Associations of Regular Marijuana Use by Adolescent Boys With Verbal Memory and Perseveration

Abstract

Many American and Dutch adolescents use marijuana regularly. There is concern that such use may impair cognitive function more in adolescents than adults. We examined effects of regular marijuana use on long-term memory and perseveration among American and Dutch adolescents. We administered Buschke's Selective Reminding Test (BSRT) to assess long-term memory and the Wisconsin Card Sorting Test (WCST) to assess perseveration in male teenagers. Usable test data were obtained for 12 American marijuana users, 13 American controls, 9 Dutch marijuana users, and 12 Dutch controls. In BSRT, users showed lower overall long-term storage than controls (adjusted means ± SE's for numbers of words per trial of 9.4 ± 0.2, 13.4 ± 0.3, 11.7 ± 0.2, and 12.4 ± 0.2 for American users, Dutch users, American controls, and Dutch controls, respectively). Marijuana was associated with memory effects only in American, not Dutch, users. Bivariate Pearson correlations for American and Dutch users combined showed associations of lower total recall with more uses in the previous year and lifetime (r = –0.61 and r = –0.53, respectively); and more perseverative errors with more uses in the previous year (r = 0.55). Some findings were consistent with the possibility that regular adolescent marijuana use causes deficits in cognition, especially memory. However, a causal interpretation cannot be inferred from our findings and is challenging to reconcile with the observation of memory deficits only in American users. Our study was novel in examining the influence of nationality on marijuana's cognitive effects. More studies of this topic should compare effects across nationalities or cultures.

Keywords

Adolescence Buschke’s Selective Reminding Test cognition marijuana memory perseveration Wisconsin Card Sorting Test

Introduction

Many American and Dutch adolescents use marijuana regularly (ESPAD Group, 2016; Johnston et al., 2016). There is concern, supported by some human and animal research findings, that regular use may impair cognition and brain function more in adolescents than adults (Lubman et al., 2015). Adolescence is a period of substantial cognitive and brain development, and it is possible that normal maturational processes during adolescence, including synaptic pruning and white matter development, might be adversely affected by regular use (Lubman et al., 2015). This might be related to effects of regular use on the endocannabinoid system, which plays a role in brain development (Lubman et al., 2015).

We examined cognitive effects of regular adolescent use among American and Dutch users, administering Buschke's Selective Reminding Test (BSRT) (Buschke, 1973) to assess verbal long-term memory and the Wisconsin Card Sorting Test (WCST) (Heaton et al., 1993; Heaton & PAR Staff, 2005) to assess perseveration. This study was conducted as part of a funding initiative encouraging binational research collaboration on drug abuse between United States and Netherlands researchers, allowing comparisons between the two countries.

We previously found some impairments of BSRT performance in adult American marijuana users who used 7+ times weekly. Such use was associated with selective memory retrieval impairments (Block & Ghoneim, 1993). Learning and relearning a word list to a criterion of two consecutive recalls required more trials in users (Block, O'Leary, et al., 2002). We (Block, O'Leary, et al., 2002) also examined regional cerebral blood flow using positron emission tomography and found memory-related changes in several brain regions.

In a subsequent study (Jager et al., 2010) of American and Dutch regular adolescent marijuana users, we examined brain function using functional magnetic resonance imaging while verbal working memory and pictorial associative memory tests were administered. Memory-related changes in brain function were found during the former, but not the latter, test. Performance was not impaired in either test in this study or a previous study involving adult users (Jager et al., 2006; 2007).

The present report describes out-of-scanner cognitive testing that was conducted in the same session as the in-scanner testing of adolescents described previously (Jager et al., 2010). One objective was to assess memory storage and retrieval functions that might potentially be more sensitive to effects of regular marijuana use than the in-scanner memory assessments (Jager et al., 2010), based on our previous BSRT findings with adult marijuana users (Block, O'Leary, et al., 2002; Block & Ghoneim, 1993). Studies of BSRT with adolescent users are limited and inconsistent in their findings (Dougherty et al., 2013; Schwartz et al., 1989). We attempted to fill this gap in understanding how adolescent use influences BSRT performance. We hypothesized that users would show memory impairments.

Our second objective was to assess perseveration of concepts using WCST (Heaton et al., 1993; Heaton & PAR Staff, 2005). We previously found no impairments in adult American marijuana users who used 7+ times weekly in a locally developed concept formation test (Block & Ghoneim, 1993), but this test lacked the extensive research base of WCST and did not assess perseveration. The term, “perseveration,” is used in varying senses in psychology and psychiatry. In the context of WCST, it refers mainly to persistence in responding to a stimulus characteristic that is incorrect (Heaton et al., 1993). Our study focused on “perseverative errors,” which are probably the most commonly used outcome measure for WCST. WCST is a test of executive function and there is particular concern that regular adolescent marijuana use may impair executive function, which continues to develop during adolescence (Dahlgren et al., 2016). Studies of WCST with adolescent users are limited and inconsistent in their findings (dos Santos Rigoni et al., 2007; Hooper et al., 2014; Lane et al., 2007). We attempted to fill this gap in understanding how adolescent use influences WCST performance. We hypothesized that users would make more perseverative errors.

Our third objective was to compare American and Dutch users and controls in BSRT and WCST performance. Our study was novel in examining the influence of nationality on cognitive effects of marijuana. Relatively few studies of cognitive effects of drugs of abuse include cross-national or cross-cultural comparisons. Indeed, such comparisons are not common in neuropsychological studies in general for children, compared to adults, in domains other than testing of intelligence or academic achievement. In a 2008 literature review, Byrd and colleagues (Byrd et al., 2008) noted that limited data were available about culture- and ethnicity-related differences in children in cognitive domains such as learning, executive function, and attention. They found only ten articles overall (half of them American) that met their inclusion criteria and these all involved residents of a single country, unlike the present study. In our study, we had no basis to predict that American and Dutch participants would perform similarly or differently, either for users or controls.

Methods

Participants

Right-handed males aged 12 to 19 years were recruited. Controls were recruited from the community. Dutch marijuana users were recruited using a variety of strategies, including advertisements on the Internet and asking for help from schools. In the United States, however, because it would not be typical for marijuana-using adolescents to voluntarily disclose their use to parents, users had to be recruited mainly from drug abuse treatment and adolescent health and resource centers offering education and treatment programs to minors who got involved with the legal system because of possession of marijuana or other drugs.

