Multimodal MRI Reveals Brain Structural Differences and Executive Dysfunction in Early Methamphetamine Abstinence Authors: Ben Bristow, Maryam Tayebi, Paul Condron, Taylor Emsden, Tutarangi Ngarimu, Eryn Kwon, Rachael McLanachan, Gordon Liu, William Schierding, Patrick McHugh, Samantha Holdsworth, Wendy Mohi, Gil Newburn, Miriam Scadeng Published: 2026 ABSTRACT ABSTRACTBackgroundMethamphetamine use disorder (MUD) is known to have profound effects on brain structure and cognitive functions. Understanding the extent of these neurobiological changes during the early abstinence period is crucial for developing targeted rehabilitation strategies.ObjectiveThis study aimed to investigate structural brain alterations and cognitive functions in early abstinent methamphetamine users compared with healthy controls.MethodsA total of 27 participants were included, comprising 13 participants with MUD in early abstinence (<30 days—mean age: 37.4 ± 9 years, 66% female) and 14 age‐ and sex‐matched healthy controls (mean age: 39.2 ± 11.1 years, 74% female). All subjects underwent brain MRI scans using a 3T scanner, acquiring T1‐weighted and myelin‐sensitive imaging data. They also completed the Tower of London (TOL) cognitive assessment to evaluate executive functioning. Grey matter volumetric analysis was performed using T1‐weighted images, and both regional and global myelin measures were quantified for group comparison.ResultsBehaviorally, MUD participants demonstrated significantly longer execution times for correct solutions (p = 0.01), required more attempts to solve problems (p = 0.02), and achieved fewer correct solutions on the first attempt (p = 0.01). Neuroimaging analyses revealed significant cortical thinning in the left lateral occipital and right lingual cortices among MUD participants. Additionally, cortical volumes of the right superior frontal and lingual cortex were significantly reduced in the MUD group. Vertex‐wise analysis further showed a negative correlation between duration of methamphetamine use and the volume of multiple cortical regions.Conclusions:These findings indicate that methamphetamine use is linked to poorer executive function and cortical changes, while the absence of group differences in myelin measures suggests that grey matter is more vulnerable to meth‐induced damage than white matter. Introduction Methamphetamine is a highly addictive stimulant affecting ∼30 million users worldwide (UNODC, 2024) and in 2022 ranked as the third most used drug globally behind cannabis and opioids. As a dopamine agonist, methamphetamine increases synaptic dopamine release and blocks reuptake (Kish, 2008). It also impacts the broader neurotransmitter systems (Čechová and Šlamberová, 2021) by overstimulating postsynaptic neurons and blunting sensitivity to natural rewards, thus reinforcing addiction (May et al. 2020). Chronic use leads to metabolic, structural, and functional brain changes, contributing to memory deficits, emotional dysregulation, and impaired executive function (Thompson et al. 2004; Love et al. 2024). These changes are driven by neurotoxic cascades involving oxidative stress, neuroinflammation, and excitotoxicity, with microglial and astrocytic dysfunction playing key roles (Jayanthi et al. 2021). Magnetic resonance imaging (MRI) provides non‐invasive insights into brain changes associated with methamphetamine use (Caviness et al. 1999). Advances in imaging, including high‐resolution T1‐weighted and myelin‐sensitive sequences, now allow detection of subtle alterations in cortical thickness, deep grey matter (GM) volumes, and myelin composition. These advancements allow visualization of methamphetamine‐induced neurotoxicity in regions tied to cognitive dysfunction. Methamphetamine‐related neurotoxicity particularly affects dopamine‐rich regions (Ares‐Santos et al. 2014). Using MRI, areas including the striatum (Chang et al. 2007), prefrontal cortex (Jan et al. 2012a), anterior cingulate cortex (London et al. 2015), and hippocampus (Golsorkhdan et al. 2020) have consistently shown alterations. These regions are critical for motor control, executive function, habit formation, reward, and learning (Hardwick et al. 2013; Diamond et al. 2013; Graybiel and Grafton, 2015; Berridge and Kringelbach, 2015; Squire and Dede, 2015). Particularly, dysfunction in the prefrontal cortex manifests as impulsivity (Kogachi et al. 2017), poor inhibitory control (Jentsch and Pennington, 2014), and impaired planning and decision‐making (Kohno et al. 2014), which are common impairments in methamphetamine users. The dorsolateral prefrontal cortex which encompasses regions of the inferior, middle (including caudal), and superior frontal gyri, is the neural substrate encoding working memory, language processing, and cognitive control, with its dysfunction precluding risky decision making (Zare et al. 2021; Yamamoto et al. 2015). Impairments, especially in executive control and impulsivity, hinder abstinence (Dawe et al. 2004). Tests such as the Tower of London (TOL), Stroop task, and Iowa Gambling Tasks have