• Alcohol use on a given day was associated with higher odds of marijuana use that day.
  • The alcohol and marijuana link was stronger for participants with low perceived risk.
  • The daily alcohol and marijuana association was weaker for college students.



Studies looking at the association between perceived risk and simultaneous use of alcohol and marijuana are scarce. The present study has three purposes: (1) to examine the association between alcohol and marijuana use at the daily level; (2) to document how this association varies by the perceived risk of using alcohol and marijuana simultaneously; (3) to test whether the association varies by college attendance.


89 young adults (Mean Age = 18.3 years, SD = 0.5) participated between October 2012 and May 2013. Participants completed a 30-min survey followed by 14 brief daily surveys in each of three waves.


Alcohol use on a given day was associated with increased odds of marijuana use that day, especially among young adults with lower perceived risk. For college students, the association between alcohol and marijuana was weaker than for non-students.


Alcohol and marijuana use were associated at a daily level, especially among young adults with lower perceived risk and those who were not attending college.


Marijuana use;Alcohol use;Young adults;Perceived risk;Daily

1. Introduction

A recent comprehensive review summarizing marijuanas negative health effects highlights particular risks for adolescents and young adults (Volkow et al., 2014). Marijuana use in adolescence can be detrimental for the functional connectivity of the brain (Zalesky et al., 2012), and frequent consumption can lead to a decline in IQ (Meier et al., 2012). Early marijuana use has been linked to poor school performance and a higher risk of dropping out of school (Bray, J.W., et al., 2000 ;  Lynskey, M. and Hall, W., 2000). The risk of motor vehicle accidents is also greater when people drive under the influence of marijuana, and this risk is accentuated for those under the age of 21 and when marijuana and alcohol are used in combination (Peck et al., 2008).

1.1. Concurrent alcohol and marijuana use

Previous research has found that marijuana use is associated with underage alcohol use and binge drinking (Donovan, J.E., et al., 1999; Patrick, M.E. and Schulenberg, J.E., 2010 ;  Patrick, M.E., et al., 2013). In young adults and college students, alcohol use has also been linked to drug use. For example, in a study of college students, almost all of the past 30-day marijuana users reported having used other substances, including alcohol and tobacco (Mohler-Kuo et al., 2003). However, the majority of this research documents these associations between people; that is, people who use alcohol also tend to use marijuana. These between-person studies do not capture whether alcohol and marijuana are used at the same time, which has important implications for overlapping effects and consequences.

Less is known about the extent to which using alcohol on a given day may predict using marijuana that day. In a study using daily surveys of first-year college students, alcohol use on a given day had a positive but non-significant effect on illegal drug use that day (Neal and Fromme, 2007). Previous research has not yet examined the link between alcohol use and marijuana use among a more general sample of young adults both attending and not attending college.

1.2. Perceived risk

Perceived risk plays an important role in substance use. As part of the transition to adulthood and the identity formation process, experimentation is common among some people (Dworkin, 2005). However, the degree of experimentation with different substances can vary depending on the perceived risk by the users. Perceived risk among young adults has been associated with trajectories of marijuana use (Bachman et al., 1988) and cocaine use (Bachman et al., 1990). Additionally, an inverse correlation between marijuana use and the perception of its risk among adolescents has been observed for the most part of the last four decades (Johnston et al., 2014 ;  Volkow, N.D., et al., 2014). Very little is known regarding the perceived risks associated with using alcohol and marijuana simultaneously, and how perceptions of simultaneous use may be associated with using both substances on the same day.

1.3. College attendance

It has been well documented that the transition out of high school is a critical period for young adults that involves changes in responsibilities and social environments (Newcomb and Bentler, 1987). In many ways, including substance use, this transition is different for college students than for non-students. Using data from five different sources, O'Malley and Johnston (2002) concluded that the use of alcohol but not marijuana was higher among college students than among non-students. Therefore, college attendance will be examined as a moderator in the current study.

1.4. The current study

The present study uses daily survey data which allow for the investigation of the combined use of alcohol and marijuana on a given day. In addition, the study also includes a measure of the perceived risk of using alcohol and marijuana simultaneously.

