Return to Psychopharmacology 3e Student Resources

Psychopharmacology 3e Web Box 9.1 - Of Special Interest: The Gateway Theory of Drug Use

The gateway theory, more properly called the gateway hypothesis, proposes the existence of developmental stages of drug use in adolescents who progress from one substance to another over time. These stages are based on surveys of young drug users asked to report the sequence of drugs they have taken from their first use to the time of the survey. As shown in Figure 1, individuals typically begin with legal drugs, namely cigarettes and alcohol. In some but not all (of course) cases the person progresses to marijuana use. A much smaller subset of individuals then progresses to so-called “hard drugs,” like cocaine or heroin. The figure illustrates that the stages are not invariant, because some individuals try marijuana before tobacco cigarettes or alcohol. Nevertheless, according to the standard gateway hypothesis, cigarettes and alcohol are gateways to marijuana, and marijuana, in turn, is a gateway to other illicit drugs (Kandel and Yamaguchi, 2002).


Figure 1 The standard gateway sequence of drug use initiation. In this diagram, thicker lines represent more likely pathways of progression based on existing statistical findings. Thus, at stage one prior to any drug use, individuals are most likely to start with alcohol, with a smaller proportion beginning with cigarettes instead. Initial use of marijuana is most likely to occur after both alcohol and cigarettes have been used, although in a small percentage of cases initial marijuana use is preceded by only one of these legal substances. Finally, cocaine or heroin use is almost always preceded by the use of marijuana according to the gateway hypothesis. (After Kandel and Yamaguchi, 2002.)

The gateway hypothesis is usually considered to have originated in a seminal 1975 paper by Denise Kandel that reported the findings from two longitudinal surveys of drug use by high school students in New York State. However, the gateway hypothesis was actually preceded by the stepping-stone theory, an idea first proposed in the 1930s that marijuana use inevitably leads to heroin addiction (Kandel, 2002). Although this absurd (to us) proposition was eventually discredited, for a time the stepping-stone theory was sufficiently influential to be the basis of testimony to Congress in the 1950s supporting new legislation to regulate marijuana more strictly (Bonnie and Whitebread II, 1970).

Since Kandel’s first paper on the topic, confusion has existed regarding the exact meaning of the gateway hypothesis. This confusion stems, in part, from conflicting statements in the paper itself. For example, Kandel makes several strong statements like “Drug use starts with legal drugs, which are a necessary stage between nonuse and illegal drug use.” and “Marihuana is a crucial stage prior to the use of other illicit drugs, such as LSD, pills, or heroin.” (Kandel, 1975, p. 913; emphasis added). But later in the paper, she states “However, the data do establish that patterns of use are likely to follow certain paths.” (Kandel, 1975, p. 914; emphasis added). The contrast between the third quoted sentence and the first two raise the issue of whether the observed developmental stages are considered to be obligatory or merely typical. Importantly, several studies conducted either within or outside of the United States found evidence for young people following sequences different than that proposed by the gateway hypothesis (Degenhardt et al., 2010; Tarter et al., 2006; Tullis et al., 2003). These results indicate that the standard gateway stages are not obligatory but simply typical for American adolescents.

In a later summary of the gateway hypothesis, Kandel and Jessor (2002) argue that the hypothesis should be thought of as consisting of three related propositions. The first proposition is the existence of the aforementioned developmental sequence of drug use, in which legal drugs (cigarettes and alcohol) are used first, followed by marijuana and then, in some cases, by “hard drugs” like cocaine and heroin. Although exceptions do exist as noted above, most of the evidence supports this ordering as the usual developmental progression. The second proposition offered by Kandel and Jessor is that use of a drug at one stage in the sequence serves as a risk factor that predicts an increased likelihood of using a drug at the next stage. In epidemiology, a “risk factor” is a variable that is statistically associated with some outcome (e.g., contracting a particular disease). This proposition also appears to be supported by existing data, in that marijuana users are likely to have had prior experience with cigarettes and/or alcohol, and that cocaine or heroin users are likely to have had prior experience with marijuana (Fergusson et al., 2006; Mayet et al., 2012). The third proposition is that use of a drug at one stage in the sequence causes the individual to progress to the next stage. This proposition is, by far, the most contentious aspect of the gateway hypothesis and also the most difficult to prove, given the wide array of causal factors underlying complex human behaviors such as illicit drug use. In fact, Kandel and Jessor concluded that the causal proposition lacked scientific support at the time of their writing.