Subjects were tested at the University of Iowa and the University Medical Center Utrecht. Exclusion criteria were medical or neurological problems; use of psychotropic medications; regular use of drugs of abuse other than marijuana, alcohol, or nicotine (10 or more episodes in lifetime); Axis I psychiatric diagnoses, except for conduct disorder, which is a common diagnosis in marijuana-using boys; intelligence quotient (IQ) scores under 80; and contraindications for magnetic resonance imaging (claustrophobia, metal objects, and full dental braces) (Jager et al., 2010). Twenty-three regular marijuana users with at least 200 uses in their lifetime and 25 controls with no or minimal experience with marijuana were included. Written informed consent was obtained from the adolescents and a parent or legal guardian in accordance with requirements of the Helsinki Declaration of 2004 and the local institutional review boards of the University of Iowa and the University Medical Center Utrecht. Subjects were reimbursed for participation.

Procedure

As described previously (Jager et al., 2010, 2013), subjects participated in two sessions separated by one week. Eligibility based on the inclusion and exclusion criteria was ascertained in the first session through a screening procedure, including locally developed, self-report questionnaires on drug use and medical history; a semi-structured computerized psychiatric interview (NIMH Diagnostic Interview Schedule for Children; C-DISC–version IV) (Shaffer et al., 2000); and testing to estimate IQ, i.e., four subtests (Block Design, Matrix Reasoning, Similarities, and Vocabulary) of the Wechsler Intelligence Scale for Children—4th edition (WISC-IV) (Baron, 2005; Wechsler, 2003).

BSRT, WCST, and functional magnetic resonance imaging were done in the second session. Participants abstained from marijuana, alcohol, and other drugs for at least 24 hours before the first session and remained abstinent until the second session was finished. Hence all marijuana users were abstinent for at least 8 days at the time of BSRT and WCST testing, resulting in minimal (if any) acute pharmacological effects of marijuana. Tobacco smoking was allowed until two hours before the second session (to avoid nicotine withdrawal). In both sessions, urine samples were collected to screen for drugs of abuse. Subjects were excluded if any drug other than marijuana was detected in urine during the second session. As described previously (Jager et al., 2010, 2013), users were excluded if cannabinoid metabolite levels in urine did not decrease between the first and second sessions, suggesting failure to abstain.

Tests

BSRT

BSRT (Buschke, 1973) involves consecutive learning and test trials on a list of words. The subject tries to recall the whole list on each test, but on learning trials after the first, is reminded only of the words missed on the immediately preceding test. Administration and scoring of our locally developed, computerized version was similar to methods described previously (Block & Ghoneim, 1993) except: A list of 15 common, concrete nouns (e.g., “drum,” “coffee”) from the Rey Auditory Verbal Learning Test (Rey, 1964) was used. Trials continued until the subject reached the criterion of recalling all words twice in a row or until the test was discontinued, whichever occurred first. Due to a computer programming difference, the test was discontinued after 10 trials for Dutch subjects but after 30 minutes for American subjects. Ten trials were analyzed for all subjects; for American subjects who continued longer, trials after the tenth were not analyzed. It was assumed that subjects who reached criterion before 10 trials would have continued to have perfect recall on the remainder of the 10 trials. Overall learning was assessed by the number of presentations required to reach criterion and by “total recall,” i.e., the total number of words recalled on successive trials. In addition, long-term storage, long-term retrieval, and short-term retrieval on successive trials were scored.

WCST

WCST involves sorting a sequence of stimulus cards by color, form, and the number of shapes shown. The test involves repeated unannounced changes in the correct sorting principle, which requires identification of, and shifting to, the new sorting principle. In the computerized version (Heaton & PAR Staff, 2005) of WCST (Heaton et al., 1993), the subject saw a series of images of cards on the screen, one at a time. Each card contained one to four identical forms (e.g., triangle) in a single color (e.g., red). There were 64 possible combinations of four numbers, four forms, and four colors. The subject moved the cards one by one (using the computer's mouse) underneath four fixed stimulus cards (one red triangle, two green stars, three yellow crosses, and four blue circles) shown on the screen according to a principle that the subject had to deduce from feedback (“right” or “wrong”) after each decision. The principles were unidimensional, requiring either sorting by number, form, or color; e.g., if the principle was color, all red cards should be placed under the one red triangle. After 10 consecutive correct decisions, the principle was shifted (from color to form to number, etc.) without warning. Testing ended when the subject completed six runs of 10 correct responses or responded to 128 cards, whichever occurred first. The number of perseverative errors was scored. Responses were classified as “perseverative” when the subject persisted in responding based on an incorrect principle. Perseverative errors occur most commonly following 10 consecutive correct responses to a previously correct principle, when the principle has changed but the subject continues responding based on the previous principle. Thus, perseveration may reflect difficulty in giving up an old for a new principle or in recognizing new possibilities (Spreen & Strauss, 1998, p. 220).

Statistical analyses

Quantitative demographic, drug use, and other characteristics of subjects were analyzed by analyses of variance with user group (users versus controls) and nationality (American vs. Dutch) as between-subjects factors. Qualitative characteristics (conduct disorder and treatment status) were analyzed using Fisher's exact tests.

Performance in BSRT was analyzed with mixed-model repeated-measures analyses of covariance using the SAS® MIXED procedure (SAS Institute Inc., 2004), with trial (1–10) as the repeated measure, user group and nationality as between-subjects factors, and IQ as a covariate. Separate analyses were done for total recall, long-term storage, long-term retrieval, and short-term retrieval. An analysis with user group and nationality as factors and IQ as a covariate was done for the number of perseverative errors in WCST. The numbers of subjects who reached criterion in BSRT were fit to a log-linear model (using the SAS® CATMOD procedure) (SAS Institute Inc., 2004) to test the main effects of user group and nationality and their interaction.

To clarify the interpretation of the results of the analyses of covariance, stepwise regression analyses were done for users’ total recall, averaged over trials, in BSRT and the number of perseverative errors in WCST, using as predictors the following between-subject variables: numbers of uses in the previous year and ages of first use of marijuana, alcohol, and nicotine; lifetime number of uses of marijuana; chronological age; IQ; nationality; and presence of conduct disorder. Both the threshold entry probability and threshold elimination probability were set at p = 0.15. Rather than excluding from the stepwise regression analyses one marijuana user who had never used alcohol or nicotine, his ages of first use for these drugs were censored at his current chronological age. Simple linear regressions and Pearson product-moment correlations were also examined.

Results

Exclusions and numbers of subjects

Based on the urine test results (Jager et al., 2010, 2013), two Dutch users had to be excluded from the analyses, one for cocaine use and one for marijuana use. There were usable cognitive test data for 46 subjects, consisting of 12 American users, 13 American controls, 9 Dutch users, and 12 Dutch controls. However, due to computer problems, BSRT data for one American user were unusable and WCST data for one Dutch control were lost, leaving 45 subjects for analysis in each test.

