Minority stress and substance use: The role of anxiety/depression and PTSD symptoms in response to the COVID-19 pandemic

Abstract

Sexual minorities were surveyed about their experiences during the pandemic, and asked about symptoms of PTSD, minority stress, anxiety/depression, and substance use. We surveyed 392 sexual minorities who self-identified as substance users, including 70 of whom also identified as a gender minority. Participants completed questionnaires that included demographic questions, COVID-related PTSD symptoms [Post Traumatic Stress Checklist for the DSM-5 (PCL-5)], minority stress [Minority Stress Scale (MSS)], anxiety/depression [Patient Health Questionnaire − 4 (PHQ-4)], and substance use [adaption from the Alcohol Use Disorders Identification Test (AUDIT)]. A serial mediation model was used and we found an indirect relationship between minority stress and substance use through anxiety/depression and COVID-related PTSD such that minority stress positively predicted anxiety/depression, which in turn predicted COVID-related PTSD; and then predicted substance use in sexual minorities. Sexual minorities experienced greater health disparities during the COVID-19 pandemic. Sexual minorities high in minority stress experienced increased rates of psychiatric symptoms, making them more vulnerable to substance use. These results underscore the need for medical and mental health professionals to address the role of minority stress, and possible substance use and abuse as a method of dealing with psychiatric symptoms and stressors.

Keywords:

• COVID-19 pandemic
• sexual and gender minorities
• substance use
• post traumatic stress symptoms

1. Introduction

Sexual minorities experience unique stressors that impact their mental and physical health due to a lifetime of prejudice and discrimination, stigmatization, oppression, and even victimization on the basis of sexual minority status (Meyer, 2013). Given the unique stressors that sexual minorities experience (e.g., sexual prejudice, discrimination, stigma (Meyer, 2003), lack of secure employment and housing, health insurance and health care access (Charlton et al., 2018), rejection by family and/or community, discriminatory public policies (Drydakis, 2021), etc., the COVID-19 pandemic has impacted the psychological well-being of individuals, increasing the rates of mental health disorders and exacerbating those with preexisting conditions, especially in the lesbian, gay and bisexual (LGB) community (Fallahi et al., 2021). We hoped to better understand the experiences and the relationships of those who identify as sexual minorities and experienced increased rates of minority stress, anxiety and depression, COVID-related PTSD, and their risk for increases for substance use.

2. Minority stress

Meyer (2013) describes the stress associated with sexual minority status as minority stress, putting sexual minorities at risk for both distal and proximal stressors (Hatzenbuehler, 2009). Distal stressors can include violence, victimization, and discrimination, resulting in distress and negative mental health outcomes, including anxiety, depression, and substance use disorders. Proximal stressors involve self-stigma that results from negative societal opinions and includes internalized homophobia and identity concealment. Pachankis (2007) posits that sexual minorities who have a hidden stigma also face stressors that lead to significant psychological symptoms. As a result of these stressors, sexual minorities experience emotional dysregulation, interpersonal and social problems, and cognitive processes that create unique mental health and physical health disparities (Hatzenbuehler, 2009). These disparities include both internalizing and externalizing diagnoses, including depression, anxiety, and substance use in sexual minority adolescents and adults. Due to the complexity of the associations between minority stress and poorer mental health outcomes, Hatzenbuehler encourages researchers to investigate mediating variables that help to explain these relationships.

3. Sexual and gender minorities and adverse mental health symptoms

Minority stress has been associated with an increase in anxiety, depression and PTSD symptoms (Ching et al., 2022, Eldahan et al., 2016; Fulginiti et al., 2021; Griffin et al., 2018). Minority stress places individuals at increased risk for problematic substance use (Mereish et al., 2022). Increases in anxiety and depressive symptoms have been associated with increased rates of PTSD symptoms and can complicate recovery from a trauma (Fallahi et al., 2022; Livingston et al., 2020). Other populations (e.g., military personnel), have also found the relationship between anxiety and depression and an increase in PTSD (Hruby et al., 2021). Bonazza et al. (2022) found that if participants were diagnosed with a previous mood disorder, they were more likely to be diagnosed with PTSD during hospitalization for the first wave of the COVID-19 pandemic.

