14

Cannabis Use and the Abuse of Other Substances

Since the 1970s, researchers have debated about the role that cannabis may play in the “gateway hypothesis,” which suggests that individuals rarely use certain substances, such as heroin or cocaine, without first having used “gateway” substances, such as alcohol, tobacco, or cannabis (Kandel, 1975; Vanyukov et al., 2012). While some research has shown an association between cannabis use and the subsequent use of other illicit drugs, the predictors of progression from cannabis use to other illicit drugs remain largely unknown (Secades-Villa et al., 2015). Emerging animal studies have begun to explore the hypothesis that cannabis exposure may enhance the susceptibility to the addictive effects of other drugs (Panlilio et al., 2012). Researchers have also begun to explore the “reverse gateway hypothesis.” This hypothesis posits that cannabis use is a reverse gateway to the initiation of other addictive drugs such as nicotine and alcohol (Agrawal et al., 2008).

In the United States, the number of individuals 12 years and older using illicit drugs rose each year between 2002 and 2013. In 2014 alone, the National Survey on Drug Use and Health reported that in this age range 27 million individuals—or almost 1 in every 10 Americans—were found

to have used illicit drugs within the past 30 days, 66.9 million were current users of tobacco, and another 139.7 million were past-month alcohol drinkers (CBHSQ, 2015). With illicit drug use on the rise, the need for understanding and addressing when and how individuals start using illicit drugs is of the utmost importance. Of similar importance is understanding the role that cannabis might play in the use of other addictive substances such as tobacco and alcohol.

The committee responsible for the 1999 Institute of Medicine (IOM)¹ report Marijuana and Medicine: Assessing the Science Base discussed the “gateway hypotheses” but did not make any specific conclusions about its relevance to cannabis use. That report questioned the designation of cannabis as a “gateway” drug because its use is often preceded by underage drinking and tobacco use, and no conclusive evidence supporting a causal link between cannabis use and the use of other illicit drugs was found at that time (IOM, 1999).

In this chapter, the committee reviews the research evidence that most directly addresses the prioritized research questions related to the associations among cannabis use and (1) the initiation of use of other substances, (2) changes in the rates and use patterns of other substances and, (3) and the development of other substance dependence or substance abuse disorder. Due to the time constraints of the study, additional search constraints were added to prioritize the types of studies that would likely produce the clearest research conclusions. For example, literature searches were limited to articles that included the search terms “longitudinal,” “prospective,” and “case-control,” and the committee did not review controlled laboratory studies with cannabis. Although the committee did not find any fair- or good-quality systematic reviews covering these issues, 12 primary articles published since the 1999 IOM report were identified and are reviewed in this chapter.

ABUSE OF OTHER SUBSTANCES

Is There an Association Between Cannabis Use and the Initiation of Use of Other Substances?

Tobacco/Nicotine

Systematic Reviews The committee did not identify a good- or fair-quality systematic review that reported on the association between cannabis use and the initiation of tobacco/nicotine use.

___________________

¹ As of March 2016, the Health and Medicine Division continues the consensus studies and convening activities previously carried out by the Institute of Medicine (IOM).

Primary Literature Mayet and colleagues (2016) conducted a retrospective cohort study of the transitions between tobacco, cannabis, and other illicit drugs initiations. Data on 16,421 adults ages 18 to 34 were collected from two French nationwide health and behavior studies conducted in 2005 and 2010. The data used included the age of initiation of substance use (cannabis, tobacco, alcohol, other illicit drugs), current use, and a number of other variables (e.g., gender, socioeconomic level). A total of 436,206 observations based on yearly measures were provided by the study subjects, including 17,510 transitions from one state of use to another. A Markov multistate model was constructed to examine transitions from cannabis use to the use of other drugs. The model’s results show that the probability of initiating tobacco after cannabis use (10.39 percent) was significantly greater (p <0.001) than the probability of initiating cannabis after tobacco use (3.47 percent). The primary study limitations include potential recall bias on the age of initiation and the usual issues surrounding the self-reporting of substance use.

