The health and psychological consequences of cannabis use - chapter 7

THIS DOCUMENT HAS BEEN RESCINDED: National Drug Strategy Monograph Series No. 25.

Page last updated: 1994

Please note the information in this publication may no longer be current but is retained on our website for historical or research purposes.

<< Previous Chapter | Contents | Next Chapter>>

7. The psychological effects of chronic cannabis use

A major concern about the psychological consequences of cannabis use has been the possible effects of its chronic use on psychological adjustment in general, and its impact upon motivation and performance in occupational and social roles in particular. There have been two variations on this concern depending upon the age of the cannabis user. Among adults, an "amotivational syndrome" has been described, in which chronic cannabis users become apathetic, socially withdrawn, and perform at a level of everyday functioning well below their capacity prior to their cannabis use. Among adolescents, the concern has been about the effects of heavy cannabis use on motivation to undertake the educational and other psychological tasks that are an essential part of the transition from childhood to adulthood. The evidence for each of these adverse outcomes of heavy cannabis use will be considered separately, beginning with the effects on adolescent development, which have understandably provoked the greatest concern, and prompted the most research.

7.1 Effects on adolescent development

The effects of heavy cannabis use on adolescent development are of special concern for a number of reasons. First, adolescents are minors whose decisions about whether or not to use drugs are not conventionally regarded as free and informed in the way that adult choices are (Kleiman, 1989). Second, adolescence is an important period of transition from childhood to adulthood, in which regular cannabis intoxication may be expected to interfere with educational achievement, the process of disengagement from dependence upon parents, the development of relationships with peers, and making important life choices, such as whether, whom and when to marry, and what occupation to pursue (Baumrind and Moselle, 1985; Polich, Ellickson, Reuter and Kahan, 1984). Third, the age at which drug use begins has implications for subsequent drug use and health and well-being. Early initiation of cannabis use predicts an increased risk of escalation to heavier cannabis use, and to the use of other illicit drugs. It also means a longer period of heavy use, and hence, an increased risk of experiencing any adverse health effects that chronic cannabis use may have in later adult life (Kleiman, 1989; Polich, Ellickson, Reuter and Kahan, 1984). Fourth, since adolescence is a time of risk-taking, the use of any intoxicant, whether alcohol or cannabis while driving a car, increases the risks of accidental injury, and hence of premature death (Kleiman, 1989; Polich, Ellickson, Reuter and Kahan, 1984).

The type of evidence that initially excited concern about the effects of chronic cannabis use on adolescents came from clinical case studies in which bright adolescents' use of cannabis escalated to daily cannabis use, and the use of other illicit drugs, leading to declining social and educational performance, as evidenced by high school drop-out, and immersion in the illicit drug subculture (e.g. Kolansky and Moore, 1971; Lantner, 1982; Milman, 1982; Smith and Seymour, 1982). In some of these cases, the syndrome remitted after the adolescent had been abstinent from cannabis for some months (Meeks, 1982; Smith and Seymour, 1982). Nonetheless, the evidence was largely anecdotal and so of limited value in making causal inferences about the contribution that cannabis made to the development of these outcomes. It did not, that is, permit a decision to be made as to what extent cannabis use was a symptom rather than a cause of personality, or other psychiatric disorders, or a form of adolescent rebellion against parental values.

The concern about the adverse effects of cannabis use on adolescent development in the late 1970s prompted a number of large-scale prospective epidemiological studies of the antecedents, and to a limited degree, the consequences of adolescent drug use (e.g. Kandel, 1988; Kaplan, Martin and Robbins, 1982; Newcombe and Bentler, 1988). These studies have attempted to tease out the contributions of the users' pre-existing personal and social characteristics from the specific effects of drug use. Some of these studies have also attempted to examine the impact of illicit drug use in adolescence upon a number of social and personal outcomes in early adult life (e.g. Kandel, 1988; Newcombe and Bentler, 1988). The most important of these studies are reviewed below.

