Discriminative Properties of Hippocampal Hypoperfusion in Marijuana Users Compared to Healthy Controls: Implications for Marijuana Administration in Alzheimer’s Dementia

Abstract

Background: Few studies have evaluated the impact of marijuana use on regional cerebral blood flow.

Objective: To determine whether perfusion in specific brain regions on functional neuroimaging, including those affected by Alzheimer’s disease pathology, are abnormal in marijuana users compared to controls.

Method: Persons with a diagnosis of cannabis use disorder by DSM-IV and DSM-V criteria (n = 982) were compared to controls (n = 92) with perfusion neuroimaging with SPECT at rest and at a concentration task. Perfusion estimates were quantified using a standard atlas. Cerebral perfusion differences were calculated using one-way ANOVA. Diagnostic separation was determined with discriminant analysis of all subjects. Feature selection with a minimum redundancy maximum relevancy (mRMR) identified predictive regions in a subset of marijuana users (n = 436) with reduced psychiatric co-morbidities.

Results: Marijuana users showed lower cerebral perfusion on average (p < 0.05). Discriminant analysis distinguished marijuana users from controls with correct classification of 96% and leave one out cross-validation of 92%. With concentration SPECT regions, there was correct classification of 95% with a leave-one-out cross validation of 90%. AUC analysis for concentration SPECT regions showed 95% accuracy, 90% sensitivity, and 83% specificity. The mRMR analysis showed right hippocampal hypoperfusion on concentration SPECT imaging was the most predictive in separating marijuana subjects from controls.

Conclusion: Multiple brain regions show low perfusion on SPECT in marijuana users. The most predictive region distinguishing marijuana users from healthy controls, the hippocampus, is a key target of Alzheimer’s disease pathology. This study raises the possibility of deleterious brain effects of marijuana use.

Keywords

Hippocampus imaging marijuana SPECT

INTRODUCTION

Reported rates of marijuana use have more than doubled in the past decade [1]. Medical marijuana is now legal in over half of the United States [2], increasing access to the drug for current and potential future users. As marijuana is often portrayed as a harmless drug in popular culture, a conception that marijuana is a benign substance with therapeutic value for some diseases is commonplace [3]. Concurrently, 10 of the 26 US states that have legalized marijuana include Alzheimer’s dementia (AD) as a valid indication for its medical use [4]. This therapeutic application of marijuana exists despite lack of conclusive evidence regarding benefit of this drug in a variety of disorders including AD [5, 6]. Understanding the relevance of marijuana use for AD necessitates investigation of its effects on the human brain compared to non-marijuana users. Currently, few studies have been published analyzing the impact that marijuana use has on the human brain.

The available literature assessing the influence of marijuana use on the brain has predominantly shown depressive effects on neurophysiology [7, 8]. A recent study using positron emission tomography (PET) imaging demonstrated decreased dopamine release in the striatum, a region of the brain that is involved in working memory, impulsive behavior, and attention [9]. These results show that heavy marijuana use has an effect on dopamine release similar to hard drugs, such as heroin and cocaine. Another study using diffusion tensor imaging showed damage to the corpus callosum, a white matter tract responsible for connecting the left and right hemispheres, in 59 individuals who smoke high-potency cannabis [10]. A third study using structural magnetic resonance imaging (MRI) demonstrated a decrease in orbitofrontal cortex volume in chronic marijuana users [11].

While a recent study suggested that marijuana can reduce amyloid-β-related inflammation by activation of cannabinoid receptors [12], other imaging studies have suggested structural damage in areas vulnerable to AD pathology. Several studies in chronic cannabis users showed decreased hippocampal volumes related to the amount of cannabis used [13 –15]. Most striking is that these structural changes are shown to be long lasting as the volume reductions seen with heavy use persisted after six months of abstinence.

There is lack of functional brain imaging data in large cohorts of persons with history of marijuana use. Such information is important given the improved sensitivity of functional imaging to physiological alterations. The purpose of this study is to further investigate specific regions of the brain that are neurophysiologically impaired in marijuana users compared to non-users. We hypothesize that marijuana users will exhibit hypoperfusion compared to controls on brain SPECT, including areas known to be affected by AD pathology such as the hippocampus.

