Structural and metabolic brain changes in the striatum associated with methamphetamine abuse


  • Linda Chang,

    Corresponding author
    1. Department of Medicine, John A. Burns School of Medicine, University of Hawaii, Honolulu, HI, USA and
    Search for more papers by this author
    • All authors declare no conflict of interests.

  • Daniel Alicata,

    1. Department of Medicine, John A. Burns School of Medicine, University of Hawaii, Honolulu, HI, USA and
    Search for more papers by this author
    • All authors declare no conflict of interests.

  • Thomas Ernst,

    1. Department of Medicine, John A. Burns School of Medicine, University of Hawaii, Honolulu, HI, USA and
    Search for more papers by this author
    • All authors declare no conflict of interests.

  • Nora Volkow

    1. National Institute on Drug Abuse and National Institute on Alcohol Abuse and Alcoholism, National Institutes of Health, Rockville, MD, USA
    Search for more papers by this author
    • All authors declare no conflict of interests.

Linda Chang MD, 1356 Lusitana Street, 7th Floor, UH Tower, Department of Medicine, John A. Burns School of Medicine, University of Hawaii, Honolulu, HI 96813, USA. E-mail:


Aims  To review structural, chemical and metabolic brain changes, particularly those in the basal ganglia, in individuals who used methamphetamine, as well as in children with prenatal methamphetamine exposure.

Methods  Magnetic resonance imaging (MRI) and positron emission tomography (PET) studies that evaluated brain structural, chemical and metabolite changes in methamphetamine subjects, or children with prenatal methamphetamine exposure, were reviewed and summarized. Relevant pre-clinical studies that provided insights to the interpretations of these imaging studies were also reviewed.

Results  In adults who used methamphetamine, MRI demonstrates enlarged striatal volumes, while MR spectroscopy shows reduced concentrations of the neuronal marker N-acetylasparate and total creatine in the basal ganglia. In contrast, children with prenatal methamphetamine exposure show smaller striatal structures and elevated total creatine. Furthermore, PET studies consistently showed reduced dopamine transporter (DAT) density and reduced dopamine D2 receptors in the striatum of methamphetamine subjects. PET studies also found lower levels of serotonergic transporter density and vesicular monoamine transporter (VMAT2) across striatal subregions, as well as altered brain glucose metabolism that correlated with severity of psychiatric symptoms in the limbic and orbitofrontal regions.

Conclusion  Neuroimaging studies demonstrate abnormalities in brain structure and chemistry convincingly in individuals who used methamphetamine and in children with prenatal methamphetamine exposure, especially in the striatum. However, many important questions remain and larger sample sizes are needed to validate these preliminary observations. Furthermore, longitudinal studies are needed to evaluate the effects of treatment and abstinence on these brain changes and to determine whether imaging, and possibly genetic, markers can be used to predict treatment outcome or relapse.


Methamphetamine induces both acute and chronic neurotoxic changes in dopaminergic and serotonergic neurons in preclinical models, including rodents and non-human primates [1–3]. Similarly, clinical observations indicate that chronic methamphetamine abuse is associated with persistent neurotoxicity, as some individuals who chronically used methamphetamine developed psychosis or persistent paranoia even years after cessation of their drug use. Recent neuroimaging studies in humans who used methamphetamine indeed demonstrated significant brain changes involving the dopaminergic [4,5] and serotonergic systems [6], glucose metabolism [7,8], neurometabolite levels [9,10] and even gross structural abnormalities [11–13]. These changes are particularly prominent in the basal ganglia, also called the striata, which include the putamen and globus pallidus dorsally, and the nucleus accumbens ventrally. Because striatal structures have the highest densities of dopaminergic synapses, which are the major sites of action for methamphetamine, neurotoxic effects may be expected to be most pronounced in these regions. While some of the neurotoxic effects of methamphetamine appear to be persistent, as the effects are observed even during abstinence [4,5,7,9,12,14]; some studies documented at least partial recovery of these neurochemical abnormalities in methamphetamine users withprotracted abstinence [15,16]. This review will discuss brain changes, particularly those found in the basal ganglia, in chronic, active and abstinent methamphetamine users, as well as in children with prenatal methamphetamine exposure. Areas that need additional research will be identified and discussed.


Several magnetic resonance imaging (MRI) studies found abnormalities in cortical and subcortical brain regions in recently abstinent and active methamphetamine users [11–13] ( Table 1). One study evaluated primarily individuals who were recently (average 14 days) abstinent from methamphetamine use with an automated segmentation program and found that the right frontal lateral ventricle and overall white matter volumes were enlarged, while the right medial cingulate gyrus, subgenual cortex paralimbic belts and bilateral hippocampus were smaller compared to control subjects. Furthermore, those with smaller hippocampal volumes had poorer memory performance on the word-recall test [11]. In contrast, another study from Korea of intravenous methamphetamine users who had been abstinent for longer periods (average 20 months) found no changes in brain volumes, except for a minor alteration in the shape of the corpus callosum [17] (Table 1). Given the different sex-proportion of the groups in the latter study, however, additional studies are needed to evaluate possible white matter abnormalities in methamphetamine users. Furthermore, these two reports did not evaluate the basal ganglia, which are major regions of interest in terms of methamphetamine-induced neurotoxicity.

