The effects of cannabis abstinence on cognition and resting state network activity in people with multiple sclerosis: A preliminary study
aSunnybrook Research Institute, Division of Psychiatry, University of Toronto, Toronto, Ontario, Canada
bClinical Laboratory and Diagnostic Services, Centre for Addiction and Mental Health, Toronto, Ontario, Canada
cDepartment of Kinesiology and Health Sciences, University of Waterloo, Waterloo, Ontario, Canada
Highlights
- •Cannabis use may exacerbate cognitive impairment in MS patients, whereas abstinence can result in improved brain activation and cognitive performance.
- •Cannabis-related cognitive impairment in MS may disrupt the default mode network.
- •Abstinence from cannabis was linked to better cognitive outcomes but did not show a significant impact on depression, anxiety, or fatigue.
Untitled section
Keywords: Multiple sclerosis, Resting state, Cognitive impairment, Cannabis
Abstract
We previously reported that people with multiple sclerosis (pwMS) who have been using cannabis frequently over many years can have significant cognitive improvements accompanied by concomitant task-specific changes in brain activation following 28 days of cannabis abstinence. We now hypothesize that the default Mode Network (DMN), known to modulate cognition, would also show an improved pattern of activation align with cognitive improvement following 28 days of drug abstinence. Thirty three cognitively impaired pwMS who were frequent cannabis users underwent a neuropsychological assessment and fMRI at baseline. Individuals were then assigned to a cannabis continuation (CC, n = 15) or withdrawal (CW, n = 18) group and the cognitive and imaging assessments were repeated after 28 days. Compliance with cannabis withdrawal was checked with regular urine monitoring. Following acquisition of resting state fMRI (rs-fMRI), data were processed using independent component analysis (ICA) to identify the DMN spatial map. Between and within group analyses were carried out using dual regression for voxel-wise comparisons of the DMN. Clusters of voxels were considered statistically significant if they survived threshold-free cluster enhancement (TFCE) correction at p < 0.05. The two groups were well matched demographically and neurologically at baseline. The dual regression analysis revealed no between group differences at baseline in the DMN. By day 28, the CW group in comparison to the CC group had increased activation in the left posterior cingulate, and right, angular gyrus (p < 0.05 for both, TFCE). A within group analysis for the CC group revealed no changes in resting state (RS) networks. Within group analysis of the CW group revealed increased activation at day 28 versus baseline in the left posterior cingulate, right angular gyrus, left hippocampus (BA 36), and the right medial prefrontal cortex (p < 0.05). The CW group showed significant improvements in multiple cognitive domains. In summary, our study revealed that abstaining from cannabis for 28 days reverses activation of DMN activity in pwMS in association with improved cognition across several domains.
Article notes
Untitled section
Received 2023 Nov 15; Revised 2024 May 1; Accepted 2024 May 23; Collection date 2024.
1.Introduction
Cognitive impairment (CI) occurs in 34 % to 91 % of people with MS (pwMS) according to disease type (Ruano et al., 2016), with impaired information processing speed (IPS) being the most common deficit (Benedict et al., 2020). The effects of cognitive decline correlate with reduced quality of life and activities of daily living (Gil-González et al., 2020, Dehghani et al., 2019). Emerging evidence reveals that smoking or vaping cannabis may further hinder cognition (Patel & Feinstein, 2017), a finding given added salience by data showing that cannabis use among pwMS has more than doubled in the last decade (Link et al., 2023).
Recent MRI studies show that pwMS who use cannabis have greater structural morphological alterations (Romero et al., 2015) and maladaptive compensatory circuitry in the brain in comparison to pwMS who are non-cannabis users (Pavisian et al., 2014). Furthermore, a dysfunctional pattern of cerebral activity is observed during IPS tasks, which is reversible following abstinence from cannabis (Feinstein et al., 2019). The role of resting state in changes like these is unknown and deserving of further inquiry, more so as cannabis use in healthy individuals is associated with abnormal functional activation and connectivity (Lorenzetti et al., 2023). Adding to the complexity of these potentially inter-related factors, cognitive impairment in pwMS by itself is also associated with altered resting state functional connectivity (Jandric et al., 2021). In particular, the default mode network has emerged as a prominent region implicated in cognitive processing in cognitively compromised pwMS (Rocca et al., 2022).
No study to date has investigated potential resting state changes linked to cannabis use in pwMS. Given the deleterious effect of cannabis on cognition in this population, a better understanding of whether cannabis further alters cognitive specific resting state networks is warranted. This forms the primary focus of our report in which we hypothesize that coming off cannabis will lead to beneficial changes in the default mode network in keeping with improvements in cognitive performance.
2.Materials and methods
2.1.Subjects
Details of study participants have been reported previously (Feinstein et al., 2019). To summarize, all participants had to have a confirmed diagnosis of MS according to revised McDonald criteria (Thompson et al., 2018), to have begun smoking cannabis after being diagnosed with MS and to have global cognitive impairment. Reason for only including cognitively impaired individuals was to see whether cognition could improve in pwMS following cannabis abstinence. Exclusion criteria included another disease of the CNS, an inability to give consent because of an intellectual disability, steroid treatment in the last three months, visual acuity of less than 20/70 and a positive urine test for an illicit drug.