Mayet and colleagues (2011) analyzed data collected from a cross-sectional survey of 29,393 17-year-old adolescents attending a compulsory military information session to assess transitions from cannabis use to the use of other substances. Data from study participants were captured via a self-administered questionnaire on substance use; thus, participants were considered followed from birth through 2011 by way of recall data. Substance use was categorized as “no lifetime use of tobacco and cannabis,” “tobacco initiation only,” “cannabis initiation only,” “daily use of tobacco only,” “daily use of cannabis only,” or “daily use of both tobacco and cannabis” (Mayet et al., 2011, p. 1102). A Markov multistate model was constructed to examine the transition states among the first-substance-use cohorts from no use/initial substance use to other substance use states.

Study participants were more likely to use tobacco (72.2 percent) than cannabis (49.4 percent), and only 2 percent of those using both tobacco and cannabis reported having used cannabis before tobacco (Mayet et al., 2011). With respect to transitions from initial substance use, the risk of initiating tobacco use from no lifetime use was 17.6 times greater (95% confidence interval [CI] = 16.5–18.9) than first initiating cannabis use. Among individuals initiating use with cannabis, the transition to first tobacco use was 3.2 times greater (95% CI = 2.9–3.6) than the transition from no lifetime use of cannabis to first tobacco use (Mayet et al., 2011). However, the transition of first tobacco use to cannabis was 42.1 times greater (95% CI = 39.3–45.1) than the transition from no lifetime use of tobacco to first cannabis use. The transition from daily use of cannabis to daily use of both cannabis and tobacco was 3.0 times greater (95% CI = 2.0–4.4) than the transition from daily tobacco use to daily use of both cannabis and tobacco (Mayet et al., 2011). The authors also found that

cannabis initiation did not increase the risk of a tobacco user transitioning to a daily cannabis smoker. The study’s limitations include potential problems with recall bias and self-reporting of substance use.

Opioids

Systematic Reviews The committee did not identify a good- or fair-quality systematic review that reported on the association between cannabis use and the initiation of use of opioids.

Primary Literature In the retrospective cohort study described earlier, Mayet and colleagues (2016) also explored the transition from cannabis use to the use of other illicit drugs. They found that the probability of initiating other illicit drugs after cannabis did not differ significantly from the probability of starting with other illicit drugs.

Mixed Drug Use

Systematic Reviews The committee did not identify a good- or fair-quality systematic review that reported on the association between cannabis use and the initiation of use of other substances.

Primary Literature Novins and Barón (2004) reported on risk factors for the initiation of substance use and transition to other substance use among American Indian adolescents living west of the Mississippi. Survey data collected as part of the Voices of Indian Teens longitudinal study from 1993 to 1996 were used to calculate the conditional probability that an adolescent who reported lifetime use of cannabis (Stage 1) would progress to report a lifetime use of stimulants, sedatives, cocaine, or other drugs such as hallucinogens, phencyclidine, or heroin (Stage 2).

For analysis purposes, the initial sample of 2,356 adolescents was reduced to 1,244 adolescents due to exclusions related to continued lifetime abstention or transition to Stage 2 before the study began and to inconsistent responses between the two waves of data collection, as well as those lost to follow-up (Novins and Barón, 2004). The hazard ratio (HR) for the progression of cannabis use (Stage 1) to a harder substance (Stage 2) was 2.737 (p <0.01). The authors noted that the study had a number of limitations, including generalizability to other populations, the self-reporting of substance use data, an inability to include tobacco use in the analysis because the survey did not differentiate between ceremonial and non-ceremonial tobacco use, and the potential for underestimating the results because of the potential bias created by individuals lost to

follow-up who may have had different (higher) patterns of substance use than those remaining in the study.

Discussion of Findings

The small number of studies reviewed provide limited evidence that cannabis use increases the rates of initiation of other drug use, mainly tobacco. Two studies had relatively large samples. The data do not provide compelling evidence that cannabis is associated with the initiation of other drugs of abuse, although this is one possibility. Other possibilities that could explain these findings include easier access to cannabis than to other illicit substances and common risk factors for both cannabis use and the use of other substances. Although cannabis use is associated with increased odds of transitioning to tobacco use relative to non-cannabis users, tobacco use was associated with far greater odds of transitioning to cannabis use relative to non-tobacco users. These data highlight tobacco use as a key risk factor for the initiation of cannabis use.