Top of page

7.1.1 Is cannabis a gateway drug?

A major concern about cannabis has been that its use in adolescence may lead to, or increase the risk of using other more dangerous illicit drugs, such as cocaine and heroin (DuPont, 1984; Goode, 1974; Kleiman, 1992). The most popular evidence for this hypothesis is the fact that the majority of heroin and cocaine users used cannabis before heroin and cocaine. Such evidence is weak. In the absence of comparative data on the prevalence of cannabis use by non-heroin addicts we are unable to decide if there is an association between cannabis and heroin use. Even if there is an association, alternative explanations of its possible causal significance have to be evaluated and excluded (Goode, 1974).

There is now abundant evidence of an association between cannabis and heroin use from a series of cross-sectional studies of adolescent drug use in the United States and elsewhere, including Australia. In the late 1970s and into the 1990s in the United States there was a strong relationship between degree of current involvement with cannabis and the use of other illicit drugs such as heroin and cocaine users. Kandel (1984), for example, found that the prevalence of other illicit drug use increased with current degree of marijuana involvement: 7 per cent of those who had never used marijuana, 33 per cent of those who had used in the past, and 84 per cent of those who were currently daily cannabis users, had used other illicit drugs. Current cannabis users were also likely to have used a larger number of different types of illicit drugs.

Cross-sectional data on drug use among Australian adults in 1993 have also shown that those who have tried cannabis are more likely to have used heroin, and the greater the frequency of cannabis use, the higher the probability of their having tried heroin (see Donnelly and Hall, 1994). In the 1993 NCADA survey of drug use in Australia, for example, the crude risk of using heroin was approximately 30 times higher among those who have used cannabis than those who have not (even though 96 per cent of cannabis users had not used heroin) (see Donnelly and Hall, 1994).

The relationships between cannabis and heroin use observed in the cross-sectional studies have also been observed in the small number of longitudinal studies of drug use. In one of the first such studies Robins, Darvish and Murphy (1970) followed up a cohort of 222 African-American adolescents identified from school records at age 33, and interviewed them retrospectively about their drug use in adolescence and young adulthood, and their adult adjustment. They found a higher rate of progression to heroin use among the young men who had used cannabis before age 20.

These early results have been confirmed and elaborated upon in the extensive research on adolescent drug use by Kandel and her colleagues (e.g. Kandel et al, 1986). These investigators have identified a predictable sequence of involvement with licit and illicit drugs among American adolescents, in which progressively fewer adolescents tried each drug class, but in which almost all of those who tried drug types later in the sequence had used all drugs earlier in the sequence (Kandel and Faust, 1975). Typically, psychoactive drug use began with the use of the legal drugs alcohol and tobacco, which were almost universally used. A smaller group of the alcohol and tobacco users (although often the majority of adolescents) initiated cannabis use, and those whose progressed to regular cannabis use were more likely to use the hallucinogens and "pills" (amphetamines and tranquillisers). The heaviest users of "pills", in turn, were more likely to use cocaine and heroin. Generally, the earlier the initiation of any drug use, and the heavier the use of any particular drug in the sequence, the more likely the user was to use the next drug type in the sequence (Kandel, 1978; Kandel et al, 1984; Kandel, 1988).

This sequence of drug involvement has largely been confirmed by other researchers. Donovan and Jessor (1983), for example, found much the same sequence of initiation, with the variation that when problematic alcohol use was distinguished from non-problem alcohol use, then marijuana use preceded problem drinking in the sequence of progression. These sequences have also been observed in the small number of prospective studies which have followed a cohort of adolescents into early adulthood and examined the patterns of progression in drug use (e.g. Kaplan et al, 1982; Yamaguchi and Kandel, 1984a, b). For the majority (87 per cent) of men "the pattern of progression is one in which the use of alcohol precedes marijuana; alcohol and marijuana precede other illicit drugs; and alcohol, cigarettes and marijuana precede the use of prescribed psychoactive drugs" (Yamaguchi and Kandel, 1984a, p671). Among the majority of women (86 per cent) the sequence was such that "either alcohol or cigarettes precedes marijuana; alcohol, cigarettes and marijuana precede other illicit drugs; alcohol and either cigarettes or marijuana precede prescribed psychoactive drugs" (Yamaguchi and Kandel, 1984a, p671).