Characteristics of the subjects

Demographic, drug use, and other characteristics of the subjects are shown in Table 1. Relative to controls, marijuana users had lower IQ scores, higher use of marijuana in their lifetime, and higher use of marijuana, alcohol and nicotine in the previous year. For use of nicotine in the previous year, there was a greater difference between marijuana users and controls for American than Dutch subjects. The age of first use of alcohol use was later for American than Dutch subjects. Some subjects had very limited experience in their lifetime (1–9 times) with drugs of abuse other than marijuana, alcohol, and nicotine, i.e., 3,4-methylenedioxy-methamphetamine (MDMA), psilocybin, and nitrous oxide for both marijuana users and controls; and amphetamines, cocaine, and lysergic acid diethylamide (LSD) for marijuana users, but not controls (Jager et al., 2010, 2013). Conduct disorder and being in treatment were both more frequent among marijuana users than controls and among American than Dutch subjects. Among American users, 9/12 = 75% had conduct disorder and were in treatment. Among the other groups, only one subject (an American control, 1/13 = 8%) had conduct disorder and only one subject (a Dutch user, 1/9 = 11%) was in treatment.

Table 1.

Demographic, drug use, and other characteristics of subjects.

	Marijuana users		Controls
	American	Dutch	American	Dutch
N	12	9	13	12
Age (years)	16.7 ± 0.2	17.9 ± 0.3	16.7 ± 0.4	16.7 ± 0.3
IQ**	99.6 ± 3.5	104.1 ± 2.7	113.3 ± 3.3	109.8 ± 3.6
Marijuana
Lifetime (number of uses)*	6,428.5 ± 2,921.8	1,044.6 ± 167.9	0.5 ± 0.5	3.0 ± 1.5
Previous year (number of uses)****	984.6 ± 294.9	442.7 ± 66.0	0.2 ± 0.2	1.9 ± 1.1
Age of first use (years)	12.5 ± 0.7	14.6 ± 0.4	16.0^a	14.8 ± 0.7
Alcohol
Previous year (number of uses)**	754.9 ± 252.9	609.8 ± 140.6	9.2 ± 7.6	343.3 ± 103.3
Age of first use (years)^††	13.3 ± 0.6	12.6 ± 0.5	15.6 ± 0.4	12.3 ± 0.7
Nicotine
Previous year (number of uses)****^,‡	4,079.2 ± 807.0	2,303.8 ± 851.6	0.3 ± 0.3	637.0 ± 294.3
Age of first use (years)	11.6 ± 0.8	14.3 ± 0.4	16.0^a	13.7 ± 0.8
Conduct disorder (percent)**^,†††	75.0	0.0	7.7	0.0
In treatment (percent)****^,†	75.0	11.1	0.0	0.0

Note. The values shown are means ± standard errors for quantitative characteristics and percentages for qualitative characteristics. The numbers of uses of marijuana, alcohol, and nicotine were defined for subjects in terms of consumption of numbers of "joints," drinks, and cigarettes, respectively. The values for the numbers of uses of marijuana in the previous year and lifetime were deemed unreliable and considered missing for one American marijuana user. The means for ages of first use of marijuana, alcohol, and nicotine were based on subjects who had used each drug; these were fewer than 75% of the subjects for marijuana, alcohol, and nicotine for American controls (N = 1, 5, and 1, respectively) and marijuana for Dutch controls (N = 6). Significance was determined for quantitative characteristics by analysis of variance with user group and nationality as between-subjects factors and for qualitative characteristics by Fisher's exact tests, and is indicated as follows: *p < 0.05, **p < 0.01, ****p < 0.0001 for the difference between marijuana users and controls; ^†p < 0.05, ^††p < 0.01, ^†††p < 0.001 for the difference between American and Dutch subjects; ^‡p < 0.05 for the interaction of user group and nationality.

^aThe standard error is undefined because N = 1.

BSRT

Subjects who reached criterion

The percentages of subjects who reached the criterion of two consecutive perfect recalls, demonstrating reliable recall of the entire list, within 10 trials varied depending on user group and nationality. None of the American users reached criterion (0/11 = 0%), in contrast with substantial percentages who reached criterion among American controls (6/13 = 46%), Dutch users (5/9 = 56%), and Dutch controls (6/12 = 50%). The interaction of user group and nationality was significant, χ²(1)=6.5, p < 0.05, indicating that the American users differed from the other combinations of user group and nationality in the percentages of subjects who reached criterion. The overall differences between users and controls and between American and Dutch subjects fell short of significance, χ²(1)=2.5, p = 0.12 and χ²(1)=3.5, p = 0.06, respectively.

Analyses of covariance

Total recall is shown in Figure 1. Total recall varied according to user group and nationality, with the interaction of these factors being significant, F(1,40) = 59.0, p < 0.0001. The overall difference between American and Dutch subjects was significant, F(1,40) = 90.9, p < 0.0001, while the overall difference between users and controls was not. As Figure 1 illustrates, American users recalled fewer words than any of the other combinations of user group and nationality. The other combinations of user group and nationality differed relatively less from one another. The poorer recall of the American users, relative to the other combinations of user group and nationality, was evident during the first trial and remained relatively constant over successive trials, although total recall naturally increased over successive trials, F(9,369) = 34.5, p < 0.0001. The covariate, IQ, significantly influenced total recall, F(1,40) = 106.6, p < 0.0001.

Figure 1.

Adjusted means ± standard errors for total recall in BSRT by trial for American and Dutch marijuana users and controls. The values were derived from analyses of covariance that included IQ as a between-subjects covariate (see text) and are adjusted to the mean IQ. Thus, the standard errors differ by user group and nationality, but not by trial.

Table 2 shows means pooled over trials for American and Dutch users and controls for total recall, long-term storage, long-term retrieval, and short-term retrieval. The pattern of significance for long-term storage, long-term retrieval, and short-term retrieval was the same as for total recall, except that the overall difference between users and controls was significant for long-term storage, F(1,40) = 6.3, p < 0.05. For long-term retrieval and short-term retrieval, this difference fell just short of significance, F(1,40) = 4.0, p = 0.052 for each. All the other effects were significant at the p < 0.0001 level: for the overall differences between American and Dutch subjects, F(1,40) = 104.2, 120.5, and 62.0 for long-term storage, long-term retrieval, and short-term retrieval, respectively; for the interaction of user group and nationality, F(1,40) = 48.0, 68.1, and 28.5, respectively; for trial, F(9,369) = 53.1, 35.9, and 13.8, respectively; and for IQ, F(1,40) = 92.7, 58.0, and 40.2, respectively. As Table 2 shows, the patterns for long-term storage and long-term retrieval resembled the pattern for total recall, i.e., American users showed lower long-term storage and long-term retrieval than any of the other combinations of user group and nationality. In contrast, the pattern for short-term retrieval was reversed, i.e., American users showed higher short-term retrieval than any of the other combinations.

Table 2.

Performance in BSRT and WCST.

