1. Introduction

The decriminalization of cannabis has gained momentum (UNODC, 2022), accompanied by a shift towards more positive cannabis attitudes and reduced perceptions of harm (Von Sydow et al., 2002; Wu et al., 2015). These factors interconnect with cannabis use initiation, frequency of use, and the development of cannabis use disorder (CUD; Askari et al., 2021; Philbin et al., 2019). Understanding the relationship between cannabis policy, attitudes, and addiction biomarkers is crucial. This cross-cultural collaborative study between research sites in Texas (TX), USA and Amsterdam, The Netherlands (NL) investigated how cannabis culture relates to cannabis cue-reactivity in individuals with CUD.

Recreational cannabis use is illegal in TX, whereas NL decriminalized use in 1976. In the US, the prevalence of past-year use has risen, particularly among adults, following the implementation of more permissive state-level cannabis policies (Bailey et al., 2020; Philbin et al., 2019). Attitudes about cannabis effects may contribute to rising use (Fleming et al., 2016; Holm et al., 2014; Martínez-Vispo and Dias, 2022; Turna et al., 2022). In Texas, past year adult cannabis use is lower (15.46%, 95%CI: 14.16–16.87) and the perception of ‘great risk’ of monthly use is higher (25.31%, 95%CI: 23.51–27.20) than the US average (past year adult use: 21.43%, 95%CI: 20.90–21.96; perceived great risk: 21.04%, 95%CI: 20.44–21.64), but differences are small and perceived risk in more regular cannabis users is under investigated (SAMHSA, 2024). Notably, in Washington state, the association between lower perception of harm and more cannabis use strengthened after recreational legalization (Fleming et al., 2016). Community attitudes can also influence the development of CUD: disapproval is protective, while approval acts as a risk factor (Berridge and Robinson., 2016; Wu et al., 2015). However, the impact of legal policies, personal attitudes, and perceived community attitudes on cannabis cue-reactivity—an addiction biomarker—in regular cannabis users remains unclear.

Cannabis cue-reactivity is the heightened neurophysiological response to cannabis cues resulting from repeated pairings of cues (e.g. joints, aroma) and the rewarding effects of use through incentive sensitization (Berridge and Robinson., 2016). Alterations in the salience, reward, motivation, and cognitive control networks during exposure to substance-related cues is associated with escalating use and the transition to disordered use (Jasinska et al., 2014). Neural cue-reactivity has been utilized as a biomarker of treatment outcomes and relapse risk due to consistent evidence across substances (Courtney et al., 2016). In cannabis users, cue-reactivity is consistently observed in the striatum, the anterior cingulate cortex (ACC), middle frontal gyrus (MFG), and precuneus, with more severe users showing greater activity in striatal regions and associations with self-reported craving and cannabis-related problems (Sehl et al., 2021).

Cultural neuroscience research has revealed that cultural factors impact neural processes related to attention, perception, self-awareness, and emotion (Prashad et al., 2017). However, the connection between cultural variations in these processes and neural mechanisms underlying addiction is unclear. Negative outcome expectancies can modulate alcohol cue-reactivity (Eddie et al., 2013), supporting the hypothesis that substance-related beliefs, influenced by legal and cultural contexts, can alter cue-reactivity in the brain.

The objectives of the current study were: 1) compare positive and negative cannabis attitudes (personal and perceived attitudes in their social environment) between individuals with CUD and matched controls across two sites (TX and NL) with different legal policies (see cannabis statement in Box 1; Cousijn et al., 2024), 2) examine differences in neural cannabis cue-reactivity, 3) investigate whether cannabis attitudes are associated with observed site differences in neural cannabis cue-reactivity. We hypothesized that individuals with CUD in NL would exhibit more positive cannabis-related attitudes and perceive a more positive social and cultural environment compared to TX, aligning with the more permissive legal environment. Additionally, we hypothesized that more positive attitudes would be associated with higher cue-reactivity, with individuals with CUD in NL displaying higher cannabis cue-reactivity and stronger positive associations with cannabis use measures - including CUD symptom count and cannabis craving - compared to individuals in TX.

2. Methods

2.1. Participants and procedures

Through flyers (i.e., on campus, throughout the city, and in cannabis shops (NL only)) and social media (i.e., Facebook, Instagram), a total of 221 individuals were enrolled in the study. All participants had to be 18–30 years old at the time of the screening, right-handed, have no MRI contraindications, history of regular (i.e. monthly) illicit drug use, excessive alcohol use, known neurological disorders, brain damage, or other major health conditions (e.g. cancer). Participants in the cannabis group had to report at least two CUD symptoms (DSM-5) during the online and phone screening and use cannabis 6–7 days per week for the past year with no active plan to stop use or seek treatment. Participants in the non-using group had to have used cannabis less than 50 times, with no more than five uses in the past year, and no use in the past three months. A rapid urine test was conducted to exclude for illicit drug use during the lab visit (NL-control: N = 2; NL-CUD: N = 4, TX-CUD: N = 3). Additional participants were excluded for excessive motion during scanning (>4.5 mm; TX-control: N = 6, NL-CUD: N = 10, TX-CUD: N = 3), structural abnormalities (NL-control: N = 1), too much cannabis use (NL-control: N = 1), experimenter error (NL-control: N = 2), and participant distraction (NL-CUD: N = 2) resulting in 34 exclusions. The final sample (N = 187) consisted of 54 NL cannabis users (NL-CUD), 50 TX cannabis users (TX-CUD), 50 NL non-using controls (NL-CON), and 33 TX non-using controls (TX-CON).

The protocols were approved by the University of Texas Dallas Institutional Review Board (19–107) and the ethics committee of the Faculty of Social and Behavioral Sciences at the University of Amsterdam (2018-DP-9616). Each participant gave informed consent and received monetary compensation (TX: $20/hour, NL: €12.50/hour; in line with institutional compensation requirements).

2.2. Tactile cannabis cue exposure paradigm

The previously validated paradigm (Filbey et al., 2016) consisted of two runs with 18 tactile-visual cues pseudorandomly presented using E-Prime 3.0. The cues consisted of a cannabis joint (NL; rolled with a placebo herbal mix) or cannabis pipe (TX) to match local preference, a pen (neutral control), and the participant's favorite fruit (appetitive control) from a selection (orange, apple, banana, kiwi, grapes). The cues were consistent across trials. Cues were placed in the left hand of the participant by the experimenter when a photo of the corresponding cue held by the participant (photos taken before the test session) was presented on the screen. After each trial, participants rated their current urge to use cannabis on a 1–10 VAS scale (see supplement for details).

2.3. Questionnaires

2.3.1. Cannabis-related measures

To evaluate the presence of DSM-5 CUD symptoms, the Mini International Neuropsychiatric Interview (MINI; Sheehan et al., 1998) was conducted (semi-structured interview, conducted for inclusion of DSM-5 CUD symptom scores in the analysis). History and current patterns of cannabis use were measured with a substance use history questionnaire. Positive (8-items) and negative (9-items) attitudes about the effects of cannabis were assessed with the Cannabis Culture Questionnaire (CCQ; Holm et al., 2016). Participants completed the scale from three perspectives: personal attitudes, perceived family and friends’ attitudes, and perceived larger society (Texas/Netherlands) attitudes.

2.3.2. Substance use, mental health, demographics, and other measures

A demographic questionnaire was administered to assess gender, age, and education years. IQ was estimated with the vocabulary and matrix reasoning tasks from the Wechsler Adult Intelligence Scale (WAIS-IV; Wechsler, 2012; Coalson et al., 2010). Participants self-reported alcohol use and related problems (AUDIT, Saunders et al., 1993), cigarette use, lifetime use of illicit substances, symptoms of anxiety (STAI; Spiegelberger, 2010), depression (BDI-II; Beck et al., 1996), and attention deficit and hyperactivity disorder (ADHD – ASRS; Ustun et al., 2017). General mental health symptoms were assessed with the DSM-5 self-rated level 1 cross-cutting symptom checklist (DSM5-CCSM; American Psychiatric Association, 2013).

2.4. Neuroimaging data collection and preprocessing

Anatomical and structural scans were collected at the University of Amsterdam (3T Philips Achieva MRI with 32-channel SENSE head coil) and at the University of Texas Dallas (3T Siemens MAGNETOM Prisma MRI with 64-channel head coil). Anatomical scans were collected for registration purposes (TR/TE = 8.3/3.9 ms, FOV = 188 × 240 × 220 mm3, 1x1x1 mm3, flip angle = 8°). During the cannabis cue exposure paradigm, two T2∗ single-shot multiband accelerated EPI sequences were collected (multiband factor = 4, TR/TE = 550/30 ms, FOV = 240x240x118.5, voxel size = 3x3x3 mm3, interslice gap = 0.3 mm, flip angle = 55°). Skull-stripping, spatial smoothing, motion correction, and registration of functional to anatomic images were conducted by the fMRIprep preprocessing pipeline using Enigma Halfpipe software (Esteban et al., 2019; Waller et al., 2022; see supplementary materials).

2.5. Data analysis

2.5.1. Behavioral data

To isolate changes in craving following exposure to cannabis cues during the tactile cannabis cue exposure paradigm specifically, craving scores were calculated by subtracting the average craving following the neutral and fruit trials from the cannabis trials.

Site and group differences in sample characteristics (demographics, mental health, and substance use measures) were assessed with ANOVA, Mann-Whitney U-tests, or chi square tests as applicable. A linear mixed effect (LME; lme4 package (Bates et al., 2015) in R version 3.6.3 (R Core Team, 2022)) model analysis with random intercepts and slope for participant was conducted to examine the fixed effects of cue type, group, and site on craving. To assess group and site differences in cannabis attitudes, LME models with crossed random effects for subject and perspective were calculated for the positive and negative CCQ scores separately.

2.5.2. fMRI data

First-level subject models were computed with FSL FEAT (FMRIB's Software Library, version 6.0). A general linear model was estimated with predictors for each cue condition convolved with a double-gamma hemodynamic response function for each run. Second-level analyses were computed to pool the activity across both runs. The contrasts of interests were Cannabis > Neutral and Cannabis Cue > Neutral + Fruit cues to identify cannabis-specific activation controlling for neutral trials and additionally controlling for food-related appetitive activity.

Whole Brain Analyses. FSL's FEAT FLAME 1 mixed effect models were conducted for each contrast to identify clusters with group differences (two sample unpaired t-test), site by group interactions (two-way between-subject ANOVA), and associations with cannabis use measures (symptom count, grams per week, cue-induced craving) in the CUD group (single group with additional covariate). The associations with cannabis use measures were also compared between sites (two groups with continuous covariate interaction).

For all analyses, site was added as a covariate to adjust for the scanner difference. Hence, as site differences were controlled for using this approach, we only interpreted interactions including site rather than main effects of site. Multiple comparison correction was applied at a Z-threshold of 3.1 and a cluster p-significance threshold of 0.05. The mean peak activity of significant clusters was extracted to visualize the effects. For clusters with significant site interactions, follow-up multiple regressions assessing the association with cannabis attitudes were performed. CCQ scores were used as predictors of mean peak activity with and without site as a moderator to examine whether site differences were driven in part by cannabis attitudes (Bonferroni-corrected p-value threshold of 0.008, correcting the 0.05 threshold for testing each outcome for the six different CCQ scores).

