Disparities in cannabis use outcomes, perceived risks and social norms across sexual orientation groups of US young adult women and men Free

Abstract

Little research has examined mechanisms driving cannabis use disparities among sexual minority (SM; versus heterosexual) young adults (YAs). Participants were US YA women (N = 1345; 24.8% bisexual, 6.1% lesbian) and men (N = 998: 8.9% bisexual, 13.8% gay). Bivariate analyses examined associations between sexual orientation (heterosexual, gay/lesbian, bisexual) and cannabis use outcomes (past 30-day [current] use, use frequency, cannabis-tobacco dual use, use intentions), perceived risks (harm, addictiveness) and social norms (social acceptability, peer use). Regressions examined sexual orientation, perceived risks and social norms as correlates of cannabis outcomes. Bisexual and lesbian (versus heterosexual) women reported lower cannabis-related perceived risks, greater social norms and higher odds of current use. Bisexual (versus heterosexual) women reported more days of use, higher odds of cannabis-tobacco dual use and greater use intentions. Gay (versus heterosexual) men reported lower perceived addictiveness, greater social norms and higher odds of current use. Lower perceived risks and greater social norms predicted greater use outcomes for women and men. Bisexual women displayed higher odds of several cannabis use outcomes, whereas lesbian women and gay men displayed higher odds of current cannabis use only. Cannabis-related perceptions and social norms may be important targets for public health messaging for SMYAs.

Introduction

In 2021, 29% of US young adults (YAs; ages 19–30) reported past-month cannabis use, a 71% increase from 17% in 2011 [1]. Increased use rates have occurred alongside the rapid expansion of cannabis legalization for medical and recreational purposes in the United States [2]. This is concerning, as cannabis use is associated with a variety of adverse consequences, including cognitive impairment, motor vehicle crashes, mental health problems and subsequent or concurrent use of other substances (e.g. tobacco, alcohol) [3, 4].

Many minoritized groups, including sexual minority (SM)-identifying individuals (i.e. bisexual, gay/lesbian or another non-heterosexual identity), report disproportionately high rates of cannabis use [5–9]. In 2020, US national data suggested that 29.4% of SM adults reported past-month cannabis use relative to only 10.7% of heterosexual adults [10]. These disparities may be especially pronounced during young adulthood, which entails significant life changes (e.g. greater independence) and greater risk for onset and progression of substance use [11–16]. Young adulthood may be an especially difficult time for SM individuals due to experiences of stigma, discrimination and potentially rejecting family, peer and other social relationships [17]. Among YAs, specifically, findings suggest that SMYAs are more likely to engage in past 30-day cannabis use [8, 9], report more days of cannabis use [7, 18] and display higher rates of cannabis-tobacco dual use [19], which is associated with poor physical, cognitive and emotional health [20] relative to heterosexual YAs.

Recent research has emerged regarding specific sexual orientation and gender differences in relation to cannabis use, albeit with varied findings. For instance, some research suggests that bisexual women are more likely than both lesbian and heterosexual women to engage in a variety of substance use behaviors [5, 21], including cannabis use [8], whereas others indicate that both bisexual and lesbian women report higher rates of cannabis use relative to heterosexual women [6]. Findings among SM men are inconsistent, with some findings suggesting that gay and bisexual (versus heterosexual) men are more likely to use cannabis [7] and others suggesting no differences [9]. Furthermore, there is a paucity of research examining differences in a wider range of cannabis use outcomes (e.g. use frequency, dual use with other substances) across specific sexual orientation-by-gender subgroups or potential mechanisms contributing to such differences.

The primary explanation for disparate rates of substance use among SM individuals has been informed by Minority Stress Theory, which posits that SM individuals face stressors (e.g. discrimination, internalized stigma) due to the stigmatization of their non-heterosexual identity, which may lead to coping-related substance use [22, 23]. Although minority stress likely plays a role in SM substance use disparities, other theory-driven factors, including perceived risks and social norms, may contribute to cannabis use disparities among SM individuals. Among YAs, social norms, including both descriptive (i.e. friend or peer use) and injunctive (i.e. perceived social acceptability of use) norms, are a strong predictor of any current cannabis use [24, 25] and greater frequency of use [26]. Cannabis-related social norms may be higher among SM individuals who typically associate with more SM-identifying peers (who demonstrate higher cannabis use rates [27, 28]) and frequent SM safe spaces (such as LGBTQ+ bars) which may be less restrictive about cannabis use [29]. Research examining the role of social norms on SM-related cannabis use disparities is limited, with only one study suggesting that cannabis-related disparities were linked to SM adolescents’ greater reports of social acceptability and peer cannabis use [30].

