- Email: [email protected]
Hippocampal glutamate is increased and associated with risky drinking in young adults with major depression
Journal of Affective Disorders 186 (2015) 95–98
Contents lists available at ScienceDirect
Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad
Short communication
Hippocampal glutamate is increased and associated with risky drinking in young adults with major depression Daniel F. Hermens a,n, Kate M. Chitty a, Rico SC Lee a, Ashleigh Tickell a, Paul S. Haber b,c, Sharon L. Naismith a, Ian B. Hickie a, Jim Lagopoulos a a
Clinical Research Unit, Brain and Mind Research Institute, University of Sydney, Camperdown, Australia Drug Health Services, Royal Prince Alfred Hospital, Camperdown, Australia c Sydney Medical School, University of Sydney, Camperdown, Australia b
art ic l e i nf o
a b s t r a c t
Article history: Received 10 April 2015 Received in revised form 2 July 2015 Accepted 6 July 2015 Available online 26 July 2015
Background: Risky drinking in young people is harmful, highly prevalent and often complicated by comorbid mental health problems that compound alcohol-induced impairment. The hippocampus and the glutamate system have been implicated in the pathophysiology of alcoholism and depression. This study aimed to determine whether risky drinking is associated with glutamate levels recorded within the hippocampus of young adults with major depression. Methods: Sixty-three young persons with major depression (22.1 73.1 years; 65% female) and 38 healthy controls were recruited. Participants completed the alcohol use disorder identification test and underwent proton magnetic resonance spectroscopy to measure in vivo glutamate levels within the hippocampus following a period of at least 48 h of abstinence. Results: Young adults with depression had significantly increased hippocampal glutamate levels and a positive association between the level of alcohol use and glutamate. Regression analysis revealed that higher levels of hippocampal glutamate were predicted by having increased levels of risky drinking and depression. Limitations: Small sample sizes for testing diagnosis by risky drinking interaction and use of creatine ratios rather than the absolute concentrations of glutamate. Discussion: The hippocampus is a critical region; given its role in learning and memory as well as mood regulation, and the neurochemical changes observed in this study may precede structural changes, which are commonly observed in both depression and alcohol misuse. These findings suggest that young adults with major depression who engage in risky drinking may be at increased risk of glutamate excitotoxicity. & 2015 Published by Elsevier B.V.
Keywords: Hippocampus Depression Alcohol Glutamate Magnetic resonance spectroscopy
1. Introduction Beyond determining the high prevalence of comorbid depression and risky drinking, little is known about the early neurobiological changes associated with this comorbidity. Importantly, the hippocampus has been implicated in the pathogenesis of both depression (Mannie et al., 2014) and alcohol use disorders (White and Swartzwelder, 2004). Acute alcohol exposure inhibits glutamate binding and reduces the transmission efficacy of cortical neurons via the suppression of N-methyl-D-aspartate receptor (NMDAr) activity (Strelnikov, 2007). With chronic alcohol use, NMDAr binding sites increase in number and level of functioning; as demonstrated in rodents who show increased glutamate transmission in the hippocampus after repeated ethanol n
Corresponding author. Fax: þ 61 2 93510652. E-mail address: [email protected] (D.F. Hermens).
http://dx.doi.org/10.1016/j.jad.2015.07.009 0165-0327/& 2015 Published by Elsevier B.V.
administration (Chefer et al., 2011). Moreover, converging evidence indicates that aberrations in glutamate homoeostasis and neurotransmission have a significant role in the development of depression. Post mortem investigation of glutamate derived from depressed patients has revealed elevated glutamate in the frontal cortex (Hashimoto et al., 2007) and studies on brain tissue from depressed patients have found down-regulation of genes that code for the excitatory amino acid transporters which reside on the glia and are responsible for clearing glutamate from the synapse (Beart and O'Shea, 2007). As a consequence of altered sensitivity, glutamate may accumulate at the synapse resulting in excitotoxicity (Meyerhoff et al., 2013). As a means to assess in vivo neurochemistry, proton magnetic resonance spectroscopy (1H-MRS) has provided important neurobiological insights into both depression and alcohol use. In terms of glutamate, 1H-MRS studies of depression samples have been mixed with evidence of decreased frontal (Hasler et al., 2007), normal frontal (Taylor et al., 2009), normal occipital (Godlewska
96
D.F. Hermens et al. / Journal of Affective Disorders 186 (2015) 95–98
et al., 2015) and increased occipital (Sanacora et al., 2004) levels. Of note, increased hippocampal glutamate has been reported in young people with a family history of depression (Mannie et al., 2014). In 1H-MRS studies of alcohol misuse, increased frontal glutamate has been reported in dependant humans (Hermann et al., 2012; Lee et al., 2007) and rats (Hermann et al., 2012). Similarly, treatment-seeking heavy drinkers show increased frontal glutamate; associated with duration of drinking (Yeo et al., 2013). In contrast, decreased glutamate levels have been found in the visual cortex of alcohol-dependent (Bagga et al., 2014) and frontal regions of binge (Silveri et al., 2014) and heavy (Ende et al., 2013) drinkers. Overall, the variability in the aforementioned 1H-MRS findings may be explained by key factors such as region of interest and duration of exposure/illness. In this study, we hypothesised that young adult outpatients with major depressive disorder (MDD) would show increased levels of in vivo hippocampal glutamate levels and this would be associated with their levels of alcohol use.