Is There an Association Between Cannabis Use and the Rates and Use Patterns of Other Substances?

Alcohol

Systematic Reviews The committee did not identify a good- or fair-quality systematic review that reported on the association between cannabis use and the rates and use patterns of drinking alcohol.

Primary Literature Buu and colleagues (2015) conducted a secondary data analysis of eight waves of data collected from 850 high-risk adolescents participating in the longitudinal Flint Adolescent Study to assess risk and protective factors for substance use and other health risk behaviors through adulthood (i.e., ages 14–24 years). The impact of early or later onset (i.e., age at first use) and of the quantity and frequency of cannabis use on heavy drinking were specific research questions. A linear mixed model was used to determine the longitudinal effects of nicotine and marijuana on heavy drinking while controlling for the early onset of alcohol use. Model results indicate that both early onset cannabis users (β, 0.2263; standard error [SE] = 0.0445; p <0.0001), late onset cannabis users (β, 0.1838; SE = 0.0461; p <0.0001), and those who used cannabis more

frequently (β, 0.2667; SE = 0.0119; p <0.0001) were all at a higher risk of heavy alcohol drinking than those who did not use cannabis at all (Buu et al., 2015). Among this population, about 80 percent of the study participants were black and had grade point averages of 3.0 or below and thus are not representative of the general youth population. Furthermore, the lifetime prevalence of substance use was higher in the study population than in the general population. The impact of cannabis use on nicotine use was not reported.

Opioids

Systematic Reviews The committee did not identify a good- or fair-quality systematic review that reported on the association between cannabis use and the rates and use patterns of opioids.

Primary Literature In a longitudinal study of a random sample of 120 adolescents ages 12 to 18 years who were admitted to a level 1 trauma center or an emergency department for injury, Whiteside and colleagues (2016) found that preinjury cannabis use was an independent predictor of continued prescription opioid use up to 12 months after discharge (relative risk, 1.69; 95% CI = 1.09–2.6). Cannabis use was assessed via a single-item question regarding cannabis use (yes/no) in the year before the injury, and the use of a range of prescription opioids (codeine, hydrocodone, oxycodone, etc.) was assessed and categorized as yes or no at months 2, 5, and 12. At 1 year post injury, 12.5 percent of adolescents were still using prescribed opioids. The study’s limitations include the use of self-reported data to determine preinjury cannabis use and opioid use, the reliance on a small study sample, and the fact that the sample was collected at an urban academic trauma center, which thus limited the generalizability of the findings.

Tobacco/Nicotine

Systematic Reviews The committee did not identify a good- or fair-quality systematic review that reported on the association between cannabis use and the rates and use patterns of tobacco and nicotine dependence.

Primary Literature Agrawal and colleagues (2008) studied women cannabis users and patterns of smoking and nicotine dependence. Data were collected as part of the Missouri Adolescent Female Twin Study (MOAFTS), a cohort study of 3,787 young adult twin females ages 18 to 29 years, who were originally interviewed in 1994–1999 and subsequently reinterviewed in 2002–2005. Data collection included lifetime cannabis

use (ever used cannabis and other measures of frequency and use) and cannabis dependence (determined by a lifetime history of one or more of four Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition [DSM-IV] abuse critera or one or more of six DSM-IV dependence criteria). Regular cigarette smoking among participants was determined by responding positively to having smoked 100 or more cigarettes lifetime and smoking 20–99 cigarettes at least once per week for a period of 2 months or longer. Nicotine dependence was defined using the seven DSM-IV dependence criteria, with at least three symptoms clustering within the same 12-month period. Data on a number of covariates were also collected, including measures of behavioral disinhibition, negative affect regulation, and other measures of psychopathology. In this sample, 44.2 percent of the participants were cannabis users, 34.7 percent were classified as regular cigarette smokers, and 17.4 percent were designated as nicotine dependent based on DSM-IV criteria. It is also important to note that only 6.8 percent of participants reported smoking their first cigarette before using cannabis for the first time.

