Internalized Homophobia and Cigarette Use Among Latino and Black Sexual Minority Men: Protective and Risk Factors

Abstract

Latino and Black sexual minority men are at elevated risk of cigarette use compared to their heterosexual and White SMM counterparts. Internalized homophobia may affect substance use disparities. However, the research linking internalized homophobia and substance use has been inconsistent. The purpose of the present study was to clarify the association between internalized homophobia and daily cigarette use by testing the roles of internalized racism and ethnic identity acceptance as potential moderators of this link. This study collected data from 165 Black and/or Latino SMMs across the United States (M age = 23.72, SD = 3.85) as part of a larger study. Data were collected from December 2020 to February 2021 via Qualtrics Panels. Zero-inflated Poisson regression was conducted to examine the association between internalized homophobia and daily cigarette use and whether internalized racism and/or ethnic identity acceptance would moderate the association between internalized homophobia and daily cigarette use. Internalized homophobia was negatively associated with daily cigarette use; however, this association was significantly moderated by internalized racism and ethnic identity acceptance. Simple slope analyses revealed that low levels of internalized racism and high levels of ethnic identity acceptance attenuated the positive association between internalized homophobia and daily cigarette use. At low levels of ethnic identity acceptance, the positive association between internalized homophobia and daily cigarette use was strengthened. This research aids in contextualizing the association between internalized homophobia and daily cigarette use among Latino and Black SMM. Implications for smoking treatment and prevention programs are discussed.

Despite a decline in cigarette smoking in the United States since 1964, there are currently 30.8 million adult smokers in the country, 16 million of whom are presently living with smoking-related illnesses (Cornelius et al., 2022; Cummings & Proctor, 2014; U.S. Department of Health and Human Services, 2014). Cigarette smoking continues to be the leading cause of preventable illnesses and death among Americans, with estimates suggesting that smoking contributes to 480,000 deaths annually (Cornelius et al., 2022). Additionally, current smokers are at elevated risk of mental health concerns, such as anxiety and depression symptoms, compared to former smokers (Fluharty et al., 2016; Taylor et al., 2014). Individuals often initiate cigarette smoking due to peer pressure, coping with stress, social events, and social desirability (e.g., ‘looking cool’; Freedman et al., 2012). While cigarette use is high among all Americans, some communities are disproportionally impacted by cigarette smoking, such as Latino and Black SMM (Cornelius et al., 2022; King et al., 2022; Kim et al., 2024).

Sexual minority individuals are at heightened risk of cigarette smoking compared to their heterosexual counterparts (Cornelius et al., 2020; King et al., 2022). Smoking is associated with masculine gender role expression (Kodriati et al., 2018), and sexual minority men (SMM) may initiate smoking to cope with discrimination and prejudice and to appear more traditionally masculine (Pachankis et al., 2011; Rivera et al., 2024). Furthermore, SMM may also engage in smoking because it is a popular activity in social circles and bars (Li et al., 2024). Notably, one in six sexual minority individuals are current smokers, compared to one in eight heterosexual adults (Cornelius et al., 2022). Additionally, one study found that sexual minority Latino adults were more likely to smoke compared to their heterosexual Hispanic/Latino adult counterparts (King et al., 2022). Additionally, individuals with multiple marginalized identities (e.g., sexual, racial, and ethnic minority individuals) may be at an even higher risk of smoking due to the potential exposure to discrimination and the internalization of those experiences (e.g., internalized homophobia, internalized racism). Kim et al. (2024) found that among sexual minority adults, Latinos were more likely to smoke compared to their White and Black peers; however, Black adults did not differ from White individuals. Thus, it is important to identify whether salient sociocultural factors (e.g., ethnic identity acceptance and internalized racism) serve as protective factors or exacerbate the effects of internalized homophobia on cigarette use.

The minority stress model (Brooks, 1981; Meyer, 2003; Rich et al., 2020) posits that sexual minority individuals are at heightened risk of adverse psychological and behavioral conditions due to the compounded effect of proximal and distal stressors present in their environments. Proximal stressors include internalized homophobia, concealment, and expectation of rejection. Distal stressors include microaggressions, heterosexist discrimination, and victimization. The minority stress framework has been applied broadly and has guided research on minoritized individuals’ alcohol use and smoking (Lehavot & Simoni, 2011), binge drinking (Pittmann et al., 2017), cannabis and illicit drug use (Feinstein & Newcomb, 2016; Gonzalez et al., 2017), appearance and performance enhancing drug misuse (Convertino et al., 2021), disordered eating (Calzo et al., 2017), HIV sexual transmission risk behaviors (Dentato et al., 2013), and various other adverse psychological and behavioral outcomes. One component of the minority stress model that has received particular attention in the literature is internalized homophobia, or the belief that negative stereotypes about sexual minority individuals are true and applicable to oneself.