It is also important to examine intersectionality recognizing that persons may belong to more than one oppressed group (e.g., race, gender, and sexual orientation) where they experience even greater minority stress. Both anxiety and depression and PTSD symptoms put sexual and gender minorities at increased risk for substance use (Downing et al., 2020; Hinds et al., 2022; Medley et al., 2016). Wittgens et al. (2022) conducted a meta analysis examining sexual minority status and risk for anxiety, depression, alcohol use disorder and suicidality. They found higher risk for these psychiatric disorders with sexual minorities as compared to heterosexuals, with the highest risk seen for those who identify as female and bisexual. Similar results were found in a sample by Graham et al. (2011) who examined black sexual minority males who experienced discrimination and harassment. Over half of the variance for both anxiety and depression was explained by discrimination and harassment and internalized homonegativity. Borgogna et al. (2019) examined both sexual and gender minority college students, and found significant rates of depression and anxiety symptoms, including an interaction of gender and sexual identity. Those participants with both gender and sexual identity minority status had worse mental health outcomes as compared to those with only one minority status. Björkenstam et al. (2017) examined sexual orientation and self-reported symptoms of anxiety, and a history of depression for a sample in Sweden. They found that bisexual women and gay men reported higher rates of anxiety and higher rates of a previous diagnosis of depression as compared to heterosexuals.

Sexual and gender minorities disproportionately face traumatic events that may result in higher rates of Acute Stress Disorder (ASD) and Post Traumatic Stress Disorder (PTSD) (Lehavot & Simoni, 2011; Scheer et al., 2020). Minority stressors that include victimization and internalized homophobia place them at risk for negative health outcomes. Scheer et al. (2020) found that in a sample of sexual and gender minorities, many had a least one traumatic event in childhood that was associated with shame, leading to increased rates of both mental health and physical symptoms. Dworkin et al. (2018) emphasized that female sexual minorities have a greater risk for trauma, and therefore, a greater risk for PTSD. Further, they point out that the experiences of daily heterosexism, trauma-related conditions, or exposure to distal stressors, can either increase, maintain, or exacerbate PTSD symptoms. Because sexual minority women are more likely to be exposed to victimization, discrimination, and heterosexism, their rates of ASD and PTSD are significantly higher, as compared to heterosexual women.

4. Substance use

Substance use and minority stress has been found to be higher in LGB populations compared to heterosexual persons (Bränström, 2017; Livingston, 2017), and has been used as a way to cope with minority stress (Felner et al., 2019). S. E. McCabe et al. (2013) found that engaging in alcohol and substance use also increased the risk for minority stress, thus contributing to a negative cycle perpetuating the link between minority stress and substance use. C. J. McCabe et al. (2021) examined sexual minority status, social stress, emotional dysregulation and substance use and found that sexual minority women experienced higher levels of stress and emotional dysregulation, with higher levels of alcohol, cigarettes, and marijuana use within each age cohort.

Parent et al. (2019) found that sexual and gender minorities are often bullied, harassed, or assaulted with little to no help from others. Additionally, microaggressions aimed at sexual or gender minorities can bring on stress and feelings of endangerment. These stressors can eventually lead to an individual turning to substances as a way to cope. For example, Rhew et al. (2017) found that in those cases of sexual assault for sexual minority women was associated with significantly more drinking and alcohol-related consequences. Dworkin et al. (2021) argue that sexual minority women have higher rates of posttraumatic stress from elevated exposure to trauma, which in turn leads to increased alcohol use noting the self-medication model. In another study by Dworkin et al. (2017), they followed sexual minority women over a two year period to evaluate the relationship between PTSD and cannabis use and found that they were more at risk for a general use of cannabis, but not in response to the fluctuation of PTSD symptoms. In other words, the women in this sample had higher rates of regular cannabis use associated with higher rates of PTSD.

5. COVID-19 pandemic

Early on in the COVID-19 pandemic, researchers found that people were reporting increases in anxiety, depression, psychological distress, social isolation, and posttraumatic symptoms (Cao et al., 2020; Fallahi et al., 2021; Grech & Grech, 2020; Marazziti et al., 2020). The intersectionality with race and gender highlighted additional groups in experiencing even more significant symptoms. For example, Novacek et al. (2020) found that Black Americans were disproportionately affected by the pandemic with higher hospitalizations and mortality rates and anxiety and depression-related symptoms. Underserved communities reported a higher risk of developing psychological disorders as a result of the stress associated with the pandemic and the confinement measures utilized (Torres-Pagán & Terepka, 2020). Alonzi et al. (2020) reported that women and nonbinary individuals reported higher rates of depression and anxiety.

The research community is now examining the experiences of sexual minorities during the COVID-19 pandemic. Sexual minorities and other marginalized groups are showing disproportionate rates and exacerbation of mental health symptoms and disorders (Drabble & Eliason, 2021; Peterson et al., 2020), including depression, anxiety and trauma (Kneale & Bécares, 2020; Moore et al., 2021; Peterson et al., 2020). Harkness et al. (2021) found that Latinx sexual minority men experienced higher rates of anxiety, depression, alcohol and substance use, loneliness and sleep disturbance during the COVID-19 pandemic. Some of the risk factors include those sexual minorities who have histories of stressful childhood experiences (Filipčić et al., 2020), as well as family conflict (Fallahi et al., 2022).

