the cultural context surrounding LGB inclusion and thus alter the life experiences of those coming of age at the time of the event. To be included in the current study, an individual had to be roughly between the ages 6 and 13 at the time of one of these three events. Thus, those in the Older cohort (defined by the Generations researchers as the Pride cohort) were 52–59 years old at the time of study recruitment and came of age at a time when homosexuality was classified as a mental disorder but also at a time that witnessed the onset of gay liberation and a gay rights movement. The Middle cohort (defined by the Generations researchers as the Visibility cohort) were 34–41 years old at the time of study recruitment and came of age in an era marked by the AIDS epidemic, which was also a time when the number and prominence of LGB-focused organizations rapidly grew. Finally, the Younger cohort (defined by the Generations researchers as the Equality cohort) were 18–25 years old at the time of study recruitment and were coming of age at a time where there was a larger national push for LGB equality, best characterized by the fight for marriage equality which involved both the taking away of rights as well as the establishment (or reestablishment) of those rights.
Initial recruitment occurred between March 2016 and March 2017. Of the 366,644 individuals screened, approximately 3.5% identified as LGBT, with 27% of these sexual minority participants meeting all the other inclusion criteria. Of those eligible, approximately 80% agreed to participate, with 48% of these individuals actually completing the baseline survey. In order to increase representation of Black and LatinX participants, an enhancement sample was recruited between April 2017 and March 2018 using the same procedures used to recruit the initial sample.
Participants for the current study included 1,507 individuals (803 women, 704 men) who ranged in age from 18 to 60 (M = 36.52, SD = 14.71). In terms of race/ethnicity, 15.6% of the participants were categorized as Black, 19.6% as LatinX, and 64.8% as White. In terms of sexual orientation identity, 55.3% of the participants were classified as lesbian or gay, 32.7% as bisexual, and 12% as some other minority sexual orientation including queer, pansexual, and same-gender loving. In terms of socioeconomic status, 13.5% of the participants met the U. S. Census criterion for poverty [26]; additionally, 19.9% of the participants had a high school education or less. Finally, 44% participants were part of the Younger cohort, 24.5% were part of the Middle cohort, and 31.5% were part of the Older cohort.
Those who met eligibility requirements and agreed to participate completed a self-report survey online or via paper and pencil and received $25 dollars for participation. Participants who completed the initial survey were then contacted at two additional time points, each one year apart to complete subsequent waves of the study. The data examined in this paper come only from the first wave of data collection.
Measures
The key predictor variable, Healthcare Stereotype Threat, was assessed using a measure adapted from Marx and Goff [27] and Abdou et al. [14], which uses self-report to assess experiences of stereotype threat. Using a 1 (Strongly Disagree) to 5 (Strongly Agree) scale, participants rated the extent to which they agreed with four items about thoughts and emotions they may have when seeking healthcare. The items were as follows: “I worry about being negatively judged because of my sexual orientation or gender identity”; “I worry that evaluations of me may be negatively affected by my sexual orientation or gender identity”; “I worry that diagnoses of me/my health may be negatively affected by my sexual orientation or gender identity”; and “I worry that I might confirm negative stereotypes about LGB people.” Responses across the items proved highly reliable (α = .90; M = 2.57, SD = 1.07).
With regard to outcomes, two measures of physical health were used, assessing health across different timeframes. For the first, which we call Bad Physical Health Days, participants responded to the following prompt: “Now thinking about your physical health, which includes physical illness and injury, for how many days during the past 30 days was your physical heath not good?” This is an item taken from the CDC’s Health Related Quality of Life Healthy Days measure [28]. Answers in this sample ranged from 0 – 31 days (M = 5.41, SD = 7.58). The second measure, Health Diagnoses, was modified from the National Health Interview Study [29] and required participants to indicate whether they had ever been told by a doctor or health professional that they had any of 23 health issues. Because healthcare stereotype threat is considered a stressor, we were interested only in physical health outcomes that may be stress sensitive. Based on a review article concerning links among life events, stress, and illness [30], 12 of the 23 health issues were selected. Examples include: hypertension, heart disease, asthma, ulcer, and sleep disorder. A sum of diagnoses was calculated for each participant and ranged from 0 to 8 in the sample (M = 1.08, SD = 1.28).
As with the outcomes for physical health, the outcomes for mental health included two measures, again assessing health across different timeframes. Psychological Distress was assessed using the Kessler Psychological Distress Scale (K6) [31, 32]. Using a 1 (all of the time) to 5 (none of the time) scale, participants indicated how often in the past 30 days they felt (i) nervous, (ii) hopeless, (iii) restless or fidgety, (iv) so depressed that nothing could cheer them up, (v) worthless, and (vi) that everything was an effort. Answers were all reverse-coded and a sum score was created such that higher scores indicated more distress (α = .89, M = 7.66, SD = 5.46, Range: 0–24). The second measure of mental health was the widely used Satisfaction with Life scale [33]. Using a 1 (Strongly Disagree) to 7 (Strongly Agree) scale, participants rated the extent to which they agreed or disagreed with five items assessing one’s feelings about the general conditions of their life. Sample items include: “I am satisfied with my life” and “In most ways my life is close to ideal.” Responses across the items proved highly reliable (α = .91; M = 4.32, SD = 1.63).