Resting state fMRI data were available in 33 participants who had been assigned to the cannabis withdrawal (CW, n = 18) and cannabis continuation (CC, n = 15) groups according to odd and even case numbers. Imaging data from three people in the CC group were degraded preventing analysis, hence the smaller sample size. The tester was blinded to group allocation prior to the index assessment. A cognitive battery was completed at baseline and day 28. Structural and functional MRI data, including metrics pertaining to resting state were obtained at both time points as well. The follow-up period was determined by the pharmacokinetics of cannabis which reveals that excretion of drug in our study sample is generally complete 28 days after last use (Feinstein et al., 2019).
2.2.Data were collected on the following
Demographic and neurological data including EDSS, disease course and duration.
Cannabis use history including frequency and duration of use.
Urine screening at baseline and follow up to check compliance with abstinence. This was done for the two main cannabinoids, delta-9 −tetrahydrocannabinol (THC) and cannabidiol. Ratios of the two main cannabinoids to creatinine were obtained to control for alterations in fluid intake that can affect metabolite concentrations in the urine (Huestis and Cone, 1998). Subjects were asked to refrain from using cannabis for at least 12 h prior to cognitive testing to avoid assessing individuals who were acutely intoxicated. An objective marker of cannabis abstinence for up to 6 h was obtained from a saliva sample using NarcoCheck® (Kappa City Bio Tech, 2022) prior to cognitive testing at baseline and Day 28 in all subjects.
Cognitive data were obtained with the Brief Repeatable Neuropsychological Battery for MS (BRNB) that assesses verbal (Selective Reminding Test) and visual (10/36 Test) memory; processing speed (Symbol Digit Modalities Test (SDMT), 2 and 3 s Paced Auditory Serial Addition Task (PASAT)), verbal fluency and executive functioning (Controlled Oral word Association Test) (Rao et al., 1991). Premorbid IQ was measured with the Wechsler Test of Adult Reading (Wechsler, 2001). Impairment on individual tests was defined as a score of 1.5 SD below age, sex and education matched normative data while global impairment was defined, by convention, as failure on two or more of the five cognitive indices comprising the BRNB. (Rao et al., 1991). Normative values were derived from studies by Boringa et al. (2001) and Walker et al. (2017).
Symptoms of depression and anxiety were recorded with the Hospital Anxiety and Depression Scale (HADS) (Zigmond & Snaith, 1983).
2.3.Consent and ethics approval
Informed consent was obtained from all participants in accordance with the Declaration of Helsinki (World Medical Association, 2013). The study was approved by the Research and Ethics Board at Sunnybrook Health Sciences Centre, affiliated with the University of Toronto.
2.4.MRI scanning parameters
The Siemens Prisma 3 T system (Erlangen, Germany) was used to collect MRI images, using a 32-channel head coil at index assessment and day 28. High resolution anatomical brain scans were acquired using a sagittal 3D T1 MPRAGE sequence with the following parameters: echo time (TE) = 1.94 ms, repetition time (TR) = 2300 ms, inversion time (TI) = 900 ms, flip angle = 9 degrees, GRAPPA x2, field of view (FoV) = 256 mm × 248 mm, 192 slices 1 mm thick, matrix = 256 × 248 (1.0 mm × 1.0 mm × 1.0 mm voxels).
2.5.Resting state acquisition
Resting state (eyes open) functional MRI data were acquired using T2-weighted axial gradient echo planar imaging (EPI), TE = 30 ms, TR = 2000 ms, flip angle = 40 deg, FoV = 204 mm × 204 mm, 32 slices 3.5 mm thick, 1.0 mm gap, matrix = 68 × 68 (3.0 mm × 3.0 mm × 3.5 mm voxels), scan time 9:20 min to obtain blood oxygen level dependent (BOLD) contrast. In total 280 volumes were obtained.
2.6.Fluid attenuated inversion recovery acquisition
The FLAIR images were acquired using the following parameters for axial PD/T2 TSE sequence: TE = 11 ms/95 ms, TR = 3050 ms, FA = 165 deg, GRAPPA x2, FoV = 224 mm × 189 mm, 48 slices with a thickness of 3.0 mm, matrix = 256 × 216 (0.875 mm × 0.875 × 3.0 mm voxels), with a scan time of 2:34 min.
2.7.Resting state fMRI pre-processing
Resting-state analysis was performed using Independent Component Analysis (ICA) using the multisession temporal concatenation tool (Beckmann and Smith, 2005) in MELODIC (Multivariate Exploratory Linear Decomposition into Independent Components) version 3.14, FSL (FMRIB's Software Library https://www.fmrib.ox.ac.uk/fsl). The T1 image files were reoriented to match the orientation of the standard Montreal Neurological Institute (MNI) 152 2-mm brain template, using the fslorient tool. Followed by BET (brain extraction tool) which was used to delete non-brain tissue from an image of the whole head (Smith, 2002). Registration to high-resolution structural and standard-space images was performed using FLIRT (FMRIB's Linear Image Registration Tool). The functional data were registered first to each subject's T1 structural images and then into the MNI 2-mm standard space. Parameters included high-pass filter cutoff of 100 (seconds), MCFLIRT motion correction with a slice timing correction, spatial smoothing with a full-width half-maximum value of 5 mm. By default, Melodic will automatically estimate the number of components from the data (Beckmann, 2012). Using FSLeyes, all the ICA components were visually examined and compared with those reported in the resting state literature (Beckmann and Smith, 2005). The DMN ICA component along with our covariates was introduced into a dual regression for between and within group analysis. The dual regression algorithm allows for voxel-wise comparisons of resting-state fMRI (Beckmann et al., 2009), correcting for multiple comparisons at p < 0.05 using threshold-free cluster enhancement. After the dual regression, we created masks for the DMN regions that survived at p < 0.05, for both the between and within group, in order to avoid whole brain analysis. These masks were created using a non-probabilistic atlas (JuelichAtlas.nii.gz); the fslmaths command was used to binarize masks and fsl-randomise command to apply the mask. We utilized the FSL cluster command to extract cluster details. The cluster intensity output obtained was utilized in our correlation analysis. Additionally, the number of voxels and cluster coordinates were recorded in Table 3, Table 4.