We were interested in looking at the experiences of sexual minorities, and the relationship between minority stress and substance use, and seeing if anxiety/depression and experiencing trauma symptoms in response to COVID would mediate that relationship. We were interested in a serial mediation model that examines minority stress, anxiety and depression, PTSD symptoms, and substance use.

6. Hypotheses

1. We predicted in a serial mediation that there would be an indirect effect between minority stress and substance use through anxiety, depression, and PTSD symptoms during the COVID pandemic. Specifically, those participants who experience minority stress are more vulnerable to symptoms of anxiety and depression, and in turn, PTSD, resulting in increases in substance use. Those that are already anxious and depressed due to minority stress are more susceptible to develop COVID-related PTSD symptoms.

2. We also predicted that sexual, and gender minorities (who identify as a sexual minority) that also identify as racial and ethnic minorities would have higher substance use.

7. Method

7.1. Participants

Initially 500 participants were surveyed. The eligibility for participation included 18 years of age or older and identification as a sexual and/or gender minority. Of those 500 participants, 36 were removed for incomplete surveys, 20 were removed as they did not identify as a sexual or gender minority, and an additional 9 were removed as likely BOT completions (all responses were completed in less than 5 minutes). Our sample consisted of 435 participants who identified as sexual minorities (including gender minorities). Similar to other studies, we were interested in the intersection between sexual orientation and gender identity, which is why we included gender minorities who also identified as sexual minorities, and not just those who identified as a sexual minority (Fox et al., 2020; Scheer et al., 2022). Minority stress is specific to sexual orientation; we only included those sexual and gender minorities who identified as sexual minorities. In addition, those participants who had most of the survey incomplete were also dropped during data cleaning. We then deleted all of those participants where there was missing data on substance use. That brought us down to a sample of 350. The remaining participants sometimes skipped questions, and so sample size varies for the correlational analyses. For the serial mediation analyses, missing data was eliminated listwise (the PROCESS macro assumes complete data and will exclude cases that are missing on any of the variables in the model cases (Hayes, 2022)), ending with a sample of 275 (maintaining 80% of the data), thus for this analysis, there was no missing data. See Tables 1 and 2 for demographic and descriptive statistics. See Appendix I for the data transparency statement.

Table 1. Demographic Characteristics of Participants (N = 392)

Characteristics	n	%
Sexual Orientation
Bisexual	183	46.7
Gay/Lesbian/Same-gender attraction	100	25.5
Asexual	54	13.8
Pansexual	32	8.2
Queer	19	4.8
Gender
Cisgender Female	190	48.5
Cisgender Male	126	32.1
Genderqueer	15	3.8
Non-binary	18	4.6
Transgender MTF	15	3.8
Transgender FTM	13	3.3
Agender	4	1.0
Two-Spirit	5	1.3
Race
White (Caucasian)	275	70.2
Black (African/Caribbean American)	50	12.8
Asian or PacificIslander	16	4.1
Hispanic or Latino/a	14	3.6
American Indian orAlaskan Native	3	0.8
Arab or MiddleEastern	2	0.5
Multiracial or MixedRace	30	7.7
Household Income
Less than $25,000	61	15.6
$25,000–$50,000	102	26.0
$50,001–$75,000	87	22.2
$75,001–$100,000	64	16.3
More than $100,000	64	16.3

Note. Those who chose not to respond were removed from this table.

Table 2. Correlation of Study Variables

Variable	M	SD	N	1	2	3	4	5	6	7	8
1. SubUs	10.4	4.6	345	–
2. Alc	3.0	1.2	348	.404***	–
3. Halluc	2.3	1.5	350	.684***	.089	–
4. Stimul	1.7	1.2	349	.846***	.142**	.413***	–
5. Opiates	1.7	1.2	348	.860***	.130*	.413***	.826***	–
6. Depres	1.7	1.2	350	.852***	.167**	.429***	.728***	.799***	–
7. MSS	2.7	0.9	294	.699***	.215***	.389***	.661***	.654***	.619***	–
8. PHQ-4	5.2	3.4	350	.369***	.069	.331***	.291***	.319***	.321***	.397***	–
9. PCL-5	28.7	20.4	345	.648***	.221***	.437***	.560***	.568***	.567***	.705***	.714***

Note. SubUse = Substance Use; Alc = Alcohol Use; Halluc = Hallucinogen Use; Stimul = Stimulant Use; Depres = Depressants Use; MSS = Minority Stress Scale; PHQ-4 = The Patient Health Questionnaire − 4; PCL-5 = Posttraumatic Stress Disorder Checklist for DSM-5 in response to COVID.

