Cannabis and cognitive deficits in youth – reviewing the evidence

Studies have demonstrated that adolescents and young adults with frequent or heavy cannabis use have impaired ability to think clearly and solve problems. However, the magnitude of the cognitive difficulties seen with heavy cannabis use is dependent in part on study design. When studies require longer periods of initial abstinence before cognitive testing, youth demonstrate less cognitive impairment than when testing closer in time after recent use. But study findings are mixed. What does the latest science in this area tell us as a whole? 


The increasing legalization of cannabis (both medicinal and recreational) across the US and internationally may lead the public to underestimate the possible harms associated with chronic cannabis use. Meta-analyses in adults have shown frequent or heavy cannabis use is associated with decreased attention, memory, learning and executive functioningGiven that the brain systems underlying these functions are still developing during late adolescence, one would expect the negative effects of heavy or chronic cannabis use to be even greater in adolescence (e.g., ages 12-17) and early adulthood (e.g., ages 18-25)Though there have been many studies of cannabis use and cognitive functioning in adolescents and young adults, they have not been analyzed as a group to test which findings are found consistently across different studies (i.e. a meta-analysis). In this study, Scott and colleagues conducted a meta-analysis – a quantitative synthesis of many studies at once – to determine what areas of cognition might be impacted by cannabis use, and whether these impacts are more likely related to cannabis intoxication (short-term) or cannabis use over time (long-term).


In this meta-analysis of observational cross-sectional studies, authors analyzed 69 reports published between 1973 and 2017conducted primarily in the U.S., United Kingdom, Europe and Australia that focused on heavy, frequent or problematic cannabis use among youth, looking beyond acute effects of intoxication. The review included a total of 2,152 cannabis users with a mean age of 21 years as compared with 6,575 controls (e.g., noncannabis users). The mean age at cannabis initiation was 15 years old and 68% of the cannabis users were male, as compared with 56% of the controls. The magnitude of difference between cannabis and non-cannabis users (e.g., effect size) was determined and then analyzed as a group for each neurocognitive domain. The domains analyzed included learning, speed of information processing, delayed memory, attention, verbal/language abilities, visuospatial skills, motor abilities, and three types of executive functioning (abstraction/shifting, inhibition, & working memory). 


Cannabis use is associated with worse learning, attention, and executive functioning.

Compared to individuals with minimal cannabis exposure, individuals with heavy, frequent, or problematic cannabis use had worse functioning in a number of neurocognitive domains. Cannabis users had slightly worse (i.e., effect sizes were meaningful, but small in magnitude) performance on learning, abstraction and set-shifting, speed of processing, delayed memory, inhibition, working memory, and attention. As might be expected, these differences were even larger in treatment-seeking cannabis users, as they are likely to have more social and emotional challenges which may be related, in part, to their cannabis use. On the other hand, verbal/language, visuospatial and motor domains were similar among cannabis users and controls 

Cannabis users and controls had similar cognitive performance when cannabis users were abstinent for 3 days or longer before testing.

As more time passes since the most recent cannabis use, cognitive functioning is less impaired. When studies were separated by the abstinence period required before cognitive testing, the cognitive declines were largest when the period of no cannabis use was less than or equal to days. Importantly, there were no differences between heavy/problematic cannabis users and non-users when the cannabis using group were abstinent for 3 or more days before testing. This finding suggests that some of the cognitive impairments for cannabis users highlighted here may be due to acute cannabis intoxication or withdrawal 

Source: Scott et al. 2018. The figure above shows the 10 areas of cognitive function and the average effect size (d) of studies that addressed each area. Positive numbers indicate better performance among the cannabis groups and negative numbers indicate worse performance among the cannabis groups, relative to the comparison groups. A value of -0.2 or 0.2 indicates a small effect size (i.e. small difference between groups across studies), -0.5 and 0.5 indicate medium effect sizes, and -0.8 and 0.8 indicate large effect sizes. As illustrated, the average effect sizes across articles, depicted as blue dots, show that cognitive functions were generally worse among the cannabis-use groups than the comparison groups. If the error bars do not touch 0, that means the effect is large enough to be statistically reliable, or significant. In three areas (verbal/language, visuospatial, and motor performance) the error bar overlaps with 0, indicating cannabis users and non-users were statistically similar – the effect is not large enough to conclude that cannabis users truly had worse performance. It is important to note that cannabis users had meaningfully worse performance on several domains, these effects were consistently small in magnitude (e.g., Cohen’s d of approximately .2-.3.)