| DMN | Number of Voxels | TAL X | TAL Y | TAL Z | *TFCE |
|---|---|---|---|---|---|
| Left, Posterior Cingulate (BA 23) | 652 | 49 | 36 | 43 | P < 0.05 |
| Right, Angular Gyrus, Inferior parietal lobe (BA 39) | 325 | 22 | 31 | 54 | P < 0.05 |
| DMN | Number of Voxels | TAL X | TAL Y | TAL Z | *TFCE |
|---|---|---|---|---|---|
| Left, Hippocampus (BA 36) | 528 | −28 | −20 | −9 | P < 0.05 |
| Right, Medial prefrontal cortex (BA 11) | 338 | 0 | 42 | −19 | P < 0.05 |
2.8.Structural acquisition
Total hyper-intense lesions volumes were obtained using a SPM segmentation tool, (Schmidt et al., 2013) and hypointense lesion volumes were obtained using Lesion Mapper, a FSL script specifically designed for multiple sclerosis hypo-intense lesion analysis (Wetter et al., 2016). Cortical reconstruction and volumetric segmentation were performed with the Freesurfer image analysis suite, which is available for download online (https://surfer.nmr.mgh.harvard.edu/). The technical details of these procedures are described in a prior publication (Dale et al., 1999).
2.9.Statistical analysis of demographic, cognitive and behavioral data
Data analysis was carried out on IBM SPSS Statistics for Windows, version 29.0 (Armonk, NY). The Shapiro-Wilk Test was used to test for normality of data distribution. For non-parametric data we used a Mann-Whitney Test. To determine between group differences in our demographic variables, we employed a two-tailed t test for continuous data and chi square analysis for ordinal data where appropriate. To determine group difference in our cognitive and structural data, we utilized ANCOVA to control for covariates. For within group analysis, we utilized a paired t-test as all data were normally distributed. To determine the strength, and direction between cognitive measures and DMN activation, a partial correlation was undertaken as this allowed us to control for the potential influence of covariates, namely EDSS, sex, disease duration, duration of cannabis use, THC-ratio and CBD ratio at baseline. Prior to running this analysis, Spearman rank correlations between the covariates were assessed to avoid entering pairs of highly correlated variables into the analysis. Given that disease duration and duration of cannabis use correlated significantly (r = 0.775; p = 0.0001), disease duration was omitted from the partial correlation as the between group difference was not significant for this variable compared to the duration of cannabis use (see Table 1). Additionally, to determine which covariate influenced cognitive results at the Day 28 assessment, we employed a multiple regression analysis.
| Cannabis Continuation group | Cannabis Withdrawal group | |||||
|---|---|---|---|---|---|---|
| Mean/frequency (N = 15) | SD | Mean/frequency (N = 18) | SD | t-test/x(2) | Sig. | |
| Age (yrs.) | 38.53 | 8.51 | 36.67 | 11.89 | t = 0.51 | 0.62 |
| Sex (% female) | 7 (53.33 %) | 7 (61.11 %) | X2 = 0.20 | 0.65 | ||
| *EDSS | 2.9 | 1.71 | 2.25 | 1.91 | t = 0.97 | 0.34 |
| Disease Course, n (%) | ||||||
| RRMS | 12 (80.0 %) | 13 (72.2 %) | X2 = 0.27 | 0.60 | ||
| SPMS | 3 (20.0 %) | 5 (27.8 %) | − | − | ||
| Disease Duration (yrs.) | 11.27 | 6.36 | 8.78 | 8.01 | t = 0.97 | 0.34 |
| Marital status (% married) | 11 (26.7 %) | 11 (38.9 %) | X2 = 0.55 | 0.46 | ||
| Employment (% unemployed) | 5 (66.7 %) | 8 (55.6 %) | X2 = 0.42 | 0.52 | ||
| Education | 14.27 | 1.87 | 14.50 | 1.65 | t = −0.38 | 0.71 |
| Daily cannabis use, n (%) | 14 (93.3 %) | 17 (94.4 %) | X2 = 0.02 | 0.89 | ||
| Cannabis use in a week (in grams) | 17.39 | 10.18 | 14.76 | 8.96 | t = 0.78 | 0.44 |
| Duration of cannabis use (yrs.) | 10.22 | 6.03 | 5.76 | 5.21 | t = 2.28 | 0.03 |
| Disease modifying drug (% Yes) | 11 (26.7 %) | 9 (50.0 %) | X2 = 1.87 | 0.17 | ||
3.Results
3.2.Cognitive results
There were no between group differences at baseline assessment on any of the individual cognitive indices. Following 28 days of cannabis abstinence, the CW groups had statistically significantly better cognitive results across all indices than the CC group (see Table 2A).
