A meta‐analysis of sexual orientation inequities in substance use among youth
Lavender Lab, Department of Psychology, University of Maryland, College Park, MD, USA
Department of Health and Exercise Science, University of Oklahoma, Norman, OK, USA
Washington University in St. Louis, St. Louis, MO, USA
Harvard University, Cambridge, MA, USA
Recovery Research Institute, Harvard Medical School, Boston, MA, USA
Center for Addiction Medicine, Massachusetts General Hospital, Boston, MA, USA
Abstract
Background and Aims
Sexual minority youth report significantly higher rates of substance use than heterosexual youth, yet a comprehensive and systematic evaluation and synthesis of the magnitude of this inequity has not been conducted. The purpose of this study was to conduct a meta‐analysis to assess the magnitude of overall inequities in substance use between sexual minority and heterosexual youth and examine demographic and methodological moderating factors that might impact variability in substance use patterns.
Methods
A meta‐analysis of studies that examined sexual orientation inequities in substance use. We conducted a comprehensive literature search across four electronic databases (PubMed, APA PsycINFO, Web of Science, ProQuest; January 2008–July 2024). Studies were eligible if they included a youth participant population (mean sample age of 25 years or younger), examined differences in substance use between sexual minority and heterosexual groups, and published between 2008 and 2024 in English. Primary outcomes were any measures of substance use; secondary outcome was age of substance use initiation. Following PRISMA guidelines, reviewers independently reviewed and extracted data. Analyses employed random‐effects models, with robust variance estimation to account for dependency among multiple effect sizes within studies. Analyses examined continuous and dichotomous outcomes separately. Bivariate meta‐regression examined moderators.
Results
Among 304 studies of 5 928 282 youth, sexual minority youth reported greater quantity and frequency of all assessed substances (i.e. alcohol, nicotine, cannabis, prescription drugs, powder cocaine, crack cocaine, meth/amphetamine, 3,4‐methylenedioxymethamphetamine, heroin, and other substances) and engaged in more polysubstance use than heterosexual youth. For continuous outcomes, Hedges' g ranged from 0.10 (95% confidence interval [CI], 0.05–0.15, I2 = 99.52, participants = 354 201, studies = 48) for alcohol to 0.40 (95% CI, 0.21–0.59, I2 = 100.00, participants = 30 679, studies = 5) for mixed/polysubstance use. Dichotomous outcomes showed consistently elevated odds ratios, ranging from 1.34 (95% CI: 1.24–1.46, I2 = 96.01, participants = 3 500 203, studies = 128) for alcohol to 4.63 (95% CI, 2.91–7.38, I2 = 86.35, participants = 391 827, studies = 10) for heroin use. Sexual minority youth also had earlier ages of initiation (all substance outcomes: odds ratio, 1.45; 95% CI, 1.04–2.03, I2 = 95.39, participants = 619 187, studies = 11). Moderation results indicated that inequities were larger for plurisexual youth (e.g. bisexual, pansexual), sexual minority girls and young women, and adolescents 18 years of age or younger. The magnitude of inequities was also larger for lifetime measures of use compared with measures of recent use.
Conclusions
Sexual minority youth — particularly those who are plurisexual, sexual minority girls and young women, and adolescents 18 years of age or younger — appear to engage in substance use at higher rates than heterosexual youth.
Untitled section
Keywords: adolescents, disparities, inequities, meta‐analysis, sexual minority, substance use, youth
Article notes
Untitled section
Received 2025 Jun 24; Accepted 2025 Nov 21; Issue date 2026 May.
INTRODUCTION
Hazardous substance use among youth (i.e. adolescents and young adults up to 25 years of age) is a serious public health issue that has a myriad of negative effects on health in adolescence and into adulthood [1, 2, 3, 4]. Sexual minority youth (SMY) (i.e. youth who identify as lesbian, gay, bisexual, queer, and/or youth who have romantic/sexual attractions to or engage in romantic/sexual behaviors with the same‐sex or same‐gender and/or multiple sexes or genders) in particular report significantly higher rates of substance use compared to heterosexual youth [5, 6]. SMY use substances at younger ages, use more frequently and at greater quantities and have increased risk for developing substance use disorders (SUDs) than heterosexual youth in adolescence and into adulthood [5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15]. Disparities in substance use, and particularly hazardous use, are considered inequities as they are rooted in intersecting systems of oppression (e.g. heterosexism, anti‐bisexual stigma, cissexism and racism), and hazardous use is a preventable outcome of socially unjust systems [6]. Consistent with motivational models of substance use and the oppression framework, SMY may engage in substance use for a variety of reasons (e.g. coping, social, enhancement and conformity), and some reflect their coping with and resistance of oppression‐based stressors (e.g. find community, feel liberated), lack of social safety and socially unjust conditions [5, 6, 16].
Substance use prevalence among SMY vary substantially by socio‐demographic and measurement factors [15, 17, 18, 19, 20, 21, 22]. Plurisexual youth (i.e. youth who are attracted to more than one sex or gender; e.g. bisexual and pansexual) and sexual minority girls and young women have the highest rates of substance use [5, 6, 15, 20]. Additionally, there is limited, mixed evidence that indicates that questioning youth (i.e. youth who are questioning or unsure of their sexual orientation) have higher rates of substance use than their heterosexual peers [23, 24, 25]. Moreover, sexual orientation is a multi‐dimensional construct, encompassing identity (e.g. lesbian), attraction (e.g. romantic or sexual attractions to a similar sex or gender or multiple sexes or genders) and/or behavioral (e.g. sexual behaviors with a similar sex or gender or multiple sexes or genders) dimensions [26]. Inequities in substance use vary based on the dimension(s) of sexual orientation measured [21].
Despite well‐documented sexual orientation inequities in substance use among youth, there has been only one systematic review and meta‐analysis of this literature [13]. This study is outdated and limited as it was published in 2008 and only included 18 studies. There has been a significant proliferation of research on sexual orientation inequities over the past 17 years, as these inequities have been acknowledged as federal public health priorities in some countries [e.g. United States (US), Australia], and there has been an increase in the measurement of sexual orientation in research [26, 27, 28]. An updated and precise understanding of sexual orientation inequities in substance use is warranted. Therefore, the purpose of this pre‐registered systematic review and meta‐analysis was to assess the magnitude of overall inequities in substance use and age of initiation between SMY and heterosexual youth, examine the variability in substance use inequities, and identify demographic and methodological factors that might influence this variability.
METHODS
This meta‐analysis is reported in line with the Preferred Reporting Items for Systematic Reviews and Meta‐analyses (PRISMA) reporting guidelines [29]. The protocol, analysis plan and coding manual were pre‐registered on the Open Science Framework (https://osf.io/9r68z). Although we followed the pre‐registration, additional analyses were conducted in response to the peer review process.
Data sources and search strategy
We conducted a comprehensive literature search across four electronic databases (PubMed, APA PsycINFO, Web of Science and ProQuest) from January 2008 through July 2024. The search began in 2008 because the last meta‐analysis of sexual orientation disparities among youth was published at that time [13]. Search terms were developed through consultation with a university librarian and based on previous meta‐analyses of substance use among sexual minority populations [13, 16, 20, 30, 31, 32]. Complete database search strategies are provided in the Appendix S1 of the Supporting information.
