Associations between the Self-Reported Likelihood of Receiving the COVID-19 Vaccine, Likelihood of Contracting COVID-19, Discrimination, and Anxiety/Depression by Sexual Orientation

There is limited evolving literature on COVID-19 vaccine uptake and its barriers among sexual minority populations (lesbian, gay, bisexual, transgender, and queer [LGBTQ]), despite their increased COVID-19 risk factors. We assessed the differences in intention to receive the COVID-19 vaccine by self-reported likelihood of contracting COVID-19, anxiety/depression, discrimination frequency, social distancing stress, and sociodemographic factors across sexual orientation. An online national cross-sectional survey was conducted in the United States between 13 May 2021, and 9 January 2022, among adults aged ≥18 (n = 5404). Sexual minority individuals had a lower intention of receiving the COVID-19 vaccine (65.62%) than heterosexual individuals (67.56%). Disaggregation by sexual orientation, however, showed that gay participants had a higher intention of COVID-19 vaccination (80.41%) and lesbian (62.63%), bisexual (64.08%), and non-heterosexual, non-LGB sexual minority (56.34%) respondents had lower intentions of receiving the COVID-19 vaccine than heterosexual respondents. Sexual orientation significantly moderated the association between the perceived likelihood of receiving the COVID-19 vaccine and the self-reported likelihood of contracting COVID-19, anxiety/depression symptoms, and discrimination. Our findings further underline the importance of improving vaccination efforts and access among sexual minority individuals and other vulnerable groups.


Introduction
Since the beginning of the COVID-19 pandemic, marginalized communities, communities of color, and other minoritized groups in the United States (US) have faced a disproportionate burden of COVID-19 risk, morbidity, and mortality [1][2][3]. Among these disproportionately at-risk groups are lesbian, gay, bisexual, transgender, and queer (LGBTQ) communities [4][5][6][7]. LGBTQ persons have a greater incidence asthma, cancer, diabetes, heart disease, substance use, and mental health symptoms-all of which are associated with minority stressors (e.g., discrimination, stigma) and predispose them to COVID-19 morbidity and mortality [4,[8][9][10][11]. Furthermore, it has been reported in studies conducted in the US that LGBTQ adults were more likely to experience worsening physical and mental health compared to non-LGBTQ adults during the pandemic [6,8,12]. For instance, a study utilizing a national sample of US adults found loneliness, stress, psychological distress, and quality of life were worse among sexual minority individuals compared with non-sexual minority persons [8].
The increased rate of COVID-19 risk factors among LGBTQ communities is further compounded by already-existing and sustained financial, social, and structural barriers and moderates the association between intention to receive COVID-19 vaccine and perceived likelihood of contracting COVID-19, anxiety/depression, and discrimination frequency.

Study Design and Samples
This US national cross-sectional survey was conducted between 13 May 2021 and 9 January 2022 among adults aged 18 years or older. The National Institutes of Health conducted it to assess COVID-19 vaccination perceptions, perceived likelihood of contracting COVID-19, social distancing stress, mental health, discrimination frequency, and participants' sociodemographic characteristics. A total of 5938 participants completed the survey that Qualtrics LLC distributed. After data cleaning, 5413 samples were achieved. Participants (n = 5404) from whom complete data was given on responses to a question about the perceived likelihood of receiving the COVID-19 vaccine were included in the analysis. Actual vaccination status was not assessed with the perceived likelihood of receiving the COVID-19 vaccine.

Measures
The perceived likelihood of receiving the COVID-19 vaccine was the dependent variable and was measured using the question, "Now that we have a Coronavirus/COVID-19 vaccine available, have you been/planning to be vaccinated?" (Not at all likely, slightly likely, moderately likely, very likely, and extremely likely). We recategorized the response options into two categories: not at all likely and likely (i.e., slightly likely, moderately likely, very likely, and extremely likely). The recategorization was based on the stratification of sexual orientation and small cell counts or samples within subgroups.
Self-reported sexual orientation, the likelihood of contracting COVID-19, anxiety/depression symptoms, discrimination frequency, or social distancing stress were the main independent variables. Sexual orientation is a spectrum and includes but is not limited to heterosexual, lesbian, gay, bisexual, queer, pansexual, and asexual [24]. However, due to sample size, we dichotomized sexual orientation into heterosexual and sexual minority (lesbian, gay, bisexual, or something else [write-in option]). Perceived likelihood of contracting COVID-19 was determined by asking the participants whether they feel they will contract COVID-19 based on their overall self-rated health (not at all likely or likely). Discrimination frequency was measured based on how often (never, about once a month, about once a week, 2 to 3 times a week, and daily or almost daily) the participants experienced any discrimination during the COVID-19 pandemic. To determine social distancing stress status, the participants were asked, "How stressful has socially distancing been for you?" Response options, which included very stressful, somewhat stressful, a little stressful, or not at all stressful, were recategorized into two groups: (1) stressful if the participants selected very stressful, somewhat stressful, or a little stressful; and (2) not at all stressful.
The sociodemographic factors that may be related to perceived likelihood of receiving the COVID-19 vaccine are age in years (18-25, 26-34, 35-49, 50 or more), gender identity (Man, Something else [write-in]/non-binary/transgender, Woman), race/ethnicity (Another race/ethnicity [i.e., American Indian, Alaskan Native, Middle Eastern, Pacific Islander], Asian, Black/African American, Hispanic/Latino, White), level of education completed, and annual household income. Thus, this study examined these variables as covariates of the perceived likelihood of receiving the COVID-19 vaccine.