The key moderating variable, Cohort, was determined by responses to the question “In what year were you born?” This was then subtracted from the current year at the time the survey was completed. Again, the cohorts were meant to reflect individuals between the ages of 18–25, 34–41, and 52–59. However, due to potential reporting errors and to account for age changes that may have occurred between screening and completion of the survey, the cohort age ranges were expanded ±2. As a result, the actual cohorts consisted of individuals between the ages of 18–27 (Younger cohort), 32–43 (Middle cohort), and 50–61 (Older cohort).
Finally, a variety of covariates were assessed and included in the analyses below. These included race/ethnicity (Black, LatinX, White), sex at birth (Female, Male), education (less than a high school degree, high school degree or more), and poverty status (at or below poverty, above poverty). In addition, we included a measure of perceived stigma to show that healthcare stereotype threat was linked with the outcomes above and beyond a general sense of being stigmatized due to minority sexual orientation. Specifically, perceptions of stigma were assessed via a brief Felt Stigma scale [34]. Using a 1 (Strongly Disagree) to 5 (Strongly Agree) scale, participants rated the extent to which they agreed or disagreed with three items assessing one’s beliefs about how non-LGB people think of LGB people (e.g., “Most people where I live think less of a person who is LGB”). As with the other scales, responses across the items proved reliable (α = .71; M = 2.66, SD = .94).
Analytic Strategy
Because of the nature of the sample, probability weights could be used in the analyses as a way to generalize the sample data to Black, LatinX, and White LGB men and women in the United States in the three distinct age cohorts included. A weighting scheme was created to account for various stages of selection and non-response. Stage 1 accounted for individuals who were contacted by Gallup but chose not to be contacted for follow-up studies; stage 2 accounted for individuals who were deemed eligible for the Generations Study who did not wish to participate; and stage 3 accounted for those who were included in the study but did not complete participation. Additionally, survey weights accounted for the fact that Black and LatinX individuals were oversampled. For information on the weighting scheme used for the quantitative data in the Generations Study, see http://www.generations-study.com/methods. In order to take advantage of these weights, the Complex Samples module in SPSS had to be used.
Figure 1 illustrates the model we tested. Separate analyses were conducted to first examine the direct links between healthcare stereotype threat and the outcomes and then to examine differences in these links based on cohort (i.e., to examine the moderating effect of cohort). In order to get comparisons between all three cohorts, two analyses had to be run for each outcome: in the first, the Older cohort served as the comparison group allowing for comparisons between the Younger and Older cohorts and the Middle and Older cohorts; in the second, the Middle cohort served as the comparison, allowing a direct comparison between the Younger and Middle cohorts. As a final set of analyses, the direct link between cohort and healthcare stereotype threat was examined, again with two analyses being run to examine comparisons among all three cohorts. In all the analyses, sex assigned at birth (0 = female; 1 = male), race/ethnicity (1 = Black; 2 = Latinx; 3 = White), education (0 = less than high school degree; 1 = high school degree or more), poverty classification (0 = at or below poverty; 1 = above poverty), cohort (1 = Younger; 2 = Middle; 3 = Older), and felt stigma all served as control variables.
Fig. 1.
Model showing direct links between healthcare stereotype threat and outcomes and the moderating role of cohort.
Results
Table 1 provides demographic data and Table 2 provides the descriptive statistics for each of the key variables; in both tables, data are presented for the entire sample as well as broken down by cohort. Table 3 provides the correlations among the key variables for the entire sample.
Table 1.
Demographics for the complete sample and each age cohort
| Complete sample | Younger cohort (N = 664; % wtd = 61.81) | Middle cohort (N = 369; % wtd = 20.71) | Older cohort (N = 474; % wtd = 17.48) | |||||
|---|---|---|---|---|---|---|---|---|
| % (unweighted) | % (weighted) | % (unweighted) | % (weighted) | % (unweighted) | % (weighted) | % (unweighted) | % (weighted) | |
| 1. Sex at birth (female) | 53.3 | 59.6 | 59.3 | 65.5 | 54.7 | 58.0 | 43.7 | 40.6 |
| 2. Race/ethnicity | ||||||||
| Black | 15.6 | 16.4 | 18.5 | 17.8 | 18.4 | 17.4 | 9.3 | 10.4 |
| LatinX | 19.6 | 21.2 | 26.8 | 25.7 | 18.4 | 17.2 | 10.3 | 10.2 |
| White | 64.8 | 62.3 | 54.7 | 56.5 | 63.1 | 65.4 | 80.4 | 79.3 |
| 3. Sexual orientation | ||||||||
| Lesbian or gay | 55.3 | 46.9 | 37.0 | 36.5 | 55.3 | 50.2 | 80.8 | 79.9 |
| Bisexual | 32.7 | 40.6 | 45.6 | 48.0 | 33.6 | 40.1 | 13.9 | 14.8 |
| Other | 12.0 | 12.5 | 17.3 | 15.6 | 11.1 | 9.6 | 5.3 | 5.3 |
| 4. Education (high school degree or less) | 19.9 | 41.8 | 32.7 | 54.2 | 8.9 | 21.9 | 10.5 | 21.7 |
| 5. Poverty (at or below poverty) | 13.5 | 18.5 | 17.9 | 20.5 | 11.9 | 18.7 | 8.6 | 11.4 |
Table 2.