A within group analysis for the CC group revealed no statistically significant differences, apart from visuospatial memory getting worse at day 28 [10/36 spatial recall (t = 4.49, p = 0.001)]. A within group analysis for the CW group revealed statistically significant improvements at day 28 in comparison to baseline across all cognitive indices (see Table 2B).
| Baseline | Vs | Day 28 | ||||
|---|---|---|---|---|---|---|
| mean | SD | mean | SD | Paired t-test | Sig. | |
| Cannabis Continuation group | ||||||
| SRT-LTS | 30.80 | 11.40 | 28.33 | 10.85 | t = 1.38 | 0.19 |
| 10/36 | 18.33 | 6.15 | 13.80 | 5.55 | t = 4.49 | 0.001 |
| PASAT3 | 32.87 | 6.62 | 31.60 | 6.47 | t = 0.82 | 0.43 |
| PASAT2 | 23.13 | 4.66 | 22.47 | 5.11 | t = 0.36 | 0.72 |
| SDMT | 36.40 | 7.77 | 36.53 | 7.09 | t = −0.07 | 0.94 |
| COWAT | 36.60 | 5.10 | 37.13 | 5.30 | t = −0.49 | 0.63 |
| CWS | 41.33 | 33.69 | 40.80 | 35.05 | t = 0.09 | 0.93 |
| HADS-A | 8.07 | 5.13 | 8.53 | 4.07 | t = −0.54 | 0.60 |
| HADS-D | 6.87 | 4.26 | 6.93 | 4.45 | t = −0.12 | 0.91 |
| MFIS | 44.00 | 20.60 | 40.40 | 21.43 | t = 0.89 | 0.39 |
| THC ratio | 124.27 | 101.22 | 137.96 | 131.20 | t = −0.27 | 0.79 |
| CBD ratio | 2.48 | 4.77 | 2.66 | 5.99 | t = −0.31 | 0.76 |
| Cannabis Withdrawal group | ||||||
| SRT-LTS | 29.56 | 13.44 | 44.67 | 8.77 | t = −5.52 | 0.000 |
| 10/36 | 18.33 | 6.14 | 24.78 | 3.19 | t = −4.40 | 0.000 |
| PASAT3 | 36.00 | 10.24 | 48.22 | 7.54 | t = −6.59 | 0.000 |
| PASAT2 | 25.78 | 9.00 | 37.78 | 6.41 | t = −6.97 | 0.000 |
| SDMT | 35.33 | 12.21 | 51.94 | 11.26 | t = −7.46 | 0.000 |
| COWAT | 39.44 | 6.06 | 45.00 | 6.72 | t = −4.54 | 0.000 |
| CWS | 41.33 | 28.26 | 64.06 | 34.03 | t = −2.99 | 0.008 |
| HADS-A | 8.33 | 4.74 | 7.56 | 4.08 | t = 0.78 | 0.447 |
| HADS-D | 7.06 | 4.70 | 5.39 | 3.07 | t = 1.93 | 0.070 |
| MFIS | 37.72 | 16.84 | 37.33 | 15.71 | t = 0.13 | 0.899 |
| THC ratio | 72.48 | 56.89 | 3.10 | 3.15 | t = 5.26 | 0.001 |
| CBD ratio | 0.07 | 0.20 | 0.00 | 0.00 | t = 1.46 | 0.162 |
Before conducting a multiple regression analysis, we conducted a Spearman rank correlation between the covariates, as outlined in our statistical analysis section. Following this, upon identifying any overlaps and eliminating irrelevant covariates, we proceeded with the multiple regression analysis to ascertain which covariate influenced cognitive outcomes at the post-day 28 assessment. Our results revealed that group assignment alone was independently predictive of all cognitive indices at day 28 [group assignment (t = 3.86, p = 0.001); duration of cannabis use (t = −0.35, p = 0.73); THC- creatinine (t = 0.87, p = 0.39); CBD-creatinine (t = −0.70, p = 0.49)].
3.3.Neuropsychiatric symptoms
At baseline there were no between group differences for depression, anxiety, fatigue or cannabis withdrawal symptoms (CWS). Post 28 days, there were no between groups differences for depression, anxiety, or fatigue. Cannabis withdrawal symptoms were significantly higher in the CW group compared to the CC group at Day 28 [CW mean = 64.06 (SD = 34.03); CC mean 40.80 (SD = 35.05); p < 0.01].
A within group analysis for the CC group revealed no changes for anxiety, depression, fatigue or cannabis withdrawal symptoms. A within group analysis for the CW group revealed no statistically significant changes for anxiety, depression and fatigue. Cannabis withdrawal symptoms were found to be significant higher following 28 days of abstinence in the CW group (t = −2.99, p = 0.008).