2. Methods Sixty-three outpatients (aged 18–30 years) with a primary DSM-IV diagnosis of MDD were recruited from a specialised service for young people with mental health problems (Scott et al., 2012). At the time of assessment, all patients were receiving clinician-based case management and their psychotropic medication regimens were as follows: 27% (17/63) were not taking psychotropic medications; 63% (40/63) were taking a third-generation anti-depressant; 17% (11/63) an atypical antipsychotic medication; and 9% (6/63) were taking a mood stabiliser. Healthy controls (N ¼ 38; aged 18–29 years) were recruited from the community and screened to ensure they had no history of a mental disorder. Exclusion criteria for all participants were medical instability, history of neurological disease, medical illness known to impact cognitive and brain function, intellectual and/or developmental disability, current substance dependence (not including alcohol) and insufficient English for assessment. All participants were asked to abstain from drug or alcohol use for 48 h prior to testing and informed that they may be asked to under-take an alcohol breath test and/or a saliva drug screen. The University of Sydney Human Research Ethics Committee approved the study and all participants gave written informed consent. A trained research psychologist conducted the BMRI Structured Interview for Neurobiological Studies (Lee et al., 2013) to determine the nature and history of any mental health problems. The interview included the Hamilton Depression Rating Scale (HDRS, 17-item) to quantify mood symptoms at the time of assessment; the Brief Psychiatric Rating Scale (BPRS) to quantify general psychiatric symptoms at the time of assessment; and, the Social and Occupational Functioning Assessment Scale (SOFAS); where a patient's functioning is rated from 0 to 100, with lower scores suggesting more severe impairment. As part of a self-report questionnaire, participants completed: the Kessler-10 (K-10), as a measure of psychological distress; the first two items of the World Health Organisation's (WHO) 'alcohol, smoking and substance involvement screening test' to assess lifetime and current substance use; and the WHO Alcohol Use Disorders Identification Test (AUDIT) to assess level of risky drinking in the past year, as well lifetime familiarity. 1 H-MRS data was acquired on a 3Tesla GE Discovery MR750 MRI scanner (GE Medical Systems, Milwaukee, WI), using an 8-channel phased array headcoil. The protocol comprised of a 3D sagittal whole-brain scout (TR ¼ 50 ms; TE ¼4 ms; matrix ¼256; no averaging, z ¼5 mm thickness), a T1-weighted MPRAGE sequence for anatomical localisation (TR ¼7.2 ms; TE ¼2.8 ms; flip
angle ¼10°; matrix 256 256; 0.9 mm isotropic voxels, 196 slices) and single-voxel 1H-MRS using PRESS acquired from a 1.5 3.0 1.0 cm3 voxel placed in the left hippocampus. Unsuppressed water scans (acquired from the same voxel) were collected prior to acquisition of the metabolite scans. Eddy current correction as implemented by LCModel package (Provencher, 1993) was undertaken on the metabolite data using the unsuppressed water FIDs as reference. Spectra were then quantified with LCModel, using a PRESS TE ¼35 basis set of 15 metabolites. Anatomical localisation of voxel placement was based on the Talaraich brain atlas and positioning was guided by the MPRAGE image. All spectra were shimmed to achieve full-width half maximum o13 Hz. Poorly fitted metabolite peaks as reflected by Cramer–Rao lower bounds 420 were excluded from further analysis. Statistical analyses were then conducted on glutamate (GLU) level as a ratio over water-scaled creatine (Cr) concentration. Segmentation of left hippocampal volumes was also undertaken as per our previously published methods (Hermens et al., 2015). Two-tailed independent t-tests were used to assess group differences for demographic, clinical and spectroscopic variables. If homogeneity of variance was violated Welch's corrected degrees of freedom and p-values were reported. Analysis of Covariance (ANCOVA) was employed to control for the potential effects of age. Pearson product moment correlations were used to examine for significant associations between total AUDIT scores and GLU levels, with partial correlations controlling for age. In order to further analyse the relationships between alcohol use (AUDIT total score), diagnosis, demographics (sex, age, family history) and glutamate, we performed multiple linear regression analyses with six predictor variables ('enter' method) in the model. Current nicotine smoking status was included as a predictor as this has been shown to affect 1H-MRS variables (Meyerhoff et al., 2013).
3. Results Table 1 summarises between-group comparisons across variables. There was no difference in gender ratio among groups; however, there was a significant difference in age (po .05; control group older). Furthermore, the groups differed significantly in years of education but did not differ in predicted IQ scores. In terms of alcohol-related variables, there were no group differences in age of first alcoholic drink or AUDIT total score, however the depression group had a significantly (p o.05) higher proportion of individuals with a family history of alcoholism. Controls had significantly higher social and occupational functioning (SOFAS) as well as lower self-reported psychological distress (K-10), current depressive (HDRS) symptoms and general psychiatric (BPRS) clinical ratings (all p o.001). The groups did not differ in the proportion of daily nicotine smokers. The groups did not differ in terms of creatine levels or hippocampal volumes; however, they were significantly (p o.05) different in GLU/Cr, with the depression group having increased levels (Table 1). ANCOVA confirmed this main effect remained significant when controlling for age (p o.05), years of education (p o.05), family history of alcoholism (po .05) or hippocampal volume (po .05). For the depression group, total AUDIT score was positively associated with GLU/Cr levels (r ¼ 0.429, n ¼62, po 0.001), and remained significant (po .001) after controlling for age or hippocampal volume. The controls showed no significant correlations between these measures. Fig. 1 depicts the associations between GLU/Cr levels and AUDIT scores for each group. The multiple linear regression model for hippocampal GLU/Cr level was significant [F(6, 97) ¼3.3, p o.01] and explained 18.1% of the variance (adjusted R2 ¼0.126) with three significant predictors:
D.F. Hermens et al. / Journal of Affective Disorders 186 (2015) 95–98
97
Table 1 Mean scores (7 standard deviation) for demographic, clinical and spectroscopic variables for depression versus controls.
Sex (% female) FH Alc (% þ ve) Current nicotine use (% daily) Age, yrs Education, yrs Predicted IQ Age, first drink AUDIT, total SOFAS K-10 total HDRS total BPRS total GLU/Cr
Depression (N ¼ 63)
Controls (N ¼38)
Between-group differences
65.1% 19.7% 17.5% 22.17 3.1 13.0 72.0 104.77 8.2 15.2 72.1 7.0 7 6.7 65.4 7 10.8 27.2 7 7.3 13.8 76.2 38.0 7 6.3 1.58 7 0.20
65.8% 2.6% 13.2% 23.3 7 2.6 14.4 7 2.3 105.2 7 7.2 15.2 7 1.6 8.9 7 4.9 89.3 7 5.9 14.8 7 3.9 1.9 7 2.6 25.9 7 2.8 1.477 0.22
χ2¼ 0.1, df¼ 101, p ¼ .942 χ2¼ 6.0, df ¼99, p¼ .015 χ2¼ 0.3, df¼ 101, p ¼ .566 t¼ 2.1, df ¼99, p¼ .035 t¼ 3.0, df ¼ 99, p¼ .003 t¼ 0.3, df ¼95, p¼ .771 t¼ 0.1, df ¼92, p ¼ .953 t¼ 1.5, df¼ 99, p ¼.127 t¼ 13.9, df ¼ 92.7, p¼ .000 t¼ 11.1, df ¼97.8, p¼ .000 t¼ 13.1, df ¼88.8, p ¼.000 t¼ 12.2, df ¼88.2, p¼ .000 t¼ 2.5, df ¼98, p¼ .016
Between group differences were tested by chi-square or t-test. Abbreviations are as follows: FH Alc ¼ family history of alcoholism; AUDIT ¼ alcohol use disorders identification test; SOFAS¼ social and occupational functioning assessment scale; K-10¼ Kessler 10-item; HDRS ¼Hamilton depression rating scale; BPRS ¼ brief psychiatry rating scale; GLU ¼ glutamate; Cr¼ creatine.
