Skip to main content
PMC home page
Alcohol and Alcoholism (Oxford, Oxfordshire) logoLink to Alcohol and Alcoholism (Oxford, Oxfordshire)
. 2020 Apr 24;55(4):416–423. doi: 10.1093/alcalc/agaa023

Inclusion of Cannabis Users in Alcohol Research Samples: Screening In, Screening Out, and Implications

Alexandra Venegas 1, Lindsay R Meredith 1, Ziva D Cooper 2,3, Brandon Towns 1, Lara A Ray 1,2,✉
PMCID: PMC7307319  PMID: 32328657

Abstract

Background

Alcohol and cannabis are frequently co-used, as 20–50% of those who drink alcohol report co-using cannabis. This study is based on the argument that alcohol researchers should enroll cannabis users in human laboratory studies of alcohol use disorder (AUD) to strengthen generalizability. This study examines how heavy drinking cannabis users differ from non-cannabis using heavy drinkers.

Methods

In a community sample of non-treatment-seeking heavy drinkers (n = 551, 35% female), cannabis users were identified through: (a) self-reported cannabis use in the past 6 months and (b) positive urine toxicology test for tetrahydrocannabinol (THC). Cannabis users, identified as described previously, were compared with non-cannabis users on demographic and clinical characteristics.

Results

Those who endorsed cannabis use in the past 6 months reported more binge drinking days. Participants who tested positive for THC had higher Alcohol Use Disorder Identification Test scores and more binge drinking days. Younger age and being a tobacco smoker were associated with an increased likelihood of cannabis use in the past 6 months, whereas male gender and being a tobacco use were associated with a greater likelihood of testing positive for THC. Individuals with cannabis use disorder (CUD) endorsed more depression and anxiety and had higher AUD symptom counts than cannabis users without CUD.

Conclusions

The inclusion of cannabis users in AUD samples allows for increased clinical severity. Excluding cannabis users from AUD studies may limit representativeness and expend unnecessary study resources. Lastly, tobacco use may explain a large portion of the effects of cannabis use on sample characteristics.

Short Summary

Alcohol and cannabis are frequently co-used substances. In a sample of non-treatment-seeking heavy drinkers (n = 551, 35% female), cannabis users reported higher alcohol use and higher likelihood of tobacco use than non-cannabis users. Including cannabis users in alcohol research studies will improve representativeness and likely increase clinical severity.

INTRODUCTION

Cannabis remains the most widely used drug among those who drink alcohol (SAMHSA, 2017), such that among those who abuse alcohol, 20–50% also report cannabis co-use (Petry, 2001). Individuals who report using both alcohol and cannabis often use them at the same time (Subbaraman and Kerr, 2015; Midanik et al., 2007), and such simultaneous use is associated with an increased risk for a host of negative outcomes (Volkow et al., 2014), including comorbid psychiatric disorders, poorer clinical treatment outcomes, increases in risky behaviors including heavy drinking and driving while intoxicated, and other adverse social sequelae (Midanik et al., 2007; Brière et al., 2011; Staiger et al., 2013; Metrik et al., 2018; Subbaraman et al., 2017). Further, cannabis use is predictive of both heavy drinking and the development and maintenance of alcohol use disorder (AUD) (Blanco et al., 2016a; Sullivan et al., 1989; Weinberger et al., 2016; Hayley et al., 2017; Regier et al., 1990; Lopez-Quintero et al., 2011), along with poor prognoses of treatment for AUD (Subbaraman, 2016; Sullivan et al., 1989; Mojarrad et al., 2014; Aharonovich et al., 2005). In fact, it has been estimated that 68% of individuals with a current cannabis use disorder (CUD) diagnosis and >86% of those with a lifetime CUD diagnosis will also meet criteria for a lifetime AUD (Agrawal et al., 2007; World Health Organization, 2014). Notably, however, these negative effects are not uniformly found in the literature (Mallett et al., 2019).

Movements to legalize and decriminalize cannabis use have changed the legal and political landscape, leading to an increased availability of cannabis, possibly influencing alcohol, tobacco and cannabis co-use patterns. In fact, the rate of co-use of cannabis and tobacco among tobacco users has increased in states where cannabis has been legalized (Wang and Cataldo, 2016). Despite recent increases in cannabis availability, there is a paucity of information regarding the profiles of regular drinkers who also use cannabis and potential implications for the inclusion of cannabis users in research studies of AUD. However, the high rates of alcohol and cannabis co-use, along with a host of clinical correlates including psychiatric severity, suggest that excluding cannabis users from clinical studies of AUD may result in non-representative samples of drinkers. For example, considering the fact that alcohol and tobacco are commonly consumed concurrently (Falk et al., 2006), it is well established that excluding tobacco users from AUD clinical studies would severely bias sample representativeness. Much as tobacco use has become standard inclusion criteria in alcohol studies, cannabis use may be moving in the same direction. Therefore, in order to elucidate the impact of cannabis and alcohol co-use in clinical research, vis-a-vis their inclusion or exclusion in AUD clinical research, we must first examine the rate of co-use in addition to their clinical correlates in AUD clinical research samples.