3.4.FSL MELODIC output
FSL MELODIC generated 112 independent components, one of which (ICA 16) showed a spatial map consistent with the Default Mode Network (DMN) (see Fig. 1). The group ICA DMN was composed of the following brain regions; right and left posterior cingulate cortex (BA 23), right and left anterior cingulate (BA 24), right medial prefrontal cortex (BA 11), right and left angular gyrus/ inferior parietal lobe (BA 39), right and left precuneus/ superior parietal lobe (BA 7), right and left temporal lobe (BA 21), and right and left hippocampal formation (BA 36).
3.5.FSL dual regression
The dual regression analysis revealed no between group differences at index assessment for the default mode network. At day 28, the CW group compared to the CC group had a statistically significant increase in activation in the left precuneus/ posterior cingulate (CW, MNI coordinates X = −8, Y = −56, Z = 12) (p < 0.05, TFCE corrected) and right angular gyrus (CW, MNI X = 50, Y = −68, Z = 34) (p < 0.05, TFCE corrected) of the DMN (see Table 3).
The within group analysis for the CC group, revealed no changes in the DMN at day 28 in comparison to baseline activation. A within group analysis of the CW group revealed a statistically significant increase in activation at day 28 in comparison to baseline for the left hippocampus (MNI, X = −29, Y = −19, Z = −15), and the right medial prefrontal cortex (right, MNI, X = 0, Y = 48, Z = −18), in addition to in the left posterior cingulate and right angular gyrus (p < 0.05, TFCE corrected) (see Table 4, and Fig. 2).
3.7.Structural results
There were no structural differences (T2 lesion load, total grey and white matter volumes) at baseline nor at day 28 for the CC group or the CW group (see Table 5).
| CC group Mean (SD) | CW group Mean (SD) | F | Significance (2-tailed) | |
|---|---|---|---|---|
| Baseline | n = 15 | n = 18 | ||
| Total lesion volume | 0.0063 (0.0047) | 0.0066 (0.0067) | 0.102 | 0.752 |
| Total white matter volume | 0.2956 (0.0294) | 0.2943 (0.0413) | 0.114 | 0.738 |
| Total grey matter volume | 0.3984 (0.0255) | 0.3970 (0.0361) | 0.200 | 0.658 |
| Day 28 | ||||
| Total lesion volume | 0.0064 (0.0048) | 0.0064 (0.0066) | 0.213 | 0.648 |
| Total white matter volume | 0.2937 (0.0300) | 0.2924 (0.0292) | 0.193 | 0.663 |
| Total grey matter volume | 0.3956 (0.0264) | 0.4053 (0.0301) | 0.205 | 0.164 |
Our within group analysis revealed no structural differences for either the CC group or the CW group (see Table 6).
| Baseline | Day 28 | Paired t-test | Significance (2-tailed) | |
|---|---|---|---|---|
| CC group Mean (SD), n = 15 | ||||
| Total Lesion volume | 0.0063 (0.0047) | 0.0064 (0.0048) | 0.219 | 0.438 |
| Total white matter volume | 0.2956 (0.0294) | 0.2937 (0.0300) | 0.161 | 0.322 |
| Total grey matter volume | 0.3984 (0.0255) | 0.3956 (0.0264) | 0.075 | 0.150 |
| CW group Mean (SD), n = 18 | ||||
| Total Lesion Volumes | 0.0066 (0.0067) | 0.0064 (0.0066) | 0.117 | 0.235 |
| Total white matter volume | 0.2943 (0.0413) | 0.2924 (0.0292) | 0.368 | 0.737 |
| Total grey matter volume | 0.3970 (0.0361) | 0.4053 (0.0301) | 0.063 | 0.126 |
4.Discussion
In this study we investigated the relationship between cannabis use, cognition and resting state activity in people with multiple sclerosis. In keeping with our hypothesis, we demonstrated that abstaining from cannabis for a minimum of 28 days increased cerebral activation in the DMN, specifically within the precuneus/ posterior cingulate cortex and angular gyrus, and was associated with improvements in cognitive performance. Conversely, the CC group did not show any significant longitudinal changes in resting state DMN and there was no improvement in cognitive deficits over time. These results align with our previous work that had two significant main findings: First, abstaining from cannabis significantly improved learning and memory, processing speed, executive function, and verbal fluency. Second, performance on an fMRI compatible Symbol Digit Modalities Test following 28 days of cannabis abstinence confirmed the evidence for cognitive improvement and linked this to increased cerebral activation in a well described SDMT-specific neural network (Feinstein et al., 2019).
It is widely accepted that cognitive deficits are a common feature of multiple sclerosis and worsen over time (Brochet et al., 2022). Recent imaging literature suggests that thalamic atrophy, total lesion volume and a dysfunctional pattern of cerebral activity in pwMS are closely tied to cognitive deficits (Benedict et al., 2020). When it comes to the mechanistic understanding of how cannabis may further hinder cognition in people with MS, the literature is sparse (Landrigan et al., 2022). What data there are suggest that cannabis further disrupts cerebral activation during cognitive tasks, most notably when it comes to processing speed, a change that is reversible following abstinence from the drug (Feinstein et al., 2019). The role of resting state networks in changes like these was not known and deserving of further inquiry given that cannabis use in healthy individuals is consistently associated with altered neural activity, particularly in both the frontal parietal network and DMN regions, important for decision making and planning (Thomson et al., 2022). Our data replicates this finding by further implicating the DMN in pwMS.