3. Results

3.1. Sample characteristics

Within the CUD group, individuals in TX reported more CUD symptoms, more days of cannabis use per week, longer duration of weekly use, more self-reported grams per week of use, and reported using cannabis in the previous 24 h more than individuals in NL (Table 1). The CUD and control groups were well-matched on age and gender. However, the CUD group reported significantly fewer years of education, more lifetime illicit substance use episodes, higher depression, and anxiety-related symptoms, were more often daily cigarette smokers, and scored lower on IQ tests. Significant group by site interactions were observed in AUDIT scores and DSM-5 cross-cutting mental health symptoms. The NL-CUD group and TX controls had lower AUDIT scores than NL controls, while the TX-CUD group scored higher than TX controls. The NL-CUD group reported more cross-cutting mental health symptoms than the TX-CUD group, and within both sites the CUD group reported more symptoms than the controls.

Table 1. Sample characteristics.

3.2. Cannabis attitudes

For both positive and negative attitudes, significant group by site (pos: β = −3.18, t = −2.65, p = .008; neg: β = −2.95, t = 2.14, p = .03) and group by perspective (pos(personal): β = −2.65, t = −2.73, p = .007; neg(personal): β = 2.96, t = 2.24, p = .02; neg(state/country): 2.11, t = 2.14, p = .03) interactions emerged. For positive attitudes only, a significant site by perspective interaction also emerged (β = −2.26, t = −2.33, p = .02; Table S1; Fig. 1).

Fig. 1.

_{Overview of cannabis culture questionnaire (CCQ) scores per group and outcome type. Error bars reflecting standard error of the mean. Pos: positive, Neg: negative; P: personal, FF: family/friends, SC: state/country; NL: Netherlands, TX: Texas – USA; CAN: cannabis group, CON: control group.}

Follow-up pairwise comparisons revealed that the CUD group reported more positive personal attitudes (U = 7537.00, p < .001, d = 0.75), and more positive perceptions of both family/friends (t = 5.83, p < .001, d = 0.86) and state/country cannabis attitudes (t = 2.42, p = .017, d = 0.36) compared to controls. The CUD group also reported less negative personal cannabis attitudes (t = −6.57, p < .001, d = −0.97) and perceived their family/friends to be less negative (−3.21, p = .002, d = −0.47). No group differences were observed in perceptions of negative country/state cannabis attitudes.

Regardless of perspective, the NL-CUD group reported significantly more positive cannabis attitudes than the NL controls (t = 4.78, p < .001, d = 0.94) but no differences emerged in negative attitudes (t = −1.15, p = −0.25). The TX-CUD group reported significantly more positive (t = 8.26, p < .001, d = 1.85) and less negative attitudes (t = −5.1, p < .001, d = 1.14) than the TX controls. Furthermore, the TX-CUD group reported significantly more positive (t = 4.62, p < .001, d = 0.91) and less negative attitudes (t = 4.9, p < .001, d = 0.96) than the NL-CUD group. In comparison, the TX controls did not report significantly different positive or negative attitudes compared to the NL controls (pos: t = 0.47, p = .64; neg: t = 0.10, p = .92).

Regardless of group, TX participants reported more positive personal and perceived family/friend attitudes (personal: t = −4.5, p < .001, d = −0.66; family/friends: t = −2.96, p = .004, d = −0.43), but no significant site differences were observed in perceptions of the state/country attitudes (t = 0.157, p = .88).

3.3. Cue-elicited craving ratings

Significant group by condition and group by site interactions emerged in self-reported craving ratings. Post-hoc Bonferroni-corrected pairwise comparisons revealed that the CUD group reported higher craving following the cannabis cues (M = 3.99, SD = 2.97) compared to both neutral (M = 2.93, SD = 2.55; t = 11.21, p_bonf < 0.001) and fruit (M = 3.03, SD = 2.60; t = 10.12, p_bonf < 0.001; Fig. 2) cues. In contrast, no significant differences in craving ratings between conditions were observed in the control group (lowest p-value = 0.147).

Fig. 2.

_{Overview of cue induced craving per group and cue type. Error bars reflecting standard error of the mean. NL: Netherlands, TX: Texas – USA; CAN: cannabis group, CON: control group.}

Following all trial types, the NL-CUD group reported higher craving than the TX-CUD group (t = 3.41, p_bonf = 0.005). Conversely, the TX controls reported higher craving (M = 1.94) than the NL controls (M = 0.34; t = −3.10, p_bonf = 0.014). Furthermore, craving was higher after all trial types in the NL-CUD group (M = 4.03) compared to the NL controls (M = 0.40; t = 8.31, p_bonf < 0.001) and the TX controls (M = 1.94, t = 4.24, p_bonf < 0.001). However, the TX-CUD group did not report significantly higher craving (M = 2.54) than the TX controls (M = 1.94; t = 1.20, p_bonf = 1.00).

3.4. Whole brain analyses

3.4.1. Can > neutral

The CUD group demonstrated heightened activity compared to the control group in four clusters spanning regions including the inferior and superior lateral occipital cortex, occipital pole, and occipital fusiform gyrus (Table 2; Fig. 3). No clusters were associated with quantity of use, severity of symptoms, or in-scanner cannabis craving in the CUD group overall or differentially between sites.

Table 2. Significant clusters from whole-brain exploratory analyses.

Fig. 3.

_{Significant clusters of activation in which the CUD group shows heightened activity compared to the control group in whole brain exploratory analyses. When controlling only for neutral cues, heightened cannabis cue reactivity is observed in four clusters in the occipital cortex (Can > Neu). Additional regions including the nucleus accumbens and regions of the medial frontal cortex emerge when also controlling for differences in activity between groups in responsivity to fruit cues (Can > Fruit + Neu).}

3.4.2. Can > neutral + appetitive

The CUD group demonstrated heightened activity compared to the control group in eight clusters encompassing parts of the precuneus, frontal medial cortex and frontal pole, thalamus, nucleus accumbens, inferior and superior lateral occipital cortex, occipital pole, fusiform gyrus, and cerebellum (Table 2; Fig. 3). While activity was not related to quantity of use in the CUD group, differential associations between Can > Neutral + Appetitive activity with CUD symptom severity and in-scanner craving ratings emerged (Table 2). Higher CUD symptom count was related to higher cannabis cue-reactivity in the TX group and fewer CUD symptoms was related to higher cannabis cue-reactivity in the NL group in the precuneus, superior and inferior lateral occipital cortex, calcarine cortex, lingual gyrus, and occipital pole (Fig. 4a). Additionally, in the TX-CUD group, higher cannabis craving after cannabis cues (controlling for craving after fruit and neutral cues) was related to higher cannabis cue-reactivity in the superior and inferior lateral occipital cortex, while the reverse was observed in the NL-CUD group (Fig. 4b).

Fig. 4.

_{A) Differential association between mean peak cannabis cue reactivity (Can > Neu + Fruit; whole-brain analysis) and CUD symptom count. B) Differential association between mean peak cannabis cue reactivity (Can > Neu + Fruit; whole-brain analysis) and in scanner cannabis craving following cannabis trials (controlled for craving following neutral and fruit trials). C) Significant interaction between positive personal cannabis attitudes and cannabis craving in mean peak voxel of site by craving whole brain analysis.}

Follow-up regression analyses examined whether cannabis attitudes interacted with craving or symptom count in the mean peak activity of clusters with significant site differences (Table 3). A significant interaction was observed between personal positive attitudes and craving. Individuals with CUD who report fewer positive attitudes about cannabis demonstrated lower cannabis cue-reactivity with increasing craving, while individuals who report more positive attitudes showed higher cannabis cue-reactivity as craving increased (Fig. 4c). This effect remained significant when controlling for gender and site differences in cannabis use characteristics and mental well-being, and when excluding potential outliers (mean peak values ±3SD of the mean). No other effects of cannabis attitudes reached significance.

Table 3.

4. Discussion

Applying a novel cross-cultural neuroscience approach, this study found distinct neural responses to cue-elicited craving, a biomarker of addiction, in individuals with CUD within contexts where cannabis use legality is prohibitive (TX) versus permissive (NL). Despite the stricter legal environment, the TX-CUD group reported more positive and less negative cannabis attitudes than the NL-CUD group. Notably, site differences emerged in the associations between self-reported craving, CUD severity, and cue-reactivity (Cannabis > Neutral + Appetitive), with evidence for the role of positive cannabis attitudes in site differences.

Contrary to our expectations, individuals with CUD in TX held more positive and less negative attitudes about cannabis, both personally and perceptions of their social environment's beliefs, compared to the NL group. It is possible that individuals who continue using cannabis in a prohibitive environment may have more positive perceptions of its effects, as those who hold more negative views may choose to abstain due the legal consequences. Additionally, while individuals with CUD in TX reside in a prohibitive state, their more positive and less negative attitudes may reflect the cultural shift towards more permissive policies in other US states and the rising prevalence of low-risk perceptions of cannabis in the US as a whole (Levy et al., 2021). Hence, it remains crucial to monitor changes in risk perceptions at both the state and country level – especially following changes in legislation - to inform future policy aimed at minimizing harm of cannabis use.

The TX-CUD group exhibited positive associations between cue-elicited craving, CUD severity, and neural cue-reactivity in clusters of the lateral occipital cortex, precuneus, supracalcerine cortex, intracalcerine cortex, and lingual gyrus. In contrast, as self-reported craving and CUD severity increased in the NL-CUD group, neural cue-reactivity in these regions decreased. The precuneus is often activated during cue reactivity and is involved in complex cognitive functions including episodic memory retrieval and self-referential processing (Cavanna and Trimble, 2006). The lingual gyrus, calcarine cortex, and lateral occipital cortex are involved in visual processing, which can be heightened by greater attention and emotion (Vuilleumier and Driver, 2007). These differential associations between neural cue-reactivity, CUD severity, and self-reported craving suggest that this addiction biomarker may function differently across cultures. Greater cue-reactivity is thought to reflect greater severity of cue-elicited craving, contributing to greater risk of relapse in real life when cues are encountered (Courtney et al., 2016). Therefore, blunted cue-elicited craving after treatment has been used as a proxy of an individual's reduced relapse liability. The negative association observed between self-reported craving and cue-reactivity suggests that that heavy users in NL may have better control over their use compared to their TX counterparts. Longitudinal data is needed to test this speculative hypothesis. These findings highlight the issue of generalizability: we cannot assume that mechanisms function the same across every cultural context.

Cultural influences likely shape personal cannabis attitudes, which in turn affect the relationship between cue-reactivity and craving in lateral occipital clusters, in which site-related differences in cue reactivity were observed. Individuals with more positive personal cannabis attitudes exhibited greater cue-reactivity as self-reported craving increased, whereas those with less positive attitudes showed decreased cue-reactivity as self-reported craving increased. These findings emphasize the connection between explicit motivational processes (e.g. cannabis attitudes and craving) and cue-reactivity in the brain, highlighting the need to raise awareness of the potential importance of perceived harms and benefits of cannabis use in clinical practice. Future research examining whether positive cannabis attitudes modulate the predictive value of cue-induced craving and neural cue reactivity in cannabis use desistance could shed light on the clinical implications of these findings. It is important to note that cannabis attitudes did not modulate cannabis cue-reactivity in regions with significant differential associations with CUD severity across sites. It remains unclear what site-related differences may underlie this effect.

