Contexts of social alcohol and cannabis use among sexual minority cisgender women and gender diverse individuals: Event-level differences in alcohol and cannabis use patterns based on the sexual orientations and gender identities of substance use companions

Abstract

Purpose:

Sexual minority women and gender diverse individuals assigned female at birth (SMWGD) are at elevated risk for alcohol and cannabis use disorders. It has been posited that characteristics of SMWGD’s substance use companions (i.e., sexual orientation, gender identity) may influence their own use, but few studies have tested this. The current study aimed to examine whether quantity and consequences of substance use varied based on sexual orientations and gender identities (SOGI) of SMWGD’s substance use companions.

Methods:

We utilized a 30-day ecological momentary assessment study of substance use among a sample of 429 SMWGD. We examined event-level associations between characteristics of substance use companions and quantity and consequences of substance use.

Results:

When SMWGD used alcohol/cannabis with most SOGI groups, they engaged in heavier drinking and cannabis use. Drinking with heterosexual men and SMW was associated with more drinking consequences. Drinking in settings with more SOGI groups and with both LGBTQ+ and heterosexual individuals was associated with heavier drinking and more consequences. Many contextual aspects of cannabis use settings predicted an increased likelihood of alcohol and cannabis co-use, while few contextual aspects of drinking settings were associated with co-use.

Conclusion:

Results suggest that SMWGD engage in heavier substance use when individuals from a range of SOGI groups are present, highlighting that heavy alcohol and cannabis use is not limited to use with SMW and nonbinary individuals. Drinking in setting with more SOGI groups present (e.g., parties and bars) appears to be associated with particularly heavy drinking and consequences.

Introduction

Alcohol and cannabis use disorders are more prevalent among sexual minority women (i.e., lesbian, bisexual, queer, and other non-heterosexual individuals women) compared to heterosexual women (Kerridge et al., 2017; Substance Abuse and Mental Health Services Administration, 2023). Further, sexual minority gender diverse individuals assigned female at birth (i.e., sexual minorities who identify outside of the gender binary) use alcohol and cannabis at levels similar to sexual minority cisgender women (Watson et al., 2020). Research has predominately focused on minority stress (i.e., chronic stress experienced by sexual and gender minorities due to the stigmatization of non-heterosexuality and gender diversity) as an explanation for these disparities (Dyar, Dworkin, et al., 2021; Dyar et al., 2022; Livingston et al., 2017; Wilson et al., 2016). However, researchers have posited that social norms and social contexts of substance use may also play a contributing role (Condit et al., 2011).

Social norms theory posits that substance use is influenced by descriptive norms, with individuals matching their use to what they think similar others use (Borsari & Carey, 2003). The more similar individuals perceive the group for whom norms apply (i.e., index group) to be to them, the more strongly they are influenced by the norm (Lewis & Neighbors, 2007; Litt et al., 2015). For example, college women are more strongly influenced by perceptions of what a “typical female student” drinks compared to a “typical student.” Based on social norms theory, sexual minority women (SMW) have been theorized to be more strongly influenced by drinking norms for SMW than drinking norms for heterosexual women (Borsari & Carey, 2003; Lewis & Neighbors, 2006; Litt et al., 2015). Consistent with this theory, Litt et al. (2015) found that SMW’s perceptions of other SMW’s drinking quantity predicting subsequent increases in their own drinking quantity, while perceptions of heterosexual women’s drinking did not.