A small literature focusing on resting state, multiple sclerosis and cognition indicates that the DMN plays a crucial role in cognitive processing (Rocca et al., 2022). Major DMN hubs identified include the posterior cingulate cortex, precuneus, medial prefrontal cortex, and the angular gyrus (Rocca et al., 2010, Bonavita et al., 2011). Our study confirms the importance of three of these regions, namely the precuneus, posterior cingulate cortex and angular gyrus in relation to the cannabis mediated effects of cognition. In particular, cannabis appears to reduce activation in the DMN (causing hypo-activation), an effect that underpins cognitive dysfunction encompassing processing speed, memory and executive function and which is alleviated with abstinence leading to concomitant cognitive improvement across these domains. Given or limited sample size, we focused on the default mode network but recognize that other networks may be implicated as well.
Our resting state data also complement the results of an earlier report − incorporating the current participants as part of a larger sample in which we demonstrated that cannabis reduces cerebral activation during the completion of the SDMT leading to a more impaired score relative to individuals who had come off the drug (Feinstein et al., 2019). The deleterious effects of cannabis on processing speed were shown to be reversible and driven by improved functional connectivity evident 28 days following cessation of drug use. This pattern of longitudinal change observed in the SDMT network is now matched by the serial resting state DMN changes. The composite functional imaging timeline is therefore one of cannabis reducing activation the DMN resting state with further evidence of network reduction when it comes to completing a processing speed task, followed by increased DMN activation and a return of task specific network functional connectivity with abstinence, these serial changes in turn linked to improved processing speed.
One plausible mechanistic explanation for our findings is that using and abstaining from cannabis can influence brain perfusion which in turn could result in improved cognitive scores and RS signal. The literature however is mixed on whether perfusion increases or decreases. For example, looking beyond MS, a recent study evaluated the effects of heavy cannabis use on cognition in a group of 982 individuals with cannabis use disorder compared to a control group of healthy non-cannabis users. They found that cannabis users exhibited lower cerebral perfusion in the hippocampus which was associated with poorer immediate and delayed story recall (Amen et al., 2017). On the other hand, a different study of 74 chronic cannabis users and 101 non-users reported increased perfusion indices (global and regional resting cerebral blood flow, oxygen extraction fraction and cerebral metabolic rate of oxygen) after 72 h of abstinence. The authors postulated that the brief period of abstinence may have led to a rebound overshoot in brain perfusion in the cannabis group that could account for the differences with the non-user group (Filbey et al., 2017).
A further intriguing observation arises from our cannabis depression data. The cessation of cannabis usage coincided with lower HADS-D scores, although it did not quite reach statistical significance, potentially due to our small sample size, which might have led to a Type 2 error. Our longitudinal findings partially coincide with cross-sectional data from a study involving 16 depressed individuals with MS and a control group of 17 non-depressed pwMS. In this study, symptoms of depression were linked to decreased functional connectivity in the DMN, particularly within the anterior and posterior cingulate cortices. These cerebral changes were part of a broader pattern of reduced resting-state activity associated with cognitive impairment as well (Bonavita et al., 2017).
It is unclear to what degree our resting state DMN findings complement or differ from what would be seen in healthy subjects who smoke cannabis. Comparisons are hindered by few studies and samples limited to adolescents and young adults. A study following adolescent and young adults who were regular cannabis users revealed abnormal connectivity within the dorsal and attentional network, even after two weeks of monitored abstinence (Harris et al., 2022). Another study in a sample of adults between the ages of 16–26 years who were regular cannabis users, revealed weaker connectivity between the left posterior cingulate cortex and various DMN nodes when compared to a matched group of non-cannabis users (Ritchay et al., 2021).
Our study has certain limitations. Our sample size was small. We could not control for the toxicity of the cannabis THC strains nor the amount of cannabis used during the 28 day follow-up period by the CC group. Our sample also comprised mostly younger, mildly disabled individuals with relapsing-remitting MS and as such, our results are not necessarily applicable to a broader range of pwMS. Additionally, we did not include cognitively preserved participants in our study, which meant we do not know whether imaging metrics can also improve with abstinence in this group. Finally, there are no data in the MS-cognition-cannabis-related imaging literature implicating sex differences. For this reason, we did not perform sex related analysis according to the SAGER guidelines.
In conclusion, our study adds to a growing literature identifying putative biomarkers of cognitive decline that are sensitive to cannabis exposure in pwMS. The finding that frequent and long term cannabis use reduces activation in the DMN with associated decline on an array of cognitive variables adds to concerns about the deleterious effects of the drug in individuals who are already vulnerable to cognitive decline as a consequence of their disease. The potential reversibility of these findings with abstinence from the drug, while noteworthy, should not diminish these concerns, for we have shown previously that pwMS will continue to use cannabis even when confronted with data revealing how their individual cognitive results are negatively impacted by their continuing usage (Feinstein et al., 2021). Our results, while interesting, should be considered preliminary given sample size concerns. Future research with a larger sample is therefore required for replication.
Funding
This study was supported by a grant from the Multiple Sclerosis Society of Canada (EGID: 2626).
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Footnotes
Footnote Group
Appendix A.Supplementary data
The following are the Supplementary data to this article:
Data availability
Data will be made available on request.