Descriptives for key variables for the complete sample and each age cohort
| Complete sample | Younger cohort (N = 664; % wtd = 61.81) | Middle cohort (N = 369; % wtd = 20.71) | Older cohort (N = 474; % wtd = 17.48) | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| M | SD | Est. Pop. M | SE | M | SD | Est. Pop. M | SE | M | SD | Est. Pop. M | SE | M | SD | Est. Pop. M | SE | |
| 1. Healthcare stereotype threat | 2.57 | 1.07 | 2.62 | .03 | 2.66a | 1.08 | 2.64 | .05 | 2.65b | 1.03 | 2.69 | .06 | 2.40a,b | 1.06 | 2.45 | .06 |
| 2. Felt stigma | 2.66 | .94 | 2.71 | .03 | 2.66 | .95 | 2.72 | .04 | 2.66 | .95 | 2.70 | .05 | 2.67 | .94 | 2.69 | .05 |
| 3. Bad physical health days | 5.41 | 7.58 | 5.43 | .23 | 4.74c,d | 6.12 | 4.87 | .29 | 5.28d | 7.51 | 5.92 | .52 | 6.45 | 9.22 | 6.87 | .52 |
| 4. Health diagnoses | 1.08 | 1.28 | .93 | .03 | .61c,e | .82 | .62 | .04 | 1.11e,f | 1.26 | 1.18 | .09 | 1.71c,f | 1.51 | 1.73 | .08 |
| 5. Psychological distress | 7.66 | 5.46 | 8.83 | 0.18 | 9.83c,e | 5.29 | 10.20 | .24 | 6.89e,f | 4.96 | 7.65 | .37 | 5.22c,f | 4.85 | 5.39 | .26 |
| 6. Life satisfaction | 4.32 | 1.63 | 4.15 | .05 | 4.15a | 1.52 | 4.04 | .07 | 4.37 | 1.69 | 4.18 | .11 | 4.52a | 1.71 | 4.50 | .09 |
a p < .05 for comparison between Younger and Older;
b p < .01 for comparison between Younger and Middle;
c p < .001 for comparison between Younger and Older;
d p < .05 for comparison between Younger and Middle;
e p < .001 for comparison between Younger and Middle;
f p < .001 for comparison between Middle and Older. All comparisons control for sex assigned at birth, race/ethnicity, education, and poverty classification.
Table 3.
Correlations among key variables
| 1 | 2 | 3 | 4 | 5 | |
|---|---|---|---|---|---|
| 1. Healthcare stereotype threat | - | ||||
| 2. Felt stigma | .32*** | - | |||
| 3. Bad physical health days | .11*** | .18*** | - | ||
| 4. Health diagnoses | .01 | .12*** | .33*** | - | |
| 5. Psychological distress | .22*** | .21*** | .27*** | .09*** | - |
| 6. Life satisfaction | −.15*** | −.28*** | −.26*** | −.15*** | −.53*** |
*p < .05; **p < .01; ***p < .001.
Central to this research was an examination of direct links between healthcare stereotype threat and physical and mental health. Table 4 shows the results for the full models. Partially confirming the prediction, healthcare stereotype threat was positively related to reported Bad Physical Health Days (B = .65, p < .05) and psychological distress (B = .67, p < .001). However, healthcare stereotype threat was not associated with lifetime physical health diagnoses or with life satisfaction.
Table 4.
Direct link between healthcare stereotype threat and outcomes
| Bad physical health days | Health diagnoses | Psychological distress | Life satisfaction | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| B | SE | t | B | SE | t | B | SE | t | B | SE | t | |
| Intercept | 5.83 | 0.51 | 11.37*** | 1.71 | 0.08 | 20.26*** | 4.43 | 0.29 | 15.30*** | 4.74 | 0.09 | 51.83*** |
| Sex1 | 1.68 | 0.46 | 3.65*** | 0.07 | 0.06 | 1.09 | 1.32 | 0.34 | 3.93*** | −0.08 | 0.10 | −0.83 |
| Race/ethnicity | ||||||||||||
| Bl2 v. Wh3 | −0.30 | 0.64 | −0.47 | 0.07 | 0.08 | 0.82 | −0.61 | 0.49 | −1.25 | −0.48 | 0.13 | −3.69*** |
| Lx4 v. Wh3 | 0.73 | 0.63 | 1.17 | 0.04 | 0.09 | 0.48 | 0.03 | 0.42 | 0.07 | −0.11 | 0.12 | −0.90 |
| Education5 | 0.66 | 0.53 | 1.25 | −0.01 | 0.07 | −0.18 | 0.58 | 0.39 | 1.49 | −0.21 | 0.11 | −1.88† |
| Poverty6 | 0.36 | 0.72 | 0.50 | 0.05 | 0.09 | 0.48 | 2.52 | 0.54 | 4.69*** | −0.63 | 0.14 | −4.54*** |
| Cohort | ||||||||||||
| Y7 v. O8 | −2.74 | 0.64 | −4.26*** | −1.18 | 0.10 | −12.18*** | 4.12 | 0.36 | 11.31*** | −0.28 | 0.11 | −2.47* |
| M9 v. O8 | −1.29 | 0.72 | −1.78† | −0.59 | 0.12 | −5.07*** | 1.87 | 0.42 | 4.47*** | −0.23 | 0.13 | −1.73† |
| Felt stigma | 1.00 | 0.27 | 3.65*** | 0.12 | 0.04 | 3.43** | 0.74 | 0.19 | 3.88*** | −0.36 | 0.05 | −7.11*** |
| HCST10 | 0.65 | 0.28 | 2.31* | −0.03 | 0.04 | −0.79 | 0.67 | 0.19 | 3.51*** | −0.07 | 0.05 | −1.23 |
†p < .10; *p < .05; **p < .01; ***p < .001.