Concerns surrounding the extent to which tightly controlled trials generalize to individuals in community settings have persisted across many different areas of clinical research, including anxiety disorders (Goldstein-Piekarski et al., 2016; Hoertel et al., 2012), cannabis dependence (Okuda et al., 2010), major depression in adults (Blanco et al., 2008a; Halvorson and Humphreys, 2015) and children (Blanco et al., 2017), borderline personality disorder (Hoertel et al., 2015), post-traumatic stress disorder (Franco et al., 2016), schizophrenia (Humphreys, 2017), nicotine dependence and other substance use disorders (Blanco et al., 2016b; King et al., 2011; Le Strat et al., 2011; Melberg and Humphreys, 2010; Moberg and Humphreys, 2017; Robinson et al., 2006; Sofuoglu et al., 2000), neurological disorders (Trivedi and Humphreys, 2015) and randomized controlled trials of psychotherapy (Stirman et al., 2003; Stirman et al., 2005). Efforts to increase representativeness of alcohol-using samples while not compromising internal validity of findings has long been a topic of discussion in the field of alcohol research as well (Blanco et al., 2008b; Hoertel et al., 2014; Humphreys et al., 2007; Humphreys and Weisner, 2000; Maisto et al., 2001; Moberg and Humphreys, 2017; Storbjörk, 2014; Velasquez et al., 2000). It is well documented that individuals with AUD are a rather heterogeneous group with complex clinical presentations (Grant et al., 2015). In clinical trials, however, individuals with many medical and psychiatric comorbidities are often excluded in an effort to increase internal validity, ensure participant safety and increase the likelihood of treatment success (Humphreys et al., 2005). These efforts to increase the integrity of research protocols via stringent inclusion/exclusion criteria have been questioned (Van Spall et al., 2007) and have been thought to increase the risk of biases in outcome estimates without gains in statistical power (Humphreys et al., 2008). In fact, Humphreys and Williams (2018) provide a compelling summary of the magnitude of exclusion rates across disorders. They estimate that researchers exclude between 72 and 92% of real-world individuals suffering from anxiety disorders, 75% of those with bipolar disorders and between 75 and 85% of those with depression. This number is especially striking for substance use disorder research, with investigators excluding ~64–96% of those with substance use disorders, including AUD. As such, it might be useful to consider the implications of adjusting inclusion/exclusion criteria in the context of cannabis use within alcohol research studies in order to reduce selection bias, bolster generalizability of findings, and narrow the gap between research and clinical practice. It is also important to consider the logistical implications of such widespread exclusion of individuals within clinical research, as researchers are forced to grapple with the trade-off between tightly controlling inclusion standards so as to increase internal validity, at the cost of increased recruitment efforts, expenditure of study resources and reductions in external validity.

The present study is based on the argument that alcohol researchers should consider allowing cannabis users to enroll in AUD clinical studies to strengthen generalizability of the findings. To inform the field, this study examines how heavy drinking cannabis users differ from non-cannabis using heavy drinkers in terms of demographic and clinical characteristics—the first study to do so within the context of human laboratory research. To do so, we identify cannabis users through two distinct methods: (a) self-reported cannabis use in the past 6 months per the Cannabis Use Disorder Identification Test—Revised (CUDIT) (Adamson et al., 2010) and (b) positive urine toxicology test for tetrahydrocannabinol (THC), and systematically compare their alcohol use, along with their scores on other individual differences measures, to non-cannabis using heavy drinkers. A second goal of the study was to investigate the hypothesis that cannabis-using heavy drinkers report greater frequencies of alcohol use and associated problems than their non-cannabis using heavy-drinking counterparts. Given that a positive toxicology result likely indicates higher frequency of cannabis use, we hypothesized that those who tested positive for THC via toxicology testing would also show higher clinical severity of AUD.

PARTICIPANTS AND METHODS

Participants

The current sample is a combination of subsamples representing research participants from four separate human laboratory studies with similar inclusion criteria and recruitment methods conducted at the University of California, Los Angeles. All four studies recruited community samples of non-treatment-seeking heavy drinkers from the Greater Los Angeles area. Three of these published studies examined pharmacotherapies for alcohol use: naltrexone (n = 199; Ray et al., 2018), ibudilast (n = 138; Ray et al., 2017a) and ivermectin (n = 140; Roche et al., 2016); the fourth study was an alcohol self-administration study (n = 140; Bujarski et al., 2018). The combination of these subsamples resulted in a final sample size for the present study of 551. Participation in multiple studies was not allowed.

Screening procedures

All study procedures were approved by the University of California, Los Angeles Institutional Review Board, and all participants provided written informed consent after receiving a full explanation of the study procedures. Participants were recruited via online and print advertisements. Interested individuals called the laboratory and completed a telephone interview for the following inclusion criteria based on previously defined criteria of hazardous alcohol drinking (US Department of Agriculture, 2015; National Institute on Alcohol Abuse and Alcoholism, 2010; Reinert and Allen, 2007): >4 drinks per occasion or 7 drinks per week for females and >6 drinks per occasion or 14 drinks per week for males, as well as an Alcohol Use Disorder Identification Test (AUDIT) (Saunders et al., 1993) score of 8 of higher for the naltrexone study; >48 drinks per month and a score of 2 or higher on the CAGE questionnaire (Ewing, 1984) for the ibudilast and ivermectin studies; or >7 drinks per week for females and >14 drinks per week for males for the self-administration study.

All studies had the following exclusion criteria: (a) current involvement in treatment programs for alcohol use or have received treatment in the prior 30 days to study participation; (b) use of non-prescription psychoactive drugs or use of prescription medications for recreational purposes, except for cannabis; (c) self-reported lifetime and/or current history of severe mental illness (e.g. bipolar disorder or psychotic disorders); (d) current use of antidepressants, mood stabilizers, sedatives, anti-anxiety medications, seizure medications or prescription painkillers and (e) self-reported current use of contraindicated medical conditions (e.g. chronic liver disease, cardiac disease).

Participants who were deemed eligible after completing the telephone interview were assessed for further exclusionary criteria as part of an in-person assessment as follows: (a) exclusion, for females, if pregnant (as verified by a urine sample), nursing or planning to get pregnant in the next 6 months or refusal to use a reliable method of birth control; (b) exclusion if breath alcohol concentration (BAC) was >0.000 g/dl as measured by the Dräger Inc. Alcotest® 6510 and (c) exclusion if participant provided a positive urine toxicology screen for any drug (other than cannabis), as measured by Medimpex United Inc. 10-panel drug test. Participants in the current analysis passed both the telephone interview and the in-person assessment of BAC, toxicology test and pregnancy screening (if female).