References
Untitled section
References
- Amen D.G., Darmal B., Raji C.A., Bao W., Jorandby L., Meysami S., Raghavendra C.S. Discriminative properties of hippocampal hypoperfusion in marijuana users compared to healthy controls: Implications for marijuana administration in alzheimer’s dementia. J. Alzheimer’s Disease. 2017;56(1):261–273. doi: 10.3233/jad-160833.
- Beckmann C.F. Modelling with independent components. Neuroimage. 2012;62(2):891–901. doi: 10.1016/j.neuroimage.2012.02.020.
- Beckmann C., Mackay C., Filippini N., Smith S. Group comparison of resting-state FMRI data using multi-subject Ica and dual regression. Neuroimage. 2009;47
- Beckmann C.F., Smith S.M. Tensorial extensions of independent component analysis for multisubject FMRI analysis. Neuroimage. 2005;25(1):294–311. doi: 10.1016/j.neuroimage.2004.10.043.
- Benedict R.H., Amato M.P., DeLuca J., Geurts J.J. Cognitive impairment in multiple sclerosis: Clinical Management, MRI, and Therapeutic Avenues. Lancet Neurol. 2020;19(10):860–871. doi: 10.1016/s1474-4422(20)30277-5.
- Bonavita S., Gallo A., Sacco R., Corte M.D., Bisecco A., Docimo R., Lavorgna L., Corbo D., Costanzo A.D., Tortora F., Cirillo M., Esposito F., Tedeschi G. Distributed changes in default-mode resting-state connectivity in multiple sclerosis. Mult. Scler. J. 2011;17(4):411–422. doi: 10.1177/1352458510394609.
- Bonavita S., Sacco R., Esposito S., d’Ambrosio A., Della Corte M., Corbo D., Docimo R., Gallo A., Lavorgna L., Cirillo M., Bisecco A., Esposito F., Tedeschi G. Default mode network changes in multiple sclerosis: A link between Depression and cognitive impairment? Eur. J. Neurol. 2017;24(1):27–36. doi: 10.1111/ene.13112.
- Brochet B., Clavelou P., Defer G., De Seze J., Louapre C., Magnin E., Ruet A., Thomas-Anterion C., Vermersch P. Cognitive impairment in secondary progressive multiple sclerosis: Effect of disease duration, age, and Progressive Phenotype. Brain Sci. 2022;12(2):183. doi: 10.3390/brainsci12020183.
- Dale A.M., Fischl B., Sereno M.I. Cortical surface-based analysis: I: Segmentation and Surface Reconstruction. Neuroimage. 1999;9(2):179–194. doi: 10.1006/nimg.1998.0395.
- Dehghani A., Khoramkish M., Shahsavari Isfahani S. Challenges in the Daily Living Activities of Patients with Multiple Sclerosis: A Qualitative Content Analysis. Int. J. Commun. Based Nurs. Midwifery. 2019;7(3):201–210. doi: 10.30476/ijcbnm.2019.44995.
- Feinstein A., Meza C., Stefan C., Staines R.W. Coming off cannabis: A cognitive and magnetic resonance imaging study in patients with multiple sclerosis. Brain. 2019;142(9):2800–2812. doi: 10.1093/brain/awz213.
- Feinstein A., Meza C., Stefan C., Staines W.R. Impaired awareness: Why people with multiple sclerosis continue using cannabis despite evidence to the contrary. Brain Behavior. 2021;11(8) doi: 10.1002/brb3.2220.
- Filbey F.M., Aslan S., Lu H., Peng S.-L. Residual effects of THC via novel measures of brain perfusion and metabolism in a large group of chronic cannabis users. Neuropsychopharmacology. 2017;43(4):700–707. doi: 10.1038/npp.2017.44.
- Gil-González I., Martín-Rodríguez A., Conrad R., Pérez-San-Gregorio M.Á. Quality of life in adults with multiple sclerosis: A systematic review. BMJ Open. 2020;10(11) doi: 10.1136/bmjopen-2020-041249.
- Harris J.C., Wallace A.L., Thomas A.M., Wirtz H.G., Kaiver C.M., Lisdahl K.M. Disrupted resting state attentional network connectivity in adolescent and young adult cannabis users following two-weeks of monitored abstinence. Brain Sci. 2022;12(2):287. doi: 10.3390/brainsci12020287.
- Huestis M.A., Cone E.J. Differentiating new marijuana use from residual drug excretion in occasional marijuana users. J. Anal. Toxicol. 1998;22(6):445–454. doi: 10.1093/jat/22.6.445.
- Jandric D., Doshi A., Scott R., Paling D., Rog D., Chataway J., Schoonheim M.M., Parker G., Muhlert N. A systematic review of resting-state functional MRI connectivity changes and cognitive impairment in multiple sclerosis. Brain Connect. 2021 doi: 10.1089/brain.2021.0104.
- Kappa City Biotech. (2022, April 21). TESTS SALIVAIRES. NarcoCheck. Retrieved April 1, 2024, from https://www.narcocheck.com/en/saliva-drug-tests.
- Landrigan J., Bessenyei K., Leitner D., Yakovenko I., Fisk J.D., Prentice J.L. A systematic review of the effects of cannabis on cognition in people with multiple sclerosis. Mult. Scler. Relat. Disord. 2022;57 doi: 10.1016/j.msard.2021.103338.