Substance use is perceived to be more normative among SMW compared to heterosexual women (Hatzenbuehler et al., 2008; Litt et al., 2015). These less restrictive substance use norms and their high level of relevance to other SMW are theorized to encourage heavier substance use when SMW use substances with other SMW (Dworkin et al., 2018; Dyar, Feinstein, Newcomb, et al., 2021). There is a small but growing literature examining differences in SMW’s substance use based on the sexual orientations and gender identities of the individuals they use substances with (Dworkin et al., 2018; Dyar, Feinstein, Newcomb, et al., 2021). In a daily diary study, Dworkin et al. (2018) found that SMW drank more in mixed sexual orientation settings (i.e., both sexual minority and heterosexual individuals present) compared to settings with only heterosexuals. However, they found no differences between drinking with only sexual minorities compared to only heterosexuals. This provides partial support for the hypothesis that drinking with sexual minorities is associated with heavier consumption. In a semi-annual longitudinal study with sexual minority cisgender women and sexual minority gender diverse individuals assigned female at birth (SMWGD), Dyar, Feinstein, Newcomb, et al. (2021) demonstrated that using cannabis with SMW and nonbinary individuals was associated with subsequent increases in cannabis use disorder symptoms, but using cannabis with sexual minority men and heterosexuals was not. These associations did not differ for cisgender women compared to gender diverse participants, suggesting norms for both groups may be important for this broader population. These studies each have a major strength. Dworkin et al. (2018) used daily diary data, allowing for an understanding of how drinking companions may influence SMW’s drinking on a day to day basis. Dyar, Feinstein, Newcomb, et al. (2021) incorporated substance use companions’ genders in analyses. The current study aimed to combine the strengths of these two studies to examine whether sexual orientations and genders of substance use companions are associated with quantity and consequences of alcohol and cannabis use among SMWGD.

Same-Day Use of Alcohol and Cannabis

In addition to being at elevated risk for alcohol and cannabis use disorders, sexual minorities are also more likely to report using alcohol and cannabis on the same day compared to heterosexual individuals (Nguyen et al., 2021). This is concerning because same-day and simultaneous use of alcohol and cannabis (i.e., using alcohol and cannabis at the same time so that their effects overlap) is associated with more negative consequences (e.g., black outs, driving under the influence, and academic consequences) than using both substances separately (i.e., on different days or so their effects do not overlap; for a review see Lee et al., 2022). Studies with the general population indicate that simultaneous alcohol and cannabis use is more likely to occur when individuals use at a friend’s place, a party, or when more people are present (Gunn et al., 2021). Despite their elevated risk for simultaneous or same-day use of alcohol and cannabis, no existing studies have examined which contexts are associated with a higher likelihood of simultaneous or same-day use among sexual minorities. The current study aims to address this limitation by examining whether the sexual orientations and gender identities (SOGIs) of individuals present during social alcohol or cannabis use are associated with an increased likelihood of also using the other substance during the same observation. Further, we will examine whether the overall composition of one’s substance use companions (i.e., number of SOGI groups represented; presence of both LGBTQ+ and heterosexual individuals) is associated with a higher likelihood of simultaneous use.

Current Study

We used ecological momentary assessment (EMA) to examine how the identities of SMWGD’s substance use companions were associated with substance use outcomes in near-real time. First, we examined event-level associations between the SOGIs of substance use companions (i.e., SMW, sexual minority men, nonbinary individuals, heterosexual women, heterosexual men) and substance use outcomes (i.e., quantity of alcohol or cannabis; consequences). We hypothesized that when SMWGD used alcohol or cannabis with SMW and nonbinary individuals, they would consume more alcohol or cannabis and experience more consequences compared to using in contexts when these groups were not present. We expected that using alcohol or cannabis with sexual minority men and heterosexual individuals would not be significantly associated with substance use outcomes or would be associated with a lower quantity and fewer consequences.

The analyses described above allow for the examination of whether the presence or absence of alcohol/cannabis use companions from different SOGI groups are associated with substance use outcomes, but not how the overall composition of all substance use companions present might impact patterns of use. To examine this, we tested whether alcohol and cannabis use outcomes differed as a function of the number of SOGI groups present and whether the substance use setting included both LGBTQ+ and heterosexual individuals (i.e., mixed settings) or only LGBTQ+ or only heterosexual individuals. Given prior work demonstrating that drinking in mixed-orientation settings was associated with heavier drinking (Dworkin et al., 2018), we expected that using alcohol or cannabis in settings with more SOGI groups and using in mixed settings would be associated with heavier consumption and more consequences.