*p < .05, **p < .01, ***p < .001

7.2. Measures

7.2.1. Posttraumatic stress disorder checklist for DSM-5 (PCL-5)

Blevins et al. (2015) Posttraumatic Stress Disorder Checklist for DSM-5 (PCL-5) 20-item scale was used to measure self-reported posttraumatic stress disorder symptoms and acute stress disorder symptoms modified for the COVID-19 pandemic. Participants answered each item on a 5-point Likert scale (1=Not al all to 5=Often). Participants were instructed to self-rate on each item. A typical item on this scale is, Since the pandemic, have you had more trouble sleeping— a high score on this item may indicate the participant is experiencing symptoms. A total symptom severity score was obtained by summing the scores for each of the items, with a suggested cutoff score between 31–33 to indicate the probability of PTSD. To obtain a provisional PTSD diagnosis, participants must also have a moderate or higher symptom endorsement in each of the DSM-5 clusters. The scale shows internal reliability (α=.963).

7.2.2. Minority Stress Scale (MSS)

Norcini Pala et al. (2017) Minority Stress Scale is a 42-item scale used to measure the following seven theoretical dimensions of minority stress: three items for Structural Stigma, three items for Enacted Stigma, 14 items for Expectations of Discrimination (3 of which were specific to Family Members), seven items for Sexual Orientation Concealment, six items for Internalized Homophobia Towards Others, six items for Internalized Homophobia Towards Oneself, and three items for Stigma Awareness. Expectations of Discrimination were split to assess family expectations from expectations in general. Participants answered each item on a 5-point Likert scale (1=Completely Disagree to 5=Completely Agree). A typical item on this scale is, Because of my sexual orientation, I will be discriminated against— a high score on this item may indicate the participant is high in Enacted Stigma. There was one reversed scored item on this scale, Because of my sexual orientation, society welcomes me— a high score in this item may indicate the participant is low in Expectations of Discrimination. The items of each subscale were averaged for a total score. The calculated internal reliability scores for the subscales from our study are as follows: Structural Stigma (α=.896), Enacted Stigma (α=.833), Expectations of Discrimination from Family Members (α=.896), Expectations of Discrimination, without Family Members items (α=.884), Sexual Orientation Concealment (α=.682), Internalized Homophobia Towards Others (α=.968), Internalized Homophobia Towards Oneself (α=.946), Stigma Awareness (α=.777). The scale shows internal reliability (α=.948).

7.2.3. Anxiety and depression: Patient health questionnaire − 4 (PHQ-4)

Kroenke et al. (2009) Patient Health Questionnaire-4 (PHQ-4) 4-item scale was used to measure anxiety and depression. The PHQ-4 was designed as a screening tool to indicate the need for further analysis with two items for depression and two for anxiety. Participants answered each item on a 4-point Likert scale (0=Not at all to 3= Nearly every day). Participants were instructed to answer based on the last two weeks. An example item includes, Feeling nervous, anxious or on edge— a low score may indicate that the participant is low in Anxiety. Items’ scores are totaled from 0–12, with psychological distress measured by: None (0–2), Mild (3–5), Moderate (6–8), Severe (9–12). On each subscale, a score of 3 or greater is considered positive for screening purposes, so a sum of each subscale was taken for each participant. The scale shows internal reliability (α=.881). The subscales are also internally reliable: Anxiety (α=.872), Depression (α=.836).

7.2.4. Substance use

The substance use measure was adapted from the Alcohol Use Disorders Identification Test (AUDIT; Saunders et al., 1993). We created a 5-item scale to measure Alcohol, Hallucinogens (e.g., marijuana, ecstasy, molly, etc), Stimulants (e.g., Crystal Meth speed, etc), Opiates (e.g., Oxycodone, OxyContin, heroin, fentanyl, etc), and Depressants (e.g., valium, Librium, etc). Participants were instructed to respond on each item with the instruction, How often do you use the following substances? using a 5-point Likert scale (0=Never to 4=4 or more times a week). An item on this scale is, Alcohol— a high score on this item may indicate the participant is high in Alcohol use. The items were summed for a total score. The scale shows internal reliability (α=.770).

7.3. Procedure

The survey was developed using SurveyMonkey (https://www.surveymonkey.com) and was approved by the Institutional Review Board at a Northeastern University (#20102). Settings were utilized through Survey Monkey that prevent IP or computer information of participants from being tracked or stored by the online site.

The sample was recruited in February, 2021, using Amazon’s Mechanical Turk (MTurk) because it enhanced our ability to increase numbers of LGB participants. The MTurk samples often yield a more representative sample than college students, especially with varying age, race, and ethnicity (Litman, n.d..). Financial compensation of $1.25 per survey as well as a chance for all completed survey participants to be entered in a drawing for a $50.00 Amazon gift card which allowed us to increase our sample size. The data collected yielded a large database where portions of the data was used in a previous study on sexual minorities during the pandemic who experienced family conflict yielding psychological distress and is published in the Journal of Homosexuality (Fallahi et al., 2022). Please see Appendix I for information about data transparency.