Eligibility criteria and screening
Studies were eligible if they met the following inclusion criteria: (1) included a participant sample of youth with a mean age of 25 years or younger; (2) assessed substance use and examined differences in substance use between heterosexual and sexual minority groups (defined broadly as any sexual orientation dimension based on identity, attraction or behaviors); (3) published between January 2008 and July 2024; and (4) written in English. All included studies used observational designs, and no experimental studies met our inclusion criteria. Our search included both peer‐reviewed journal articles and gray literature (i.e. dissertations, theses). If data from the same sample were published in a peer‐reviewed article and dissertation, we retained only the peer‐reviewed article. We excluded duplicates, reviews and manuscripts without accessible full text or relevant effect size information identified. As we did not have the resources to search for gray literature through other types of website searches, conference abstracts, white papers and book chapters were not included. Reviewers independently screened titles and abstracts of all identified records, followed by full‐text reviews of potentially eligible studies.
Outcome
The primary outcomes of interest encompassed any measure of substance use, including alcohol, nicotine, cannabis, prescription drugs, cocaine, methamphetamine/amphetamine (meth/amphetamine), 3,4‐methylenedioxymethamphetamine (MDMA) (e.g. molly), crack, heroin, other substance use (e.g. inhalants, sedatives and hallucinogens) or mixed/polysubstance use. Quantity measures captured how much substance was used (e.g. number of drinks per day, cigarettes smoked per day) and included binary prevalence measures (any use vs. no use). Frequency measures captured how often substances were used over time (e.g. number of days used per month, percentage of use days over past 30 days). The secondary outcome was age of substance use initiation, which referred to the age at which participants first used a substance.
Data extraction
Using a standardized coding form, each study was extracted by two independent reviewers. Extracted information included: study (e.g. sampling strategy, inclusion/exclusion criteria), sample demographic (e.g. age, sex, gender, race/ethnicity and sexual orientation), outcome (e.g. type of substance use, duration of use) characteristics and statistical data needed for effect size calculation (e.g. sample sizes, means). Coders were trained by the first author, who is a doctoral‐level researcher with expertise in SMY's substance use, and coders met weekly to discuss the coding process. Discrepancies were resolved through discussion and consensus. We contacted corresponding authors of studies with unreported or partially reported effect size data for eligible outcomes.
Effect size estimation
For continuous outcomes, we first calculated Cohen's d and then transformed them to Hedges' g, which applies small study bias correction, to measure standardized mean differences in substance use between SMY and heterosexual youth. Positive values indicate higher substance use among SMY than heterosexual youth. Effect sizes are interpreted within the context of substance use research rather than using generic interpretation guidelines, with consideration given to the magnitude of differences observed and their practical significance. For dichotomous outcomes, we calculated and synthesized log odds ratios (ORs) (exponentiated after the synthesis for interpretation) to quantify the relative likelihood of substance use between SMY and heterosexual youth (i.e. OR >1 indicates higher likelihood for SMY; OR <1 indicates lower likelihood for SMY). Following established recommendations [33, 34], we winsorized effect sizes that were three or more standard deviations from the mean to minimize the disproportionate influence of extreme outliers on pooled estimates.
Statistical analysis
Analyses used random‐effects models using the robumeta command (which conducts meta‐analysis with robust variance estimation) for STATA/MP 16.0 [35], with robust variance estimation (RVE) to account for dependency among multiple effect sizes within studies. Effect sizes were calculated and analyzed separately for each substance type (e.g. alcohol, nicotine and cannabis) to allow for substance‐specific comparisons. We analyzed continuous and dichotomous outcomes separately because they require different effect size metrics representing their underlying distributions (Hedges' g vs. log OR). Although transformations are available to put these in the same type of metric (e.g. transform g to OR), this is not appropriate given their different distributional properties. Random effects inverse variance‐weighted mean effect sizes and 95% CI were calculated using an assumed average within‐study correlation between effects of 0.70 based on current RVE guidelines [35]. For samples that appeared in multiple studies, we included the larger sample size when there was complete overlap, while samples with less than 50% overlap were treated as distinct. For the analyses with too few studies/effects for RVE (i.e. when k < 40 studies and on average <5 effect sizes per study) [35], we used the metan command (which performs standard meta‐analysis for independent effect sizes). In these cases, rather than averaging within‐study effects, we selected appropriate effect sizes to synthesize across studies and only selected one from each study. 1 This approach ensured statistical independence of effect sizes as required by the metan command, while maintaining the distinction between different outcome types (i.e. quantity, frequency and age of initiation). Missing data on effect sizes and subgroups were handled using complete case analysis.
Moderator analyses
Moderator analyses examined potential effect size variations through bivariate meta‐regression with RVE to identify sources of heterogeneity and populations with greatest disparities. Study‐level moderators examined whether aggregate study characteristics predicted effect size magnitude, including publication year, sampling strategy (probability vs. convenience sampling), publication type (journal article vs. dissertation/thesis), country (United States vs. other countries), mean age (≤18 vs. >18 years), percentage of White participants and college student sample (yes/no). 2 Subgroup analyses examined participant moderators where effect sizes could be calculated separately for participant subgroups within studies, including sex/gender 3 (effect sizes for female vs. male participants), sexual orientation category and sexual orientation measurement type (identity vs. behavior vs. attraction). For sexual orientation category, we combined across identity, attraction and behavior dimensions. This means that regardless of whether sexual orientation was measured via identity, attraction and/or behavior in the original studies, we categorized participants into four groups: (1) studies that aggregated all sexual minority individuals combined into one group; (2) lesbian/gay individuals or those with same‐sex attraction/sexual partners; (3) bisexual/plurisexual individuals or those with multiple‐sex attractions/sexual partners or who identified as mostly heterosexual; and (4) those who were unsure/questioning. This approach allows us to analyze differences between sexual orientation subgroups while accommodating the diverse measurement approaches across studies. Substance‐related moderators examined whether outcome characteristics influenced effect sizes, including outcome measurement type (quantity vs. frequency) and duration (lifetime vs. recent use). Recent use encompassed any past‐year use, including past 12 months, 6 months, 90 days, 30 days, 2 weeks, week or daily use, while lifetime use referred to any use throughout participants' lives. Moderator analyses used per‐analysis missing data exclusion, with only studies missing the relevant moderator variable excluded from each specific analysis, to maximize the available data for each individual moderator test.
Heterogeneity and reporting biases
Heterogeneity was assessed using τ2, I2 and the Q‐statistic [36]. The τ2 statistic measures the estimated between‐study variance (>0.50 suggesting substantial heterogeneity), I2 estimates what proportion of the variation in observed effects is because of variation in true effects rather than chance [37] and the Q‐statistic tests whether observed differences exceed sampling error (significant P‐values indicate heterogeneity). Additionally, prediction intervals were calculated to estimate the likely relationships examined in this set of studies in future studies [38]. Publication bias was evaluated through a meta‐regression of effect sizes on standard errors using RVE [39] to account for effect size dependencies (Table S1) and visually examined through contour‐enhanced funnel plots (Figures S2–S3) [40], which should be interpreted cautiously given the dependent effect size structure of our data.