Fig. 1. Scatterplot of hippocampal glutamate (relative to creatine) level vs. the alcohol use disorder identification test (AUDIT) total score in the depression (black triangles; black solid line) and control (grey circles; grey dashed line) groups.
AUDIT total (β ¼0.268, p o.05), sex (β ¼0.238, p o.05) and diagnosis (β ¼ 0.244, p o.05). In this model, increased levels of risky drinking, being female and having a diagnosis of depression were all associated with increased levels of glutamate. A follow-up regression model was performed with three interaction variables (diagnosis AUDIT total; sex AUDIT total; diagnosis sex) and the three significant predictors (diagnosis, sex, AUDIT total). This model was also significant [F(5, 99) ¼5.7, po .01] and explained 23.3% of the variance (adjusted R2 ¼0.192) with two significant variables: diagnosis (β ¼ 0.208, p o.05) and the interaction of diagnosis and AUDIT total (β ¼ 0.245, p o.05). Of note, sex was excluded from the model.
4. Discussion These findings are consistent with other 1H-MRS studies reporting increased glutamate levels in depression (Li et al., 2014; Sanacora et al., 2004), family history of depression (Mannie et al.,
2014) and alcohol misuse (Lee et al., 2007; Yeo et al., 2013). Of these studies, one targeted the hippocampus (Mannie et al., 2014) while the others assessed either frontal or occipital regions. However, there is also evidence (primarily in frontal regions) of decreased glutamate in depression/alcohol misuse, highlighting the need for further studies of in vivo hippocampal glutamate. Importantly, in rodents, there is evidence that short-term, repeated exposure to moderate doses of alcohol can lead to increases in extracellular glutamate levels in the CA3 region of the hippocampus (Chefer et al., 2011), which is important for the encoding of new spatial information (Kesner, 2007). Prolonged abstinence may also lead to decreased hippocampal glutamate levels in depressed, risky drinkers since there is evidence, in both human and rodent studies, showing that despite a hyper-glutamatergic state during withdrawal there is a subsequent normalisation of glutamate levels (Meyerhoff et al., 2013). It is important to note, that we did not specifically monitor for withdrawal symptoms in the current study as no subjects were diagnosed with alcohol dependence disorder. Thus, future studies of individuals with depression and alcohol dependence who abstain (and hence undergo withdrawal) at the time of 1H-MRS scan are warranted. There were several limitations associated with this study. Firstly, the sample sizes in our study were relatively modest, particularly with respect to potential sex differences. Another limitation is the use of creatine ratios rather than the absolute concentrations of glutamate; although creatine ratios are a robust method of normalising metabolite levels across subjects and have been used extensively in the literature. 1H-MRS is only able to measure the absolute pool of available neurometabolites within a voxel, thus limiting any inferences about whether increased glutamate levels are primarily neuronal, extracellular or post-synaptic. Our 1H-MRS protocol did not allow quantification of glutamine (GLN). However by indexing glutamate-to-glutamine cycling (via GLU/GLN) future studies may shed even more light on the combined effects of depression and alcohol on this system. Finally, it has been suggested that the therapeutic effects of many psychotropic medications are potentially a result of glutamatergic interactions (Sanacora et al., 2008). As such, the contributing effects on the glutamatergic system of the medications taken by subjects in this study cannot be fully ascertained. In conclusion, data from this study suggests that younger people with MDD should be cautious about the excessive use of alcohol. During this period of critical brain development, excessive alcohol use appears to be associated with markers of excitotoxicity, which in turn, is likely to have detrimental effects on cognitive and social functioning. The hippocampus is a particularly
98
D.F. Hermens et al. / Journal of Affective Disorders 186 (2015) 95–98
critical region; given its role in learning and memory as well as mood regulation, and the neurochemical changes observed in this study may precede structural changes, which are commonly observed in both depression and alcohol misuse. Indeed, alcohol may exacerbate the neurobiological changes inherent to MDD, which in turn, is associated with pronounced disability. In order to strengthen and extend these findings, future studies should longitudinally track the concomitant impact of alcohol on the emerging neurobiological changes (including those specifically glutamatergic in nature) seen in young people with depression.
Conflict of interest No conflict declared.
Author disclosures Dr. Hermens has received honoraria for educational seminars from Janssen-Cilag and Eli Lilly. Prof. Hickie has led a range of community-based and pharmaceutical industry-supported depression awareness and education and training programmes. He has led depression and other mental health research service evaluation or investigator-initiated research projects that have been supported by a variety of pharmaceutical partners. Servier and Pfizer support current investigator-initiated studies. He has received honoraria for his contributions to professional educational seminars related to depression, youth mental health and circadian-rhythms research. He has received travel support from Servier to attend scientific meetings related specifically to circadianrhythm disorders. The authors report no other biomedical financial interests or potential conflicts of interest.
Contributors DFH, JL, SLN and IBH contributed to the conception of the study; DFH, JL, KMC and AT undertook the data processing and statistical analysis; DFH wrote the first draft of the manuscript. All authors contributed to data interpretation, discussion and have approved the final manuscript.
Role of the funding source This work was funded by a grant from the NSW Ministry of Health, Mental Health and Drug and Alcohol Office as well as National Health and Medical Research Council program (No. 566529) and Centres of Clinical Research Excellence (No. 264611) grants. DFH was supported by a grant from the NSW Ministry of Health, Mental Health and Drug and Alcohol Office; SLN was supported by an NHMRC Career Development Award (No. 402864); and IBH was supported by an NHMRC Senior Principal Research Fellowship (No. 1046899). These funding agencies had no role in the analysis or interpretation of these data; or in the writing of this report; or the decision to submit the paper for publication.
Acknowledgement The authors would like to express their gratitude to individuals who participated in this study.