8. Data analyses

We first tested for correlations between main study variables and then conducted a serial mediation model. The serial mediation model was tested using Hayes (2022) PROCESS macro model 6. We investigated the indirect effect of minority stress on substance use through anxiety/depressive symptoms and post-traumatic disorder symptoms in response to the COVID-19 pandemic and the serial mediation of these two variables. Using the bias-corrected bootstrap method with 10,000 resamples we calculated the 95% confidence interval (CI). If the confidence interval does not include 0 then the indirect effect is significant.

9. Results

To test for multicollinearity, we examined the VIF in a multiple regression including all the variables including covariates. None of the relationships had a VIF over 5, most were between 1–2, therefore we did not standardize the variables before running the analyses.

10. Descriptive statistics and Pearson correlations

The descriptive statistics and bivariate correlations between the main study variables can be found in Table 2. Substance use (SU) was strongly, positive correlated with minority stress (MSS) (r =.70, p=.000), anxiety/depression (PHQ-4) (r =.37, p=.000), and the Post-traumatic Stress Checklist for DSM-5 in response to COVID (PCL-5) (r =.65, p=.000). We also found that MS was highly positively correlated with the PHQ-4 (r =.40, p=.000), and the PCL-5 (r =.71, p=.000). The PHQ-4 was highly positively correlated with the PCL-5 (r =.71, p=.000). We also ran correlations between the study variables and each type of substance use.

Alcohol use was positively correlated with the MSS (r = .215, p < .001), and the PCL-5 (r = .22, p < .001). Hallucinogen use was positive correlated with the MSS (r = .39, p < .001), the PHQ-4 (r = .33, p < .001), and the PCL-5 (r = .44, p < .001). Stimulant use was positive correlated with the MSS (r = .66, p < .001), the PHQ-4 (r = .29, p < .001), and the PCL-5 (r = .56, p < .001). Opiate use was positive correlated with the MSS (r = .65, p < .001), the PHQ-4 (r = .32, p < .001), and the PCL-5 (r = .57, p < .001). Depressant use was positive correlated with the MSS (r = .62, p < .001), the PHQ-4 (r = .32, p < .001), and the PCL-5 (r = .57, p < .001). Since all study variables were related to each form of substance use except the PHQ-4 and alcohol use, a total score of substance use was calculated by summing each of the types of substances. All further analyses used this total score.

11. Serial mediation analysis

Prior to running the serial mediation analyses we recoded several variables. Gender minority was recoded to 1=yes, 2=no; race was recoded to 1=white, 2=ethnic/racial minority; sex orientation was recoded to 1=bisexual, asexual, pansexual, queer, other, 2=gay lesbian/same gender attraction. Bisexual was coded together with asexual, pansexual, queer, and other because the categories were very small and could not be analyzed separately. Research has found that plurisexual (bisexual, queer, pansexual, etc) identified adults have been found to have higher levels of mental health symptoms as compared to gays and lesbians (Wilson et al., 2021). Similarly, race was categorized into 2 categories, white and racial/ethnic minority persons. One of the reasons we did this was because the categories among racial and ethnic minorities were too small to analyze separately.

To test our main hypothesis, we conducted a serial mediation model analysis using the PROCESS macro model 6 (Hayes, 2022). In this model age, race, sexual orientation, gender minority status, annual personal income, worry COVID in US will increase in the next year, worry you or a family member will get sick with COVID, personally diagnosed with COVID, family member diagnosed with COVID, family member died of COVID, lost job or furloughed during COVID, worry about finances and ability to support self/family, and worry about finances and ability to return to a normal lifestyle were included as covariates. Standardized coefficients are reported.

First in the serial mediation analysis, minority stress positively predicted anxiety/depressive symptoms (β=.30, p = .000). Anxiety/depressive symptoms were also related to race (ethnic minorities were more anxious and depressed than whites), (β = −.10, p < .05), sexual orientation (bisexual, asexual, pansexual, queer, other were more anxious and depressed than those who identified as lesbian or gay, single sex attraction) (β = −.12, p < .05), gender minority status (those who identified as a gender minority were less likely to report anxiety/depression than cisgender individuals), (β =.13, p < .05), worry about finances and supporting self/family during the COVID pandemic (β =.18, p < .05), and worry about finances and ability to return to a normal life after the COVID pandemic (β =.17, p < .05) explaining 38% of the variance in anxiety/depressive symptoms.

Secondly minority stress (β=.41, p = .000) and anxiety/depression positively predicted post-traumatic disorder symptoms (β=.46, p = .000). Worry about COVID increasing in the next year (β= .11, p < .05) and lost or furloughed from job due to COVID (β= −.10, p < .01) were also linked to post-traumatic stress disorder symptoms explaining 75% of the variance in post-traumatic symptoms.