RESULTS
Study characteristics
Among 304 studies (5 928 282 participants) meeting the inclusion criteria, 72 reported continuous and 260 reported dichotomous outcomes, with 28 reporting both types of outcomes (Figure 1; Table 1; Appendix S2 and Appendix S3). The mean participant age was 19.61 years old for continuous outcomes and 19.19 years for dichotomous outcomes, with approximately 36% to 43% under 18 years. Among sexual orientation dimensions, the combined sexual minority group represented 58.9% of continuous and 29.7% of dichotomous outcome studies. Sexual orientation was primarily assessed with identity measures. For substance use outcomes, alcohol was most frequently reported in continuous outcomes, although nicotine was most frequently reported in dichotomous outcomes. Few studies examined age of initiation. Recent use patterns were more commonly assessed than lifetime use history. The number of studies published per year has increased dramatically from 2008 to 2024 (Figure S1).
| Study characteristics | Continuous outcome (n = 392, k = 72) | Dichotomous outcome (n = 4433, k = 260) | ||||
|---|---|---|---|---|---|---|
| N (%)/mean (SD) | Range | Missingness | N (%)/mean (SD) | Range | Missingness | |
| Study characteristics | ||||||
| Publication y | 2017.61 (5.18) | 2008–2024 | n = 0 | 2018.21 (3.63) | 2008–2024 | n = 0 |
| Study sampling strategy | n = 0 | n = 0 | ||||
| Convenience sampling | 227 (57.9) | 1629 (36.8) | ||||
| Probability sampling | 165 (42.1) | 2804 (63.3) | ||||
| Publication type | n = 0 | n = 0 | ||||
| Journal article | 346 (88.3) | 4281 (96.6) | ||||
| Dissertation/thesis | 46 (11.7) | 152 (3.4) | ||||
| Country | n = 0 | n = 0 | ||||
| United States | 366 (93.4) | 3597 (81.1) | ||||
| Other countries | 26 (6.6) | 836 (18.9) | ||||
| Mean age (y) | 15.39–24.60 | n = 208 | 11.50–25.00 | n = 2662 | ||
| Continuous variable | 19.61 (3.08) | 19.19 (3.29) | ||||
| 18 y or younger (vs. older than 18 y) | 67 (36.4) | 788 (43.1) | ||||
| Transgender identity | n = 346 | n = 3815 | ||||
| % Transgender | 0.03 (0.03) | 0.00–0.13 | 0.04 (0.05) | 0.00–0.21 | ||
| Race/ethnicity | n = 32 | n = 1206 | ||||
| % White, non‐Hispanic/Latine | 0.59 (0.18) | 0.00–0.93 | 0.56 (0.21) | 0.00–0.95 | ||
| College student | n = 0 | n = 49 | ||||
| Yes | 111 (28.3) | 727 (16.6) | ||||
| Participant characteristics | ||||||
| Sex/gender | n = 244 | n = 1584 | ||||
| Effect sizes for female | 104 (70.3) | 1517 (53.3) | ||||
| Effect sizes for male | 44 (29.7) | 1332 (46.8) | ||||
| Sexual orientation category | n = 0 | n = 37 | ||||
| All sexual minority participants combined into one group | 231 (58.9) | 1307 (29.7) | ||||
| Lesbian/gay, participants with same‐sex attraction/sexual partners | 56 (14.3) | 1176 (26.8) | ||||
| Bisexual/plurisexual, participants with both‐sex attractions/sexual partners/mostly heterosexual | 81 (20.7) | 1315 (29.9) | ||||
| Unsure/questioning | 24 (6.1) | 598 (13.6) | ||||
| Sexual orientation measurement type | n = 0 | n = 0 | ||||
| Identity | 357 (91.1) | 4054 (91.5) | ||||
| Attraction | 15 (3.8) | 194 (4.4) | ||||
| Behavior | 14 (3.6) | 98 (2.2) | ||||
| Mixed | 6 (1.5) | 87 (2.0) | ||||
| Substance outcomes | n = 0 | n = 3 | ||||
| Alcohol | 174 (44.4) | n = 6 | 1147 (25.9) | n = 92 | ||
| Any alcohol use | 107 (63.7) | 648 (61.4) | ||||
| Hazardous alcohol use a | 61 (36.3) | 407 (38.6) | ||||
| Nicotine | 89 (22.7) | 1290 (29.1) | ||||
| Cannabis | 68 (17.4) | 550 (12.4) | ||||
| Prescription drugs | 13 (3.3) | 286 (6.5) | ||||
| Cocaine | 2 (0.5) | 110 (2.5) | ||||
| Meth/amphetamines | 2 (0.5) | 144 (3.3) | ||||
| Ecstasy/MDMA/molly | 0 (0.0) | 172 (3.9) | ||||
| Crack | 1 (0.3) | 2 (0.1) | ||||
| Heroin | 1 (0.3) | 77 (1.7) | ||||
| Other specific substance use b | 26 (6.6) | 546 (12.3) | ||||
| Mixed/poly substance use c | 16 (4.1) | 106 (2.4) | ||||
| Substance characteristics | ||||||
| Outcome measurement type | n = 9 | n = 0 | ||||
| Frequency | 205 (53.5) | 1889 (42.6) | ||||
| Quantity | 148 (38.6) | 2422 (54.6) | ||||
| Age of initiation | 30 (7.8) | 122 (2.8) | ||||
| Outcome measurement duration | n = 51 | n = 316 | ||||
| Lifetime use | 33 (9.7) | 1469 (36.3) | ||||
| Recent use | 308 (90.3) | n = 0 | 2621 (63.7) | n = 0 | ||
| Past 12 months/6 months/90 days | 107 (31.4) | 742 (28.3) | ||||
| Past 30 days/2 weeks/week, daily | 201 (58.9) | 1879 (71.7) | ||||
Reporting biases
Visual inspection of contour‐enhanced funnel plots conducted separately for continuous (Figure S2) and dichotomous outcomes (Figure S3) revealed some asymmetry, particularly for studies with larger standard errors, suggesting possible small‐study effects. RVE meta‐regressions of effect sizes on standard errors for continuous and dichotomous outcomes showed statistically significant relationships, indicating potential small‐study effects (Table S1).
Meta‐analysis results by substance outcomes
SMY had significantly elevated rates of substance use compared to heterosexual youth across all substance outcomes (Table 2). For continuous outcomes, Hedges' g ranged from 0.10 (95% CI = 0.05–0.15, I2 = 99.52, participants = 354 201, studies = 48) for alcohol to 0.40 (95% CI = 0.21–0.59, I2 = 100.00, participants = 30 679, studies = 5) for mixed/polysubstance use. To provide concrete clinical interpretation for the significant alcohol use findings, a supplementary analysis of studies measuring frequency as days per month (studies = 10, effect sizes = 41) estimated that SMY use alcohol approximately 0.36 more days per month compared to heterosexual youth. Dichotomous outcomes showed consistently elevated ORs, ranging from 1.34 (95% CI = 1.24–1.46, I2 = 96.01, participants = 3 500 203, studies = 128) for alcohol to 4.63 (95% CI = 2.91–7.38, I2 = 86.35, participants = 391 827, studies = 10) for heroin use, indicating SMY had 34% to 363% higher odds of substance use compared to heterosexual peers. Between‐study heterogeneity was high, with Q‐statistics indicating significant variation in both outcome types.