been used to quantify these deficits (Chang et al. 2002), with TOL performance closely linked to planning ability (van der Plas et al. 2009). Beyond GM atrophy, myelin‐sensitive imaging techniques, such as MAGiC (Nasser et al. 2023) and diffusion MRI (dMRI) (Aung et al. 2013), offer insight into myelin and white matter integrity. Myelin facilitates rapid electrical conduction along axons and supports neuronal metabolism (de Faria et al. 2019). Myelin's involvement in normal neuronal function is well established (Khelfaoui et al. 2024), reflected by the diverse range of psychiatric and nervous system disorders that are characterized by damaged myelin (Fields, 2008). Methamphetamine use is known to induce chronic neuroinflammation via microglial activation (Sekine et al. 2008) and oxidative stress responses (Ravera et al. 2015), both of which may indirectly contribute to myelin damage (Ding et al. 2021). Disrupted myelin integrity in methamphetamine users is supported by dMRI studies showing increased radial diffusivity, a marker of myelin damage (Liu et al. 2025; Alicata et al. 2009). A mouse study by Bowyer et al. (2008) observed myelin damage in the caudate nucleus, and Sharma and Kiyatkin (2009) identified disrupted myelinated fibres partially in the cortex, hippocampus, thalamus, and hypothalamus through immunostaining in rats, both from acute methamphetamine exposure. This supports the findings that methamphetamine exposure can disrupt integrity and potentially damage myelin. This study investigates whether significant differences in GM volume and myelin content exist between methamphetamine users and healthy controls, alongside assessments of planning and executive cognitive functions. Identifying structural targets of METH neurotoxicity may inform future interventions aimed at supporting grey and white matter recovery during abstinence. Methods Participants The participants for this study were recruited through local rehabilitation centers from Tairāwhiti (Gisborne), New Zealand. Thirteen participants who met DSM‐V criteria for methamphetamine use disorder (MUD) and had been abstinent for less than 1 month were included. The control group consisted of 14 participants (age, sex, and ethnicity matched) with no history of methamphetamine use. Ethics approval was obtained (2022 EXP 11360) through the Health and Disability Ethics Committee (HDEC) in New Zealand. Participants were excluded if they had any contraindications to MRI (metal exposure, non‐compatible implants, claustrophobia), had major structural brain abnormalities unrelated to methamphetamine use, a history of other substance use disorders or current intoxication, or any current or prior psychiatric disorders. Executive Function Test The TOL task was administered to assess participants' planning and problem‐solving abilities (Shallice, 1982). This involves the movement of balls placed on pegs to match a certain configuration and in a specific number of moves, following the procedure described and discussed in Farhadian et al. 2017, sec. 2.3.3 and detailed in our supplementary material (Figure 1). Key metrics are reported in the supplementary material. Additionally, we calculated other parameters, including mean first‐move time (the average latency before initiating a move, calculated as total first‐move time/total attempts), and the impulsivity index (the ratio of mean first‐move time to execution time on correct trials). An image showing the Tower of London task layout. A computerized executive function task used to assess planning and problem‐solving abilities. MRI Acquisition MRI data were acquired at Mātai Medical Research Institute using a 3.0 T GE Healthcare Signa Premier scanner with a 48‐channel Air head coil. The imaging protocol incorporated T1‐weighted BRAVO structural imaging alongside MAGiC, a myelin‐sensitive imaging sequence (Table 1). TABLE 1MRI acquisition parameters used to acquire 3D T1‐weighted and myelin‐sensitive images.Parameter3D T1‐weighted IR‐prep, fast IR‐SPGR3D MAGiC PROMORepetition Time (TR)6.8 ms5.72 msEcho Time (TE)2.7 ms2.2 ms (max 22.2 ms)Flip Angle12°4°Matrix Size256 × 256256 × 256Field of View (FOV)25.6 cm25.6 cmSlices320312Voxel Size1 mm3 1 mm3 Slice Thickness0.5 mm1 mmScan Time1 min 41 s4 min 49 sJohn Wiley & Sons, Ltd. MRI acquisition parameters used to acquire 3D T1‐weighted and myelin‐sensitive images. Image Processing All images were visually checked, and those with significant head motion or other artifacts were excluded. Figure 2 shows an in‐depth analysis pipeline overview. No participants were excluded following a visual check. The acquired T1‐weighted images were processed using FreeSurfer (version 7.4.1, Fischl, 2012) for both cortical and subcortical segmentation. Cortical thickness and volumetric measures were extracted for the whole brain using FreeSurfer's automated parcellation based on the Desikan–Killiany atlas. Group comparisons were conducted across all cortical regions. For volumetric analyses, regional volumes were adjusted for total intracranial volume (TIV) to control for individual