We also examined associations between contextual factors and the likelihood of same-day alcohol and cannabis use. We examined whether the likelihood of cannabis use differed depending on the presence or absence of individuals from each of our five SOGI groups in drinking contexts as well as whether the likelihood of drinking use differed depending on the presence or absence of individuals from each of our five SOGI groups in cannabis use contexts. We hypothesized that when SMWGD used alcohol or cannabis with SMW and nonbinary individuals, they would be more likely to also consume the other substance than when these groups were not present. We expected that using alcohol or cannabis with sexual minority men and heterosexual individuals would not be significantly associated with the likelihood of using the other substance. We also tested whether using one substance (e.g., alcohol) in contexts with more SOGI groups present or in settings that included both LGBTQ+ and heterosexual substance use companions would predict an increased likelihood of using the other substance (e.g., cannabis).

Finally, we explored whether associations differed for SMW and nonbinary individuals. We did not expect any significant differences based on a prior study (Dyar, Feinstein, Newcomb, et al., 2021).

Methods

Participants and Procedures

The current analyses used data from a longitudinal study of substance use among 429 SMWGD conducted between August 2020 and May 2021. Participants were recruited via online advertisements on social media. Eligible participants were 1) U.S. residents, 2) 18–25 years old, 3) assigned female at birth, 4) identified as women or under the nonbinary umbrella (e.g., nonbinary, genderqueer,), 5) identified as lesbian, bisexual, pansexual, or queer, and 6) met alcohol or cannabis use criteria (i.e., four or more drinks at least twice and/or using cannabis on at least three days in past month).¹ Participants who appeared eligible based on their responses to the eligibility survey were text messaged by study team members to verify their eligibility and their access to a mobile phone with text message capabilities. To verify their eligibility, participants were asked to text back demographic and contact information (i.e., age, state of residence, email address), which was cross-checked with their responses in the eligibility survey. Stratified sampling was used to ensure that the sample was diverse in race/ethnicity, gender identity, and sexual identity.

The study included a baseline assessment (day 0), a 30-day EMA study (days 1–30), and a follow-up assessment. This study uses data from the 30-day EMA study. During the EMA period, participants were asked to complete two surveys per day. Invitations for morning surveys were sent via email or text message (based on participant preference) at 8:00am and participants had until 1:00pm in their time zone to complete the survey. Invitations for evening surveys were sent at 6:00pm and participants had until 12:00am in their time zone to complete the survey. Participants who had not completed the survey by three hours after the survey invitation were sent a reminder. Those who missed more than three surveys in a row were contacted by study staff to check in and re-engage participants. Surveys were programmed in REDCap and took approximately two minutes to complete. Participants were paid up to $150: $20 for baseline, $20 for follow-up, $1 for each EMA survey, and a $5 bonus for every six EMA surveys completed in a row. The study received Institutional Review Board approval at Northwestern University (where data was collected) and Ohio State University (where data is stored) and participants provided informed consent. See Dyar et al. (2022) for additional study details.

The analytic sample included participants with at least one occasion of social alcohol or cannabis use during the EMA period (n=385). The sample was comprised predominately of people of color (34.5% non-Latinx White; Table 1), with a sizeable proportion of gender diverse participants (27.0%).

Table 1.

Demographics of Analytic Sample at Baseline (N = 385)

Demographic Variable	n	%
Sexual Identity
Lesbian	95	24.7%
Bisexual	101	26.2%
Pansexual	86	22.3%
Queer	103	26.8%
Race/Ethnicitya
White	218	56.6%
Black	83	21.6%
Latine	116	30.1%
Asian	50	13.0%
Other Race/Ethnicity	31	8.1%
Gender Identity
Cisgender Women	281	73.0%
Gender Diverse	104	27.0%
Substance Use Criteria Met
Alcohol Only	100	26.0%
Cannabis Only	83	21.6%
Alcohol and Cannabis	202	52.5%
Age (M, SD)	22.31 (1.98)

^aPercentages add up to more than 100% because participants could select multiple racial/ethnic identities.