Heightened cannabis cue-reactivity was observed in the CUD group compared to controls in both contrasts. When controlling for neutral cues, heightened cue-reactivity emerged in four clusters in the occipital cortex. When further controlling for appetitive cues, additional clusters emerged spanning the precuneus, nucleus accumbens, thalamus, and medial frontal cortex. Together, these findings indicate a relative deactivation to a natural reward cue (i.e. fruit) in individuals with CUD in regions of the salience network implicated in motivational processes of addiction (Seeley, 2019; Koob and Volkow, 2016). This aligns with the incentive salience theory of addiction in which sensitization to drug-related cues occurs in parallel with desensitization to natural rewards (Koob and Volkow, 2010). However, the deactivation to fruit cues could be specific to the neurobiological effects of cannabis on food behavior (i.e. the ‘munchies’) mediated by CB1 receptor activity (Koch et al., 2015), rather than a general addiction process.

The current study has several limitations. First, TX CUD group reported greater cannabis use frequency, quantity, duration, past 24-h use, and CUD symptoms. Although statistically significant, both groups reported using cannabis six to seven days per week and fell within the moderate severity classification for symptoms, suggesting the samples reflected similar populations across sites. Notably, site effects remained significant in sensitivity analyses controlling for years of weekly use, abstinence in the last 24 h, typical grams of use per week, MINI CUD score, DSM-5 cross-cutting symptoms, and gender, and exclusion of potential outliers did not change the results. Additionally, grams of cannabis used per week does not account for actual cannabinoid exposure, which is likely to differ between these regions based on available cannabis product potency (Chandra et al., 2019). Future cross-cultural research should incorporate biospecimen analyses as recommended in the iCannToolkit (Lorenzetti et al., 2021) to examine the role of cannabinoid exposure and research focusing on differences in cannabis use characteristics across sites should be encouraged. Furthermore, our study only included cannabis users within a limited age range (18–30), limiting the generalizability of the findings to younger and older cannabis users. Also, it is crucial that future studies assessing the role of cultural attitudes in cannabis use include a balanced - or at least representative – gender sample with sufficient power to assess gender differences. Finally, we were unable to control for differences across sites in co-use of tobacco products, which is highly prevalent in cannabis users in Europe and much less so in the US (Hindocha et al., 2016). While our sample was ecologically valid, over-recruiting uncommon subgroups (e.g. non-tobacco-using cannabis users in NL and tobacco-using cannabis users in TX) can shed light on interactive neurobiological effects and account for cross-cultural differences.

In conclusion, this study provides preliminary evidence of diverging relationships between neural cannabis cue-reactivity and cue-induced craving in individuals with CUD across distinct cannabis environments. Positive attitudes about cannabis’ effects appears to drive these differences. These findings highlight the potential of applying a cross-cultural framework to the neuroscience of addiction processes to better understand the factors that may contribute to the maintenance of CUD.

Introduction

The acceptance of cannabis use has increased, along with a shift toward more positive views and fewer perceived risks. These changes are connected to when people start using cannabis, how often they use it, and whether they develop a cannabis use disorder (CUD), which is a problematic pattern of cannabis use. It is important to understand how cannabis policies, public attitudes, and signs of addiction are related. This study was a collaborative effort between research sites in Texas, USA, and Amsterdam, The Netherlands. It explored how cannabis culture is linked to brain responses to cannabis cues in individuals with CUD.

In Texas, using cannabis for recreation is against the law, while in the Netherlands, it has been decriminalized since 1976. In the United States, more adults have used cannabis in the past year, especially after states adopted more relaxed cannabis laws. People's beliefs about the effects of cannabis might contribute to this increase in use. In Texas, past-year adult cannabis use is lower, and the belief that monthly use carries a "great risk" is higher compared to the overall U.S. average. However, these differences are small, and how regular cannabis users perceive risk is not fully understood. It is notable that in Washington state, after recreational cannabis became legal, the link between lower perceived harm and increased cannabis use became stronger. Community attitudes can also affect the development of CUD: disapproval tends to offer protection, while approval can be a risk factor. However, the impact of laws, personal beliefs, and perceived community attitudes on cannabis cue-reactivity—a biological sign of addiction—in regular cannabis users remains unclear.

Cannabis cue-reactivity is a strong brain response to things linked to cannabis (like joints or its smell). This happens because these cues have been repeatedly paired with the rewarding effects of using cannabis. Changes in brain networks related to importance, reward, motivation, and self-control when exposed to drug-related cues are linked to increased use and the development of a disorder. Brain responses to cues have been used as a way to predict treatment outcomes and relapse risk, as evidence for this is consistent across different substances. In cannabis users, cue-reactivity is often seen in specific brain areas, including the striatum, anterior cingulate cortex, middle frontal gyrus, and precuneus. Users with more severe conditions show greater activity in striatal regions, and this activity is connected to reported cravings and cannabis-related problems.

Research in cultural neuroscience has shown that cultural factors influence how the brain handles attention, perception, self-awareness, and emotions. However, the connection between these cultural differences in brain processes and the brain mechanisms behind addiction is not well understood. Negative expectations about outcomes can change how the brain reacts to alcohol cues, which supports the idea that beliefs about substances, shaped by legal and cultural settings, can alter brain responses to cues.

The goals of this study were: 1) to compare positive and negative attitudes toward cannabis (personal beliefs and perceived attitudes of others) among individuals with CUD and matched control groups across two sites (Texas and the Netherlands) with different legal policies; 2) to examine differences in brain responses to cannabis cues; and 3) to investigate whether cannabis attitudes are linked to observed differences in brain responses between the two sites. It was predicted that individuals with CUD in the Netherlands would have more positive cannabis-related attitudes and would perceive a more positive social and cultural environment compared to those in Texas, which would align with the more permissive legal setting. Additionally, it was hypothesized that more positive attitudes would be linked to higher cue-reactivity. It was expected that individuals with CUD in the Netherlands would show higher cannabis cue-reactivity and stronger positive links with cannabis use measures—including CUD symptom count and cannabis craving—compared to individuals in Texas.

Methods

Participants and procedures

A total of 221 individuals were recruited for the study through flyers and social media. All participants needed to be between 18 and 30 years old at the time of screening, right-handed, and without any conditions that would prevent an MRI scan. They also could not have a history of regular illicit drug use, excessive alcohol use, known neurological disorders, brain damage, or other major health problems. Participants in the cannabis group had to report at least two CUD symptoms and use cannabis 6–7 days per week for the past year, with no active plans to quit or seek treatment. Participants in the non-using group had to have used cannabis fewer than 50 times in their lives, no more than five times in the past year, and no use in the past three months. A urine test was conducted during the lab visit to check for illicit drug use. Additional participants were excluded for excessive motion during scanning, structural abnormalities, too much cannabis use, experimenter error, or participant distraction. The final study included 187 participants: 54 cannabis users from the Netherlands (NL-CUD), 50 cannabis users from Texas (TX-CUD), 50 non-using control participants from the Netherlands (NL-CON), and 33 non-using control participants from Texas (TX-CON).

The study procedures were approved by the ethics boards at the University of Texas Dallas and the University of Amsterdam. All participants provided informed consent and received payment for their time.

Tactile cannabis cue exposure paradigm

A previously tested method was used, involving two sessions with 18 different tactile-visual cues presented in a random order. The cues included a cannabis joint (in the Netherlands) or a cannabis pipe (in Texas) to match local preferences, a pen (as a neutral control), and the participant's favorite fruit (as an appetitive control). The cues were placed in the participant's left hand while a photo of the cue was shown on a screen. After each trial, participants rated their current urge to use cannabis on a scale from 1 to 10.

Questionnaires

The presence of CUD symptoms was assessed using a semi-structured interview. Information on past and current cannabis use patterns was gathered using a substance use history questionnaire. Positive and negative attitudes about the effects of cannabis were measured with the Cannabis Culture Questionnaire. Participants completed this scale from three different viewpoints: their personal attitudes, what they believed their family and friends thought, and what they believed society at large (in Texas or the Netherlands) thought.

In addition, a demographic questionnaire collected information on gender, age, and years of education. An estimate of IQ was obtained using specific tasks from a standard intelligence scale. Participants also reported their alcohol use and related problems, cigarette use, lifetime use of illicit substances, symptoms of anxiety, depression, and attention deficit and hyperactivity disorder. General mental health symptoms were assessed using a self-rated checklist.

Neuroimaging data collection and preprocessing

Brain scans were collected at the University of Amsterdam and the University of Texas Dallas. Anatomical and structural scans were collected for registration purposes. During the cannabis cue exposure study, two functional brain imaging sequences were collected. Standard methods were used to prepare the brain imaging data for analysis, which included removing skull tissue, smoothing the images, correcting for head movement, and aligning the functional images with the anatomical scans.

Data analysis

To specifically identify changes in craving after exposure to cannabis cues, craving scores were calculated by subtracting the average craving ratings from the neutral and fruit trials from the cannabis trials.

Differences between sites and groups in participant characteristics (such as demographics, mental health, and substance use measures) were analyzed using appropriate statistical tests. A statistical model was used to examine how cue type, participant group, and study site affected craving. To assess group and site differences in cannabis attitudes, separate statistical models were calculated for positive and negative cannabis attitude scores, taking into account individual participants and different viewpoints.

For the brain imaging data, initial individual brain models were computed. A general statistical model was used to estimate brain activity for each cue condition. A second level of analysis combined activity across both sessions. The main comparisons of interest were cannabis cues versus neutral cues, and cannabis cues versus both neutral and fruit cues. This helped identify brain activation specific to cannabis, while also controlling for general reactions to non-drug cues and food-related cues.

Whole-brain analyses were conducted to find brain regions with differences between groups, interactions between site and group, and connections with cannabis use measures (like symptom count, grams per week, and cue-induced craving) within the CUD group. These associations with cannabis use measures were also compared between sites.

For all analyses, the study site was included as a factor to account for differences between the MRI scanners. This means that only interactions involving site, rather than main effects of site, were interpreted. Statistical correction was applied to account for multiple comparisons, ensuring the reliability of the findings. The average brain activity in significant clusters was extracted to help visualize the effects. For brain regions with significant site interactions, further statistical analyses were performed to assess how cannabis attitudes were linked to these effects. Cannabis attitude scores were used to predict average brain activity, both with and without site as a modifying factor, to see if cannabis attitudes partly explained the site differences.

Results

Sample characteristics

Within the group of individuals with CUD, those in Texas reported more CUD symptoms, more days of cannabis use per week, a longer history of weekly use, more self-reported grams of use per week, and reported using cannabis in the previous 24 hours more often than those in the Netherlands. The CUD and control groups were well-matched for age and gender. However, the CUD group reported fewer years of education, more lifetime illicit substance use episodes, higher depression and anxiety symptoms, were more often daily cigarette smokers, and scored lower on IQ tests. Significant interactions between group and site were observed in alcohol use scores and general mental health symptoms. The CUD group in the Netherlands and the control group in Texas had lower alcohol use scores than the control group in the Netherlands, while the CUD group in Texas scored higher than the control group in Texas. The CUD group in the Netherlands reported more general mental health symptoms than the CUD group in Texas, and within both sites, the CUD group reported more symptoms than the control groups.

