The effect of transcranial Direct Current Stimulation on gamma activity and working memory in schizophrenia

Author's Accepted Manuscript

The effect of transcranial direct current stimulation on gamma activity and working memory in schizophrenia Kate E. Hoy, Neil W. Bailey, Sara L. Arnold, Paul B. Fitzgerald

www.elsevier.com/locate/psychres

PII: DOI: Reference:

S0165-1781(15)00235-8 http://dx.doi.org/10.1016/j.psychres.2015.04.032 PSY8869

To appear in:

Psychiatry Research

Received date: 12 September 2014 Revised date: 26 February 2015 Accepted date: 16 April 2015 Cite this article as: Kate E. Hoy, Neil W. Bailey, Sara L. Arnold, Paul B. Fitzgerald, The effect of transcranial direct current stimulation on gamma activity and working memory in schizophrenia, Psychiatry Research, http://dx.doi. org/10.1016/j.psychres.2015.04.032 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

The effect of transcranial Direct Current Stimulation on gamma activity and working memory in schizophrenia

Kate E Hoy* Neil W Bailey Sara L Arnold Paul B Fitzgerald

Monash Alfred Psychiatry Research Centre, The Alfred and Monash University, Central Clinical School, Victoria, Australia.

*Address Correspondence to: Dr Kate Hoy Monash Alfred Psychiatry Research Centre, Level 4 607 St Kilda Road, Melbourne, 3004 Ph: 61 3 9076 5034 Fax: 61 3 9076 6588 [email protected]

1

The effect of transcranial Direct Current Stimulation on gamma activity and working memory in schizophrenia

Kate E Hoy* Neil W Bailey Sara L Arnold Paul B Fitzgerald

Monash Alfred Psychiatry Research Centre, The Alfred and Monash University, Central Clinical School, Victoria, Australia.

*Address Correspondence to: Dr Kate Hoy Monash Alfred Psychiatry Research Centre, Level 4 607 St Kilda Road, Melbourne, 3004 Ph: 61 3 9076 5034 Fax: 61 3 9076 6588 [email protected]

2

Abstract Working memory impairments in schizophrenia have been strongly associated with abnormalities in gamma oscillations within the dorsolateral prefrontal cortex (DLFPC). We recently published the first ever study showing that anodal transcranial direct current stimulation (tDCS) to the left DLPFC was able to significantly improve working memory performance in schizophrenia. In the current paper we present a secondary analysis from this study, specifically looking at the effect of tDCS on gamma activity and its relationship to working memory. In a repeated measures design we assessed the impact of anodal tDCS (1mA, 2mA , sham) on gamma activity in the left DLPFC at three time-points post stimulation (0 min, 20 min, 40 min). EEG data was available for 16 participants in the 2mA condition, 13 in the 1mA condition and 12 in the sham condition. Following 2mA stimulation we found a significant increase in gamma event-related synchronisation in the left DLPFC, this was in the context of a significantly improved working memory performance. There was also a significant decrease in gamma event-related synchronisation, with no changes in working memory, following sham stimulation. The current study provides preliminary evidence that tDCS may enhance working memory in schizophrenia by restoring normal gamma oscillatory function.

Keywords: transcranial Direct Current Stimulation; gamma oscillations; working memory; schizophrenia.

3

1. Introduction Cognitive impairments are a core feature of schizophrenia. They are highly prevalent, result in considerable functional disability, and are not effectively treated by current approaches (Insel, 2010). Pharmacotherapy, despite its effectiveness for the positive symptoms of schizophrenia, has shown little to no effect on the cognitive impairments (Kreyenbuhl et al., 2010). Cognitive remediation has generally resulted in only modest improvements in cognition following many hours of therapy (Vinogradov et al., 2012).

Working memory refers to the process of keeping information ‘in mind’ for short periods of time. It is an essential component of higher level cognitive functions (for example language, learning, problem solving), and indeed improvements in working memory have been shown to enhance more complex thought and action (Jaušovec and Jaušovec, 2012). In non-clinical populations working memory functioning has been consistently associated with activity in the Dorsolateral Prefrontal Cortex (DLPFC) (For review see Curtis and D’Esposito, 2013). While working memory is subserved by a number of brain regions, it has been shown that the DLPFC is a central node for the systems responsible for the manipulation of information (Barbey et al., 2013). Indeed, impairments in working memory in schizophrenia have been reliably associated with impaired functioning DLPFC; more specifically it is abnormalities in neural synchrony within this brain region that are believed to underlie the working memory deficits (Chen et al., 2014; Haenschel et al., 2009; Lett et al., 2014). Neural synchrony, referring to large populations of neurons firing simultaneously at specific frequencies, has been shown to be essential for successful cognitive functioning (Uhlhaas and Singer, 2006). Working memory in particular is associated with synchronous activity at the gamma frequency (>40Hz), with increased cognitive effort associated with increased gamma synchrony in healthy populations (Basar-Erglou et al., 2007; Howard et al., 2003).