Perceived risks of cannabis use, including lower perceived harm to health and addictiveness, are also consistent correlates and predictors of current cannabis use among YAs [25, 31, 32]. Less research has examined differences in risk perceptions based on sexual orientation. Some research suggests that SM men and women perceive fewer health-related substance use risks relative to heterosexual individuals [33, 34]. Lower risk perceptions among SMYAs may be related to greater cannabis-related social norms [35, 36] as well as greater cannabis-related marketing exposure (which often include health claims) [37, 38].

Although disparities in cannabis use among SM relative to heterosexual individuals are well-documented, there is a dearth of research examining specific SM-by-gender subgroups in relation to a wider range of cannabis use outcomes (e.g. use frequency, dual use with other substances) or theory-driven correlates of use. Thus, the current study examined cannabis use outcomes (i.e. past 30-day [current] use, future use intentions, frequency of current use, cannabis-tobacco dual use) among YA women and men, separately, and examined as correlates: (i) group membership, defined by gender and sexual orientation (distinguishing heterosexual, gay/lesbian and bisexual) and (ii) theoretically based factors, including perceived risks (i.e. harm, addictiveness) and social norms (i.e. social acceptability, peer use).

Methods

Participants and procedures

This study analyzed survey data from a 2 year longitudinal study of YAs (aged 18–34) designed to examine correlates of cigarette and e-cigarette use among individuals recruited from six metropolitan statistical areas (MSAs: Atlanta, Boston, Minneapolis, Oklahoma City, San Diego, Seattle) with varied tobacco and cannabis legislative contexts [39]. This study, detailed elsewhere [40], was approved by the George Washington University Institutional Review Board.

Participants were recruited in Fall 2018 and surveyed biannually until Fall 2020 (total of five waves). To recruit participants, ads posted on Facebook and Reddit targeted individuals by using indicators reflecting those eligible (i.e. ages 18–34, residing in one of the six MSAs, English speaking). After clicking on an ad, individuals were directed to a webpage with a study description, consent form and eligibility screener, and then eligible individuals completed the online Wave 1 survey via Alchemer. Upon completion, participants were asked to confirm their participation in the study a week later. Purposive, quota-based sampling ensured the sample represented sufficient numbers of e-cigarette and cigarette users (roughly one-third each), roughly equal numbers of men and women and 40% racial/ethnic minority (subgroup enrollment was capped by MSA).

Of the 10433 individuals who clicked on ads, 9847 consented, of which 2751 (27.9%) were excluded due to: (i) ineligibility (n = 1472) and/or (ii) their subgroup target being met (n = 1279). Among the remaining 7096 individuals, 48.8% (n = 3460) provided complete data, and 86.9% (n = 3006) confirmed participation [40]. The current analyses involve the use of baseline sociodemographic data (as sociodemographics were assessed at baseline only) and Fall 2020 data, which assessed cannabis use outcomes and cannabis-related perceived risks and social norms.

Measures

Stratification variable: gender

Participants were asked ‘What is your gender: male; female; or other?’. Participants who selected ‘other’ (n = 69) were excluded from primary analyses but explored in descriptive analyses.

Primary correlate of interest: sexual orientation

Participants were asked ‘Do you consider yourself: heterosexual or straight; gay or lesbian; or bisexual?’. These responses were recoded into dummy variables. Participants could also respond ‘other’ (n = 43) or ‘prefer not to answer’ (n = 37); participants who selected these options were excluded from primary analyses due to small sample sizes but were explored in descriptive analyses.

Theory-based correlates: perceived risks and social norms

All participants provided ratings of their ‘perceptions of cannabis harm, addictiveness and social acceptability’ (‘How [harmful to your health; addictive; socially acceptable among your peers] is the use of cannabis?’ 1 = not at all to 7 = extremely) and ‘peer cannabis use’ (‘Out of your 5 closest friends, how many of them use marijuana [containing THC]?’).