Cannabis attitudes

For both positive and negative attitudes, significant interactions between group and site, and between group and perspective emerged. For positive attitudes only, a significant interaction between site and perspective was also observed.

Further comparisons showed that the CUD group reported more positive personal attitudes and more positive perceptions of cannabis attitudes among their family/friends and in society compared to control groups. The CUD group also reported less negative personal cannabis attitudes and perceived their family/friends to be less negative. No group differences were found in perceptions of negative societal cannabis attitudes.

Regardless of their viewpoint, the CUD group in the Netherlands reported significantly more positive cannabis attitudes than the control group in the Netherlands, but no differences were found in negative attitudes. The CUD group in Texas reported significantly more positive and less negative attitudes than the control group in Texas. Furthermore, the CUD group in Texas reported significantly more positive and less negative attitudes than the CUD group in the Netherlands. In comparison, the control group in Texas did not report significantly different positive or negative attitudes compared to the control group in the Netherlands.

Regardless of the group, participants in Texas reported more positive personal and perceived family/friend attitudes, but no significant site differences were observed in perceptions of general societal attitudes.

Cue-elicited craving ratings

Significant interactions between group and cue type, and between group and site were found in self-reported craving ratings. Specific comparisons showed that the CUD group reported higher craving after cannabis cues compared to both neutral and fruit cues. In contrast, no significant differences in craving ratings between cue types were observed in the control group.

After all cue types, the CUD group in the Netherlands reported higher craving than the CUD group in Texas. Conversely, the control group in Texas reported higher craving than the control group in the Netherlands. Moreover, craving was higher after all cue types in the CUD group in the Netherlands compared to both the control group in the Netherlands and the control group in Texas. However, the CUD group in Texas did not report significantly higher craving than the control group in Texas.

Whole brain analyses

The CUD group showed increased brain activity compared to the control group in areas of the occipital cortex when only controlling for neutral cues. No specific brain regions were linked to the amount of cannabis used, the severity of symptoms, or craving experienced during the scan in the CUD group overall, nor did these links differ between sites.

When also controlling for differences in brain activity related to fruit cues, the CUD group showed increased activity compared to the control group in more brain regions, including parts of the precuneus, frontal medial cortex, frontal pole, thalamus, nucleus accumbens, inferior and superior lateral occipital cortex, occipital pole, fusiform gyrus, and cerebellum. While brain activity was not related to the amount of cannabis used in the CUD group, different associations emerged between this activity, CUD symptom severity, and craving experienced during the scan. Higher CUD symptom count was related to higher cannabis cue-reactivity in the Texas group, while fewer CUD symptoms were related to higher cannabis cue-reactivity in the Netherlands group in specific brain regions. Additionally, in the Texas CUD group, higher cannabis craving after cannabis cues was related to higher cannabis cue-reactivity in the superior and inferior lateral occipital cortex, while the opposite was observed in the Netherlands CUD group.

Further analyses investigated whether cannabis attitudes influenced craving or symptom count in brain areas with significant site differences. A significant interaction was observed between personal positive attitudes and craving. Individuals with CUD who reported fewer positive attitudes about cannabis showed lower cannabis cue-reactivity as craving increased, while those with more positive attitudes showed higher cannabis cue-reactivity as craving increased. This finding remained significant even after accounting for gender and site differences in cannabis use characteristics and mental well-being, and when removing potential extreme data points. No other effects related to cannabis attitudes reached significance.

Discussion

Using a new cross-cultural approach in neuroscience, this study found different brain responses to cue-elicited craving, which is a sign of addiction, in individuals with CUD across regions with strict cannabis laws (Texas) versus more open laws (Netherlands). Despite the stricter legal environment, the Texas CUD group reported more positive and less negative attitudes about cannabis than the Netherlands CUD group. Importantly, site differences appeared in how self-reported craving, CUD severity, and brain responses to cues were linked, with evidence suggesting that positive cannabis attitudes played a role in these site differences.

Contrary to expectations, individuals with CUD in Texas held more positive and less negative personal attitudes about cannabis, and they also perceived their social environment's beliefs to be more positive, compared to the Netherlands group. It is possible that individuals who continue to use cannabis in a restrictive environment may have more positive views of its effects, as those with more negative views might choose not to use it due to legal consequences. Additionally, while individuals with CUD in Texas live in a restrictive state, their more positive and less negative attitudes may reflect a cultural shift toward more open policies in other U.S. states and a rising perception of low risk associated with cannabis across the U.S. Therefore, it is important to track changes in risk perceptions at both state and national levels—especially after changes in laws—to help shape future policies aimed at reducing the harm from cannabis use.

The Texas CUD group showed positive links between craving prompted by cues, CUD severity, and brain responses to cues in specific areas of the brain. In contrast, as self-reported craving and CUD severity increased in the Netherlands CUD group, brain responses to cues in these regions decreased. The precuneus is often active during cue reactivity and is involved in complex thought processes, including remembering past events and thinking about oneself. The lingual gyrus, calcarine cortex, and lateral occipital cortex are involved in visual processing, which can be heightened by greater attention and emotion. These different links between brain responses to cues, CUD severity, and self-reported craving suggest that this sign of addiction may work differently across cultures. Greater brain response to cues is thought to reflect more severe cue-elicited craving, which increases the risk of relapse when cues are encountered in daily life. Therefore, a reduced brain response to cues after treatment has been used as a sign of an individual's lower likelihood of relapse. The negative link observed between self-reported craving and brain response to cues suggests that heavy users in the Netherlands might have better control over their use compared to their Texas counterparts. More long-term data is needed to test this idea. These findings highlight a challenge: it cannot be assumed that these brain mechanisms function the same way in every cultural context.

Cultural influences likely shape personal cannabis attitudes, which in turn affect the relationship between brain responses to cues and craving in specific brain regions where site-related differences in cue reactivity were observed. Individuals with more positive personal cannabis attitudes showed a greater brain response to cues as self-reported craving increased, whereas those with less positive attitudes showed a decreased brain response to cues as self-reported craving increased. These findings emphasize the connection between conscious motivational processes (such as cannabis attitudes and craving) and brain responses to cues. This highlights the need to increase awareness of the potential importance of perceived harms and benefits of cannabis use in clinical practice. Future research should examine whether positive cannabis attitudes change how well cue-induced craving and brain responses to cues predict whether someone will stop using cannabis. This could shed light on the practical implications of these findings for treatment. It is important to note that cannabis attitudes did not change brain responses to cues in regions where there were significant differences in how CUD severity was linked across sites. It remains unclear what site-related differences might explain this effect.

Increased brain response to cannabis cues was observed in the CUD group compared to control groups in both comparisons. When controlling for neutral cues, increased cue-reactivity appeared in four clusters in the occipital cortex. When further controlling for appetitive cues, additional clusters emerged across several brain regions. Together, these findings indicate a relative decrease in response to a natural reward cue (like fruit) in individuals with CUD in brain regions involved in the motivational processes of addiction. This aligns with a theory of addiction where sensitivity to drug-related cues increases while sensitivity to natural rewards decreases. However, the decreased response to fruit cues could be specific to how cannabis affects food behavior, rather than a general addiction process.

This study has several limitations. First, the Texas CUD group reported higher cannabis use frequency, quantity, duration, recent use, and more CUD symptoms. Although statistically significant, both groups reported using cannabis six to seven days per week and fell within the moderate severity range for symptoms, suggesting the samples were similar across sites. Notably, site effects remained significant even after accounting for differences in cannabis use characteristics and mental well-being, and removing potential extreme data points did not change the results. Additionally, the reported grams of cannabis used per week do not account for actual cannabinoid exposure, which likely differs between these regions based on the potency of available cannabis products. Future cross-cultural research should include analyses of biological samples to examine the role of cannabinoid exposure, and more research on differences in cannabis use characteristics across sites should be encouraged. Furthermore, this study only included cannabis users within a limited age range (18–30), which limits how broadly the findings can be applied to younger and older cannabis users. Also, it is crucial that future studies assessing the role of cultural attitudes in cannabis use include a balanced or representative sample of genders with enough participants to assess gender differences. Finally, it was not possible to control for differences across sites in the co-use of tobacco products, which is very common among cannabis users in Europe but much less so in the U.S. While the study sample was realistic, recruiting specific subgroups (e.g., cannabis users in the Netherlands who do not use tobacco, and cannabis users in Texas who do) could help understand how neurobiological effects interact and account for cross-cultural differences.

In summary, this study provides initial evidence of different relationships between brain responses to cannabis cues and craving in individuals with CUD across distinct cannabis environments. Positive attitudes about the effects of cannabis appear to drive these differences. These findings highlight the potential of using a cross-cultural approach in the neuroscience of addiction processes to better understand the factors that may help maintain problematic cannabis use.

Introduction

Laws related to cannabis use are changing, and many people now view cannabis more positively, seeing it as less harmful. These changes are connected to how people start using cannabis, how often they use it, and if they develop a cannabis use disorder. It is important to understand the links between cannabis laws, people's attitudes, and measurable signs of addiction. This study, a collaboration between research teams in Texas, USA, and Amsterdam, The Netherlands, looked at how cannabis culture relates to brain reactions to cannabis cues in individuals with cannabis use disorder.

In Texas, recreational cannabis use is against the law, but in the Netherlands, it has been decriminalized since 1976. In the United States, more adults are using cannabis, especially after states passed laws making it more permissive. People's beliefs about cannabis effects might contribute to this increase in use. In Texas, fewer adults report using cannabis in the past year, and more perceive a high risk from monthly use compared to the U.S. average, though these differences are small. It is worth noting that in Washington state, after recreational cannabis became legal, the link between seeing less harm and increased cannabis use grew stronger. Community attitudes also influence the development of cannabis use disorder: disapproval protects against it, while approval can be a risk factor. However, it is still unclear how legal policies, personal attitudes, and perceived community attitudes affect cannabis cue-reactivity, which is a sign of addiction.

Cannabis cue-reactivity is a stronger brain response to cannabis cues, such as the sight or smell of a joint, which happens because these cues have been repeatedly linked to the rewarding effects of using cannabis. Changes in brain networks involved in attention, reward, motivation, and cognitive control when exposed to substance-related cues are connected to increased use and the development of a disorder. Brain responses to cues have been used as a way to predict treatment success and relapse risk because consistent evidence supports this across various substances. In cannabis users, cue-reactivity is consistently observed in specific brain areas like the striatum and anterior cingulate cortex. More severe users often show greater activity in parts of the striatum, which is linked to self-reported cravings and other cannabis-related problems.

Research in cultural neuroscience has shown that cultural factors affect brain processes related to attention, perception, self-awareness, and emotion. However, the connection between these cultural differences in brain processes and the brain mechanisms underlying addiction is not well understood. For example, believing in negative outcomes can change how the brain reacts to alcohol cues. This suggests that beliefs about substances, which are shaped by legal and cultural surroundings, can alter how the brain reacts to cues.