Lastly, minority stress (β =.34, p=.000), and post-traumatic stress symptoms (β =.35, p=.000) were positively related to substance use. In addition, age (older individuals reported greater substance use than younger) (β =.12, p < .01), race (ethnic/racial minorities were less likely to use substances than whites) (β = −.13, p < .01), gender minority status (those who identified as a gender minority reported more substance use than cisgender individuals) (β = −.10, p < .05) and worry about you or family member will get sick with COVID (β= −.011, p < .05) all predicted substance use explaining 61% of the variance in substance use. See Table 3.

Table 3. The results of the regression analysis of substance use (N = 275)

Predictors	B	SE B	ß	t
MSS	1.82***	.32	.34	5.63
PHQ-4	−.01	.08	−.01	−.13
PCL-5	.08***	.02	.35	4.48
Age	.06**	.02	.12	2.80
Race	−1.57**	.49	−.13	−3.22
Sexual Orient	−.45	.43	−.04	−1.04
Gender Minority	−.89*	.42	−.10	−2.13
Personal Inc	.30	.18	.08	1.66
COV Worry	.06	.22	.02	.29
COV Fam	−.45*	.21	−.11	−2.13
COV Had	−1.15	.63	−.09	−1.84
COV FmHad	.46	.43	.05	1.06
COV Death	−.84	.64	−.07	−1.32
Personal Job	−.18	.43	−.02	−.41
Finance Sppt	.03	.20	.01	.14
Finance Rtrn	−.11	.20	−.03	−.55

Note. MSS = Minority Stress Scale; PHQ-4 = The Patient Health Questionnaire − 4; PCL-5 = Post traumatic Stress Disorder Checklist for DSM-5 in response to COVID; Sexual Orient = Sexual Orientation; Personal Inc = Personal Income; COV Worry (How worried increase in US COVID cases next 12 months); COV Fam (How worried you or a family member will get sick from COVID); COV Had (Personally diagnosed with COVID); FmHad (Family member diagnosed with COVID); COV Death (Family member died of COVID); Pers Job (Lost job or furloughed during COVID); Finance Sppt (How worried about finances and ability to support self/family); Finance Rtrn (How worried about finances and ability to return to a normal lifestyle). Race (1=white, 2=ethnic/racial minority); Sexual Orient (1= bisexual, asexual, pansexual, queer, other, 2= lesbian/gay/same sex attraction); Gender minority (1=yes 2=no)

*** p = .000, **p < .01, *p < .05

In Table 4 and Figure 1 you can find the indirect effects of anxiety/depressive symptoms and post-traumatic stress symptoms on substance use. The results indicate that both the total effect (β =.54, p=.000) and the direct effect (β =.34 p=.000) of minority stress on substance use were significant; however, the direct effect was less than the total effect, therefore any indirect relationships were partial mediators. Two significant indirect relationships out of three paths were significant. If the bootstrapping confidence interval includes 0 the indirect effect is not significant (Preacher & Hayes AF, 2008). The indirect effect of minority stress on substance use through anxiety/depressive symptoms was not significant (β= −.00, SE = .02, 95% CI= [−.03, .03]). The indirect effect of minority stress on substance use, through post-traumatic stress symptoms (β =.14, SE =.04, 95% CI= [.09, .21]) was significant. The indirect effect of minority stress on substance abuse through anxiety/depressive symptoms and post-traumatic symptoms, respectively (β =.05, SE =.02, 95% CI= [.03, .08]) was significant indicating serial mediation. Both anxiety/depressive symptoms and post-traumatic disorder symptoms in response to COVID partially mediated the relationship between minority stress and substance use. Minority stress positively predicted anxiety/depressive symptoms, in turn, anxiety/depressive symptoms positively predicted post-traumatic stress disorder symptoms and ultimately substance use.

Figure 1. Serial Mediation Model (N = 275).

Note. Standardized coefficients included. Dotted line is not a significant path. Solid lines are significant paths. Total effect = .53, Direct effect = .34

***p < .001

Table 4. Indirect Effect of Minority Stress on Substance Use via Anxiety/Depression Symptoms and PTSD Symptoms (N = 275)

Path	Indirect	Boot SE	95 % CI
	Effect		LL	UL
Minority Stress→Anx/Dep →SubUse	−.00	.02	−.03	.03
Minority Stress→PTSD →SubUse	.14	.04	.09	.21^a
Minority Stress→Anx/Depr Symps →PTSD→SubUse	.05	.02	.02	.08^a

Note. Standardized coefficients are presented. CI = bootstrapping confidence interval, LL lower limit CI, UL upper limit CI; ^a CI does not include zero. Anx/Dep= anxiety and depressive symptoms; PTSD = Post-traumatic stress disorder symptoms in response to COVID.