Survival analyses were used to examine whether women who smoked cannabis were at an increased risk of moving from experimenting with smoking (but not first time smoking) to becoming a regular smoker. After adjusting for covariates, the results indicate that women with a prior history of cannabis use were 4.4 times more likely (HR, 4.41; 95% CI = 3.57–5.44) to transition from experimenting with smoking to becoming regular smokers. An additional analysis was conducted to assess spurious associations caused by the onset of cannabis use and regular smoking in the same year. The results of this analysis showed a diminished effect size; the effect size of the hazards of regular smoking in cannabis users was reduced to 1.8 (95% CI = 1.5–2.2) for those meeting this condition.

Mixed Drug Use

Systematic Reviews The committee did not identify a good- or fair-quality systematic review that reported on the association between cannabis use and the rates and use patterns of substances other than cannabis.

Primary Literature To examine trajectories of adolescent cannabis and alcohol users Patton and colleagues (2007) analyzed data from a 10-year cohort study of health in 2,032 adolescents and young adults living in Victoria, Australia. Data were collected in eight waves over the study period from an initial study sample of adolescents who were in mid-secondary school in 1992. About 95 percent of students from the initial study sample participated in Waves 1 through 6, and 75 percent of the students participated in Wave 8. The frequency of cannabis and alcohol

use was categorized in three categories: “any alcohol or cannabis use,” “at least moderate-risk alcohol or cannabis use,” and “high-risk alcohol or cannabis use” (Patton et al., 2007, p. 609). Logistic regression modeling was used to explore the associations between substance use in adolescence (at about 15 years old in Wave 1) and later substance use in early adulthood (at about 25 years old in Wave 8).

After adjusting for a number of social and behavioral factors and persistent substance use measures, the researchers found that adolescents with moderate-risk cannabis use were seven times as likely to develop high-risk cannabis use (odds ratio [OR], 7.4; 95% CI = 3.3–17) and twice as likely to develop high-risk alcohol use in early adulthood (OR, 2.2; 95% CI = 1.1–4.5) compared with students with no hazardous alcohol use or daily cannabis use (Patton et al., 2007).

Among this population, the risk was also elevated for daily cigarette smoking (OR, 3.0; 95% CI = 1.7–5.4), for the use of amphetamines (OR, 6.0; 95% CI = 3.6–10.0), for the use of ecstasy (OR, 7.2; 95% CI = 4.3–12.0), and for the use of cocaine (OR, 4.7; 95% CI = 2.3–9.7) within the past 12 months, as reported in Wave 8 (Patton et al., 2007). The study’s limitations include a 25 percent reduction in the initial sample between Wave 1 and Wave 8 (imputation techniques were used to mitigate potential bias related to students missing waves of the survey), the use of self-reports to determine substance use, and questions about the generalizability of the study to other populations.

The use of cannabis and relapse after discharge from a substance abuse program were the focus of a study conducted by Aharonovich and colleagues (2005). This longitudinal study followed 349 patients who had undergone and successfully completed inpatient treatment for a DSM-IV diagnosis of alcohol, cocaine, or heroin dependence; patients had not experienced mania or non-affective psychosis. Patients were followed up after discharge at months 6, 12, and 18 to update the Psychiatric Research Interview for Substance and Mental Disorders. Responses were analyzed to assess cannabis use and return to substance abuse, sustained remission from substance abuse, and relapse to substance abuse after sustained remission. Of the 349 patients participating in the study, 250 contributed data through at least one follow-up interview; the study results are based on this subset of patients. Of the 250 patients dependent on alcohol, cocaine, or heroin at baseline who did not achieve sustained remission from using these substances, 41.4 percent used cannabis during followup after hospital discharge compared to 15.4 percent of those who had achieved remission (p <0.0001) (Aharonovich et al., 2005). Among the patients dependent on alcohol at baseline who failed to achieve sustained remission, 38.7 percent used cannabis (p <0.004), and among patients dependent on cocaine at baseline who failed to achieve sustained remis-

sion, 52.5 percent used cannabis during follow-up after hospital discharge (p <0.03). Relapse after sustained remission was also seen among patients who used cannabis during follow-up.