Internalized Homophobia and Cigarette Use

Across studies sampling SMM, the association between internalized homophobia and cigarette use, and substance use more generally, is inconsistent. Internalized homophobia has previously been found to be significantly positively correlated with cigarette use among SMM and other sexual minority populations (D’Avanzo et al., 2016; Gamarel et al., 2015; Lehavot & Simoni, 2011). In an examination of HIV-serodiscordant sexual minority male couples, higher levels of internalized homophobia were associated with greater likelihood of self-reported smoking (Gamarel et al., 2015). One potential explanation for these findings is that when sexual minority individuals internalize societal stigma, they may experience guilt when they attempt to pursue their same-sex desires, which may subsequently influence their use of substances, such as nicotine, to regulate negative affect (Huynh et al., 2022; Khantzian, 1985, 1997). However, other studies have suggested that there is a negative or null association between internalized homophobia and cigarette use. For example, Amadio and Chung (2004) found a non-significant association between internalized homophobia and cigarette use among SMM, and Holloway et al. (2012) found that higher levels of internalized homophobia were associated with lower use of cigarettes among SMM. Taken together, these findings suggest a need for further research on the association between internalized homophobia and cigarette use. Examining different moderators relating to attitudes towards social phenomena may provide insight into the subtleties and complications of these inconsistent findings.

Internalized Racism and Intersectionality

One potential moderator of the association between internalized homophobia and cigarette use is internalized racism. Internalized racism refers to the acceptance of racial stereotypes of an individual’s racial or ethnic group and felt pressure to adapt to White cultural norms (David et al., 2019). Internalized racism may be experienced alongside and in conjunction with other forms of internalized oppression, as posited through Kimberly Crenshaw’s theory of intersectionality (Crenshaw, 1991). Rooted initially in Black feminist theory and critical race theory, intersectionality theory posits that every individual sits at multiple axes of privilege and oppression, which are interconnected and influence one another. Inherent to an individual’s social location are elements of power and inequity, and these elements are fluid and dynamic, having the potential to change at any time (Crenshaw, 1991; Else-Quest et al., 2022; Else-Quest, 2023; Else-Quest & Hyde, 2016). In addition to foundational qualitative research, the intersectional approach in psychology and public health research has used quantitative methods (e.g., additive effects, multiplicative effects; Cole, 2009) to explore and generate specific and testable hypotheses and deepen the field’s understanding of how these social categories are interconnected and contribute to adverse health outcomes. For example, an intersectional approach to quantitative scholarship may reveal additive effects, defined as the cumulative impact of the interplay between multiple minoritized characteristics, which interact with one another and accumulate. Each social category may contribute to cumulative effects that may amplify or deepen inequality (Else-Quest & Hyde, 2016; Turan et al., 2019). Given that both experiencing race-based prejudice and experiencing sexuality-based prejudice are associated with adverse mental health and behavioral outcomes (English et al., 2018; Layland et al., 2022), it is possible that the internalization of these multiple forms of oppression may result in additive effects, uniquely heightening risk for such outcomes, including cigarette use.

While research on the consequences of intersectional internalized stigma is still in its infancy, there is preliminary evidence to support the notion that interactive effects of racism and homophobia may lead to an increase in substance use among Latino and Black SMM. Notably, Layland et al. (2022) examined the association of intersectional stigma subgroups with alcohol and marijuana use among Black and Latino SMM. Results showed that Black and Latino SMM who experienced both frequent racist and heterosexist stigma were at increased odds of drinking and marijuana use disorder symptoms, compared to subgroups who experienced a single form of stigma or low to no stigma. Given that Layland et al. (2022) did not examine cigarette use or the inconsistent links between internalized homophobia and cigarette use, examining the potential role of internalized racism on the link between internalized homophobia and daily cigarette use is warranted.