Measures

Sexual orientation and gender identity of substance use companions

Sexual orientation and gender identity of substance use companions were measured during EMA surveys when participants indicated having used alcohol or cannabis with other people. Participants were asked “which of the following best describes those you were using marijuana with.” “Drinking” was substituted for “using marijuana” on days when alcohol was consumed with other people. Response options: LGBTQ+ women, LGBTQ+ men, nonbinary individuals, straight women, and straight men. Participants could select multiple responses. We also used these responses to create two additional variables: the number of sexual orientation by gender identity groups present and a binary variable indicating whether or not the setting included both LGBTQ+ and heterosexual individuals.

Duration of cannabis intoxication

Duration of cannabis intoxication was measured during EMA surveys when participants indicated having used cannabis. They were asked “How many hours were you high?” Responses were provided in increments of one hour from 0 to 11 hours and participants could indicate being high for 12+ hours. Duration of intoxication acts as a parsimonious proxy for cannabis quantity that functions across modes of use at the event-level (Calhoun et al., 2022).

Cannabis consequences

Cannabis consequences were measured only during morning assessments by using six items from two existing measures (Lee et al., 2021; Simons et al., 2012). Participants were asked “Which of the following things happened to you as a result of your marijuana use yesterday?” (e.g., “I said or did something that embarrassed me”) and could select multiple responses.

Number of drinks

Number of drinks consumed was assessed during surveys when participants indicated having consumed alcohol by asking participants “How many drinks did you have since the last survey?” Participants could indicate the specific number of drinks they consumed from 0 to 24 or indicate that they consumed 25+ drinks. Instructions in each EMA survey asked participants to consider their experiences since the start of the last EMA survey window (e.g., 8am or 6pm) if they missed their prior survey.

Alcohol consequences

Alcohol consequences were measured only during morning assessments using a five item version of a measure used in a previous EMA study of alcohol use among SMW (Dyar, Dworkin, et al., 2021). Participants were asked “Did any of the following things happen to you yesterday as a result of drinking?” (e.g., “I had a hangover”) and could select multiple responses.

Analytic Plan

Because the identities of alcohol and cannabis use companions were only available when alcohol or cannabis use occurred with other people, only those observations were included in analyses: 2,588 observations of social cannabis use from 267 participants and 2,571 observations of social drinking from 291 participants. The median completion rate was 83.3% and 60.1% of participants completed at least 75% of surveys. Within completed surveys, less than 1% of data were missing. Missing data were handled using Bayesian methods (Asparouhov & Muthén, 2010).

Analyses were conducted in Mplus version 8.9. Bayesian multilevel structural equation modeling (MSEM) with informative priors was used. MSEM utilizes latent variables, rather than person-mean centering, to separate within- and between-person variance, and evidence suggests that it outperforms traditional multilevel modeling techniques (Lüdtke et al., 2008). As a result of disaggregating within and between-person variance, event-level variables indicate the extent to which an individual was experiencing more/less of a construct than usual (above/below their person mean) during a particular assessment.

We examined three sets of models and two sets of outcomes. In the first set of models, we tested associations between using alcohol or cannabis use with five sexual orientation and gender (SOGI) groups (SMW, sexual minority men, nonbinary individuals, heterosexual women, and heterosexual men) and alcohol and cannabis outcomes (duration of cannabis intoxication, number of drinks, alcohol and cannabis consequences). All substance use companion groups were entered simultaneously and their associations with alcohol and cannabis use outcomes at the same assessment were examined. We also conducted sensitivity analyses to determine whether associations between using alcohol/cannabis with specific SOGI groups remained significant when we controlled for the number of SOGI groups present as well as moderation analyses in which the participants’ gender identity was a between-person moderator of the event-level associations between the identities of substance use companions and substance use outcomes described above.

In the second set of models, we examined whether the number of SOGI groups represented among the participants’ substance use companions predicted alcohol and cannabis outcomes. In the third, we examined whether there were differences in alcohol and cannabis outcomes during observations when both LGBTQ+ and heterosexual substance use companions were present compared to when only LGBTQ+ or only heterosexual substance use companions were present.