A Cox proportional model that adjusted for sociodemographic variables and diagnoses of substance dependence and a number of psychiatric symptoms and disorders was developed to examine the effects of cannabis use on a number of outcomes, including the return to substance use (multiple substance use, alcohol only, cocaine only, and heroin only), sustained remission from substance use, and relapse to substance use. HRs were significant (p <0.0001) for cannabis use and a return to the use of multiple substances, alcohol only, and cocaine only. Cannabis use was associated with a statistically significant reduced hazard of achieving a sustained remission from multiple substance use and cocaine use specifically (p <0.05). In addition, cannabis use was found to increase the hazard of relapse to alcohol use (p <0.05).

Discussion of Findings

The primary literature reviewed present limited evidence that cannabis use affects the rates and patterns of the use of other substances. With regard to alcohol use, cannabis users were found to be at a higher risk for heavy drinking than nonusers. With regard to opioids, cannabis use predicted continued opioid prescriptions 1 year after injury. Finally, cannabis use was associated with reduced odds of achieving abstinence from alcohol, cocaine, or polysubstance use after inpatient hospitalization and treatment for substance use disorders. The limitations of these studies include their lack of generalizability due to their use of restricted study populations, their limited assessment of cannabis use, the lack of dose–response relationships, and the potential for self-report bias.

Is There an Association Between Cannabis Use and the Development of Other Substance Dependence or Other Substance Abuse Disorder?

Alcohol

Systematic Reviews The committee did not identify a good- or fair-quality systematic review that reported on the association between cannabis use and the development of alcohol dependence or alcohol use disorder.

Primary Literature Buu and colleagues (2014) assessed the long-term effects of cannabis use on alcohol problems and alcohol use disorder (AUD) using data from the Michigan Longitudinal Study. The researchers followed a sample of 160 female–male sibling pairs from high-risk families (sample total of 320 individuals) from ages 3–5 to 21–23 years, assessing the participants every 3 years using the Drinking and Drug History Questionnaire, Diagnostic Interview Schedule, Diagnostic Interview Schedule for Children, and the Health and Daily Living Questionnaire. Data were collected on age at first use of alcohol, cannabis, and nicotine as well as the quantity and frequency of use and were analyzed using a linear mixed model. The authors concluded that a higher frequency of cannabis use was related to greater odds of developing drinking problems (β, 0.55; SE = 0.08; p <0.05) and to meeting an AUD diagnosis (β, 0.59; SE = 0.09; p <0.05) (Buu et al., 2014). However, the odds were not as high as those associated with the frequency of alcohol consumption on the odds of developing drinking problems (β, 1.90; SE = 0.10; p <0.05) and the odds of meeting an AUD diagnosis (β, 1.75; SE = 0.31; p <0.05) (Buu et al., 2014). Furthermore, an early onset of cannabis use was not found to contribute to AUD. A major limitation of this study is that the participant population included children who had intact families in early childhood, families that were at high risk for developing AUD, and families of minority race/ ethnicity, thus limiting the generalizability of the study results.

Tobacco/Nicotine

Systematic Reviews The committee did not identify a good- or fair-quality systematic review that reported on the association between cannabis use and the development of tobacco or nicotine dependence or tobacco or nicotine abuse disorder.

Primary Literature Timberlake and colleagues (2007) conducted a study to examine the role of cannabis use in adolescence and the likelihood of developing nicotine dependence and initiating daily tobacco smoking at an earlier age. Survey data were collected from 90,118 students participating in the National Longitudinal Study of Adolescent Health conducted in 132 U.S. schools (public and private) between September 1994 and April 1995. A subsample of participants was followed up at three points with more in-depth surveys, a baseline survey (Wave I), and two subsequent surveys (Wave II, 1 year after the baseline survey, and Wave III, 6 years later). Of these, 5,963 unrelated participants formed the primary sample and included individuals who had not smoked cigarettes by the baseline survey (Wave I) but smoked at least one cigarette by Wave III. Participants ranged in age from 18.3 to 27.7 years. Data on lifetime use of

cannabis and prior-month use at Wave I, age at daily cigarette smoking, and lifetime and current nicotine dependence at Wave III were available for these participants. A smaller sample of 1,447 participants who had tried cannabis by Wave I and for which data on the age of first use was available was used to examine lifetime and current nicotine dependence 6 years later. Cannabis use was classified as no lifetime use, experimental use (1–10 times), and regular use (greater than 10 times). Age at first use was also collected from adolescents who had experimented with cannabis by Wave I of the survey. Nicotine dependence was defined using the Fagerstrom Test for Nicotine Dependence. Demographic risk factor data were also collected. Survey-based logistic regression analysis and censored regression techniques were used to predict outcomes.