Ethnic Identity Acceptance and Cultural Resilience

Other potential moderators of the association between internalized homophobia and daily cigarette use include identity characteristics that contribute to cultural resilience. Cultural resilience refers to how an individual’s cultural background influences how they cope and overcome social adversity (Clauss-Ehlers, 2010). Cultural resilience mechanisms have been shown to improve overall social well-being and serve as protective factors to mitigate the association between stressors and mental health (Thorpe et al., 2022). One identity characteristic that may contribute to cultural resilience is ethnic identity. Ethnic identity often consists of an individual’s sense of belongingness to an ethnic group, which is defined by cultural heritage (e.g., values, traditions, language). Ethnic identity is a multidimensional construct that often consists of exploration (i.e., understanding one’s ethnic group), resolution (i.e., commitment regarding one’s ethnic group), and affirmation (i.e., ethnic identity pride; Phinney & Ong, 2007; Umaña-Taylor et al., 2004). These different dimensions of ethnic identity are associated with a variety of adverse mental and physical health outcomes. In a recent meta-analysis of 51 studies examining the association between discrimination and health outcomes (e.g., depression, sleep disturbance), the negative association between discrimination and health outcomes was moderated by different elements of racial and ethnic identity. While ethnic/racial identity commitment and a composite ethnic/racial identity variable served as buffers in the association between discrimination and health outcomes, ethnic/racial identity exploration exacerbated the negative association between discrimination and health outcomes, specifically at higher levels of exploration (Yip et al., 2019). Furthermore, higher levels of total composite score of ethnic identity were negatively associated with cigarette smoking, and a strong ethnic identity is often noted as a protective factor for other forms of substance use such as alcohol and marijuana use (Nguyen et al., 2012; Pugh & Bry, 2007). In a recent study of Black SMM, high levels of total composite score of ethnic identity buffered the association between experiences of heterosexist microaggressions and cannabis use, as well as substance use overall (Kalinowski et al., 2022). Given past literature showing elements of ethnic identity acceptance as a protective factor, an examination of the construct’s role on the relation between internalized homophobia and daily cigarette use is warranted.

The Current Study

The current study aimed to enhance clarity surrounding the association between internalized homophobia and cigarette use by exploring the effect of potential moderators (i.e., internalized racism and ethnic identity acceptance) that may strengthen or attenuate the association. Specifically, we hypothesized that, in a sample of Latino and Black SMM:

H1) The association between internalized homophobia and daily cigarette use would be positive and significant.

H2) The association between internalized homophobia and daily cigarette use would be significantly moderated by internalized racism such that at higher levels of internalized racism, the positive association between internalized homophobia and daily cigarette use would be strengthened.

H3) The association between internalized homophobia and daily cigarette use would be significantly moderated by ethnic identity acceptance such that at higher levels of ethnic identity acceptance, the positive association between internalized homophobia and daily cigarette use would be buffered.

Method

Participants

Participants were part of a larger study comprising 647 participants. After reducing the full sample to exclusively Latino and Black cisgender SMM, the final sample consisted of 165 participants recruited across the West (N = 30), Midwest (N = 28), Northeast (N = 30), and South (N = 77) of the U.S. between 18–30 years of age (M age = 23.72, SD = 3.85). All participants identified as SMM, with most participants identifying as gay (37.0%) or bisexual (58.8%). Participants were not required to have smoked a cigarette to be included in the study. Demographics for study participants are presented in Table 1.

Table 1.

Sociodemographic Characteristics of Participants (N = 165)

Characteristics	N (%)
Age
18–22	79 (47.9)
23–26	40 (24.2)
27–30	46 (27.9
Education
Less than high school	6 (3.7)
High school or GED	47 (28.0)
Some college	51 (30.5)
College degree (AA/BS/BA)	39 (23.8)
Some graduate work	8 (4.9)
Graduate/professional degree	15 (9.1)
Employment status (all that apply)
Full time (30 hours or more)	81 (49.1)
Part time (less than 30 hours)	25 (15.2)
Unemployed	24 (14.5)
Disabled	7 (4.2)
Full-time student	31 (18.8)
Part time student	10 (6.1)
Income
Less than $6,000	31 (18.9)
$6,000–$11,999	18 (11.0)
$12,000–$17,999	17 (10.4)
$18,000–$23,999	16 (9.8)
$24,000–$29,999	29 (17.7)
$30,000–59,999	30 (18.3)
$60,000 or more	23 (14.0)
U.S region
West	30 (16.96)
Midwest	28 (18.18)
Northeast	30 (46.66)
South	77 (18.18)
Have you had unstable housing in the past 6 months?
No	122 (73.9)
Yes	43 (26.1)
Race/Ethnicity
Hispanic/Latino	88 (53.3)
Non-Hispanic Black	77 (46.7)
Sexual orientation
Bisexual	97 (58.8)
Gay	61(37.0)
Other (e.g., queer, pan)	7 (4.2)

Procedure

Procedures were approved by the Institutional Review Board at San Diego State University. As part of a larger study of emotional dysregulation among heterosexual and SMM participants were recruited through Qualtrics Panels. Qualtrics Panels is a service provided through Qualtrics in which the platform advertises a survey through targeted mailing lists and social media. Eligible participants were given an email invitation to participate in a one-time, anonymous/de-identified online questionnaire. Participants were required to read the informed consent document on Qualtrics to continue the survey. All measures regarding sexual orientation, racial identity, ethnic identity acceptance, and substance use were presented in a counterbalanced order and took approximately 30–50 minutes to complete. Participants who completed the majority of items (see validity check items section below) were compensated with $3.50 Qualtrics e-rewards currency, which can be redeemed in the form of gift cards.