Measures

At the in-person screening visit participants completed a comprehensive battery of individual differences, clinical and substance use measures, including: (a) the CUDIT captured cannabis-related problems; (b) the Timeline Follow-Back (TLFB) (Sobell and Sobell, 1992) interview for the past 30 days measured alcohol and cannabis use; (c) the AUDIT measured alcohol-related problems; (d) the Beck Depression Inventory (BDI-II) (Beck et al., 1996) and the (e) Beck Anxiety Inventory (BAI) (Beck et al., 1988) measured depression and anxiety symptoms, respectively; (f) the Fagerstrom Test for Nicotine Dependence (Heatherton et al., 1991) was used to determine smoking status; and (g) the Structured Clinical Interview for DSM-5 (SCID-5) (adapted from First, 2014) was used to determine current (past 3-month) diagnosis of AUD and lifetime diagnoses of exclusionary psychiatric disorders. Participants who reported using cannabis at least 10 times per month over the previous 3 months also completed the CUD module of the SCID-5.

Data analysis

Comparisons consisted of univariate independent samples t-tests for continuous outcomes or chi-square tests for dichotomous outcomes. This study identified and compared cannabis users versus non-cannabis users based on the following two definitions: (a) self-reported cannabis use over the past 6 months per the CUDIT (n = 286 endorsed cannabis use) and (b) positive urine toxicology screen for THC at the time of assessment (n = 107 tested positive for THC). Cannabis users and non-cannabis users were compared on demographic (i.e. age, gender and education level), psychiatric (i.e. BDI-II and BAI scores) and substance use (i.e. drinking days, drinks per drinking day, binge drinking days, AUDIT score and tobacco smoking status) variables. Slight variation in sample sizes presented in the results and in Tables 1 and 2 are due to missing data generated from the full sample.

Table 1.

Group differences between non-cannabis users and cannabis users identified by CUDIT (i.e. endorsed use in the past 6 months)

Variable Non-cannabis users (n = 252)a,c Cannabis users (n = 286)a,c Test for difference
Age 32.73 (9.92) 28.10 (7.42) t = 6.066, P < 0.001
Gender
 Female (%) 95 (50.53) 93 (49.47) χ2 = 1.52, P = 0.22
Education
 Some college or higher (%) 164 (65.34) 180 (62.94) χ2 = 0.34, P = 0.56
Drinking daysd 15.96 (8.00) 16.14 (7.22) t = −0.28, P = 0.78
Drinks per drinking dayd 5.70 (3.89) 5.81 (3.37) t = −0.36, P = 0.72
Binge drinking daysd 7.70 (7.32) 10.91 (8.57) t = −3.55, P = 0.001
AUDIT 15.61 (7.66) 15.70 (7.13) t = −0.14, P = 0.89
Tobacco smoker
 Yes (%) 83 (32.94) 154 (53.85) χ2 = 23.76, P < 0.001
BDI-II 9.28 (9.50) 9.63 (9.17) t = −0.43, P = 0.67
BAI 5.88 (7.62) 5.92 (6.23) t = −0.077, P = 0.94
Open in a new tab

aStandard deviations appear within parentheses for continuous variables.

bInconsistent sample sizes due to missing data generated from the full sample (n = 551).

cGroup identification determined by the CUDIT.

dAssessed by TLFB interview for the past 30 days.

Table 2.

Group differences between non-cannabis users and cannabis users identified by positive urine toxicology screen for THC

Variableb Negative toxicology (n = 440)a,c Positive toxicology (n = 107)a,c Test for difference
Age 30.70 (9.37) 29.42 (7.49) t = 1.44, P = 0.15
Gender
 Female (%) 172 (39.18) 18 (17.14) χ2 = 18.10, P < 0.001
Education
 Some college or higher (%) 294 (67.74) 51 (48.11) χ2 = 14.23, P < 0.001
Drinking daysd 15.46 (7.67) 17.64 (7.72) t = −2.64, P = 0.009
Drinks per drinking dayd 5.46 (3.60) 6.71 (3.71) t = −3.20, P = 0.001
Binge drinking daysd 7.70 (7.32) 10.91 (8.57) t = −3.55, P = 0.001
AUDIT 15.00 (7.47) 17.04 (7.83) t = −2.50, P = 0.013
Tobacco smoker
 Yes (%) 165 (38.19) 72 (67.92) χ2 = 30.53, P < 0.001
BDI-II 9.12 (9.05) 11.53 (10.78) t = −2.12, P = 0.035
BAI 5.78 (6.91) 6.40 (6.92) t = −0.82, P = 0.41
Open in a new tab

aStandard deviations appear within parentheses for continuous variables.

bInconsistent sample sizes due to missing data generated from the full sample (n = 551).

cGroup identification determined by urine toxicology screens.

dAssessed by TLFB interview for the past 30 days.

As a follow-up to the univariate models (i.e. t-tests), multivariate logistic regressions were performed in which the significant univariate predictors were tested concomitantly. Results from the logistic regression analyses are reported as a function of odds ratios (ORs) with 95% confidence intervals (CI). Analyses were performed using SPSS version 25. For all comparisons, statistical significance was set at P < 0.05, and all tests were two-tailed.

RESULTS

Of the 551 total participants who completed the in-person assessment battery, 286 (53.7%) reported using any cannabis in the past 6 months on the CUDIT and/or past 30 days on the TLFB. Of the 551 total participants, 185 reported using cannabis ≥10 times per month in the past 3 months and as such underwent the CUD module of the SCID-5, whereas 366 did not. Of the 185 individuals who underwent the CUD module of the SCID-5, 131 (23.8%) did not meet diagnostic criteria for a CUD, whereas 34 (6.2%), 11 (2.0%) and 9 (1.6%) met diagnostic criteria for mild, moderate and severe CUDs, respectively. All participants also underwent the current (past 3-month) AUD module of the SCID-5, although an AUD diagnosis was not an inclusion criterion for any study; individuals with no, mild, moderate and severe AUD enrolled in the studies. A total of 221 (40.1%) did not meet diagnostic criteria, 152 (27.6%) met criteria for mild, 95 (17.2%) met criteria for moderate and 83 (15.1%) met criteria for a severe current (past 3-month) AUD.