Top of page

Yamaguchi and Kandel (1984b) also examined variables which predicted progression to illicit drug use beyond cannabis use. They were specifically interested in "whether the use of certain drugs lower in the sequence influences the initiation of higher drugs" (p673) and used sophisticated statistical methods to discover if the statistical relationship between cannabis use and subsequent illicit drug use persisted after controlling for temporally prior variables, such as pre-existing adolescent behaviours and attitudes, interpersonal factors, and age of initiation into drug use. If the relationship persisted after controlling for these variables, confidence was increased that the relationship was a causal one.

Yamaguchi and Kandel found that the relationship between marijuana use and progression to the use of other illicit drugs was not only explained by friends' marijuana use (which also predicted progression). Among men, the age of initiation of marijuana was an important modifier of this relationship: men who initiated marijuana use under the age of 16, were "even more likely to initiate other illicit drugs than is expected from the longer period of risk resulting from an early age of onset" (p677). Most importantly, "persons who have not used marijuana have very small probabilities of initiating other drugs, ranging from 0.01 to 0.03 (men) or 0.02 (women)" indicating that in their cohort, "marijuana appears to be a necessary condition for the initiation of other illicit drugs" (p677).

The work of Kandel and her colleagues and that of other researchers (e.g. O'Donnell and Clayton, 1982) has been interpreted by some as confirming the "gateway drug" hypothesis or "the stepping stone theory of drug use" (DuPont, 1984). Although it is not always clear what is being claimed by proponents of this hypothesis, it does not imply that a high proportion of those who experiment with marijuana will go on to use heroin. Indeed, the overwhelming majority of cannabis users do not use harder drugs like heroin. Kandel has explicitly disavowed this interpretation of her work:

The notion of stages in drug behavior does not imply that these stages are either obligatory or universal ... the model is not meant to be a variant of the controversial `stepping-stone' theory of drug addiction in which use of marijuana was assumed inexorably to lead to the use of other illicit 'hard' drugs, especially heroin (Kandel, 1988, pp58-61).

The view that cannabis use generally leads to the use of other illicit drugs is contradicted by the evidence from the studies of Kandel and her colleagues. Cannabis use is largely a behaviour of late adolescence and early adulthood. Kandel's research has shown that it has been initiated by the age of 19 in 90 per cent of those who ever used cannabis, and initiation is rare after 20 years. The frequency of its use peaks in the early 20s, when 50 per cent of males and 33 per cent of females reported using, and rapidly declines by age 23, with "the assumption of the roles of adulthood .. getting married, entering the labor force, or becoming a parent .. that may be incompatible with involvement in illicit drugs and deviant lifestyles" (Kandel and Logan, 1984, p665). Hence, although those who use cannabis are more likely to use other illicit drugs than those who do not, it is more usual for cannabis use to decline in early adult life, with only a minority continuing to use regularly, or progressing to the use of more dangerous illicit drugs. Even in the case of the minority (about one in four) who progress to daily cannabis use, the majority cease their use by the mid to late 20s (Kandel and Davies, 1992).

A better supported hypothesis is that cannabis use, especially heavy cannabis use, greatly increases the chances of progressing to the use of other illicit drugs. But even this type of relationship does not necessarily mean that cannabis use "causes" heroin use. As Kandel (1988) has stressed, the existence of sequential stages of progression does not "necessarily imply causal linkages among different drugs". The sequences "could simply reflect the association of each class of drugs with different ages of initiation or [with pre-existing] individual attributes, rather than the specific effects of the use of one class of drug on the use of another" (Kandel, 1988, p61).

Top of page

A plausible alternative hypothesis is that of selective recruitment. That is, there is a selective recruitment to cannabis use of deviant and nonconformist persons with a predilection for the use of intoxicating substances. On this hypothesis, the sequence in which drugs are typically used reflects their differential availability and societal disapproval (e.g. Donovan and Jessor, 1983). Further, the sequence of initiation into drug use is held to be a consequence of the availability of different drugs at different ages, with the use of the least available, and most strongly socially disapproved "hard" drugs being last. This hypothesis exculpates cannabis use as a cause of progression to other illicit drug use, since cannabis use and other illicit drug use are the common consequences of adolescent deviance and nonconformity (Kaplan et al, 1982; Newcombe and Bentler, 1988).