Measures

Sociodemographic.

Participants reported demographic information such as age, race, ethnicity, education, employment status, income, sexual orientation, and housing insecurity. To assess race, participants were asked “What is your race?” with the following response options: 1) White, 2) Black or African American, 3) Native American or American Indian, 4) Asian/Pacific Islander, and 5) Other (please specify). To assess ethnicity, participants were asked, “What is your ethnicity?” with the following response options: 1) Hispanic or Latino and 2) Not Hispanic or Latino. To assess education, participants were asked “What is the highest grade you reached in school?” with the following response options: 1) less than high school, 2) high school or GED, 3) some college, 4) college degree (Associate’s or Bachelor’s), 5) some graduate work (no degree to date) and 6) Graduate/professional degree. To assess employment status, participants were asked to select all that applied for the following options: 1) employed full-time (30 or more hours per week), 2) employed part-time (less than 30 hours per week), 3) unemployed, 4) disabled, 5) retired, 6) student. To assess income, participants were asked “what is your annual income as an individual before taxes?” with the following options: 1) less than $6,000, 2) $6,000-$11,999, 3) $12,000-$17,999, 4) $18,000-$23,999, 5) $24,000-$29,999, 6) $30,000-$59,999, and 7) $60,000 or more. To assess sexual orientation, participants were asked “how do you identify your sexual orientation?” with the following options: 1) Lesbian/gay, 2) bisexual, 3) heterosexual, 4) asexual, and 5) other (please specify). To assess housing insecurity, participants were asked “Have you had unstable in the past 6 months?” with a “yes” or “no” response option.

Daily Cigarette Use.

Participants were asked one self-report free-response item regarding their frequency of cigarette use: “How many cigarettes do you smoke per day?” Participants responded with a numerical value. This single-item measure for daily cigarette use is a reliable and valid way to measure the frequency of cigarette use, and a large body of literature has used this single-item measure as an outcome variable (Heatherton et al., 1989; Khaled et al., 2011; Lawrence et al., 2007; Rivera et al., 2024; Rozzell et al., 2020); however, respondents may be subjected to potential digit bias (i.e., participants may tend to round responses to the nearest approximate number that ends in zero or five; Blank et al., 2017; Gariti et al., 1998; Lewis-Esquerre et al., 2005; Shiffman, 2010).

Internalized Homophobias.

Internalized homophobia was assessed using the Internalized Homophobia Scale (Herek et al., 1998), a unidimensional 9-item self-report scale used to measure internalized negative attitudes and beliefs about SMM. Items were rated on a 5-point Likert scale ranging from 1 (Strongly disagree) to 5 (Strongly agree). Items are summed, with higher scores reflecting greater internalized homophobia. Example items include “I often feel it best to avoid personal or social involvement with other gay/bisexual males” and “I have tried to stop being attracted to males in general.” In the original study, Herek et al. (1998) reported strong internal consistency (α = .83) for men in his sample of sexual minority individuals. Similar studies have reported Cronbach’s alpha values greater or equal to .70 (Lewis et al., 2003; Meyer, 1995). Internalized homophobia scores have been positively correlated with increased depression symptoms, lower self-esteem, and increased anxiety symptoms (Herek et al., 1998; Igartua et al., 2009; Newcomb & Mustanski, 2010). In the current study, internal consistency was high, α = .91.

Internalized Racism.