This study aimed to: 1) compare positive and negative cannabis attitudes (both personal and perceived attitudes of their social environment) between individuals with cannabis use disorder and similar control groups in Texas and the Netherlands, given their different legal policies; 2) examine differences in brain responses to cannabis cues; and 3) investigate if cannabis attitudes are linked to any observed differences in brain responses between the two locations. Researchers expected that individuals with cannabis use disorder in the Netherlands would have more positive cannabis-related attitudes and perceive a more positive social and cultural environment, which aligns with the less restrictive legal setting there. It was also hypothesized that more positive attitudes would be linked to higher cue-reactivity. Therefore, individuals with cannabis use disorder in the Netherlands were expected to show higher cannabis cue-reactivity and stronger positive links with cannabis use measures, such as the number of cannabis use disorder symptoms and cannabis craving, compared to individuals in Texas.

Methods

Participants and procedures

A total of 221 individuals were recruited for the study through flyers and social media. All participants needed to be between 18 and 30 years old, right-handed, and have no issues with MRI scanning, history of regular illicit drug use, excessive alcohol use, known neurological disorders, brain damage, or other major health conditions. Participants in the cannabis group had to report at least two cannabis use disorder symptoms (according to DSM-5 criteria) during online and phone screenings, use cannabis 6–7 days per week for the past year, and have no current plans to stop using or seek treatment. Participants in the non-using group had to have used cannabis fewer than 50 times in their lives, with no more than five uses in the past year, and no use in the past three months. Urine tests were conducted to check for illicit drug use during lab visits, leading to some exclusions. Additional participants were excluded for excessive movement during scanning, structural brain abnormalities, too much cannabis use, experimenter error, and participant distraction. The final study group consisted of 187 individuals: 54 cannabis users from the Netherlands, 50 cannabis users from Texas, 50 non-using controls from the Netherlands, and 33 non-using controls from Texas.

The study procedures were approved by the University of Texas Dallas Institutional Review Board and the ethics committee of the University of Amsterdam. All participants provided informed consent and received payment for their time.

Tactile cannabis cue exposure paradigm

A previously established method was used, involving two sessions with 18 tactile-visual cues presented in a random order. The cues included a cannabis joint (in the Netherlands, rolled with a placebo mix) or a cannabis pipe (in Texas) to match local preferences. Other cues were a pen (neutral control) and the participant's favorite fruit (appetitive control) from a selection. The same cues were used throughout the trials. An experimenter placed the cue in the participant's left hand while a photo of the corresponding cue being held by the participant was shown on a screen. After each trial, participants rated their current urge to use cannabis on a scale from 1 to 10.

Questionnaires

Cannabis-related measures

To determine the presence of DSM-5 cannabis use disorder symptoms, a semi-structured interview called the Mini International Neuropsychiatric Interview was conducted. Information on past and current cannabis use patterns was gathered using a substance use history questionnaire. Positive (8 items) and negative (9 items) attitudes about the effects of cannabis were assessed using the Cannabis Culture Questionnaire. Participants completed this scale from three viewpoints: their personal attitudes, their perceived attitudes of family and friends, and their perceived attitudes of the wider society in Texas or the Netherlands.

Substance use, mental health, demographics, and other measures

A demographic questionnaire collected information on gender, age, and years of education. IQ was estimated using specific tasks from the Wechsler Adult Intelligence Scale. Participants reported their alcohol use and related problems, cigarette use, lifetime use of illicit substances, and symptoms of anxiety, depression, and attention deficit and hyperactivity disorder. General mental health symptoms were assessed with the DSM-5 self-rated level 1 cross-cutting symptom checklist.

Neuroimaging data collection and preprocessing

Brain scans (anatomical and structural) were collected at the University of Amsterdam and the University of Texas Dallas using different MRI machines. Anatomical scans were taken for registration purposes. During the cannabis cue exposure paradigm, two functional MRI sequences were collected. Standard preprocessing steps were performed on the brain imaging data, including skull removal, spatial smoothing, motion correction, and aligning functional images to anatomical images using specialized software.

Data analysis

Behavioral data

To understand changes in craving specifically due to cannabis cues, craving scores were calculated by subtracting the average craving reported after neutral and fruit trials from the craving reported after cannabis trials.

Differences between locations (Texas vs. Netherlands) and groups (cannabis users vs. controls) in participant characteristics (like demographics, mental health, and substance use) were assessed using appropriate statistical tests. A statistical model was used to examine the effects of cue type, group, and location on craving, taking into account individual differences. To assess group and location differences in cannabis attitudes, statistical models were used for positive and negative Cannabis Culture Questionnaire scores separately, also accounting for individual and perspective variations.

fMRI data

Initial analyses of individual brain scan data were computed using specialized software. A general linear model estimated predictors for each cue condition, accounting for brain responses over time. Subsequent analyses combined activity across both scanning sessions. The main comparisons of interest were "Cannabis > Neutral" and "Cannabis Cue > Neutral + Fruit cues" to identify brain activation specific to cannabis, controlling for both neutral cues and food-related appetitive activity.

Whole Brain Analyses

Statistical models were run to identify brain regions with differences between groups (cannabis users vs. controls), interactions between location and group, and associations with cannabis use measures (symptom count, grams per week, cue-induced craving) within the cannabis use disorder group. These associations with cannabis use measures were also compared between locations.

For all analyses, location was included as a control variable to adjust for scanner differences. Therefore, only interactions involving location were interpreted, rather than main effects of location. Multiple comparison correction was applied using specific statistical thresholds. The average peak activity of significant brain clusters was extracted for visualization. For clusters showing significant interactions involving location, further analyses examined whether cannabis attitudes were associated with these effects. Cannabis Culture Questionnaire scores were used to predict mean peak activity, considering location as a potential moderating factor, with a corrected statistical threshold applied.

Results

Sample characteristics

Within the cannabis use disorder group, individuals in Texas reported more cannabis use disorder symptoms, more days of cannabis use per week, longer duration of weekly use, more self-reported grams of use per week, and reported using cannabis in the previous 24 hours more than individuals in the Netherlands. The cannabis use disorder and control groups were well-matched for age and gender. However, the cannabis use disorder group reported fewer years of education, more lifetime illicit substance use episodes, higher depression and anxiety symptoms, were more often daily cigarette smokers, and scored lower on IQ tests. Significant differences between groups and locations were found for alcohol use scores and general mental health symptoms. The cannabis use disorder group in the Netherlands and the control group in Texas had lower alcohol use scores than controls in the Netherlands, while the cannabis use disorder group in Texas scored higher than Texas controls. The cannabis use disorder group in the Netherlands reported more general mental health symptoms than the Texas cannabis use disorder group, and within both locations, the cannabis use disorder group reported more symptoms than the controls.

Cannabis attitudes

For both positive and negative attitudes, significant interactions were observed between group and location, and between group and the perspective from which attitudes were reported (personal, family/friends, state/country). For positive attitudes, a significant interaction between location and perspective also emerged.

Further comparisons showed that the cannabis use disorder group reported more positive personal attitudes, and perceived both their family/friends and the wider state/country as having more positive cannabis attitudes compared to controls. The cannabis use disorder group also reported less negative personal cannabis attitudes and perceived their family/friends as having less negative attitudes. No differences were observed between groups in perceptions of negative country/state cannabis attitudes.

Regardless of their perspective, the cannabis use disorder group in the Netherlands reported significantly more positive cannabis attitudes than the controls in the Netherlands, but no differences were found in negative attitudes. The cannabis use disorder group in Texas reported significantly more positive and less negative attitudes than the controls in Texas. Furthermore, the cannabis use disorder group in Texas reported significantly more positive and less negative attitudes than the cannabis use disorder group in the Netherlands. In contrast, the controls in Texas did not report significantly different positive or negative attitudes compared to the controls in the Netherlands.

Regardless of the group, participants in Texas reported more positive personal and perceived family/friend attitudes, but no significant location differences were observed in perceptions of state/country attitudes.

Cue-elicited craving ratings

Significant interactions between group and cue condition, and between group and location, were found for self-reported craving ratings. Detailed comparisons revealed that the cannabis use disorder group reported higher craving after cannabis cues compared to both neutral and fruit cues. In contrast, no significant differences in craving ratings between conditions were observed in the control group.

Following all types of cues, the cannabis use disorder group in the Netherlands reported higher craving than the cannabis use disorder group in Texas. Conversely, the Texas controls reported higher craving than the Netherlands controls. Furthermore, craving was higher after all cue types in the Netherlands cannabis use disorder group compared to both Netherlands controls and Texas controls. However, the Texas cannabis use disorder group did not report significantly higher craving than the Texas controls.

Whole brain analyses

Can > neutral

The cannabis use disorder group showed increased brain activity compared to the control group in four clusters within the occipital cortex, which is involved in visual processing. No brain regions were found where activity was linked to the amount of cannabis used, symptom severity, or self-reported craving during the scan, either in the cannabis use disorder group overall or differently between locations.

Can > neutral + appetitive

The cannabis use disorder group showed increased brain activity compared to the control group in eight clusters, including parts of the precuneus, frontal medial cortex, thalamus, nucleus accumbens, and cerebellum. While this activity was not related to the quantity of cannabis use in the cannabis use disorder group, different associations between brain activity (cannabis cues compared to neutral and appetitive cues) and cannabis use disorder symptom severity and in-scanner craving ratings emerged. Higher cannabis use disorder symptom count was linked to higher cannabis cue-reactivity in the Texas group, while fewer cannabis use disorder symptoms were related to higher cannabis cue-reactivity in the Netherlands group in several visual processing brain regions. Additionally, in the Texas cannabis use disorder group, higher cannabis craving after cannabis cues was linked to higher cannabis cue-reactivity in the occipital cortex, while the opposite was observed in the Netherlands cannabis use disorder group.

Further statistical analyses examined whether cannabis attitudes interacted with craving or symptom count in the peak brain activity of clusters with significant location differences. A significant interaction was observed between personal positive attitudes and craving. Individuals with cannabis use disorder who reported fewer positive attitudes about cannabis showed lower cannabis cue-reactivity as craving increased, while those who reported more positive attitudes showed higher cannabis cue-reactivity as craving increased. This effect remained significant even after controlling for demographic and other cannabis use characteristics. No other effects related to cannabis attitudes reached significance.

Discussion

This study, using a new cross-cultural neuroscience approach, found different brain responses to cue-induced craving, a sign of addiction, in individuals with cannabis use disorder across environments where cannabis use is illegal (Texas) versus permissive (Netherlands). Even with stricter laws, the Texas cannabis use disorder group reported more positive and less negative cannabis attitudes than the Netherlands cannabis use disorder group. Significantly, differences emerged between locations in how self-reported craving, cannabis use disorder severity, and brain cue-reactivity were linked, suggesting that positive cannabis attitudes play a role in these differences.

Contrary to expectations, individuals with cannabis use disorder in Texas had more positive and less negative attitudes about cannabis, both personally and regarding their social environment's beliefs, compared to the Netherlands group. It is possible that individuals who continue using cannabis in a restrictive environment may have more positive views of its effects, as those with more negative views might choose to avoid use due to legal consequences. Also, while individuals with cannabis use disorder in Texas live in a state with strict laws, their more positive and less negative attitudes might reflect a broader cultural shift towards more permissive policies in other U.S. states and a rising perception of low risk from cannabis in the U.S. generally. Therefore, it is essential to monitor changes in risk perceptions at both state and national levels, especially after changes in laws, to help shape future policies aimed at reducing the harm of cannabis use.