| Substance outcome types | n (k) | Participants N | Hedges' g | SE | 95% CI | 95% PI | P‐value | τ2 | I2 (%) | Q‐statistic | Q‐statistic (P‐value) |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Continuous outcome | |||||||||||
| Outcome type: quantity and frequency | |||||||||||
| All substance outcomes | 360 (70) | 551 512 | 0.18 | 0.03 | 0.12–0.24 | −0.79‐1.15 | <0.001 | 0.24 | 99.97 | 210 000 | <0.001 |
| Alcohol | 168 (48) | 354 201 | 0.10 | 0.03 | 0.05–0.15 | −0.26‐0.46 | <0.001 | 0.03 | 99.52 | 9406.70 | <0.001 |
| Any alcohol use | 107 (34) | 299 918 | 0.10 | 0.03 | 0.04–0.16 | −0.14‐0.33 | 0.003 | 0.01 | 98.54 | 2073.01 | <0.001 |
| Hazardous alcohol use a | 61 (24) | 202 770 | 0.10 | 0.04 | 0.01–0.18 | −0.34‐0.54 | 0.02 | 0.04 | 99.70 | 7423.35 | <0.001 |
| Nicotine | 82 (22) | 251 767 | 0.19 | 0.03 | 0.12–0.26 | −0.11‐0.48 | <0.001 | 0.02 | 94.39 | 350.93 | <0.001 |
| Cannabis | 57 (22) | 79 678 | 0.21 | 0.05 | 0.10–0.32 | −0.53‐0.95 | <0.001 | 0.12 | 99.50 | 4236.27 | <0.001 |
| Other specific substance use b | 26 (12) | 46 332 | 0.32 | 0.08 | 0.15–0.49 | −0.70‐1.34 | 0.002 | 0.21 | 99.99 | 77035.19 | <0.001 |
| Mixed/polysubstance usec * | 5 (5) | 30 679 | 0.40 | 0.10 | 0.21–0.59 | −0.22‐1.02 | <0.001 | 0.04 | 100.00 | 424978.01 | <0.001 |
| Outcome type: age of initiation | |||||||||||
| Cannabis initiation* | 3 (3) | 350 114 | −0.31 | 0.20 | −0.70‐0.08 | −2.02‐1.40 | 0.119 | 0.12 | 99.97 | 5796.06 | <0.001 |
Age of initiation analyses for dichotomous outcomes demonstrated that SMY initiated substance use at younger ages than heterosexual youth (all substance outcomes: OR = 1.45; 95% CI = 1.04–2.03, I2 = 95.39, participants = 619 187, studies = 11, respectively), indicating SMY had 45% higher odds of earlier substance use initiation compared to heterosexual peers, and studies demonstrated high heterogeneity.
Given that the age inclusion criterion for this review was studies with samples with a mean age of 25 years or younger, there were some studies that had wider age ranges. As such, we also conducted sensitivity analyses excluding studies with upper age ranges in the 30s and older to ensure the findings were specific to studies with youth samples. We found that patterns of results remained consistent for substance use outcomes, age of initiation and the below moderation findings.
Moderator analyses results
Analyses revealed significant moderating factors across both continuous and dichotomous outcomes. For continuous outcomes (Table 3), significant moderating effects were observed across study participant‐level characteristics (publication year, age and sexual orientation) and measurement‐specific characteristics (duration). Modest temporal changes in substance use inequities were observed over publication years, with disparities decreasing by −0.11 per decade (g = −0.11; 95% CI = −0.22 to −0.01, I2 = 99.95, participants = 551 512, studies = 70). Participants 18 years or younger revealed significantly larger disparities compared to those with older participants (g= 0.21; 95% CI = 0.03–0.40, I2 = 99.98, participants = 315 647, studies = 44). Identity‐based sexual orientation measurement revealed consistent patterns across subgroups. Specifically, lesbian/gay (g = 0.17; 95% CI = 0.05–0.29, I2 = 93.22, participants = 282 868, studies = 17) and bisexual/plurisexual (g = 0.19; 95% CI = 0.10–0.28, I2 = 93.78, participants = 285 932, studies = 19) participants showed higher substance use compared to heterosexual participants. Last, greater disparities were observed for lifetime substance use than recent use (g = 0.18; 95% CI = 0.02–0.34, I2 = 99.81, participants = 421 642, studies = 59).
| Moderator variable | n (k) | Participants N | Hedges' g | SE | 95% CI | P‐value | τ2 | I2 (%) | Q‐statistic | Q‐statistic (P‐value) | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Study characteristics | |||||||||||
| Publication y (per decade) | 360 (70) | 551 512 | −0.11 | 0.05 | −0.22 | −0.01 | 0.04 | 0.09 | 99.95 | 62306.04 | <0.001 |
| Study sampling strategy | |||||||||||
| Probability (vs. convenience) | 360 (70) | 551 512 | 0.13 | 0.07 | −0.01 | 0.28 | 0.06 | 0.13 | 99.96 | 98294.68 | <0.001 |
| Publication type | |||||||||||
| Journal article (vs. dissertation/thesis) | 360 (70) | 551 512 | −0.06 | 0.07 | −0.23 | 0.11 | 0.41 | 0.34 | 100.00 | 210 000 | <0.001 |
| Country | |||||||||||
| US (vs. non‐US) | 360 (70) | 551 512 | −0.08 | 0.05 | −0.19 | 0.03 | 0.14 | 0.22 | 99.99 | 190 000 | <0.001 |
| Mean age (y) | |||||||||||
| 18 y or younger (vs. older than 18 y) | 178 (44) | 315 647 | 0.21 | 0.09 | 0.03 | 0.40 | 0.03 | 0.18 | 99.98 | 110 000 | <0.001 |
| Race/ethnicity | |||||||||||
| % White, non‐Hispanic/Latine | 328 (61) | 393 469 | −0.04 | 0.17 | −0.39 | 0.30 | 0.80 | 0.19 | 99.99 | 140 000 | <0.001 |
| College student | |||||||||||
| Yes (vs. no) | 360 (70) | 551 512 | −0.03 | 0.06 | −0.15 | 0.10 | 0.69 | 0.20 | 99.96 | 150 000 | <0.001 |
| Sample participant characteristics | |||||||||||
| Sex/gender | |||||||||||
| Effect sizes for female (vs. male) | 124 (29) | 154 566 | −0.14 | 0.13 | −0.42 | 0.14 | 0.32 | 0.12 | 99.86 | 16373.57 | <0.001 |
| Sexual orientation category (ref: studies that aggregated all sexual minority participants combined into one group) | 360 (70) | 551 512 | 0.24 | 99.99 | 210 000 | <0.001 | |||||
| Lesbian/gay, participants with same‐sex attraction/sexual partners | 0.05 | 0.08 | −0.12 | 0.22 | 0.53 | ||||||
| Bisexual/plurisexual, participants with both‐sex attractions/sexual partners/mostly heterosexual | 0.03 | 0.06 | −0.09 | 0.15 | 0.60 | ||||||
| Unsure/questioning | 0.45 | 0.18 | 0.01 | 0.88 | 0.05 | ||||||
| Sexual orientation category (ref: lesbian/gay, participants with same‐sex attraction/sexual partners) | 360 (70) | 551 512 | 0.24 | 99.99 | 210 000 | <0.001 | |||||
| Studies that aggregated all sexual minority participants combined into one group | −0.05 | 0.08 | −0.22 | 0.12 | 0.53 | ||||||
| Bisexual/plurisexual, participants with both‐sex attractions/sexual partners/mostly heterosexual | −0.02 | 0.08 | −0.18 | 0.14 | 0.81 | ||||||
| Unsure/questioning | 0.40 | 0.18 | −0.02 | 0.81 | 0.06 | ||||||
| Sexual orientation measurement type (only among lesbian/gay, participants with same‐sex attractions/sexual partners) | |||||||||||
| Identity | 47 (17) | 282 868 | 0.17 | 0.06 | 0.05 | 0.29 | 0.008 | 0.03 | 93.22 | 215.35 | <0.001 |
| Sexual orientation measurement type (only among bisexual/plurisexual, participants with both‐sex attractions/sexual partners/mostly heterosexual) | |||||||||||
| Identity | 71 (19) | 285 932 | 0.19 | 0.04 | 0.10 | 0.28 | <0.001 | 0.02 | 93.78 | 265.03 | <0.001 |
| Substance characteristics | |||||||||||
| Outcome measurement type | |||||||||||
| Quantity (vs. frequency) | 351 (70) | 551 512 | 0.04 | 0.06 | −0.07 | 0.16 | 0.45 | 0.26 | 99.96 | 180 000 | <0.001 |
| Outcome measurement duration | |||||||||||
| Lifetime (vs. recent a ) | 327 (59) | 421 642 | 0.16 | 0.07 | 0.00 | 0.32 | 0.05 | 0.02 | 99.84 | 4692.31 | <0.001 |
| Outcome measurement duration (ref: past 30 days/2 weeks/week, daily) | 327 (59) | 421 642 | 0.02 | 99.81 | 4692.24 | <0.001 | |||||
| Lifetime | 0.18 | 0.07 | 0.02 | 0.34 | 0.04 | ||||||