Internalized racism was assessed using the Appropriated Racial Oppression Scale (Campón & Carter, 2015), which is a 4-factor, 24-item scale that is used to measure attitudes, beliefs, and emotional responses to internalized racism among people of color. For the purposes of this study, we used the 8-item Devaluation of Own Group subscale (Campón & Carter, 2015). This subscale was selected as it assessed how individuals judge their own race based on White cultural standards, leading to internalized negative beliefs and discrimination towards their own race. This scale has been validated in young sexual minority Black men (Wade et al., 2022). Participants were asked to use a 7-point Likert scale ranging from one (Strongly Disagree) to seven (Strongly Agree) to answer questions such as: “Because of my race, I feel useless at times,” “I wish I were not a member of my race,” and “Whenever I think a lot about being a member of my racial group, I feel depressed” (Campón & Carter, 2015). The subscale items were summed, with higher scores indicating elevated levels of internalized racism. Previous research has found significant positive correlations between internalized racism and poor self-esteem, depression, anxiety, denial of white privilege, unawareness of institutional racism, and denial of blatant racial discrimination (Campón & Carter, 2015; Roberson & Pieterse, 2021). Campón and Carter (2015) and Roberson and Pieterse (2021) reported Cronbach’s alpha values of .87 and .86, respectively. In the current study, α = .92.

Ethnic Identity Acceptance.

Ethnic identity acceptance was assessed using Multi-Ethnic Identity Measure Revised scale (Phinney & Ong, 2007), which is a 6-item scale that measures an individual’s exploration and commitment to their racial or ethnic group. For the purposes of this study, the total score was used to assess the overall strength of ethnic identity acceptance. Previous studies have validated the use of the total score of the current and previous versions of the Multi-Ethnic Identity Measure scale among Black SMM and women populations (Kalinowski et al., 2022; Thorpe et al., 2022). Participants use a 5-point Likert scale ranging from one (Strongly Disagree) to five (Strongly Agree) to answer questions such as: “I have a strong sense of belonging to my own ethnic group,” “I understand pretty well what my ethnic group membership means to me,” and “I feel a strong attachment towards my own ethnic group” (Phinney & Ong, 2007). The items were averaged to analyze the overall strength of ethnic identity acceptance, with higher mean scores showing stronger attachment to one’s racial or ethnic group. Previous studies have reported Cronbach’s alpha values of .80 and above (Brown et al., 2014; Phinney & Ong, 2007; Thorpe et al., 2022). In the current study, α = .92.

Validity Check Items.

Participants were presented three validity items. Example items included: “For this item, please select strongly disagree.” Qualtrics’s internal data quality flags and removes low-quality participants (Belliveau & Yakovenko, 2022). Participants were excluded if two or more of the validity checks were not answered correctly.

Planned Analysis

SPSS 28.0 and R 4.4.2 were used for analyses in this study. Descriptive statistics were computed for demographic information and means for all variables. Bivariate correlations were calculated between main study variables. Pairwise deletion was used to handle missing data. The current study conducted a zero-inflated Poisson regression analysis to examine the associations between internalized homophobia and the count outcome of daily cigarette use moderated by internalized racism and ethnic identity acceptance (in one model). This type of analysis was selected given overdispersion of excess zeros, which is often caused by structural and sampling factors. Zero-inflated Poisson models are common in nicotine and tobacco research, and the value of zero could be due to sampling factors (e.g., a smoker attempting to quit reporting zero cigarette) or structural factors (e.g., cigarette counts among nonsmokers; Pittman et al., 2022). Since the presence of zeros could be due to both factors, the regression generates two models: a logit model that predicts excess zeros and a count model that predicts the number of daily cigarettes the respondents reported. The unstandardized regression coefficients for both models were exponentiated to yield an incident rate ratio (IRR). All analyses were set with a predetermined α = 0.05. Simple slope analyses were conducted to probe the effects of internalized homophobia on daily cigarette use at one standard deviation above and below the mean for internalized racism and ethnic identity acceptance.

Results

Descriptive Statistics

Regarding income, 67.7% had an annual income, before taxes, of $29,999 or less. Regarding educational level, 3.7% had less than high school, 28.0% had graduated high school or had a GED, 30.5% had some college-level education, 23.8% had graduated with a college degree, 4.9% had some graduate degree work, and 9.1% had graduate/professional degrees. Regarding employment (all that apply), 49.1% had a full-time job (30 hours or more), 15.2% had a part-time job (less than 30 hours), 14.5% were unemployed, 4.2% reported disability, 18.8% were a full-time student, and 6.1% were part-time student. Regarding house insecurity, 26.1% reported having housing insecurity within the last 6 months. See Table 1 for the sociodemographic characteristics for participants. In total, 164 participants completed Devaluation of Own Group subscale of internalized racism with a mean score of 19.6 (out of possible 56.0). All 165 participants filled out the measures of internalized homophobia, ethnic identity acceptance, and daily cigarette use scales, and the mean scores for those scales were 21.4 (out of a possible 45.0), 3.6 (out of a possible 5.0), and 2.0 (out of a possible 23), respectively. See Table 2 for means, standard deviations, and intercorrelations between main study variables.