The Texas cannabis use disorder group showed positive links between cue-induced craving, cannabis use disorder severity, and brain cue-reactivity in brain areas involved in visual processing and memory. In contrast, as self-reported craving and cannabis use disorder severity increased in the Netherlands cannabis use disorder group, brain cue-reactivity in these regions decreased. These brain areas are involved in complex cognitive functions including visual processing, memory, and self-referential thought. These different associations suggest that this addiction sign may function differently across cultures. Greater cue-reactivity is generally believed to reflect more severe cue-induced craving, leading to a higher risk of relapse when cues are encountered in real life. Therefore, reduced cue-induced craving after treatment has been seen as a sign of an individual's lower risk of relapse. The negative link observed between self-reported craving and cue-reactivity in the Netherlands suggests that heavy users there might have better control over their use compared to their Texas counterparts. More long-term studies are needed to test this idea. These findings highlight that we cannot assume brain mechanisms work the same way in every cultural context.

Cultural influences likely shape personal cannabis attitudes, which in turn affect the relationship between brain cue-reactivity and craving in visual processing areas, where location-related differences in cue reactivity were observed. Individuals with more positive personal cannabis attitudes showed greater cue-reactivity as self-reported craving increased, whereas those with less positive attitudes showed decreased cue-reactivity as self-reported craving increased. These findings emphasize the connection between conscious motivational processes (like cannabis attitudes and craving) and brain cue-reactivity, highlighting the need to consider the potential importance of perceived harms and benefits of cannabis use in clinical settings. Future research should explore whether positive cannabis attitudes influence how well cue-induced craving and brain cue-reactivity predict whether someone stops using cannabis. It is important to note that cannabis attitudes did not change cannabis cue-reactivity in brain regions that showed significant differences in association with cannabis use disorder severity across locations. It remains unclear what location-related factors might explain this.

Increased cannabis cue-reactivity was observed in the cannabis use disorder group compared to controls in both comparisons. When controlling for neutral cues, increased cue-reactivity appeared in areas of the occipital cortex. When also controlling for appetitive cues, additional brain areas like the precuneus, nucleus accumbens, thalamus, and medial frontal cortex also showed increased activity. These findings suggest a relative decrease in response to natural reward cues (like fruit) in individuals with cannabis use disorder in brain regions linked to motivation in addiction. This aligns with the theory of incentive salience in addiction, where sensitivity to drug-related cues increases while sensitivity to natural rewards decreases. However, the decreased response to fruit cues could be specific to the effects of cannabis on food behavior (such as the "munchies"), rather than a general addiction process.

This study has several limitations. First, the Texas cannabis use disorder group reported more frequent, higher quantity, longer duration, and more recent cannabis use, as well as more cannabis use disorder symptoms. While statistically significant, both groups reported using cannabis six to seven days per week and were classified as having moderate symptom severity, suggesting the samples were generally similar across locations. Notably, location effects remained significant even after controlling for various cannabis use characteristics and mental well-being, and removing potential outliers did not change the results. Additionally, reported grams of cannabis used per week does not account for actual cannabinoid exposure, which likely differs between these regions due to varying cannabis product potency. Future cross-cultural research should include biological sample analyses to examine the role of cannabinoid exposure and encourage studies focusing on differences in cannabis use characteristics across locations. Furthermore, this study only included cannabis users aged 18–30, limiting how broadly the findings apply to younger and older users. It is also crucial that future studies assessing the role of cultural attitudes in cannabis use include a balanced or representative gender sample with enough participants to assess gender differences. Finally, researchers could not control for differences across locations in the co-use of tobacco products, which is very common among cannabis users in Europe but less so in the U.S. While the study sample was realistic, recruiting less common subgroups (e.g., non-tobacco-using cannabis users in the Netherlands and tobacco-using cannabis users in Texas) could help understand interactive brain effects and account for cross-cultural differences.

In summary, this study provides initial evidence of differing relationships between brain cannabis cue-reactivity and cue-induced craving in individuals with cannabis use disorder across distinct cannabis environments. Positive attitudes about the effects of cannabis appear to drive these differences. These findings highlight the potential of using a cross-cultural approach in addiction neuroscience to better understand the factors that may contribute to the ongoing nature of cannabis use disorder.

Introduction

The laws around cannabis use have become less strict in many places, and people's views on cannabis have also become more positive. This shift has led to people seeing cannabis as less harmful. These changes are connected to how often people start using cannabis, how frequently they use it, and whether they develop a cannabis use disorder, which is a problematic pattern of use. It is important to understand how cannabis laws, people's beliefs, and signs of addiction in the body are connected. This study, conducted in Texas, USA, and Amsterdam, The Netherlands, looked at how cannabis culture affects how people with cannabis use disorder react to things that remind them of cannabis.

Recreational cannabis use is against the law in Texas, while The Netherlands made it legal to some extent in 1976. Across the US, more adults have reported using cannabis in the past year, especially after states adopted more relaxed cannabis laws. People's beliefs about cannabis effects might lead to this increase in use. In Texas, fewer adults use cannabis each year compared to the national average, and more people there see a high risk in using it monthly. However, these differences are slight, and not much is known about how regular cannabis users in Texas perceive risk. For example, in Washington state, after recreational cannabis became legal, people who saw less harm in cannabis were more likely to use it. The opinions of a community can also affect whether someone develops a cannabis use disorder: if the community disapproves, it can offer protection, but if it approves, it can increase the risk. Still, it is not clear how laws, personal beliefs, and community views affect how regular cannabis users react to cannabis cues, which is a sign of addiction.

Cannabis cue-reactivity refers to the strong physical and brain responses someone has when they encounter things that remind them of cannabis, like a joint or its smell. This happens because the brain has learned to link these cues with the pleasurable effects of cannabis. Changes in brain areas related to attention, reward, motivation, and decision-making when seeing drug-related cues are linked to increased drug use and developing a disorder. Brain reactions to cues are used as a sign to predict how well treatment might work or if someone might relapse, because this has been seen with many different substances. In cannabis users, this brain activity is often seen in specific brain areas like the striatum, anterior cingulate cortex, middle frontal gyrus, and precuneus. Users with more severe problems show greater activity in the striatum, and this activity is linked to their reported cravings and cannabis-related issues.

Studies in cultural neuroscience show that cultural differences affect brain processes related to how people pay attention, see things, understand themselves, and feel emotions. However, it is not clear how these cultural differences in brain processes connect to the brain changes involved in addiction. For example, believing that alcohol will have bad effects can change how the brain reacts to alcohol cues. This suggests that beliefs about substances, which are shaped by laws and culture, can change how the brain reacts to those substances.

This study had three main goals: 1) to compare positive and negative attitudes about cannabis between people with cannabis use disorder and a control group in Texas and The Netherlands, considering different laws in each place; 2) to look at differences in how the brain reacts to cannabis cues; and 3) to see if cannabis attitudes are linked to any differences in brain reactions found between the two locations. Researchers expected that people with cannabis use disorder in The Netherlands would have more positive views on cannabis and see their social environment as more accepting, because the laws there are more relaxed. They also expected that more positive attitudes would lead to stronger reactions to cannabis cues, and that people with cannabis use disorder in The Netherlands would show stronger brain reactions to cannabis cues, linked to their cannabis use and craving, compared to those in Texas.

Methods

Participants and Procedures

Researchers recruited 221 individuals through flyers and social media. To participate, individuals had to be between 18 and 30 years old, right-handed, and have no medical conditions or drug use history that would interfere with the study. Those in the cannabis use disorder (CUD) group reported at least two CUD symptoms and used cannabis 6–7 days a week for the past year without plans to stop. The control group used cannabis fewer than 50 times ever, with no more than five uses in the last year, and none in the last three months. Some participants were excluded after drug tests or due to issues during the brain scans. In total, 187 individuals completed the study, divided into groups from The Netherlands (NL) and Texas (TX) who either used cannabis or were controls.

The study procedures were approved by ethics committees at both universities. All participants agreed to take part after being fully informed about the study and received payment for their time.

Tactile Cannabis Cue Exposure Paradigm

The study used a method where participants were shown and touched objects related to cannabis. Each participant completed two sessions, with 18 different items shown in a random order. These items included a cannabis joint (for NL participants) or a pipe (for TX participants), a pen (as a neutral object), and their favorite fruit (such as an orange or apple, as a desirable non-drug item). As a picture of the item appeared on a screen, an experimenter placed the actual item in the participant's left hand. After each item, participants rated their urge to use cannabis on a scale from 1 to 10.

Questionnaires

Participants filled out several questionnaires to gather information. To understand their cannabis use and attitudes, they completed a survey on their history of substance use and a "Cannabis Culture Questionnaire." This questionnaire measured both positive and negative beliefs about cannabis effects from three viewpoints: their own personal attitudes, what they thought their family and friends believed, and what they thought society in general (Texas or The Netherlands) believed. A structured interview also helped identify symptoms of cannabis use disorder.

Additional questionnaires collected details about demographics such as gender, age, and education. Their intelligence (IQ) was estimated using standard tests. Participants also reported their alcohol use and related problems, cigarette use, lifetime use of other illegal substances, and symptoms of mental health conditions like anxiety, depression, and attention deficit and hyperactivity disorder. A general checklist was used to assess overall mental health symptoms.

Neuroimaging Data Collection and Preprocessing

Brain scans were performed at both study sites using MRI machines. These scans captured detailed images of the brain's structure and activity. During the cannabis cue task, specific types of functional MRI scans were taken to measure changes in blood flow in the brain, which indicate brain activity. Before analyzing the data, the raw brain images underwent several steps to prepare them. These steps included removing parts of the skull from the images, smoothing the images, correcting for any head movements during the scan, and aligning the functional images with the structural images. This preparation ensures the data is accurate for analysis.

Data Analysis

To analyze the behavioral data, researchers calculated how much craving increased after seeing cannabis cues by subtracting craving for neutral and fruit items. Statistical tests were used to look for differences between groups and sites in participant characteristics like age, mental health, and substance use. More advanced statistical models were used to understand how cue type, participant group, and study site affected craving and cannabis attitudes.

For the fMRI brain scan data, activity related to each type of cue was first calculated for each person. Then, overall activity across all sessions was combined. Researchers specifically looked at brain areas that showed more activity when participants saw cannabis cues compared to neutral items, and also compared to both neutral and fruit items. This helped identify brain responses specifically linked to cannabis, even when controlling for general interest in appealing things.

In the whole-brain analyses, further statistical models were used to find brain regions where activity differed between groups, or where the effect of group varied by site. They also looked at how brain activity was linked to cannabis use measures like symptom count, amount used per week, and craving for those with cannabis use disorder. To account for differences between the MRI scanners at each site, the study site was included in the analysis. If there were significant interactions involving the study site, further analyses were done to see if cannabis attitudes explained these differences in brain activity.