	Marijuana users		Controls
	American	Dutch	American	Dutch
BSRT
Total recall	10.3 ± 0.2	13.3 ± 0.2	11.9 ± 0.2	12.2 ± 0.2
Long-term storage	9.4 ± 0.2	13.4 ± 0.3	11.7 ± 0.2	12.4 ± 0.2
Long-term retrieval	8.3 ± 0.2	12.9 ± 0.3	10.8 ± 0.2	11.4 ± 0.2
Short-term retrieval	2.0 ± 0.1	0.5 ± 0.1	1.1 ± 0.1	0.8 ± 0.1
WCST
Perseverative errors	11.3 ± 1.6	6.0 ± 1.7	9.8 ± 1.5	9.6 ± 1.6

Note. Adjusted means ± standard errors are shown for the numbers of words in BSRT and perseverative errors in WCST. For BSRT, adjusted means are averaged over trials. The values were derived from analyses of covariance that included IQ as a covariate (see text) and are adjusted to the mean IQ.

Stepwise regression and correlation analyses

In the stepwise regression analysis for users for total recall, the predictor variables entered into the model, in order, were nationality (1 = Dutch, 0 = American), the number of uses of marijuana in the previous year, and presence of conduct disorder (1 = present, 0 = absent). No predictor variables met the criterion for elimination. Based on the final model, the regression equation for predicting total recall was 4.2 × nationality + 2.1 × conduct disorder – 0.002 × marijuana use in the previous year + 9.6. The significance levels associated with nationality, conduct disorder, and marijuana use in the previous year were p < 0.01, p < 0.10, and p < 0.01, respectively; for the overall model, F(3,15) = 12.6, p < 0.001. Dutch nationality and more marijuana use in the previous year were associated with higher and lower total recall, respectively. There was a marginal trend toward conduct disorder being associated with higher total recall.

For the users, the one quantitative characteristic that was selected in the stepwise regression analysis, the number of uses of marijuana in the previous year, showed a significant Pearson product-moment correlation with total recall, r = –0.61, p < 0.01 (with other characteristics disregarded). The only other quantitative characteristics listed in Table 1 that correlated significantly with total recall were lifetime number of uses of marijuana, r = –0.53, p < 0.05 and age of first use of nicotine, r = 0.54, p < 0.05.

The effects of nationality and the number of uses of marijuana in the previous year on total recall of users are shown in Figure 2. All of the individual values for American users fell below the mean for controls, indicated by the horizontal line; whereas all the individual values for Dutch users fell above or near this line. The lines of best fit obtained by simple linear regression separately for American and Dutch users showed very similar slopes for predicting total recall from the number of uses of marijuana in the previous year, –0.0016 and –0.0012, respectively. When American and Dutch users were combined, the slope was a little steeper, –0.0020.

Figure 2.

Effects of nationality and the number of uses of marijuana in the previous year on total recall in BSRT for marijuana users. The values of total recall are means over trials. In addition to values for each subject, the lines of best fit obtained by simple linear regression for predicting total recall from the number of uses of marijuana in the previous year are shown for American and Dutch users separately and for both nationalities combined. The mean for controls for both nationalities combined is indicated by the horizontal line.

Another feature illustrated by Figure 2 is the much wider range of the number of uses of marijuana in the previous year for American than Dutch users. The users who reported more than two uses of marijuana per day on average during the previous year were all American and all showed relatively low recall (70% or fewer of the words on average). Only 13% (2/15) of the less frequent users showed such low recall.

Stepwise regression analyses were not done for long-term storage, long-term retrieval, and short-term retrieval because Pearson product-moment correlations of these variables with total recall were extremely high in magnitude (r = 0.97, 0.99, and –0.93, respectively, p < 0.0001 for all).

WCST

Analysis of covariance

Table 2 shows means for the number of perseverative errors. There were no significant effects, although there were trends toward effects of nationality, the interaction of user group and nationality, and IQ, F(1,40) = 3.2, 2.7, and 4.0, respectively; p's = 0.08, 0.11, and 0.052, respectively. Similar to the findings in BSRT, American users tended to have more perseverative errors than any of the other combinations of user group and nationality.

Stepwise regression and correlation analyses

In the stepwise regression analysis for users for the number of perseverative errors, the predictor variables entered into the model, in order, were conduct disorder and the number of uses of marijuana in the previous year. No predictor variables met the criterion for elimination. Based on the final model, the regression equation for predicting the number of perseverative errors was 5.0 × conduct disorder + 0.003 × marijuana use in the previous year + 5.0. The significance level associated with both conduct disorder and marijuana use in the previous year was p < 0.10; for the overall model, F(2,17) = 6.1, p < 0.05. Presence of conduct disorder and more marijuana use in the previous year were both marginally associated with more perseverative errors.

For the users, the one quantitative characteristic that was selected in the stepwise regression analysis, the number of uses of marijuana in the previous year, showed a significant Pearson product-moment correlation with the number of perseverative errors, r = 0.55, p < 0.05 (with other characteristics disregarded). None of the other quantitative characteristics listed in Table 1 correlated significantly with the number of perseverative errors. Nor did the number of perseverative errors correlate significantly with total recall in BSRT, r = –0.35, p = 0.13.

Summary of overall and frequency-related effects of regular adolescent marijuana use

In summary, there were some statistically significant overall and frequency-related effects of regular adolescent marijuana use. In BSRT: (1) Users showed lower overall long-term storage than controls. (2) Bivariate Pearson correlations showed that lower total recall was associated with more uses of marijuana in the previous year and lifetime (the former, but not the latter, was significant in the final stepwise regression model). (3) In WCST, a bivariate correlation showed that more perseverative errors were associated with more uses in the previous year (although this association was marginal in the final stepwise regression model).

Discussion

Considerations complicating the interpretation of marijuana's effects

Were the findings concerning overall and frequency-related effects of regular adolescent marijuana use described above truly caused by this use? Inferring causality in an observational study, such as the present one, is always uncertain, but especially so in the present study, considering the interactions of user group and nationality on all five measures of memory function in BSRT. Associations of regular adolescent marijuana use with memory were observed only in American, not Dutch, users. These interactions may have been related, at least in part, to distinctive characteristics of the American users. The four most frequent users were all American, not Dutch, and all showed markedly low recall. Moreover, the American users included high percentages (75%) who had conduct disorder and were in treatment, relative to small percentages of Dutch users and American and Dutch controls. The higher percentage of American than Dutch users with conduct disorder was probably largely attributable to site differences in recruitment strategies, with most American, but not Dutch, users recruited from treatment programs. Boys ending up in such programs may display more externalizing behavior problems and more often meet criteria for conduct disorder.

Overall, the present results suggest that memory was more sensitive than perseverative errors to any consequences of regular adolescent marijuana use–but, even regarding memory, the complicated findings mandate great caution in interpretation and the associations with marijuana should be considered tentative until further verification.