Table 1.  Brain structural abnormalities evaluated by structural MRI.
StudySubjectsGender (male/female)Age (years)Meth. usage variablesAnalysis methodRegions examinedFindings
Thompson et al. [11]22 active meth. users15M/7F35.3 ± 1.7Duration: 10.5 ± 1.1 year; usage: 3.4 ± 0.8 g/week;Automated segmentation: cortical pattern matchingCortical gray matter and white matter, ventricles, hippocampusEnlarged right frontal lateral ventricle and bilateral white matter; decreased volumes of the right medial cingulate gyrus, subgenual cortex paralimbic belts and bilateral hippocampus; smaller hippocampal volumes correlated with decreased memory performance on the word-recall test
21 comparison subjects10M/11F31.9 ± 1.5Used in last 30 days: 18.9 ± 1.8 days
   Age first use: 26.1 ± 1.8 years
Chang et al. [12]50 abstinent meth. users24M/26F32.1 ± 7.1Life-time used: 4519 ± 5730 g; at 1 g/day (median)Manual morphometry with regions of interest; semi-automatic segmentation of corpus callosum (CC)Caudate, putamen, globus pallidus, thalamus, midbrain, cerebellum, CCMeth. users had larger putamen +10%) and globus pallidus (+7–9%) volumes, but those with smaller striatal (putamen and globus pallidus) volumes performed more poorly on neuropsychological tests and had longer life-time meth. use
50 non-drug users24M/26F31.7 ± 7.4Abstinent: 4.0 ± 6.2 months
Jernigan et al. [13]21 abstinent meth.Meth.: 17M/4F38.2 ± 7.7Life-time used: 4930 ± 93.6 g;Semi-automated morphometryCaudate, putamen, nucleus accumbens, thalamus, hippocampus, amygdalaMeth. users had larger caudate, lenticulate, accumbens and parietal cortices; younger meth. users had larger nucleus accumbens volumes; meth. users with larger parietal cortical volumes had more severe cognitive impairment
22 abstinent meth. + HIV 30 HIV 30 neitherHIV-negative non-drug: 17M/13F38.1 ± 10.5Abstinent 94 ± 89 days (∼3.0) monthsCerebral cortices (all lobes)
Oh et al. [17]27 abstinent meth. users23 M/4F36.7 ± 5.6Life-time used: 334 ± 506 g intravenously 0.63 ± 0.5/daySemi-automatic segmentationCorpus callosumIncreased curvature in the genu (4.1 degrees) and decreased width in posterior midbody (0.77 mm) and isthmus area (0.86 mm) of the CC were observed in meth. users relative to healthy comparison subjects
18 comparison subjects14 M/4F33.6 ± 6.7Abstinence: 20.5 ± 35.4 months
Chang et al. [35]13 meth. exposed children4M/9F6.9 ± 3.5Meth. exposed for at least 2/3 of pregnancyAutomated segmentation: cortical pattern matchingCaudate, putamen, globus pallidus, thalamus, midbrain, cerebellum, hippocampusMeth. exposed children had smaller putamen (−17.7% bilateral), globus palidus (left −27%; right −30%) and hippocampus (left −19%; right −20%) volumes
15 unexposed children6M/9F7.8 ± 3.2

Two additional MRI studies documented enlargement of the basal ganglia volumes, including the caudate nucleus, globus pallidus and putamen [12,13], and the parietal cortex [13] in adults who used methamphetamine chronically but were abstinent (average 3–4 months) at the time of the studies (Table 1, Fig. 1). Other subcortical regions evaluated, including the thalamus, midbrain, cerebellum (vermis) and the corpus callosum, were relatively normal [12]. Based on these cross-sectional studies, larger white matter, parietal cortical and basal ganglia volumes were observed during early abstinence (< 4 months) from methamphetamine use [11–13], but relatively normal volumes in those with longer abstinence (> 20 months). The early enlargement of these brain structures suggests methamphetamine-induced inflammation or reactive gliosis, which may normalize with longer abstinence from methamphetamine use. Longitudinal studies are needed to determine whether these enlarged structures would normalize with prolonged abstinence in the same individuals that used methamphetamine. However, the volumes of putamen and globus pallidus were smaller in those with greater cumulative methamphetamine use, suggesting an eventual loss of basal ganglia volumes with greater methamphetamine exposure [12].

Figure 1.

Left top: structural MRI showing striatal brain regions measured in the morphometry studies. Right: bargraphs showing methamphetamine subjects, as a group, had larger putamen and globus pallidus than comparison subjects. Left bottom: FDG image (co-registered with MRI and atrophy corrected) showing decreased relative metabolism in the striatal region. (Part of this figure is modified from Chang et al. [12])


One possible mechanism for striatal enlargement observed on MRI is increased water content or cell volumes association with inflammation. Glial activation, a form of neuroinflammatory response, associated with methamphetamine is well documented in pre-clinical models. In vitro studies demonstrated that methamphetamine contributes to central nervous system (CNS) inflammation by stimulating increased release and/or activation of matrix degrading proteinases [18], which would lead to breakdown of the blood–brain barrier and the influx of inflammatory cytokines, chemokines and macrophages into the brain.

One of the MRI studies that reported enlarged striata in individuals who used methamphetamine chronically also found better cognitive performance in those with larger striata [12], which suggests that the enlarged striata may be a compensatory response to the effects of methamphetamine. Microglial and astroglial responses are indeed normal compensatory neural responses to brain injury. However, excess neuroinflammation may lead to further brain injury. In a separate study, Jernigan et al. found that subjects chronically used methamphetamine with poorer performance had larger volumes of parietal cortices [13], suggesting deleterious effects of inflammation in the parietal region on cognitive function. Hence, neuroinflammatory response associated with methamphetamine exposure in different brain regions may lead to differential effects on cognitive function.

In vivo studies of methamphetamine administration to rodents and non-human primates have demonstrated significant activation of astroglia, microglia and mediators of inflammation including prostaglandins, cytokines, interleukin, neurotrophic factors and reactive oxygen species [3,19–23]. These changes may be responsible for blood–brain barrier and cellular damage leading to increased metabolic demands on these vulnerable regions. Conversely, pre-treatment with the non-steroidal anti-inflammatory agent indomethacin [24], NMDA receptor antagonists dizocilpine (MK-801) or dextromethorphan [25] and the antioxidizing agent, n-acetyl-l-cysteine [26,27], all have been shown to attenuate methamphetamine-associated microglial activation and the depletion of dopamine and dopamine transporters (DAT) in the terminal regions of striatal neurons following repeated toxic doses of methamphetamine. Therefore, anti-inflammatory agents, glutamatergic antagonists or antioxidants might have significant benefits on the cognitive outcomes of recovering methamphetamine addicts. Imaging studies such as MRI may be used to monitor possible improvement on the inflammatory changes by assessing changes in brain volumes, especially the striata and the parietal cortices.