Cannabis use outcomes

‘Current (past 30-day) cannabis use’ and ‘number of days used in the past 30 days’ were assessed by asking participants to report the number of days they used cannabis in the past 30 days. Those who reported ≥1 day of use were classified as current users; those who reported 0 days of use were classified as nonusers. All participants also indicated their ‘future cannabis use intentions’ (‘How likely are you to try or continue to use marijuana in the next year?’ 1 = not at all to 7 = extremely). Among current cannabis users, a dichotomous variable was created for ‘current cannabis-tobacco dual use’, using data from past 30-day assessments of tobacco use (specifically cigarettes, e-cigarettes, large or little cigars, smokeless and hookah). Blunt use was not included as tobacco use, as it is a specific form of combined tobacco and cannabis use.

Covariates

Sociodemographic covariates included age (continuous variable), race (categorized as White, Black, Asian or other due to group sizes) and ethnicity (Hispanic versus non-Hispanic). Participants were classified as residing in an MSA with a legalized recreational cannabis market (Boston, San Diego, Seattle) or not (Atlanta, Minneapolis, Oklahoma City) in Fall 2020 based on their residential zip code. Cannabis-related covariates among current cannabis users included: ‘age of initiation’ and ‘mode of use’ (‘How do you use marijuana most of the time: smoke in a joint, smoke in a bowl, roll in cigar papers [i.e. blunt], smoke in a waterpipe/bong, ingest/drink, vape, other’).

Data analysis

All analyses were conducted among women (N = 1345) and men (N = 998), separately using SPSS v29. First, bivariate analyses (i.e. Chi-square tests, one-way ANOVAs) examined associations between sexual orientation (i.e. heterosexual, gay/lesbian, bisexual) and cannabis use outcomes, perceived risks and social norms and sociodemographic covariates. Due to non-normality of days of cannabis use (Kolmogorov-Smirnov = 0.17, P < 0.001), negative binomial regressions were examined for all models predicting days of cannabis use. Multivariable regression models (controlling for sociodemographics) examined associations between correlates of interest and: (i) current cannabis use (logistic) among all participants; (ii) future cannabis use intentions (linear) among nonusers; (iii) number of days used in the past 30 days (negative binomial) among current users and (iv) current cannabis-tobacco dual use (logistic) among current users. Secondary cannabis-use characteristics (i.e. age of initiation, mode of use) were included as covariates in models among cannabis users.

Results

Participant characteristics

In this sample of 2476 participants, primary analyses excluded 133 participants who reported ‘other’ or ‘prefer not to answer’ for sexual orientation or gender (who were characterized to better understand this group, as detailed below). In the total analytic sample (N = 2343; 57.4% women, 42.6% men), among women, 69.1% (N = 930) identified as heterosexual, 6.1% (N = 82) lesbian and 24.8% (N = 333) bisexual. On average, women were 24.65 years old (SD = 4.68), 35.3% racial/ethnic minority and 51.1% resided in an MSA with legalized recreational cannabis. Among men, 77.3% (N = 771) identified as heterosexual, 13.8% (N = 137) gay and 8.9% (N = 89) bisexual. On average, men were 24.74 years old (SD = 4.73), 37.3% racial/ethnic minority and 52.8% resided in an MSA with legalized recreational cannabis (Table I).

Cannabis use among women

Among women, 39.5% reported current cannabis use. Current users reported using an average of 14.03 days (SD = 11.91) of the past 30 days (average of 3.14 [SD = 7.81] times per day), and 43.1% reported current cannabis-tobacco dual use (Table I).

Bivariate analyses are reported in Table I. Regarding theory-based correlates, lesbian and bisexual (versus heterosexual) women reported lower perceived harm and addictiveness and greater social acceptability and peer use. Bisexual (versus lesbian) women reported greater social acceptability and peer use. In addition, supplementary analyses examining sexual orientation and sociodemographic correlates of cannabis-related perceived risks and social norms indicated that bisexual and lesbian women reported lower perceived harm (bisexual: B = –0.69, SE = 0.11; lesbian: B = –0.49, SE = 0.19) and addictiveness (bisexual: B = –0.32, SE = 0.12; lesbian: B = –0.52, SE = -.21) and greater perceived social acceptability (bisexual: B = 0.69, SE = 0.11; lesbian: B = 0.27, SE = 0.19) and peer cannabis use (bisexual: B = 0.93, SE = 0.11; lesbian: B = 0.55, SE = 0.19) relative to heterosexual women after accounting for sociodemographics (see Supplementary Table 1).