References Bagga, D., Khushu, S., Modi, S., Kaur, P., Bhattacharya, D., Garg, M.L., Singh, N., 2014. Impaired visual information processing in alcohol-dependent subjects: a proton magnetic resonance spectroscopy study of the primary visual cortex. J. Stud. Alcohol Drugs 75, 817–826. Beart, P.M., O'Shea, R.D., 2007. Transporters for L-glutamate: an update on their molecular pharmacology and pathological involvement. Br. J. Pharmacol. 150, 5–17. Chefer, V., Meis, J., Wang, G., Kuzmin, A., Bakalkin, G., Shippenberg, T., 2011. Repeated exposure to moderate doses of ethanol augments hippocampal glutamate neurotransmission by increasing release. Addict. Biol. 16, 229–237. Ende, G., Hermann, D., Demirakca, T., Hoerst, M., Tunc-Skarka, N., Weber-Fahr, W., Wichert, S., Rabinstein, J., Frischknecht, U., Mann, K., Vollstadt-Klein, S., 2013. Loss of control of alcohol use and severity of alcohol dependence in nontreatment-seeking heavy drinkers are related to lower glutamate in frontal white matter. Alcohol. Clin. Exp. Res. 37, 1643–1649. Godlewska, B.R., Near, J., Cowen, P.J., 2015. Neurochemistry of major depression: a study using magnetic resonance spectroscopy. Psychopharmacology (Berl) 232, 501–507. Hashimoto, K., Sawa, A., Iyo, M., 2007. Increased levels of glutamate in brains from patients with mood disorders. Biol. Psychiatry 62, 1310–1316. Hasler, G., van der Veen, J.W., Tumonis, T., Meyers, N., Shen, J., Drevets, W.C., 2007. Reduced prefrontal glutamate/glutamine and gamma-aminobutyric acid levels in major depression determined using proton magnetic resonance spectroscopy. Arch. Gen. Psychiatry 64, 193–200. Hermann, D., Weber-Fahr, W., Sartorius, A., Hoerst, M., Frischknecht, U., TuncSkarka, N., Perreau-Lenz, S., Hansson, A.C., Krumm, B., Kiefer, F., Spanagel, R., Mann, K., Ende, G., Sommer, W.H., 2012. Translational magnetic resonance spectroscopy reveals excessive central glutamate levels during alcohol withdrawal in humans and rats. Biol. Psychiatry 71, 1015–1021. Hermens, D.F., Naismith, S.L., Chitty, K.M., Lee, R.S.C., Tickell, A., Duffy, S.L., Paquola, C., White, D., Hickie, I.B., Lagopoulos, J., 2015. Cluster analysis reveals abnormal hippocampal neurometabolic profiles in young people with mood disorders. Eur. Neuropsychopharmacol. 25, 836–845. Kesner, R.P., 2007. Behavioral functions of the CA3 subregion of the hippocampus. Learn. Mem. 14, 771–781. Lee, E., Jang, D.-P., Kim, J.-J., An, S.K., Park, S., Kim, I.-Y., Kim, S.I., Yoon, K.-J., Namkoong, K., 2007. Alteration of brain metabolites in young alcoholics without structural changes. Neuroreport 18, 1511–1514. Lee, R.S.C., Hermens, D.F., Redoblado-Hodge, M.A., Naismith, S.L., Porter, M.A., Kaur, M., White, D., Scott, E.M., Hickie, I.B., 2013. Neuropsychological and socio-occupational functioning in young psychiatric outpatients: a longitudinal investigation. PLoS ONE 8, e58176. Li, M., Metzger, C.D., Li, W., Safron, A., van Tol, M.J., Lord, A., Krause, A.L., Borchardt, V., Dou, W., Genz, A., Heinze, H.J., He, H., Walter, M., 2014. Dissociation of glutamate and cortical thickness is restricted to regions subserving trait but not state markers in major depressive disorder. J. Affect. Disord. 169, 91–100. Mannie, Z.N., Filippini, N., Williams, C., Near, J., Mackay, C.E., Cowen, P.J., 2014. Structural and functional imaging of the hippocampus in young people at familial risk of depression. Psychol. Med. 44, 2939–2948. Meyerhoff, D.J., Durazzo, T.C., Ende, G., 2013. Chronic alcohol consumption, abstinence and relapse: brain proton magnetic resonance spectroscopy studies in animals and humans. Curr. Top. Behav. Neurosci. 13, 511–540. Provencher, S.W., 1993. Estimation of metabolite concentrations from localized in vivo proton NMR spectra. Magn. Reson. Med. 30, 672–679. Sanacora, G., Gueorguieva, R., Epperson, C.N., Wu, Y.-T., Appel, M., Rothman, D.L., Krystal, J.H., Mason, G.F., 2004. Subtype-specific alterations of {gamma}-aminobutyric acid and glutamate in patients with major depression. Arch. Gen. Psychiatry 61, 705–713. Sanacora, G., Zarate, C.A., Krystal, J.H., Manji, H.K., 2008. Targeting the glutamatergic system to develop novel, improved therapeutics for mood disorders. Nat. Rev. Drug Discov. 7, 426–437. Scott, E.M., Hermens, D.F., Glozier, N., Naismith, S.L., Guastella, A.J., Hickie, I.B., 2012. Targeted primary care-based mental health services for young Australians. Med. J. Aust. 196, 136–140. Silveri, M.M., Cohen-Gilbert, J., Crowley, D.J., Rosso, I.M., Jensen, J.E., Sneider, J.T., 2014. Altered anterior cingulate neurochemistry in emerging adult binge drinkers with a history of alcohol-induced blackouts. Alcohol. Clin. Exp. Res. 38, 969–979. Strelnikov, K., 2007. Can mismatch negativity be linked to synaptic processes? A glutamatergic approach to deviance detection. Brain Cogn. 65, 244–251. Taylor, M.J., Selvaraj, S., Norbury, R., Jezzard, P., Cowen, P.J., 2009. Normal glutamate but elevated myo-inositol in anterior cingulate cortex in recovered depressed patients. J. Affect. Disord. 119, 186–189. White, A.M., Swartzwelder, H.S., 2004. Hippocampal function during adolescence: a unique target of ethanol effects. Ann. N.Y. Acad. Sci. 1021, 206–220. Yeo, R.A., Thoma, R.J., Gasparovic, C., Monnig, M., Harlaar, N., Calhoun, V.D., Kalyanam, R., Mayer, A.R., Durazzo, T.C., Hutchison, K.E., 2013. Neurometabolite concentration and clinical features of chronic alcohol use: a proton magnetic resonance spectroscopy study. Psychiatry Res. 211, 141–147.