Results

Sample Characteristics

Among those with cannabis use disorder (CUD), individuals in Texas (TX) reported more CUD symptoms, used cannabis more days per week, used it for a longer period weekly, used more grams per week, and had used it more recently than individuals in The Netherlands (NL). The CUD and control groups were similar in age and gender. However, the CUD group generally had less education, more history of illegal drug use, higher levels of depression and anxiety, were more often daily cigarette smokers, and scored lower on IQ tests. Differences related to alcohol use and overall mental health symptoms varied by group and location. For example, the NL CUD group had more general mental health symptoms than the TX CUD group, and in both locations, the CUD group had more symptoms than the control groups.

Cannabis Attitudes

The study found that both positive and negative cannabis attitudes varied depending on the group (CUD vs. control) and the study site (TX vs. NL), as well as by whose perspective was being considered (personal, family/friends, or society). Overall, the CUD group reported more positive personal attitudes about cannabis and believed their family, friends, and society also held more positive views compared to the control group. The CUD group also held fewer negative personal attitudes and thought their family and friends were less negative about cannabis.

When comparing sites, the NL CUD group reported more positive cannabis attitudes than NL controls, but their negative attitudes were similar. The TX CUD group reported both more positive and less negative attitudes than TX controls. Interestingly, the TX CUD group had significantly more positive and less negative attitudes than the NL CUD group. However, the TX control group's attitudes were similar to the NL control group's. In general, Texas participants, regardless of their group, reported more positive personal attitudes and beliefs about their family and friends' attitudes than participants in The Netherlands.

Cue-Elicited Craving Ratings

Participants' self-reported craving levels changed differently depending on their group (CUD or control) and their location. The cannabis use disorder (CUD) group reported much higher craving after seeing cannabis cues compared to neutral objects or fruit. The control group, however, showed no significant difference in craving regardless of the type of object they saw.

The NL CUD group generally reported higher craving across all types of objects than the TX CUD group. On the other hand, the TX control group reported higher craving than the NL control group. The NL CUD group showed much higher craving than both the NL and TX control groups. However, the TX CUD group did not report significantly higher craving than the TX control group.

Whole Brain Analyses

When looking at brain activity, the cannabis use disorder (CUD) group showed more activity than the control group in several brain regions when exposed to cannabis cues, even when compared to neutral objects. These areas included parts of the visual cortex. No link was found between this brain activity and the amount of cannabis used, symptom severity, or craving.

However, when comparing cannabis cue activity to both neutral and appealing (fruit) cues, the CUD group showed increased activity in additional brain areas, including those involved in memory, emotion, and reward. While the amount of cannabis used was still not linked to brain activity, the severity of CUD symptoms and reported craving showed different patterns between sites. In the Texas CUD group, more severe symptoms were linked to higher brain reactions to cannabis cues. But in the NL CUD group, fewer symptoms were linked to higher brain reactions in similar areas. Similarly, in the Texas CUD group, more craving was linked to higher brain reactions, while the opposite was true for the NL CUD group.

Further analysis showed that personal positive attitudes about cannabis played a role. Individuals with CUD who had fewer positive attitudes about cannabis showed less brain reaction to cues as their craving increased. In contrast, those with more positive attitudes showed greater brain reactions as their craving increased. This specific finding was strong and remained even after considering other factors like gender and differences in cannabis use or mental health.

Discussion

This study used a new approach to compare brain responses to cannabis cues in different cultures. It found clear differences in how individuals with cannabis use disorder (CUD) reacted to these cues, depending on whether they lived in a place where cannabis use is illegal (Texas) or more accepted (The Netherlands). Surprisingly, even though Texas has stricter laws, people with CUD there reported more positive and fewer negative attitudes about cannabis than those in The Netherlands. Importantly, the connections between reported craving, CUD severity, and brain reactions to cannabis cues also varied between the two locations, with positive cannabis attitudes playing a role in these differences.

Against what was expected, individuals with cannabis use disorder in Texas held more positive and less negative views about cannabis, both personally and in terms of what they thought their social circles believed, compared to the group in The Netherlands. One reason for this might be that people who continue using cannabis in a place with strict laws may simply have more positive views about its effects, while those with negative views might stop using it because of the legal risks. Also, even though Texas has strict laws, the positive attitudes there might reflect a broader cultural shift across other US states where cannabis laws are becoming more relaxed and people see it as less risky. Therefore, it is important to keep track of how people perceive the risks of cannabis at both state and national levels, especially after new laws are passed, to help create policies that reduce harm from cannabis use.

In the Texas CUD group, higher craving and more severe CUD symptoms were linked to stronger brain reactions to cannabis cues in areas like the lateral occipital cortex and precuneus. However, in the NL CUD group, as craving and symptom severity increased, the brain's reaction in these same areas actually decreased. The precuneus is known for its role in memory and self-awareness, while the other areas mentioned are involved in processing visual information, which can be more active with increased attention or emotion. These different patterns suggest that the brain's response to cues, which is a sign of addiction, might work differently depending on the culture. Stronger brain reactions to cues are usually thought to mean stronger cravings and a higher risk of relapse. Therefore, a weaker reaction to cues after treatment is seen as a sign that someone is less likely to relapse. The opposite link seen in the NL group might mean that heavy users there have more control over their cannabis use compared to those in Texas. More long-term studies are needed to confirm this idea. These results show that what we learn in one cultural setting might not apply directly to another.

It is likely that cultural factors shape a person's attitudes about cannabis, and these attitudes then influence how the brain reacts to cannabis cues in response to craving. For example, people with more positive personal views on cannabis showed stronger brain reactions to cues as their craving grew. In contrast, those with less positive attitudes showed weaker brain reactions as craving increased. These results highlight how conscious motivations, like attitudes and craving, are linked to brain responses. This suggests that in clinical settings, it is important to discuss people's beliefs about the harms and benefits of cannabis use. Future studies could explore if positive cannabis attitudes influence how well cue-induced craving and brain reactions predict whether someone will stop using cannabis. It is worth noting that while attitudes affected the link between craving and brain reactions, they did not explain the differences in brain reactions related to CUD severity between the two locations. The reasons for these site-specific differences remain unknown.

This study had several limitations. First, the Texas cannabis use disorder (CUD) group reported using cannabis more often, in greater amounts, for longer periods, and had more symptoms than the NL CUD group. While these differences were statistically significant, both groups used cannabis frequently (six to seven days a week) and had a moderate level of symptoms, suggesting they were still comparable. Even when these differences were accounted for in the analysis, the main findings related to location remained strong. Also, simply measuring grams used per week does not fully show how much of the active chemicals in cannabis a person was exposed to, which can vary by region due to product strength. Future studies should test biological samples to better understand this. Additionally, the study only included cannabis users aged 18–30, so the results may not apply to younger or older users. Future research should also include a diverse gender sample to study potential differences. Finally, the study could not account for differences in tobacco use between sites, which is common among cannabis users in Europe but less so in the US.

In summary, this study offers early evidence that the brain's reactions to cannabis cues and related cravings differ in people with cannabis use disorder, depending on the legal and cultural environment they live in. Positive attitudes about cannabis seem to be a key factor in these differences. These results show that studying addiction through a cultural lens can help us better understand what keeps cannabis use disorder going.

Introduction

Making cannabis less illegal has gained speed. Along with this, people are starting to think more positively about cannabis and see it as less harmful. These changes are linked to how people start using cannabis, how often they use it, and if they develop a cannabis use problem (CUD). It is very important to understand how cannabis laws, people's thoughts about cannabis, and signs of addiction in the body are connected. This study brought together researchers from Texas (USA) and Amsterdam (Netherlands) to see how cannabis culture affects how the brain reacts to cannabis in people with CUD.

In Texas, using cannabis for fun is against the law. But in the Netherlands, the laws became less strict in 1976. In the USA, more adults have used cannabis in the past year, especially after states made cannabis laws easier. People's beliefs about how cannabis affects them might be causing this rise in use. In Texas, fewer adults (about 15%) used cannabis in the past year compared to the average across the USA (about 21%). Also, more people in Texas thought using cannabis monthly was very risky (about 25%) compared to the USA average (about 21%). However, these differences are small, and not enough is known about how often regular users see risks. In one state, Washington, thinking cannabis was less harmful became even more linked to using it more after it became legal for fun.

What a community thinks can also affect if someone develops CUD. If people do not approve of cannabis, it can help protect others from using it too much. If people approve, it can make it more likely to have problems. But it is still not clear how laws, what a person thinks, and what a community thinks affects how the brain reacts to cannabis cues, which is a sign of addiction, in people who use cannabis regularly.

Cannabis cue-reactivity means a strong brain reaction to things linked to cannabis, like a joint or its smell. This happens when the brain repeatedly links these cues to the good feelings of using cannabis. When people see things linked to drugs, changes happen in brain areas that deal with noticing things, rewards, wanting things, and clear thinking. These changes are linked to using more and more of the drug and developing a drug problem. How the brain reacts to drug cues has been used as a sign of how well treatment will work and the risk of starting to use again, because this has been seen with many different drugs. In cannabis users, this brain reaction is often seen in certain brain areas. People with more severe problems show more activity in some of these areas, and this is linked to how much they want cannabis and problems they report from using it.

Studies on culture and the brain show that culture affects how the brain pays attention, sees things, understands oneself, and feels emotions. But it is not clear how these cultural differences in brain processes are connected to how addiction works in the brain. Expecting bad results can change how the brain reacts to alcohol cues. This supports the idea that what people believe about drugs, which is shaped by laws and culture, can change how the brain reacts to drug cues.

This study aimed to do three things: 1) compare positive and negative thoughts about cannabis (both personal and what they thought others believed) between people with CUD and similar control groups in Texas and the Netherlands, which have different cannabis laws; 2) look at differences in how the brain reacts to cannabis cues; and 3) see if thoughts about cannabis were linked to the brain reaction differences seen between the places. Researchers thought that people with CUD in the Netherlands would have more positive thoughts about cannabis and feel that their friends and culture were more okay with it, because of the less strict laws there. They also thought that more positive thoughts would be linked to a stronger brain reaction. They expected people with CUD in the Netherlands to show a stronger brain reaction to cannabis cues and stronger links to how much cannabis was used and wanted, compared to people in Texas.

Methods

People for the study were found through flyers and social media. In total, 221 people joined. All participants had to be 18–30 years old, right-handed, and have no problems with MRI scans, no history of regular illegal drug use, no excessive alcohol use, no known brain problems, or other major health issues. People in the cannabis group had to report at least two signs of cannabis use disorder and use cannabis 6–7 days a week for the past year, with no plan to stop or get help. People in the non-using group had to have used cannabis fewer than 50 times in their life, used it five times or less in the past year, and not used it in the past three months. A quick urine test was done to check for illegal drug use. Some people were not included due to moving too much during the scan, unusual brain parts, mistakes by the researchers, or being distracted. In the end, 187 people were part of the study: 54 cannabis users from the Netherlands, 50 cannabis users from Texas, 50 non-using controls from the Netherlands, and 33 non-using controls from Texas. The study rules were approved by research review boards in both Texas and Amsterdam. Each person agreed to be in the study and was paid for their time.