Multivariable logistic regression indicated that lesbian (aOR = 2.19, 95% CI = 1.38–3.48) and bisexual (aOR = 2.36, 95% CI = 1.82–3.06) (versus heterosexual) women, those with lower perceived harm (aOR = 0.84, 95% CI = 0.76–0.93) and addictiveness (aOR = 0.88, 95% CI = 0.81–0.96), and those with greater perceived social acceptability (aOR = 1.13, 95% CI = 1.02, 1.25) and peer cannabis use (aOR = 1.93, 95% CI = 1.75–2.12) had higher odds of reporting current use. Among nonusers, bisexual (versus heterosexual) women (B = 0.74, SE = 0.17) and those with greater peer cannabis use (B = 0.39, SE = 0.05) reported greater use intentions (Table II). Among current users, multivariable regression found that those identifying as bisexual (B = 0.82, SE = 0.09) (versus heterosexual), reporting lower harm (B = –0.16, SE = 0.03) and greater peer use (B = 0.11, SE = 0.03) reported more days of cannabis use, and that bisexual (aOR = 1.42, 95% CI = 1.07, 1.62) (versus heterosexual) women, those with lower perceived harm (aOR = 0.79, 95% CI = 0.67, 0.93) and those with greater peer use (aOR = 1.15, 95% CI = 1.07, 1.21) had higher odds of cannabis-tobacco dual use (Table III).

Cannabis use among men

Among men, 30.0% reported current cannabis use. Current cannabis users reported using an average of 14.41 days (SD = 11.50) in the past 30 days (average of 3.60 [SD = 5.65] times per day), and 57.1% reported current cannabis-tobacco dual use (Table I).

Bivariate analyses are reported in Table I. Regarding theory-based correlates, gay (versus heterosexual) men reported lower perceived addictiveness, whereas gay and bisexual men reported greater social acceptability and peer use. In addition, supplementary analyses examining sexual orientation and sociodemographic correlates of cannabis-related perceived risks and social norms indicated that bisexual and gay men reported greater perceived social acceptability (bisexual: B = 0.58, SE = 0.20; gay: 0.54, SE = 0.17) and peer cannabis use (bisexual: B = 0.52, SE = 0.19; gay: B = 0.43, SE = 0.16) relative to heterosexual men after accounting for sociodemographics; sexual orientation was not associated with perceived harm or addictiveness (see Supplementary Table 1).

Multivariable logistic regression indicated that bisexual (aOR = 2.34, 95% CI = 1.47–3.73) (versus heterosexual) men, those with lower perceived harm (aOR = 0.85, 95% CI = 0.76–0.96) and those with greater peer use (aOR = 2.11, 95% CI = 1.88–2.38) had higher odds of cannabis use; among nonusers, those reporting lower perceived harm (B = –0.15, SE = 0.05) and greater social acceptability (B = 0.10, SE = 0.04) and peer use (B = 0.14, SE = 0.06) reported greater use intentions (Table II). Among current users, multivariable regression indicated that those reporting greater peer use (B = 0.17, SE = 0.04) reported more days of use and that gay (aOR = 0.34, 95% CI = 0.19, 0.62) (versus heterosexual) men had lower odds of dual cannabis-tobacco use (Table III).

Sub-analyses: ‘other’ and ‘prefer not to answer’ reports for sexual orientation and/or gender

Of the 133 participants excluded from primary analyses, 43 (32.3%) reported ‘other’ for sexual orientation (26 queer, 13 asexual, 4 not sure), 37 (27.8%) reported ‘prefer not to answer’ and 53 reported ‘other’ for gender (who did not report ‘other’ or ‘prefer not to answer’ for sexual orientation; see Supplementary Table 2 for summary of sociodemographics and cannabis use outcomes). Of those reporting ‘other’ sexual orientation, 27.3% reported current cannabis use; current users reported an average of 16.67 (SD = 11.28) days of use, and 26.7% reported cannabis-tobacco dual use. Of those reporting ‘prefer not to answer’ to sexual orientation, 37.8% reported current cannabis use; current users reported an average of 14.92 (SD = 12.02) days of use, and 71.4% reported cannabis-tobacco dual use. Of those reporting ‘other’ gender, 56.6% reported current cannabis use; current users reported an average of 15.43 (SD = 11.31) days of use, and 53.3% reported cannabis-tobacco dual use.