Another possible mechanism for enlarged striatal structures might involve trophic substances released by microglia or astrocytes and the trophic effects of dopamine. A recent study found that plasma levels of brain derived neurotrophic factor (BDNF) were elevated in a cohort of 50 adult who used methamphetamine compared to control subjects, although the BDNF increase did not correlate with any of the drug usage variables [28]. Studies in rodents found increases in dopamine neuron spine density in the striatum after cocaine administration [29], which might be related to the cocaine-mediated increased extracellular dopamine. Hence, the trophic effects of dopamine also might explain the opposite effects of enlarged (inflammation in) basal ganglia and parietal cortices on cognitive performance. Dopaminergic terminals, which release dopamine, are concentrated in the basal ganglia but not in the parietal cortices. Therefore, the neurotrophic effects of dopamine may counteract the injurious effects of inflammation (i.e. astroglial and microglial activation) in the basal ganglia.

An additional possible etiology for striatal enlargement may be related to chronic D1 or D2 occupancy. Schizophrenia patients treated with typical antipsychotics that have high affinity binding to D2 receptors [30,31], but not those treated with atypical antipsychotic medications with low affinity for the D2 receptors [32], showed enlarged striatal volumes. Stimulant-dependent individuals, including those who used methamphetamine or cocaine [33] that cause dopamine release and binding to D1 or D2 receptors, also showed enlarged basal ganglia structures on MRI. However, a recent study in transgenic mice that expressed enhanced green fluorescent protein (EGFP) under the control of either the D1 or D2 receptor promoter (Drd1-EGFP or Drd2-EGFP) found that the increased spine density observed after chronic cocaine treatment is stable only in D1-receptor-containing neurons and a similar pattern of changes was observed in DeltaFosB expression D1 receptor-containing neurons in the nucleus accumbens [29]. These changes appear to last several months (1–3.5 months) following the last dose of cocaine suggesting long-lasting neuroplasticity associated with chronic stimulant use. These studies suggest that striatal enlargement in cocaine users is mediated more probably by chronic D1-receptor occupancy while basal ganglia enlargement in schizophrenic patients is related primarily to D2 receptor binding (by typical antipsychotics). However, the mechanism of how D1 and D2 receptors might lead to the morphological changes remains unclear. Further studies are needed to determine whether D1 or D2 receptors, or both, are involved in these changes in the striatum of methamphetamine subjects.


Relatively little is known about the effects of methamphetamine on the developing brain, either in adolescents or in children exposed to methamphetamine in utero. In a study of adults abstinent from methamphetamine use, enlarged nucleus accumbens was observed primarily in the younger methamphetamine subjects (between ages 20–30 years) [13]. These young adult methamphetamine subjects had similar volumes of nucleus accumbens as healthy 10–15-year-old children [34]. Because these young methamphetamine subjects also initiated their regular methamphetamine use at an earlier age, primarily during adolescence, methamphetamine abuse might have altered the normal developmental process of myelination and volume loss of the striatal structures.

Further implication for the potential neurotoxic effects of methamphetamine on brain development is derived from a small MRI study of children who had prenatal methamphetamine exposure compared to unexposed children [35] (Table 1). In contrast to adult methamphetamine subjects who showed enlarged basal ganglia structures, these children with prenatal methamphetamine exposure, for at least two-thirds of the pregnancy, had smaller putamen bilaterally (−17.7%) and smaller globus pallidus (−27% left; −30% right). In addition, smaller hippocampal volumes (−19% left; −20% right) was observed [35], similar to that in adult methamphetamine subjects [11]. As these children were evaluated post-natally at ages 3–16 years, these brain changes may reflect a pattern of neuroadaptation and recovery from prenatal methamphetamine exposure. Whether these volume changes will normalize with continued brain maturation into adulthood will require further longitudinal follow-up studies. These findings also illustrate that the effects of chronic methamphetamine exposure on the immature developing brain differs from that in the mature brains of adults who used methamphetamine (smaller rather than larger basal ganglia structures). The more pronounced decrease (−20%) in hippocampal volumes in these children compared to −7% in the adults also suggests that the developing brain may be more vulnerable to the neurotoxic effects of methamphetamine. Additional factors that may contribute to the in utero effects of methamphetamine may include the direct effects of methamphetamine on uterine vascular resistance, blood flow and fetal hypoxia [35]. Other associated factors, such as nutritional deficiencies [36] or maternal stress with high cortisol levels during pregnancy [37], may also contribute to the smaller brain structures observed in these children.


[1H] MRS is an in vivo technique that provides sensitive measurement of major metabolites in brain tissue, including N-acetylaspartate (NAA), a marker for neuronal integrity and density; creatine and phosphocreatine (CR), which reflect cellular energy metabolism; choline-containing compounds (CHO), a marker of membrane turnover (from synthesis and degradation); and myo-inositol (MI). The concentrations of CR as well as CHO are threefold higher in glia than in neurons, while MI is found only in glial cells. Therefore MI, and to a lesser extent CHO and CR, are considered putative markers of glial content [38].

[1H] MRS studies of individuals who abstained recently from chronic methamphetamine use demonstrated decreased concentrations of NAA and CR in the basal ganglia and to a lesser extent in the frontal white matter, while the concentrations of CHO and MI were increased in the medial frontal gray matter, where the NAA level was relatively normal [9,39] (Table 2). The reduced NAA concentration in the basal ganglia indicates reduced neuronal density, injury or loss, while the normal NAA level in the frontal gray matter suggests that neurons in this region may be less sensitive to the damaging effects of methamphetamine. Moreover, the increased concentration of CHO and MI in the frontal gray matter may reflect inflammation and glial activation or proliferation [9]. These findings are consistent with another small study that showed increased glial metabolite ratio of CHO/CR and decreased NAA/CR in the medial frontal gray matter (anterior cingulate cortex) in methamphetamine subjects during early abstinence (mean 9.3 weeks) [40]. A longitudinal follow-up study by the same group demonstrated further that the decreased NAA/Cr in the cingulate cortex during early abstinence (∼3 months) persisted with protracted abstinence (∼3 years) [10].