METHODS

Study participants

This study was conducted in accordance with the STARD guidelines (http://www.stard-statement.org/). All subjects were obtained for retrospective analysis from a large multisite psychiatric database, involving 26,268 patients who came for evaluation of complex, treatment resistant issues to one of nine outpatient clinics (Newport Beach, Costa Mesa, Fairfield, and Brisbane, CA; Tacoma and Bellevue, WA; Reston, VA; Atlanta, GA; and New York, NY) between 1995–2014. Diagnoses were made by board certified or eligible psychiatrists, using all of the data available to them, including detailed clinical history, mental status examination, and DSM-IV or V criteria, consistent with the current standard of care. Anonymized data was extracted from a research database using a data mining technique within a protocol deemed appropriate by an independent IRB IntegReview (http://www.integreview.com/) to be exempt from human subjects review in accordance with 45 CFR 46.101(b)(4) (IRB #004).

Included in the database were healthy adult volunteers who had single-photon emission computed tomography (SPECT) studies. The exclusion criteria for the healthy subjects were: 1) current or past evidence of psychiatric illnesses as determined by clinical history, mental status examinations, and the Structured Clinical Interview for Diagnosis for DSM-IV; 2) current reported medical illnesses or medication; 3) history of brain trauma; 4) current or past drug or alcohol abuse; 5) first degree relative with a psychiatric illness. Written informed consent was obtained from all healthy subjects under an approved IRB protocol (IRB# 20021714). A subset of 982 (ages 18–84, males 83%) patients with a diagnosis of cannabis use disorder by DSM-IV and DSM-V criteria were identified and compared to a healthy control population (n = 92, age 18–84, 42% male) with perfusion neuroimaging using SPECT. The most common comorbidities in in the cannabis use group were i) attention-deficit/hyperactivity disorder (62%); ii) traumatic brain injury (TBI) (47%); iii) and major depressive disorder (35%).

SPECT neuroimaging acquisition

All subjects received functional perfusion neuroimaging with SPECT as described in previous studies [16, 17]. Briefly, brain SPECT scans were performed using a high resolution Picker (Philips) Prism XP 3000 triple-headed gamma camera (Picker Int. Inc., Ohio Nuclear Medicine Division, Bedford Hills, OH, USA) with low energy high-resolution fan beam collimators. Each participant received an age and weight appropriate dose of technetium–99 m hexamethylpropyleneamine (HMPAO). For rest scans, the injection was done intravenously in a quiet room with low level ambient lighting while awake with eyes open. Brain SPECT scanning was then done 30 minutes after injection. For concentration scans, patients were injected three minutes after starting the Conners Continuous Performance Test (Conners Continuous Performance Test, CCPT-II, Multi-Health Systems, Toronto, Ontario). SPECT scanning was subsequently done 30 minutes after injection.

Data was acquired in 128×128 matrices, resulting in 120 images per scan with each image separated by three degrees spanning 360 degrees. The original image matrix obtained at 128×128×29 with voxel sizes of 2.16 mm×2.16 mm×6.48 mm were transformed and resliced to a 79×95×68 matrix with voxel sizes of 2 mm×2 mm×2 mm consistent with the Montreal Neurological Institute template [18]. Images were smoothed using an 8 mm FWHM isotropic Gaussian kernel with slice thickness of 6 mm. A low pass filter was applied with a high cutoff. Chang attenuation correction was performed [19]. The total number of counts in a given brain SPECT study was 10 million. Transaxial slices oriented horizontal to the AC-PC line were created along with coronal and sagittal images at 6.6 mm apart, unsmoothed.

Perfusion quantification

Regional cerebral blood flow was quantified in all 256 regions (128 baseline, 128 concentration) using a standard neuroanatomical atlas, the AAL atlas [20], as previously described [16, 17]. To summarize, counts in each region of interest were quantified using trimmed means. Trimmed means were calculated using all scores in a 98% confidence interval (–2.58 < Z < –2.58). Perfusion for each region was then estimated with the trimmed mean using the following formula:

T = 10*((subject ROI_mean – trimmed regional_avg)/trimmed regional_stdev)+50.

Predictive data analytics

All statistical analyses were performed in Statistical Package for Social Science (SPSS, version 22, IBM, Armonk, NY) and were controlled for age, gender, and race. To identify specific brain regions that may be predictive in distinguishing marijuana users from controls, a one-way ANOVA was conducted to such areas from the AAL atlas [20] of computed perfusion values [21]. Multiple pairwise comparisons were accounted for with a false discovery rate [22]. We hypothesized that regional cerebral hypoperfusion, including in areas known to be affected by AD such as the hippocampus, would reliably classify marijuana users. For the hippocampus specifically, Cohen’s D effect size calculations were rendered based on mean differences and standarddeviations [23].