Comparisons across the study source (i.e. the naltrexone, ibudilast, ivermectin or alcohol self-administration study) revealed significant differences in variables of interest across study source (P’s < 0.03). As such, all the multivariate logistic regression analyses described below adjust for study source.

Correlates of self-reported cannabis use per CUDIT

Table 1 presents group comparisons based on self-reported cannabis use (endorsed use in the past 6 months per CUDIT). Compared to non-cannabis users (n = 252), those who reported using cannabis in the past 6 months (n = 286) were younger, were more likely to be cigarette smokers and reported a higher number of binge drinking days (P’s < 0.002). Self-reported use in the past 6 months (per the CUDIT) was significantly correlated with tobacco smoking status (r = 0.21, P < 0.0001). Multivariate logistic regression models following up on these univariate results predicted self-reported cannabis use in the past 6 months (per the CUDIT) as a function of age, binge drinking days and smoking status, while controlling for study source. This model was significant (P < 0.05) and explained 16.5% of the variance in reported cannabis use and correctly classified 64.9% of cases. Age and smoking status remained statistically significant predictors (P’s < 0.05), such that younger age and being a tobacco smoker were associated with an increased likelihood of being classified as a cannabis user (ORs = 0.92 and 2.98, respectively); number of binge drinking days was not a significant predictor (P > 0.05).

Correlates of positive urine toxicology screens for THC

Table 2 presents group comparisons based on toxicology test results. Cannabis users, as determined by a positive urine toxicology screen for THC (n = 107), were more likely to be male, score higher on the AUDIT and the BDI-II, have less educational achievement, identify as cigarette smokers and report heavier drinking in the past 30 days (more drinking days, drinks per drinking days and binge drinking days) than non-cannabis users (n = 440, P’s < 0.035). A positive urine toxicology screen for THC was significantly correlated with various substance use measures, including AUDIT score (r = 0.11, P = 0.013), drinking days in the past 30 days (r = 0.11, P < 0.0001), drinks per drinking day (r = 0.14, P = 0.001), binge drinking days (r = 0.17, P < 0.0001) and tobacco smoking status (r = 0.24, P < 0.0001). A follow-up multivariate logistic regression model predicted cannabis use (i.e. positive urine toxicology screen for THC) as a function of gender, education level, drinking days, drinks per drinking day, binge drinking days, AUDIT score, tobacco smoking status and BDI score (which was log transformed due to the non-normality of the data across all groups), while controlling for study source. This model was also significant (P < 0.05) and explained 18.1% of the variance in positive urine toxicology screens for THC and correctly classified 80.3% of cases. Gender, binge drinking days and tobacco smoking status remained significant predictors of cannabis use (P’s < 0.05), such that male gender, more binge drinking days and being a tobacco smoker were associated with a greater likelihood of testing positive for THC (ORs = 0.40, 1.06 and 2.98, respectively); drinking variables, AUDIT score, log-transformed BDI score and education were not significant predictors (P’s > 0.05).

Exploratory analyses: CUD diagnosis

In order to examine the role of CUD diagnosis, as a comorbidity to AUD, we selected individuals who met diagnostic criteria for current AUD and who reported regular cannabis use (i.e. reported using cannabis ≥10 times per month over the previous 3 months) and therefore completed a diagnostic interview for CUD (n = 130). In this sample, 69.2% (n = 90) met criteria for AUD only and 30.8% (n = 40) met criteria for AUD + CUD. Regarding the CUD diagnosis, 24 individuals met criteria for mild CUD, 10 met for moderate CUD and 6 met for severe CUD. For the purpose of these analyses, and due to the small sample sizes, we collapsed CUD into present versus not, such that all severity levels were culled into a single CUD+ group. We compared the two groups on demographics (i.e. sex, age, education), alcohol use in the past 30 days (i.e. drinking days, total drinks, drinks per drinking day and binge drinking days) and on clinical variables of interest (i.e. depressive symptoms, anxiety symptoms, alcohol withdrawal and AUD symptom count). Univariate analyses (t-tests and chi-squares) revealed that CUD+ group reported significantly higher anxiety symptoms on the BAI (M = 8.97 (SD = 7.74)) than CUD− group (M = 5.98 (SD = 6.72); t(127) = −2.22, P < 0.05). A similar pattern emerged for depressive symptoms, such that the CUD+ group exhibited significantly higher BDI-II scores (M = 15.80 (SD = 11.86)) than the CUD− group (M = 9.61 (SD = 9.00); t(128) = −2.95, P < 0.01). This also held true for AUD symptom count, such that the comorbid group endorsed significantly more AUD symptoms in the diagnostic interview (M = 4.95 (SD = 2.64)) than those without a comorbid CUD diagnosis (M = 3.98 (SD = 2.19); t(128) = −2.19, P < 0.05). No other significant group differences were found in the remaining demographic or alcohol use variables (P’s > 0.10). A multivariate logistic regression analysis, including depression, anxiety and AUD symptom count simultaneously, indicated that the overall model was significant (χ2 = 9.69, P < 0.05), yet no single variable was significant over and above the others.