1Sex at Birth: 0 = Female, 1 = Male; 2Black; 3White; 4LatinX; 5Education: 0 = less than high school degree, 1 = high school degree or more; 6Poverty: 0 = at or below poverty, 1 = above poverty; 7Younger Cohort; 8Older Cohort; 9Middle Cohort; 10Healthcare Stereotype Threat.
Comparisons between the Younger and Middle cohorts are not shown as they were run in a separate model. There were no differences in links between healthcare stereotype threat and the outcomes between these two cohorts.
Before turning to the potential moderating role of cohort, direct links between cohort and healthcare stereotype threat were examined. Table 5 shows the results for the comparisons among cohorts looking at healthcare stereotype threat as an outcome. Again, these analyses controlled not only for demographic variables but for felt stigma as well; this was done as a way to examine potential group differences as they related to healthcare-related identity threat irrespective of an overall sense of being stigmatized. Those in both the Younger and Middle cohorts reported significantly more healthcare stereotype threat than did those in the Older cohort (B = .21, p < .01, and B = .27, p < .001, respectively). In contrast, those in the Younger and Middle cohorts did not differ in their reports of healthcare stereotype threat (B = −.06, p = ns).
Table 5.
Cohort predicting healthcare stereotype threat
| Healthcare stereotype threat | |||
|---|---|---|---|
| B | SE | t | |
| Intercept | 2.52 | 0.06 | 41.31*** |
| Sex1 | −0.20 | 0.07 | −2.96* |
| Race/ethnicity | |||
| Bl2 v. Wh3 | −0.02 | 0.09 | −0.25 |
| Lx4 v. Wh3 | 0.11 | 0.08 | 1.45 |
| Education5 | 0.02 | 0.08 | 0.32 |
| Poverty6 | −0.10 | 0.09 | −1.14 |
| Felt stigma | 0.36 | 0.03 | 11.10*** |
| Cohort | |||
| Y7 v. O8 | 0.21 | 0.08 | 2.79** |
| M9 v. O8 | 0.27 | 0.08 | 3.41*** |
*p < .05; **p < .01; ***p < .001.
1Sex at birth: 0 = Female, 1 = Male; 2Black; 3White; 4LatinX; 5Education: 0 = less than high school degree, 1 = high school degree or more; 6Poverty: 0 = at or below poverty, 1 = above poverty; 7Younger cohort; 8Older cohort; 9Middle cohort.
Comparisons between the Younger and Middle cohorts are not shown as they were run in a separate model. There were no differences in reports of healthcare stereotype threat between these two cohorts.
With regard to cohort as a moderator, there were only significant differences based on cohort in the link between healthcare stereotype threat and life satisfaction (see Table 6 and Fig. 2). For the Older cohort, a 1-unit increase in stereotype threat was associated with a −.31 decrease in life satisfaction; however, for the Younger cohort, a 1-unit increase in stereotype threat was associated with a −.03 decrease in life satisfaction (p < .05), and for the Middle cohort, a 1-unit increase in stereotype threat was associated with a .05 increase in life satisfaction (p < .05). There was no significant difference in links between healthcare stereotype threat and the outcomes between the Younger and Middle cohorts. Cohort did not moderate the link between healthcare stereotype threat and reported days of bad physical health, lifetime number of physical health diagnoses, or psychological distress.
Table 6.