Results from this study indicate that regular lifetime users of cannabis at Wave I were 1.89 times more likely to develop lifetime nicotine dependence (t = 2.3 p <0.05, adjusted odds ratio [aOR], 1.89; 95% CI = 1.09–3.30) than nonusers. Past-month users (both experimental and regular) at Wave I were 1.83 times more likely to develop lifetime nicotine dependence (t = 2.3 p <0.05, aOR, 1.83; 95% CI = 1.08–3.11) than nonusers. Furthermore, lifetime users who began using at later ages (23–27) were less likely to develop nicotine dependence at Wave III compared to those who began using at earlier ages (t = −3.3 p <0.01, aOR, 0.82; 95% CI = 0.73−0.93). Limitations associated with this study include self-reported data on substance use, and recall bias.

Agrawal and colleagues (2008), as described in the above section, studied women cannabis users and patterns of smoking and nicotine dependence. Survival analyses were used to examine whether women who smoked cannabis were at an increased risk of moving from regular smoker to nicotine dependent. After adjusting for covariates, the results indicate that women with a prior history of cannabis use were 2.8 times more likely (HR, 2.80; 95% CI = 1.84–4.26) to transition from regular smoking to nicotine dependence. Limitations associated with this study include the lack of generalizability to men, self-reported data on substance use, and recall bias.

Mixed Drug Use

Systematic Reviews The committee did not identify a good- or fair-quality systematic review that reported on the association between cannabis use and the development of substance dependence or substance abuse disorder.

Primary Literature In a longitudinal U.S. study of a nationally representative sample of 34,653 adults 18 years or older, Blanco and colleagues

TABLE 14-1 Cannabis Use in the Past 12 Months and Incident Psychiatric Disorders in Wave 2

NOTE: CI = confidence interval; OR = odds ratio.

SOURCE: Adapted from Blanco et al., 2016.

(2016) examined the association between cannabis use and the risk of developing substance abuse and other mental health disorders. This study investigated the potential association between cannabis use in the past year (Wave 1) with incident substance use disorders, alcohol abuse and dependence, other drug abuse and dependence, and nicotine dependence 3 years later (Wave 2). Both Wave 1 and Wave 2 adjusted for sociodemographic characteristics, a family history of substance use disorder, a disturbed family environment, childhood parental loss, low self-esteem, social deviance, education, recent trauma, past and present psychiatric disorders, and the respondent’s history of divorce. The researchers found that, after adjusting for covariates, cannabis use in the 12 months preceding the interview was associated with an increased risk of developing any substance use disorders, including Cannabis Use Disorder (OR, 6.2; 95% CI = 4.1−9.4) (Blanco et al., 2016). The adjusted ORs for all incident psychiatric disorders in Wave 2 are presented in Table 14-1.

The frequency of cannabis use in Wave 1 was also associated with an incidence of any substance use disorder in Wave 2 (aOR, 1.9; 95% CI = 1.7−2.1), indicating a dose−response association between cannabis use and substance use disorder.² Some of the limitations of this study included the fact that substance use was ascertained by self-report, that there was a possibility of residual confounding, and that the follow-up period was limited to 3 years (Blanco et al., 2016).

Palmer and colleagues (2009) analyzed the substance use experiences

___________________

² Frequency of cannabis use was measured as “no use,” “some use but less than one use per month,” and “one or more uses per month.”

of 1,733 individuals (ages 12–25) who participated in the Colorado Community Twin Study. Data on substance use experimentation and repeated use were collected via self-reported questionnaires and psychiatric interviews in two waves about 5 years apart. Substance abuse and dependence were assessed using the Composite International Diagnostic Interview–Substance Abuse Module (CIDI–SAM) structured interview. With respect to substance use, experimentation was defined as “having used a substance one or more times in a person’s lifetime”; repeated marijuana use was defined as having used cannabis “six or more times in a respondent’s lifetime”; and cannabis abuse and dependence were defined based on the DSM-IV as having compulsive use without generally developing physiological dependence (APA, 1994, p. 216; Palmer et al., 2009, pp. 79–80).