Although proving the causal proposition of the gateway hypothesis has, thus far, been an intractable problem for human clinical researchers, some animal studies have looked for neurobiological mechanisms that could predispose an organism to progress from one drug to another in the gateway sequence. A study by Levine and coworkers (2011) illustrates one type of approach to this issue. These investigators developed an experimental paradigm in which mice were given nicotine in their drinking water for either 24 hours or 7 days. These mice, along with control animals given water only, were then tested for the ability of cocaine to elicit either locomotor sensitization (increased locomotor activation following repeated cocaine administration) or a conditioned place preference (which tests the rewarding effects of cocaine; see the discussion of drug reward later in this chapter). Figure 2 shows that 7 days of nicotine pre-exposure greatly increased the strength of cocaine-induced locomotor sensitization, whereas 24 h of nicotine had no effect on this measure (data not shown). When the rewarding effects of cocaine were tested using a place-conditioning procedure, 7-day nicotine pre-exposure again exerted a pronounced effect compared with the water-only condition (Figure 3). Additional experiments performed to determine the mechanism underlying these effects showed that nicotine enhanced cocaine induction of the transcription factor FosB in the striatum by means of epigenetic mechanisms (see later section in this chapter for more on the neurobiology of addiction and for a discussion of FosB and epigenetics). The conclusion from these findings is that use of one drug, in this case a substance thought to be a gateway drug, can significantly alter neurobiological and behavioral responsiveness to a drug that follows it in the gateway sequence (for additional discussion, see E. Kandel and D. Kandel, 2014;  D. Kandel and E. Kandel, 2015).


Figure 2 Enhancement of cocaine sensitization by prior nicotine exposure. Mice were given nicotine (50 μl/ml) in their drinking water for 7 consecutive days. For the next 4 days, the mice were injected IP once daily with 20 mg/kg cocaine or saline and their locomotion (distance traveled) in an activity chamber was measured each day. Control mice given no drugs (Water → saline) and mice given nicotine but no cocaine (Nicotine → saline) exhibited little or no change in activity when the 4th test day was compared to the 1st test day (mean values approximating 1 in the graph). Animals given daily cocaine but no nicotine (Water → cocaine) exhibited significant locomotor sensitization, as their activity on day 4 was about 80% higher than their activity on day 1. However, the cocaine-treated animals with nicotine pre-exposure (Nicotine → cocaine) showed much greater sensitization, with a day 4-to-day 1 ratio of over 350%. (After Levine et al., 2011.)


Figure 3 Enhancement of cocaine conditioned place preference by prior nicotine exposure.Mice were given the same drug treatments as in Fig. 2. On day 8, the animals were given free access to both sides of a place conditioning apparatus (see Chapter 4) to determine their initial preference. On days 9-11, the animals were given two sessions in the apparatus, 4 h apart, in which they were first given saline IP and confined for 30 min to the initially preferred side and then later given cocaine (20 mg/kg IP) and confined for another 30 min to the non-preferred side. On day 12, the mice were again given free access to both sides of the apparatus and the amount of time spent on each side was recorded. The graph depicts the change in amount of time spent in the cocaine-paired side on day 12 compared to day 8. The Water → saline and Nicotine → saline groups showed a small, statistically nonsignificant reduction in time spent on the cocaine-paired side. In contrast, the Water → cocaine group showed a clear conditioned place preference as indicated by the large increase in time spent on the cocaine-paired side. Importantly, the Nicotine → cocaine group showed even more robust cocaine place-conditioning than the Water → cocaine group. (After Levine et al., 2011.)

The study just described is by no means the first to demonstrate that exposure to one drug can influence the sensitivity to another drug taken later. Indeed, the phenomena of cross-tolerance and cross-sensitization have been known for many years (see Chapter 1). However, this is one of the first examples of cross-sensitization placed within the context of the gateway hypothesis. Is the hypothesis proven, therefore? Clearly, the answer is no. The study was conducted in mice, not humans, and the researchers did not determine whether nicotine pre-exposure affected the propensity of the animals to initiate cocaine self-administration. (This would have been a better model of increased motivation of human smokers to try cocaine).