Hypoperfused brain areas in the marijuana group were then used to classify this group from controls with discriminant analysis using a leave-one-out cross validation [24]. For this step, baseline perfusion measures were inputted into the discriminant analysis followed by a separate analysis with concentration SPECT regions. This was followed by extraction of predicted probabilities from this model. These probabilities were then used to determine diagnostic sensitivity, specificity, and accuracy with receiver operating characteristic (ROC) and area under the curve (AUC) analyses.

Feature selection using a reduced co-morbidity analysis

Because almost half of the original sample had persons with TBI, we reduced co-morbidity burden by excluding these subjects in this analysis as TBI could potentially mimic hypoperfusion seen in in our marijuana subjects. This resulted in 436 marijuana users compared to 92 controls. A different machine learning algorithm, support vector machine (SVM) was applied [25]. As part of this analysis an oversampling method was applied, k = 0.5. A minimum redundancy maximum relevancy (mRmR) method was then applied to identify the top 10 most predictive SPECT regions (baseline or concentration) [26]. All 256 ROIs, 128 baseline and 128 concentration, were inputted as variables into the SVM mRMR analysis.

RESULTS

All regions with lower perfusion on both baseline and concentration SPECT scans are listed in Table 1. Of the 256 regions assessed, the majority was statistically significant in showing lower perfusion among marijuana users compared to controls (p < 0.05). Hypoperfused regions in marijuana users compared to controls include areas known to be targeted by AD pathology including: i) hippocampus; ii) parahippocampal gyrus; iii) precuneus; iv) posterior cingulate; and v) medial temporal lobes.

The discriminant analysis of 982 marijuana users from 92 controls using baseline SPECT regions yielded a correct classification rate of 96% with the leave-one-out cross validation at 92%. The corresponding accuracy of the linear discriminant probabilities in identifying marijuana users from controls was 95% based upon AUC analysis with 90% sensitivity and 85% specificity (Fig. 1). When using concentration SPECT regions in the discriminant analysis, the results were similar with a correct classification of 95% with a leave-one-out cross validation of 90%. The AUC analysis for concentration SPECT regions from the discriminant analysis was also similar to the baseline regions with 95% accuracy, 90% sensitivity, and 83% specificity.

The top 10 predictive regions from the 256 ROIs (128 baseline and 128 concentration) inputted into the subgroup SVM mRMR analysis are listed in Table 2. Notably, the right hippocampus on concentration SPECT scan is the most predictive region in distinguishing marijuana users from controls. Hippocampal perfusion in marijuana users is 13% lower on concentration scans compared to controls (F = 81.8 for left hippocampus, p < 0.001; F = 99.4 for right hippocampus, p < 0.001) with a large effect size (Cohen’s D = 0.99 for left and 1.03 for right). Baseline hippocampal perfusion was reduced by about 17% in marijuana users compared to controls (F = 114.3, p < 0.001 for left hippocampus; F = 123.9, p < 0.001 for right hippocampus). The results also had large effect size (Cohen’s D = 1.1 for left hippocampus and 1.2 for right hippocampus).

DISCUSSION

Regional, specifically hippocampal, hypoperfusion in marijuana users reliably distinguishes this group from healthy controls. The right hippocampus during concentration task was the single most predictive region in distinguishing marijuana users from their normal counterparts. This finding is important given a prior multi-modal study showing 3% reduced hippocampal volume in cannabis users compared to non-users and increased N-acetylaspartate on magnetic resonance spectroscopy [27]. Another study of 15 chronic cannabis users utilized functional MRI (fMRI) to reveal impaired activation during learning tasks compared to non-users [28]. Individuals who use both marijuana and nicotine also have lower hippocampal volumes and lower immediate/delayed story recall compared to non-users [29]. Additionally, cannabis use is thought to interfere with memory formation by inhibiting long-term potentiation [30].

While the hippocampus is the most predictive region in our study, it is also seen as abnormal in marijuana users based on the literature. Other brain areas are affected as well. A recent systematic review concluded that structural and functional defects in the cerebellum are a common feature in neuroimaging studies of cannabis users [31]. This is also seen with our work given that three of the ten most predictive regions in our results were cerebellar sub-regions. These findings are consistent with descriptions in the literature of coordination deficits in marijuana users [32].