Table 2.  Brain metabolite abnormalities on proton MR spectroscopy in METH users and METH-exposed subjects.
StudySubjectsGender (male/female)Age (years)Meth. usage variablesBrain regions evaluatedFindings
Ernst et al. [9]26 meth. users13M/13F33.4 ± 7.9Life-time exposure 3640 g (median);Frontal gray matter (FGM: at anterior cingulate)Decreased NAA (−6%) and decreased CR (−8%) in BG; increased CHO (+13%) and (+11%) in FGM
24 non-drug user controls12M/12F30.2 ± 4.6Abstinence 4.25 months (median)Frontal white matter (FWM), basal ganglia (BG)Greater cumulative life-time meth. usage correlated with lower NAA in the frontal white matter
Sekine et al. [41]13 meth. users8M/5F25.7 ± 3.7Duration of use: 3.2 ± 2.5 yearsRight and left basal gangliaDecreased CR in bilateral BG; CR correlated inversely with duration of meth. use and with positive symptom subscore on BPRS
11 comparison7M/4F26.0 ± 2.0Abstinence 1.5 ± 1.2 years
Nordahl et al. [40]9 meth. users;9 M32.5 ± 6.4Duration of use: 15.1 ± 7.2 yearsAnterior cingulated cortex (ACC), visual cortexDecreased NAA/CR and increased CHO/CR in anterior cingulated in meth. users
9 comparison9 M32.7 ± 6.8Abstinence 9.3 ± 3.3 weeks
Nordahl et al. [10]16 recently abstinent meth.5M/11F37.2 ± 2.3Abstinence: 2.95 ± 0.4 monthsAnterior cingulated cortex (ACC), visual cortexDecreased ACC NAA in both recently and distantly abstinent meth. users; NAA correlated negatively with duration of use; increased CHO in recently abstinent meth. group only; CHO correlated negatively with duration of abstinence
8 distantly abstinent meth.4M/4F36.5 ± 2.8Abstinence: 37.50 ± 5.87 months
16 comparison5M/8F34.7 ± 2.3 
Chang et al. [39]60 meth. users (24 with HIV)18M/18M34.9 ± 6.7Duration: 108 ± 71 months; 6.6 ± 0.9 day/week; 2.2 ± 3.2 g/day; life-timeFrontal gray matter (anterior cingulate), frontal white matter, basal gangliaMeth. users showed decreased NAA in BG (−4%) and FWM (−5%); increased CHO in FGM (+12%); increased MI in FGM (+10%)
83 non-drug users (44 with HIV)23M/16F35.5 ± 8.6usage: 8241 ± 16 850 g; abstinence 6.3 ± 7.8 monthsMeth. users with HIV had additional decreased NAA in BG (−9%) and increased CHO and MI in FWM
Smith et al. [42]12 meth. exposed children;Not classified8.1 ± 0.8Prenatal meth. exposure for 2/3 of pregnancyFrontal white matter Basal gangliaIncreased CR in BG (+10%); non-significantly decreased NAA in FWM (−8%)
14 unexposed children7.3 ± 1.1

Abnormal brain metabolites in the basal ganglia also correlated with psychiatric symptoms and estimated indices of methamphetamine exposure [41]. A small study of individuals who used methamphetamine in Japan found a lower metabolite ratio of CR/CHO bilaterally in the basal ganglia that correlated negatively with residual psychiatric symptoms, as measured by the positive symptom subscale of the Brief Psychiatric Rating Scale (BPRS) [41]. This subscale includes measures for symptoms of conceptual disorganization, tension, mannerism and posturing, grandiosity, hostility, suspiciousness, hallucinatory behavior, uncooperativeness, unusual thought and excitement. The findings from this small study suggest that chronic methamphetamine use might lead to persistent metabolite abnormalities in the basal ganglia that are associated with the generation and persistency of psychiatric symptoms in abstinent users. However, this study might have missed concurrent decreases in NAA and CR in the basal ganglia, because only the metabolite ratio of NAA/CR was measured. Furthermore, the subjects in the Japanese study were younger and had used methamphetamine for only 3 years, whereas the subjects in the earlier report had used the drug for more than 10 years and were older in age [9] (Table 2). In the studies that measured NAA and CR concentrations, subjects who had greatest cumulative life-time methamphetamine usage showed the lowest NAA in the frontal white matter [9]. In the longitudinal study described above [10], NAA/Cr correlated negatively with duration of methamphetamine use while CHO/NAA correlated negatively with duration of abstinence. The findings are difficult to interpret, as changes in these metabolites might be due to either the numerator, or the denominator metabolites, or both. Additional studies are needed to further clarify these observations.

In contrast to adults who abused methamphetamine, children with a history of prenatal methamphetamine exposure (aged 3–16 years) had higher levels of CR in the basal ganglia (+10%) and a trend for lower NA (−8%) in the frontal white matter [42]. The increased CR level may reflect methamphetamine-induced changes in cellular energy metabolism or methamphetamine-induced glial activation as the CR concentration is threefold higher in glia than in neurons. However, the glial markers CHO and MI were normal. This small study suggests that methamphetamine may affect the immature developing brain differently than the mature adult brain. More studies are needed to verify the findings from this pilot study.


Cognitive impairment or deficits have been reported in individuals actively using methamphetamine, as well as during early and prolonged periods of abstinence. Active methamphetamine users performed poorer than control subjects on tests that evaluated learning and memory (word recognition and word and picture recall), attention, psychomotor speed (Symbol Digit Modalities Test and Stroop color test) and executive functioning (Trailmaking Test, part B) [43]. Although these deficits were not related to the duration of methamphetamine use, those who used the drug intravenously tended to perform poorer than the other users. These findings suggest reduced attention ability, decreased capacity for working memory and processing speed and reduced ability to manipulate new information (episodic memory) as a result of active methamphetamine use.

During early abstinence (5–14 days of last use), poorer performance in attention, psychomotor speed, learning, memory, verbal and non-verbal fluency were also reported [44]. Drug withdrawal effects could have contributed to the poorer cognitive performance. However, several studies demonstrated that individuals who chronically used methamphetamine performed more poorly on cognitive tests even after prolonged abstinence (over 1 year) [5]. Methamphetamine subjects during both early [45] and protracted periods of abstinence [46] were found to be more easily distracted and performed more poorly on selective attention tasks, due possibly to their inability to suppress irrelevant information.