DISCUSSION

The present study evaluated the clinical characteristics of cannabis versus non-cannabis users within a sample of heavy drinkers in order to elucidate the clinical correlates of co-use and inform clinical research practices for alcohol research studies. Of note, not all participants met diagnostic criteria for an AUD. This likely indicates that the current sample is not likely to be entirely representative of those seeking treatment for AUD; however, it is reasonable to assume that this sample is largely indicative of heavy drinkers who enroll in human laboratory studies of AUD. Additionally, >50% of the sample endorsed using cannabis. As such, the prevalence of cannabis use in heavy drinkers participating in clinical research underscores the issue of representativeness when cannabis users are included or excluded from such studies. Participants who endorsed cannabis use in the past 6 months (which includes those who tested positive for THC) reported higher levels of alcohol use compared to non-cannabis users, as indicated by higher numbers of binge drinking days. Conversely, participants who tested positive for THC at the time of assessment, exhibited higher clinical severity profiles for a host of drinking outcomes (e.g. AUDIT, drinking days, binge drinking days) compared with those who tested negative for THC. These findings are generally consistent with the emerging literature suggesting that the co-use of alcohol and cannabis is associated with greater clinical severity and poorer outcomes (Sullivan et al., 1989; Blanco et al., 2016a; Weinberger et al., 2016; Subbaraman, 2016; Mojarrad et al., 2014; Aharonovich et al., 2005). Follow-up multivariate models suggested that age and smoking status consistently predicted self-reported cannabis use in the past 6 months, over and above other significant predictors. A similar pattern of multivariate results was obtained for predicting positive toxicology test such that smoking status, sex and BDI scores remained significant predictors. It is noteworthy that drinking variables were significant in the univariate models but did not survive the multivariate comparison, which may be, at least in part, due to them accounting for shared variance (i.e. suppression effect). Moreover, a common factor across these analyses is the role of smoking status as consistently associated with higher frequency of cannabis use among participants in our study.

Notably, several other group differences were characteristic of the cannabis-using group but only for those identified through positive toxicology screens, including education, sex and depressive symptomatology. For example, the cannabis-using group had 20% higher proportion of males and 20% less having received any college education, demarcating potentially relevant demographic and socioeconomic factors. Cannabis users also displayed higher levels of depressive symptomatology, which is consistent with literature finding greater rates of psychiatric comorbidities among co-users of alcohol, cannabis and other substances (Midanik et al., 2007) as well as heightened risk for the development of depression among cannabis users specifically (Lev-Ran et al., 2014). Further, these patterns are seen in areas where recreational or even medicinal cannabis use is not legalized, as alleviation of psychiatric symptoms (i.e. anxiety, depression) was reported as the most common reason for medicinal cannabis use in the Southeastern USA (Salazar et al., 2019). This result also was found in the multivariate analyses, such that increases in depressive symptomatology per the BDI-II were associated with a greater likelihood of testing positive for THC on the urine toxicology screen. Given that a positive urine toxicology screen definitively indicates more recent cannabis use, it is also possible that it identifies those with more frequent use as well. In fact, those who tested positive for THC on the urine toxicology screen also reported more frequent cannabis use in the 30 days prior to assessment (M = 17.26, SD = 10.78) than those who did not (M = 1.25, SD = 3.63; t = −14.81, P < 0.0001). These findings indicate that more frequent cannabis and alcohol co-use patterns are associated with higher levels of clinical severity, which holds important implications for inclusion/exclusion criteria for clinical studies of AUD. Specifically, allowing for inclusion of individuals with a positive urine toxicology screen would potentially result in a more clinically severe drinking sample, albeit more likely to be representative of populations of those who seek treatment for AUD in the community (Blanco et al., 2008b; Grant et al., 2015; Humphreys, 2003).

Notably, younger age and higher likelihood of cigarette smoking was consistently different among cannabis users, identified through either self-report or toxicology test. In keeping with the changing landscape for cannabis legalization and access, as well as the changing market of tobacco products targeting youth (e.g. electronic nicotine delivery systems, e-cigarettes) (Fadus et al., 2019), these findings highlight possible cohort effects that may influence the enrollment and generalizability of AUD clinical samples for years to come, such that the inclusion of younger drinkers who are more likely to also be concurrent tobacco and cannabis users, might also increase the clinical severity of study samples. With regard to the multivariate analyses, it is important to note that group differences with regard to alcohol use outcomes are non-significant after accounting for tobacco smoking status. This suggests that tobacco use may be driving some of the observed differences between cannabis users and non-cannabis users in this sample.

While these results are informative, they must be interpreted in light of the study’s strengths and limitations. Strengths include the large data set drawing from multiple human laboratory studies of AUD. Limitations include the studies being conducted in a state in which cannabis is legalized both recreationally and medicinally. Profiles of cannabis use may differ in states that regulate cannabis more restrictively. However, exploring cannabis use as a moderator of outcomes of clinical studies of AUD may prove beneficial, as findings from the present study suggest that although allowing for cannabis use results in a more representative sample of heavy drinkers in community settings, they appear to inherently differ on a host of clinical and demographic correlates from those who drink alcohol only. Another important limitation of the present study is that these data come from samples of non-treatment seekers for AUD. As such, this likely indicates that the present sample is younger and has less severe alcohol problems than those who seek treatment for AUD (Ray et al., 2017b; Rohn et al., 2017). However, whether those who seek treatment for AUD are more likely to be concurrent cannabis users and display similar clinical correlates to those reported herein are beyond the scope of this manuscript.

In conclusion, given that cannabis and alcohol co-users have distinct demographic characteristics, report more severe drinking profiles, have higher rates of nicotine use and report higher depressive symptomatology, it appears that the inclusion of cannabis users in AUD samples allows for higher levels of clinical severity. In fact, the results from the exploratory analyses suggest that affective symptoms, namely depression and anxiety, may be associated with the comorbidity of AUD + CUD, as compared to AUD and regular cannabis use. The comorbidity was also associated with higher symptom count for AUD, possibly indicating a worsened clinical presentation. However, excluding cannabis users from AUD clinical studies may lead to a less representative sample, whereas allowing cannabis use among AUD participants enables researchers to learn more about this sizeable subgroup. Additionally, allowing for the inclusion of cannabis users in alcohol human laboratory research poses a potential cost and efficiency advantage, such that their exclusion would significantly diminish the sampling pool from which to recruit, resulting in longer and therefore more expensive enrollment periods and study timelines.