Pallidum hypoperfusion found in our study is corroborated by subcortical white matter abnormalities seen on diffusion tensor imaging data of marijuana users in the Human Connectome Project [33]. An fMRI study also showed attenuation of frontal-subcortical circuits in heavy cannabis users as well as reduced performance on motivation driven tasks [34]. The anterior temporal lobe hypoperfusion detected in our study is also consistent with prior fMRI data showing temporal lobe deactivation as a function of negative emotionality in marijuana users [35].

Several studies of perfusion imaging in marijuana users have shown similar results compared to ours. A small O15 PET study in a sample of 12 marijuana users used a randomized clinical trial design to examine brain perfusion before and after marijuana. The study results found frontal, temporal, and occipital lobe hypoperfusion— all findings concordant with our study [36]. That same study showed increased perfusion in several regions, such as the cerebellum, that we did not observe in our larger sample. Another comparatively recent paper from the same group showed marijuana users recruited the cerebellum in a monetary decision task on O15 PET imaging. [37]. However, a recent study showed reduced FDG-PET glucose metabolism in marijuana users in the frontal lobes, inferior parietal cortex, and fusiform gyrus at baseline [38]. This same study also showed impaired response to methyphenidate stimulant challenge in female cannabis users in the cerebellum, medial frontal gyrus, pons, hippocampus, thalamus, and midbrain. A perfusion MRI study with arterial spin labeling also showed temporal lobe hypoperfusion in 23 adolescent marijuana users compared to 23 age-matched controls [39]. These perfusion differences persisted after four weeks of abstinence suggesting that the effects of marijuana on brain hypoperfusion may be long term.

A trimmed means approach was used to estimate cerebral perfusion from the brain SPECT scans of our study participants. The main advantage of this method is in its robustness, lowering partial volume effects while protecting against outliers [40, 41]. In excluding such information, the resulting dataset may not as closely resemble the original data but this issue is less likely in larger datasets such as in our sample.

There are several implications for this work. First, marijuana users have lower cerebral perfusion than non-users. Second, the most predictive region separating these two groups is hippocampal hypoperfusion on concentration SPECT imaging. This work suggests that marijuana use has potentially deleterious influences in the brain— particularly regions important in memory and learning and known to be affected by AD. The main advantage of our study is the large sample size, application of different algorithms to confirm consistency of results, and subgroup analysis to account for co-morbidities. The main disadvantage of our study is its cross sectional design. An additional caveat to consider is that brain atrophy from marijuana use could be driving the hippocampal hypoperfusion results given its vulnerability to atrophy [42]. Future work combining structural and functional imaging may better elucidate this relationship.

However, such results may give pause to the continued promotion of a substance that may be harmful to users, especially if used for the intention of therapeutic purposes. Longitudinal studies are required at the pre-teen, high school, and college level to better characterize the time course of such alterations. Based upon the results of our study, doubts remain as to the application of marijuana in treatment of AD, as some have suggested [4].

DISCLOSURE STATEMENT

Authors’ disclosures available online (https://www.j-alz.com/manuscript-disclosures/16-0833r3).

References

Hasin

, Saha

, Kerridge

, Goldstein

, Chou

, Zhang

, Jung

, Pickering

, Ruan

, Smith

, Huang

, Grant

(2015) Prevalence of marijuana use disorders in the United States Between 2001-2002 and 2012-2013. JAMA Psychiatry 72, 1235–1242.

Maas

(2016) Traumatic brain injury: Changing concepts and approaches. Chin J Traumatol 19, 3–6.

Grotenhermen

, Muller-Vahl

(2012) The therapeutic potential of cannabis and cannabinoids. Dtsch Arztebl Int 109, 495–501.

Maust

, Bonar

, Ilgen

, Blow

, Kales

(2016) Agitation in Alzheimer disease as a qualifying condition for medical marijuana in the United States. Am J Geriatr Psychiatry 24, 1000–1003.

American Academy of Neurology, Position Statement: Use of Medical Marijuana for Neurologic Disorders., https://www.aan.com/uploadedFiles/Website_Library_Assets/Documents/6.Public_Policy/1.Stay_Informed/2.Position_Statements/3.PDFs_of_all_Position_Statements/Final%20Medical%20Marijuana%20Position%20Statement.pdf, Accessed 7/31-2016.

Antonsdottir

, Makino

, Porsteinsson

(2016) Dazed and confused: Medical cannabis in Alzheimer disease. Am J Geriatr Psychiatry 24, 1004–1006.