Contents lists available at ScienceDirect
Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad
Short communication
Hippocampal glutamate is increased and associated with risky drinking in young adults with major depression Daniel F. Hermens a,n, Kate M. Chitty a, Rico SC Lee a, Ashleigh Tickell a, Paul S. Haber b,c, Sharon L. Naismith a, Ian B. Hickie a, Jim Lagopoulos a a
Clinical Research Unit, Brain and Mind Research Institute, University of Sydney, Camperdown, Australia Drug Health Services, Royal Prince Alfred Hospital, Camperdown, Australia c Sydney Medical School, University of Sydney, Camperdown, Australia b
art ic l e i nf o
a b s t r a c t
Article history: Received 10 April 2015 Received in revised form 2 July 2015 Accepted 6 July 2015 Available online 26 July 2015
Background: Risky drinking in young people is harmful, highly prevalent and often complicated by comorbid mental health problems that compound alcohol-induced impairment. The hippocampus and the glutamate system have been implicated in the pathophysiology of alcoholism and depression. This study aimed to determine whether risky drinking is associated with glutamate levels recorded within the hippocampus of young adults with major depression. Methods: Sixty-three young persons with major depression (22.1 73.1 years; 65% female) and 38 healthy controls were recruited. Participants completed the alcohol use disorder identification test and underwent proton magnetic resonance spectroscopy to measure in vivo glutamate levels within the hippocampus following a period of at least 48 h of abstinence. Results: Young adults with depression had significantly increased hippocampal glutamate levels and a positive association between the level of alcohol use and glutamate. Regression analysis revealed that higher levels of hippocampal glutamate were predicted by having increased levels of risky drinking and depression. Limitations: Small sample sizes for testing diagnosis by risky drinking interaction and use of creatine ratios rather than the absolute concentrations of glutamate. Discussion: The hippocampus is a critical region; given its role in learning and memory as well as mood regulation, and the neurochemical changes observed in this study may precede structural changes, which are commonly observed in both depression and alcohol misuse. These findings suggest that young adults with major depression who engage in risky drinking may be at increased risk of glutamate excitotoxicity. & 2015 Published by Elsevier B.V.
Keywords: Hippocampus Depression Alcohol Glutamate Magnetic resonance spectroscopy
1. Introduction Beyond determining the high prevalence of comorbid depression and risky drinking, little is known about the early neurobiological changes associated with this comorbidity. Importantly, the hippocampus has been implicated in the pathogenesis of both depression (Mannie et al., 2014) and alcohol use disorders (White and Swartzwelder, 2004). Acute alcohol exposure inhibits glutamate binding and reduces the transmission efficacy of cortical neurons via the suppression of N-methyl-D-aspartate receptor (NMDAr) activity (Strelnikov, 2007). With chronic alcohol use, NMDAr binding sites increase in number and level of functioning; as demonstrated in rodents who show increased glutamate transmission in the hippocampus after repeated ethanol n
Corresponding author. Fax: þ 61 2 93510652. E-mail address: [email protected] (D.F. Hermens).
http://dx.doi.org/10.1016/j.jad.2015.07.009 0165-0327/& 2015 Published by Elsevier B.V.
administration (Chefer et al., 2011). Moreover, converging evidence indicates that aberrations in glutamate homoeostasis and neurotransmission have a significant role in the development of depression. Post mortem investigation of glutamate derived from depressed patients has revealed elevated glutamate in the frontal cortex (Hashimoto et al., 2007) and studies on brain tissue from depressed patients have found down-regulation of genes that code for the excitatory amino acid transporters which reside on the glia and are responsible for clearing glutamate from the synapse (Beart and O'Shea, 2007). As a consequence of altered sensitivity, glutamate may accumulate at the synapse resulting in excitotoxicity (Meyerhoff et al., 2013). As a means to assess in vivo neurochemistry, proton magnetic resonance spectroscopy (1H-MRS) has provided important neurobiological insights into both depression and alcohol use. In terms of glutamate, 1H-MRS studies of depression samples have been mixed with evidence of decreased frontal (Hasler et al., 2007), normal frontal (Taylor et al., 2009), normal occipital (Godlewska
96
D.F. Hermens et al. / Journal of Affective Disorders 186 (2015) 95–98
et al., 2015) and increased occipital (Sanacora et al., 2004) levels. Of note, increased hippocampal glutamate has been reported in young people with a family history of depression (Mannie et al., 2014). In 1H-MRS studies of alcohol misuse, increased frontal glutamate has been reported in dependant humans (Hermann et al., 2012; Lee et al., 2007) and rats (Hermann et al., 2012). Similarly, treatment-seeking heavy drinkers show increased frontal glutamate; associated with duration of drinking (Yeo et al., 2013). In contrast, decreased glutamate levels have been found in the visual cortex of alcohol-dependent (Bagga et al., 2014) and frontal regions of binge (Silveri et al., 2014) and heavy (Ende et al., 2013) drinkers. Overall, the variability in the aforementioned 1H-MRS findings may be explained by key factors such as region of interest and duration of exposure/illness. In this study, we hypothesised that young adult outpatients with major depressive disorder (MDD) would show increased levels of in vivo hippocampal glutamate levels and this would be associated with their levels of alcohol use.