A second set of outcomes were also examined in relation to each of the proposed predictors described above. These outcomes are the likelihood of cannabis use on drinking days and the likelihood of drinking on cannabis use days. For these models, we examined whether one’s drinking context (i.e., SOGI of drinking companions; number of SOGI groups present when drinking, if the drinking setting included both LGBTQ+ and heterosexual individuals) predicted an increase likelihood of also using cannabis during that observation. Only individuals who reported at least some cannabis use in the past month at baseline were included in these analyses. We also conducted a parallel set of models examining whether one’s cannabis use context (i.e., SOGI of cannabis use companions; number of SOGI groups present when using cannabis, if the cannabis use setting included both LGBTQ+ and heterosexual individuals) predicted an increased likelihood of also drinking during that observation.

In all models, associations described were modeled at both within- and between-person levels. We controlled for day of assessment, weekend or weekday assessment, and morning or evening assessment. All event-level associations were allowed to vary across individuals. As alcohol and cannabis consequences were only assessed once per day, analyses that involved these variables used day-level composites of sexual orientations and gender identities of substance use companions. Age, sexual identity, gender identity, and race/ethnicity were included as covariates at the between-person level.

Results

See Table 2 for descriptive information about contexts of alcohol and cannabis use. Overall, 79% of alcohol use was social and drinking with SMW was most common (45%), followed by drinking with heterosexual women (37%) and heterosexual men (34%). Drinking with sexual minority men (16%) and nonbinary individuals (12%) was the least frequently endorsed. More than half (55.6%) of social alcohol use included companions of only one SOGI group and social drinking settings included both LGBTQ+ and heterosexual individuals 27.1% of the time. A majority (57%) of cannabis use was social. Cannabis use with SMW was the most common (34%), followed by using cannabis with heterosexual men (20%) and heterosexual women (16%). Using cannabis with sexual minority men (12%) and nonbinary individuals (12%) were the least frequently endorsed. Cannabis use was more likely to occur with individuals from only one SOGI group (72.5%) than with individuals from multiple groups and cannabis use settings included both LGBTQ+ and heterosexual individuals 16.1% of the time.

Table 2.

Descriptive Statistics for Contexts of Use

	M	SD	N reporting any use with this group
Alcohol Use With
Any Companions	.79	.29	291
Sexual Minority Women	.45	.37	279
Sexual Minority Men	.16	.27	159
Non-Binary Individuals	.12	.25	108
Heterosexual Women	.37	.33	271
Heterosexual Men	.34	.33	257
Multiple Groups Present	1.68	.92	-
Mixed Setting	.27	.44	-
Cannabis Use With
Any Companions	.57	.37	267
Sexual Minority Women	.34	.37	209
Sexual Minority Men	.12	.24	120
Non-Binary Individuals	.12	.26	97
Heterosexual Women	.16	.25	172
Heterosexual Men	.20	.29	172
Multiple Groups Present	1.38	.72	-
Mixed Setting	.16	.37	-

Means represent the average total proportion of observations which included cannabis use during which participants reported using alcohol or cannabis with companions of each gender and sexual orientation.

Event-level Associations with Substance Use Companions

When participants used cannabis with SMW, sexual minority men, nonbinary individuals, or heterosexual men, they remained intoxicated for longer than when they used cannabis in social contexts that did not include individuals from these groups (Table 3). When participants used in settings that included both LGBTQ+ and heterosexual individuals, they tended to be high for longer; however, the number of SOGI groups present among one’s substance use companions did not predict duration of intoxication. Cannabis consequences did not vary based on the sexual orientations and gender identities of other individuals present, the number of SOGI groups represented among substance use companions, or whether the setting included both LGBTQ+ and heterosexual individuals.

Table 3.