12. Discussion

Sexual minorities are at risk for minority stress because of exposure to prejudice and discrimination, stigma, oppression, and victimization due to their sexual minority status (Meyer, 2013). As predicted, we found that minority stress places sexual minorities more at risk for anxiety and depression, PTSD, and increased substance use (Eldahan et al., 2016; Ching et al., 2020; Griffin et al., 2018; Fulginiti et al., 2020; Mereish et al., 2022). In the serial mediation analysis, we found, as hypothesized, that there was an indirect relationship between minority stress and substance use through anxiety/depression and COVID-related PTSD symptoms. Those higher in minority stress were more vulnerable to anxiety/depression and in turn PTSD symptoms during the COVID-19 pandemic contributing to greater substance use. This supports past relationships found in the literature between PTSD symptoms and increased substance use (Downing et al., 2020; Hinds et al., 2022).

In examining intersectionality in our sample, recognizing that sexual minorities also may be from other oppressed communities (e.g., race, ethnicity, gender, etc.), we found that non-white sexual minorities reported less substance use as compared to white sexual minorities contrary to our second hypothesis. This is consistent in the literature in regard to racial/ethnic minorities and whites (Center for Behavioral Health Statistics and Quality, 2021). Further, those with plurisexual identities (e.g., bisexual minorities) have greater risk behaviors (DiPlacido & Fallahi, 2020), such as substance use as compared to gay or lesbian identities (Bränström, 2017; Verrastro et al., 2016) with the exception of those who identify as asexual (Hill et al., 2022). Further, gender minorities engaged in more substance use compared to cisgender minorities, findings that were similar to the previous work of Hendricks and Testa (2012) and Gonzalez et al. (2017). Interestingly, we found that those who identified as a gender minority were less likely to report anxiety/depression than cisgender individuals. It could be that trangender individual’s stress is more potent than their sexual minority stress.

Sexual minorities in the United States face a disproportionate burden of social inequities that intersects with mental health, minority stress, and substance use (Fish et al., 2021) creating an exacerbation in symptoms for sexual minorities. As a result, sexual minorities are more likely to show increased rates of mental health disorders, trauma, and substance use during the pandemic.

Among the LGBTQ+ community, substance use increases among minority stress risk factors. The highest risk factors linked to substance use include lack of support, victimization, internalized and externalized behavior problems, lack of or unsafe housing, and negative reactions to their disclosure (Goldbach et al., 2014). All of these risk factors contribute to higher minority stress which leads to an increase in substance use from alcohol and nicotine to marijuana and even illegal drugs, such as cocaine and ecstasy. Internalized homophobia, specifically, increases with a rise in minority stress which is associated with an uptick in substance use (Lehavot & Simoni, 2011). As internalized homophobia, discrimination, and the expectation of rejection changes over time, so does the use of substances (Hatzenbuehler et al., 2008). Similarly, sexual minorities who experience an increase in negative experiences throughout the day are more likely to binge drink at the end of that day, it should be noted that this association is also found among the general population (Livingston, 2017). However, minority stress has a stronger association with substance use than general stress does (Avery Desmarais et al., 2020).

Fish et al. (2021) argued that these findings require action as part of our disaster responsiveness during the pandemic. Outreach and education are needed for medical and mental health professionals on the unique needs of sexual minorities. Recognition, proper assessment, and access to treatment are needed, despite barriers that can include a lack of access because of the burden on the healthcare system due to the pandemic (Jemberie et al., 2020) or because of stigma and discrimination that may deter sexual minorities from seeking treatment (Biancarelli et al., 2019).

12.1. Future directions

Researchers are beginning to document the specific vulnerabilities among the sexual minority population during the COVID-19 pandemic and social support may be one of the key buffers between minority stress and substance use. As many were forced into isolation, they experienced the loss of their support system (Hafi & Uvais, 2020), the very thing that may help buffer sexual minorities from minority stress. Thoits (2011) found that perceived social support decreased the need for professional mental health treatment. For sexual minorities, the pandemic has impacted every part of their lives (Barrientos et al., 2021). With limited social support, discrimination is faced alone for gender and sexual minorities. Physical and psychological distress is higher in sexual minorities compared to heterosexuals (Moore et al., 2021); and the pandemic has exacerbated stressors experienced by sexual minorities (Oginni et al., 2021). During this time, perceived isolation has further exacerbated their mental health symptoms (Rodriguez-Seijas et al., 2020). A lack of community is one possible explanation, as community involvement is a predictor of decreased substance use (Avery Desmarais et al., 2020; Goldbach et al., 2015). Another explanation is that minority stress leads to difficulties in emotion regulation which leads to increased alcohol and substance use (Rogers et al., 2017). Substance use is clearly a risk factor in response to high sexual minority stress, especially during the COVID-19 pandemic. As our study has found, sexual minorities who are experiencing minority stress during the COVID-19 pandemic are at higher risk for psychiatric conditions and may be self-medicating to relieve their psychological symptoms.