| Past 12 months/6 months/90 days | 0.05 | 0.06 | −0.07 | 0.16 | 0.40 | ||||||
For dichotomous outcomes (Table 4), significant moderating factors were age, sex and sexual orientation and substance measurement duration. Specifically, participants 18 years old or younger showed 25% higher odds of disparities compared to participants above 18 years (OR = 1.25; 95% CI = 1.05–1.48, I2 = 95.80, participants = 1 796 870, studies = 135), indicating that sexual orientation inequities in substance use are particularly pronounced during adolescence. For studies where sex/gender‐specific effect sizes could be calculated, female/women participants demonstrated 21% higher odds of disparities compared to male/men participants (OR = 1.21; 95% CI = 1.05–1.40, I2 = 95.09, participants = 2 471 116, studies = 101). When examining sexual orientation categories (I2 = 97.54, participants = 5 675 993, studies = 232), aggregated across identity, attraction and behavior, bisexual/plurisexual participants had a greater magnitude in disparities with 28% higher odds (OR = 1.28; 95% CI = 1.16–1.42) compared to lesbian/gay participants. Studies examining unsure/questioning participants showed smaller effect sizes with 31% lower odds compared to studies that combined sexual minority participants into one group (OR = 0.69; 95% CI = 0.58–0.83) and 18% lower odds compared to lesbian/gay participants (OR = 0.82; 95% CI = 0.70–0.96). The analysis of sexual orientation measurement types revealed that among lesbian/gay participants, identity‐based measures showed 45% higher odds (OR = 1.45; 95% CI = 1.30–1.62, I2 = 94.53, participants = 2 386 871, studies = 81) and attraction‐based measures showed 68% higher odds (OR = 1.68; 95% CI = 1.14–2.49, I2 = 82.17, participants = 43 440, studies = 7) of substance use compared to their heterosexual counterparts, while behavior‐based measures did not reach statistical significance (OR = 1.26; 95% CI = 0.89–1.78, I2 = 87.03, participants = 523 110, studies = 10). For bisexual/plurisexual participants, identity‐ and behavior‐based measures also revealed consistent disparities between bisexual/plurisexual participants compared to heterosexual participants, and attraction was not reported as there were too few reports. For bisexual/pansexual participants, effect sizes were larger when sexual orientation was assessed on a behavioral dimension with 139% higher odds (OR = 2.39; 95% CI = 1.74–3.28, I2 = 91.41, participants = 520 710, studies = 10) versus an identity dimension with 78% higher odds (OR = 1.78; 95% CI = 1.65–1.92, I2 = 94.82, participants = 2 906 772, studies = 89) of substance use than heterosexual participants. Studies examining lifetime substance use showed 14% higher odds of disparities than those examining recent use (OR = 1.14; 95% CI = 1.01–1.29, I2 = 97.30, participants = 5 185 079, studies = 221), suggesting that sexual orientation inequities are more pronounced when considering cumulative lifetime experiences rather than current use patterns. Most study‐level characteristics including sampling strategy, publication type, country, race/ethnicity composition, college student status and outcome measurement type (quantity vs. frequency) did not significantly moderate effect sizes across outcome types.
| Moderator variable | n (k) | Participants N | OR | SE | 95% CI | P‐value | τ2 | I2 (%) | Q‐statistic | Q‐statistic (P‐value) | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Study characteristics | |||||||||||
| Publication y (per decade) | 4311 (232) | 5 675 993 | 0.87 | 0.07 | 0.75 | 1.01 | 0.06 | 0.25 | 98.20 | 4732.93 | <0.001 |
| Study sampling strategy | |||||||||||
| Probability (vs. convenience) | 4311 (232) | 5 675 993 | 1.09 | 0.06 | 0.97 | 1.22 | 0.15 | 0.26 | 95.62 | 4709.76 | <0.001 |
| Publication type | |||||||||||
| Journal article (vs. dissertation/thesis) | 4311 (232) | 5 675 993 | 0.96 | 0.11 | 0.76 | 1.23 | 0.74 | 0.24 | 99.72 | 4750.63 | <0.001 |
| Country | |||||||||||
| US (vs. non‐US) | 4311 (232) | 5 675 993 | 0.91 | 0.08 | 0.77 | 1.08 | 0.27 | 0.23 | 98.90 | 4652.51 | <0.001 |
| Mean age (y) | |||||||||||
| 18 y or below (vs. above 18 y) | 1776 (135) | 1 796 870 | 1.25 | 0.09 | 1.05 | 1.48 | 0.01 | 0.25 | 95.80 | 2404.71 | <0.001 |
| Transgender identity | |||||||||||
| % Transgender | 612 (56) | 1 683 823 | 0.16 | 0.93 | 0.02 | 1.55 | 0.10 | 0.22 | 99.71 | 2110.71 | <0.001 |
| Race/ethnicity | |||||||||||
| % White, non‐Hispanic/Latine | 3135 (182) | 5 008 660 | 0.76 | 0.15 | 0.56 | 1.03 | 0.08 | 0.21 | 98.06 | 3591.27 | <0.001 |
| College student | |||||||||||
| Yes (vs. no) | 4262 (231) | 5 672 198 | 0.92 | 0.07 | 0.80 | 1.06 | 0.25 | 0.26 | 98.75 | 4746.15 | <0.001 |
| Participant characteristics | |||||||||||
| Sex | |||||||||||
| Effect sizes for female (vs. male) | 2760 (101) | 2 471 116 | 1.21 | 0.07 | 1.05 | 1.40 | 0.01 | 0.42 | 95.09 | 1772.70 | <0.001 |
| Sexual orientation category (ref: studies that aggregated all sexual minority participants combined into one group) | 4274 (232) | 5 675 993 | 0.23 | 97.54 | 4541.20 | <0.001 | |||||
| Lesbian/gay, participants with same‐sex attraction/sexual partners | 0.85 | 0.06 | 0.75 | 0.96 | 0.01 | ||||||
| Bisexual/plurisexual, participants with both‐sex attractions/sexual partners/mostly heterosexual | 1.09 | 0.06 | 0.97 | 1.21 | 0.14 | ||||||
| Unsure/questioning | 0.69 | 0.09 | 0.58 | 0.83 | <0.001 | ||||||
| Sexual orientation category (ref: lesbian/gay, participants with same‐sex attraction/sexual partners) | 4274 (232) | 5 675 993 | 0.23 | 97.54 | 4541.20 | <0.001 | |||||
| Studies that aggregated all sexual minority participants combined into one group | 1.18 | 0.06 | 1.04 | 1.34 | 0.01 | ||||||
| Bisexual/plurisexual, participants with both‐sex attractions/sexual partners/mostly heterosexual | 1.28 | 0.05 | 1.16 | 1.42 | <0.001 | ||||||
| Unsure/questioning | 0.82 | 0.08 | 0.70 | 0.96 | 0.01 | ||||||
| Sexual orientation measurement type (only among lesbian/gay, participants with same‐sex attractions/sexual partners) | |||||||||||
| Identity | 1013 (81) | 2 386 871 | 1.45 | 0.06 | 1.30 | 1.62 | <0.001 | 0.37 | 94.53 | 1258.59 | <0.001 |
| Behavior | 60 (10) | 523 110 | 1.26 | 0.15 | 0.89 | 1.78 | 0.16 | 0.25 | 87.03 | 63.40 | <0.001 |
| Attraction | 37 (7) | 43 440 | 1.68 | 0.16 | 1.14 | 2.49 | 0.02 | 0.17 | 82.17 | 29.94 | <0.001 |
| Sexual orientation measurement type (only among bisexual/plurisexual, participants with both‐sex attractions/sexual partners/mostly heterosexual) | |||||||||||
| Identity | 1170 (89) | 2 906 772 | 1.78 | 0.04 | 1.65 | 1.92 | <0.001 | 0.24 | 94.82 | 1520.34 | <0.001 |
| Behavior | 50 (10) | 520 710 | 2.39 | 0.14 | 1.74 | 3.28 | <0.001 | 0.19 | 91.41 | 84.75 | <0.001 |
| Substance characteristics | |||||||||||
| Outcome measurement type | |||||||||||
| Quantity (vs. frequency) | 4311 (232) | 5 675 993 | 1.00 | 0.06 | 0.89 | 1.12 | 0.99 | 0.24 | 97.00 | 4620.53 | <0.001 |
| Outcome measurement duration | |||||||||||
| Lifetime (vs. recent a ) | 4051 (221) | 5 185 079 | 1.11 | 0.06 | 0.99 | 1.25 | 0.08 | 0.24 | 97.40 | 5340.46 | <0.001 |
| Outcome measurement duration (ref: past 30 days/2 weeks/week, daily) | 4051 (221) | 5 185 079 | 0.25 | 97.30 | 5336.68 | <0.001 | |||||