Despite the lack of proof of the causal proposition of the gateway hypothesis, the hypothesis has become widely accepted by many researchers, counselors, and drug policy makers. A massive effort has been made to reduce adolescent use of proposed gateway drugs (for example, the DARE [Drug Abuse Resistance Education] Program), in the hope that this will diminish the number of people who later become dependent on substances like cocaine or heroin. However, objections to the gateway hypothesis can be raised on at least two different grounds. First, the studies upon which the hypothesis is based have been carried out mainly using school surveys. The majority of drug users in a typical high school are probably occasional rather than heavy or “hard-core” users, particularly if we focus on illicit drugs other than marijuana. This is not only because hard-core users represent a relatively small percentage of the adolescent population, but also because such individuals often drop out of school.

Second, even if we stick to the adolescent population usually considered in the gateway hypothesis, alternative theories (other than one drug causing use of another) have been offered to explain the gateway sequence (Hall and Lynskey, 2005). For example, Morral and coworkers (2002) found that the relationship between marijuana use and progression to hard drugs could be accounted for by a common factor they called “drug use propensity,” which refers to the tendency to use both marijuana and other illicit drugs. Other researchers have proposed a more general alternative to the gateway hypothesis termed either the “common syndrome theory” or the “problem behavior theory.” According to this idea, some adolescents exhibit a number of problem behaviors, such as delinquency, sexual promiscuity, misconduct, parental defiance, and substance use, all of which “reflect a single, underlying factor” (Donovan and Jessor, 1985). This proposed personality factor would increase an individual’s likelihood of experimenting with more-addictive substances. Further, the observed progression that is usually taken as support for the gateway hypothesis might be due to differences in drug availability (termed “exposure opportunity”; Wagner and Anthony, 2002) and perceived risk. Even though alcohol and cigarettes cannot be purchased legally by minors, they are nevertheless readily available and are often considered harmless by young people. Likewise, marijuana is more easily obtained than drugs like cocaine and heroin, and its use would be considered less risky. Finally, Vanyukov and coworkers have proposed a “common liability to addiction” theory, which argues that some individuals possess a complex personality trait predisposing them to develop a substance use disorder, and that the gateway sequence simply reflects a typical pattern of drug use initiation that emerges from this trait (Vanyukov et al., 2003; 2012). In response to these objections, Kandel and Kandel (2015) have proposed that the gateway hypothesis and the common liability model are complementary rather than competitive. Their idea is that “Common factors explain the use of drugs in general, while specific factors explain why young people use specific drugs and do so in a particular sequence.” (emphasis added; Kandel and Kandel, 2015, p. 131).

The gateway hypothesis has recently been applied to the debate surrounding the risks versus benefits of e-cigarettes (Bell and Keane, 2014). One side of this debate points out that e-cigarette use (i.e., “vaping”) should not be discouraged because it is safer than tobacco smoking due to the lack of cancer-causing tar in e-cigarette vapor (see Chapter 13 for additional discussion). The other side argues against e-cigarettes because in young people, they may serve as a gateway to subsequent use of tobacco cigarettes, marijuana, or hard drugs (Kandel and Kandel, 2015; but see Phillips, 2015, for an opposing view).

In summary, there is a well-described progression of substance use among adolescents, although this progression is neither invariant within the U.S. nor universal across all countries. The gateway hypothesis proposes that initial use of alcohol and/or cigarettes increases an adolescent’s risk of progressing to marijuana and that marijuana use is a risk factor for hard drugs such as cocaine and heroin. Yet it remains to be demonstrated that there is a causal connection involved in the progression of drug use. Moreover, there are alternative theories that may be able to account for a progression of drug use without one substance serving as a “gateway” for another.



Bell, K., and Keane, H. (2014). All gates lead to smoking: The ‘gateway theory’, e-cigarettes and the remaking of nicotine. Soc. Sci. Med., 119, 45–52.

Bonnie, R. J., and Whitebread II, C. H. (1970). The Forbidden Fruit and the Tree of Knowledge: An Inquiry Into the Legal History of American Marijuana Prohibition. Virginia Law Rev., vol. 56, no. 6. Virginia Law Review Association, Charlottesville, VA.