Thinking in terms of a continuum from “not at all caused by marijuana use” to “completely caused by marijuana use” may help in interpreting the present results. An interpretation toward the causal end is that there is a threshold frequency of marijuana use below which there are no or negligible adverse effects on memory, and above which there are. This threshold might be, roughly, more than two uses per day on average during the previous year. The four users who exceeded this threshold had poor memory, and, because they were all American, interactions of user group and nationality occurred for all measures of memory.

An interpretation toward the “not at all caused by marijuana use” end of the continuum is that characteristics other than marijuana use or its frequency were responsible for our findings. The prime candidates were the disproportionate percentages of American users, relative to the other groups, who had conduct disorder and were in treatment. The findings that conduct disorder was only marginally related to total recall and perseverative errors in the final stepwise regression models, and that its marginal trend for total recall was to be associated with better performance, undercut this interpretation.

Primarily because it was difficult to recruit adolescent marijuana users, there was little opportunity to go beyond rejecting those who did not meet the inclusion/exclusion criteria, and, additionally, reject others to achieve good matching of users and controls, or of American and Dutch users. Users and controls differed not only in conduct disorder and being in treatment, but in IQ and experience with alcohol and nicotine. Additionally, we did not obtain information on other characteristics such as level of education, socioeconomic status, and family relationships. There might have been differences in these and other characteristics between users and controls, or between American and Dutch users, that influenced performance on WCST and BSRT. Many subjects were below the age at which children can drop out of school (generally 16 years in both Iowa and the Netherlands), possibly attenuating differences in years of education. A study with better matching among the combinations of user group and nationality would be needed to distinguish possible influences of marijuana use, compared to other characteristics, on test performance. Even such a study, if it had a cross-sectional design like ours, would likely not prove whether marijuana use caused certain cognitive effects or whether these cognitive differences were predictors of marijuana use. Because the present study cannot resolve questions of causality, the observed effects may best be conceptualized as correlates of a “marijuana-using lifestyle,” particularly among adolescent male American users. We also cannot know whether these effects would generalize to female users.

Certain social and cultural differences between the United States and the Netherlands might have contributed to the interactions of user group and nationality. In the 1970s, the Netherlands adopted a policy of not enforcing violations involving small amounts of marijuana (MacCoun, 2011). “Coffeeshops” selling marijuana became tolerated, in contrast to the legal sanctions for marijuana use in the United States. The likelihood of being treated for marijuana use is greater in the United States, mainly because of more criminal justice referrals (MacCoun, 2011). There are also broader social and cultural differences between the two countries (Hofstede Insights, 2020; Hofstede et al., 2010). For example, the two of Hofstede et al's six cultural dimensions that most notably distinguish the United States from the Netherlands are the former's higher masculinity and lower long-term orientation (Hofstede Insights, 2020).

Although neither the cultural differences regarding marijuana between the two countries, nor the broader cultural differences, pertain directly to memory function, ways by which they might impact such function can be envisioned, including differences in expectations about the cognitive effects of marijuana, or differences related to language, self-image of ability, goal directedness, or motivation to perform well. In the present study, the users showed some changes in brain function during a monetary incentive delay test, possibly indicative of altered motivational brain circuitry (Jager et al., 2013). Potential confounding by cultural divergence of apparent effects of long-term use of marijuana or other psychoactive drugs on BSRT performance has been discussed and illustrated by a study comparing BSRT performance of Navajos who were or were not peyote users (Pope & Yurgelun-Todd, 2004). Conceivably, the American marijuana users in the present study might have diverged more culturally from the other three groups than these groups did from one another. The possibility of subcultural differences precludes any simplistic attribution of findings to differences between American versus Dutch culture that assumes a uniform and representative character of participants within each nationality, which was neither the case among subjects in the present study, nor among the populations of either country as a whole (particularly a country as large and diverse as the United States).

American and Dutch controls in our study performed comparably on both WCST and BSRT. This was unsurprising, although we know of no other studies providing evidence of comparable performance between American and Dutch subjects. Although WCST and BSRT were developed and widely used in the United States, they have been used in the Netherlands (Bruins et al., 1990; van den Berg et al., 2020). More articles on WCST have actually been published by Netherlands-based publishers (one large one in particular) than those of any other country (Silva-Filho et al., 2011). WCST is not heavily language-dependent compared to BSRT. BSRT norms for a Dutch population are unavailable, but Flemish norms for Belgians have been reported (Thielen et al., 2019).

Although five measures of memory function in BSRT were analyzed, we did not adjust the p values for significance reported above for the number of analyses done, e.g., by dividing by five. These measures were intended to reflect different aspects of memory function, but were not statistically independent in the sense that distinct, uncorrelated cognitive tests would be. Had we divided unadjusted p values by five, the interactions of user group and nationality for four of these five measures, reported as p < 0.0001, would have remained significant, i.e., p < 0.0005. The interaction of user group and nationality for the percentages of subjects who reached criterion, and other effects reported as p < 0.05, would have lost statistical significance.

The sample size in the present study was modest. If we had been able to achieve a larger sample size, some effects that did not attain statistical significance might have done so. The sample size proposed in the grant applications for this research (N = 12 for each combination of user group and nationality) was calculated based on the functional magnetic resonance imaging aspects (Jager et al., 2010, 2013) of the research, not the WCST and BSRT. It was proposed to provide adequate power (power = 0.80, α = 0.05) for comparisons of user groups for each nationality separately, assuming that the activation differences observed were similar in magnitude to relatively large activation differences in certain regions of association cortex in previous working memory studies by the Netherlands research group. The proposed sample size was also influenced by the cost of imaging relative to the financial constraints of the funding opportunities. For both the imaging aspects (Jager et al., 2010, 2013) and the WCST and BSRT, data from the two nationalities were pooled to increase statistical power. For a simple comparison of user groups pooled over nationalities, post hoc achieved power (α = 0.05), based on the observed effect size for long-term storage in BSRT, which showed a significant difference between users and controls, was about 0.95.

Comparison with findings of other studies

Memory as assessed by BSRT

Two past reports are most directly pertinent to the present study: A study of regular adolescent marijuana users showed impairment on total recall on BSRT, but not consistent long-term retrieval, following brief abstinence (Dougherty et al., 2013). This report does not indicate whether any users were in treatment. A study of regular adolescent marijuana users who were in treatment showed no impairment on BSRT after several days of abstinence or when retested 6 weeks later (Schwartz et al., 1989).

Of intermediate relevance to the present study, because it involved adults rather than adolescents, is a study about effects of regular marijuana use on BSRT described in several reports (Pope et al., 2001, 2002; 2003; Pope & Yurgelun-Todd, 2004; Tzilos et al., 2005). Adult, frequent, long-term, current marijuana users showed impairments after seven days, but not 28 days, of abstinence. No association of performance of current users with duration of use was found. Long-term users who had used minimally for the previous three months did not show impairments. Users who began using before age 17 showed impairments after 28 days of abstinence before, but not after, adjustment for differences in verbal intelligence. Magnetic resonance imaging showed no associations of performance with hippocampal volumes.