2. Methods Sixty-three outpatients (aged 18–30 years) with a primary DSM-IV diagnosis of MDD were recruited from a specialised service for young people with mental health problems (Scott et al., 2012). At the time of assessment, all patients were receiving clinician-based case management and their psychotropic medication regimens were as follows: 27% (17/63) were not taking psychotropic medications; 63% (40/63) were taking a third-generation anti-depressant; 17% (11/63) an atypical antipsychotic medication; and 9% (6/63) were taking a mood stabiliser. Healthy controls (N ¼ 38; aged 18–29 years) were recruited from the community and screened to ensure they had no history of a mental disorder. Exclusion criteria for all participants were medical instability, history of neurological disease, medical illness known to impact cognitive and brain function, intellectual and/or developmental disability, current substance dependence (not including alcohol) and insufficient English for assessment. All participants were asked to abstain from drug or alcohol use for 48 h prior to testing and informed that they may be asked to under-take an alcohol breath test and/or a saliva drug screen. The University of Sydney Human Research Ethics Committee approved the study and all participants gave written informed consent. A trained research psychologist conducted the BMRI Structured Interview for Neurobiological Studies (Lee et al., 2013) to determine the nature and history of any mental health problems. The interview included the Hamilton Depression Rating Scale (HDRS, 17-item) to quantify mood symptoms at the time of assessment; the Brief Psychiatric Rating Scale (BPRS) to quantify general psychiatric symptoms at the time of assessment; and, the Social and Occupational Functioning Assessment Scale (SOFAS); where a patient's functioning is rated from 0 to 100, with lower scores suggesting more severe impairment. As part of a self-report questionnaire, participants completed: the Kessler-10 (K-10), as a measure of psychological distress; the first two items of the World Health Organisation's (WHO) 'alcohol, smoking and substance involvement screening test' to assess lifetime and current substance use; and the WHO Alcohol Use Disorders Identification Test (AUDIT) to assess level of risky drinking in the past year, as well lifetime familiarity. 1 H-MRS data was acquired on a 3Tesla GE Discovery MR750 MRI scanner (GE Medical Systems, Milwaukee, WI), using an 8-channel phased array headcoil. The protocol comprised of a 3D sagittal whole-brain scout (TR ¼ 50 ms; TE ¼4 ms; matrix ¼256; no averaging, z ¼5 mm thickness), a T1-weighted MPRAGE sequence for anatomical localisation (TR ¼7.2 ms; TE ¼2.8 ms; flip
angle ¼10°; matrix 256 256; 0.9 mm isotropic voxels, 196 slices) and single-voxel 1H-MRS using PRESS acquired from a 1.5 3.0 1.0 cm3 voxel placed in the left hippocampus. Unsuppressed water scans (acquired from the same voxel) were collected prior to acquisition of the metabolite scans. Eddy current correction as implemented by LCModel package (Provencher, 1993) was undertaken on the metabolite data using the unsuppressed water FIDs as reference. Spectra were then quantified with LCModel, using a PRESS TE ¼35 basis set of 15 metabolites. Anatomical localisation of voxel placement was based on the Talaraich brain atlas and positioning was guided by the MPRAGE image. All spectra were shimmed to achieve full-width half maximum o13 Hz. Poorly fitted metabolite peaks as reflected by Cramer–Rao lower bounds 420 were excluded from further analysis. Statistical analyses were then conducted on glutamate (GLU) level as a ratio over water-scaled creatine (Cr) concentration. Segmentation of left hippocampal volumes was also undertaken as per our previously published methods (Hermens et al., 2015). Two-tailed independent t-tests were used to assess group differences for demographic, clinical and spectroscopic variables. If homogeneity of variance was violated Welch's corrected degrees of freedom and p-values were reported. Analysis of Covariance (ANCOVA) was employed to control for the potential effects of age. Pearson product moment correlations were used to examine for significant associations between total AUDIT scores and GLU levels, with partial correlations controlling for age. In order to further analyse the relationships between alcohol use (AUDIT total score), diagnosis, demographics (sex, age, family history) and glutamate, we performed multiple linear regression analyses with six predictor variables ('enter' method) in the model. Current nicotine smoking status was included as a predictor as this has been shown to affect 1H-MRS variables (Meyerhoff et al., 2013).
3. Results Table 1 summarises between-group comparisons across variables. There was no difference in gender ratio among groups; however, there was a significant difference in age (po .05; control group older). Furthermore, the groups differed significantly in years of education but did not differ in predicted IQ scores. In terms of alcohol-related variables, there were no group differences in age of first alcoholic drink or AUDIT total score, however the depression group had a significantly (p o.05) higher proportion of individuals with a family history of alcoholism. Controls had significantly higher social and occupational functioning (SOFAS) as well as lower self-reported psychological distress (K-10), current depressive (HDRS) symptoms and general psychiatric (BPRS) clinical ratings (all p o.001). The groups did not differ in the proportion of daily nicotine smokers. The groups did not differ in terms of creatine levels or hippocampal volumes; however, they were significantly (p o.05) different in GLU/Cr, with the depression group having increased levels (Table 1). ANCOVA confirmed this main effect remained significant when controlling for age (p o.05), years of education (p o.05), family history of alcoholism (po .05) or hippocampal volume (po .05). For the depression group, total AUDIT score was positively associated with GLU/Cr levels (r ¼ 0.429, n ¼62, po 0.001), and remained significant (po .001) after controlling for age or hippocampal volume. The controls showed no significant correlations between these measures. Fig. 1 depicts the associations between GLU/Cr levels and AUDIT scores for each group. The multiple linear regression model for hippocampal GLU/Cr level was significant [F(6, 97) ¼3.3, p o.01] and explained 18.1% of the variance (adjusted R2 ¼0.126) with three significant predictors:
D.F. Hermens et al. / Journal of Affective Disorders 186 (2015) 95–98
97
Table 1 Mean scores (7 standard deviation) for demographic, clinical and spectroscopic variables for depression versus controls.
Sex (% female) FH Alc (% þ ve) Current nicotine use (% daily) Age, yrs Education, yrs Predicted IQ Age, first drink AUDIT, total SOFAS K-10 total HDRS total BPRS total GLU/Cr
Depression (N ¼ 63)
Controls (N ¼38)
Between-group differences
65.1% 19.7% 17.5% 22.17 3.1 13.0 72.0 104.77 8.2 15.2 72.1 7.0 7 6.7 65.4 7 10.8 27.2 7 7.3 13.8 76.2 38.0 7 6.3 1.58 7 0.20
65.8% 2.6% 13.2% 23.3 7 2.6 14.4 7 2.3 105.2 7 7.2 15.2 7 1.6 8.9 7 4.9 89.3 7 5.9 14.8 7 3.9 1.9 7 2.6 25.9 7 2.8 1.477 0.22
χ2¼ 0.1, df¼ 101, p ¼ .942 χ2¼ 6.0, df ¼99, p¼ .015 χ2¼ 0.3, df¼ 101, p ¼ .566 t¼ 2.1, df ¼99, p¼ .035 t¼ 3.0, df ¼ 99, p¼ .003 t¼ 0.3, df ¼95, p¼ .771 t¼ 0.1, df ¼92, p ¼ .953 t¼ 1.5, df¼ 99, p ¼.127 t¼ 13.9, df ¼ 92.7, p¼ .000 t¼ 11.1, df ¼97.8, p¼ .000 t¼ 13.1, df ¼88.8, p ¼.000 t¼ 12.2, df ¼88.2, p¼ .000 t¼ 2.5, df ¼98, p¼ .016
Between group differences were tested by chi-square or t-test. Abbreviations are as follows: FH Alc ¼ family history of alcoholism; AUDIT ¼ alcohol use disorders identification test; SOFAS¼ social and occupational functioning assessment scale; K-10¼ Kessler 10-item; HDRS ¼Hamilton depression rating scale; BPRS ¼ brief psychiatry rating scale; GLU ¼ glutamate; Cr¼ creatine.