- Link K., Knowles L.M., Alschuler K.N., Ehde D.M. Characterizing cannabis use in a sample of adults with multiple sclerosis and chronic pain: An observational study. Mult. Scler. Relat. Disord. 2023;75 doi: 10.1016/j.msard.2023.104742.
- Lorenzetti V., Gaillard A., Thomson D., Englund A., Freeman T.P. Effects of cannabinoids on resting state functional brain connectivity: A systematic review. Neurosci. Biobehav. Rev. 2023;145 doi: 10.1016/j.neubiorev.2022.105014.
- Patel V.P., Feinstein A. Cannabis and cognitive functioning in multiple sclerosis: The role of gender. Multiple Sclerosis J. – Exp., Transl. Clinical. 2017;3(2) doi: 10.1177/2055217317713027.
- Pavisian B., MacIntosh B.J., Szilagyi G., Staines R.W., O’Connor P., Feinstein A. Effects of cannabis on cognition in patients with MS. Neurology. 2014;82(21):1879–1887. doi: 10.1212/wnl.0000000000000446.
- Rao S.M., Leo G.J., Bernardin L., Unverzagt F. Cognitive dysfunction in multiple sclerosis. Neurology. 1991;41(5):685–691. doi: 10.1212/wnl.41.5.685.
- Ritchay M.M., Huggins A.A., Wallace A.L., Larson C.L., Lisdahl K.M. Resting state functional connectivity in the default mode network: Relationships between cannabis use, gender, and cognition in adolescents and young adults. Neuroimage: Clinical. 2021;30 doi: 10.1016/j.nicl.2021.102664.
- Rocca M.A., Valsasina P., Absinta M., Riccitelli G., Rodegher M.E., Misci P., Rossi P., Falini A., Comi G., Filippi M. Default-mode network dysfunction and cognitive impairment in Progressive MS. Neurology. 2010;74(16):1252–1259. doi: 10.1212/wnl.0b013e3181d9ed91.
- Rocca M.A., Schoonheim M.M., Valsasina P., Geurts J.J.G., Filippi M. Task- and resting-state fmri studies in multiple sclerosis: From regions to systems and time-varying analysis Current Status and Future Perspective. Neuroimage: Clinical. 2022;35 doi: 10.1016/j.nicl.2022.103076.
- Romero K., Pavisian B., Staines W.R., Feinstein A. Multiple sclerosis, cannabis, and cognition: A Structural MRI study. Neuroimage: Clinical. 2015;8:140–147. doi: 10.1016/j.nicl.2015.04.006.
- Ruano L., Portaccio E., Goretti B., Niccolai C., Severo M., Patti F., Cilia S., Gallo P., Grossi P., Ghezzi A., Roscio M., Mattioli F., Stampatori C., Trojano M., Viterbo R.G., Amato M.P. Age and Disability Drive cognitive impairment in multiple sclerosis across disease subtypes. Mult. Scler. J. 2016;23(9):1258–1267. doi: 10.1177/1352458516674367.
- Schmidt, P., Mühlau, M., Gaser, C., & Wink, L. (2013). LS : A Lesion Segmentation Tool for SPM, 1–12.
- Smith S.M. Fast Robust Automated Brain Extraction. Hum. Brain Mapp. 2002;17(3):143–155. doi: 10.1002/hbm.10062.
- Thompson A.J., Banwell B.L., Barkhof F., Carroll W.M., Coetzee T., Comi G., Correale J., Fazekas F., Filippi M., Freedman M.S., Fujihara K., Galetta S.L., Hartung H.P., Kappos L., Lublin F.D., Marrie R.A., Miller A.E., Miller D.H., Montalban X., Cohen J.A. Diagnosis of multiple sclerosis: 2017 revisions of the McDonald Criteria. Lancet Neurol. 2018;17(2):162–173. doi: 10.1016/s1474-4422(17)30470-2.
- Thomson H., Labuschagne I., Greenwood L.-M., Robinson E., Sehl H., Suo C., Lorenzetti V. Is resting-state functional connectivity altered in regular cannabis users? A systematic review of the literature. Psychopharmacology. 2022;239(5):1191–1209. doi: 10.1007/s00213-021-05938-0.
- Wechsler D. Psychological Corporation; 2001. Wechsler Test of Adult Reading: WTAR.
- Wetter N.C., Hubbard E.A., Motl R.W., Sutton B.P. Fully automated open-source lesion mapping of T2-FLAIR images with FSL correlates with clinical disability in MS. Brain Behavior. 2016 doi: 10.1002/brb3.440.
- World Medical Association World Medical Association Declaration of Helsinki. J. Am. Med. Assoc. 2013;310(20):2191. doi: 10.1001/jama.2013.281053.
- Zigmond A.S., Snaith R.P. The hospital anxiety and depression scale. Acta Psychiatr Scand. 1983;67(6):361–370. doi: 10.1111/j.1600-0447.1983.tb09716.x. PMID: 6880820.
Further reading
Untitled section
References
- Feinstein A., Meza C., Stefan C., Staines W.R. Discontinuing cannabis improves depression in people with multiple sclerosis: A short report. Mult. Scler. J. 2020;27(4):636–639. doi: 10.1177/1352458520934070.
Associated Data
Supplementary Materials
Data Availability Statement
Data will be made available on request.