differences in overall brain size. Structural and myelin‐sensitive MRI data were processed using two complementary pipelines. T1‐weighted images were analyzed using FreeSurfer's recon‐all pipeline to extract cortical thickness, cortical volume, and subcortical volume. The Desikan–Killiany atlas was used for cortical parcellation, and FreeSurfer's aseg atlas for subcortical segmentation. The ‐qcache flag enabled smoothed cortical surface data in fsaverage space. Group‐level analyses were conducted using mri_glmfit, applying a general linear model with cluster‐wise p‐thresholding (p < 0.05) and 5 mm smoothing. Myelin‐sensitive images were acquired using a MAGiC sequence and processed with SyMRI 15.0 to generate quantitative myelin maps. These maps were registered to MNI152 space using FSL's flirt, and regional myelin values were extracted using the JHU white matter tractography atlas with fslmaths. Myelin‐sensitive images were acquired using a MAGiC sequence and processed using SyMRI 15.0 (SyntheticMR, Linköping, Sweden) to generate quantitative myelin maps (Warntjes et al. 2016). The FSL command, flirt, was used for linear image registration of the myelin images (Jenkinson and Smith, 2001; Jenkinson et al. 2002). Statistics Cortical volume and thickness differences between MUD and control groups were assessed using FreeSurfer's GLM after surface alignment to a common template and 5 mm FWHM smoothing. Vertex‐wise analyses included age and sex as covariates, with cluster‐wise Monte Carlo correction (p < 0.05). The rest of the statistical analyses were conducted using RStudio (version 2024.12.1+563; R version 4.4.1). Multiple linear regression models investigated group differences in neuroanatomical and cognitive measures—separate models were fitted for each outcome (subcortical volumes, cortical metrics, myelin values, and cognitive indices)—with group (MUD vs. control) as the primary predictor, adjusting for age and sex. For analyses of methamphetamine‐use duration, age and sex effects were first regressed out, and the resulting residuals were subjected to Pearson correlations to test associations with brain measures independent of those covariates. All p values with p < 0.05 considered significant. Results Participants The study included a total of 27 participants (13 MUD,14 healthy controls). Shown in Table 2, there was no significant difference in age between the MUD (38.08 ± 9.53 years) and control (39.21 ± 11.13 years) groups. Years of education were significantly higher (t = −3.42, p = 0.009) in the control group. Participants were predominantly Māori, with a higher proportion of females in both groups; MUD group showed greater rates of head injury (p = 0.016), smoking, and poly‐drug use, had used methamphetamine for 16.54 ± 21.26 years, and had been abstinent for 14.55 ± 5.73 days. TABLE 2Participant demographic information.CharacteristicCase (N = 13)Control (N = 14) p valueAge (years)38.08 ± 9.5339.21 ± 11.130.777Ethnicity (Maori/NZ European)12/114/00.481Sex (M/F)5/84/100.695Education (years)12.00 ± 1.7314.71 ± 2.400.009 * Abstinence (days)8.64 ± 9.80Years of Methamphetamine Use14.55 ± 5.73Head Injury (Yes)500.016 * Alcohol (Yes)6110.120Smoking (Yes)1050.077Poly Drug Use (Yes)960.322 Note: Group comparisons were performed using an independent samples t‐test.*p < 0.05.John Wiley & Sons, Ltd. Participant demographic information. Note: Group comparisons were performed using an independent samples t‐test. *p < 0.05. Cognitive Function The MUD group showed significantly longer execution times (t = −2.65, p = 0.01), more total number of attempts (t = 2.47, p = 0.02), and fewer correct solutions at first attempt (t = −2.5, p = 0.01) (Table 3). The TOL performance revealed no significant group differences for total score, mean solution time, total attempts, or first‐move time. Although controls scored slightly higher total scores and marginally faster mean solution time on correct trials, these differences were minimal. TABLE 3Cognitive performance on the Tower of London task for MUD and control groups.Tower of London performanceCase (M ± SD)Control (M ± SD) t value p value Cohen's d [95% CI] Total score30.22 ± 3.5333.09 ± 3.70−1.390.18−0.79 [−0.13 −1.71]Mean solution time—correct solutions (ms)18770 ± 613622738 ± 1828−0.400.690.28 [−0.61 1.16]Mean execution time—correct solutions (ms)8442 ± 17496827 ± 11002.650.01*−1.13 [−2.09 −0.18]Total number of attempts16.46 ± 2.6614 ± 2.482.470.02*0.95 [0.10 1.79]Total first move time (ms)164831 ± 56432234779 ± 251643−0.940.36−0.40 [−1.21 0.40]Mean first move time (ms)10330 ± 429416661 ± 18324−1.170.26−0.50 [−1.31 0.30]Impulsivity index0.0019 ± 0.0010.0015 ± 0.00081.120.270.42 [−0.37 1.23]Correct solutions at first attempt0.71 ± 0.170.87 ± 0.16−2.50.01*−0.96 [−1.80 −0.11]Mean attempts (2‐3 moves)1.41 ± 0.731.14 ± 0.361.240.220.45 [−0.35 1.25]Mean attempts (4‐5 moves)1.47 ± 0.361.28 ± 0.391.310.200.51 [−0.29 1.32]Independent t‐test * = p < 0.05.John Wiley & Sons, Ltd. Cognitive performance on the Tower of London task for MUD and control groups. Cohen's d [95% CI] Independent t‐test * = p < 0.05. Imaging Data Group Difference—Cortical Thickness When comparing cortical thickness group differences, significant clusters were identified in the left lateral occipital cortex (LOC) (peak = 4.287, area = 993.44 mm2, cluster‐wise corrected p = 0.015) and in the right lingual gyrus (LG) (peak = 4.887, area = 1227.49 mm2, cluster‐wise corrected p = 0.002), indicating that the control group exhibited significantly greater cortical thickness in these regions compared to the MUD group (see Figure 3). Group differences in cortical thickness between methamphetamine (case) and control participants, identified using FreeSurfer vertex‐wise analysis. After regressing out age and sex, the case group exhibited significantly reduced cortical thickness in (a) the left lateral occipital cortex (p = 0.015, cluster‐wise corrected) and (b) right lingual gyrus (p = 0.002, cluster‐wise corrected), compared to the control group. The left lateral occipital cortex, involved in object recognition and visuospatial processing, showed reduced thickness in the case group, suggesting possible impairments in visual perception. Similarly, reduced thickness in the right lingual gyrus may reflect disruptions in visual memory and learning associated with methamphetamine use. Group Difference—Cortical Volume Group comparisons revealed significantly lower cortical volumes in the MUD group than in the control group in two clusters: the right superior frontal cortex (peak = 4.065 within a 1205.87 mm2 cluster (cluster‐wise corrected p = 0.005)), and the right lingual gyrus (peak = 4.142 within a 1081.3 mm2 cluster (cluster‐wise corrected p = 0.018)) (see Figure 4). Group differences in cortical volume between methamphetamine (case) and control participants, identified using FreeSurfer vertex‐wise analysis. After regressing out age and sex, the case group exhibited significantly reduced cortical volume in (a) the right superior frontal cortex (p = 0.007, cluster‐wise corrected) and (b) right lingual gyrus (p = 0.018, cluster‐wise corrected), compared to the control group. The right superior frontal cortex, involved in executive functions such as working memory and decision‐making, showed reduced volume in the case group, aligning with cognitive deficits linked to substance use. The right lingual gyrus, important for visual processing and memory, also showed reduced volume, suggesting potential impairments in visual information processing among methamphetamine users. Correlation Between Cortical Metrics and Years of Methamphetamine Usage Cortical thickness did not show any correlation with length of methamphetamine use. Using FreeSurfer's vertex‐wise analysis with cluster‐wise correction, we identified a significant cluster in the volume of the left middle temporal cortex that was negatively associated with years of methamphetamine use (peak = –4.40 within a 1,274.10 mm2 cluster [cluster‐wise correction p = 0.002]). We also performed ROI‐based Pearson correlations; as summarized in Table 4, several cortical volumes exhibited moderate, dose‐dependent negative correlations with duration of use. The right caudal middle frontal cortex showed the strongest relationship (r = –0.754, p = 0.001), followed by the left caudal middle frontal (r = –0.673, p = 0.008) and the left lateral occipital cortex (r = –0.598, p = 0.023) (Figure 5). TABLE 4Correlation results between years of methamphetamine use and cortical volume after adjusting for age and gender.Cortical regionCorrelation [95% CI] p valueL. Caudal middle frontal−0.673 [−0.887 0.222]0.008L. Inferior temporal−0.561[−0.841 0.043]0.036L. Lateral occipital−0.598[−0.857 −0.099]0.023R. Caudal middle frontal−0.754[−0.917 0.373]0.001R. Middle temporal−0.532[−0.829 −0.003]0.049R. Superior parietal−0.565[−0.843 0.049]0.035John Wiley & Sons, Ltd. Correlation results between years of methamphetamine use and cortical volume after adjusting for age and gender. The scatter plots illustrate the negative associations between years of methamphetamine use and cortical volume in specific brain regions, after adjusting for age and gender. These findings indicate that prolonged methamphetamine exposure is associated with reduced cortical volume in regions implicated in executive function and visual processing. Myelin Analysis of myelin content across various white matter tracts revealed no significant differences between the MUD and control groups (Table 5). Similarly, cortical myelin volume did not differ significantly between groups. Although the global myelin volume appeared slightly reduced in the MUD group (M = 166.73 ± 22.52) compared to controls (M = 172.10 ± 29.87), this difference was not statistically significant (p = 0.81), and once corrected for intracranial volume and brain parenchyma, this effect diminished. The correlational analysis revealed no significant association between myelin values in any white‐matter region and the duration of methamphetamine use. TABLE 5Myelin content values across various white matter tracts and global myelin volume for the MUD and control