Several of the studies discussed above evaluated the relationship between brain volume or metabolism in relation to cognitive function. Thompson et al. [11] demonstrated in a cohort of adults who used methamphetamine actively that smaller bilateral hippocampal volumes correlated with poorer memory on the word-recall test. This finding suggests that chronic methamphetamine use may lead to injury in the hippocampus and hence impairment in episodic memory function.

In the study that found enlarged striatal (putamen and globus pallidus) volumes in adults who recently abstained from methamphetamine use, those with greater cumulative methamphetamine use, or longer duration of use, had smaller striatal structures [12]. Furthermore, individuals with smaller striatal volumes also performed more poorly on several tests that involved executive function (verbal fluency) and fine motor function (non-dominant grooved pegboard). These findings suggest that although methamphetamine use may be associated initially with enlargement of the striatal structures, probably as a compensatory (inflammatory) response, and preserved cognitive function, the volumes of the striatum ultimately decrease with greater methamphetamine usage, accompanied by cognitive impairment. Finally, Jernigan et al. [13] found enlarged striatal volumes in adults who used methamphetamine during early abstinence; however, poorer performance on cognitive function was associated with larger volumes in the parietal cortices, while there was no specific relationship with basal ganglia volumes.

In the small pilot study, children exposed to methamphetamine in utero performed more poorly on visual motor integration, sustained attention (for both errors of omission and commission and response time variability) and long-delay verbal memory [35]. Poorer cognitive performance was correlated with smaller striatal as well as smaller hippocampal volumes. Specifically, poorer performance on sustained attention (omission and commission) was correlated with smaller hippocampus and putamen, and poorer performance of verbal delayed memory was correlated with smaller putamen [35]. Due to the small sample size of this pilot study, a NIDA-supported study is being performed in a larger cohort of children with and without prenatal methamphetamine exposure in order to determine whether these initial observations can be replicated.


Table 3.  Changes in dopaminergic and serotonergic markers in chronic methamphetamine users.
StudySubjectsGender (male/female)Age (years)Meth usage variablesPET tracer(s) and binding siteRegions examinedFindings
McCann et al. [4]6 abstinent meth. users3M/3F37 ± 8Duration: 3–20 years;[11C]WIN-35 428 (dopamine transporters)Caudate, putamenDecreased DAT density in the caudate nucleus (−23%) and putamen (−25%)
10 controls4 M/6 F30 ± 10Abstinence ∼3 years
Volkow et al. [7]15 abstinent meth. users6M/9F32 ± 6Duration 11 ± 6 years; used ∼1.6 ± 3 g/day; last life-time exposure 12 290 ± 22 396 g; last use ∼5.9 ± 9 months[11C] D-threomethyl-phenidate (dopamine transporters)Caudate, putamen, ventral striatumReduced striatal DAT density in caudate (−27.8%) and putamen (−21%); DAT density correlated with years of meth. use in caudate
18 non-drug users12M/9F31 ± 7
Volkow et al. [15]5 abstinent meth. users2M/3F29 ± 3Early abstinence 3 ± 1.6 months;[11C]-D-threomethyl phenidate (dopamine transporters)Caudate, putamen, ventral striatumNormalization (increased) striatal DAT density from early to protracted abstinence of + 19% in CA; + 16% in putamen; greater meth. used (gm/day) correlated with lesser DAT change (recovery); DAT recovery correlated with duration of abstinence
11 non-drug users4M/7F31 ± 7Longer abstinence: 14 ± 2 months; abstinence 17 (10) months
 1M/4F35 ± 3
Volkow et al. [52]15 abstinent meth. users6M/9F32 ± 7Duration of use 10 ± 6 years;[11C] raclopride (D2 receptors) and [18]FDGD2 receptors in caudate and putamen in relation to glucose metabolism in whole brainDecrease D2 receptor density caudate (−16%) and putamen (−10%); decrease D2 receptor density in putamen correlated with increase obitofrontal cortex metabolism
21 controls15M/6F31 ± 7Life-time used: 13 ± 20 kg
Sekine et al. [47]11 meth. users11 M27 ± 6Duration of use: 4.8 ± 4.8 years; duration of abstinence 5.6 ± 5.7 months[11C]WIN-35 428 (dopamine transporters)Striatum, NAC, prefrontal cortexDecreased DAT density in caudate/putamen (−20.2%), in NAC (−29.6%) and in prefrontal cortex (−33.3%) in meth. users; DAT correlated negatively with duration of meth. use, and with positive symptoms of BPRS (subscale and total BPRS score)
11 comparison11 M27 ± 4
Sekine et al. [48]11 meth. users11 M27 ± 6Duration of use 4.8 (4.8) years; abstinence 5.6 (5.7) months[11C]WIN-35 428 (dopamine transporters)Orbitofrontal cortex (OFC), prefrontal cortex (PFC)Decrease DAT density in OFC, PFC and amygdala in meth. users; DAT density correlated negatively with duration of METH use and score on subscale for positive symptoms on BPRS
9 comparison9 M27 ± 4
Sekine et al. [6]12 meth. users7M/5F31 ± 7Duration of use 6.7 (3.2) years; duration of abstinence 1.6 (1.3) years[11C](+)McN-5652 (serotonin transporters)Midbrain, thalamus, caudate, putamen amygdala, anterior cingulate temporal cortex, OFC, DLPFC cerebellumDecrease in 5HT-T density in all 10 regions; duration of meth. use correlated inversely with 5-HT-T density in 5 brain regions; 5HT-T density also correlated negatively with scores on aggression scale in 8 brain regions
12 comparison7M/5F32 ± 7
Johanson et al. [50]15 meth. users11M/4F32 (21–48)Toxic symptoms at least 3 months;[11C]methyl-phenidate (dopamine transporters)Caudate, anterior putamen, posterior putamenMeth. users showed ∼15% lower DAT and 10% lower VMAT2 across striatal subregions; meth. subjects performed more poorly than controls on 3 of the 12 tasks (digit symbol substitution test, California verbal learning test, and rapid visual information processing)
16 controls10M/6F31 (18–48)Duration ∼10 years Abstinent ∼3 years (3 months–18 years); ∼1/3 also dependent on cocaine, alcohol and marijuana[11C]dihydrotetrabenazine (VMAT2)

Positron emission tomography (PET) studies can be used to assess dopaminergic function (D2-receptors, dopamine and vesicular monoamine transporter or VMAT), serotonergic function and brain glucose metabolism in individuals who used methamphetamine. The same techniques have also been applied to pre-clinical studies of non-human primates.