4

Dysfunctional gamma activity in schizophrenia has been repeatedly reported in the literature, believed to be related to the well-established GABA impairments seen in the illness (Chen et al., 2014; Lett et al., 2014;). GABA is the brains’ primary inhibitory neurotransmitter and, amongst other functions, has a central role in both generating and modulating synchronous gamma activity (Chen et al., 2014). The literature is somewhat mixed with respect to the nature of the gamma abnormalities in schizophrenia, namely whether gamma is excessive or impaired (Basar-Erglou et al., 2007; Chen et al., 2014; Gonzalez-Burgos et al., 2011; Haenschel et al., 2009;). In reviewing this literature Sun and colleagues (2011) concluded that gamma activity is in fact not optimally regulated in patients with schizophrenia wherein patients are not able to increase gamma when the level of cognitive effort requires it and that below a certain level of cognitive demand, (when gamma suppression is thought to be adaptive), they instead show an increase. Indeed, research has indicated that patients with schizophrenia are not able to modulate gamma activity in response to cognitive task demands (Basar-Erglou et al., 2007; Chen et al., 2014; Gonzalez-Burgos et al., 2011; Haenschel et al., 2009; Moran et al., 2011;; Sun et al., 2011). Treatment approaches which target such processes would have considerable potential for significantly improving working memory function in schizophrenia.

Transcranial Direct Current Stimulation (tDCS) is a non-invasive form of brain stimulation which has shown considerable promise for the enhancement of cognition (Utz, 2010; Jacobson 2012). tDCS involves the application of a very weak electrical current applied using two surface electrodes (anode and cathode) applied to the scalp. This current alters the excitability of brain cells by shifting their membrane potentials in a de- or hyperpolarising direction; thus making them more or less likely to fire (Nitsche and Frengi, 2007). Stimulation of brain cells under the anode appears to increase brain activity whereas

5

stimulation under the cathode generally has the opposite effect (Jacobson, 2012). tDCS is a non-polarizing form of brain stimulation, unlike Transcranial Magnetic Stimulation (TMS), and therefore is not associated with a risk of seizure induction (Nitsche and Frengi, 2007). Indeed tDCS, when provided within defined safety limits, has been shown to be a safe and well tolerated technique being associated with only minor adverse effects such as tingling or itching at the stimulation site (Poreisz et al., 2007). There is evidence from Magnetic Resonance Spectroscopy (MRS) investigations that anodal tDCS has its excitatory effects via the direct modulation of GABA-ergic activity (Stagg et al., 2011; Stagg et al., 2009), with a growing number of studies also showing enhanced neural synchrony following anodal tDCS (Hoy et al., 2013; Zahele et al., 2011), including in the gamma frequency range (Antal et al., 2004). In light of the converging evidence connecting the dysfunctional processes thought to underlie working memory deficits in schizophrenia and the proposed mechanisms of action of tDCS, we recently undertook a proof of concept study to investigate whether tDCS was able to enhance working memory in schizophrenia (Hoy et al., 2014). We revealed a significant improvement in working memory performance over time following a single 20 minute stimulation session of 2mA anodal tDCS to the left DLFPC, while 1mA and sham stimulation had no effect on performance. While we have shown tDCS to have considerable promise with respect to enhancing behavioural performance in schizophrenia, we have yet to examine the effect of tDCS on the abnormal gamma thought to underlie working memory impairments.

The aim of the current study was to investigate whether our previously reported improvements in working memory following 2mA anodal tDCS (Hoy et al., 2014) were reflected by changes in gamma activity in the left DLPFC. In Hoy and colleagues (2014) significant improvements in working memory performance were seen over time following

6

2mA tDCS as compared to no changes in performance over time post 1mA or sham stimulation. Therefore, in the current analyses we investigated the effect of 2mA, 1mA and sham stimulation on change in gamma event-related synchronization (ERS) during performance of a working memory task across the three time-points post stimulation (i.e. 0 min, 20 min and 40 min). We hypothesized that 2mA anodal tDCS would result in increased gamma activity over time, in line with the improvements seen in working memory performance. We also hypothesised that neither 1mA nor sham stimulation would result in significant changes in gamma, consistent with the lack of behavioural improvement in these conditions.