Table 2.

Means, Standard Deviations, and Intercorrelations Between Main Study Variables

Variables	M	SD	1	2	3	4
1. Internalized Racism	19.60	12.29	—	—	—	—
2. Internalized Homophobia	21.35	9.25	.56**	—	—	—
3. Ethnic Identity Acceptance	3.63	1.07	−.06	−.04	—	—
4. Daily Cigarette Use	1.98	4.24	.16*	.09	−.15	—

Note.

^**= p <.01

^*= .05

Preliminary Analysis

Upon initial examination of the dataset, one extreme outlier for daily cigarette use (100 daily cigarettes) was identified. Values with a SD of +/−3.33 from the mean were considered outliers. Therefore, the single value of 100 daily cigarettes was winsorized, which is a method that minimizes the influence of outliers in the dataset by transforming the value so it is closer to other values in the set (i.e., the outlier value of 100 was transformed to the next highest ordinal value that was not an outlier). A bivariate correlation analysis was conducted first to determine the intercorrelations between main study variables. Internalized racism and internalized homophobia were significantly intercorrelated (r = .56, p = .01). Additionally, internalized racism and daily cigarette use were significantly intercorrelated (r = .16, p = .05). Correlations between other study variables did not achieve statistical significance. Data is considered skewed when skewness values are below or above the range of −2 to +2, or when kurtosis values are below or above −7 to 7 (Byrne, 2010). Skew and kurtosis did not exceed these values for the continuous variables (internalized racism, internalized racism, and ethnic identity acceptance). Daily cigarette use was a count variable with a range of 0 to 23.

Moderation Analyses

Contrary to our hypotheses, internalized homophobia was inversely associated with daily cigarette use such that higher levels of internalized homophobia were associated with lower daily cigarette use, B = −.04, [95% CI: −.06, −.02], IRR = .96, [95% CI: .94, .98], SE < 0.01, p < .001. Internalized racism was positively associated with daily cigarette use such that higher levels of internalized racism were associated with higher daily cigarette use, B = .02, [95% CI: .01, .04], IRR = 1.02, [95% CI: 1.01, 1.04], SE < 0.01, p < .001. Lastly, ethnic identity acceptance was inversely associated with daily cigarette use such that higher levels of ethnic identity acceptance were associated with lower daily cigarette use, B = −.25, [95% CI: −.36, −.14], IRR = .78, [95% CI: .69, .87], SE < 0.01, p < .001. See Table 3 for the count model of main variables and interactions.

Table 3.

Count Model of Main Variables and Interactions

	Estimate	SE	z-score	95% CI	p-value	IRR	95% CI
Intercept	1.68	0.08	21.17	1.53–1.84	<.001	5.37	4.61–6.29
IHP	−0.04	0.01	−3.81	−0.06–−0.02	<.001	0.96	0.94–0.98
IR	0.02	0.01	3.03	0.01–0.04	<.001	1.02	1.01–1.04
EI	−0.25	0.06	−4.48	−0.36–−0.14	<.001	0.78	0.69–0.87
IHP:IR	0.002	0.001	3.15	0.001–0.002	<.001	1.00	1.00–1.0002
IHP:EI	−0.07	0.01	−6.98	−0.09–−0.05	<.001	0.93	0.91–0.95

Note. IHP = Internalized Homophobia; IR = Internalized Racism; EI = Ethnic Identity; IRR = Incident Rate Ratio.

Internalized Racism

The negative association between internalized homophobia and daily cigarette use was significantly moderated by internalized racism, B = .002, [95% CI: .0006, .0026], IRR = 1.002, [95% CI: 1.006, 1.0027], SE < 0.01, p = .002. Among individuals with high internalized racism (one SD above the mean), internalized homophobia was not significantly associated with daily cigarette use, B = −.0021, [95% CI: −.045, .003], IRR = .979, [95% CI: .955, 1.003], SE <0.01, p = .086. Among individuals with low internalized racism (one SD below the mean), the negative association between internalized homophobia and daily cigarette use was strengthened, B = −.061, [95% CI: −.086, −.036], IRR = .940, [95% CI: .917, .964], SE <.001, p < .01. See Table 3 for the count model of main variables and interactions. Internalized racism was not a significant moderator in the logit model. See Table 4 for the logit model of the main variables and interactions.

Table 4.