PET studies were performed in four different laboratories in various methamphetamine subject populations (Table 3). All these studies found decreased DAT density (Fig. 2), between 15 and 28% in the striatum, associated with chronic methamphetamine use. The tracers used to evaluate DAT included [11C]WIN-35 428 [4,47,48] and [11C]-D-threomethylphenidate [5]. The finding of decreased DAT in those who used methamphetamine is remarkably consistent in these cross-sectional samples, despite the differences in radiotracers and in subject characteristics. For instance, the period of abstinence from methamphetamine use ranged between 3 months and 20 years among studies. These in vivo findings are also in agreement with post-mortem studies of humans who used methamphetamine chronically that found significant reductions in DAT, as well as tyrosine hydroxylase and post-synaptic D2-receptors. However, the post-mortem studies found no significant loss of DOPA decarboxylase, VMAT or pigmented neuronal cell bodies. Therefore, the data suggest that methamphetamine toxicity may not permanently damage dopaminergic neurons; rather, the changes in DAT and D2 receptors reflected primarily neuroadaptive changes [49]. One in vivo study that followed a group of subjects who used methamphetamine longitudinally during abstinence indeed found that striatal DAT density normalized (increased) from early (< 6 months) to protracted (12–17 months) abstinence (+19% in CA; +16% in putamen) [15]. Therefore, it is possible that the initially decreased DAT density represents neuroadaptation, such as down-regulation of the DAT or internalization into the cell membrane with loss of visibility to PET tracers. In this scenario, the transporter density then slowly returns to near-normal values during prolonged abstinence.

Figure 2.

Left: PET images with C-11 cocaine ligands showing lower dopamine transporter (DAT) binding in an individual who abused methamphetamine compared to a non-drug-user control subject; both are 33-year-old males. Middle: bargraphs showing decreased in DAT in the methamphetamine subject compared to controls in the caudate and putamen; right: correlations between neuropsychological tests and DAT showing slower gross motor function and poor memory recall in subjects with lower DAT. (This figure is modified from Volkow et al. [52])

A recent in vivo PET study used [11C]dihydrotetrabenazine to evaluate VMAT2, and found 10% lower VMAT2 across striatal subregions of individuals who used methamphetamine chronically (for an average of 10 years), but had been abstinent for an average of 3 years [50]. The finding of decreased VMAT contrasts with the post-mortem study of human methamphetamine users discussed above [49]. One possible explanation for the difference in findings between the in vivo and the post-mortem study might be related to differences in the subject population. Specifically, the methamphetamine subjects in the in vivo study were dependent upon other drugs of abuse (including alcohol, cocaine and marijuana), and those in the post-mortem study might have been younger.

In most of these PET studies, decreased DAT density was associated with greater amount and/or duration of methamphetamine usage [15,47]. Furthermore, individuals who had used the greatest amount of methamphetamine (gm/day) showed the least recovery in DAT during abstinence [15]. The decreased DAT was also found to be associated with poorer performance on several cognitive tests, including timed gait, grooved pegboard and performance on the Auditory Verbal Learning Test [5]. However, in contrast to the recovery of the DAT after protracted abstinence (mean = 15 months), cognitive deficits persisted in these individuals who used methamphetamine, which suggests either insufficient neuronal terminal recovery or that other neural network or systems that affect neurocognitive functioning may take yet longer to recover from the toxic effects of methamphetamine.

Chronic methamphetamine use is often accompanied by psychiatric symptoms that are thought to be associated with changes in specific neurochemicals, such as DAT. Sekine et al., using the DAT ligand [11C]WIN 35 428, also found decreased striatal DAT during early abstinence in adults who chronically abused methamphetamine in Japan [47]. In addition, these investigators reported decreased DAT in the prefrontal and orbitofrontal cortices [48]. Furthermore, the decreased DAT correlated with the BPRS (described above). This suggests that changes in DAT in the striatum may be associated with the severity and duration of residual psychiatric symptoms in these individuals.

In addition to the dopamine system, the serotonergic system has been implicated in pre-clinical studies of chronic methamphetamine exposure. Only one study by Sekine et al. used the ligand ([11C]McN-5652), which has a high binding specificity to the pre-synaptic serotonin transporter (SERT), to evaluate a cohort of adults who chronically abused methamphetamine during protracted abstinence (mean = 1.6 years). SERT density was decreased in all 10 pre-defined brain regions (Table 3). Decreased SERT density in five of these brain regions (midbrain, thalamus, caudate nucleus, putamen and orbitofrontal cortex) was associated with longer duration of methamphetamine use, while SERT density in eight brain regions (thalamus, caudate nucleus, putamen, anterior cingulate cortex, temporal cortex, orbitofrontal cortex, dorsolateral prefrontal cortex and cerebellar cortex) correlated negatively with the severity of aggression score measured by the Aggression Questionnaire scale. Furthermore, whole-brain statistical parametric mapping analysis showed that the severity of aggression was associated with decreased SERT density in the orbitofrontal cortex, anterior cingulate cortex and temporal cortex. Surprisingly, changes in SERT density did not correlate with other clinical measures of psychiatric symptoms including psychosis (BPRS), depression (Hamilton Rating Scale for Depression) or anxiety (Hamilton Rating Scale for Anxiety).