The study used a proven method where people touched and saw cannabis cues. Each person had two turns. During these turns, they were given a cannabis joint (in the Netherlands, made with a fake herbal mix) or a cannabis pipe (in Texas) to hold. These matched what people used locally. They also held a pen (a normal item for comparison) and their favorite fruit (something that usually makes people want it). Pictures of these items were shown on a screen. After each test, people rated how much they wanted to use cannabis on a scale from 1 to 10.

To learn about cannabis use, people answered questions from the Mini International Neuropsychiatric Interview to check for signs of cannabis use disorder. They also filled out a form about their past and current cannabis use. A survey called the Cannabis Culture Questionnaire (CCQ) asked about positive and negative thoughts about cannabis effects. People answered these questions from three viewpoints: their own thoughts, what they thought their family and friends thought, and what they thought the wider society (Texas/Netherlands) thought. Other questions asked about age, gender, and how many years of schooling they had. Their intelligence (IQ) was estimated using word and puzzle tests. People also reported their alcohol use, cigarette use, if they had ever used illegal drugs, and signs of worry, sadness, and ADHD. A general mental health checklist was also used.

Brain scans were done in Amsterdam and Texas. Detailed pictures of the brain were taken to help line up the images. During the cannabis cue tests, two types of brain scans were taken quickly. The brain scan data was then cleaned and prepared using special computer programs.

To find out about changes in wanting cannabis only when seeing cannabis cues, the average wanting from the normal and fruit tests was taken away from the cannabis test wanting scores. Differences between places and groups in age, gender, mental health, and drug use were checked using special math tests. Another math model was used to look at how the type of cue, group, and place affected wanting. To check for group and place differences in cannabis thoughts, other math models were used for positive and negative CCQ scores separately.

Brain scan data was analyzed in steps. First, brain activity for each person was looked at. Then, activities from both test turns were combined. The main comparisons were cannabis cues versus normal cues, and cannabis cues versus normal and fruit cues. This helped find brain activity specifically for cannabis, while also accounting for normal items and food-related items. For all analyses, "place" (Texas or Netherlands) was included to account for differences in scanning machines. This meant that the focus was on how place interacted with other factors, not just on place itself. Special math rules were used to make sure the results were real. The average highest activity in important brain areas was noted to show the results clearly. For brain areas with important differences by place, further math tests were done to see if cannabis thoughts were linked. CCQ scores were used to predict brain activity, with and without place being a factor, to see if cannabis thoughts helped explain differences by place.

Results

Within the group of cannabis users, people in Texas reported more signs of cannabis use problems, used cannabis more days per week, used it for a longer time each week, reported using more grams per week, and said they used cannabis in the past 24 hours more than people in the Netherlands. The cannabis user and control groups were similar in age and gender. However, the cannabis user group had fewer years of schooling, more illegal drug use incidents in their life, more signs of sadness and worry, were more often daily cigarette smokers, and scored lower on intelligence tests. Important links between group and place were seen in alcohol problem scores and general mental health symptoms. The cannabis user group from the Netherlands and the control group from Texas had lower alcohol problem scores than the control group from the Netherlands. The cannabis user group from Texas scored higher than the control group from Texas. The cannabis user group from the Netherlands reported more general mental health problems than the cannabis user group from Texas. In both places, the cannabis user group reported more mental health problems than the control groups. (See Table 1 for sample details.)

For both positive and negative thoughts about cannabis, important connections between group and place, and group and viewpoint (personal, family/friends, state/country) were found. For positive thoughts only, an important connection between place and viewpoint was also seen (see Figure 1). A closer look showed that the cannabis user group had more positive personal thoughts about cannabis, and thought their family/friends and the wider society also had more positive views, compared to the control groups. The cannabis user group also had less negative personal thoughts about cannabis and thought their family/friends were less negative. No group differences were seen in what people thought about the state's or country's negative views on cannabis.

No matter the viewpoint, the cannabis user group from the Netherlands reported much more positive cannabis thoughts than the control group from the Netherlands, but no differences were seen in negative thoughts. The cannabis user group from Texas reported much more positive and less negative thoughts than the control group from Texas. Also, the Texas cannabis user group reported much more positive and less negative thoughts than the Netherlands cannabis user group. In comparison, the control group from Texas did not report very different positive or negative thoughts compared to the control group from the Netherlands. No matter the group, people in Texas reported more positive personal thoughts and what they thought their family/friends thought. But no major differences were seen in what they thought the state or country believed.

Important links between group, test type, and place were seen in how much people wanted cannabis. A closer look showed that the cannabis user group wanted cannabis more after seeing cannabis cues, compared to both normal and fruit cues (see Figure 2). In contrast, the control group did not show major differences in wanting for different cues. After all types of tests, the cannabis user group from the Netherlands wanted cannabis more than the cannabis user group from Texas. However, the control group from Texas wanted cannabis more than the control group from the Netherlands. Also, the cannabis user group from the Netherlands wanted cannabis more after all types of tests compared to both the Netherlands and Texas control groups. But the Texas cannabis user group did not want cannabis much more than the Texas control group.

Whole Brain Analyses

Can > neutral

The cannabis user group showed more brain activity than the control group in four brain areas in the back of the head (see Table 2 and Figure 3). These brain areas were not linked to how much cannabis was used, how severe the problems were, or how much cannabis was wanted during the scan, either in general or differently between the places.

Can > neutral + appetitive

The cannabis user group showed more brain activity than the control group in eight brain areas, including parts of the brain related to memory, control, reward, and vision (see Table 2 and Figure 3). While this brain activity was not linked to how much cannabis was used in the cannabis user group, different links appeared between this brain activity, how severe the cannabis problems were, and how much cannabis was wanted during the scan. More cannabis problems were linked to more brain reaction in the Texas group, but fewer cannabis problems were linked to more brain reaction in the Netherlands group, in certain brain areas (Figure 4a). Also, in the Texas cannabis user group, wanting cannabis more after seeing cannabis cues was linked to more brain reaction, while the opposite was seen in the Netherlands cannabis user group (Figure 4b).

Further math tests looked at whether cannabis thoughts worked together with wanting or problem severity in the brain areas that showed important differences by place (see Table 3). An important link was seen between personal positive thoughts and wanting. People with cannabis use disorder who had fewer positive thoughts about cannabis showed less brain reaction as they wanted it more. On the other hand, people with more positive thoughts showed more brain reaction as they wanted it more (Figure 4c). This result stayed true even after considering gender, cannabis use habits, and mental health differences by place, and even after removing unusual data points. No other effects of cannabis thoughts were important.

Discussion

Using a new way to study the brain across cultures, this study found different brain reactions to wanting cannabis when shown cues. This was seen in people with cannabis use disorder in places where cannabis use is against the law (Texas) versus where laws are less strict (Netherlands). Even with stricter laws, the cannabis user group in Texas reported more positive and less negative thoughts about cannabis than the group in the Netherlands. Importantly, differences by place appeared in how wanting, how severe the cannabis problems were, and how the brain reacted to cues were linked. This shows that positive thoughts about cannabis played a part in these differences by place.

This was not what researchers expected. People with cannabis use disorder in Texas had more positive and less negative thoughts about cannabis, both personally and what they thought their friends and family believed, compared to the group in the Netherlands. It might be that people who keep using cannabis where it's against the law have more positive ideas about it, because those who think it's bad might choose to stop due to legal reasons. Also, even though people with CUD in Texas live where laws are strict, their more positive thoughts might show a general shift towards less strict cannabis laws in other US states and more people thinking cannabis is not risky across the whole USA. So, it is very important to keep track of how people see risks in both states and the whole country, especially after laws change, to help make future rules that reduce problems from cannabis use.

The Texas cannabis user group showed links where wanting cannabis, how bad the problems were, and brain reaction all went up together in certain brain areas. But, in contrast, as wanting cannabis and problems got worse in the Netherlands cannabis user group, brain reaction in these areas went down. The precuneus brain area is often active when seeing cues and helps with memory and thinking about oneself. Other brain areas help with seeing things, and this can be stronger with more focus and feeling. These different links between brain reaction, how severe the cannabis problems were, and how much people wanted cannabis suggest that this sign of addiction might work differently in different cultures. More brain reaction is usually thought to mean more wanting, leading to a higher chance of starting to use again when seeing cues in real life. So, less wanting after treatment has been used to show that a person is less likely to start using again. The opposite link seen in the Netherlands suggests that heavy users there might control their use better than those in Texas. More studies over time are needed to check this idea. These results show that what is learned in one place might not be true everywhere; brain processes might work differently in different cultures.

Culture probably shapes what people think about cannabis, which then affects how brain reaction and wanting are linked in certain brain areas, where differences in brain reaction by place were seen. People with more positive personal thoughts about cannabis showed more brain reaction as they wanted it more. On the other hand, those with less positive thoughts showed less brain reaction as they wanted it more. These results highlight the link between clear reasons for wanting something (like thoughts about cannabis and wanting it) and how the brain reacts to cues. This shows that it is important to understand how people see the good and bad sides of cannabis in helping them. Future studies could look at how positive cannabis thoughts change how well wanting and brain reaction predict stopping cannabis use, which could help with treatment. It is important to note that cannabis thoughts did not change brain reactions in areas where brain reactions were linked to how bad the cannabis problems were across places. It is still not clear what differences by place might cause this.

More brain reaction to cannabis cues was seen in the cannabis user group compared to control groups in both comparisons. When only comparing to normal cues, more brain reaction appeared in four areas in the back of the brain. When also comparing to food cues, more brain areas showed up, including parts of the brain related to reward and control. These results together show less brain activity for a normal reward (like fruit) in people with cannabis use disorder, in brain areas important for wanting things related to addiction. This fits with a theory that says the brain gets more excited by drug cues at the same time it gets less excited by normal rewards. However, the less activity for fruit might be just about how cannabis affects eating (like the 'munchies') and not a general addiction process.

This study has some weaknesses. First, the cannabis user group in Texas reported using cannabis more often, in larger amounts, for longer, in the last 24 hours, and had more problems. Even though these differences were real, both groups used cannabis most days of the week and had medium problems, meaning the groups were mostly alike. Importantly, the differences by place were still strong even after checking for how long they used, if they used recently, how many grams they used, problem scores, other mental health issues, and gender. Removing unusual data did not change things. Also, how many grams used per week does not show how much actual cannabis chemical they took in, which is probably different in these places due to how strong the cannabis is. Future studies across cultures should include tests of body samples to see how much cannabis chemical they took in, and studies on cannabis use differences between places should be done. Also, this study only looked at cannabis users between 18 and 30, so the results might not apply to younger or older users. It is also very important that future studies on how culture affects cannabis use have an equal number of men and women, or enough people to see gender differences. Lastly, researchers could not control for how much people also used tobacco, which is common with cannabis users in Europe but not as much in the US. While this study group was realistic, getting more people from less common groups (like cannabis users in the Netherlands who do not use tobacco, or cannabis users in Texas who do) could help understand how the brain works differently and why cultures differ.

To sum up, this study shows early signs that how the brain reacts to cannabis cues and how much a person wants cannabis are different for people with cannabis use disorder in places with different cannabis laws. Thinking positively about cannabis seems to be why these differences exist. These results show that looking at addiction through a cultural lens can help understand what keeps cannabis use disorder going.