The above-described studies of BSRT and marijuana use were conducted in the United States. Their participants were likely more comparable to the American than Dutch users in the present study. The findings of these past studies were mixed, with some indicating impairments on BSRT associated with marijuana use, and others not. The period of abstention from marijuana use before administration of BSRT appeared to influence the findings, but this influence was not clear-cut, either. The abstention period in the present study was variable (minimum of 8 days and mean of 5.1 weeks), but longer than in some past studies. There is also a study showing some impairments on BSRT in adult, frequent, very long-term marijuana users from Costa Rica (Fletcher et al., 1996; Page et al., 1988). The cultural contexts and demographic characteristics of these users differed from both the American and Dutch users in the present study.

Association of marijuana use with changes in retrieval processes in BSRT

In the present study, the American users showed higher short-term retrieval, but lower long-term retrieval, compared to any of the other combinations of user group and nationality, i.e., they showed an apparently compensatory increase in reliance on short-term retrieval. We observed a similar, though equivocal, compensatory increase in reliance on short-term retrieval in BSRT in a previous study in regular adult users who used 7+ times weekly (Block & Ghoneim, 1993). We also observed apparently increased reliance on short-term memory in such users in single-trial recall of a list of words, as indicated by increased recall of words toward the end, relative to the middle, of the list (Block, O'Leary, et al., 2002). We found similar trends towards compensatory increases in short-term retrieval in BSRT in regular users of various drugs of abuse (Block, Erwin, et al., 2002) and volunteers experiencing acute effects of benzodiazepines (Block & Berchou, 1984).

Findings based on retrieval measures in BSRT have sometimes diverged from findings based on other techniques for assessing effects of drugs. Some studies comparing recall of materials learned before versus after drug administration suggested that acute administration of drugs such as marijuana and benzodiazepines impaired long-term storage, but not long-term retrieval, in contrast with BSRT, which implies that long-term retrieval is also affected. We have discussed this issue elsewhere (Block & Berchou, 1984), as have others (Lister, 1985). Such differences might derive, in part, from the way Buschke's scoring procedures conceptualize and operationalize “retrieval” (Block & Berchou, 1984).

WCST

Some studies of regular adolescent marijuana use showed more perseverative errors and fewer categories achieved on WCST (dos Santos Rigoni et al., 2007; Lane et al., 2007). Another study did not find effects of regular adolescent marijuana use on tests of executive function, including WCST (Hooper et al., 2014).

Other reports are of intermediate relevance to the present study because they involved adult rather than adolescent marijuana users (Bolla et al., 2002; Dahlgren et al., 2016; Fontes et al., 2011; Gruber et al., 2012; Hermann et al., 2007; Lamers et al., 2006; Lyons et al., 2004; Pope et al., 2001, 2002, 2003; Pope & Yurgelun-Todd, 1996; Solowij et al., 1997, 2002). Many, though not all (Hermann et al., 2007; Lamers et al., 2006; Lyons et al., 2004; Solowij et al., 2002), found some impairments associated with marijuana use in perseverative errors or other scores. There have been tendencies for impairments to be associated with earlier onset of use, more frequent use, and shorter abstinence periods prior to testing, but findings have not been consistent, e.g., long-term, regular users were not impaired after brief abstinence (Solowij et al., 2002). Our observation of a bivariate correlation indicating more perseverative errors with more uses in the previous year provides one more finding suggesting that impairment is related to frequency of use.

Conclusion

Some of our findings, i.e., the overall and frequency-related associations observed in the tests, are consistent with the possibility that regular adolescent marijuana use causes deficits in cognition–more so for memory decrements than perseverative errors. It is uncommon when studying cognitive effects of marijuana or other drugs of abuse to conduct cross-national studies and compare effects across nationalities or cultures, as we did. Our cross-national findings are harder to reconcile with a causal interpretation of cognitive effects of marijuana use, especially the findings that the marijuana-related memory decrements occurred only in American, not Dutch, users. The present study may be the first one to test for (and find) an influence of nationality on the cognitive effects of regular adolescent marijuana use. Additional studies of this kind would be valuable. Consideration of nationality and culture might help in making sense of the many discrepant findings among studies concerning cognitive effects of regular marijuana use, both in adolescents and adults.

Footnotes

Authors' Note

Gerry Jager is currently affiliated with the Department of Agrotechnology and Food Sciences, Wageningen University, Wageningen, the Netherlands and Maartje Luijten is currently affiliated with the Behavioural Science Institute and Orthopedagogics: Family and Behaviour, Radboud University, Nijmegen, the Netherlands.

Acknowledgments

The authors thank Jordan Zuccarelli, Megan Becker, and Catherine Fruehling-Wall for data acquisition in Iowa; C. Kice Brown for statistical analyses; Alane Tranel for assistance with the literature review; and all American and Dutch institutions that assisted with recruitment of the participants.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by grants of the National Institute on Drug Abuse of the National Institutes of Health (United States) under National Institute on Drug Abuse grant number R01DA019338 and the Netherlands Organization for Health Research and Development under ZonMW grant number 31100003 as part of their Netherlands–U.S. Binational Research Collaboration on Drug Abuse and Addiction.

ORCID iD

Robert I. Block

Author Biographies

Robert I. Block PhD, is an associate professor in the Department of Anesthesia at the Roy J. and Lucille A. Carver College of Medicine, University of Iowa, Iowa City, IA, USA.

Gerry Jager PhD, is an associate professor in the Department of Agrotechnology and Food Sciences at Wageningen University, Wageningen, the Netherlands. At the time the research described in this article was conducted, she was affiliated with Rudolf Magnus Institute of Neuroscience, University Medical Center Utrecht, Utrecht, the Netherlands.

Maartje Luijten, PhD, is an assistant professor in the Behavioural Science Institute and Orthopedagogics: Family and Behaviour at Radboud University, Nijmegen, the Netherlands. At the time the research described in this article was conducted, she was affiliated with Rudolf Magnus Institute of Neuroscience, University Medical Center Utrecht, Utrecht, the Netherlands.

Nick F. Ramsey PhD, is a professor in the Rudolf Magnus Institute of Neuroscience at the University Medical Center Utrecht, Utrecht, the Netherlands.