| Lifetime | 1.14 | 0.06 | 1.01 | 1.29 | 0.03 | ||||||
| Past 12 months/6 months/90 days | 1.09 | 0.06 | 0.96 | 1.24 | 0.16 | ||||||
DISCUSSION
In this systematic review and meta‐analysis of 304 studies, including 5 928 282 youth, the findings demonstrated that SMY initiated substance use at younger ages, were more likely to use all substances (e.g. alcohol, nicotine, cannabis, opioids, stimulants and other drugs) at a greater frequency and quantity and engaged in more mixed and polysubstance and hazardous substance use than heterosexual youth. These inequities were accentuated based on developmental, demographic and methodological factors. Our findings are consistent with the only previous meta‐analysis of sexual orientation disparities in substance among youth from 2008 [13], and provide an updated, comprehensive and rigorous estimation of sexual orientation inequities in substance use among youth. The magnitude of inequities documented in our study ranged widely, with more accessible and common substances (e.g. alcohol, nicotine) demonstrating smaller effects and illicit drugs (e.g. heroin) having the largest effects. Although the prior meta‐analysis found larger effects across all substances compared to our study, it similarly documented the largest associations for illicit drugs [13]. Our study analyzed inequities separately for continuous and dichotomous outcomes, with more documented significant findings among dichotomous outcomes, likely because there were substantially more reports with dichotomous outcomes, providing more statistical power to detect effects.
Our study's results highlight key demographic differences in substance use inequities. First, we found that plurisexual youth had the highest inequities in substance use compared to monosexual youth (e.g. heterosexual, lesbian and gay). Second, we found that inequities in substance use were greater among SMY girls and young women compared to SMY boys and young men. These finding are consistent with the prior meta‐analysis of substance use disparities [13] and other studies demonstrating that plurisexual and SMY girls and young women have elevated rates of numerous poor health outcomes [5, 41]. Third, our findings demonstrate developmental differences, where inequities were more pronounced in adolescence compared to young adulthood. Given that prior longitudinal research has shown that substance use inequities have increased, particularly among sexual minority girls and young women [15, 42], and more recent work has indicated potentially elevated rates for SMY of color [18, 19, 22], prevention and intervention research is urgently needed to address the unique biopsychosocial and health needs of these particular subgroups.
Several etiological factors shape SMY's substance use. According to the oppression framework, inequities in substance use are rooted in systems of oppression [6]. SMY experience and navigate pervasive oppression throughout their development (e.g. heterosexist‐ and anti‐bisexual‐based victimization or discrimination, identity concealment, internalized oppression and heterosexist laws) [5, 6, 43, 44]. Therefore, SMY may use substances to cope with distress, shame, harm and lack of safety inflicted by systems of oppression and/or to resist oppression (e.g. use substances to be in community, celebrate, feel liberated and be sexual) [6, 20, 45, 46]. Furthermore, oppression shapes cultural and social norms related to substance use, which then create permissive substance use norms in the lesbian, gay, bisexual, trans, queer (LGBTQ) community [6, 45, 47, 48]. In fact, studies have documented that alcohol and nicotine companies specifically target the LGBTQ community [6, 49], creating further risk for hazardous substance use among SMY. Last, common etiological factors that influence substance use and development of SUDs (e.g. traumatic life experiences, genetic factors) among the broader youth population also play a role for SMY and intersect with their experiences with oppression [6, 20].
Some methodological factors played a role in our findings. Inequities in substance use were elevated when lifetime versus recent substance use was assessed. This finding is unsurprising as substance use experimentation is more common in adolescence than recent or repeated use. Therefore, more SMY and heterosexual youth would have had an opportunity to use, further magnifying the disparity. Additionally, this finding might suggest that SMY are more likely to experiment in substance use, which theory suggests is because of oppression‐based stress experiences and differences in substance use norms and willingness to use substances [5, 6, 45, 47, 48]. Although underpowered to estimate the magnitude of inequities along the attraction dimension of sexual orientation, our results also indicated that substance use inequities were documented across both identity and behavioral dimensions of sexual orientation. Adolescence is a hallmark period of identity exploration and development [41], therefore, assessing only sexual identification or behavior may miss youth who have not yet identified or do not feel comfortable identifying as a sexual minority, but may be aware of their attractions. These results highlight the importance of comprehensive assessment of sexual orientation among youth, including fluidity [50], especially as only a few studies measured more than one sexual orientation dimension.
The findings of this study have research and clinical implications. Additional research is needed to examine developmental and etiological factors that impact substance use initiation. More longitudinal studies are needed to document inequities and their trajectories from adolescence and throughout the lifespan. Moreover, effective intervention and prevention strategies targeting substance use among SMY are lacking [5,41]. SMY face several barriers when they are in treatment, including providers' inadequate cultural competency, lack of autonomy over their own treatment and fears of confidentiality (e.g. providers outing SMY to their parents/guardians) [51]. Therefore, policies and interventions that increase access to SUD treatment and increase cultural humility among providers are needed. Furthermore, routine screenings for harmful substance use across settings (e.g. schools, primary care, emergency departments and community organizations) and other universal prevention interventions may help prevent and identify harmful substance use among SMY. However, screenings must also be coupled with accessible, de‐stigmatizing, culturally affirming and confidential services to ensure their success and in order not pathologize or harm youth who engage in hazardous substances use. Last, multi‐level interventions that increase affirmation, resilience, resistance and intervene on oppression are essential to reducing and eliminating inequities in substance use [6, 44]. These interventions should target oppression and stigma targeting SMY (e.g. heterosexism, anti‐bisexual prejudice) as well as stigma related to substance use and addiction that marginalizes youth who use substances or have SUDs.