The selective recruitment hypothesis has received support from a number of studies. There are substantial correlations between various forms of nonconforming adolescent behaviour, such as, high school drop-out, early premarital sexual experience and pregnancy, delinquency, and alcohol and illicit drug use (Jessor and Jessor, 1977; Osgood et al, 1988). All such behaviours are correlated with nonconformist and rebellious attitudes and anti-social conduct in childhood (Shedler and Block, 1990) and early adolescence (Jessor and Jessor, 1977; Newcombe and Bentler, 1988). Recent research indicates that those who are most likely to use other illicit drugs, namely, those who become regular cannabis users (Kandel and Davies, 1992), are more likely to have a history of anti-social behaviour (Brook et al, 1992; McGee and Feehan, 1993), nonconformity and alienation (Brook et al, 1992; Jessor and Jessor, 1978; Shedler and Block, 1990), perform more poorly at school (Bailey et al, 1992; Hawkins et al, 1992; Kandel and Davies, 1992), and use drugs to deal with personal distress and negative affect (Kaplan and Johnson, 1992; Shedler and Block, 1990). In general, the more of these risk factors that adolescents have, the more likely they are to progress to more intensive involvement with cannabis, and hence, to use other illicit drugs (Brook et al, 1992; Newcombe, 1992; Scheier and Newcombe, 1991).

One way of testing the selective recruitment hypothesis is to discover whether cannabis use continues to predict progression to "harder" illicit drugs after statistically controlling for pre-existing differences in personality and other characteristics (e.g. deviance) between cannabis users and non-users. In several such studies (e.g. Kandel et al, 1986; O'Donnell and Clayton, 1982; Robins et al, 1970) the relationship between cannabis and heroin use has been reduced when pre-existing differences have been controlled for, but in all cases the relationship has persisted, albeit in attenuated form. O'Donnell and Clayton (1982) have interpreted this type of finding as strong evidence in favour of a causal connection between cannabis and heroin use.

The credibility of such an argument for a causal interpretation of the relationship between cannabis and heroin use depends upon whether the most important prior characteristics have been adequately measured and statistically controlled for in these studies. It would be difficult to argue that this has been the case. Kandel et al (1986), for example, were unable to measure the users' attitudes and family characteristics at the time of drug initiation, or differential drug availability, either or both of which "may account for the observed relationships between the early and late stage drugs" (p679). In both the studies of O'Donnell and Clayton (1982) and Robins et al (1970) the measures of deviance "prior" to drug use were assessed retrospectively with unknown validity. Baumrind (1983) has contested the ability of these studies to exclude the alternative hypothesis that personality differences which preceded cannabis use were the causes of the progression to heroin use. She has argued that "it is safer in the absence of evidence of external validity" of these measures to assume that the relationship between marijuana use and heroin use is spurious.

Even if we assume for the purpose of argument that the association between cannabis and heroin use is not wholly explained by pre-existing differences in deviant behaviour between cannabis users and non-users, it remains to be explained how cannabis use "causes" heroin use. It may seem superficially plausible to suggest that there is something about the pharmacological effect of cannabis which predisposes heavy users to progress to the use of other intoxicants, but there is no obvious pharmacological mechanism for such progression. Is it the development of tolerance to the positive effects of cannabis, or to some form of experiential satiation with its effects? Does the euphoria of cannabis awaken appetite for intoxication by other drugs? These possibilities are difficult to test.

Any pharmacological explanation in which more potent illicit drugs serve as "substitutes" for less potent drugs like alcohol and cannabis has to contend with a number of facts. As already indicated, there are relatively low rates of progression from cannabis use to the sustained use of other illicit drugs; experimentation and abandonment is more the norm. Even those heavy cannabis users who use other illicit drugs continue to use cannabis as well as the new illicit drugs. As Donovan and Jessor (1983) have noted: "...`harder' drugs do not serve as substitutes for `softer' drugs. Rather, a deepening of regular substance use appears to go along with a widening of experience in the drug domain" (p548-549).

Top of page

There is also good reason for believing that the pattern of progression observed among American adolescents in the 1970s was conditioned by historical differences in drug availability (Kandel, 1978). Historical evidence from among earlier cohorts of heroin users indicated that prior involvement with cannabis was confined to those geographic areas of the US in which it was readily available (Goode, 1974). Research on African-American adolescents also showed a variation in the sequence of drug use, with the use of more readily available cocaine and heroin preceding the use of the less readily available hallucinogens and "pills" (Kandel, 1978). Most dramatically, American soldiers in Vietnam were more likely to use heroin than alcohol because heroin was cheaper and more freely available than alcohol to most American troops who were younger than the minimum drinking age of 21 (Robins, 1993).