The aim of this study is to understand the consumption of alcohol and marijuana among young adults, and the role that perceived risk plays in this use. Specifically, we examined three research questions: (RQ1) Does alcohol use on a given day predict marijuana use on the same day? (RQ2) Is this association moderated by perceived risk of simultaneous alcohol and marijuana use? (RQ3) Is this association moderated by college attendance?

2. Methods

2.1. Procedure

During the spring of 2012, 12th graders from three high schools (rural, suburban, and urban) in the Midwest were recruited to participate in a baseline survey (see also Griffin and Patrick, 2014). Out of 440 eligible students, 318 (72.3%) completed the baseline survey, of whom 300 (94.3%) provided their contact information for follow-up. Of the 300 students, about two-thirds (N = 202) were randomized into an intensive measurement group (IMG). In October 2012 (wave 1), January 2013 (wave 2), and May 2013 (wave 3) young adults in the IMG received a 30-min web survey followed by 14 consecutive web-based daily surveys. For each of the 14 daily surveys, participants were sent a link to a questionnaire about the previous day, (e.g., “This survey is about Thursday from the time you woke up until you went to sleep”). At the end of the study, each participant could have a maximum of 42 daily surveys (3-waves × 14-days). Participants who were under age 18 were not invited to participate in the follow-up surveys until they turned 18. Participants provided informed consent once they turned 18; all procedures were approved by the IRB.

The response rates were: 87 out of 193 (45.1%) eligible respondents for wave 1, 70 out of 201 (34.8%) for wave 2, and 69 out of 202 (34.2%) for wave 3. The mean number of completed daily surveys ranged from 6.36 (SD = 6.15) to 8.20 (SD = 5.50) within waves. Across all waves 48.3% of participants completed at least half (i.e., 21) of the daily surveys.

Attrition analysis indicated that participants in the current analysis did not significantly differ from attriters based on gender, age, alcohol use, marijuana use, or perceived risk of alcohol or marijuana. Attriters were more likely than participants to be Black, have parents who did not attend college, and attend an urban high school.

2.2. Participants

For the current analysis, data from 89 participants who completed at least one follow-up wave and at least one daily survey were used. Of 89 participants in this study, 38.2% were male, 78.8% were White, and 70.8% had a parent with at least some college education. The average age at baseline was 18.3 years (SD = 0.5, range = 17.6 to 20.1). About half (46.1%) were enrolled full-time in a four-year college during at least one of the waves.

2.3. Measures

2.3.1. Level 1 (daily level)

Marijuana use indicates whether participants used marijuana or hashish each day (0 = no, 1 = yes).

Number of drinks indicates the number of alcoholic drinks consumed each day, on a scale from 1 drink to 25 or more drinks. For participants not drinking on a given day, this variable was coded as 0.

Weekend days were coded as Thursday, Friday, or Saturday, based on research regarding substance use patterns among young adults ( Del Boca, F.K., et al., 2004 ;  Maggs, J.L., et al., 2011).

2.3.2. Level 2 (wave level)

Wave number was coded as 0 (Wave 1), 1 (Wave 2), and 2 (Wave 3).

Wave-level mean number of drinks was constructed by averaging the number of drinks within each wave.

2.3.3. Level 3 (person level)

College status was a time-invariant variable created using information from the three wave questionnaires, due to lack of variability across waves. Participants were coded as full-time 4-year college students (1) if they indicated that they were attending a 4-year college as a full-time student at any wave; all others were coded as non-students (0).

Perceived risk was measured with the question, “How much do you think people risk harming themselves (physically or in other ways) if they use alcohol and marijuana at the same time, so their effects overlap?” Response options included: (1) no risk, (2) slight risk, (3) moderate risk, (4) great risk. Perceived risk was used as a time-invariant variable (i.e., the mean of responses from the three waves for each person) because there was not enough variability to be modeled at the wave-level.

Person mean number of drinks was created by averaging the number of drinks across all available days.