Statistical Analyses
To analyze the data, we first estimated the prevalence of the perceived likelihood of receiving the COVID-19 vaccine based on sexual orientation (heterosexual, lesbian, gay, bisexual, and something else) using Chi-Squared tests. Second, we performed descriptive and bivariate statistics to describe the prevalence of and differences in the perceived likelihood of receiving the COVID-19 vaccine by sociodemographic characteristics, perceived likelihood of contracting COVID-19, anxiety/depression, discrimination frequency, and social distancing stress across sexual orientation (heterosexual vs. sexual minority). Chi-Squared tests were used to determine the differences. Furthermore, we conducted moderation analyses to determine whether sexual orientation moderates the association between the perceived likelihood of receiving the COVID-19 vaccine and anxiety/depression, discrimination frequency, perceived likelihood of contracting COVID-19, or social distancing stress, respectively.
Finally, to evaluate the association between the perceived likelihood of receiving the COVID-19 vaccine and anxiety/depression, discrimination frequency, perceived likelihood of contracting COVID-19, and social distancing stress, adjusting for the sociodemographic characteristics based on sexual orientation (heterosexual vs. sexual minority), multivariable logistic regression models were utilized. The models were conducted at an alpha level of 0.05, 95% confidence intervals (CIs), and adjusted odds ratios (AORs). We examined multicollinearity among the independent variables, but no significant multicollinearity issues were observed because the mean-variance inflation factor (VIF) of 1.15 was less than the 10 thresholds for serious multicollinearity [27]. The analyses were performed using STATA/SE version 16 [28]. We applied listwise deletion method to handle the missing values. This study received Institutional Review Board (IRB) approval on 23 December 2020 (IRB #000308) as an exempt protocol from the IRB-Human Research Protections Program-Office of Human Subjects Research Protections. Figure 1 shows the prevalence of the perceived likelihood of receiving the COVID-19 vaccine by sexual orientation. There was a statistically significant difference in the perceived likelihood of receiving the COVID-19 vaccine among the sexual orientation groups (chi2[4] = 14.05, p = 0.007). Gay individuals had a higher prevalence of the likelihood of receiving the vaccine, followed by those who identified as heterosexual, bisexual, lesbian, and something else. Table 1 presents the descriptive and bivariate analysis results based on sexual orientation. The highest proportion of heterosexual participants were aged 35-49 years (23.06%), women (62.85%), White (43.26%), had college or higher education (39.89%), had an annual household income of $75,000 or more (27.40%), perceived likelihood of contracting COVID-19 as likely (68.55%), and indicated social distancing as stressful (68.16%). A proportion of the participants also experienced mild (21.86%), moderate (11.98%), and severe (10.75%) anxiety/depression symptoms. The participants experienced discrimination about once a month (15.52%), once a week a week (10.16%), 2-3 times a week (9.78%), and daily or almost daily (11.95%). About 67.56% of the heterosexual participants perceived receiving the COVID-19 vaccine. The perceived likelihood of receiving the COVID-19 vaccine varied among subgroups of heterosexual participants. The highest prevalence of the individuals that reported that they were likely to receive the COVID-19 vaccine was among those who were 50 years or older (87.68%), Asian (81.94%), had college or higher education (79.16%), had an annual household income of $75,000 or more (80.48%), or reported being likely to contract COVID-19 (69.75%).  Table 1 presents the descriptive and bivariate analysis results based on sexual orientation. The highest proportion of heterosexual participants were aged 35-49 years (23.06%), women (62.85%), White (43.26%), had college or higher education (39.89%), had an annual household income of $75,000 or more (27.40%), perceived likelihood of contracting COVID-19 as likely (68.55%), and indicated social distancing as stressful (68.16%). A proportion of the participants also experienced mild (21.86%), moderate (11.98%), and severe (10.75%) anxiety/depression symptoms. The participants experienced discrimination about once a month (15.52%), once a week a week (10.16%), 2-3 times a week (9.78%), and daily or almost daily (11.95%). About 67.56% of the heterosexual participants perceived receiving the COVID-19 vaccine. The perceived likelihood of receiving the COVID-19 vaccine varied among subgroups of heterosexual participants. The highest prevalence of the individuals that reported that they were likely to receive the COVID-19 vaccine was among those who were 50 years or older (87.68%), Asian (81.94%), had college or higher education (79.16%), had an annual household income of $75,000 or more (80.48%), or reported being likely to contract COVID-19 (69.75%).