Table 4.  Changes in brain glucose metabolism in chronic methamphetamine users.
StudySubjectsGender (male/female)Age (years)Meth usage variablesPET tracer(s) and binding siteRegions examinedFindings
Volkow et al. [52]15 abstinent meth. users6M/9F32 ± 712 early abstinent (range 0.5–5 months);[18]FDG (glucose metabolism)Striatum (caudate and putamen), thalamus, cortical regions, cerebellumGlobal glucose metabolism higher (+14%) in meth. users; parietal metabolism (+20%); striatal metabolism (caudate −12%, putamen −6%); thalamus (−17%); increased parietal metabolism associated with poorer performance on grooved pegboard
21 controls15M/6F31 ± 73 protracted abstinent (11–35 months)
Wang et al. [16]12 early abstinent meth. users; 5 after protracted abstinence; 11 non-drug controls2M/3F29 ± 3Early abstinence −3 ± 1.6 months[18]FDG (glucose metabolism)Striatum (caudate and putamen) and thalamusMetabolism of thalamus and striatum less in meth. users; significant increased metabolism in thalamus (+12%) after protracted abstinence and similar to controls; striatum metabolism less in meth. users, with no significant changes after protracted abstinence; improved metabolism in thalamus correlated with improved timed gait, symbol digit modalities and delayed recall
4M/7F31 ± 7Protracted abstinence −14 ± 2 months
London et al. [8]SPM: 17 patients11M/6F34.7(1.87)Days used in last 30–18.85(1.95) days;[18]FDG (glucose metabolism)Striatum, cingulateDecreased MRglc in anterior cingulate and insula and increased CMRglc in lateral orbitofrontal cortex, middle and posterior cingulate cortex, amygdala, ventral striatum and cerebellar vermis associated with higher Beck Depression Inventory and state–trait anxiety scores in meth. users
CMRglu: 14 patients10M/4F34.5(2.14)days used in last 30–18.1 (2.25) days
SPM: 18 comparison10M/8F32.3(1.91) 
CMRglc: 13 controls8M/5F32.6(2.48) 
London et al. [56]17 abstinent meth.11M/7F35 ± 2Meth. used: ∼9.3 ± 1.1 years; ∼3.6 g/week; ∼19 of last 30 days[18]FDG (glucose metabolism)Cingulate regions (anterior and middle); insulaMeth. users differ from controls in the relationships between auditory CPT performance and relative glucose metabolism of the anterior nd middle cingulated gyrus and the insula
16 controls10M/6F33 ± 2

Brain function, as assessed by glucose metabolism, was also evaluated in subjects who chronically abused methamphetamine in several studies. The first study documented global increases in glucose metabolism in methamphetamine subjects. After normalization of regional values to whole-brain metabolism, increased glucose metabolism reached significance in the parietal brain regions but relative decreases in metabolism were seen in the striatum (greater decrease in caudate than in putamen) and in the thalamus [7] (Fig. 1). The finding of increased parietal metabolism was interpreted as reflecting a glial response to brain injury, as glial cells have higher metabolism than neurons and methamphetamine administration has been shown to cause damage to pyramidal glutamatergic neurons in the parietal cortex [51]. Furthermore, individuals with higher parietal metabolism had slower performance on the grooved pegboard test, which suggests that inflammatory changes in the parietal region may lead to poor performance on this task and possibly other tasks that would require parietal function (e.g. attention and visuospatial skills). This observation might also be related to the finding of poorer cognitive performance in subjects with increased parietal volumes, as observed in another cohort of chronic methamphetamine users [13] (see discussion above). Together, these results indicate that an inflammatory or glial response may occur with chronic methamphetamine use, especially in the parietal cortices. A follow-up study of these subjects after protracted abstinence revealed recovery of thalamic metabolism but significant long-lasting reductions in striatal metabolism, which remained depressed even after > 9 months of detoxification [16]. In these subjects the recovery in thalamic glucose metabolism correlated with improved performance on cognitive tests (including timed gait, symbol digit modalities and delayed recall on the Rey Auditory Verbal Learning Test). In these subjects, the increased metabolic activity in parietal cortex persisted afterdetoxification and was most accentuated in the precuneus. Inasmuch as the precuneus is a brain regions involved with the level of arousal, the persistent abnormalities would suggest long-lasting disruption in alertness in methamphetamine abusers. The results from these brain metabolic studies corroborate the findings from DAT studies showing that some of the brain abnormalities improve during prolonged abstinence, but others do not. Moreover, in the case of the studies where subjects were evaluated both with a DAT ligand and with brain glucose metabolism showing recovery of DAT levels in striatum but persistent abnormalities in metabolism in striatum indicate that the increases in DAT levels with abstinence are insufficient to return striatal function to normal [15]. This again is reflected by the improper recovery of motor and memory performance in these subjects.

In order to determine whether changes in the dopaminergic system are related to brain function in specific brain regions, Volkow et al. measured D2 receptor binding, using C-11 raclopride, and explored its relationship to brain glucose metabolism in various brain regions [52]. D2 receptors are typically located post-synaptically from dopaminergic neurons, such as on γ-aminobutyric acid (GABA) cells in the striatum or on glia. Individuals chronically used methamphetamine had decreased D2 receptors (Fig. 3), due possibly to down-regulation of the receptors in response to chronic, methamphetamine-mediated release of dopamine into the synapses [52]. Alternatively, individuals who used methamphetamine could have had low levels of D2 receptors which predisposed them to drug use [52], as had been observed in individuals addicted to other drugs, such as cocaine [53], heroin [54] and alcohol [55]. Furthermore, the D2 receptor level in those who abused methamphetamine was associated with orbitofrontal glucose metabolism; this suggests that methamphetamine-induced changes in the dopaminergic system might in turn alter the dopamine-mediated striatal regulation of the orbitofrontal cortex, via the striato–thalamo–cortical pathway. However, abnormalities in the reciprocal projections from the orbitofrontal cortex to the nucleus accumbens may also influence the dopaminergic system [52]. In either scenario, the combined pathology of abnormal orbitofrontal glucose metabolism and decreased D2 receptors might perpetuate the compulsive drug-taking behavior in subjects who chronically abuse methamphetamine [52].

Figure 3.