2. Method 2.1 Participants 18 participants with schizophrenia were recruited into the repeated measures within subjects study, our previous manuscript reported the behavioural findings only from all 18 participants (Hoy et al., 2014). The current paper reports a secondary analysis from this study which examined neurophysiological changes. Due to the presence of excessive noise, the EEG data from a number of participants across the conditions was not able to be used. Analysable EEG data for the current analysis was available for 16 participants in the 2mA condition, 13 in the 1mA condition and 12 in the sham condition. (See Table 1 for demographic and clinical data by stimulation condition).

Diagnosis was confirmed using the Mini-International Neuropsychiatric Interview (MINI), administered by research staff experienced in its use (KH/SA) (Sheehan et al., 1998). All recruited participants were regularly taking atypical antipsychotics (i.e. 3 Risperidone; 3 Aripiprazole; 2 Olanzapine; 2 Amisulpride;1 Clozapine; 1 Paliperidone; 1 Olanzapine +

7

Paliperidone; 1 Olanzapine + Aripriprazole; 1 Aripriprazole+ Quetiapine; 1 Quetiapine + Ziprasidone; 1 Pericyazine+Asenapine and 1 Quetiapine +Asenapine), 9 were additionally taking antidepressant medication (i.e. 2 SNRI; 4 SSRI; 1 TCA and 1 SNRI + TCA). There are a number of known, as well as many potential unknown, interaction effects between medication and tDCS (Brunoni et al., 2012). Therefore, in addition to excluding participants on medication that has been shown to influence the effects of tDCS (see Brunoni et al., 2012), we also required participants to remain on their current dose and type of medication throughout the study. Exclusion criteria consisted of the presence of any neurological or serious medical conditions, or current pregnancy. Written consent was obtained from participants prior to commencement of the study. Ethical approval was granted by Monash University and the Alfred Hospital ethics committees.

2.2 Procedure This was a randomised repeated-measures double-blind study design. Participants attended for three sessions which were held at least 72 hours apart (See Figure 1 for study design). Sessions were randomised and counterbalanced, and each involved the provision of 20 minutes of 1mA, 2mA or sham tDCS followed by working memory assessment with concurrent EEG recordings. An electrode array which allowed recording of frontal activity was applied, namely F3, FZ, and F4; facial electrodes were used for measurement of eye movements (positioned adjacent to the left and right outer canthus of each eye and above and below the left orbit) and left and right mastoids were employed for referencing. The limited EEG array was used to ensure we were able to quickly replace the required electrodes to allow for immediate recording following stimulation. The EEG array was set up and impedances of less than 5kȍ were achieved before tDCS was applied. Immediately prior to stimulation, the F3 EEG electrode was removed and the anodal tDCS electrode was

8

positioned over the F3 position and the cathode over the right supraorbital region. While montages can vary, particularly with respect to position of the reference electrode, the montage used in the current study is widely accepted as a standard for anodal stimulation of the left DLPFC stimulation (Nitsche et al., 2008). Immediately following stimulation the F3 EEG electrode was replaced and impedances were re-checked, this was achieved within 60 seconds of the end of stimulation for all participants. Participants undertook the 2-back working memory task immediately following stimulation and at 20 and 40 minutes post stimulation. EEG was sampled at 1000Hz (band pass 0.1-100Hz) using a SynAmps 2 amplifier (Compumedics, Melbourne Australia). Impedances were maintained at less than 5kȍ. Electrodes were single Ag/AgCl electrodes.

2.3 Transcranial Direct Current Stimulation tDCS was applied using an Eldith Stimulator Plus (neuroConn GBH) delivering direct current through two surface electrodes (35 cm² saline soaked sponges). Saline solution was 0.9% saline. Anodal stimulation was applied at the left DLPFC for 20 minutes (ramp up of 120s and ramp down of 15s) across three conditions: 1 mA, 2 mA, and sham. The tDCS administrator entered a stimulation code into the tDCS machine at the start of the session; when a placebo, or ‘Sham,’ code was entered stimulation is turned off following fade-in of 30 seconds. This switching off of stimulation occurs within the software of the tDCS device, allowing blinding of the administrator to active or sham conditions. The order of the tDCS sessions was counterbalanced. Tolerability of tDCS was assessed via self-report following questioning from an experienced tDCS administrator (SA).

9

2.4 Working Memory Task: 2 back A series of letters from A to J were presented consecutively and participants were required to respond with a button press when the presented letter was the same as the letter presented 2 trials earlier. The order of the presentation of letters (from A to J) was random. The task consisted of 130 trials containing 25% targets. Each letter was presented for 500ms with a 1500ms delay between stimuli presentations. Each participant undertook the task three times per session (0mins, 20mins and 40mins), over the three sessions this equated to a total of nine blocks. Alternate stimuli were used for each of these blocks, so that participants never repeated the exact same 5 minute 2 back block.