Demographic variables included gender (female [reference group], male), race/ethnicity (White [reference], non-White), and parent education (parent[s] did not attend college [reference], parent[s] attended at least some college).

2.4. Data analysis plan

To model the nested structure of the data (i.e., daily surveys at level 1, nested within waves at level 2, nested within persons at level 3), a three-level multilevel model (Raudenbush, S.W. and Bryk, A.S., 2002 ;  Snidjers and Bosker, 1994) was used. A Bernoulli distribution with over-dispersion was employed to model the dichotomous outcome (marijuana use). The level-3 variables measuring perceived risk and number of drinks were grand-mean centered; the level-2 variable number of drinks was group-mean centered. All analyses were run in HLM 7.0 (Raudenbush et al., 2011) using full penalized quasilikelihood (PQL) estimation. The simple slopes method was used to probe the significant interactions (Bauer, D. J. and Curran, P. J., 2005 ;  Preacher, K. J., et al., 2006).

3. Results

3.1. Preliminary analysis

The intraclass correlation coefficient (ICC) indicated that 72.3% of the variation in marijuana use was person-level variation (level 3), 26.4% was wave-level (level 2), and 1.3% was daily-level (level 1).

3.2. Multilevel models

Table 1 shows the results of the multilevel model predicting marijuana use. At level 3, greater perceived risk of the simultaneous use of marijuana and alcohol was significantly associated with lower odds of marijuana use. Gender, race, parent education, college status, and person mean number of drinks were not significantly associated with marijuana use.

Table 1. Perceived risk predicting marijuana use on a given day.
Marijuana use
Odds ratio (OR) Confidence interval (CI)
Level 3: Person-level (89 subjects)
 Intercept 0.022 [0.002,0.241]
 Male gender 0.152 [0.017,1.366]
 White race 0.387 [0.041,3.640]
 Parent college education 0.585 [0.062,5.522]
 4-year full-time college 0.411 [0.057,2.945]
 Perceived risk 0.231 [0.068,0.782]
 Person mean number of drinks 3.051 [0.919,10.126]
Level 2: wave level (3 time points)
 Wave number 0.855 [0.415,1.773]
 Wave-level mean number of drinks 1.473 [0.453,4.792]
Level 1: daily-level (42 diaries)
 Weekend 0.470⁎⁎ [0.322,0.688]
  4-year full-time college 2.899⁎⁎ [1.653,5.087]
 Number of drinks 1.433⁎⁎ [1.282,1.603]
  Perceived risk 0.705 [0.612,0.812]
  4-year full-time college 0.833 [0.704,0.985]

Note. The indented columns under the “Weekend” and “Number of drinks” predictors for level 1 represent interaction effects.

⁎. p < .05.

⁎⁎. p < .001.

At level 2, wave number and mean wave-level number of drinks were not significant predictors of marijuana use on a given day.

At level 1, the interaction between weekend days and college status was significant, where non-students had lower odds of consuming marijuana on weekends (OR = 0.47, p < .001) and students had a non-significant trend of higher odds of consuming marijuana on weekends (OR = 1.37, p = .37). With respect to RQ1, the odds of using marijuana increased when participants reported consuming a greater number of drinks that day. With respect to RQ2, perceived risk moderated this association. Specifically, the association between alcohol use and marijuana use at the daily level was significantly stronger among individuals with lower levels of perceived risk (OR = 1.63, p < .001). With respect to RQ3, college status also moderated the association between alcohol and marijuana use. The association between alcohol use and marijuana use at the daily level was stronger among individuals who were not in college (OR = 1.43, p < .001).

4. Discussion

Our findings suggest that alcohol use predicts marijuana use at the daily level. This expands existing work based on between-person findings. For participants reporting a lower perceived risk of the combined use of alcohol and marijuana, this association was stronger. That is, young adults who perceived simultaneous alcohol and marijuana use as less risky were more likely to use marijuana on days they had a greater number of drinks, compared to people who perceived simultaneous alcohol and marijuana use to be riskier. This expands previous research documenting that the perceived risk of marijuana use plays an important role in substance use (Bachman, J.G., et al., 1988 ;  Dworkin, J., 2005). Special attention should be paid to young adults who perceive the combined use of alcohol and marijuana as having a low risk.