Interaction between cohort and stereotype threat on life satisfaction
| Life satisfaction | |||
|---|---|---|---|
| B | SE | t | |
| Intercept | 4.71 | 0.09 | 51.43*** |
| Sex1 | −0.08 | 0.10 | −0.79 |
| Race/ethnicity | |||
| Bl2 v. Wh3 | −0.47 | 0.13 | −3.62*** |
| Lx4 v. Wh3 | −0.11 | 0.12 | −0.90 |
| Education5 | −0.19 | 0.11 | −1.79† |
| Poverty6 | −0.64 | 0.14 | −4.63*** |
| Cohort | |||
| Y7 v. O8 | −0.26 | 0.11 | −2.31* |
| M9 v. O8 | −0.22 | 0.13 | −1.64 |
| Felt stigma | −0.36 | 0.05 | −7.12*** |
| HCST10 | −0.31 | 0.09 | −3.42*** |
| HCST10*Cohort | |||
| Y7 v. O8 | 0.27 | 0.11 | 2.50* |
| M9 v. O8 | 0.35 | 0.14 | 2.55* |
†p < .10; *p < .05; **p < .01; ***p < .001.
1Sex at Birth: 0 = Female, 1 = Male; 2Black; 3White; 4LatinX; 5Education: 0 = less than high school degree, 1 = high school degree or more; 6Poverty: 0 = at or below poverty, 1 = above poverty; 7Younger cohort; 8Older cohort; 9Middle cohort; 10Healthcare stereotype threat.
Comparisons between the Younger and Middle cohorts are not shown as they were run in a separate model. There were no differences in the link between healthcare stereotype threat and life satisfaction between these two cohorts.
Fig. 2.
Links between healthcare stereotype threat and life satisfaction at ±1 standard deviation from the mean for healthcare stereotype threat for the Younger, Middle, and Older cohorts of LGB individuals.
Discussion
Health disparities between LGB individuals and heterosexuals exist and persist across a range of domains including mental health, physical health, and substance use. Using a national probability sample of Black, LatinX, and White LGB men and women from three distinct age cohorts, the study presented in this paper empirically examined the potential role that stereotype threat may play in these health differences. In line with hypotheses, higher levels of self-reported healthcare stereotype threat were associated with poorer health outcomes when assessed in a proximal temporal fashion concerning health in the past 30 days.
Furthermore, there were differences in reports of healthcare stereotype threat based on cohort such that the Younger and Middle cohorts reported higher levels of healthcare stereotype threat than those in the Older cohort. This is an interesting finding given that the minority stress model, which emphasizes the role of social conditions on health, might predict that improved social conditions around LGB inclusion for the Middle and certainly the Younger cohorts should relate to less exposure to minority stressors [23]. It is important to note that age and cohort are conflated in this cross-sectional study, and that the cohort differences in healthcare stereotype threat could be reflective of development and maturity across time. Perhaps those in the Older cohort do not report as much worry about being judged by healthcare providers because over time they have developed a set of coping skills and a sense of resilience with regard to stigma. Those in the Younger and Middle cohorts, in contrast, have had less time to develop effective coping in the face of stigma and so may experience heightened worries. This potential explanation is undermined, however, by the fact that there were not differences across the cohorts in reports of felt stigma more generally (see Table 2). Instead, it may be the case that those in the Older cohort have more established relationships with healthcare providers, both because time has permitted such relationship building and also due to an increased rate of seeing doctors in light of increasing healthcare needs. Increased familiarity with one’s healthcare providers likely facilitates less worry about being judged based on identity. Future research is needed to explore these possibilities and to better understand properties associated with both age and cohort that may exacerbate or ameliorate healthcare stereotype threat.
In addition to predicting differences in exposure to healthcare stereotype threat, cohort moderated the link between healthcare stereotype threat and life satisfaction such that those in the Older cohort showed worse outcomes (i.e., lower life satisfaction) in connection with healthcare stereotype threat than their younger peers. As mentioned earlier, the cohorts were defined by the occurrence of significant LGBT-related events that would have occurred during key developmental time periods in the lives of potential participants. As such, these events may have long-lasting impacts on how experience with gay-related stressors are processed. For individuals in the Older cohort who were developing a sexual minority identity at a time where homosexuality was both medicalized and criminalized, experiences of gay-related stress, particularly stressors associated with the healthcare context, may have particularly deleterious effects. For these individuals, stress exposure may exacerbate feelings of otherness and concerns about being rejected, ultimately harming outcomes such as life satisfaction. In contrast, Younger individuals in the study may be less reactive in terms of life satisfaction in the face of stressors because they may be better able to attribute the stress experience to something about an outlier’s prejudice as opposed to about some internal flaw about oneself or about a larger societal disdain of sexual minorities. Of course, such an interpretation must be taken with caution as cohort did not moderate links between healthcare stereotype threat and the other outcomes.
The data presented here are promising in shedding light on the phenomenon of healthcare stereotype threat among LGB people; at the same time, many more research questions remain. To begin, the correlational nature of the data raises questions about the temporal and/or causal relationship between the constructs. Though minority stress theory would posit that the stressor causes the poorer health outcomes, it could be the case that the experience of poor health leads to hypervigilance and an increased fear of stigmatization in the healthcare context. It is also the case that felt stigma more broadly could independently cause healthcare stereotype threat and poor health outcomes. Thus, studies are needed that demonstrate a causal effect of healthcare stereotype threat on health outcomes among LGB individuals. Experiments like the one by Abdou and Fingerhut [15] in which Black and White women were randomly assigned to a healthcare setting eliciting stereotypes regarding Black women or one that was seemingly identity neutral serve as a model of the kind of work that is needed.