Extending our model, we suggest the importance of studying social isolation as it is a frequently listed concern in the face of limited social support during the pandemic (Toze et al., 2021). However, those who have maintained virtual relationships are showing less distress during isolation. Some organizations have improved social support circles as a result of the pandemic which are helping improve well-being amongst the sexual and gender minority communities (Oginni et al., 2021). Sexual and gender minorities who perceived strong social support from friends and family showed less physical and mental health decline during the pandemic (Zhang et al., 2022). Social support is essential in the well-being of minority groups during the pandemic, especially from family (Barrientos et al., 2021; Zhang et al., 2022).

12.2. Limitations

Because of the cross-sectional nature of our sample, we can not determine the direction of the relationships in our serial mediation model. Ideally since there is an implied order of the variables in the model when testing mediation a longitudinal design would be the ideal methodology to determine the temporality of the relationships between variables. For example, it could be that substance use predicts minority stress other than what we hypothesized. We suggest the importance of studying the serial mediation model found in this study over time. Also, the fact that our study was administered on-line allowing for many people to view the invitation to participate and decide for themselves whether to take the survey, there are limitations to the generalizability of our results. It could be that those who were struggling with substance use and minority stress did not want to be exposed to a study looking at the psychological vulnerabilities during the COVID-10 pandemic contributing to substance use. There is also the limitation of self-report measures potentially subject to bias. Although our study did find support for the hypothesized indirect effect with these self report measures. Further, another possible limitation involves the use of the MTurk sample. When utilizing an MTurk sample, there is the potential that some of the people participating might also be participating in multiple studies, yielding the potential for a small sample (Litman, n.d.d.). While MTurk samples tend to be more diverse than traditional college student samples, they are still more likely to be homogeneous as the participants are often more educated and tend to be younger.

13. Conclusion

Sexual minorities experience stress because of their minority status resulting in negative outcomes that may include unique mental health and physical health disorders. Sexual minorities experience greater health disparities, especially in terms of substance use. This outcome was again highlighted during the COVID-19 pandemic utilizing a serial mediation analysis, where sexual minorities who were high in minority stress experienced increased rates of psychiatric symptoms, e.g. anxiety/depression, and resulting COVID-19 PTSD symptoms, making them more vulnerable to substance use.

14. Clinical significance

Sexual minorities experience greater health disparities, and this outcome was again highlighted during the COVID-19 pandemic. Sexual minorities who were high in minority stressors (including distal and proximal stressors), also experienced increased rates of psychiatric symptoms, e.g. anxiety/depression, and PTSD symptoms, making them more vulnerable to substance use as a possible coping strategy (Felner et al., 2019). This underscores the need for medical and mental health professionals to address the role of minority stress, psychiatric symptoms and disorders, and possible substance use and abuse as a method of dealing with those symptoms and stressors.

Data availability statement

The raw data required to reproduce the above findings are available to download from the Center for Open Science (OSF.IO). The processed data required to reproduce the above findings are available to download from https://osf.io/34r5j/.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

No funding was provided for this research.

Notes on contributors

Joanne DiPlacido

Joanne DiPlacido, Ph.D. is a professor in the Department of Psychological Science at Central Connecticut State University. She is Co-director of the Sexual and Gender Minority Stress and Health Lab. Her area of expertise is in Health Psychology. Her major research area is on minority stress and physical and mental health, especially among sexual and gender minorities.

Carolyn R. Fallahi

Carolyn R. Fallahi, Ph.D. is a professor and chair of the Department of Psychological Science at Central Connecticut State University. Co-director of the Sexual and Gender Minority Stress and Health Lab, she is committed to understanding the struggles of sexual and gender minorities face in today’s world. In addition, she is engaged in research that examines substance use disorders, post-traumatic stress, and the scholarship of teaching in psychology.

Carissa D. Daigle

Carissa Daigle has a bachelor’s degree in Psychological Science and is currently pursuing a master’s degree in Applied Clinical Psychology. Carissa currently lives in Pennsylvania with her husband, child, and family dog, where enjoys going for walks, reading, and exploring local attractions with her family.

Appendix I.

Data Transparency

The data reported in this manuscript was collected from a large sample of sexual minority participants utilizing participants collected from Amazon Mechanical Turk. A previous paper utilizing the same database was published in the Journal of Homosexuality examining family conflict as well as physical symptoms around sexual minority stress during the COVID-19 pandemic. Substance use, the focus of this study, was not included in the previous publication.