Within-person associations between social context and substance use outcomes

	Duration of Cannabis Intoxication		Cannabis Consequences		Number of Drinks		Alcohol Consequences		Drinking on Cannabis Use Days		Cannabis Use on Drinking Days
Predictor	b	95% CI	b	95% CI	b	95% CI	b	95% CI	b	95% CI	b	95% CI
Sexual Minority Women	.32	.02, .61	−.03	−.18, .12	.69	.44, .94	.18	.06, .29	.52	.21, .84	.34	.05, .64
Sexual Minority Men	.51	.16, .85	−.07	−.25, .10	.71	.34, 1.08	.11	−.04, .26	.54	.17, .91	.23	−.17, .63
Non-Binary Individuals	.58	.10, 1.07	.03	−.18, .24	.22	−.15, .57	.15	−.03, .32	.14	−.31, .58	.01	−.47, .46
Heterosexual Women	.11	−.17, .37	−.01	−.14, .12	.48	.25, .71	.06	−.05, .16	.44	.14, .75	−.31	−.59, −.04
Heterosexual Men	.39	.10, .68	.05	−.08, .19	.43	.21, .65	.13	.03, .24	.51	.21, .81	.12	−.15, .39
Number of SOGI Groups	−.15	−.33, .02	−.02	−.08, .04	.51	.40, .63	.14	.09, .19	.47	.30, .66	.08	−.07, .22
Mixed Setting	.44	.17, .62	.04	−.09, .16	.86	.63, 1.08	.26	.15, .36	1.02	.52, 1.56	−.03	−.29, .22

Within-person covariates included day of assessment and assessment type (i.e., weekend or weekday assessment and morning or evening assessment). Age, sexual identity, gender identity, and race/ethnicity were included as covariates at the between-person level. Bolding is used to highlight significant associations.

When participants drank with SMW, sexual minority men, heterosexual men, or heterosexual women, they consumed more alcohol than when they drank in social contexts that did not include individuals from these groups. Further, when participants drank with SMW or heterosexual men, they experienced more drinking consequences. When participants drank in settings where more SOGI groups were represented and in settings with both LGBTQ+ and heterosexual individuals, they consumed more alcohol and experienced more drinking consequences.

When participants used cannabis with SMW, sexual minority men, heterosexual women, and heterosexual men, they were more likely to also drink than when they used cannabis in social contexts that did not include individuals from these groups. Further, using cannabis in settings where more SOGI groups were represented and in settings with both LGBTQ+ and heterosexual individuals was associated with an increased likelihood of also drinking during the same observation. On the other hand, only one drinking context variable was associated with cannabis use. When participants drank with SMW, they were more likely to also consume cannabis during the same observation.

We conducted sensitivity analyses to determine whether associations between drinking/using cannabis with specific SOGI groups remained significantly associated with substance use outcomes when we controlled for the number of SOGI groups present. The pattern of results remained the same, suggesting that the number of SOGI groups present did not explain associations between specific SOGI groups and substance use outcomes.

Moderation by Participant Gender

To determine whether effects of using a substance with individuals from specific SOGI groups, quantity/consequences of use, and likelihood of co-use differed for SMW compared to gender diverse individuals, we tested for moderation by gender identity. Associations between using cannabis with heterosexual men and duration of intoxication (b = .82; 95% CI: .17, 1.48) as well as cannabis consequences (b = .33; 95% CI: .02, .64) differed for SMW and gender diverse individuals. Among SMW, using cannabis with heterosexual men was not significantly associated with duration of intoxication (b = .23; 95% CI: −.11, .56) or number of consequences (b = −.10; 95% CI: −.26,.06). Among gender diverse individuals, using cannabis with heterosexual men was associated with being intoxicated for longer (b = 1.06; 95% CI: .45, 1.65) and experiencing more consequences (b = .32; 95% CI: .02, .60).

Discussion

These results add to the few existing studies that have examined whether substance use differs based on the identities of the individuals SMWGD use substances with. Overall, results suggest that using alcohol and cannabis with most SOGI groups is associated with a longer duration of cannabis intoxication, heavier alcohol consumption, and a higher likelihood of drinking on cannabis use days. Alcohol consequences and the likelihood of cannabis use on drinking days were only associated with drinking with companions from a smaller subset of SOGI groups (i.e., SMW and/or heterosexual men). Findings also indicate that alcohol use and consequences were elevated on days when more SOGI groups were represented among an individual’s substance use companions as well as when both LGBTQ+ and heterosexual people were represented among individual’s substance use companions.