Results show that the risks of alcohol abuse/dependence (OR, 3.44; 95% CI = 1.93–6.12) and tobacco dependence (OR, 4.12; 95% CI = 2.26–7.51) were greater in individuals who used cannabis more than once in their lifetime (without meeting a diagnosis of cannabis substance use disorder) compared to those who did not use cannabis (Palmer et al., 2009). Individuals diagnosed with cannabis use disorder had higher odds of being diagnosed with alcohol abuse/dependence (OR, 8.78; 95% CI = 3.15–24.53) and tobacco dependence (OR, 8.61; 95% CI = 3.15–23.56) than those who did not use cannabis. However, once the logistic regression models were adjusted for the individuals’ involvement with alcohol and tobacco, the ORs no longer reached significance (Palmer et al., 2009).

The researchers found that individuals with cannabis use disorder were not at higher risk for alcohol abuse/dependence (OR, 1.77; 95% CI = 0.54–5.78) or tobacco dependence (OR, 2.61; 95% CI = 0.78–8.72) compared with those who had used cannabis more than once in their lifetime but did not have cannabis use disorder (Palmer et al., 2009). They note that the cannabis and other substance use results indicate “a model of generalized risk since substance use disorders on any substance in young adulthood could be predicted by involvement with any of the three substances in adolescence” (Palmer et al., 2009, p. 78). Study limitations include the difficulty capturing the more severe cases in the cohort, as they are generally not reported; questions about the reliability of self-reporting; of the fact that covariates of substance abuse were not included in the logistic regression models; and the failure of the authors to impose clustering criteria or to distinguish between dependence with or without physiological symptoms (Palmer et al., 2009).

Using data from 1,265 participants of the Christchurch Health and Development longitudinal birth cohort study, Fergusson and colleagues (2008) explored factors associated with illicit drug use, abuse, or dependence among study participants at ages 16 to 25. Cannabis use data were collected for each year and were classified into four levels of frequency:

“did not use cannabis,” “used less than monthly on average (1–11 times),” “used at least monthly on average (12-50 times),” and “used at least weekly (>50 times)” (Fergusson et al., 2008, p. 169). Annual frequency of cannabis use was the strongest predictor of illicit drug use (β, 1.58; SE = 0.06, p <0.0001) and drug abuse or dependence (β, 1.73; SE = 12, p <0.0001) across age groups (Fergusson et al., 2008). The interaction between cannabis use and age was also explored and the association was found to diminish with increasing age. The adjusted odds ratios for the risk of illicit drug use and abuse/dependence for participants who used cannabis at least weekly are presented in Table 14-2 below. Study limitations include questions about the generalizability of the study and the fact that the assessments were based on self-reported data. The confidence intervals for some results are wide.

Discussion of Findings

Most of the studies reviewed indicate an association between cannabis use and use of or dependence on other substances, with some data indicating this effect is more pronounced in younger individuals and is dependent on the dose or frequency of cannabis use. The strengths of some studies cited include the study designs (longitudinal cohort stud-

TABLE 14-2 Adjusted Odds Ratios (and 95% Confidence Intervals) for at Least Weekly Cannabis Use and Risk Factors for Cannabis Use and Illicit Drug Abuse/Dependence, at Ages 16–17, 20–21, and 24–25

NOTE: CI = confidence interval; aOR = adjusted odds ratio.

SOURCE: Adapted from Fergusson et al., 2008.

ies), the existence of large sample sizes, and the fact that adjustments were made for a variety of confounders. The magnitude of the associations appears in the moderate range. The limitations of the studies include, in most cases, the use of self-report for cannabis use, recall bias, and, in some cases, the limited duration of follow-up.

