In what we believe is the first study to examine the causal effects of acute cannabis administration in human adolescence and adulthood, we found two differing profiles of effects. Compared with adults, adolescents experienced blunted subjective, physiological and psychotomimetic effects of cannabis, while cannabis impaired inhibitory processes in adolescents but not adults. Specifically, on cannabis adolescents reported feeling less stoned, feeling less effect of the drug, less dry mouth and less cognitive disorganization than adults. The adults were also markedly more anxious and less alert during the cannabis session than the placebo session, while no session difference was found for the adolescents (however, since these group differences did not differ over time, these may be session effects rather than effects of cannabis). Indeed, there was no subjective rating on which adolescents reported greater drug effect than adults. Further, adults' but not adolescents' diastolic BP rose after cannabis.

Intriguingly, we found opposing effects between age groups on wanting of cannabis following drug administration. The adolescents did not show a typical satiety effect, wanting more cannabis post drug regardless of whether they had taken cannabis or placebo. Meanwhile the adults wanted less cannabis post drug, an effect that appears to be driven by a decrease in wanting following cannabis but not after placebo (although this putative interpretation remains tentative in the absence of a group × drug × time interaction).

In terms of cognitive effects, when intoxicated with cannabis adults showed greater impaired recall of prose following a delay than adolescents. After adjusting for drug order, the adults also had longer response times on the spatial working memory task following cannabis, while the adolescents were not affected. Although neither group was impaired at inhibiting a pre-potent response following cannabis, the adolescents but not adults were less accurate on the inhibition task after cannabis.

These results are in line with our first hypothesis that adolescents would be less sensitive to physiological, intoxication and anxiogenic effects compared with adults. These findings accord with the preclinical evidence that shows reduced anxiogenic, aversive and locomotor effects in adolescent rodents.^{35, 42, 43, 44} Further, while our second hypothesis predicted a greater degree of psychotomimetic effects following cannabis in the adolescents compared with the adults, we instead found the opposite: cognitive disorganization was especially elevated in adults compared with adolescents after cannabis. This unexpected finding is however in agreement with our first hypothesis of lesser intoxication effects in adolescents, perhaps suggesting a common mechanism by which adolescents are resilient to the acute negative effects of cannabis. It may also reflect an awareness in adults of the greater cognitive impairments they were experiencing, rather than amplified psychotic-like effects of cannabis per se. We also found that cannabis increased anhedonia symptoms in adults but not in adolescents; interestingly however, on placebo the adolescents had (non-significantly) higher levels of anhedonia than the adults.

Lastly, partial support for our third hypothesis, that we would see greater cognitive impairment following cannabis in adolescents than adults, was seen in greater impairment of response inhibition accuracy following cannabis in the adolescents compared with adults. However, contrary to expectations we did not see greater cannabis-related memory impairment in the adolescents, instead finding evidence of greater impairment in adults. Preclinical evidence for greater adolescent sensitivity to acute memory-impairing effects of cannabis is however inconsistent.⁷⁷ In adult humans cannabis appears to selectively impair episodic and working memory domains,⁷⁸ leaving other memory domains intact, while rodents typically become impaired on a wide range of memory tasks across domains including object recognition and spatial learning, implying that preclinical findings for cannabis and memory may be somewhat limited in translation.

These findings have important implications for public health, especially given the current changes in legislation that are making cannabis more available and may influence adolescent use in several parts of the globe. If adolescents do not feel satiated after an acute dose of the drug while also experiencing fewer negative effects, they may well use more cannabis in a smoking session than adults,⁴⁴ potentially contributing to the increased risk of long-term harms associated with younger age of use, including addiction.¹⁶ In turn, adults' experience of more negative effects of cannabis may limit their use and reduce their risk of harms, which would concur with the declining prevalence of cannabis use seen from early adulthood.⁴ A clear next step from these findings is therefore replication (importantly with females as well as males) and then assessment of naturalistic use of cannabis in different age groups, using measures that clearly record weight and potency of cannabis smoked,⁷⁹ topography of inhalation,⁵⁶ alongside ratings of subjective negative and positive intoxication effects. Tracking these participants longitudinally would be important in determining how these age-related sensitivities may impact in the long term on cannabis use patterns and mental and physical health outcomes.

Our study has several critical strengths. Importantly our groups were well matched on baseline measures including premorbid IQ and levels of anxiety, depression, impulsivity and schizotypy. This increases our confidence that participants in the two age groups were drawn from similar populations, and maximizes comparability between groups. Further the use of cannabis plant material, rather than extracted or synthetic cannabinoids, via an ecologically valid administration procedure (that is, inhalation) enhances the relevance of our findings to the real world use of this drug. Administering a known THC dosage that closely corresponds to that contained in about a third of a typical joint,⁵⁶ which was weight adjusted to allow for weight differences in adolescents and adults, are both strengths of this controlled study.