Contrary to our hypothesis that only using alcohol/cannabis with SMW and nonbinary individuals would be associated with heavier use, using alcohol/cannabis with individuals from most SOGI groups was associated with heavier use. Given the operationalization of presence/absence of specific SOGI groups and their simultaneous entry into models, this indicates that SMWGD engage in heavier alcohol/cannabis use when SMW, sexual minority men, and/or heterosexual men are present (compared to when they are not), controlling for the presence or absence of substance use companions from all other SOGI groups. Cannabis use was also heavier when SMWGD used cannabis with nonbinary individuals and when they drank with heterosexual women. One possible explanation for this pattern of findings is that it represents an additive effect of the number of groups present. In other words, it may indicate that SMWGD simply use substances more heavily when they are in more diverse settings where more people are likely to be present. However, when we controlled for the number of SOGI groups present in sensitivity analyses, the pattern of results persisted, suggesting that this is not the explanation. Another possibility is that using substances with each of these groups may be associated with heavier use, potentially for different reasons. Dworkin et al. (2018) posited that drinking with sexual and gender minority individuals may contribute to heavier drinking among SMW via the more permissive drinking norms among sexual and gender minority individuals, while drinking with heterosexual individuals may contribute to heavier drinking among SMW as result of experiences or expectations of minority stress, which are theorized to contribute to coping motives for drinking. Future research should explore whether these different processes and motives for drinking among individuals of different sexual orientations may explain why SMWGD drink more heavily in contexts with so many different SOGI groups.

In contrast to this overall pattern of heavier use with most SOGI groups, there were two outcomes that displayed more specific patterns. Drinking consequences were elevated when SMWGD drank with SMW or heterosexual men, and drinking with SMW was associated with a higher likelihood of also using cannabis during the same observation. The association between drinking with SMW and experiencing more consequences is consistent with hypotheses derived from social norms theory. It has been proposed that more permissive norms regarding substance use among SMW and the relevance of these norms to other SMW contribute to these theorized associations (Condit et al., 2011; Dyar, Feinstein, Newcomb, et al., 2021). However, it is important to note that this is one of the only findings that appears to provide support for this hypothesis in the current study and as such other potential explanations for this finding should also be considered. For example, do SMWGD tend to drink with SMW in settings that may contribute to a higher likelihood of experiencing consequences or are they less likely to engage in protective behavioral strategies when drinking with SMW? The association between drinking with heterosexual men and experiencing more alcohol consequences also has multiple potential explanations that should be explored by future research. Drawing from Dworkin et al. (2018) and the prevalence of sexual objectification of SMWGD by heterosexual men, particularly in contexts involving alcohol (Flanders et al., 2020; Flanders et al., 2021), one potential explanation is that drinking with heterosexual men may increase the likelihood of experiencing sexual harassment or anxiety about potentially experiencing sexual harassment. This may lead some SMWGD to drink to cope with the discomfort caused by experiences or expectations of sexual harassment. Of note, we do not have evidence to support this supposition in analyses, but if this finding is replicated, future research may explore this as a potential mechanistic pathway.

When SMWGD drank with SMW, they were more likely to also use cannabis during the same observation. One potential explanation for this may be more permissive norms around cannabis use or co-use among SMW. Alternatively, this pattern may also arise from systematic differences in the contexts in which SMWGD drink with SMW. For example, if drinking with SMW is more likely to occur at home or at a friend’s home, then these less public locations may increase the likelihood of using cannabis. Future research should explore factors that may help to explain this finding.