The historical and geographical variations in sequencing of illicit drug use suggest a sociological explanation of both the sequencing of illicit drug use and the higher rates of progression to heroin use among heavy cannabis users. One of the most popular sociological hypotheses is that cannabis use increases the chance of using other illicit drugs by increasing contact with other drug users as part of a drug using subculture. On this hypothesis, heavy cannabis use leads to greater involvement in a drug using subculture which, in turn, exposes cannabis users to the example of peers who have used other illicit drugs. Such exposure also increases opportunities to use other illicit drugs because of their increased availability within their social circle, and places the individual in a social context in which illicit drug use is encouraged and approved (e.g. Goode, 1974).

Although plausible, there is surprisingly little direct evidence on the drug subculture hypothesis. Goode (1974) presented data from the late 1960s indicating that the number of friends who used heroin was a stronger predictor of heroin use than was frequency of cannabis use, arguing that the "correlation between frequency of use and the use of dangerous drugs ... [is] the result of interaction and involvement with others who use" (p332). These observations have been supported by Kandel's (1984) finding that the strongest predictor of continued cannabis use in early adulthood was the number of friends who were cannabis users.

The hypotheses of selective recruitment and socialisation in a drug-using subculture are not mutually exclusive; both processes could independently contribute to the relationship between regular marijuana use and progression to heroin use (Goode, 1974). As already noted, the selective recruitment hypothesis is supported by the consistent finding of pre-existing differences between those who use marijuana and those who do not, which are most marked in those whose continued use of cannabis predicts their use of other illicit drugs. Once initiated into cannabis use, heavy users become further distinguished from non-users and those who have discontinued their use by the intensity of their social relations and activities which involve the use of marijuana, such as mixing with other drug users, and buying and selling illicit drugs. The illegality of these activities confers on the use, possession and sale of cannabis a socialising and subcultural influence not possessed by the possession and use of the legal drugs (Goode, 1974).

On the available evidence, the case for a pharmacological explanation of the role of cannabis use in progression to other illicit drug use is weak. A sociological explanation is more plausible than a pharmacological one. The predictive value of cannabis use is more likely to reflect a combination of: the selective recruitment to heavy cannabis use of persons with combination of pre-existing personality and attitudinal traits that predispose to the use of other intoxicants; and the effects of socialisation into an illicit drug subculture in which there is an increased availability of, and encouragement to use, other illicit drugs.

Top of page

7.1.2 Educational performance

A major concern about the effects of adolescent cannabis use has been the possibility that its use impairs educational performance, and increases the chances of students discontinuing their education. Such a possibility is plausible: heavy cannabis use in the high school years would impair memory and attention, thereby interfering with learning in and out of the classroom (Baumrind and Moselle, 1985). If use became chronic, persistently impaired learning would produce poorer performance in high school and later in college, and increase the chance of a student dropping out of school. If the adolescent's school performance was marginal to begin with, as research reviewed above suggests it is more likely to be among marijuana users, then regular use could increase the pre-existing risk of high school failure. Because of the importance of high school education to occupational choice, this potential effect of adolescent cannabis use could have consequences which ramified throughout the affected individual's adult life.

Such a possibility has been supported by cross-sectional studies (e.g. Kandel, 1984; Robins et al, 1970). These and other studies (see Hawkins et al, 1992) have found a positive relationship between degree of involvement with cannabis as an adult and the risk of dropping out of high school. Studies of relationships between performance in college and marijuana smoking have produced more equivocal results (see below), usually failing to find consistent evidence that the performance of cannabis users was more impaired than would be predicted by their performance prior to cannabis use. These studies have been criticised (Baumrind and Moselle, 1985; Cohen, 1982), however. Baumrind and Moselle have argued that grade point average is an insensitive measure of adverse educational effects among bright high school and college students, while Cohen has argued that students whose learning has been most adversely affected by their chronic heavy cannabis use would not be found in college samples (Cohen, 1982).