This line of research will also benefit from a broadening of the outcomes assessed, including such constructs as physician trust, likelihood of disclosing one’s sexual orientation to the physician, likelihood of adhering to a physician’s recommendations, number of actual health visits and delay in seeking care. It will also be useful to include physiological data in these studies. Experimental studies, for example, could assess heart rate, blood pressure, and cortisol, as a way to assess the biological components underlying threat and thereby leading to broader deleterious health consequences.
Though we did not have specific predictions as to whether healthcare stereotype threat would be differentially related to health outcomes based on whether these were assessed generally or in the last 30 days, results showed that reports of stereotype threat were related only to health outcomes that were temporally recent. This may reflect methodological issues in how health and well-being are assessed and demonstrate that relationships can more accurately be ascertained when constructs are measured in narrower rather than diffuse ways. In a potentially related vein, these differential relationships may offer a suggestion as to how individuals processed the healthcare stereotype threat questions, using what has happened to them recently to answer the prompts and thereby increasing the likelihood that healthcare stereotype threat would relate to health outcomes that are assessed recently as opposed to overall. Of course, it could also be the case that the lack of a relationship between healthcare stereotype threat and the health diagnoses measure is less about anything temporal in nature and instead reflects the potential lack of sensitivity in that measure. Reflective of this is the fact that 42.6% of the sample reported no health diagnoses. Despite all of this, the theory of stereotype threat would suggest that there are downstream and/or accumulated effects that come from the experience of stereotype threat. Thus, more research is needed to understand the extent and types of health consequences that potentially occur for LGB individuals reporting healthcare stereotype threat.
In the current dataset, cohort served as an important exploratory moderator; additional research should explore other potential within-group moderators of the link between healthcare stereotype threat and health outcomes. Examples of moderators that have been studied in other stereotype threat contexts which could be applied here include: stigma consciousness [35]; locus of control beliefs [36]; and self-monitoring [37]. One very relevant moderator, which Steele theorized about in early writings on stereotype threat [9], is identification with one’s group or the domain in question, with those who are highly identified being most susceptible to threat. Supporting this, Schmader [38] found that strength of one’s gender identity altered the typical performance results found in stereotype threat experiments. More specifically, women higher in gender identity showed performance decrements in the threat versus no threat condition, but women lower in gender identity did not show these differences based on threat. Similar findings have been discovered with regard to racial identity with those who are highly identified with their racial group showing the strongest reaction to healthcare stereotype threat [15]. Future research should investigate differences in the experience of healthcare stereotype threat among LGB people based on various dimensions of identity (e.g., centrality, public regard, private regard, community participation, sexual orientation label, etc.).
In addition to within group moderators, research will benefit from investigations of potential ameliorative interventions. For example, in discussing stereotype threat and ethnic health disparities, Aronson et al. [39] suggested that self-affirmation interventions that were developed to ameliorate stereotype threat in academic settings could also be used in medical settings. These researchers also point to studies showing the importance of cues that signal fairness in reducing stereotype threat. In the medical context such cues could include clear statements regarding diversity and inclusion that get posted in waiting rooms and inclusion of a range of images (e.g., representing same-sex couples as well as different-sex couples) on fliers and posters.
At the center of intervention research might be investigations into the ways in which characteristics and behaviors of healthcare providers impact stereotype threat for LGB patients. One example concerns identity concordance (or discordance) between provider and patient. As a parallel, researchers interested in ethnic health disparities have investigated the role of a physician’s race and more specifically of the racial concordance between patient and provider in predicting healthcare experiences. For example, in a field experiment, Alsan et al. [40] randomly assigned Black men to either a Black or non-Black doctor and assessed whether racial concordance (or lack thereof) predicted a patient’s willingness to utilize various preventive services (blood pressure measurement, BMI measurement, cholesterol test, diabetes test). Results showed that, patients whose race matched that of their doctor increased significantly their likelihood of requesting preventative services, particularly those involving invasive procedures, in comparison to those whose race did not match that of their doctor. The researchers went on to suggest that this finding may occur due to different types and amounts of communication that occur between patient and provider based on racial concordance. Supporting this, the researchers showed that those Black patients assigned a Black doctor were more open in terms of communication and shared personal information and information unrelated to preventative care more often than those assigned a White doctor.
Extending these findings to sexual orientation, it would be useful to examine the role that sexual orientation concordance between patient and provider has on the experience of healthcare stereotype threat and on health outcomes for LGB people. In line with this, in an exploratory study on sexual orientation disclosure to healthcare providers, Klitzman and Greenberg [41] found that such disclosure was positively related to participant’s belief that their healthcare provider was LGB. Furthermore, in a study using the same dataset as the one used in the current study, Martos et al. [42] found that a majority of participants (52%) reported interest in accessing healthcare either from LGB providers or LGB focused clinics in the future, suggesting that identity concordance may be important to sexual minority patients.