The study is not without limitations. First, we cannot speak to mechanism of the reported age-related sensitivities. Although the findings may represent age-related neural sensitivities to cannabis, there are a number of alternative explanations. Adolescents have a higher basal metabolism than adults,^{80, 81} alongside lower percentage body fat,^{82, 83} potentially affecting the speed of THC metabolism between the groups. Should THC and its by-products be metabolized more quickly in adolescents than adults, this could potentially result in the reduced subjective and episodic memory effects seen in adolescents; however, if drug metabolism in the adolescents was faster, a quicker decline of drug effects would be expected, which does not appear to be the case. Further, this would not explain the adolescent's impaired inhibition accuracy when the adults were unaffected. Group differences in the effect of cannabis on diastolic BP are also intriguing, though adolescents' diastolic BP was lower on both sessions at baseline, consistent with normative data.⁸⁴ This finding should also be viewed alongside a lack of a group difference in the more robust effect of cannabis increasing heart rate. Relatedly, participants were given a weight-adjusted dose, meaning that because adolescents typically weigh less than adults,⁸⁵ on average they received a lower dose. We cannot therefore rule out the possibility that the blunted effects seen in the adolescents are due to the reduced dose; however, again this would not explain the overall pattern of results including the adolescents' (but not adults') impaired response inhibition accuracy. Moreover, critically the weight of cannabis administered did not correlate with any outcome in either group. Groups could potentially be matched for body weight in future research, however this would result in biased samples that do not reflect the population as a whole. An important goal now is to investigate the mechanisms by which these apparent group differences occur, for instance, a first step would be to repeat key components of our protocol using an fMRI paradigm.

Second, all our participants were necessarily regular cannabis users, raising the possibility that our findings may be affected by group differences in past cannabis use. Although the groups were matched for cannabis abuse symptomology and days since last use, the adolescents did report more days of cannabis use per month than the adults (11 days versus 8 days); further while the adults had been using for more years, they had started using from an older age. Tolerance to some cannabis effects following frequent use has been reported (including for spatial working memory and episodic memory⁷⁸), however findings are inconsistent⁸⁶ and little is known about the development of cannabis tolerance and how different usage patterns affect this. As such it is possible that differing cannabis use histories and patterns may explain group differences in outcomes. Importantly however, none of our measures of cannabis use correlated with outcomes in both the adolescent and adult groups. Relatedly, the adolescents were more frequent and heavier cigarette smokers, with higher nicotine-dependence scores, and they had started tobacco smoking from a younger age than the adults. The groups were well matched for age of first alcohol use, but the adolescents were less frequent alcohol drinkers. It is possible that cross-tolerance to cannabis from previous alcohol or tobacco use may occur, though we are not aware of evidence demonstrating such an effect. A recent ecological momentary assessment study suggested that acutely tobacco use may offset acute impairment of working memory from cannabis,⁸⁷ though this has yet to be replicated in a controlled study. It is possible therefore that the age group differences in alcohol and cigarette use may be contributing to our findings.

Third, we recruited only males, due to differing age of puberty onset and potentially differing brain development trajectories between sexes, thus precluding generalization of findings to teenage girls. Samples in cannabis research are often predominantly male and gender effects have rarely been assessed, with inconsistent findings.⁷⁸ Some have shown heightened subjective⁸⁸ and working memory⁸⁹ effects in women compared with men, though others found no differences.⁹⁰ Recently it was found that younger age of cannabis use onset predicted poorer episodic memory in women but not men,⁹¹ suggesting that there may be age-dependent sex differences in the cognitive effects of cannabis. Given such findings, there is a clear evidence gap regarding the effects of cannabis in young women and girls and future research should assess whether our findings generalize to females.

Finally, since this was a novel study, with multiple statistical comparisons and limited or mixed evidence on which to base our prior hypotheses, it is important to treat these findings with caution. Replications with larger sample sizes (which can now be determined according to effect sizes reported in this paper) are required before strong conclusions can be drawn.

In conclusion, compared with adults, adolescent cannabis users experienced blunted subjective, physiological and memory impairing effects of cannabis. Further, adolescents were not satiated by cannabis and the drug impaired their inhibitory processes while leaving those of adults intact. To our knowledge, this is the first study to administer cannabis in a controlled setting to humans under 18, and it therefore represents a significant step forward in the translation of preclinical developmental psychopharmacology. In agreement with preclinical cannabinoid administration studies, we found evidence to suggest that human adolescents and adults are differentially sensitive to the acute effects of cannabis. Longitudinal research is now needed to determine the degree to which age-related sensitivities are indeed contributing to escalated use and increased risk of cannabis-related harms in adolescent cannabis users.