Fig. 1. Scatterplot of hippocampal glutamate (relative to creatine) level vs. the alcohol use disorder identification test (AUDIT) total score in the depression (black triangles; black solid line) and control (grey circles; grey dashed line) groups.
AUDIT total (β ¼0.268, p o.05), sex (β ¼0.238, p o.05) and diagnosis (β ¼ 0.244, p o.05). In this model, increased levels of risky drinking, being female and having a diagnosis of depression were all associated with increased levels of glutamate. A follow-up regression model was performed with three interaction variables (diagnosis AUDIT total; sex AUDIT total; diagnosis sex) and the three significant predictors (diagnosis, sex, AUDIT total). This model was also significant [F(5, 99) ¼5.7, po .01] and explained 23.3% of the variance (adjusted R2 ¼0.192) with two significant variables: diagnosis (β ¼ 0.208, p o.05) and the interaction of diagnosis and AUDIT total (β ¼ 0.245, p o.05). Of note, sex was excluded from the model.
4. Discussion These findings are consistent with other 1H-MRS studies reporting increased glutamate levels in depression (Li et al., 2014; Sanacora et al., 2004), family history of depression (Mannie et al.,
2014) and alcohol misuse (Lee et al., 2007; Yeo et al., 2013). Of these studies, one targeted the hippocampus (Mannie et al., 2014) while the others assessed either frontal or occipital regions. However, there is also evidence (primarily in frontal regions) of decreased glutamate in depression/alcohol misuse, highlighting the need for further studies of in vivo hippocampal glutamate. Importantly, in rodents, there is evidence that short-term, repeated exposure to moderate doses of alcohol can lead to increases in extracellular glutamate levels in the CA3 region of the hippocampus (Chefer et al., 2011), which is important for the encoding of new spatial information (Kesner, 2007). Prolonged abstinence may also lead to decreased hippocampal glutamate levels in depressed, risky drinkers since there is evidence, in both human and rodent studies, showing that despite a hyper-glutamatergic state during withdrawal there is a subsequent normalisation of glutamate levels (Meyerhoff et al., 2013). It is important to note, that we did not specifically monitor for withdrawal symptoms in the current study as no subjects were diagnosed with alcohol dependence disorder. Thus, future studies of individuals with depression and alcohol dependence who abstain (and hence undergo withdrawal) at the time of 1H-MRS scan are warranted. There were several limitations associated with this study. Firstly, the sample sizes in our study were relatively modest, particularly with respect to potential sex differences. Another limitation is the use of creatine ratios rather than the absolute concentrations of glutamate; although creatine ratios are a robust method of normalising metabolite levels across subjects and have been used extensively in the literature. 1H-MRS is only able to measure the absolute pool of available neurometabolites within a voxel, thus limiting any inferences about whether increased glutamate levels are primarily neuronal, extracellular or post-synaptic. Our 1H-MRS protocol did not allow quantification of glutamine (GLN). However by indexing glutamate-to-glutamine cycling (via GLU/GLN) future studies may shed even more light on the combined effects of depression and alcohol on this system. Finally, it has been suggested that the therapeutic effects of many psychotropic medications are potentially a result of glutamatergic interactions (Sanacora et al., 2008). As such, the contributing effects on the glutamatergic system of the medications taken by subjects in this study cannot be fully ascertained. In conclusion, data from this study suggests that younger people with MDD should be cautious about the excessive use of alcohol. During this period of critical brain development, excessive alcohol use appears to be associated with markers of excitotoxicity, which in turn, is likely to have detrimental effects on cognitive and social functioning. The hippocampus is a particularly
98
D.F. Hermens et al. / Journal of Affective Disorders 186 (2015) 95–98
critical region; given its role in learning and memory as well as mood regulation, and the neurochemical changes observed in this study may precede structural changes, which are commonly observed in both depression and alcohol misuse. Indeed, alcohol may exacerbate the neurobiological changes inherent to MDD, which in turn, is associated with pronounced disability. In order to strengthen and extend these findings, future studies should longitudinally track the concomitant impact of alcohol on the emerging neurobiological changes (including those specifically glutamatergic in nature) seen in young people with depression.
Conflict of interest No conflict declared.
Author disclosures Dr. Hermens has received honoraria for educational seminars from Janssen-Cilag and Eli Lilly. Prof. Hickie has led a range of community-based and pharmaceutical industry-supported depression awareness and education and training programmes. He has led depression and other mental health research service evaluation or investigator-initiated research projects that have been supported by a variety of pharmaceutical partners. Servier and Pfizer support current investigator-initiated studies. He has received honoraria for his contributions to professional educational seminars related to depression, youth mental health and circadian-rhythms research. He has received travel support from Servier to attend scientific meetings related specifically to circadianrhythm disorders. The authors report no other biomedical financial interests or potential conflicts of interest.
Contributors DFH, JL, SLN and IBH contributed to the conception of the study; DFH, JL, KMC and AT undertook the data processing and statistical analysis; DFH wrote the first draft of the manuscript. All authors contributed to data interpretation, discussion and have approved the final manuscript.
Role of the funding source This work was funded by a grant from the NSW Ministry of Health, Mental Health and Drug and Alcohol Office as well as National Health and Medical Research Council program (No. 566529) and Centres of Clinical Research Excellence (No. 264611) grants. DFH was supported by a grant from the NSW Ministry of Health, Mental Health and Drug and Alcohol Office; SLN was supported by an NHMRC Career Development Award (No. 402864); and IBH was supported by an NHMRC Senior Principal Research Fellowship (No. 1046899). These funding agencies had no role in the analysis or interpretation of these data; or in the writing of this report; or the decision to submit the paper for publication.
Acknowledgement The authors would like to express their gratitude to individuals who participated in this study.