Longitudinal studies of the effect of cannabis use on educational achievement have produced mixed support for the hypothesis (e.g. Kandel et al, 1986; Newcombe and Bentler, 1988). Kandel et al (1986), for example, analysed the follow-up data from the cohort on which their earlier cross-sectional finding of a relationship between cannabis use and high school drop-out had been reported. They reported a negative relationship between marijuana use in adolescence and years of education completed in early adulthood but this relationship disappeared once account was taken of the fact that those who used cannabis in adolescence had much lower educational aspirations than those who did not.

Newcombe and Bentler (1988) used a different approach to analysis in their study of the effects of adolescent drug use on educational pursuits in early adulthood. They used a composite measure of degree of drug involvement, which measured frequency of use of alcohol, cannabis and "hard drugs", and a measure of social conformity in adolescence as a control variable in the analyses, which examined the relationships between adolescent drug use and educational pursuits in early adulthood. They found negative correlations between adolescent drug use and high school completion, but after controlling for the higher nonconformity and lower academic potential among adolescent drug users, there was only a modest negative relationship between drug use and college involvement. The only specific effect of any particular type of drug use, over and above their measure of drug use involvement, was a negative relationship between hard drug use in adolescence and high school completion.

On the whole then, the available evidence from the longitudinal studies suggests that there may be a modest statistical relationship between cannabis and other illicit drug use in adolescence and poor educational performance. The apparently strong relationship between cannabis use and high school drop-out observed in cross-sectional studies exaggerates the adverse impact of cannabis use on school performance because adolescents who perform less well at school, and have lower academic aspirations, are more likely to use cannabis. But even if the relationship is statistically small, it may be substantively important, especially among those whose educational performance was marginal to begin with, because of the adverse effects that educational underachievement has on subsequent life choices, such as occupation, and the opportunities that they provide or exclude.

Top of page

7.1.3 Occupational performance

Among those young adult cannabis users who enter the work-force, the continued use of cannabis and other illicit drugs in young adulthood might impair job performance for the same reasons that it has been suspected of impairing school performance, namely, that chronic intoxication impairs work performance. There is some suggestive support for this expectation, in that cannabis users report higher rates of unemployment than non-users (e.g. Kandel, 1984; Robins et al, 1970), but this comparison is likely to be confounded by the different educational qualifications of the two groups. Longitudinal studies have suggested that there is a relationship between adolescent marijuana use and job instability among young adults which is not explained by differences in education and other characteristics which precede cannabis use (e.g. Kandel et al, 1986). Newcombe and Bentler (1988) provided a more extensive analysis of the effects of adolescent drug use on occupational performance in young adulthood. They examined the relationships between adolescent drug use and income, job instability, job satisfaction, and resort to public assistance in young adulthood, while controlling for differences between users and non-users in social conformity, academic potential and income in adolescence. Their findings supported those of Kandel and colleagues in that adolescent drug users had a larger number of changes of job than non-drug users. Newcombe and Bentler conjectured that this reflects either impaired work performance, or a failure of illicit drug users to develop responsible employment behaviours such as conscientiousness, thoroughness, and reliability.

Top of page

7.1.4 Interpersonal relationships

There are developmental and empirical reasons for suspecting that cannabis use may adversely affect interpersonal relationships. The developmental reason is that heavy adolescent drug use may produce a developmental lag, entrenching adolescent styles of thinking and coping which would impair the ability to form adult interpersonal relationships (Baumrind and Moselle, 1985). The empirical reason is the strong positive correlation between drug use, precocious sexual activity, and early marriage, which in turn predicts a high rate of relationship failure (Newcombe and Bentler, 1988).

Cross-sectional studies of drug use in young adults have indicated that a high degree of involvement with marijuana predicts a reduced probability of marriage, an increased rate of cohabiting, an increased risk of divorce or terminated de facto relationships, and a higher rate of unplanned parenthood and pregnancy termination (Kandel, 1984; Robins et al, 1970). Kandel (1984) also found that heavy cannabis users were more likely to have a social network in which friends and the spouse or partner were also cannabis users (Kandel, 1984). These findings have been largely confirmed in analyses of the longitudinal data from this cohort of young adults (Kandel et al, 1986).