Moderation Analysis
There was a significant interaction between sexual orientation and social distancing stress (chi2[2] = 6.40, p= 0.041). However, the model was not significant (chi2[3] = 7.50, p = 0.058) after including the three terms (main effects and the interaction terms). Figure 2 presents whether sexual orientation moderates the association between the perceived likelihood of receiving the COVID-19 vaccine and the likelihood of contracting COVID-19. There was a significant interaction between sexual orientation and the likelihood of contracting COVID-19 (chi2[2] = 28.33, p < 0.001). Heterosexual and sexual minority individuals who had a greater perceived likelihood of contracting COVID-19 had the highest probability of being likely to receive the COVID-19 vaccine compared to their counterparts who reported their likelihood of contracting COVID-19 as not likely. Figure 3 shows that the moderation effects of sexual orientation on anxiety/depression symptoms influence the likelihood of receiving the COVID-19 vaccine. Sexual orientation significantly moderated the association between anxiety/depression symptoms and the perceived likelihood of receiving the COVID-19 vaccine (chi2[4] = 47.45, p < 0.001). The highest probability of being likely to receive the COVID-19 vaccine was observed among sexual minority individuals who had mild anxiety/depression symptoms.
The moderation effects of sexual orientation on the association between the perceived likelihood of receiving the COVID-19 vaccine and discrimination frequency during the pandemic are presented in Figure 4. Sexual orientation significantly moderated the association between the perceived likelihood of receiving the COVID-19 vaccine and discrimination frequency during the pandemic (chi2 [5] = 84.13, p < 0.001). Sexual minority individuals who experienced discrimination about once a month during the pandemic had the highest probability of COVID-19 vaccine intention compared to their counterparts. Figure 2 presents whether sexual orientation moderates the association between the perceived likelihood of receiving the COVID-19 vaccine and the likelihood of contracting COVID-19. There was a significant interaction between sexual orientation and the likelihood of contracting COVID-19 (chi2[2] = 28.33, p < 0.001). Heterosexual and sexual minority individuals who had a greater perceived likelihood of contracting COVID-19 had the highest probability of being likely to receive the COVID-19 vaccine compared to their counterparts who reported their likelihood of contracting COVID-19 as not likely.     The moderation effects of sexual orientation on the association between the perceived likelihood of receiving the COVID-19 vaccine and discrimination frequency during the pandemic are presented in Figure 4. Sexual orientation significantly moderated the association between the perceived likelihood of receiving the COVID-19 vaccine and discrimination frequency during the pandemic (chi2 [5] = 84.13, p < 0.001). Sexual minority individuals who experienced discrimination about once a month during the pandemic had the highest probability of COVID-19 vaccine intention compared to their counterparts.

Multivariate Logistic Regression
Stratified by sexual orientation, the logistic regression analysis results of the associations between the perceived likelihood of receiving the COVID-19 vaccine and sociodemographic characteristics, perceived likelihood of contracting COVID-19, anxiety/depression symptoms, discrimination frequency, and social distancing stress are presented in Table 2.

Multivariate Logistic Regression
Stratified by sexual orientation, the logistic regression analysis results of the associations between the perceived likelihood of receiving the COVID-19 vaccine and sociodemographic characteristics, perceived likelihood of contracting COVID-19, anxiety/depression symptoms, discrimination frequency, and social distancing stress are presented in Table 2.