References Bagga, D., Khushu, S., Modi, S., Kaur, P., Bhattacharya, D., Garg, M.L., Singh, N., 2014. Impaired visual information processing in alcohol-dependent subjects: a proton magnetic resonance spectroscopy study of the primary visual cortex. J. Stud. Alcohol Drugs 75, 817–826. Beart, P.M., O'Shea, R.D., 2007. Transporters for L-glutamate: an update on their molecular pharmacology and pathological involvement. Br. J. Pharmacol. 150, 5–17. Chefer, V., Meis, J., Wang, G., Kuzmin, A., Bakalkin, G., Shippenberg, T., 2011. Repeated exposure to moderate doses of ethanol augments hippocampal glutamate neurotransmission by increasing release. Addict. Biol. 16, 229–237. Ende, G., Hermann, D., Demirakca, T., Hoerst, M., Tunc-Skarka, N., Weber-Fahr, W., Wichert, S., Rabinstein, J., Frischknecht, U., Mann, K., Vollstadt-Klein, S., 2013. Loss of control of alcohol use and severity of alcohol dependence in nontreatment-seeking heavy drinkers are related to lower glutamate in frontal white matter. Alcohol. Clin. Exp. Res. 37, 1643–1649. Godlewska, B.R., Near, J., Cowen, P.J., 2015. Neurochemistry of major depression: a study using magnetic resonance spectroscopy. Psychopharmacology (Berl) 232, 501–507. Hashimoto, K., Sawa, A., Iyo, M., 2007. Increased levels of glutamate in brains from patients with mood disorders. Biol. Psychiatry 62, 1310–1316. Hasler, G., van der Veen, J.W., Tumonis, T., Meyers, N., Shen, J., Drevets, W.C., 2007. Reduced prefrontal glutamate/glutamine and gamma-aminobutyric acid levels in major depression determined using proton magnetic resonance spectroscopy. Arch. Gen. Psychiatry 64, 193–200. Hermann, D., Weber-Fahr, W., Sartorius, A., Hoerst, M., Frischknecht, U., TuncSkarka, N., Perreau-Lenz, S., Hansson, A.C., Krumm, B., Kiefer, F., Spanagel, R., Mann, K., Ende, G., Sommer, W.H., 2012. Translational magnetic resonance spectroscopy reveals excessive central glutamate levels during alcohol withdrawal in humans and rats. Biol. Psychiatry 71, 1015–1021. Hermens, D.F., Naismith, S.L., Chitty, K.M., Lee, R.S.C., Tickell, A., Duffy, S.L., Paquola, C., White, D., Hickie, I.B., Lagopoulos, J., 2015. Cluster analysis reveals abnormal hippocampal neurometabolic profiles in young people with mood disorders. Eur. Neuropsychopharmacol. 25, 836–845. Kesner, R.P., 2007. Behavioral functions of the CA3 subregion of the hippocampus. Learn. Mem. 14, 771–781. Lee, E., Jang, D.-P., Kim, J.-J., An, S.K., Park, S., Kim, I.-Y., Kim, S.I., Yoon, K.-J., Namkoong, K., 2007. Alteration of brain metabolites in young alcoholics without structural changes. Neuroreport 18, 1511–1514. Lee, R.S.C., Hermens, D.F., Redoblado-Hodge, M.A., Naismith, S.L., Porter, M.A., Kaur, M., White, D., Scott, E.M., Hickie, I.B., 2013. Neuropsychological and socio-occupational functioning in young psychiatric outpatients: a longitudinal investigation. PLoS ONE 8, e58176. Li, M., Metzger, C.D., Li, W., Safron, A., van Tol, M.J., Lord, A., Krause, A.L., Borchardt, V., Dou, W., Genz, A., Heinze, H.J., He, H., Walter, M., 2014. Dissociation of glutamate and cortical thickness is restricted to regions subserving trait but not state markers in major depressive disorder. J. Affect. Disord. 169, 91–100. Mannie, Z.N., Filippini, N., Williams, C., Near, J., Mackay, C.E., Cowen, P.J., 2014. Structural and functional imaging of the hippocampus in young people at familial risk of depression. Psychol. Med. 44, 2939–2948. Meyerhoff, D.J., Durazzo, T.C., Ende, G., 2013. Chronic alcohol consumption, abstinence and relapse: brain proton magnetic resonance spectroscopy studies in animals and humans. Curr. Top. Behav. Neurosci. 13, 511–540. Provencher, S.W., 1993. Estimation of metabolite concentrations from localized in vivo proton NMR spectra. Magn. Reson. Med. 30, 672–679. Sanacora, G., Gueorguieva, R., Epperson, C.N., Wu, Y.-T., Appel, M., Rothman, D.L., Krystal, J.H., Mason, G.F., 2004. Subtype-specific alterations of {gamma}-aminobutyric acid and glutamate in patients with major depression. Arch. Gen. Psychiatry 61, 705–713. Sanacora, G., Zarate, C.A., Krystal, J.H., Manji, H.K., 2008. Targeting the glutamatergic system to develop novel, improved therapeutics for mood disorders. Nat. Rev. Drug Discov. 7, 426–437. Scott, E.M., Hermens, D.F., Glozier, N., Naismith, S.L., Guastella, A.J., Hickie, I.B., 2012. Targeted primary care-based mental health services for young Australians. Med. J. Aust. 196, 136–140. Silveri, M.M., Cohen-Gilbert, J., Crowley, D.J., Rosso, I.M., Jensen, J.E., Sneider, J.T., 2014. Altered anterior cingulate neurochemistry in emerging adult binge drinkers with a history of alcohol-induced blackouts. Alcohol. Clin. Exp. Res. 38, 969–979. Strelnikov, K., 2007. Can mismatch negativity be linked to synaptic processes? A glutamatergic approach to deviance detection. Brain Cogn. 65, 244–251. Taylor, M.J., Selvaraj, S., Norbury, R., Jezzard, P., Cowen, P.J., 2009. Normal glutamate but elevated myo-inositol in anterior cingulate cortex in recovered depressed patients. J. Affect. Disord. 119, 186–189. White, A.M., Swartzwelder, H.S., 2004. Hippocampal function during adolescence: a unique target of ethanol effects. Ann. N.Y. Acad. Sci. 1021, 206–220. Yeo, R.A., Thoma, R.J., Gasparovic, C., Monnig, M., Harlaar, N., Calhoun, V.D., Kalyanam, R., Mayer, A.R., Durazzo, T.C., Hutchison, K.E., 2013. Neurometabolite concentration and clinical features of chronic alcohol use: a proton magnetic resonance spectroscopy study. Psychiatry Res. 211, 141–147.