This study has some noteworthy limitations. Although SMY consistently showed elevated substance use compared to their heterosexual peers, the substantial heterogeneity observed in our primary analyses (I2 > 90% across most substance outcomes in Table 2) indicates considerable variation in effect sizes across studies. Similarly, the wide prediction intervals suggest that the magnitude of disparities varies depending on different populations, settings and study characteristics. Although we accounted for sampling differences through moderator analyses, the integration of effect sizes derived from weighted estimates of probability samples versus estimates from convenience samples was a methodological limitation in our meta‐analysis. Our moderator analyses were limited to bivariate models examining one variable at a time, which may yield confounded associations when moderators are correlated, particularly when interpreting differences because of study‐level characteristics. Moreover, because of resource constraints, our search was limited to studies in published in English, limiting our understanding of published work in other languages. Although we assessed the percentage of transgender participants in studies where gender identity was reported, these studies were quite limited, reducing our power to adequately test for gender identity effects. Similarly, studies were not clear on whether they assessed sex or gender, as such, our findings mirror limitations found in the reporting of primary studies that conflate sex and gender. Last, although we examined proxies for study quality, such as study sampling strategy (probability vs. convenience sampling), publication type (published studies vs. dissertation/thesis) and measurement duration (lifetime vs. recent use), a formal assessment of study quality using a standardized tool was not conducted because of limited resources, especially given the large number of studies included.
This systematic review and meta‐analysis provide the most rigorous and comprehensive evidence documenting significant and consistent sexual orientation inequities in substance use among youth, indicating that SMY have significantly elevated rates of substance use compared to their heterosexual peers. Multi‐level approaches are needed to reduce and eliminate these inequities, especially targeting oppression, increasing safety and affirmation and improving access to culturally affirming treatment for SMY.
DECLARATION OF INTERESTS
None.
Supporting information
ACKNOWLEDGEMENTS
We dedicate this manuscript in memory of Dr. Jeremy Goldbach who passed away on 7 June 2025. We are fortunate for Dr. Goldbach's contributions to this manuscript as well as grateful for his impactful legacy and career, which focused on improving the wellbeing and health of LGBTQ youth. We also acknowledge Nikhat Ghouse who was the university librarian who provided guidance and consultation for identifying search terms, conducting the search and article retrieval. We also acknowledge Dr. Emily Tanner‐Smith for her initial feedback on the protocol and assistance with the initial stages of data analysis. We thank the Lavender Lab's research assistants at American University who helped with the initial abstract and full‐text review.
Untitled section
Mereish EH, Lee H, Goldbach JT, Hathaway S, Hennessy EA. A meta‐analysis of sexual orientation inequities in substance use among youth. Addiction. 2026;121(5):1048–1063. 10.1111/add.70301
Untitled section
Footnote Group
Footnotes
Footnote Group
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are openly available through the Open Science Framework at: https://osf.io/t6ywu/files/osfstorage.
REFERENCES
Untitled section
References
- 1. Health UDo, Services H . Preventing tobacco use among youth and young adults: a report of the Surgeon General 3 Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, office on smoking and health; 2012.
- 2. Siegel RL, Jacobs EJ, Newton CC, Feskanich D, Freedman ND, Prentice RL, et al. Deaths due to cigarette smoking for 12 smoking‐related cancers in the United States. JAMA Intern Med. 2015;175(9):1574–1576. 10.1001/jamainternmed.2015.2398
- 3. Jha P, Ramasundarahettige C, Landsman V, Rostron B, Thun M, Anderson RN, et al. 21st‐century hazards of smoking and benefits of cessation in the United States. New England Journal of Medicine. 2013;368(4):341–350. 10.1056/NEJMsa1211128
- 4. Kendler KS, Myers J, Damaj MI, Chen X. Early smoking onset and risk for subsequent nicotine dependence: a monozygotic co‐twin control study. Am J Psychiatry. 2013;170(4):408–413. 10.1176/appi.ajp.2012.12030321
- 5. Mereish EH. Substance use and misuse among sexual and gender minority youth. Curr Opin Psychol. 2019;30:123–127. 10.1016/j.copsyc.2019.05.002
- 6. Mereish EH. Oppression‐based stress and alcohol inequities among sexual and gender minority people: an intersectional multi‐level framework. Alcohol Res. 2024;44(1):05. 10.35946/arcr.v44.1.05
- 7. Phillips G II, Turner B, Salamanca P, Birkett M, Hatzenbuehler ML, Newcomb ME, et al. Victimization as a mediator of alcohol use disparities between sexual minority subgroups and sexual majority youth using the 2015 National Youth Risk Behavior Survey. Drug Alcohol Depend. 2017;178:355–362. 10.1016/j.drugalcdep.2017.05.040
- 8. Johnson SE, O'Brien EK, Coleman B, Tessman GK, Hoffman L, Delahanty J. Sexual and gender minority US youth tobacco use: Population assessment of tobacco and health (PATH) study wave 3, 2015–2016. Am J Prev Med. 2019;57(2):256–261. 10.1016/j.amepre.2019.03.021
- 9. Dunbar MS, Siconolfi D, Rodriguez A, Seelam R, Davis JP, Tucker JS, et al. Alcohol use and cannabis use trajectories and sexual/gender minority disparities in young adulthood. Psychol Addict Behav. 2022;36(5):477–490. 10.1037/adb0000806
- 10. Marshal MP, King KM, Stepp SD, Hipwell A, Smith H, Chung T, et al. Trajectories of alcohol and cigarette use among sexual minority and heterosexual girls. J Adolesc Health. 2012;50(1):97–99. 10.1016/j.jadohealth.2011.05.008
- 11. Corliss HL, Wadler BM, Jun H‐J, Rosario M, Wypij D, Frazier AL, et al. Sexual‐orientation disparities in cigarette smoking in a longitudinal cohort study of adolescents. Nicotine Tob Res. 2013;15(1):213–222. 10.1093/ntr/nts114
- 12. Goldbach JT, Mereish EH, Burgess C. Sexual orientation disparities in the use of emerging drugs. Subst Use Misuse. 2017;52(2):265–271. 10.1080/10826084.2016.1223691
- 13. Marshal MP, Friedman MS, Stall R, King KM, Miles J, Gold MA, et al. Sexual orientation and adolescent substance use: a meta‐analysis and methodological review*. Addiction. 2008;103(4):546–556. 10.1111/j.1360-0443.2008.02149.x
- 14. Talley AE, Turner B, Foster AM, Phillips G. Sexual minority youth at risk of early and persistent alcohol, tobacco, and marijuana use. Arch Sex Behav. 2019;48(4):1073–1086. 10.1007/s10508-018-1275-7
- 15. Fish JN, Bishop MD, Russell ST. Developmental differences in sexual orientation and gender identity–related substance use disparities: Findings from population‐based data. J Adolesc Health. 2021;68(6):1162–1169. 10.1016/j.jadohealth.2020.10.023
- 16. Goldbach JT, Tanner‐Smith EE, Bagwell M, Dunlap S. Minority stress and substance use in sexual minority adolescents: a meta‐analysis. Prev Sci. 2014;15(3):350–363. 10.1007/s11121-013-0393-7
- 17. Eisenberg ME, Gower AL, Watson RJ, Rider GN, Thomas D, Russell ST. Substance use behaviors among LGBTQ+ youth of color: Identification of the populations bearing the greatest burden in three large samples. J Adolesc Health. 2022;71(3):317–323. 10.1016/j.jadohealth.2022.04.007
- 18. Andrzejewski J, Calzo JP, Smith LR, et al. Binge drinking disparities by gender identity, race, and ethnicity in California secondary schools. J LGBT Youth. 2022;1–22. 10.1080/19361653.2022.2113198
- 19. Pollitt AM, Mallory AB, Fish JN. Homophobic bullying and sexual minority youth alcohol use: Do sex and race/ethnicity matter? LGBT Health. 2018;5(7):412–420. 10.1089/lgbt.2018.0031
- 20. Kidd JD, Jackman KB, Wolff M, Veldhuis CB, Hughes TL. Risk and protective factors for substance use among sexual and gender minority youth: a scoping review. Curr Addict Rep. 2018;5(2):158–173. 10.1007/s40429-018-0196-9
- 21. Evans‐Polce RJ, Kcomt L, Veliz P, Boyd CJ, McCabe SE. Associations of sexual identity change and identity–attraction discordance with symptomatic alcohol and other drug use. Soc Psychiatry Psychiatr Epidemiol. 2024;59(8):1389–1399. 10.1007/s00127-023-02600-2
- 22. Bishop MD, Moran SA, Turpin RE, Aparicio EM, Mereish EH, Russell ST, et al. Binge drinking disparities by grade, race and ethnicity, sexual orientation, and gender identity. Pediatrics. 2024;154(3):e2023063978. 10.1542/peds.2023-063978
- 23. D'Agati D, Kahn GD, Swartz KL. Preteen behaviors and sexual orientation of high school students who report depressive symptoms, United States, 2015‐2017. Public Health Rep. 2021;136(2):132–135. 10.1177/0033354920971718
- 24. Kann L, McManus T, Harris WA. Youth risk behavior surveillance — United States, 2017. MMWR Surveill Summ. 2018;67(SS‐8):1–114. 10.15585/mmwr.ss6708a1
- 25. Talley AE, Hughes TL, Aranda F, Birkett M, Marshal MP. Exploring alcohol‐use behaviors among heterosexual and sexual minority adolescents: intersections with sex, age, and race/ethnicity. Am J Public Health. 2014;104(2):295–303. 10.2105/ajph.2013.301627
- 26. National Academies of Sciences E, Medicine . Measuring sex, gender identity, and sexual orientation2022.