2.5 Statistical Analysis Of the 18 participants tested, useable EEG data was available for 16 participants in 2mA, 13 in 1mA and 12 in sham conditions. For the subsequent behavioural analysis we only included these participants to allow for a valid comparison between working memory performance and gamma ERS/ERD in each condition. Therefore, the behavioural data reported here is a slightly reduced subset from the data reported in Hoy and colleagues (2014).

Offline EEG analysis was performed using Scan 4.3 (Compumedics, Melbourne, Australia). Ocular artefact correction was conducted using an off-line algorithm (Semlitsch et al., 1986). Data were visually inspected and noisy data were rejected. Automatic artefact rejection was also conducted to exclude epochs with deflections larger than +/-100µv, and epochs with more than 20µv in the 60-100Hz band (in order to exclude muscle artefact). Data were rereferenced offline to averaged mastoids. ERS was calculated for correct trials only in the gamma band using ERS% = (A – R)/R x 100, where A and R are total gamma power (both

10

evoked and induced, zero phase shift filtered while computing a signal envelope around 37.5Hz, with a 7.5Hz half bandwidth and 48dB/octave roll-off) during the active interval (A the first 1000ms following the stimulus), and the reference interval (R - the 500ms preceding the stimulus). Using this formula, positive values represent synchronization, and negative values represent desynchronisation. We utilised the outlier labeling method whereby the difference between the first and third quartile of the distribution is multiplied by g (g = 2.2) and the resulting value is added to the third quartile and subtracted from the first quartile with outliers falling outside this (Hoaglin and Iglewicz, 1987). Outliers were subsequently windsorised, which was required for 4.4% of data and is considered to be within acceptable limits, i.e. whereby p values are not effected (Hoaglin and Iglewicz, 1987).

To investigate the impact of tDCS on gamma ERS in the left DLPFC following stimulation we used paired-samples t-tests to compare gamma ERD/ERS at F3 immediately post stimulation to the two subsequent time points (20 minutes and 40 minutes) for each stimulation condition. We also assessed the the impact of tDCS on learning effects post stimulation by comparing d prime immediately post stimulation to the two subsequent time points (20 minutes and 40 minutes post stimulation ) with paired-sample t-tests for each stimulation condition. d prime is a discriminability index which takes into account the ability to correctly identify targets and to minimise false alarms, and has been shown to have high sensitivity in schizophrenia (Haaveit et al., 2010). We also conducted a series of Spearman’s correlations to investigate any relationship between change in gamma ERS/ERD and dprime across the time points for each stimulation condition. 1-tailed statistical tests were used in line with our directional hypotheses (that 2mA tDCS would result in increased gamma ERS and improved behavioural performance). All results were assessed using an alpha level of less than 0.05.

11

3. Results 3.1 2mA 3.1.1 Gamma ERS When compared to the immediately post stimulation time point (0 min) there was no difference in gamma ERS at 20 minutes post stimulation (t(15) = 0.945, p = 0. 165), there was however a significant increase in gamma at 40 minutes post stimulation 2mA (t(15) = 1.851, p = 0.042; d =0.68,). See Figure 2a and Table 2.

3.1.2 d prime Following 2mA there was a significant improvement in 2-back performance at 20 minutes post stimulation compared to the 0 minute time point ( t (15) = 2.302, p = 0.018 ; d =0.84), with this level of improved performance maintained at 40 minutes post 2mA stimulation (See Figure 2a). There was also a trend towards improved performance at 40 minutes as compared to 0 minutes post stimulation (t (15) = 1.558 , p = 0.070). Means and standard deviations are provided in Table 2.

3.1.3 Correlation There was a near significant positive correlation between increase in gamma ERS and improvement in d prime from 0 min to 20 min post stimulation (rho = 0.426, p = 0.050). There were no other significant correlations between gamma change and improvement in d prime following 2mA stimulation.

3.2 1mA 3.2.1 Gamma ERS

12

There was no change in gamma ERS at either the 20 minute (t (12) = -0.429, p = 0.675) or 40 minute time point (t (12) = - 1.375, p = 0.194) compared to immediately post stimulation. See Figure 2b. Means and standard deviations are provided in Table 2.

3.2.2 d prime Similarly, there was no change in 2back performance at either time point (20mins: t (12) = 0.210, p = 0.300; 40 mins: t (12) = -1.084, p = 0.300) compared to 0 minutes post stimulation. See Figure 2b. Means and standard deviations are provided in Table 2.

3.2.3 Correlation There were no significant correlations between change in gamma and d prime following 1mA stimulation.

3.3 Sham 3.3.1 Gamma ERS Compared to the 0 minutes post stimulation time point there was no difference in gamma ERS at 20 minutes post stimulation (t(11) = -0.998, p = 0.338), however there was a significant decrease in gamma at 40 minutes post stimulation (t(11) = -3.716, p = 0.003; d =1.57). See Figure 2c. Means and standard deviations are provided in Table 2.