groups.White matter tractsCase (M ± SD)Control (M ± SD) t value p valueCohen's d [95% CI]Body Corpus Callosum23.16 ± 1.9522.82 ± 1.74−0.560.580.18 [−0.58 0.94]Genu Corpus Callosum24.85 ± 2.1324.66 ± 1.89−0.240.810.09 [−0.67 0.85]Splenium Corpus Callosum23.21 ± 1.9523.04 ± 1.95−0.270.790.09 [−0.67 0.84]L. Superior Longitudinal Fasciculus24.57 ± 1.8624.28 ± 1.80−0.330.740.16 [−0.60 0.92]R. Superior Longitudinal Fasciculus23.45 ± 2.0022.98 ± 1.95−0.730.470.24 [−0.53 1.00]L. Cingulum23.83 ± 1.9423.54 ± 1.86−0.410.690.15 [−0.60 0.92]R. Cingulum22.71 ± 2.0322.70 ± 1.94−0.080.930.01 [−0.75 0.76]L. Superior Corona Radiation30.98 ± 2.4630.97 ± 2.49−0.210.840.00 [−0.76 0.76]R. Superior Corona Radiation23.35 ± 1.8822.98 ± 1.85−0.660.520.20 [−0.57 0.96]L. Anterior Thalamic Radiation29.65 ± 2.5030.47 ± 2.450.690.50−0.33 [−1.10 0.43]R. Anterior Thalamic Radiation30.89 ± 2.2530.68 ± 1.88−0.120.900.10 [−0.66 0.86]L. Inferior Frontal Occipital Fasciculus17.46 ± 1.6017.50 ± 1.57−0.270.79−0.03 [−0.79 0.73]R. Inferior Frontal Occipital Fasciculus31.45 ± 2.1932.10 ± 2.140.770.45−0.30 [−1.06 0.46]L. Inferior Longitudinal Fasciculus31.53 ± 2.5631.91 ± 2.550.360.72−0.15 [0.91 0.61]R. Inferior Longitudinal Fasciculus18.36 ± 1.7118.34 ± 1.63−0.160.870.02 [−0.74 0.78]L. Uncinate Fasciculus22.68 ± 1.8222.52 ± 1.84−0.310.760.09 [−0.67 0.85]R. Uncinate Fasciculus23.45 ± 2.0022.98 ± 1.95−0.730.470.24 [−0.53 1.00]Global myelin volume166.73 ± 22.52172.10 ± 29.870.230.81−0.21 [−0.99 0.57] Myelin volume (%) (Brain parenchyma normalised) 13.02 ± 1.4713.15 ± 1.45−0.270.78−0.09 [−0.87 0.69] Myelin volume (%) (Brain intracranial normalised) 11.81 ± 1.5311.87 ± 1.38−0.210.83−0.04 [−0.82 0.73]John Wiley & Sons, Ltd. Myelin content values across various white matter tracts and global myelin volume for the MUD and control groups. Myelin volume (%) (Brain parenchyma normalised) Myelin volume (%) (Brain intracranial normalised) Discussion This study investigated methamphetamine‐related brain structural alterations and executive dysfunction, recruiting individuals in early abstinence (<1 month) for multimodal MRI and Tower of London task assessment. MUD group exhibited focal cortical thinning and dose‐dependent volume reductions, especially in frontal and occipital regions, along with impairments in executive and other cognitive functions. No significant differences emerged for myelin content or subcortical volume. Prefrontal Executive Region Cortical analysis identified two closely related dorsolateral prefrontal (DLPFC) regions that are compromised in very‐early abstinent methamphetamine users (Kohno et al. 2014; Zare et al. 2021). The reduced volume in the right superior frontal gyrus (SFG), together with the dose‐dependent association between years of methamphetamine use and smaller volumes in the bilateral caudal middle frontal gyri (cMFG), indicates DLPFC involvement in a progressive impact of methamphetamine use on this region. (Kohno et al. 2014; Schwartz et al. 2010). Together, these regions constitute the DLPFC system, which underpins working‐memory manipulation, set‐shifting, impulse control, and strategic planning (Zhang et al. 2013; Falquez et al. 2014; Hu et al. 2016; Yee et al. 2010). Although these regions are commonly described within the dorsolateral prefrontal cortex framework, the present study identified volumetric differences primarily in a more medial portion of the prefrontal cortex, specifically within the superior frontal gyrus. The SFG occupies a medial–dorsal position and forms part of an extended prefrontal executive network, supporting higher‐order cognitive control processes that overlap substantially with classical DLPFC functions. The frontal cortex is a susceptible region for stimulant neurotoxicity, and structural MRI studies exploring methamphetamine use consistently demonstrate prefrontal cortical compromise (Kohno et al. 2014). Voxel‐based morphometry in active users shows GM loss in both medial frontal (Thompson et al. 2004) and SFG (Jan et al. 2012a). Short‐term abstinent individuals (<6 months) still show reduced density in the right middle frontal gyrus compared with those abstinent longer than 6 months (Kim et al. 2006). Schwartz et al. (2010) reported lower density in left middle‐frontal cortex, while Kogachi et al. (2017) showed sex‐specific effects where male users had relatively larger, and female users smaller, frontal volumes—particularly in the medial orbitofrontal and superior frontal cortices—compared with their respective controls. Kogachi et al. (2017) suggested that sex may modulate the effects of methamphetamine on brain morphometry. We have age‐ and sex‐matched the cohorts as well as added age and sex as covariates in the appropriate models to attempt to control for the effects of age and sex to better power the analysis to see the subtle cortical differences between MUD and control groups. Emerging evidence points to partial prefrontal recovery, with Nie et al. (2020) showing that longer abstinence was linked to thicker bilateral SFG—interpreted as reactive gliosis—and Morales et