RESEARCH GAPS

To address the research gaps relevant to cannabis use and the abuse of other substances, the committee suggests the following:

• Additional studies are needed to determine whether cannabis use is an independent risk factor for, or causally contributes to, the initiation or use of and dependence on other drugs of abuse later in life.
• In states with legalized recreational cannabis, there need to be longitudinal studies that examine whether the prevalence of use of other drugs parallels the increase in prevalence of cannabis use.

SUMMARY

This chapter summarizes current research evidence on the association between cannabis use and the potential for abusing other substances. Several important research conclusions were reached (see Box 14-1); however, it is important that these conclusions be interpreted within the context of the limitations discussed in the Discussion of Findings sections above.

REFERENCES

Fergusson, D. M., J. M. Boden, and L. J. Horwood. 2008. The developmental antecedents of illicit drug use: Evidence from a 25-year longitudinal study. Drug and Alcohol Dependence 96(1–2):1–2.

IOM (Institute of Medicine). 1999. Marijuana and medicine: Assessing the science base. Washington, DC: National Academy Press.

Kandel, D. 1975. Stages in adolescent involvement in drug use. Science 190:912–914.

Mayet, A., S. Legleye, N. Chau, and B. Falissard. 2011. Transitions between tobacco and cannabis use among adolescents: A multi-state modeling of progression from onset to daily use. Addictive Behaviors 36(11):1101–1105.

Mayet, A., S. Legleye, F. Beck, B. Falissard, and N. Chau. 2016. The gateway hypothesis, common liability to addictions or the route of administration model: A modelling process linking the three theories. European Addiction Research 22(2):107–117.

Novins, D. K., and A. E. Barón. 2004. American Indian substance use: The hazards for substance use initiation and progression for adolescents aged 14 to 20 years. Journal of the American Academy of Child & Adolescent Psychiatry 43(3):316–324.

Palmer, R. H., S. E. Young, C. J. Hopfer, R. P. Corley, M. C. Stallings, T. J. Crowley, and J. K. Hewitt. 2009. Developmental epidemiology of drug use and abuse in adolescence and young adulthood: Evidence of generalized risk. Drug and Alcohol Dependence 102(1–3):1–3.

Panlilio, L. V., Z. Justinova, P. Mascia, M. Pistis, A. Luchicchi, S. Lecca, C. Barnes, G. H. Redhi, J. Adair, S. J. Heishman, S. Yasar, M. Aliczki, J. Haller, and S. R. Goldberg. 2012. Novel use of a lipid-lowering fibrate medication to prevent nicotine reward and relapse: Preclinical findings. Neuropsychopharmacology 37:1838–1847.

Patton, G. C., C. Coffey, M. T. Lynskey, S. Reid, S. Hemphill, J. B. Carlin, and W. Hall. 2007. Trajectories of adolescent alcohol and cannabis use into young adulthood. Addiction 102(4):607–615.

Secades-Villa, R., O. Garcia-Rodriguez, C. J. Jin, S. Wang, and C. Blanco. 2015. Probability and predictors of the cannabis gateway effect: A national study. Journal of Drug Policy 26(2):135–142.

Timberlake, D. S., B. C. Haberstick, C. J. Hopfer, J. Brickerm, J. R. Sakai, J. M. Lessem, and J. K. Hewitt. 2007. Progression from marijuana use to daily smoking and nicotine dependence in a national sample of U.S. adolescents. Drug and Alcohol Dependence 88(2–3):272–281.

Vanyukov, M. M., R. E. Tarter, G. P. Kirillova, L. Kirisci, M. D. Reynolds, M. J. Kreek, K. P. Conway, B. S. Maher, W. G. Iacono, L. Bierut, M. C. Neale, D. B. Clark, and T. A. Ridenour. 2012. Common liability to addiction and “gateway hypothesis”: Theoretical, empirical and evolutionary perspective. Drug and Alcohol Dependence 123(Suppl 1):S3–S17.

Whiteside, L. K., J. Russo, J. Wang, M. L. Ranney, V. Neam, and D. F. Zatzick. 2016. Predictors of sustained prescription opioid use after admission for trauma in adolescents. Journal of Adolescent Health 58(1):92–97.

This page intentionally left blank.

Part IV

Research Barriers and Recommendations

This page intentionally left blank.