Logit Model of Main Variables and Interactions

	Estimate	SE	z-score	95% CI	p-value	IRR	95% CI
Intercept	0.83	0.21	3.92	0.41–1.24	<.001	2.29	1.50–3.46
IHP	−0.05	0.03	−1.67	−0.10–0.008	0.09	0.95	0.90–1.00
IR	−0.02	0.02	−0.76	−0.06–0.02	0.45	0.98	0.94–1.02
EI	−0.33	0.19	−1.72	−0.71–0.47	0.09	0.71	0.49–1.60
IHP:IR	−0.001	0.002	−0.328	−0.003–0.002	0.74	0.99	0.99–1.00
IHP:EI	−0.03	0.02	−1.221	−0.08–0.02	0.22	0.97	0.92–1.02

Note. IHP = Internalized Homophobia; IR = Internalized Racism; EI = Ethnic Identity; IRR = Incident Rate Ratio.

Ethnic Identity Acceptance

The association between internalized homophobia and daily cigarette use was also statistically significantly moderated by ethnic identity acceptance, B = −.068, [95% CI: −.086, −.048], IRR = .934, [95% CI: .918, .953], SE < 0.01, p < .01. Among individuals with high levels of ethnic identity acceptance (one SD above the mean), the negative association between internalized homophobia and daily cigarette use was strengthened, B = −.113, [95% CI: −.141, −.085], IRR = .893, [95% CI: .868, .918], SE < 0.01, p < .01. Among individuals with low levels of ethnic identity acceptance (one SD below the mean), internalized homophobia was positively associated with daily cigarette use, B = .030, [95% CI: .0002, .061], IRR = 1.03, [95% CI: 1.0002, 1.062], SE < 0.01, p = .047. See Table 3 for the count model of main variables and interactions. Ethnic identity acceptance was not a significant moderator in the logit model. See Table 4 for the logit model of the main variables and interactions.

Discussion

Cigarette use is associated with various adverse health outcomes, and Latino and Black SMM are at elevated risk of daily cigarette use. To our knowledge, this is one of the first studies to examine potential risk (i.e., internalized racism) and protective (i.e., ethnic identity acceptance) factors that may qualify the association between internalized homophobia and daily cigarette use among Latino and Black SMM. The current study found that internalized homophobia was negatively associated with daily cigarette use. This was not consistent with our original hypothesis. However, the association between internalized homophobia and daily cigarette use has been inconsistent in past literature (D’Avanzo et al., 2016; Gamarel et al., 2015; Holloway et al., 2012; Lehavot & Simoni, 2011). Further, the findings from our moderation analyses illuminate and help to clarify the complicated association between these constructs, and the ways in which risk and protective factors strengthen and attenuate this relation.

The association between internalized homophobia and daily cigarette use was significantly moderated by internalized racism. Minority stress theory does not explicitly indicate how multiple minoritized identity characteristics may intersect and compound upon one another. As such, a minority stress perspective assuming additive effects of minority stressors might suggest that experiencing stressors related to multiple minority statuses (i.e., internalized racism and internalized homophobia) would consistently increase individuals’ risk for experiencing adverse health behaviors and outcomes. However, our intersectional approach – one focusing on the complicated interrelations and interwoven impacts of multiple forms of oppression -- revealed more complicated relations between these constructs in this study. Despite the negative main effect of internalized homophobia on daily cigarette use, high levels of internalized racism did not exacerbate the association between internalized homophobia and daily cigarette use. These findings suggest that the effect was not multiplicative and is inconsistent with previous literature (Layland et al., 2022). In contrast, low levels of racism did strengthen the negative association between internalized homophobia and daily cigarette use, suggesting that low levels of internalized racism served as a protective factor and decreased daily cigarette use, in the context of internalized homophobia. Latino and Black SMM may benefit from working to uproot internalized racism to experience the benefits of minimized internalized racism. Relatedly, clinicians may wish to consider examining ways to reduce internalized stigma in at least one domain (e.g., internalized racism) to reduce the risk of cigarette use, which may be higher for individuals with higher levels of internalized racism.

The findings also showed that the association between internalized homophobia and daily cigarette use was significantly moderated by ethnic identity acceptance. Internalized homophobia was negatively associated with daily cigarette use at high levels of ethnic identity acceptance, suggesting a potentially protective effect of high levels of ethnic identity acceptance. However, internalized homophobia was positively associated with daily cigarette use at low levels of ethnic identity acceptance, fully flipping the negative association. Consistent with our findings, Kalinowski et al. (2022) found that stronger ethnic identity acceptance buffered the association between heterosexist microaggressions and marijuana and overall substance use. Black and Latino SMM’s commitment to their respective racial and ethnic communities may help them focus on the positive experiences of their community to alleviate experiences of distal and proximal minority stressors, resulting in decreased cigarette use. On the contrary, Latino and Black SMM’s exploration of their identity may result in exposure to homophobic messaging and stigmatization of sexual minority people (MacCarthy et al., 2021). These findings demonstrate that the combination of high levels of internalized homophobia and low levels of ethnic identity acceptance may be detrimental to one’s health.