al. (2012) likewise reporting GM increases in the inferior frontal gyri within the 1st month, suggesting early and continuing cortical restoration. Collectively, these studies represent a dynamic course in which methamphetamine exposure and early abstinence are marked by pronounced dorsolateral and superior‐frontal alteration, whereas prolonged drug cessation may permit partial structural rebound. Posterior Vision Region Significant thinning was found in the right LG and left LOC. The LOC encodes object form and feeds conscious object recognition and visuospatial attention, whereas the LG contributes to scene processing, visual memory, and the integration of high‐resolution visual detail into semantic representations (van der Velden et al. 2023; Zhang et al. 2016). Damage in either region is therefore expected to degrade the fidelity of visual information available for higher‐order planning. A growing body of multimodal work supports the vulnerability of posterior visual‐association cortex to psychostimulant exposure and aligns closely with our LG‐LOC findings. MRI volumetry in chronic methamphetamine users consistently shows occipital GM loss: Nakama et al. (2011) reported accelerated volume decline in the occipital lobe relative to controls, while Nie et al. (2020) found lower right lateral occipital volume and linked thinner right LG/pericalcarine cortex to longer duration of use in long‐term abstinent users. Similarly, we identified negative correlation between the left lateral occipital volume and duration of methamphetamine use. However, there are studies that showed increased GM density in lateral occipital cortex in MUD individuals (Jan et al. 2012a; Morales et al. 2015). Functional studies show that methamphetamine and other drug exposures heighten activity or alter structure in occipital and lingual regions, suggesting a convergent drug‐related impact on visual processing networks (Van Hedger et al. 2018; Wang et al. 2012; Makris et al. 2008). The occipital cortex is highly glutamatergic and metabolically active, which may render it particularly vulnerable to meth‐induced glutamate dysregulation and excitotoxicity. This, in combination with meth‐related vascular and neuroinflammatory mechanisms, likely contributes to the vulnerability to structural and functional damage observed in visual–spatial regions (Halpin et al. 2014; Canedo et al. 2021; Self et al. 2012), although direct measures of methamphetamine‐evoked glutamate release in visual cortex are limited. Methamphetamine reliably alters glutamate levels within the striatum (Mark et al. 2004) and prefrontal cortex (Parsegian and See, 2014; Wu et al. 2018), while glutamate/glutamine (Glx) in the occipital cortex is known to be modulated during visual processing (Kurcyus et al. 2018). These mechanisms may underlie the cortical thinning we observed in the LOC and LG, highlighting the occipital cortex as a critical but underrecognized target of methamphetamine neurotoxicity. Association With Length of Methamphetamine Exposure Within the MUD group, we observed a clear correlation between duration of use and brain volume, where longer drug use was associated with smaller cortical volumes in multiple regions after age and sex adjustment. Cortical volume inversely correlated with years of methamphetamine use, particularly in DLPFC, left LOC, inferior temporal cortex, bilateral middle temporal cortex, and right superior parietal lobule, consistent with a dose‐dependent degradation of executive–visuospatial networks (Ruan et al. 2018; Nie et al. 2020). These dose–atrophy relationships appear in both active users and during early abstinence abstinent individuals for only a few weeks and are also echoed by longitudinal non‐human primate data (Howell and Murnane, 2008). Similar trajectories have been observed in MPTP‐administered (1‐methyl‐4‐phenyl‐1,2,3,6‐tetrahydropyridine) primate models, where the neurotoxin MPTP, which selectively targets dopaminergic neurons and causes neuroinflammation and subsequent GM atrophy, parallels the progressive degeneration seen in chronic methamphetamine exposure (Jeong et al. 2018). Dose‐dependent prefrontal atrophy accords with earlier morphometric work. Cumulative lifetime dose predicted reduced GM in the right superior frontal and superior temporal cortex, and the right caudate nucleus (Ruan et al. 2018). A large voxel‐based study of recently abstinent individuals likewise found that years of use correlated inversely with GM density of SFG (Schwartz et al. 2010). Beyond the frontal lobe, prolonged use has been linked to smaller LG and LOC volumes (Nie et al. 2020), echoing the occipital correlation detected in our study. Myelination No significant differences in myelin content were observed between the MUD and control groups using synthetic MRI, a novel imaging technique applied in this population. While previous studies using diffusion tensor imaging (DTI) have indirectly suggested myelin disruption—particularly via reduced fractional anisotropy in frontal and parietal regions (McKenna et al. 2016; Liu et al. 2025), myelination