Volkow

, Compton

, Weiss

(2014) Adverse health effects of marijuana use. N Engl J Med 371, 879.

Volkow

, Swanson

, Evins

, DeLisi

, Meier

, Gonzalez

, Bloomfield

, Curran

, Baler

(2016) Effects of cannabis use on human behavior, including cognition, motivation, and psychosis: A review. JAMA Psychiatry 73, 292–297.

van de Giessen

, Weinstein

, Cassidy

, Haney

, Dong

, Ghazzaoui

, Ojeil

, Kegeles

, Xu

, Vadhan

, Volkow

, Slifstein

, Abi-Dargham

(2016) Deficits in striatal dopamine release in cannabis dependence. Mol Psychiatry. doi: 10.1038/mp.2016.21 [Epub ahead of print].

10.

Rigucci

, Marques

, Di Forti

, Taylor

, Dell’Acqua

, Mondelli

, Bonaccorso

, Simmons

, David

, Girardi

, Pariante

, Murray

, Dazzan

(2016) Effect of high-potency cannabis on corpus callosum microstructure. Psychol Med 46, 841–854.

11.

Filbey

, Aslan

, Calhoun

, Spence

, Damaraju

, Caprihan

, Segall

(2014) Long-term effects of marijuana use on the brain. Proc Natl Acad Sci U S A 111, 16913–16918.

12.

Currais

, Quehenberger

, Armando

, Daugherty

, Maher

, Schubert

(2016) Amyloid proteotoxicity initiates an inflammatory response blocked by cannabinoids. NPJ Aging Mech Dis 2, 16012.

13.

Ashtari

, Avants

, Cyckowski

, Cervellione

, Roofeh

, Cook

, Gee

, Sevy

, Kumra

(2011) Medial temporal structures and memory functions in adolescents with heavy cannabis use. J Psychiatr Res 45, 1055–1066.

14.

Cousijn

, Wiers

, Ridderinkhof

, van den Brink

, Veltman

, Goudriaan

(2012) Grey matter alterations associated with cannabis use: Results of a VBM study in heavy cannabis users and healthy controls. Neuroimage 59, 3845–3851.

15.

Yucel

, Solowij

, Respondek

, Whittle

, Fornito

, Pantelis

, Lubman

(2008) Regional brain abnormalities associated with long-term heavy cannabis use. Arch Gen Psychiatry 65, 694–701.

16.

Amen

, Raji

, Willeumier

, Taylor

, Tarzwell

, Newberg

, Henderson

(2015) Functional neuroimaging distinguishes posttraumatic stress disorder from traumatic brain injury in focused and large community datasets. PLoS One 10, e0129659.

17.

Raji

, Willeumier

, Taylor

, Tarzwell

, Newberg

, Henderson

, Amen

(2015) Functional neuroimaging with default mode network regions distinguishes PTSD from TBI in a military veteran population. Brain Imaging Behav 9, 527–534.

18.

Mazziotta

, Toga

, Evans

, Fox

, Lancaster

, Zilles

, Woods

, Paus

, Simpson

, Pike

, Holmes

, Collins

, Thompson

, MacDonald

, Iacoboni

, Schormann

, Amunts

, Palomero-Gallagher

, Geyer

, Parsons

, Narr

, Kabani

, Le Goualher

, Boomsma

, Cannon

, Kawashima

, Mazoyer

(2001) A probabilistic atlas and reference system for the human brain: International Consortium for Brain Maing (ICBM). Philos Trans R Soc Lond B Biol Sci 356, 1293–1322.

19.

Chang

, Henkin

, Buddemeyer

(1984) The sources of overestimation in the quantification by SPECT of uptakes in a myocardial phantom: Concise communication. J Nucl Med 25, 788–791.

20.

Tzourio-Mazoyer

, Landeau

, Papathanassiou

, Crivello

, Etard

, Delcroix

, Mazoyer

, Joliot

(2002) Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain. Neuroimage 15, 273–289.

21.

Amen

, Willeumier

, Omalu

, Newberg

, Raghavendra

, Raji

(2016) Perfusion neuroimaging abnormalities alone distinguish National Football League players from a healthy population. J Alzheimers Dis 53, 237–241.

22.

Benjamini

, Hochberg

(1995) Controlling the false discovery rate: A practical and powerful approach to multiple testing. J R Stat Soc Series B Stat Methodol 57, 289–300.

23.