Previous culturally-tailored smoking cessation group interventions, social-media campaigns, and individual counseling have been moderately effective in encouraging sexual minority individuals to reduce smoking behaviors (Baskerville et al., 2017). However, qualitative research has revealed that sexual minority individuals who participated in group interventions desired more inclusiveness, specifically for people of color, and the incorporation of empowering messages in targeted social media campaigns (Baskerville et al., 2018). Future research may wish to consider incorporating Latino and Black sexual minority people in prevention and intervention research studies to parse differences between racial groups. It may be beneficial to explore how culturally-based identity factors (e.g., traditional machismo; Arciniega et al., 2008) are associated with smoking, as SMM are stereotyped as being more feminine than their heterosexual counterparts (Blashill & Powlishta, 2009). A previous study found that SMM who were more likely to conform to masculine expectations reported increased cigarette use, suggesting that these men may use smoking as a tool to enhance perceived masculinity to avoid possible victimization (Pachankis et al., 2011). More smoking cessation interventions are needed to address sociocultural factors and minority stressors that may be impacting cigarette use among Latino and Black sexual minority individuals.

Limitations and Future Directions

These findings should be viewed in light of this study’s limitations. First, the design was cross-sectional, which limits the assertions of temporal ordering and causality. Research should analyze longitudinal data to clarify temporal ordering of the associations identified in this work. Second, the sample’s age was restricted to 18–30 years old. This may make it difficult to generalize our findings to older Latino and Black SMM. Older Latino and Black SMM may have a different experience with smoking given the social (e.g., less social acceptance) and political (e.g., lack of legal protections) environment they developed in compared to younger Latino and Black SMM (Gordon et al., 2021). SMM are often disproportionately targeted by tobacco companies via marketing campaigns (e.g., smoking being part of the sexual minority community) and price discounts, especially among young, low-income, and people of color (Acosta-Deprez et al., 2021). Another limitation of our study is that cigarette use is limited in measuring only frequency, and the item assessed how many cigarettes individuals smoke per day with an undermined timeframe, which can lead to potential digit bias. The measurement does not incorporate other smoking behaviors (e.g., cravings). Future studies should examine other tobacco and nicotine products, such as cigars, hookah, and e-cigarettes/vaping, given the high prevalence in the LGBT+ community. The current study’s sample consisted of cisgender men, and researchers may want to incorporate transgender men and nonbinary individuals, who also experience high rates of smoking (Cornelius et al., 2022; U.S. Department of Health and Human Services, 2014). Our findings are limited to Black and/or Latino participants in the study, combined, and do not incorporate other racial/ethnic minority groups (e.g., Asian America Pacific Islanders). Additionally, our sampling approach (i.e., Qualtrics Panel) was not explicitly nationally representative in nature. As such, the generalizability of these findings to individuals who do not participate in such panels is unknown. As such, researchers should examine whether there are differences between Black and Latino SMM in terms of their cigarette use, internalized homophobia, and ethnic identity acceptance considering the differences between cultures. Future research should explore whether our findings may be replicated in other racial and ethnic groups, and sexual minority women of color. Additionally, researchers may want to explore other forms of internalized oppression (e.g., internalized sexism, internalized transphobia) among women, transgender and gender diverse people, and how they interact with race and sexual orientation on substance use. Finally, future research should explore other forms of identity/cultural experiences (e.g., spirituality) to test whether these constructs can serve as moderators to buffer the relation between internalized homophobia and substance use among Latino and Black SMM. Overall, as researchers explore protective factors on the association between minority stress and substance use, public health efforts can focus on expanding the message of ethnic identity acceptance and facilitating substance use prevention efforts and interventions that are more inclusive for people with multiple identities, including Latino and Black SMM.

Clinical Impact Statement.

Some evidence suggests that Latino and Black sexual minority men are disproportionately impacted by cigarette smoking. Internalized stigmas, such as internalized homophobia and racism, may increase the risk of cigarette smoking, whereas ethnic identity acceptance may decrease the risk of cigarette use. This present study shows evidence that different levels of ethnic identity acceptance and internalized racism play a protective role in the association of internalized homophobia and daily cigarette use.