A potential future direction for research is to compare the compositions of samples of clinical studies of AUD where cannabis concurrent use was explicitly excluded to a sample in which cannabis and alcohol-use was not exclusionary to be able to directly speak to the implications of exclusion versus inclusion of cannabis use within these studies. Additionally, additional clinical correlates that were not explored within the present analyses could also be investigated. Based on the current findings, however, cannabis use could be explored as a putative moderator of clinical outcomes in AUD studies that include co-users. Interestingly, when examining the impact of nicotine co-use on drinking outcomes, the pharmacotherapy naltrexone was found to significantly reduce drinking only among those who were nicotine co-users (Anton et al., 2018). Thus, understanding characteristics of poly-substance users, which arguably represent the norm as opposed to the exception, can inform the development of targeted interventions.

FUNDING

This work was supported by the National Institute on Alcohol Abuse and Alcoholism (R01AA021744, R21AA022214, and R21AA022752 to L.A.R.) and the UCLA Clinical and Translational Science Institute (UL1TR000124 to L.A.R.).

CONFLICT OF INTEREST STATEMENT

None declared.

References

  1. Adamson SJ, Kay-Lambkin FJ, Baker AL et al. (2010) An improved brief measure of cannabis misuse: the Cannabis Use Disorders Identification Test-Revised (CUDIT-R). Drug Alcohol Depend 110:137–43. [DOI] [PubMed] [Google Scholar]
  2. Agrawal A, Lynskey MT, Madden PA et al. (2007) A latent class analysis of illicit drug abuse/dependence: results from the National Epidemiological Survey on Alcohol and Related Conditions. Addiction 102:94–104. [DOI] [PubMed] [Google Scholar]
  3. US Department of Agriculture (2015) Dietary guidelines for Americans 2015–2020. Washington, DC: US Department of Health and Human Services and USDA. [Google Scholar]
  4. Aharonovich E, Liu X, Samet S et al. (2005) Postdischarge cannabis use and its relationship to cocaine, alcohol, and heroin use: a prospective study. Am J Psychiatry 162:1507–14. [DOI] [PubMed] [Google Scholar]
  5. Anton RF, Latham PK, Voronin KE et al. (2018) Nicotine-use/smoking is associated with the efficacy of naltrexone in the treatment of alcohol dependence. Alcohol Clin Exp Res 42:751–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Beck AT, Epstein N, Brown G et al. (1988) An inventory for measuring clinical anxiety: psychometric properties. J Consult Clin Psychol 56:893. [DOI] [PubMed] [Google Scholar]
  7. Beck AT, Steer RA, Brown GK (1996) Beck Depression Inventory-II. San Antonio 78:490–8. [Google Scholar]
  8. Blanco C, Hasin DS, Wall MM et al. (2016a) Cannabis use and risk of psychiatric disorders: Prospective evidence from a US national longitudinal study. JAMA Psychiatry 73:388–95. [DOI] [PubMed] [Google Scholar]
  9. Blanco C, Hoertel N, Franco S et al. (2017) Generalizability of clinical trial results for adolescent major depressive disorder. Pediatrics 140:e20161701. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Blanco C, Olfson M, Goodwin RD et al. (2008a) Generalizability of clinical trial results for major depression to community samples: results from the National Epidemiologic Survey on Alcohol and Related Conditions. J Clin Psychiatry 69:1276–80. [DOI] [PubMed] [Google Scholar]
  11. Blanco C, Olfson M, Okuda M et al. (2008b) Generalizability of clinical trials for alcohol dependence to community samples. Drug Alcohol Depend 98:123–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Blanco C, Wall MM, Lindquist MA. et al. (2016b) Generalizability of neuroimaging studies in 5 common psychiatric disorders based on the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC). J Clin Psychiatry 77:e1618–25. [DOI] [PubMed] [Google Scholar]
  13. Brière F, Fallu J-S, Descheneaux A et al. (2011) Predictors and consequences of simultaneous alcohol and cannabis use in adolescents. Addict Behav 36:785–8. [DOI] [PubMed] [Google Scholar]
  14. Bujarski S, Jentsch JD, Roche DJ et al. (2018) Differences in the subjective and motivational properties of alcohol across alcohol use severity: application of a novel translational human laboratory paradigm. Neuropsychopharmacology 43:1891. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Ewing JA. (1984) Detecting alcoholism: the CAGE questionnaire. JAMA 252:1905–7. [DOI] [PubMed] [Google Scholar]
  16. Fadus MC, Smith TT, Squeglia LM (2019) The rise of e-cigarettes, pod mod devices, and JUUL among youth: factors influencing use, health implications, and downstream effects. Drug Alcohol Depend 201:85–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Falk DE, Hsiao-Ye Y, Hiller-Sturmhöfel S (2006) An epidemiologic analysis of co-occurring alcohol and tobacco use and disorders. Alcohol Res 29:162. [PMC free article] [PubMed] [Google Scholar]
  18. First MB. (2014) Structured clinical interview for the DSM (SCID) The Encyclopedia of Clinical Psychology, 1–6.
  19. Franco S, Hoertel N, McMahon K et al. (2016) Generalizability of pharmacologic and psychotherapy clinical trial results for posttraumatic stress disorder to community samples. J Clin Psychiatry 77:e975–81. [DOI] [PubMed] [Google Scholar]
  20. Goldstein-Piekarski A, Williams L, Humphreys K (2016) A trans-diagnostic review of anxiety disorder comorbidity and the impact of multiple exclusion criteria on studying clinical outcomes in anxiety disorders. Transl Psychiatry 6:e847. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Grant BF, Goldstein RB, Saha TD et al. (2015) Epidemiology of DSM-5 alcohol use disorder: results from the National Epidemiologic Survey on Alcohol and Related Conditions III. JAMA Psychiatry 72:757–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Halvorson M, Humphreys K (2015) A review of the nature and impact of exclusion criteria in depression treatment outcome research. Ann Depress Anxiety 2:1058. [Google Scholar]