As a final note, we feel it is important to comment on the nature of the data used for this study and to urge government agencies to more robustly collect data concerning sexual orientation and gender identity in large-scale, population-based studies. As noted in the 2011 Institutes of Medicine (IOM) report on the health of LGBT people [7], a limitation of much of the research on sexual minorities is the use of non-probability samples. The report recognized that the existing research has been valuable in revealing certain phenomena, in testing interventions, and in bringing to light hypotheses to be tested; at the same time, the non-representativeness of the sample has limited generalizability and an understanding of the population as a whole. The IOM report ends with call for more robust collection of data concerning LGBT identity in population-based studies. Though the data presented in this paper come with their own limitations (e.g., only individuals in three cohorts were included; participation was limited to a small number of racial/ethnic groups), they demonstrate the feasibility of collecting and working with population level data concerning sexual orientation and reveal important phenomena that likely apply more broadly than studies with smaller, convenience samples could allow for.
Acknowledgment
Funding Generations is funded by a grant from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD grant 1R01HD078526).
Contributor Information
Adam W Fingerhut, Department of Psychology, Loyola Marymount University, Los Angeles, CA 90045, USA.
Alexander J Martos, The Williams Institute, School of Law, University of California Los Angeles, Los Angeles, CA, USA.
Soon Kyu Choi, The Williams Institute, School of Law, University of California Los Angeles, Los Angeles, CA, USA.
Cleopatra M Abdou, Suzanne Dworak-Peck School of Social Work, University of Southern California, Los Angeles, CA, USA.
Compliance with Ethical Standards
Conflict of Interest We have no known conflicts of interest.
References
- 1. Kerridge BT, Pickering RP, Saha TD, et al. Prevalence, sociodemographic correlates and DSM-5 substance use disorders and other psychiatric disorders among sexual minorities in the United States. Drug Alcohol Depend. 2017;170:82–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Frost DM, Lehavot K, Meyer IH. Minority stress and physical health among sexual minority individuals. J Behav Med. 2015;38:1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Burgard SA, Cochran SD, Mays VM. Alcohol and tobacco use patterns among heterosexually and homosexually experienced California women. Drug Alcohol Depend. 2005;77:61–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Lock J, Steiner H. Gay, lesbian, and bisexual youth risks for emotional, physical, and social problems: Results from a community-based survey. J Am Acad Child Adolesc Psychiatry. 1999;38:297–304. [DOI] [PubMed] [Google Scholar]
- 5. Sandfort TG, Bakker F, Schellevis FG, Vanwesenbeeck I. Sexual orientation and mental and physical health status: findings from a Dutch population survey. Am J Public Health. 2006;96:1119–1125. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Gonzales G, Henning-Smith C. Health disparities by sexual orientation: results and implications from the behavioral risk factor surveillance system. J Community Health. 2017;42:1163–1172. [DOI] [PubMed] [Google Scholar]
- 7. Institute of Medicine (US) Committee on Lesbian, Gay, Bisexual, and Transgender Health Issues and Research Gaps and Opportunities. The Health of Lesbian, Gay, Bisexual, and Transgender People: Building a Foundation for Better Understanding. Washington (DC): National Academies Press; 2011. [PubMed] [Google Scholar]
- 8. Lick DJ, Durso LE, Johnson KL. Minority stress and physical health among sexual minorities. Perspect Psychol Sci. 2013;8:521–548. [DOI] [PubMed] [Google Scholar]
- 9. Steele CM. A threat in the air. How stereotypes shape intellectual identity and performance. Am Psychol. 1997;52:613–629. [DOI] [PubMed] [Google Scholar]
- 10. Steele CM, Aronson J. Stereotype threat and the intellectual test performance of African Americans. J Pers Soc Psychol. 1995;69:797–811. [DOI] [PubMed] [Google Scholar]
- 11. Schmader T, Johns M, Forbes C. An integrated process model of stereotype threat effects on performance. Psychol Rev. 2008;115:336–356. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Spencer SJ, Steele CM, Quinn DM. Stereotype threat and women’s math performance. J Exp Soc Psychol. 1999;35:4–28. [Google Scholar]
- 13. Stone JA, Sjomeling M, Lynch CI, Darley JM. Stereotype threat effects on black and white athletic performance. J Pers Soc Psychol. 1999;77(6):1213–1227. [Google Scholar]
- 14. Abdou CM, Fingerhut AW, Jackson JS, Wheaton F. Healthcare stereotype threat in older adults in the health and retirement study. Am J Prev Med. 2016;50:191–198. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Abdou CM, Fingerhut AW. Stereotype threat among black and white women in health care settings. Cultur Divers Ethnic Minor Psychol. 2014;20:316–323. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Fingerhut AW, Abdou CM. The role of healthcare stereotype threat and social identity threat in LGB health disparities. J Soc Issues. 2017;73:493–507. [Google Scholar]