References

Baron

I. S.

(2005). Test review: Wechsler Intelligence Scale for Children—Fourth edition (WISC-IV). Child Neuropsychology: A Journal on Normal and Abnormal Development in Childhood and Adolescence, 11(5), 471–475. https://doi.org/10.1080/09297040590951587

Block

R. I.

Berchou

(1984). Alprazolam and lorazepam effects on memory acquisition and retrieval processes. Pharmacology, Biochemistry, and Behavior, 20(2), 233–241. https://doi.org/10.1016/0091-3057(84)90248-x

Block

R. I.

Erwin

W. J.

Ghoneim

M. M.

(2002). Chronic drug use and cognitive impairments. Pharmacology Biochemistry and Behavior, 73(3), 491–504. https://doi.org/10.1016/s0091-3057(02)00816-x

Block

R. I.

Ghoneim

M. M.

(1993). Effects of chronic marijuana use on human cognition. Psychopharmacology, 110(1-2), 219–228. https://doi.org/10.1007/bf02246977

Block

R. I.

O'Leary

D. S.

Hichwa

R. D.

Augustinack

J. C.

Boles Ponto

L. L.

Ghoneim

M. M.

Arndt

Hurtig

R. R.

Watkins

G. L.

Hall

J. A.

Nathan

P. E.

Andreasen

N. C.

(2002). Effects of frequent marijuana use on memory-related regional cerebral blood flow. Pharmacology, Biochemistry, and Behavior, 72(1-2), 237–250. https://doi.org/10.1016/s0091-3057(01)00771-7

Bolla

K. I.

Brown

Eldreth

Tate

Cadet

J. L.

(2002). Dose-related neurocognitive effects of marijuana use. Neurology, 59(9), 1337–1343. https://doi.org/10.1212/01.wnl.0000031422.66442.49

Bruins

Kumar

Schneider-Helmert

(1990). Influence of desglycinamide-(Arg⁸) vasopressin on memory in healthy subjects. Neuropsychobiology, 23(2), 82–88. https://doi.org/10.1159/000119432

Buschke

(1973). Selective reminding for analysis of memory and learning. Journal of Verbal Learning and Verbal Behavior, 12(5), 543–550. https://doi.org/10.1016/S0022-5371(73)80034-9

Byrd

Arentoft

Scheiner

Westerveld

Baron

I. S.

(2008). State of multicultural neuropsychological assessment in children: Current research issues. Neuropsychology Review, 18(3), 214–222. https://doi.org/10.1007/s11065-008-9065-y

10.

Dahlgren

M. K.

Sagar

K. A.

Racine

M. T.

Dreman

M. W.

Gruber

S. A.

(2016). Marijuana use predicts cognitive performance on tasks of executive function. Journal of Studies on Alcohol and Drugs, 77(2), 298–308. https://doi.org/10.15288/jsad.2016.77.298

11.

dos Santos Rigoni

da Silva Oliveira

de Moraes

J. F. D.

Zambom

L. F.

(2007). O consumo de maconha na adolescência e as conseqüências nas funções cognitivas. Psicologia em Estudo, 12(2), 267–275.

12.

Dougherty

D. M.

Mathias

C. W.

Dawes

M. A.

Furr

R. M.

Charles

N. E.

Liguori

Shannon

E. E.

Acheson

(2013). Impulsivity, attention, memory, and decision-making among adolescent marijuana users. Psychopharmacology, 226(2), 307–319. https://doi.org/10.1007/s00213-012-2908-5

13.

ESPAD Group. (2016). ESPAD report 2015: Results from the European School Survey Project on alcohol and other drugs. Publications Office of the European Union.

14.

Fletcher

J. M.

Page

J. B.

Francis

D. J.

Copeland

Naus

M. J.

Davis

C. M.

Morris

Krauskopf

Satz

(1996). Cognitive correlates of long-term cannabis use in Costa Rican men. Archives of General Psychiatry, 53(11), 1051–1057. https://doi.org/10.1001/archpsyc.1996.01830110089011

15.

Fontes

M. A.

Bolla

K. I.

Cunha

P. J.

Almeida

P. P.

Jungerman

Laranjeira

R. R.

Bressan

R. A.

Lacerda

A. L.

(2011). Cannabis use before age 15 and subsequent executive functioning. The British Journal of Psychiatry: The Journal of Mental Science, 198(6), 442–447. https://doi.org/10.1192/bjp.bp.110.077479

16.

Gruber

S. A.

Sagar

K. A.

Dahlgren

M. K.

Racine

Lukas

S. E.

(2012). Age of onset of marijuana use and executive function. Psychology of Addictive Behaviors: Journal of the Society of Psychologists in Addictive Behaviors, 26(3), 496–506. https://doi.org/10.1037/a0026269

17.

Heaton

R. K.

, & PAR Staff. (2005). WCST:CV4: WCST: Computer Version 4—Research Edition. Psychological Assessment Resources.

18.

Heaton

R. K.

Chelune

G. J.

Talley

J. L.

Kay

G. G.

Curtiss

(1993). Wisconsin card sorting test manual—Revised and expanded. Psychological Assessment Resources, Inc.

19.

Hermann

Sartorius

Welzel

Walter

Skopp

Ende

Mann

(2007). Dorsolateral prefrontal cortex N-acetylaspartate/total creatine (NAA/tCr) loss in male recreational cannabis users. Biological Psychiatry, 61(11), 1281–1289. https://doi.org/10.1016/j.biopsych.2006.08.027

20.

Hofstede Insights. (2020). Country comparison . https://www.hofstede-insights.com/country-comparison/the-netherlands,the-usa/

21.

Hofstede

G. J.

Minkov

(2010). Cultures and Organizations: Software of the Mind: Intercultural Cooperation and Its Importance for Survival: Third Edition. McGraw Hill.

22.

Hooper

S. R.

Woolley

De Bellis

M. D.

(2014). Intellectual, neurocognitive, and academic achievement in abstinent adolescents with cannabis use disorder. Psychopharmacology, 231(8), 1467–1477. https://doi.org/10.1007/s00213-014-3463-z

23.

Jager

Block

R. I.

Luijten

Ramsey

N. F.

(2010). Cannabis use and memory brain function in adolescent boys: A cross-sectional multicenter functional magnetic resonance imaging study. Journal of the American Academy of Child and Adolescent Psychiatry, 49(6), 561–572. https://doi.org/10.1016/j.jaac.2010.02.001

24.

Jager

Block

R. I.

Luijten

Ramsey

N. F.

(2013). Tentative evidence for striatal hyperactivity in adolescent cannabis-using boys: A cross-sectional multicenter fMRI study. Journal of Psychoactive Drugs, 45(2), 156–167. https://doi.org/10.1080/02791072.2013.785837

25.

Jager

Kahn

R. S.

Van Den Brink

Van Ree

J. M.

Ramsey

N. F.

(2006). Long-term effects of frequent cannabis use on working memory and attention: An fMRI study. Psychopharmacology, 185(3), 358–368. https://doi.org/10.1007/s00213-005-0298-7

26.

Jager

Van Hell

H. H.

De Win

M. M.

Kahn

R. S.

Van Den Brink

Van Ree

J. M.

Ramsey

N. F.

(2007). Effects of frequent cannabis use on hippocampal activity during an associative memory task. European Neuropsychopharmacology: The Journal of the European College of Neuropsychopharmacology, 17(4), 289–297. https://doi.org/10.1016/j.euroneuro.2006.10.003

27.