  23. Hayley AC, Stough C, Downey LA (2017) DSM-5 cannabis use disorder, substance use and DSM-5 specific substance-use disorders: Evaluating comorbidity in a population-based sample. Eur Neuropsychopharmacol 27:732–43. [DOI] [PubMed] [Google Scholar]
  24. Heatherton TF, Kozlowski LT, Frecker RC et al. (1991) The Fagerström test for nicotine dependence: a revision of the Fagerstrom Tolerance Questionnaire. Br J Addict 86:1119–27. [DOI] [PubMed] [Google Scholar]
  25. Hoertel N, Falissard B, Humphreys K et al. (2014) Do clinical trials of treatment of alcohol dependence adequately enroll participants with co-occurring independent mood and anxiety disorders? An analysis of data from the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC). J Clin Psychiatry 75:221–37. [DOI] [PubMed] [Google Scholar]
  26. Hoertel N, Le Strat Y, Blanco C et al. (2012) Generalizability of clinical trial results for generalized anxiety disorder to community samples. Depress Anxiety 29:614–20. [DOI] [PubMed] [Google Scholar]
  27. Hoertel N, López S, Wang S et al. (2015) Generalizability of pharmacological and psychotherapy clinical trial results for borderline personality disorder to community samples. Personal Disord 6:81. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Humphreys K. (2003) Datapoints: do participants in alcoholism treatment outcome studies resemble patients seen in everyday practice? Psychiatr Serv 54:1576–6. [DOI] [PubMed] [Google Scholar]
  29. Humphreys K. (2017) A review of the impact of exclusion criteria on the generalizability of schizophrenia treatment research. Clin Schizophr Relat Psychoses 11:49–57. [DOI] [PubMed] [Google Scholar]
  30. Humphreys K, Harris AH, Weingardt KR (2008) Subject eligibility criteria can substantially influence the results of alcohol-treatment outcome research. J Stud Alcohol Drugs 69:757–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Humphreys K, Weingardt KR, Harris AH (2007) Influence of subject eligibility criteria on compliance with National Institutes of Health guidelines for inclusion of women, minorities, and children in treatment research. Alcohol Clin Exp Res 31:988–95. [DOI] [PubMed] [Google Scholar]
  32. Humphreys K, Weingardt KR, Horst D et al. (2005) Prevalence and predictors of research participant eligibility criteria in alcohol treatment outcome studies, 1970–98. Addiction 100:1249–57. [DOI] [PubMed] [Google Scholar]
  33. Humphreys K, Weisner C (2000) Use of exclusion criteria in selecting research subjects and its effect on the generalizability of alcohol treatment outcome studies. Am J Psychiatry 157:588–94. [DOI] [PubMed] [Google Scholar]
  34. Humphreys K, Williams LM (2018) What can treatment research offer general practice? Lancet Psychiatry 5:295–7. [DOI] [PubMed] [Google Scholar]
  35. King AC, Cao D, Southard CC et al. (2011) Racial differences in eligibility and enrollment in a smoking cessation clinical trial. Health Psychol 30:40. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Le Strat Y, Rehm J, Le Foll B (2011) How generalisable to community samples are clinical trial results for treatment of nicotine dependence: a comparison of common eligibility criteria with respondents of a large representative general population survey. Tob Control 20:338–43. [DOI] [PubMed] [Google Scholar]
  37. Lev-Ran S, Roerecke M, Le Foll B et al. (2014) The association between cannabis use and depression: a systematic review and meta-analysis of longitudinal studies. Psychol Med 44:797–810. [DOI] [PubMed] [Google Scholar]
  38. Lopez-Quintero C, Hasin DS, De Los Cobos JP et al. (2011) Probability and predictors of remission from life-time nicotine, alcohol, cannabis or cocaine dependence: Results from the national epidemiologic survey on alcohol and related conditions. Addiction 106:657–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Maisto SA, Conigliaro J, McNeil M et al. (2001) The relationship between eligibility criteria for participation in alcohol brief intervention trials and other alcohol and health-related variables. Am J Addict 10:218–31. [DOI] [PubMed] [Google Scholar]
  40. Mallett KA, Turrisi R, Trager BM et al. (2019) An examination of consequences among college student drinkers on occasions involving alcohol-only, marijuana-only, or combined alcohol and marijuana use. Psychol Addict Behav 33:331–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Melberg HO, Humphreys K (2010) Ineligibility and refusal to participate in randomised trials of treatments for drug dependence. Drug Alcohol Rev 29:193–201. [DOI] [PubMed] [Google Scholar]
  42. Metrik J, Gunn RL, Jackson KM et al. (2018) Daily patterns of marijuana and alcohol co-use among individuals with alcohol and cannabis use disorders. Alcohol Clin Exp Res 42:1096–104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Midanik LT, Tam TW, Weisner C (2007) Concurrent and simultaneous drug and alcohol use: results of the 2000 National Alcohol Survey. Drug Alcohol Depend 90:72–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Moberg CA, Humphreys K (2017) Exclusion criteria in treatment research on alcohol, tobacco and illicit drug use disorders: A review and critical analysis. Drug Alcohol Rev 36:378–88. [DOI] [PubMed] [Google Scholar]
  45. Mojarrad M, Samet JH, Cheng DM et al. (2014) Marijuana use and achievement of abstinence from alcohol and other drugs among people with substance dependence: a prospective cohort study. Drug Alcohol Depend 142:91–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. National Institute on Alcohol Abuse and Alcoholism (U.S.) (2010) Rethinking drinking: alcohol and your health.
  47. Okuda M, Hasin DS, Olfson M et al. (2010) Generalizability of clinical trials for cannabis dependence to community samples. Drug Alcohol Depend 111:177–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Petry NM. (2001) A behavioral economic analysis of polydrug abuse in alcoholics: Asymmetrical substitution of alcohol and cocaine. Drug Alcohol Depend 62:31–9. [DOI] [PubMed] [Google Scholar]