- 17. Calabrese SK, Earnshaw VA, Magnus M, et al. Sexual stereotypes ascribed to black men who have sex with men: an intersectional analysis. Arch Sex Behav. 2018;47:143–156. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Hequembourg AL, Brallier SA. An exploration of sexual minority stress across the lines of gender and sexual identity. J Homosex. 2009;56:273–298. [DOI] [PubMed] [Google Scholar]
- 19. Geiger W, Harwood J, Hummert ML. College students’ multiple stereotypes of lesbians: a cognitive perspective. J Homosex. 2006;51:165–182. [DOI] [PubMed] [Google Scholar]
- 20. Ojeida-Leitner D, Lewis RK. Assessing health related stereotype threats and mental healthcare experiences among a LGBT sample. J Prev Interv Community. 2019; 49: 251– 265. [DOI] [PubMed] [Google Scholar]
- 21. Meyer IH. Prejudice, social stress, and mental health in lesbian, gay, and bisexual populations: conceptual issues and research evidence. Psychol Bull. 2003;129:674–697. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Hammack PL, Frost DM, Meyer IH, Pletta DR. Gay men’s health and identity: social change and the life course. Arch Sex Behav. 2018;47:59–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Meyer IH. Does an improved social environment for sexual and gender minorities have implications for a new minority stress research agenda? Psychol Sex Rev. 2016;7:81–90. [PMC free article] [PubMed] [Google Scholar]
- 24. Krueger EA, Lin A, Kittle KR, Meyer IH.. Generations - A Study of the Life and Health of LGB People in a Changing Society. Los Angeles, CA: The Williams Institute; 2020. [Google Scholar]
- 25. Frost DM, Meyer IH, Hammack PL. Health and well-being in emerging adults’ same-sex relationships: critical questions and directions for research in developmental science. Emerg Adulthood. 2015;3:3–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. US Census Bureau. Poverty Thresholds 2018. Available at https://www.census.gov/data/tables/time-series/demo/income-poverty/historicalpoverty-thresholds.html. Accessibility verified on May 4, 2021.
- 27. Marx DM, Goff PA. Clearing the air: the effect of experimenter race on target’s test performance and subjective experience. Br J Soc Psychol. 2005;44:645–657. [DOI] [PubMed] [Google Scholar]
- 28. Centers for Disease Control and Prevention. CDC HRQOL–14 Healthy Days Measure. Available at https://www.cdc.gov/hrqol/hrqol14_measure.htm. Accessibility verified on January 27, 2021.
- 29. National Health Interview Survey (NHIS). 2014 NHIS Questionnaire – Sample Adult. Atlanta, GA: Centers for Disease Control and Prevention; 2014. [Google Scholar]
- 30. Salleh MR. Life event, stress and illness. Malays J Med Sci. 2008;15:9–18. [PMC free article] [PubMed] [Google Scholar]
- 31. Kessler RC, Barker PR, Colpe LJ, et al. Screening for serious mental illness in the general population. Arch Gen Psychiatry. 2003;60:184–189. [DOI] [PubMed] [Google Scholar]
- 32. Kessler RC, Green JG, Gruber MJ, et al. Screening for serious mental illness in the general population with the K6 screening scale: results from the WHO World Mental Health (WMH) survey initiative. Int J Methods Psychiatr Res. 2010;19 Suppl 1:4–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Diener E, Emmons RA, Larsen RJ, Griffin S. The satisfaction with life scale. J Pers Assess. 1985;49:71–75. [DOI] [PubMed] [Google Scholar]
- 34. Herek GM. Hate crimes and stigma-related experiences among sexual minority adults in the United States: prevalence estimates from a national probability sample. J Interpers Violence. 2009;24:54–74. [DOI] [PubMed] [Google Scholar]
- 35. Brown RP, Pinel EC. Stigma on my mind: individual differences in the experience of stereotype threat. J Exp Soc Psychol. 2003;39:626–633. [Google Scholar]
- 36. Cadinu M, Maass A, Lombardo M, Frigerio S. Stereotype threat: the moderating role of locus of control beliefs. Eur J Soc Psychol. 2006;36:183–197. [Google Scholar]
- 37. Inzlicht M, Aronson J, Good C, McKay L. A particular resiliency to threatening environments. J Exp Soc Psychol. 2006;42:323–336. [Google Scholar]
- 38. Schmader T. Gender identification moderates stereotype threat effects on women’s math performance. J Exp Soc Psychol. 2002;38:194–201. [Google Scholar]
- 39. Aronson J, Burgess D, Phelan SM, Juarez L. Unhealthy interactions: the role of stereotype threat in health disparities. Am J Public Health. 2013;103:50–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Alsan M, Garrick O, Graziani GC. Does diversity matter for health? Experimental evidence from Oakland. Am Econ Rev. 2019;109(12):4071–4111. [Google Scholar]
- 41. Klitzman RL, Greenberg JD. Patterns of communication between gay and lesbian patients and their health care providers. J Homosex. 2002;42:65–75. [DOI] [PubMed] [Google Scholar]
- 42. Martos AJ, Fingerhut A, Wilson PA, Meyer IH. Utilization of LGBT-Specific clinics and providers across three cohorts of lesbian, gay, and bisexual people in the United States. SSM Popul Health. 2019;9:100505. [DOI] [PMC free article] [PubMed] [Google Scholar]