3.3.2 d prime There was no change in 2-back performance at either time point (20mins: t (11) = 0.521, p = 0.612; 40 mins: t (11) = -0.375, p = 0.715) compared to 0 minutes post stimulation. See Figure 2c. Means and standard deviations are provided in Table 2.

13

3.3.3 Correlation There were no significant correlations between change in gamma and d prime on the 2back following sham.

3.4 Tolerability of tDCS All participants tolerated tDCS well with no adverse events reported.

4. Discussion We present data showing that tDCS induced improvements in working memory performance in schizophrenia are consistent with in enhanced gamma ERS in the left DLPFC. Specifically, following 2mA left DLPFC anodal stimulation participants showed a significant increase, of a moderate to large effect size (d =0.68 ), in gamma synchrony from the initial (0 min) to the final (40 minutes) post stimulation time point. This was alongside a significant improvement in performance on the 2back from 0 min post to 20 min post, large effect size (d =0.84), which was maintained at final time point (40 minutes post stimulation). Correlational analysis revealed a moderate, and near significant, positive relationship between enhanced gamma and improved performance from 0 to 20 minutes following 2mA. There were no changes in either gamma or working memory performance following 1mA stimulation. Following sham stimulation there was again no change in working memory performance, there was however a significant decrease in gamma from the initial (0 min) to the final (40 minutes) time point, again with a large effect size (d =1.52). These findings are consistent with previous research showing that tDCS appears to enhance working memory performance via modulating neural synchrony. In particular, tDCS induced working memory improvements in healthy controls has been associated with increases in theta and decreases in alpha synchrony (Hoy et al., 2013; Zahele et al., 2011). To

14

the best of our knowledge, this is the first study which has looked at the EEG correlates of working memory performance following tDCS in a patient population. Our findings provide the first, preliminary, evidence that tDCS appears to be able to modulate the pathophysiology underlying working memory deficits in schizophrenia to restore neuronal functioning and behaviour.

Across the 40 minutes post 2mA stimulation, patients exhibited a significant increase in gamma alongside significantly improved performance on the repeated undertakings of the 2back task over this time. This is consistent with what is seen in healthy controls, where increasing effort in relation to working memory tasks results in increased gamma synchrony (Basar-Eroglu et al., 2007). While increased effort in working memory tasks is most often induced by increasing working memory loads (i.e. 2 back versus 3 back), studies have shown that increasing effort is also a function of practice particularly during the first few instances of repeated task engagement as was the case in the current study (Yeo et al., 2004). While previous research shows inconsistency with respect to whether gamma activity is increased or decreased in schizophrenia, as discussed earlier, there does appear to be a consensus that patients exhibit an inability to modulate gamma activity when increased cognitive effort is required (Basar-Erglou et al., 2007; Chen et al., 2014; Gonzalez-Burgos et al., 2011; Lette et al., 2014). As such, the provision of 2mA anodal tDCS may have ‘normalised’ patient’s ability to modulate gamma activity in response to increased effort, resulting in improved performance with repeated task engagement, as is seen in healthy controls under normal conditions (Basar-Erglou et al., 2007; Yeo et al., 2004). The ability of tDCS to modulate gamma in this way in patients with schizophrenia, while not directly addressed in the current study, may be perhaps due to the effect that anodal tDCS has been shown to have on GABAergic activity subsequently providing the appropriate cortical ‘environment’ for normal

15

oscillatory modulation (Stagg et al., 2009; Stagg et al., 2011). A somewhat unexpected finding was the significant decrease in gamma over time following sham stimulation, with no change in working memory performance. To the best of our knowledge there has been no investigation of the effect of repeated task engagement on gamma activity in schizophrenia. It is likely however, in light of the level of performance, that patients were exhibiting a behavioural floor effect and were thus unable to benefit from practice in the absence of stimulation. That this was reflected in decreased gamma is not inconsistent with research showing impaired gamma synchrony in patients with schizophrenia when increasing cognitive effort is required (Sun et al., 2011).

The current findings indicate that tDCS is able to enhance working memory performance in schizophrenia by modulating underlying DLPFC pathophysiology. In light of a number of limitations, however, this study should be considered preliminary. Such findings need to be confirmed and extended in larger samples using a more sophisticated EEG arrays which would provide the power to investigate the impact of tDCS across multiple brain regions and frequency bands. Additionally, if tDCS is to be of therapeutic relevance, the duration of tDCS induced cognitive enhancement, and underlying changes in neural synchrony, need to be investigated by providing repeated stimulation sessions and follow up assessments. Finally, despite the measures we took to minimise this confound, there remains a lack of clarity regarding the effects of medication on these results. Investigations of patients with schizophrenia not on medication is practically very challenging and, more importantly, raises significant generalisability issues with respect to ultimate therapeutic applications. Medication analyses would help elucidate any impact, however the current study did not have the power for this and it is recommended such analyses are undertaken in future investigations of larger samples. Despite these limitations, this study provides initial support

16

that tDCS appears to improve working memory performance in patients with schizophrenia by restoring gamma oscillatory function.