Discussion

The current study expands upon research investigating cannabis use among specific sexual orientation-by-gender subgroups of YAs across a range of cannabis use outcomes, including current cannabis use, future use intentions among nonusers and higher-risk use characteristics (more frequent use, dual use with tobacco) among current users. Moreover, theoretically grounded predictors, including perceived risk (harm, addictiveness) and social norms (social acceptability, peer use), were shown to play key roles in cannabis use and use intentions across both men and women. In the current sample, rates of current cannabis use were ∼54% for SM women, relative to ∼33% for heterosexual women. Rates of current use ranged from ∼45% (gay) to ∼58% (bisexual) for SM men, relative to ∼37% for heterosexual men. Rates of cannabis use among both SM and heterosexual women and men were higher in the current study relative to prior studies, which report current use rates of ∼25% among SM women relative to ∼6% among heterosexual women and ∼30% among SM men relative to ∼12% among heterosexual men [7].

Current findings indicated that while all SM women displayed higher odds of current cannabis use, only bisexual women were more likely to engage in more potentially problematic cannabis use outcomes (i.e. more days of use, cannabis-tobacco dual use [41–43]), even after controlling for a variety of other factors related to cannabis use. These results are inconsistent with some prior literature suggesting that only bisexual (but not lesbian) women displayed higher odds of current cannabis use and that both bisexual and lesbian women reported more days of cannabis use, relative to heterosexual women [6–8]. However, current findings are consistent with the tobacco and other substance use literature, which generally suggests that bisexual women display higher odds of substance use relative to both lesbian and heterosexual women [5, 21, 44, 45]. Results from this study suggest that efforts focused on preventing cannabis use initiation should be targeted toward all SM women, whereas intervention efforts focused on reducing more frequent use and dual use with tobacco may be most important or relevant for bisexual women, specifically.

Moreover, perceived risk and social norms correlated with cannabis use among women. While both bisexual and lesbian (versus heterosexual) women reported lower perceived risk and greater social norms, bisexual women also reported greater social norms relative to lesbian women. Prior research suggests that bisexual women may perceive greater substance use norms (versus heterosexual and lesbian women) due to higher use rates among individuals most similar to them—other bisexual women [46]. In addition, lower perceived harm and greater peer use were uniquely associated with greater use frequency and cannabis-tobacco dual use, and greater peer use was uniquely associated with use intentions. These associations may reflect that greater peer cannabis use may be uniquely associated with disparities in more problematic cannabis use behaviors and future use intentions among bisexual women. Individuals with more cannabis-using peers may have greater opportunities to use cannabis or may be more likely to access cannabis in the future. Additionally, greater peer cannabis use among bisexual women may be a proxy for frequenting LGBTQ+ safe spaces (e.g. LGBTQ+ bars, PRIDE events), which offer a sense of belonging along with greater exposure to substance use, including the use of both cannabis and tobacco [47–50].

Fewer cannabis-related disparities emerged among men, with those identifying as bisexual (versus heterosexual) reporting higher odds of any past 30-day cannabis use only. Previous findings regarding disparities in cannabis use among men have been inconsistent, with some studies suggesting that both bisexual and gay men report higher rates of current cannabis use relative to heterosexual men [7]. Neither gay nor bisexual men displayed increased likelihood of any other cannabis outcomes relative to heterosexual men, which is inconsistent with some prior work highlighting more frequent cannabis use among gay relative to heterosexual men [7]. Interestingly, gay (versus heterosexual) men displayed ‘lower’ odds for cannabis-tobacco dual use, which is consistent with recent tobacco literature indicating that gay YA men displayed lower odds for any tobacco (i.e. cigarette, e-cigarette, cigar, hookah) use relative to heterosexual YA men [21]. Thus, while bisexual men may be more likely to experiment with cannabis, they may be less likely to engage in more problematic cannabis use behaviors. Bisexual men may use cannabis intermittently when socializing with LGBTQ+ peers in which cannabis-related norms are more present [35]. However, these use patterns may not generalize outside of these settings.

Indeed, relative to heterosexual men, gay men reported lower perceived addictiveness and both gay and bisexual men reported greater social norms. Peer cannabis use was, in turn, associated with higher odds of current cannabis use. However, peer use was also associated with greater frequency of use and use intentions, and social acceptability was associated with greater use intentions. Thus, while gay and bisexual men displayed greater risk factors for more frequent days of use and future cannabis use intentions, they were not more likely to engage in such behaviors. Such findings may suggest that, relative to SM women, SM men may experience protective factors (e.g. greater social support, less rejection and discrimination [51, 52]) that may buffer the impact of perceived risks and social norms on problematic cannabis use. Future research should explore the role of these factors on associations among sexual orientation, perceived risks, social norms and cannabis use behaviors.