Degenhardt, L., Dierker, L., Chiu, W. T., Medina-Mora, M. E., Neumark, Y., Sampson, N., et al. (2010). Evaluating the drug use “gateway” theory using cross-national data: Consistency and associations of the order of initiation of drug use among participants in the WHO World Mental Health Surveys. Drug Alcohol Depend., 108, 84–97.

Donovan, J. E., and Jessor, R. (1985). Structure of problem behavior in adolescence and young adulthood. J. Consult. Clin. Psychol. 53, 890–904.

Fergusson, D. M., Boden, J. M., and Horwood, L. J. (2006). Cannabis use and other illicit drug use: Testing the cannabis gateway hypothesis. Addiction, 101, 556–569.

Hall, W. D., and Lynskey, M. (2005). Is cannabis a gateway drug? Testing hypotheses about the relationship between cannabis use and the use of other illicit drugs. Drug Alcohol Rev., 24, 39–48.

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

Kandel, D. (2002). Examining the gateway hypothesis. Stages and pathways of drug involvement. In Stages and Pathways of Drug Involvement: Examining the Gateway Hypothesis (D.B. Kandel, Ed.), pp. 3–15. Cambridge University Press, Cambridge, UK.

Kandel, D. B., and Jessor, R. (2002). The Gateway Hypothesis revisited. In Stages and Pathways of Drug Involvement: Examining the Gateway Hypothesis (D.B. Kandel, Ed.), pp. 365–372. Cambridge University Press, Cambridge, UK.

Kandel, D., and Kandel, E. (2015). The Gateway Hypothesis of substance abuse: developmental, biological and societal perspectives. Acta Paediatrica, 104, 130–137.

Kandel, D. B., and Yamaguchi, K. (2002). Stages of drug involvement in the U.S. population. In Stages and Pathways of Drug Involvement: Examining the Gateway Hypothesis (D.B. Kandel, Ed.), pp. 65–89. Cambridge University Press, Cambridge, UK.

Kandel, E. R., and Kandel, D. B. (2014). A molecular basis for nicotine as a gateway drug. N. Engl. J. Med., 371, 932–943.

Levine, A., Huang, Y., Drisaldi, B., Griffin Jr., E. A., Pollak, D. D., Xu, S., et al. (2011). Molecular mechanism for a gateway drug: Epigenetic changes initiated by nicotine prime gene expression by cocaine. Sci. Transl. Med., 3, 107ra109. doi:10.1126/scitranslmed.3003062.

Mayet, A., Legleye, S., Falissard, B., and Chau, N. (2012). Cannabis use stages as predictors of subsequent initiation with other illicit drugs among French adolescents: Use of a multi-state model. Addict. Behav., 37, 160–166.

Morral, A. R., McCaffrey, D. F., and Paddock, S. M. (2002). Reassessing the marijuana gateway effect. Addiction, 97, 1493–1504.

Phillips, C. V. (2015). Gateway effects: Why the cited evidence does not support their existence for low-risk tobacco products (and what evidence would). Int. J. Environ. Res. Public Health, 12, 5439–5464.

Tarter, R. E., Vanyukov, M., Kirisci, L., Reynolds, M., and Clark, D. B. (2006). Predictors of marijuana use in adolescents before and after licit drug use: Examination of the gateway hypothesis. Am. J. Psychiatry, 163, 2134–2140.

Tullis, L. M., DuPont, R., Frost-Pineda, K., and Gold, M. S. (2003). Marijuana and tobacco: A major connection? J. Addict. Dis., 22, 51–62.

Vanyukov, M. M., Tarter, R. E., Kirillova, G. P., Kirisci, L., Reynolds, M. D., Kreek, M. J., et al. (2012). Common liability to addiction and “gateway hypothesis”: Theoretical, empirical and evolutionary perspective. Drug Alcohol Depend., 123S, S3–S17.

Vanyukov, M. M., Tarter, R. E., Kirisci, L., Kirillova, G. P., Maher, B. S., and Clark, D. B. (2003). Liability to substance use disorders: 1. Common mechanisms and manifestations. Neurosci. Biobehav. Rev., 27, 507–515.

Wagner, F. A., and Anthony, J. C. (2002). Into the world of illegal drug use: Exposure opportunity and other mechanisms linking the use of alcohol, tobacco, marijuana, and cocaine. Am. J. Epidemiol., 155, 918–925.