Acknowledgements: KH and PF were supported by National Health and Medical Research Council (NHMRC) fellowships. This research was supported by a Monash University grant.

Financial Disclosures: PF has received equipment for research from Brainsway Ltd, Medtronic Ltd and MagVenture A/S and funding for research from Cervel Neurotech and Neuronetics Ltd. PF has received consultancy fees as a scientific advisor for Bionomics. There are no other relevant conflicts of interest.

17

5. References Antal, A., Varga, E.T., Kincses, T.Z., Nitsche, M.A., Paulus, W., 2004. Oscillatory brain activity and transcranial direct current stimulation in humans. Neuroreport 15(8), 1307-1310.

Barbey, A. K., Koenigs, M., & Grafman, J., 2013. Dorsolateral prefrontal contributions to human working memory. Cortex, 49(5), 1195-1205.

Basar-Eroglu, C., Brand, A., Hildebrandt, H., Karolina, K., Mathes, B., Schmiedt, C., 2007. Working memory related gamma oscillations in schizophrenia patients. International Journal of Psychophysiology 64(1), 39-45

Brunoni, A.R., Nitsche, M.A., Bolognini, N., Bikson, M., Wagner, T., Merabet, L., Edwards, D.J., Valero-Cabre, A., Rotendber, A., Pascual-Leone, A., Ferrucci, R., Priori, A., Boggio, P., Frengi, F., 2010. Clinical research with transcranial direct current stimulation (tDCS), challenges and future directions. Brain Stimulation 5(3), 175-195.

Chen, C.M.A., Stanford, A.D., Mao, X., Abi-Dargham, A., Shungu, D.C., Lisanby, S.H., Schroeder, C.E., Kegeles, L.S., 2014. GABA level, gamma oscillation, and working memory performance in schizophrenia. NeuroImage 4, 531-539.

Curtis, C. E., & D'Esposito, M., 2003. Persistent activity in the prefrontal cortex during working memory. Trends in cognitive sciences, 7(9), 415-423.

Haatveit, B.C., Sundet, K., Hugdahl, K., Ueland, T., Melle, I., Andreassen, O.A., 2010. The validity of d prime as a working memory index, Results from the “Bergen n-back task”. Journal of Clinical and Experimental Neuropsychology 32(8), 871-880.

Haenschel, C., Bittner, R. A., Waltz, J., Haertling, F., Wibral, M., Singer, W., Linden, D.E.J. & Rodriguez, E., 2009. Cortical oscillatory activity is critical for working memory as revealed by deficits in early-onset schizophrenia. The Journal of Neuroscience, 29(30), 9481-9489.

18

Hoaglin, D.C., Iglewicz, B., 1987. Fine tuning some resistant rules for outlier labeling, Journal of American Statistical Association 82, 1147-1149.

Howard, M.W., Rizzuto, D.S., Caplan, J.B., Madsen, J.R., Lisman, J., AschenbrennerScheibe, R., Schulze-Bonhage, A., Kahana, M.J., 2003 Gamma oscillations correlate with working memory load in humans. Cerebral Cortex 13(12), 1369-1374.

Hoy, K.E., Emonson, M.R., Arnold, S.L., Thomson, R.H., Daskalakis, Z.J., Fitzgerald, P.B., 2013. Testing the limits, Investigating the effect of tDCS dose on working memory enhancement in healthy controls. Neuropsychologia 51(9), 1777-1784.

Hoy, K.E., Arnold, S.L., Emonson, M.R., Daskalakis, Z.J., Fitzgerald, P.B., 2014. An investigation into the effects of tDCS dose on cognitive performance over time in patients with schizophrenia. Schizophrenia Research 155(1), 96-100.

Insel, T.R., 2010. Rethinking schizophrenia. Nature 468(7321), 187-193.

Jacobson, L., Koslowsky, M., Lavidor, M., 2012. tDCS polarity effects in motor and cognitive domains, a meta-analytical review. Experimental Brain Research 216(1),110.

Jaušovec, N., Jaušovec, K., 2012. Working memory training, improving intelligence– changing brain activity. Brain and Cognition 79(2), 96-106.