Left: PET images with C-11 raclopride ligands showing lower dopamine D2 receptor binding in an individual who abused methamphetamine compared to a control subject. Right: scatterplots showing lower group average of dopamine D2 receptors in the methamphetamine subject. (This figure is modified from Volkow et al. [52])

Psychiatric symptoms including depression and anxiety [8], psychosis [41,47] and aggression [6] have been reported in adults who used methamphetamine during periods of early and protracted abstinence. Signs of depression and hypersomnolence are especially common during early abstinence. London et al. studied a group of individuals who chronically used methamphetamine during early abstinence (4–7 days), and indeed found increased severity of depressive and anxious symptoms (as measured by self-report on the Beck Depression Index and State–Trait Anxiety Index, respectively). The self-rating of anxiety symptoms was associated inversely with relative glucose metabolism in the anterior cingulate cortex, insula and orbitofrontal cortex, while depressive symptoms covaried positively with relative metabolism in the limbic regions [8]. These relationships suggest that these brain regions are involved in affective dysregulation during early abstinence and withdrawal from methamphetamine use. Future studies are needed to determine whether these changes are persistent with prolonged abstinence. The investigators demonstrated further that individuals who used methamphetamine had higher error rates during the performance of a vigilance task during the PET studies. The methamphetamine subjects' performance was associated negatively with metabolism in the cingulate regions, while the control subjects had a positive association with the metabolism in this region [56]. Those with lower regional brain glucose metabolism and smaller structures in the hippocampus also had a higher error rates on the vigilance task during the PET studies. These findings document further that abnormal metabolism in the limbic, paralimbic and hippocampal regions may affect cognitive performance.

Only one study evaluated the combined effects of methamphetamine and marijuana, which is abused commonly by individuals who abused methamphetamine as well. Methamphetamine subjects with concurrent low frequency of marijuana use (∼8 days/month) had lower regional glucose metabolism in the temporal and parahippocampal regions than those without concurrent marijuana use [57]. These findings highlight the importance of controlling for, or excluding use of other concurrent drugs in imaging studies of individuals who used methamphetamine. For instance, cigarette smoking and alcohol use are also very common among drug users, and future studies need to evaluate the interactive and combined effects of nicotine and alcohol with methamphetamine on brain function.


Methamphetamine use is common among HIV-infected individuals, and has contributed to higher rates of risky sexual behavior and hence the risk of contracting sexually transmitted diseases, including HIV [58,59]. Because both methamphetamine and HIV show independent injurious effects on the basal ganglia structures, the combined effects of methamphetamine and HIV were evaluated in several recent studies. One morphometric study found that HIV infection was associated with volume loss in the striatal structures, while methamphetamine was associated with increased volume of these structures; therefore, an interaction effect was observed in those with both conditions [13].

MR spectroscopy demonstrated further comorbid effects of HIV and methamphetamine on neurometabolites in three brain regions evaluated, but the abnormalities were greatest in the basal ganglia. Compared with subjects with either HIV or methamphetamine alone, those with both conditions showed additional decreases in NAA and CR in the basal ganglia, and additional increases in CHO and MI in the frontal white matter [39] (Table 2). The additive effects on basal ganglia NAA and CR suggest that HIV patients who abuse methamphetamine have greater brain injury in the basal ganglia. This is not surprising, given that HIV patients, especially those with dementia [60], as well as individuals who abused methamphetamine show decreased dopamine markers on PET studies that evaluate dopamine transporter densities (see Table 3 regarding PET studies in methamphetamine subjects). Additive effects on dopamine transporter loss would also be expected and should be evaluated.


Neuroimaging techniques demonstrate a wide range of abnormalities in the structure and chemistry of those who abused methamphetamine, particularly in the striatal regions. On morphometry, a consistent finding is striatal enlargement in adults who used methamphetamine. MR spectroscopy additionally demonstrates reduced concentrations of NAA and CR in the basal ganglia. Furthermore, PET studies show a consistent pattern of reduced dopamine transporter (DAT) density in the striatum of subjects who used methamphetamine, as well as reductions in DA D2 receptors in the striatum of methamphetamine abusers. While the PET findings are consistent with the well-known dopaminergic effects of methamphetamine, the etiology of the structural and MRS abnormalities is less clear. Magnetic resonance studies have also explored the effects of in utero exposure to methamphetamine on brain structure and chemistry in children; however, imaging as well as clinical studies on the effects of prenatal methamphetamine exposure on brain development are still scarce.

Although the published studies demonstrate convincingly abnormalities in brain structure and chemistry in those who abused methamphetamine, many important questions remain. For instance, while some of the neuroimaging measures correlated with methamphetamine usage variables, a systematic investigation of these relationships would require larger sample sizes. Similarly, while a few studies have demonstrated apparent normalization of some imaging variables during continued abstinence, typically over many months, whether these abnormalities are truly reversible and how long this process might take would require much more detailed longitudinal studies. Another important area of investigation should include evaluations of genetic polymorphism of various monoamine markers, and their relationships to various neuroimaging parameters, in individuals who use methamphetamine. Such genetic-imaging correlations may help to determine whether particular genotypes are more susceptible to brain changes associated with methamphetamine-associated neurotoxicity, and to determine further these relationships to drug dependence or relapse.

Finally, very little is known about the relationship between imaging variables and the long-term psychiatric and cognitive sequelae of prolonged methamphetamine abuse. One important, unanswered question is whether neuroimaging markers during early abstinence can be used to predict treatment success or, conversely, risk of relapse in individual methamphetamine users. In the case of the striatal abnormalities in methamphetamine abusers, this is particularly relevant in that there is increasing evidence of the involvement of the dorsal striatum in addiction [61], both through its involvement with conditioned responses [62,63] and with habit learning [64]. Similarly, it is unclear whether neuroimaging variables might provide surrogate markers to monitor brain function and chemistry during treatment trials.


We received support for this work from the National Institute on Drug Abuse (NIDA K24-DA16170 for L. Chang; K02-DA16991 for T. Ernst) joint support from NIDA and the National Institute of Neurological Disorders and Stroke (2U54 NS3906-06 for D. Alicata).