Current findings emphasize the importance of examining sexual orientation-by-gender subgroup differences in not only cannabis use rates, but also frequency of use, dual use with tobacco and future use intentions, as well as risk correlates that contribute to cannabis use outcomes. A uniform approach to reducing cannabis use may be insufficient to reduce sexual orientation-related disparities in cannabis use outcomes. Instead, it may be necessary to develop culturally congruent and inclusive strategies tailored toward specific SM groups, as well as specific cannabis use outcomes and their related risk correlates. For instance, public health messaging tailored toward bisexual women should focus on frequency of cannabis use and dual use with tobacco and may benefit from targeting perceptions of cannabis-related norms. However, messaging tailored toward lesbian women should focus on any cannabis use and on risk perceptions.

Limitations

The current study has several limitations. First, there may be limited generalizability to other US YAs, given targeted recruitment of tobacco users (who may also be more likely to use cannabis [53]). Thus, cannabis use rates overall and among subgroups (i.e. SM and heterosexual YAs), which were higher than previous studies [7, 10], should not be interpreted as prevalence rates. In addition, our sample was not representative of national estimates on several key demographics (e.g. greater proportion of our sample identified as SM relative to national estimates) [54]. Second, sexual orientation was assessed at baseline only and cannabis use outcomes and related perceived risks and social norms were assessed at follow-up only. Thus, we are unable to examine bidirectional associations among cannabis use outcomes, perceived risks or social norms. Moreover, we were unable to account for potential changes in sexual orientation from baseline to follow-up. Finally, analyses excluded participants who used ‘other’ labels to identify their sexual orientation or gender, given that sample sizes for these groups were small, and we lacked sufficient rationale regarding how to integrate their responses with existing groups. Future research should utilize wording consistent with the National Academies of Science, Engineering, and Medicine report on measuring sex, gender identity and sexual orientation (i.e. ‘I use a different term’ rather than ‘other’) and provide participants with a write-in option when reporting these aspects of identity [55].

Conclusions

As public health authorities strive to understand cannabis use disparities among vulnerable populations, it is crucial to understand cannabis use outcomes and theory-based correlates among specific subgroups of SMYAs. Expanding upon previous research indicating that SMYAs display elevated rates of cannabis use relative to heterosexual YAs, the current study examined differences in not only cannabis use rates, but also frequency of cannabis use, cannabis-tobacco dual use and future use intentions among heterosexual, gay/lesbian and bisexual women and men, respectively. Bisexual women displayed higher odds of all cannabis use outcomes, whereas lesbian women and gay men displayed higher odds of any past 30-day cannabis use only, even after controlling for other factors related to cannabis use. Both cannabis-related perceptions and social norms may be associated with disparities in cannabis use outcomes, particularly among bisexual women.

Supplementary data

Supplementary data are available at HEAL online.

Author contributions

All authors contributed to the conceptualization of the study and reviewed manuscript drafts. Data analyses were conducted by KR. Initial manuscript drafts were led by KR and CB.

Funding

US National Cancer Institute (R01CA215155-01A1; PI: Berg). National Institute on Drug Abuse (R25DA054015; MPIs: Obasi, Reitzel); American Cancer Society (134 128-IRG-19-142; PI: Romm) to K.R. Oklahoma Tobacco Settlement Endowment Trust contract #R22-03; National Cancer Institute grant awarded to the Stephenson Cancer Center (P30CA225520) to K.R. and A.C. National Cancer Institute (R01CA239178-01A1; MPIs: Berg, Levine; R01CA179422-01; PI: Berg); Fogarty International Center (R01TW010664-01; MPIs: Berg, Kegler); National Institute of Environmental Health Sciences/Fogarty (D43ES030927-01; MPIs: Berg, Caudle, Sturua); National Institute on Drug Abuse (R01DA054751-01A1; MPIs: Berg, Cavazos-Rehg) to C.B.

Conflict of interest statement

None declared.

Ethical Approvals

Institutional Review Board approvals were obtained from George Washington University.

References