  49. Ray LA, Bujarski S, Shoptaw S et al. (2017a) Development of the neuroimmune modulator ibudilast for the treatment of alcoholism: a randomized, placebo-controlled, human laboratory trial. Neuropsychopharmacology 42:1776. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Ray LA, Bujarski S, Yardley MM et al. (2017b) Differences between treatment-seeking and non-treatment-seeking participants in medication studies for alcoholism: do they matter? Am J Drug Alcohol Abuse 43:703–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Ray LA, Green R, Roche DJ et al. (2018) Pharmacogenetic effects of naltrexone in individuals of East Asian descent: Human laboratory findings from a randomized trial. Alcohol Clin Exp Res 42:613–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Regier DA, Farmer ME, Rae DS et al. (1990) Comorbidity of mental disorders with alcohol and other drug abuse. JAMA 264:2511–8. [PubMed] [Google Scholar]
  53. Reinert DF, Allen JP (2007) The alcohol use disorders identification test: an update of research findings. Alcohol Clin Exp Res 31:185–99. [DOI] [PubMed] [Google Scholar]
  54. Robinson ML, Schroeder JR, Moolchan ET (2006) Adolescent smokers screened for a nicotine replacement treatment trial: correlates of eligibility and enrollment. Nicotine Tob Res 8:447–54. [DOI] [PubMed] [Google Scholar]
  55. Roche DJ, Yardley MM, Lunny KF et al. (2016) A pilot study of the safety and initial efficacy of ivermectin for the treatment of alcohol use disorder. Alcohol Clin Exp Res 40:1312–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Rohn MC, Lee MR, Kleuter SB et al. (2017) Differences between treatment-seeking and nontreatment-seeking alcohol-dependent research participants: an exploratory analysis. Alcohol Clin Exp Res 41:414–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Salazar CA, Tomko RL, Akbar SA et al. (2019) Medical cannabis use among adults in the southeastern United States. Cannabis 2:53. [PMC free article] [PubMed] [Google Scholar]
  58. Saunders JB, Aasland OG, Babor TF et al. (1993) Development of the alcohol use disorders identification test (AUDIT): WHO collaborative project on early detection of persons with harmful alcohol consumption-II. Addiction 88:791–804. [DOI] [PubMed] [Google Scholar]
  59. Sobell LC, Sobell MB (1992) Timeline follow-back. In Measuring Alcohol Consumption. Totowa, NJ: Humana Press, 41–72. [Google Scholar]
  60. Sofuoglu M, Dudish-Poulsen S, Nicodemus KK et al. (2000) Characteristics of research volunteers for inpatient cocaine studies: focus on selection bias. Addict Behav 25:785–90. [DOI] [PubMed] [Google Scholar]
  61. Staiger PK, Richardson B, Long CM et al. (2013) Overlooked and underestimated? Problematic alcohol use in clients recovering from drug dependence. Addiction 108:1188–93. [DOI] [PubMed] [Google Scholar]
  62. Stirman SW, DeRubeis RJ, Crits-Christoph P et al. (2003) Are samples in randomized controlled trials of psychotherapy representative of community outpatients? A new methodology and initial findings. J Consult Clin Psychol 71:963. [DOI] [PubMed] [Google Scholar]
  63. Stirman SW, DeRubeis RJ, Crits-Christoph P et al. (2005) Can the randomized controlled trial literature generalize to nonrandomized patients? J Consult Clin Psychol 73:127. [DOI] [PubMed] [Google Scholar]
  64. Storbjörk J. (2014) Implications of enrolment eligibility criteria in alcohol treatment outcome research: generalisability and potential bias in 1-and 6-year outcomes. Drug Alcohol Rev 33:604–11. [DOI] [PubMed] [Google Scholar]
  65. Subbaraman MS. (2016) Substitution and complementarity of alcohol and cannabis: a review of the literature. Subst Use Misuse 51:1399–414. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Subbaraman MS, Kerr WC (2015) Simultaneous versus concurrent use of alcohol and cannabis in the National Alcohol Survey. Alcohol Clin Exp Res 39:872–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Subbaraman MS, Metrik J, Patterson D et al. (2017) Cannabis use during treatment for alcohol use disorders predicts alcohol treatment outcomes. Addiction 112:685–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Substance Abuse and Mental Health Services Administration (SAMHSA) (2017) Results from the 2013 National Survey on Drug Use and Health: Summary of National Findings. NSDUH Series H-48, HHS Publication No.(SMA) 14–4863 Rockville, MD: Substance Abuse and Mental Health Services Administration. [Google Scholar]
  69. Sullivan JT, Sykora K, Schneiderman J et al. (1989) Assessment of alcohol withdrawal: the revised clinical institute withdrawal assessment for alcohol scale (CIWA-Ar). Br J Addict 84:1353–7. [DOI] [PubMed] [Google Scholar]
  70. Trivedi RB, Humphreys K (2015) Participant exclusion criteria in treatment research on neurological disorders: are unrepresentative study samples problematic? Contemp Clin Trials 44:20–5. [DOI] [PubMed] [Google Scholar]
  71. Van Spall HG, Toren A, Kiss A et al. (2007) Eligibility criteria of randomized controlled trials published in high-impact general medical journals: a systematic sampling review. JAMA 297:1233–40. [DOI] [PubMed] [Google Scholar]
  72. Velasquez MM, DiClemente CC, Addy RC (2000) Generalizability of project MATCH: a comparison of clients enrolled to those not enrolled in the study at one aftercare site. Drug Alcohol Depend 59:177–82. [DOI] [PubMed] [Google Scholar]
  73. Volkow ND, Baler RD, Compton WM et al. (2014) Adverse health effects of marijuana use. N Engl J Med 370:2219–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
  74. Wang JB, Cataldo JK (2016) Medical marijuana legalization and co-use in adult cigarette smokers. Am J Health Behav 40:205–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  75. Weinberger AH, Platt J, Goodwin RD (2016) Is cannabis use associated with an increased risk of onset and persistence of alcohol use disorders? A three-year prospective study among adults in the United States. Drug Alcohol Depend 161:363–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  76. World Health Organization (2014) Global status report on alcohol and health, 2014. Geneva, Switzerland: World Health Organization. [Google Scholar]

Articles from Alcohol and Alcoholism (Oxford, Oxfordshire) are provided here courtesy of Oxford University Press

RESOURCES