differences were not observed in our population. DTI infers microstructural and myelin/axonal changes from water diffusion patterns, whereas synthetic MRI directly estimates myelin content based on magnetic exchange between surrounding water compartments. Edema‐related changes linked to methamphetamine exposure, such as myelin swelling and disrupted tissue structure (Bowyer et al. 2008; Sharma and Kiyatkin, 2009), may interfere with accurate myelin quantification. Additionally, the absence of group differences may reflect limited statistical power due to small sample size. Nonetheless, significant GM loss in methamphetamine users suggests broader neurostructural disruption despite preserved myelin metrics. Subcortical Structures Although our study did not identify significant differences in subcortical structures between groups, prior research has consistently documented methamphetamine‐related alterations in key subcortical regions, particularly within the striatum (Chang et al. 2005), hippocampus (Golsorkhdan et al. 2020), and amygdala (Azimzadeh et al. 2024). For example, Chang et al. (2005) and Jan et al. (2012b) reported enlarged striatal volumes in abstinent methamphetamine users, which may reflect a compensatory neuroplastic response to chronic stimulant exposure. Animal studies further support this, with Jedynak et al. (2007) demonstrating methamphetamine‐induced structural plasticity in the dorsal striatum, and long‐term potentiation observed in the striatum of methamphetamine‐exposed rats (Nishioku et al. 1999; Furlong et al. 2017). A recent systematic review highlights inconsistencies in reporting the direction of volumetric change in subcortical structures with methamphetamine exposure (Liu et al. 2025); our cohort did not show significant differences in subcortical regions. Executive Function Performance MUD group showed slower, less efficient, and less accurate performances on the TOL task, solving fewer problems on the first attempt (71% vs. 87%), making more total attempts, and taking approximately 24% longer to execute correct solutions, despite comparable total scores. Suggesting participants are taking longer to formulate a plan, and when executing the plan, it is often inaccurate. These deficits align with structural reductions in dorsolateral prefrontal and occipital regions, which support executive and visuospatial planning. Performance diverged most clearly on complex tasks: while both groups performed similarly on simpler problems, MUD group made progressively more extra attempts as task difficulty increased, suggesting impaired cognitive flexibility and working memory. This mirrors findings in other stimulant users, where increasing task demands reveal planning inefficiencies (Ersche et al. 2006). In contrast, some studies report overall lower TOL scores in recently abstinent methamphetamine users (Farhadian et al. 2017), a difference not observed here. Despite slower execution, total problem‐solving time did not differ significantly, and MUD individuals showed slightly faster (but less accurate) first moves, hinting at impulsive, inefficient planning. Regions in which we observed volumetric reductions, right SFG and bilateral caudal middle frontal gyri, have been repeatedly implicated in planning and executive control by TOL studies and a meta‐analysis (Baker et al. 1996; Nitschke et al. 2017). This anatomical overlap is consistent with the planning inefficiencies measured on the TOL in our cohort. Additionally, thinning in the LOC and LG, which support visuospatial analysis, may impair the mental manipulation of intermediate problem states. These findings raise the possibility that methamphetamine‐related neurotoxicity disrupts the integrity of the inferior fronto‐occipital fasciculus (IFOF), a key white matter tract linking occipital and prefrontal regions. Prior diffusion studies in methamphetamine users support this hypothesis, showing altered IFOF microstructure (Huang et al. 2020; Andres et al. 2016), which may compromise communication between visual input and strategic planning systems. Potential Confounding Factors Several co‐occurring factors may confound the interpretation of neuroimaging and cognitive findings in this study. (i) Tobacco smoking, more prevalent among MUD group in our sample, has been associated with reduced GM volume in the prefrontal cortex and dorsal anterior cingulate (Brody et al. 2004). (ii) Cannabis use, also more frequent in the MUD group, has been linked to reductions in frontal and temporal GM (Battistella et al. 2014), potentially exacerbating methamphetamine‐related neurotoxicity. Additionally, (iii) attention‐deficit/hyperactivity disorder (ADHD) is more common among methamphetamine users (Obermeit et al. 2013), and both ADHD and methamphetamine use disorder are characterized by disruptions to mesolimbic and mesocortical dopamine systems (Eme, 2017). Structurally, thinner right superior frontal cortices have been reported in children with ADHD (Hai et al. 2022). (iv) Trauma can manifest through various forms of abuse, including