Heavy or frequent cannabis use during adolescence and young adulthood is associated with poorer cognitive functioning across numerous domains when individuals are tested within 3 days of their last useThese declines are potentially meaningful, but modestwhen tested within 72 hours of last useImportantly, these effects on cognition have the potential to impact academic functioning and driving ability for that period of time since last use, for several days after the feeling of being intoxicated has liftedThe studies in this meta-analysis were all cross-sectional, meaning cannabis users and non-users were compared at one point in time. Of note, while cannabis users’ worse performance may may due to cannabis use, and study authors in some cases tried to control for alternative explanations for cannabis users’ worse performance, there remain other potential reasons that cannabis users would have worse cognitive functioning. These include difficulties with planning, behavioral control, working memory (i.e., executive functioning) that were present before they ever used cannabis. Therefore, truly longitudinal studies, where individuals are measured at several points over time, are needed to fully understand the effects of cannabis on youth cognitive performance.  

Cognitive deficits in youth might be reversible or partially reversible with abstinencedifferent meta-analysis (looking at both adolescents and adults) found that the neurocognitive deficits associated with cannabis use were no longer significant after 25 days of abstinence. Separately, a systematic review found that after 14 days of cannabis abstinence, there may be sustained memory deficits with mixed findings in other neurocognitive domains. Taken together, these studies suggest that cannabis measurably impacts cognitive functioning in the days or weeks since last use, with some smaller longer-term impactsInterestingly those with earlier onset cannabis use (e.g., before age 16) did not have more impaired functioning than those with later onset cannabis use.

  1. All of the studies included were observational cross-sectional studies, of which only 699 of the 2152 cannabis users had been abstinent for 72 hours or longer (32% of the sample), meaning that the majority of the subjects included had been abstinent less than a few days, when cannabis is likely still active metabolicallyLess than 350 of the cannabis users had been abstinent for 2 weeks or longer (16% of the sample). 
  2. Most of the studies included fewer than 30 subjects in the cannabis-using arm with only two studies including greater than 100 cannabis-subjectsSince smaller studies tend to show larger effect sizes than larger ones, the effect sizes shown here may be an over-estimate. 
  3. Most studies tested less than four of the ten domains analyzed meaning that the cognitive domain data is collected from different subsets of studies per domain analyzedStudies that measure several domains in the same group of individuals are needed to gain a more comprehensive picture of how cannabis affects cognitive performance. 
  4. The proportion of THC in cannabis, the main psychoactive ingredient, has increased significantly over time. This is likely to be a confounder when pooling the results from studies conducted over a 45-year span, yet data collection year was not included as a factor in these analyses. 


  • For individuals and families seeking recovery: Frequent or heavy cannabis use seems to be associated with cognitive difficultiesRegardless of whether this is an effect of acute intoxication, early withdrawal or a long-term effect it should be considered in terms of safety (e.g. driving) and school/workplace functioning, particularly within three days of last cannabis use. This study suggests that the magnitude of the impairment decreases with increasing time since last use, suggesting that some, perhaps most, of the cognitive effects measured here might be reversible with increasing abstinence 
  • For treatment professionals and treatment systems: Motivating youth to abstain might be enhanced by reviewing the data suggesting that some, if not most, of the neurocognitive effects of sustained cannabis use might be reversed with abstinenceFurthermore, treatment-seeking cannabis users were shown to have greater neurocognitive deficits than those not seeking treatment for cannabis useIndividuals presenting to treatment may benefit from a focus on their thinking and problem solving abilities. For example, cognitive retraining may be considered as part of one’s treatment plan to address cognitive deficits and potentially enhance recovery outcomes 
  • For scientists: There are significant confounders in cross-sectional analyses, including cause and effect (for example, individuals with pre-existing cognitive deficits, particularly executive functioning, might be more prone to heavy or frequent cannabis use)The field would greatly benefit from large longitudinal studies (such as the Adolescent Brain Cognitive Development study) where cognitive functioning is studied prior to substance use and substance use data is collected in real time (rather than retrospective report)  
  • For policy makers: Disclosures about the possible neurocognitive effects of cannabis could be required at the growing number of medical marijuana dispensaries, in the same manner that risk/benefit discussions are expected prior to the initiation of any prescribed medicationGiven that the neurocognitive deficits are clearly measurable across multiple domains within 72 hours of last cannabis use, this should be taken into account in terms of the time frame to be considered for operating ‘under the influence’ of cannabis.


Scott, J. C., Slomiak, S. T., Jones, J. D., Rosen, A. F. G., Moore, T. M., & Gur, R. C. (2018). Association of Cannabis With Cognitive Functioning in Adolescents and Young Adults: A Systematic Review and Meta-analysis. JAMA Psychiatry, 75(6), 585-595. doi:10.1001/jamapsychiatry.2018.0335