- 27. Office of Disease Prevention and Health Promotion OotASfH, Office of the Secretary, U.S. Department of Health and Human Services , Healthy People 2030.
- 28. Department of Health DaAatNHaMRC, Australian government . 2025. Statement on sex, gender, variations of sex characteristics and sexual orientation in health and medical research
- 29. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta‐analyses: the PRISMA statement. Int J Surg. 2010;8(5):336–341. 10.1016/j.ijsu.2010.02.007
- 30. King M, Semlyen J, Tai SS, Killaspy H, Osborn D, Popelyuk D, et al. A systematic review of mental disorder, suicide, and deliberate self harm in lesbian, gay and bisexual people. BMC Psychiatry. 2008;8(1):1–17. 10.1186/1471-244X-8-70
- 31. Marshal MP, Dietz LJ, Friedman MS, Stall R, Smith HA, McGinley J, et al. Suicidality and depression disparities between sexual minority and heterosexual youth: a meta‐analytic review. J Adolesc Health. 2011;49(2):115–123. 10.1016/j.jadohealth.2011.02.005
- 32. Lee JGL, Ylioja T, Lackey M. Identifying lesbian, gay, bisexual, and transgender search terminology: a systematic review of health systematic reviews. PLoS ONE. 2016;11(5):e0156210. 10.1371/journal.pone.0156210
- 33. Lipsey MW, Wilson DB. Practical meta‐analysis SAGE publications, Inc; 2001.
- 34. Borenstein M, Hedges LV, Higgins JP, et al. Introduction to Meta‐analysis John Wiley & Sons; 2009.
- 35. Tanner‐Smith EE, Tipton E. Robust variance estimation with dependent effect sizes: practical considerations including a software tutorial in Stata and SPSS. Res Synth Methods. 2014;5(1):13–30. 10.1002/jrsm.1091
- 36. Johnson BT, Hennessy EA. Systematic reviews and meta‐analyses in the health sciences: best practice methods for research syntheses. Soc Sci Med. 2019;233:237–251. 10.1016/j.socscimed.2019.05.035
- 37. Borenstein M, Higgins JP, Hedges LV, Higgins JPT, Rothstein HR. Basics of meta‐analysis: I2 is not an absolute measure of heterogeneity. Res Synth Methods. 2017;8(1):5–18. 10.1002/jrsm.1230
- 38. IntHout J, Ioannidis JP, Rovers MM, Goeman JJ. Plea for routinely presenting prediction intervals in meta‐analysis. BMJ Open. 2016;6(7):e010247. 10.1136/bmjopen-2015-010247
- 39. Rodgers MA, Pustejovsky JE. Evaluating meta‐analytic methods to detect selective reporting in the presence of dependent effect sizes. Psychol Methods. 2021;26(2):141–160. 10.1037/met0000300
- 40. Palmer TM, Sutton AJ, Peters JL, Moreno SG. Contour‐enhanced funnel plots for meta‐analysis. The STATA Journal. 2008;8(2):242–254. 10.1177/1536867X0800800206
- 41. Fish JN. Future directions in understanding and addressing mental health among LGBTQ youth. Journal of Clinical Child & Adolescent Psychology. 2020;49(6):943–956. 10.1080/15374416.2020.1815207
- 42. Watson RJ, Lewis NM, Fish JN, Goodenow C. Sexual minority youth continue to smoke cigarettes earlier and more often than heterosexuals: findings from population‐based data. Drug Alcohol Depend. 2018;184:64–70. 10.1016/j.drugalcdep.2017.11.025
- 43. Goldbach JT, Gibbs JJ. A developmentally informed adaptation of minority stress for sexual minority adolescents. J Adolesc. 2017;55(1):36–50. 10.1016/j.adolescence.2016.12.007
- 44. Sullivan TR, Flynn SS, Touhey S, Mereish EH. Through their eyes: a qualitative, daily diary exploration of oppression‐based stress experiences among sexual and gender minority adolescents. Psychol Sex Orientat Gend Divers. 2024. 10.1037/sgd0000779
- 45. Boyle SC, LaBrie JW, Omoto AM. Normative substance use antecedents among sexual minorities: A scoping review and synthesis. Psychol Sex Orientat Gend Divers. 2020;7(2):117–131. 10.1037/sgd0000373
- 46. Mereish EH, Abramson JR, Lee H, et al. Intersectional oppression‐based stress, drinking to cope motives, and alcohol use and hazardous drinking among sexual and gender minority adolescents who are Black, Indigenous, and People of Color. LGBT Health. 2024;12(2):125–133. 10.1089/lgbt.2024.0023
- 47. Mereish EH, Goldbach JT, Burgess C, DiBello AM. Sexual orientation, minority stress, social norms, and substance use among racially diverse adolescents. Drug Alcohol Depend. 2017;178:49–56. 10.1016/j.drugalcdep.2017.04.013
- 48. Gamarel KE, Mereish EH, Colby SM, Barnett NP, Hayes K, Jackson KM. Sexual minority disparities in substance use willingness among youth. Subst Use Misuse. 2018;53(1):170–175. 10.1080/10826084.2017.1327977
- 49. Whiteley D, Rickards‐Hill D, Dimova E, Emslie C. Performing solidarity? A scoping review of alcohol marketing to sexual and gender minorities. Drugs Educ Prev Policy. 31(6):1–9. 10.1080/09687637.2023.2260550
- 50. Katz‐Wise SL, Ranker LR, Gordon AR, Xuan Z, Nelson K. Sociodemographic patterns in retrospective sexual orientation identity and attraction change in the sexual orientation fluidity in youth study. J Adolesc Health. 2023;72(3):437–443. 10.1016/j.jadohealth.2022.10.015
- 51. Mereish EH, Gamarel KE, Operario D. Understanding and addressing alcohol and substance use in sexual and gender minority youth. In: Monti PM, Colby SM, Tevyaw, TO , editors. Brief Interventions for Adolescent Alcohol and Substance Abuse. The Guilford Press; 2018. p. 305–327.
Associated Data
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are openly available through the Open Science Framework at: https://osf.io/t6ywu/files/osfstorage.