physical abuse, emotional or psychological abuse, neglect, and exposure to violence (Felitti et al. 1998). Particularly in childhood, the neurobiological manifestations of abuse are found to translate into abnormal development of important stress, emotional regulation, cognitive, and reward pathways (Wiet, 2017). These neurobiological changes predispose individuals who have suffered trauma to substance use disorders later in life, including methamphetamine addiction (Banducci et al. 2014; Rasche et al. 2016; Miela et al. 2018; Huang et al. 2021). Together, these factors may independently or synergistically contribute to the observed brain alterations, complicating attribution solely to methamphetamine exposure. Limitations The primary limitation of this study is the small sample size, which reduces statistical power and limits the ability to detect subtle or region‐specific group differences. This also restricts generalizability. Our study is also limited by the restricted executive function inferences that can be made, as no further executive function tests were used. Although group demographics were matched, high rates of polysubstance use, head injury, and psychiatric conditions in the MUD group were self‐reported and not clinically verified, introducing potential misclassification bias. The use of self‐report measures may result in underreporting or inaccuracies in substance use history and comorbid diagnoses. Finally, while the study controlled for age and sex, other variables such as sleep, diet, childhood trauma, comorbidity, and socioeconomic status were not accounted for and may influence brain structure or cognition. Additionally, the higher education level in the control group may have provided a potential source of cognitive reserve. These factors should be considered in future studies with larger, more diverse cohorts and standardized clinical assessments. Conclusion Our early‐abstinence methamphetamine group showed evidence of impaired planning and GM loss in the dorsolateral prefrontal and occipital cortices, with the severity of structural decline scaling with cumulative drug exposure. However, the small sample size means these findings should be considered preliminary. This pattern supports a dose‐dependent breakdown of fronto‐occipital networks critical for executive control. Given the lack of observed myelin and subcortical volume differences, cortical GM may exhibit greater susceptibility to methamphetamine‐induced structural alterations than white matter. Future multimodal studies with larger, better‐controlled samples are needed to disentangle drug‐specific effects from those of polysubstance use, comorbid conditions, and individual neurodevelopmental risk factors. Author Contributions Ben Bristow: writing – review and editing, writing – original draft, visualization, formal analysis, investigation, data curation. Maryam Tayebi: investigation, writing – original draft, writing – review and editing, visualization, methodology, formal analysis, project administration, supervision, data curation. Paul Condron: investigation, methodology, visualization, software, resources, data curation. Taylor Emsden: investigation, methodology, visualization, software, data curation. Tutarangi Ngarimu: investigation, methodology, resources. Eryn Kwon: conceptualization, investigation, funding acquisition, methodology, software, project administration, supervision, data curation, resources. Rachael McLanachan: writing – original draft. Gordon Liu: writing – review and editing, writing – original draft, methodology. William Schierding: conceptualization, writing – review and editing, methodology, visualization, software, formal analysis, resources, funding acquisition. Patrick McHugh: Methodology, Investigation, Resources, Conceptualization, Funding acquisition. Samantha Holdsworth: investigation, conceptualization, funding acquisition, methodology, writing – review and editing, project administration, supervision, resources. Wendy Mohi: investigation, funding acquisition, methodology, project administration, data curation, resources. Gil Newburn: conceptualization, investigation, funding acquisition, methodology, supervision, resources. Miriam Scadeng: conceptualization, investigation, funding acquisition, writing – review and editing, methodology, project administration, supervision, data curation, resources. Funding This research was supported by Methamphetamine Recovery Programme cornerstone funder, the Fred Lewis Enterprise Foundation and Mātai Medical Research Institute (cornerstone funder Kānoa—Regional Development Unit of New Zealand). The authors gratefully acknowledge their contribution, which made this work possible. B.B. was further supported by a University of Auckland Doctoral Scholarship; M.T. and E.K. by senior research fellowships from the Hugh Green Foundation; W.M. by the Health Research Council; and W.S. by the Vision Research Foundation. Conflicts of Interest The authors declare no conflicts of interest. Supporting information Supplementary Material: brb371288‐sup‐0001‐SuppMat.docx