With respect to differences based on college attendance, non-students were more likely to use marijuana on days they consumed a greater number of drinks and more likely to use marijuana on weekdays, compared to college students. This suggests that young adults who are not in college may have particular risks for substance use that should be addressed.

Limitations of this study include attrition and the fact that participants were sampled from a small number of high schools. Future research could extend this work to follow adolescents for longer periods of time and to include older participants as substance use reaches peak levels in young adulthood.

Role of funding source

The National Institute on Drug Abuse (NIDA) funded the study and the preparation of this manuscript (R21 DA031356 to M.E.P). NIDA had no further role in study design; in the collection, analysis and interpretation of data; in the writing of the report; or in the decision to submit the paper for publication.


G. Yeomans-Maldonado analyzed the data and wrote the initial draft of the article. M. Patrick designed the study, and provided feedback to the initial draft and all versions of the article. This included re-writing some sections of the manuscript as well as providing bibliographical references to strengthen the introduction section. All authors have contributed to and have approved the final manuscript.

Conflict of interest

No conflict declared.


  1. Bachman, J.G., et al., 1990 J.G. Bachman, L.D. Johnston, P.M.O. Malley; Explaining the recent decline in cocaine use among young adults : Further evidence that perceived risks and disapproval lead to reduced drug use; (1990)
  2. Bachman, J.G., et al., 1988 J.G. Bachman, L.D. Johnston, P.M.O. Malley, R.H. Humphrey; Explaining the recent decline in marijuana use: Differentiating the effects of perceived risks; Disapproval, and general lifestyle factors* (1988), pp. 92–113
  3. Bauer, D. J. and Curran, P. J., 2005 D.J. Bauer, P.J. Curran; Probing interactions in fixed and multilevel regression: Inferential and graphical techniques; Multivariate Behavioral Research, 40 (2005), pp. 373–400
  4. Bray, J.W., et al., 2000 J.W. Bray, G.A. Zarkin, C. Ringwalt, Q. Junfeng; The relationship between marijuana initiation and dropping out of high school; Health Economics, 9 (2000), pp. 9–18
  5. Del Boca, F.K., et al., 2004 F.K. Del Boca, J. Darkes, P.E. Greenbaum, M.S. Goldman; Up close and personal: Temporal variability in the drinking of individual college students during their first year; Journal of Consulting and Clinical Psychology, 72 (2004), pp. 155–164
  6. Donovan, J.E., et al., 1999 J.E. Donovan, R. Jessor, F.M. Costa; Adolescent problem drinking: Stability of psychosocial and behavioral correlates across a generation; Journal of Studies on Alcohol, 60 (1999), pp. 352–361
  7. Dworkin, J., 2005 J. Dworkin; Risk taking as developmentally appropriate experimentation for college students; Journal of Adolescent Research, 20 (2005), pp. 219–241
  8. Griffin, J. and Patrick, M.E., 2014 J. Griffin, M.E. Patrick; Nonresponse bias in a longitudinal measurement design examining substance use across the transition out of high school; Drug and Alcohol Dependence, 143 (2014), pp. 232–238
  9. Johnston et al., 2014 L.D. Johnston, P.M. O’Malley, R.A. Miech, J.G. Bachman, J.E. Schulenberg; Monitoring the Future National Survey Results on Drug Use, 1975–2013: Overview, Key Findings on Adolescent Drug Use; Institute for Social Research, The University of Michigan, Ann Arbor (2014)
  10. Lynskey, M. and Hall, W., 2000 M. Lynskey, W. Hall; The effects of adolescent cannabis use on educational attainment: a review; Addiction, 95 (2000), pp. 1621–1630