Only two associations between using substances with specific groups and substance use outcomes differed between cisgender women and gender diverse individuals. Using cannabis with heterosexual men was associated with heavier cannabis use and more consequences among gender diverse individuals but not among cisgender women. On average, heterosexual men have more negative attitudes toward transgender individuals broadly (Fisher et al., 2017; Kanamori & Cornelius-White, 2016, 2017; Kanamori et al., 2017), and nonbinary individuals specifically (Ocasio, 2022), than heterosexual women. Given the possibility of experiencing microaggressions from heterosexual men, gender diverse individuals may consume more cannabis as a coping strategy. Experiencing or expecting stigma based on two identities may explain why it was significant for gender diverse individuals but not cisgender women. Future research should explore ways in which motives and context may interact to predict different patterns of cannabis use.

Results also indicate that when more SOGI groups are represented among SMWGD’s substance use companions, they engaged in heavier drinking, experienced more drinking consequences, and were more likely to consume both alcohol and cannabis. The same outcomes were also elevated when SMWGD used substances in settings where both LGBTQ+ and heterosexual individuals were present. This suggests that the number of groups present and/or the diversity of the group may parallel other substance use context variable(s) that were not measured by the current study: the number of people present and drinking at a party. Both of these variables have been linked with heavier drinking (O’Donnell et al., 2019) and an increased likelihood of simultaneous alcohol and cannabis use (Lee et al., 2022). Future research should explore whether more diverse drinking settings are uniquely associated with heavier drinking when the presence of a party context and the number of people present are controlled for. This would help to determine if there is something unique about diverse settings for alcohol use among SMWGD or whether these findings are simply an artifact of its correlation with other contextual factors.

Clinical Implications

Our results suggest that SMWGD engage in heavier alcohol and cannabis use when they use substances with individuals from a range of SOGI groups. Given the high prevalence of alcohol and cannabis use disorders in this population, it is important to consider the potential clinical implications of these findings. Other studies have found that using cannabis in social settings and using in multiple different settings regularly is associated with heavier use and more alcohol and cannabis problems among SMWGD (Dyar & Feinstein, 2023; Dyar et al., 2023; Dyar, Feinstein, Crosby, et al., 2021). Together, these findings suggest that patterns of use across different settings may be particularly important to consider in this population. Clinicians working with SMWGD clients who wish to reduce their substance use or engage in harm reduction may work with their clients to determine whether social settings are associated with heavier use for them and if this generalizes broadly to all social settings or specific settings. Clinicians can then help their clients to develop strategies to reduce heavy use in these higher risk settings, such as engaging in protective behavioral strategies (e.g., setting a limit of the number of drinks or amount of cannabis consumed). Notable, while individual-level interventions are important for addressing the current needs of SMWGD, it is critically important for population and system-level interventions to be developed to reduce another known driver of heavy alcohol and cannabis use – minority stress.

Limitations

The current study’s findings should be considered in light of its limitations. First, SMWGD who used alcohol or cannabis regularly, lived in the US, and were between 18 and 25 were included in this study. As a result, it is unclear whether similar patterns will be found among other groups of sexual minorities. Second, the measure of substance use companions’ identities included transgender men and women under the broader categories of LGBTQ+ men and women. Future research may explore whether using substances with binary transgender individuals is associated with increased substance use among nonbinary individuals due to their shared transgender identities. Third, we did not assess the number of substance use companions present. Given that drinking with more companions is associated with heavier use (Kuntsche et al., 2015), future research should examine whether these findings hold when number of substance use companions is controlled for.

Conclusions

The current study adds to the nascent literature on the role of the sexual orientations and gender identities of substance use companions on substance use among SMWGD. Using alcohol or cannabis with most SOGI groups was associated with heavier alcohol and cannabis use, suggesting that heavier substance use among SMWGD is not limited to use with other SMW and nonbinary individuals. Future research should explore potential mechanisms and theories that may help explain why SMWGD drink and use cannabis more heavily with each of these SOGI groups. Results also indicated that drinking in settings where more SOGI groups were represented was also associated with heavier use and more drinking consequences. Future research should explore whether this is an artifact of associations between the number of SOGI groups represented and the number of individuals present or may indicate that drinking in diverse settings may be uniquely associated with heavier drinking.