Evaluating the cognitive effects of long-term cannabis use is difficult because of many confounding variables. The judgment as to whether a given effect is a result of cannabis use is complicated by various sources of bias that can occur in naturalistic studies. Experimental data are lacking because one cannot ethically recruit a group of individuals to consume large amounts of cannabis for 20 years, while having a comparison group abstain. The only available data come from naturalistic studies of individuals using cannabis in uncontrolled settings, which are beset by all the usual methodological problems that afflict naturalistic and retrospective studies.
First is the issue of selection bias. Cannabis users who respond to an advertisement for a study may not be representative of the source population from which they are drawn. For example, cannabis users who come into a study may be particularly worried about their cognitive functioning and want to be evaluated, or may have psychiatric symptoms and want to be evaluated by a mental health professional, or may be self-selected in some other way. So for a number of reasons, one might predict that selection factors might bias the results in favor of finding a deficit in cannabis users and rejecting the null hypothesis (type I error).
Second is the issue of information or recall bias. Cannabis users may suspect that they would be rejected from the study, and not get the associated compensation, if they reported significant use of other drugs or psychiatric disorders. This would result in a denial or minimizing of their use of other potentially toxic drugs or the presence of psychiatric disorders, which, in turn, would result in a bias away from the null hypothesis (type I error).
Third, and most serious, is the issue of unmeasured confounding variables, which impacts all naturalistic studies. Even if we could do a perfect study in which we completely eliminated selection bias and completely eliminated information bias, if we found that cannabis users exhibited cognitive deficits, how would we know that these deficits were attributable to cannabis use and not to some other confounding factor such as impaired cognitive functioning prior to cannabis use?
There have been many studies measuring the cognitive effects of long-term cannabis use. This literature lacks consensus, and a 2003 meta-analysis of existing studies that met appropriate methodological criteria failed to find significant cognitive deficits in long-term users (17
The results from a study by Pope et al. (18
), support these findings. They administered neuropsychological tests to 63 current heavy cannabis users who had smoked cannabis at least 5,000 times in their lives and to 72 control subjects who had smoked no more than 50 times in their lives. Although differences between the groups after 7 days of supervised abstinence were reported, no deficits were found after 28 days abstinence, after adjusting for various potentially confounding variables. These findings suggest that cognitive deficits associated with long-term cannabis use are reversible and related to recent cannabis exposure. If further research finds that there are, indeed, irreversible cognitive effects associated with long-term cannabis use, it is likely that the effects will be subtle and not robust.
Regardless of whether cannabis causes permanent or irreversible effects, the study of chronic long-term users by Gruber et al. (19
) produced some very disturbing findings. They asked long-term heavy users about various demographic attributes and also about their experiences with cannabis itself. Some striking results appeared when they compared 108 heavy cannabis users to 72 control subjects who had smoked cannabis fewer than 50 times, and a median of only 10 times, in their entire lives. The parental education in both groups was almost the same, with 58% of the heavy cannabis users and 51% of the controls having at least one parent who graduated from college. But the level of college graduation among the subjects themselves was dramatically different, with fewer than half of the heavy cannabis users having obtained a college degree as opposed to almost 80% of the control subjects. Similarly, the family income in both groups was roughly the same, with about 20–25% of the families reporting an income of less than $30,000 dollars a year. However, the income of the subjects themselves was dramatically different, with more than 50% of the heavy users reporting incomes of less than $30,000 per year almost twice as frequently as the controls.
One might hypothesize that the effects of cannabis use had nothing to do with these differences in educational attainment and income, that instead the differences arose from a decision by the long-term cannabis users not to pursue cultural norms like going to college or working at a high-level job. If this hypothesis were true, one would expect the long-term cannabis users to deny adverse effects of cannabis use. However, when cannabis users were asked to rate the effects of their own cannabis use as positive, neutral, or negative, they gave overwhelmingly negative ratings of the effects that cannabis had had on their social life (70%), their physical health (81%), their mental health (60%), their cognition (91%), their memory (91%), and their career (79%).
It seems a reasonable hypothesis that the negative effects these long-term heavy cannabis users reported are due to being acutely intoxicated every day. People intoxicated with cannabis demonstrate impairments in a variety of cognitive, perceptual, and psychomotor tasks. Tasks showing the most impairment involve short-term memory, sustained or divided attention, complex decision-making, and reaction time (20
). Ninety-seven percent of the heavy users in the Gruber study reported driving on a regular basis while intoxicated. Studies using driving simulators show marked impairment during acute cannabis intoxication, and accident statistics show that a disproportionate number of accidents occur in individuals intoxicated with cannabis and alcohol (21
). Forty-five percent of the heavy users have children. It is reasonable to presume that chronic cognitive impairment will adversely affect one’s ability to raise children. Moreover, 44% of the heavy users held jobs that potentially could endanger themselves or others, jobs such as electricians, nurses and pharmacists, chefs, nannies and daycare workers, and truck drivers. Other heavy users held jobs that, poorly performed, could greatly inconvenience others, jobs like postal workers and administrators of various types.