  11. Maggs, J.L., et al., 2011 J.L. Maggs, L.R. Williams, C.M. Lee; Ups and downs of alcohol use among first-year college students: Number of drinks, heavy drinking, and stumble and pass out drinking days; Addictive Behaviors, 36 (2011), pp. 197–202
  12. Meier, M.H., et al., 2012 M.H. Meier, A. Caspi, A. Ambler, H. Harrington, R. Houts, R.S.E. Keefe, K. McDonald, A. Ward, R. Poulton, T.E. Moffitt; From the cover: PNAS Plus: Persistent cannabis users show neuropsychological decline from childhood to midlife; Proceedings of the National Academy of Sciences (2012)
  13. Mohler-Kuo et al., 2003 M. Mohler-Kuo, J.E. Lee, H. Wechsler; Trends in marijuana and other illicit drug use among college students: results from 4 Harvard School of Public Health College Alcohol Study surveys: 1993–2001; Journal of American College Health, 52 (2003), pp. 17–24
  14. Neal, D.J. and Fromme, K., 2007 D.J. Neal, K. Fromme; Event-level covariation of alcohol intoxication and behavioral risks during the first year of college; Journal of Consulting and Clinical Psychology, 75 (2007), pp. 294–306
  15. Newcomb, M.D. and Bentler, P.M., 1987 M.D. Newcomb, P.M. Bentler; Changes in drug use from high school to young adulthood: Effects of living arrangement and current life pursuit; Journal of Applied Developmental Psychology (1987)
  16. O'Malley and Johnston, 2002 P.M. O'Malley, L.D. Johnston; Epidemiology of alcohol and other drug use among American college students; Journal of Studies on Alcohol and Drugs, 14 (2002), pp. 23–29
  17. Patrick, M.E. and Schulenberg, J.E., 2010 M.E. Patrick, J.E. Schulenberg; Alcohol use and heavy episodic drinking prevalence and predictors among national samples of American eighth- and tenth-grade students; Journal of Studies on Alcohol and Drugs, 71 (2010), pp. 41–45
  18. Patrick, M.E., et al., 2013 M.E. Patrick, J.E. Schulenberg, M.E. Martz, J.L. Maggs, P.M. O'Malley, L.D. Johnston; Extreme binge drinking among 12th-grade students in the United States: prevalence and predictors; JAMA Pediatrics, 167 (2013), pp. 1019–1025
  19. Peck, R.C., et al., 2008 R.C. Peck, M.A. Gebers, R.B. Voas, E. Romano; The relationship between blood alcohol concentration (BAC), age, and crash risk; Journal of Safety Research, 39 (2008), pp. 311–319
  20. Preacher, K. J., et al., 2006 K.J. Preacher, P.J. Curran, D.J. Bauer; Computational tools for probing interaction effects in multiple linear regression, multilevel modeling, and latent curve analysis; Journal of Educational and Behavioral Statistics, 31 (2006), pp. 437–448
  21. Raudenbush, S.W. and Bryk, A.S., 2002 S.W. Raudenbush, A.S. Bryk; Hierarchical linear models: Applications and data analysis methods; (2002)
  22. Raudenbush et al., 2011 S.W. Raudenbush, A.S. Bryk, R.T. Congdon; HLM 7: Hierarchical linear and nonlinear modeling; IL, Scientific Software International, Inc., Lincolnwood (2011)
  23. Snidjers and Bosker, 1994 T.A.B. Snidjers, R.J. Bosker; Modeled variance in two-level models; Sociological Methods & Research, 22 (1994), pp. 342–363
  24. Volkow, N.D., et al., 2014 N.D. Volkow, R.D. Baler, W.M. Compton, S.R.B. Weiss; Adverse health effects of marijuana use; The New England Journal of Medicine, 370 (2014), pp. 2219–2227
  25. Zalesky, A., et al., 2012 A. Zalesky, N. Solowij, M. Yücel, D.I. Lubman, M. Takagi, I.H. Harding, V. Lorenzetti, R. Wang, K. Searle, C. Pantelis, M. Seal; Effect of long-term cannabis use on axonal fibre connectivity; Brain, 135 (2012), pp. 2245–2255
Back to Top

Document information

Published on 26/05/17
Submitted on 26/05/17

Licence: Other

Document Score


Views 14
Recommendations 0

Share this document

claim authorship

Are you one of the authors of this document?