Cohen

(1988) Statistical Power Analysis for the Behavioral Sciences, 2nd Ed., Lawrence Erlbaum Associates, Hillsdale, NJ.

24.

Moller

, Pijnenburg

, van der Flier

, Versteeg

, Tijms

, de Munck

, Hafkemeijer

, Rombouts

, van der Grond

, van Swieten

, Dopper

, Scheltens

, Barkhof

, Vrenken

, Wink

(2015) Alzheimer disease and behavioral variant frontotemporal dementia: Automatic classification based on cortical atrophy for single-subject diagnosis. Radiology 279, 838–848.

25.

Iniesta

, Stahl

, McGuffin

(2016) Machine learning, statistical learning and the future of biological research in psychiatry. Psychol Med 46, 2455–2465.

26.

Ding

, Peng

(2005) Minimum redundancy feature selection from microarray gene expression data. J Bioinform Comput Biol 3, 185–205.

27.

Yucel

, Lorenzetti

, Suo

, Zalesky

, Fornito

, Takagi

, Lubman

, Solowij

(2016) Hippocampal harms, protection and recovery following regular cannabis use. Transl Psychiatry 6, e710.

28.

Carey

, Nestor

, Jones

, Garavan

, Hester

(2015) Impaired learning from errors in cannabis users: Dorsal anterior cingulate cortex and hippocampus hypoactivity. Drug Alcohol Depend 155, 175–182.

29.

Filbey

, McQueeny

, Kadamangudi

, Bice

, Ketcherside

(2015) Combined effects of marijuana and nicotine on memory performance and hippocampal volume. Behav Brain Res 293, 46–53.

30.

Navakkode

, Korte

(2014) Pharmacological activation of CB1 receptor modulates long term potentiation by interfering with protein synthesis. Neuropharmacology 79, 525–533.

31.

Ganzer

, Broning

, Kraft

, Sack

, Thomasius

(2016) Weighing the evidence: A systematic review on long-term neurocognitive effects of cannabis use in abstinent adolescents and adults. Neuropsychol Rev 26, 186–222.

32.

Schrot

, Hubbard

(2016) Cannabinoids: Medical implications. Ann Med 48, 128–141.

33.

Orr

, Paschall

, Banich

(2016) Recreational marijuana use impacts white matter integrity and subcortical (but not cortical) morphometry. Neuroimage Clin 12, 47–56.

34.

Blanco-Hinojo

, Pujol

, Harrison

, Macia

, Batalla

, Nogue

, Torrens

, Farre

, Deus

, Martin-Santos

(2016) Attenuated frontal and sensory inputs to the basal ganglia in cannabis users. Addict Biol. doi: 10.1111/adb.12370 [Epub ahead of print].

35.

Heitzeg

, Cope

, Martz

, Hardee

, Zucker

(2015) Brain activation to negative stimuli mediates a relationship between adolescent marijuana use and later emotional functioning. Dev Cogn Neurosci 16, 71–83.

36.

O’Leary

, Block

, Koeppel

, Flaum

, Schultz

, Andreasen

, Ponto

, Watkins

, Hurtig

, Hichwa

(2002) Effects of smoking marijuana on brain perfusion and cognition. Neuropsychopharmacology 26, 802–816.

37.

Vaidya

, Block

, O’Leary

, Ponto

, Ghoneim

, Bechara

(2012) Effects of chronic marijuana use on brain activity during monetary decision-making. Neuropsychopharmacology 37, 618–629.

38.

Wiers

, Shokri-Kojori

, Wong

, Abi-Dargham

, Demiral

, Tomasi

, Wang

, Volkow

(2016) Cannabis abusers show hypofrontality and blunted brain responses to a stimulant challenge in females but not in males. Neuropsychopharmacology 41, 2596–2605.

39.

Jacobus

, Goldenberg

, Wierenga

, Tolentino

, Liu

, Tapert

(2012) Altered cerebral blood flow and neurocognitive correlates in adolescent cannabis users. Psychopharmacology (Berl) 222, 675–684.

40.

(2013) Functional Magnetic Resonance Imaging Processing, Springer, Berlin.

41.

Tohka

, Zijdenbos

, Evans

(2004) Fast and robust parameter estimation for statistical partial volume models in brain MRI. Neuroimage 23, 84–97.

42.

Lorenzetti

, Solowij

, Whittle

, Fornito

, Lubman

, Pantelis

, Yucel

(2015) Gross morphological brain changes with chronic, heavy cannabis use. Br J Psychiatry 206, 77–78.