Johnston

L. D.

O'Malley

P. M.

Miech

R. A.

Bachman

J. G.

Schulenberg

J. E.

(2016). Monitoring the future national survey results on drug use, 1975–2015: Overview, key findings on adolescent drug use. Institute for Social Research, The University of Michigan.

28.

Lamers

C. T.

Bechara

Rizzo

Ramaekers

J. G.

(2006). Cognitive function and mood in MDMA/THC users, THC users and non-drug using controls. Journal of Psychopharmacology (Oxford, England), 20(2), 302–311. https://doi.org/10.1177/0269881106059495

29.

Lane

S. D.

Cherek

D. R.

Tcheremissine

O. V.

Steinberg

J. L.

Sharon

J. L.

(2007). Response perseveration and adaptation in heavy marijuana-smoking adolescents. Addictive Behaviors, 32(5), 977–990. https://doi.org/10.1016/j.addbeh.2006.07.007

30.

Lister

R. G.

(1985). The amnesic action of benzodiazepines in man. Neuroscience and Biobehavioral Reviews, 9(1), 87–94. https://doi.org/10.1016/0149-7634(85)90034-x

31.

Lubman

D. I.

Cheetham

Yucel

(2015). Cannabis and adolescent brain development. Pharmacology & Therapeutics, 148, 1–16. https://doi.org/10.1016/j.pharmthera.2014.11.009

32.

Lyons

M. J.

Bar

J. L.

Panizzon

M. S.

Toomey

Eisen

Xian

Tsuang

M. T.

(2004). Neuropsychological consequences of regular marijuana use: A twin study. Psychological Medicine, 34(7), 1239–1250. https://doi.org/10.1017/s0033291704002260

33.

MacCoun

R. J.

(2011). What can we learn from the Dutch cannabis coffeeshop system? Addiction (Abingdon, England), 106(11), 1899–1910. https://doi.org/10.1111/j.1360-0443.2011.03572.x

34.

Page

J. B.

Fletcher

True

W. R.

(1988). Psychosociocultural perspectives on chronic cannabis use: The Costa Rican follow-up. Journal of Psychoactive Drugs, 20(1), 57–65. https://doi.org/10.1080/02791072.1988.10524372

35.

Pope

H. G.

Jr. Gruber

A. J.

Hudson

J. I.

Cohane

Huestis

M. A.

Yurgelun-Todd

(2003). Early-onset cannabis use and cognitive deficits: What is the nature of the association? Drug and Alcohol Dependence, 69(3), 303–310. https://doi.org/10.1016/s0376-8716(02)00334-4

36.

Pope

H. G.

Jr. Gruber

A. J.

Hudson

J. I.

Huestis

M. A.

Yurgelun-Todd

(2001). Neuropsychological performance in long-term cannabis users. Archives of General Psychiatry, 58(10), 909–915. https://doi.org/10.1001/archpsyc.58.10.909

37.

Pope

H. G.

Jr. Gruber

A. J.

Hudson

J. I.

Huestis

M. A.

Yurgelun-Todd

(2002). Cognitive measures in long-term cannabis users. Journal of Clinical Pharmacology, 42(S1), 41S–47S. https://doi.org/10.1002/j.1552-4604.2002.tb06002.x

38.

Pope

H. G.

Jr. Yurgelun-Todd

(1996). The residual cognitive effects of heavy marijuana use in college students. JAMA, 275(7), 521–527. https://doi:10.1001/jama.1996.03530310027028

39.

Pope

H. G.

Jr. Yurgelun-Todd

(2004). Residual cognitive effects of long-term cannabis use. In Castle

Murray

(Eds.), Marijuana and madness: Psychiatry and neurobiology (pp. 198–210). Cambridge University Press.

40.

Rey

(1964). L'examen Clinique en Psychologie. Presses Universitaires de France.

41.

SAS Institute. (2004). SAS/STAT® 9.1 user's guide.

42.

Schwartz

R. H.

Gruenewald

P. J.

Klitzner

Fedio

(1989). Short-term memory impairment in cannabis-dependent adolescents. American Journal of Diseases of Children (1960), 143(10), 1214–1219. https://doi.org/10.1001/archpedi.1989.02150220110030

43.

Shaffer

Fisher

Lucas

C. P.

Dulcan

M. K.

Schwab-Stone

M. E.

(2000). NIMH diagnostic interview schedule for children version IV (NIMH DISC-IV): Description, differences from previous versions, and reliability of some common diagnoses. Journal of the American Academy of Child and Adolescent Psychiatry, 39(1), 28–38. https://doi.org/10.1097/00004583-200001000-00014

44.

Silva-Filho

J. H. d.

Pasian

S. R.

Humberto

J. S. M.

(2011). Teste Wisconsin de classificação de cartas: Uma revisão sistemática de 1952 a 2009. Psico-USF, 16(1), 107–116. https://doi.org/10.1590/S1413-82712011000100012

45.

Solowij

Grenyer

B. F. S.

Peters

Chesher

(1997). Long-term cannabis use impairs memory processes and frontal lobe function. Symposium on the Cannabinoids, Stone Mountain, Georgia.

46.

Solowij

Stephens

R. S.

Roffman

R. A.

Babor

Kadden

Miller

Christiansen

McRee

Vendetti

& Marijuana Treatment Project Research Group. (2002). Cognitive functioning of long-term heavy cannabis users seeking treatment. JAMA, 287(9), 1123–1131. https://doi.org/10.1001/jama.287.9.1123

47.

Spreen

Strauss

(1998). A compendium of neuropsychological tests: Administration, norms, and commentary: Second edition. Oxford University Press.

48.

Thielen

Verleysen

Huybrechts

Lafosse

Gillebert

C. R.

(2019). Flemish normative data for the Buschke Selective Reminding Test. Psychologica Belgica, 59(1), 58–77. https://doi.org/10.5334/pb.486

49.

Tzilos

G. K.

Cintron

C. B.

Wood

J. B.

Simpson

N. S.

Young

A. D.

Pope

H. G.

Jr. Yurgelun-Todd

D. A.

(2005). Lack of hippocampal volume change in long-term heavy cannabis users. The American Journal on Addictions, 14(1), 64–72. https://doi.org/10.1080/10550490590899862

50.

van den Berg

de Weerd

Reuvekamp

van der Meere

(2020). Cognitive control deficits in pediatric frontal lobe epilepsy. Epilepsy & Behavior, 102, 106645. https://doi.org/10.1016/j.yebeh.2019.106645

51.

Wechsler

(2003). WISC-IV Wechsler Intelligence Scale for Children: Fourth edition: Administration and Scoring: Manual. Harcourt Assessment.