Kreyenbuhl, J., Buchanan, R.W., Dickerson, F.B., Dixon, L.B., 2010. The schizophrenia patient outcomes research team (PORT), updated treatment recommendations 2009. Schizophrenia Bulletin 36(1), 94-103.

Lett, T.A., Voineskos, A.N., Kennedy, J.L., Levine, B., Daskalakis, Z.J., 2013. Treating Working Memory Deficits in Schizophrenia, A Review of the Neurobiology. Biological Psychiatry. 75(5), 361-370.

19

Moran, L.V., Hong, L.E., 2011. High vs low frequency neural oscillations in schizophrenia. Schizophrenia Bulletin 37(4), 659-663. Poreisz, C., Boros, K., Antal, A., & Paulus, W., 2007. Safety aspects of transcranial direct current stimulation concerning healthy subjects and patients. Brain research bulletin, 72(4), 208-214.

Gonzalez-Burgos, G., Fish, K.N., Lewis, D.A., 2011. GABA neuron alterations, cortical circuit dysfunction and cognitive deficits in schizophrenia. Neural Plasticity doi, 10.1155/2011/723184.

Nitsche, M.A., Fregni, F., 2007. Transcranial Direct Current Stimulation-An Adjuvant Tool for the Treatment of Neuropsychiatric Diseases?. Current Psychiatry Reviews 3(3), 222-232.

Nitsche, M.A., Cohen, L.G., Wassermann, E.M., Priori, A., Lang, N., Antal, A., Paulus, W., Humel, F., Boggio, P.S., Fregni, F., Pascual-Leone, A., 2008. Transcranial direct current stimulation, state of the art 2008. Brain Stimulation 1(3), 206-223.

Semlitsch, H.V., Anderer, P., Schuster, P., Presslich, O., 1986. A solution for reliable and valid reduction of ocular artifacts applied to the P300 ERP. Psychophysiology 23, 695-703.

Sheehan, D.V., Lecrubier, Y., Sheehan, K.H., Amorim, P., Janavs, J., Weiller, E., Keskiner, A., Schinka, J., Knapp, E., Sheehan, M.F., Dunba, G.C., 1998. The Mini-International Neuropsychiatric Interview (MINI), the development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. Journal of Clinical Psychiatry 59, 22-33. Stagg, C.J., Best, J.G., Stephenson, M.C., O'Shea, J., Wylezinska, M., Kincses, Z.T., Morris, P.G., Matthews, P.M., Johansen-Berg, H., 2009. Polarity-sensitive modulation of cortical neurotransmitters by transcranial stimulation. The Journal of Neuroscience 29(16), 5202-5206. Stagg, C.J., Bachtiar, V., Johansen-Berg, H., 2011. The role of GABA in human motor learning. Current Biology 21(6), 480-484. 20

Sun, Y., Farzan, F., Barr, M.S., Kirihara, K., Fitzgerald, P.B., Light, G.A., Daskalakis, Z.J., 2011. Gamma oscillations in schizophrenia, mechanisms and clinical significance. Brain Research 1413, 98-114.

Uhlhaas, P.J., Singer, W., 2006. Neural synchrony in brain disorders, relevance for cognitive dysfunctions and pathophysiology. Neuron 52(1), 155-168.

Utz, K.S., Dimova, V., Oppenländer, K., Kerkhoff, G., 2010. Electrified minds, transcranial direct current stimulation (tDCS) and galvanic vestibular stimulation (GVS) as methods of non-invasive brain stimulation in neuropsychology—a review of current data and future implications. Neuropsychologia 48(10), 2789-2810.

Vinogradov, S., Fisher, M., de Villers-Sidani, E., 2012. Cognitive training for impaired neural systems in neuropsychiatric illness. Neuropsychopharmacology 37(1), 43-76.

Yeo, G.B., Neal, A., 2004. A multilevel analysis of effort, practice, and performance, effects; of ability, conscientiousness, and goal orientation. Journal of Applied Psychology 89(2), 231.

Zaehle, T., Sandmann, P., Thorne, J.D., Jäncke, L., Herrmann, C.S., 2011. Transcranial direct current stimulation of the prefrontal cortex modulates working memory performance, combined behavioural and electrophysiological evidence. BMC Neuroscience 12(1), 2.

21

Figure Legends: Figure 1. Illustration of tDCS setup and protocol. (a) anodal electrode was placed over F3 (left DLPFC) and cathodal electrode over the right supraorbital space. (b) Participants each underwent three experimental sessions spaced a week apart. tDCS was randomly applied at 1mA, 2mA or sham for 20 minutes across the three weeks and followed by the 2-back immediately following, 20 minutes and 40 minutes post stimulation with concurrent EEG recordings at F3, FZ F4. (c) illustration of n back task.

Figure 2. Means and standard errors for Gamma ERS and d prime performance over time as a function of stimulation condition, 2mA (a), 1mA (b) and sham (c).