Newcombe and Bentler (1988) found similar relationships between drug use and early marriage in their analysis of the cross-sectional data from their cohort of young adults in Los Angeles. Drug use in adolescence predicted an increased rate of early family formation in late adolescence and of divorce in early adulthood, which they interpreted as evidence that: "early drug involvement leads to early marriage and having children which then results in divorce" (p97). Newcombe and Bentler argued that this finding provided evidence for their theory of "precocious development", according to which drug use accelerates development and "... drug users tend to bypass or circumvent the typical maturational sequence of school, work and marriage and become engaged in adult roles of jobs and family prematurely without the necessary growth and development to enhance success with these roles ... [developing] a pseudomaturity that ill prepares them for the real difficulties of adult life" (pp35-36).

Less attention has been paid to the possibility that cannabis use has adverse effects on the development of social relationships outside marriage. Newcombe and Bentler (1988) have reported one of the few such studies. They investigated the relationship between adolescent drug use and degree of social support and the experience of loneliness reported in young adulthood. Cross-sectional analyses of data on drug use and degree of social support in adolescence showed that drug users reported having less social support than non-users (Newcombe and Bentler, 1988). But the effects of adolescent drug use on social support and loneliness in young adulthood were minor. Alcohol use in adolescence was associated with decreased loneliness in adulthood, while only hard drug use in adolescence was associated with decreased social support and increased loneliness in early adulthood.

Top of page

7.1.5 Mental health

The impact of adolescent cannabis and other drug use on general health in early adult life has not been investigated, in large part because it will be difficult to detect any adverse effects of adolescent drug use on adult health in the longitudinal studies that have been conducted. In such cohorts, heavy cannabis use - the riskiest pattern of use from the perspective of health effects - has generally been observed to occur at low rates. In any case, young adulthood is too soon to expect any adverse health effects to be evident, because of the relatively short period of use by young adults.

For good reasons, the effects of cannabis use on mental health have been the health outcomes most studied. Cannabis is a psychoactive drug which effects the users' mood and feeling, so chronic heavy use could possibly adversely affect mental health, especially among those whose adjustment prior to their cannabis use was poor and who use cannabis to modulate and control their negative mood states and emotions. The relationships between cannabis use and the risks of developing dependence upon cannabis or major mental illnesses such as schizophrenia, are reviewed below (see pp110-122 and pp173-178 respectively). In this section attention is confined to non-psychotic symptoms of depression and distress.

A number of studies have suggested an association between cannabis use and poor mental health. Kandel's (1984) cross-sectional study found an inverse association between the intensity of marijuana involvement and degree of satisfaction with life, and a positive association between marijuana involvement and a greater likelihood of having consulted a mental health professional, and having been hospitalised for a psychiatric disorder (Kandel, 1984). Longitudinal analyses of this same cohort, however, found only weak associations between adolescent drug use and these adult outcomes; the strongest relationship between adolescent drug use and mental health, was a positive relationship between cigarette smoking in adolescence and increased symptoms of depression in adulthood (Kandel et al, 1986).

The cross sectional adult data in Newcombe and Bentler's (1988) study showed strong relationships between adolescent drug use and emotional distress, psychoticism and lack of a purpose in life. Emotional distress in adolescence predicted emotional distress in young adulthood, but there were no relationships between adolescent drug use and the experience of emotional distress, depression and lack of a sense of purpose in life in young adulthood. There were a number of small but substantively significant effects of adolescent drug use on mental health in young adulthood. Adolescent drug use predicted psychotic symptoms in young adulthood, and hard drug use in adolescence predicted increased suicidal ideation in young adulthood, after controlling for general drug use and earlier emotional distress. Newcombe and Bentler interpreted these findings as evidence that adolescent drug use "interferes with organised cognitive functioning and increases thought disorganisation into young adulthood" (p180).

Top of page

7.1.6 Delinquency and crime

Since initiation into illicit drug use and the maintenance of regular illicit drug use are both strongly related to degree of social nonconformity or deviance (e.g. Donovan and Jessor, 1980; Newcombe and Bentler, 1988; Polich et al, 1984) it is reasonable to expect adolescent illicit drug use to predict social nonconformity and various forms of delinquency and crime in young adulthood. Cross-sectional studies of adult drug users seem to support this hypothesis: they indicate that there is a relationship between the extent of marijuana use as an adult and a history of lifetime delinquency (e.g. Kandel, 1984; Robins et al, 1970), having been convicted of an offence, and having had a motor vehicle accident while intoxicated (Kandel, 1984).