Discussion
The present study elucidated interesting trends across sexual orientation for the perceived likelihood of receiving the COVID-19 vaccine. Our multivariate logistic regression analysis yielded similar findings to previous literature as individuals who reported lower income (heterosexual respondents) and education levels (heterosexual and sexual minority respondents) were less likely to report being likely to receive the COVID-19 vaccine [29][30][31][32][33]. This illustrates an important area of concern and intervention for vaccine education and promotion, given individuals of lower socioeconomic status have the highest risk for poor respiratory health [34]. While heterosexual individuals had a higher prevalence of vaccination intention compared to sexual minority participants when aggregated, further analysis of disaggregated sexual orientation showed a higher likelihood of vaccination among gay individuals compared to heterosexual, bisexual, and lesbian participants-with heterosexual participants having a higher prevalence than bisexual and lesbian participants. These findings mirror the variability in vaccination intention across sexual orientations found in previous studies [13,35]. As such, there may be vaccination disparities across sexual orientation groups and, thus, a need to tailor vaccination information, campaigns, and resources to the heterogeneity of sexual minority communities.
Among sexual minority respondents, Black/African American individuals had the lowest likelihood to report vaccination intention, which mirrors prior research [13]. The low likelihood of vaccination among Black/African American individuals may be explained by sexual minority people of color being doubly stigmatized and at risk for discrimination, medical abuse, and harassment due to their sexual orientation and race [36][37][38]. Our finding is particularly concerning given Black sexual minority individuals have also reported a higher prevalence of asthma, diabetes, hypertension, kidney disease, obesity, stroke, and smoking compared to heterosexual persons and other racial/ethnic groups [4]. These comorbidities have also been shown to increase morbidity and mortality among COVID-19-infected patients, especially among Black/African Americans [39][40][41]. Heightened fears of abuse and harassment both inside and outside the clinical setting, coupled with a higher risk of contracting COVID-19, may subsequently negatively impact vaccination coverage. Indeed, Black LGBT individuals have reported lower care access and vaccination coverage compared to other racial/ethnic groups, highlighting an urgent need for more access to healthcare, vaccine education, and vaccines among communities of color [9,13,15,22].
In line with TPB postulations, our moderation analysis showed individuals who reported a higher perceived likelihood of contracting COVID-19 were more likely to express vaccine intent, which is consistent with other studies [20,23,42,43]. Interestingly, we also found sexual minority individuals with mild anxiety/depression symptoms had the highest likelihood of receiving the COVID-19 vaccine, but individuals with moderate and severe anxiety/depression had a lower likelihood of vaccination. Prior research in this area is inconsistent, with some showing individuals with mental health conditions are more likely to be vaccinated [44,45], others showing they are less likely [43,[46][47][48], and some showing variability in vaccination by the type and severity of mental health symptoms [44]. We provide evidence of heterogeneity in vaccine intent by sexual orientation and level of anxiety/depression symptoms, illustrating individuals with mild and severe anxiety/depression may be the most at risk for being unvaccinated. Our findings may be a result of the higher severity of anxiety/depression leading to lower self-efficacy [47]. Prior research has also highlighted the higher mental health and economic burden during the pandemic [6,[49][50][51][52][53]. Moreover, individuals with mental health conditions, especially sexual minorities, are more likely to lack trust in the medical system, experience stigma, and face socioeconomic and structural barriers to care, meaning lower reported vaccine intent among this population is of urgent concern [47, [54][55][56].
Sexual minorities face a multitude of socioeconomic and structural factors that adversely impact their health outcomes, including discrimination and stigma [57][58][59][60]. Indeed, sexual minority persons have reported high levels of discrimination during the pandemic [4,61,62]. In the present study, individuals that reported discrimination once a week or more had the lowest probability of receiving the vaccine for both sexual orientation groups. Avoidance of medical care because of perceived discrimination has been documented among racial/ethnic and sexual minority groups [63][64][65]. Thus, discrimination experiences may also lead to vaccine hesitancy [48], highlighting a need to improve trust and engagement in healthcare settings with populations vulnerable to discrimination. Interestingly, sexual minority individuals who experienced discrimination once a month had the highest perceived probability of receiving the COVID-19 vaccine, indicating a deviation from prior literature [66,67]. Thus, further research is needed to elucidate the mechanism behind this finding.

Limitations
Even though the data from this cross-sectional survey provided a large sample and valuable information on the likelihood of receiving the COVID-19 vaccine among sexual minorities, it is not representative of the US population. In addition, the cross-sectional nature of the data prevents us from establishing temporal sequence or causality. Also, individuals already vaccinated may have been included in the study due to the fact that the survey was conducted without vaccination status verification. Moreover, the perception or likelihood of receiving the COVID-19 vaccine may not translate into actual vaccination among the survey respondents. The survey was conducted only in English and therefore did not include some populations of the US whose primary language is not English. Furthermore, attitude towards vaccination is very dependent on time. Subsequently, this might have shifted participants' perception regarding the COVID-19 vaccine, considering the timespan of the data collection (May 2021-January 2022). Lastly, individuals without internet access or a mobile plan might not have been recruited because the survey was conducted online.

Conclusions
Our findings further underline the disproportionately lower vaccination intention among Black/African American sexual minorities and other vulnerable groups. These findings highlight the need for targeted educational approaches for minority populations on the effectiveness of the vaccine and healthcare providers on the concerns and hesitations of sexual minorities, especially Black/African American sexual minority persons. Additionally, this study further highlights the crucial role of discrimination and anxiety/depression in the high likelihood of receiving the COVID-19 vaccine among both sexual minorities and heterosexual populations. These findings further support the need for the development of creative targeted approaches to understanding and addressing the sources and causes of anxiety/depression and discrimination among these populations to improve vaccination acceptance.