22

Table 1. Patient demographics and clinical data 2mA

1mA

Sham

(n=16)

(n=13)

(n=12)

Gender (f/m)

6/10

5/8

4/8

Handedness (r/l)

16/0

13/0

12/0

Age

41.31 ± 10.25

40.62 ± 8.81

40.42 ± 10.65

Years of education

13.88 ± 1.78

13.92 ± 1.98

13.25 ± 1.60

Years since diagnosis

15.45 ± 8.57

15.22 ± 8.30

15.00 ± 7.18

PANSS Positive

17.69 ± 7.64

17.92 ± 6.86

18.75 ± 7.92

PANSS Negative

15.50 ± 3.95

15.85 ± 4.20

15.83 ± 4.22

PANSS General

36.06 ± 8.87

35.92 ± 8.49

37.92 ± 8.21

PANSS Total

69.25 ± 17.15

69.69 ± 15.99

72.50 ± 16.98

23

Table 2. Means and standard deviations for gamma ERS and dprime at 0, 20 and 40 minutes post stimulation.

0 mins

20 mins

40 mins

Gamma ERS

Mean

SD

Mean

SD

Mean

SD

Sham

0.406

10.080

-1.193

8.714

-4.887

11.308

1mA

0.847

6.526

-1.011

7.549

-3.623

8.501

2mA

-1.944

11.034

0.884

7.346

1.759

8.928

dprime

Mean

SD

Mean

SD

Mean

SD

Sham

2.737

0.516

2.819

0.859

2.649

0.829

1mA

2.919

0.546

2.954

0.651

2.728

0.640

2mA

2.905

0.598

3.172

0.473

3.164

0.697

Highlights • We investigated the effect of tDCS dose on gamma activity in schizophrenia. •

2mA tDCS significantly increased gamma ERS and improved working memory.

•

tDCS may enhance working memory in schizophrenia by restoring gamma activity.

24

Table 1

Table 1. Patient demographics and clinical data 2mA

1mA

Sham

(n=16)

(n=13)

(n=12)

Gender (f/m)

6/10

5/8

4/8

Handedness (r/l)

16/0

13/0

12/0

Age

41.31 ± 10.25

40.62 ± 8.81

40.42 ± 10.65

Years of education

13.88 ± 1.78

13.92 ± 1.98

13.25 ± 1.60

Years since diagnosis

15.45 ± 8.57

15.22 ± 8.30

15.00 ± 7.18

PANSS Positive

17.69 ± 7.64

17.92 ± 6.86

18.75 ± 7.92

PANSS Negative

15.50 ± 3.95

15.85 ± 4.20

15.83 ± 4.22

PANSS General

36.06 ± 8.87

35.92 ± 8.49

37.92 ± 8.21

PANSS Total

69.25 ± 17.15

69.69 ± 15.99

72.50 ± 16.98

1

Table 2

Table 2. Means and standard deviations for gamma ERS and dprime at 0, 20 and 40 minutes post stimulation.

0 mins

20 mins

40 mins

Gamma ERS

Mean

SD

Mean

SD

Mean

SD

Sham

0.406

10.080

-1.193

8.714

-4.887

11.308

1mA

0.847

6.526

-1.011

7.549

-3.623

8.501

2mA

-1.944

11.034

0.884

7.346

1.759

8.928

dprime

Mean

SD

Mean

SD

Mean

SD

Sham

2.737

0.516

2.819

0.859

2.649

0.829

1mA

2.919

0.546

2.954

0.651

2.728

0.640

2mA

2.905

0.598

3.172

0.473

3.164

0.697

1

a

Figure 1

F3

Fz

F4

b tDCS/EEG Set Up (1mA, 2mA or Sham)

20 mins tDCS

c C

20 mins

A B

match

A

2back + EEG 40 mins

non- match

2back + EEG

500ms Stimulus Presentation 1500ms Inter-stimulus interval

0 mins

2back + EEG

Figure 2

c

b

a

GAMMA ERS 0

2

4

-8

-6

-4

-2

0

2

4

-8

-6

-4

-2

-8

-6

-4

-2

0

2

4

GAMMA ERS

GAMMA ERS

0 mins

0 mins

0 mins

20 mins

20 mins

20 mins

40 mins

40 mins

40 mins

Sham

1mA

2mA

2.4

2.6

2.8

3

3.2

3.4

2.4

2.6

2.8

3

3.2

3.4

2.4

2.6

2.8

3

3.2

3.4

2back dPrimje 2back dPrime

2back dPrime

0 mins

0 mins

0 mins

20 mins

20 mins

20 mins

40 mins

40 mins

40 mins

1mA

Sham

2mA