Johnston et al (1978) reported a detailed analysis of the relationship between intensity of drug use and delinquency across two waves of interviews of adolescent males undertaken as part of the "Youth in Transition" study. They found in their cross-sectional data that there was a strong relationship between involvement in delinquency and degree of involvement with illicit drugs, that is, self-reported rates of delinquent activity increased steadily with increasing degree of drug involvement. However, a series of analyses looking at changes in drug use and crime over time indicated that the groups defined on intensity of drug involvement differed strongly in their rate of delinquent acts before their drug use. Moreover, the onset of illicit drug use (including cannabis) had little effect on delinquent acts, except perhaps among those who used heroin, among whom there was a suggestion that the rates of delinquency increased. Finally, rates of delinquent acts declined over time in all drug use groups and at about the same rate. The findings were interpreted as delivering "a substantial, if not mortal, blow" to the hypothesis that "drug use somehow causes other kinds of delinquency" (p156).

Newcombe and Bentler (1988) reported a somewhat more complicated although no less plausible picture in their longitudinal study. They reported a positive relationship between drug use and criminal involvement in adolescence, but found more mixed results in the relationship between adolescent drug use and criminal activity in young adulthood. Adolescent drug use predicted drug crime involvement in young adulthood; but after controlling for other variables, it was negatively correlated with violent crime, and general criminal activities in young adulthood. Newcombe and Bentler argued that these negative correlations indicated that the correlation between different forms of delinquency in adolescence decreases with age, as criminal activities become differentiated into drug-related and non-drug-related offences. Hard drug use in adolescence also had a specific effect on young adult crime over and above that of drug use in general: it predicted an increased rate of criminal assaults in young adulthood.

Top of page

7.1.7 Conclusions

There are a number of clear outcomes of research on adolescent cannabis and other illicit drug use. First, there is strong continuity of development from adolescence into early adult life in which many of the indicators of adverse development which have been attributed to cannabis use precede its first use (Kandel, 1978). These include minor delinquency, poor educational performance, nonconformity, and poor adjustment. Second, there was a predictable sequence of initiation into the use of illicit drugs among American adolescents in the 1970s in which the use of licit drugs preceded experimentation with cannabis, which preceded the use of hallucinogens and "pills", which in turn preceded the use of heroin and cocaine. Generally, the earlier the age of initiation into drug use, and the greater the involvement with any drug in the sequence, the greater the likelihood of progression to the next drug in sequence.

The causal significance of these findings, and especially the role of cannabis in the sequence of illicit drug use, remains controversial. The hypothesis that the sequence of use represents a direct pharmacological effect of cannabis use upon the use of later drugs in the sequence is the least compelling. A more plausible and better supported explanation is that it reflects a combination of the selective recruitment into cannabis use of nonconforming and deviant adolescents who have a propensity to use illicit drugs, and the socialisation of cannabis users within an illicit drug using subculture which increases the exposure, opportunity, and encouragement to use other illicit drugs.

There has been some support for the hypothesis that heavy adolescent use of cannabis impairs educational performance. Cannabis use appears to increase the risk of failing to complete a high school education, and of job instability in young adulthood. The apparent strength of these relationships in cross-sectional studies has been exaggerated because those who are most likely to use cannabis have lower pre-existing academic aspirations and high school performance than those who do not. Even though more modest than has sometimes been supposed, the apparently adverse effects of cannabis and other drug use upon educational performance may cascade throughout young adult life, affecting choice of occupation, level of income, choice of mate, and quality of life of the user and his or her children.

There is weaker but suggestive evidence that heavy cannabis use has adverse effects upon family formation, mental health, and involvement in drug-related (but not other types of) crime. In the case of each of these outcomes, the apparently strong associations revealed in cross-sectional data are much more modest in longitudinal studies after statistically controlling for associations between cannabis use and other variables which predict these adverse outcomes.

On balance, there are sufficient indications that cannabis use in adolescence adversely affects adolescent development to conclude that it is a socially desirable goal to discourage adolescent cannabis use, and especially regular cannabis use.