Enhancing Visual Working Memory by Transcranial Direct Current Stimulation Over The Right Dorsolateral Prefrontal Cortex
Thomas Horst 10378774
Alexander Laufer & Ilja Sligte
Total word-count (including references, graphs/tables, titles): 3939 29-05-2015
Abstract
It has been found that transcranial direct current stimulation (tDCS), over the left
dorsolateral prefrontal cortex (DLPFC) could improve visual working memory (VWM) performance. However, whether VWM performance could also be enhanced by stimulating the right DLPFC has not yet been explored. This study investigated whether anodal-tDCS applied over the right DLPFC, was associated with enhanced VWM performance compared to sham stimulation. Sixteen subjects received active and sham tDCS over the right DLPFC, while performing a change-detection task. Results show that active compared to sham stimulation, did not significantly improve VWM performance. However, a significant improvement of VWM performance was found in the sham stimulation condition, indicating a placebo-effect. Future studies should continue to investigate the functional differences between the left and the right DLPFC in VWM. To further explore placebo-effects in tDCS research, the relation between tDCS and VWM should be investigated in a sham-controlled double blind manner.
Introduction
Baddeley’s influential theory of working memory refers to the term working memory as a brain system that provides temporary storage and manipulation of the information necessary for such complex cognitive tasks as language comprehension, learning, and reasoning (Baddeley & Hitch, 1974). It is nowadays widely accepted that that the prefrontal cortex, and in particular, the dorsolateral prefrontal cortex (DLPFC) plays a crucial role in working memory (WM) (Paulesu, Frith, Bench, Bottini, Grasby, Frackowiak, 1993; Salmon et al., 1996; Zimmer, 2008). Understandably, for cognitive neuroscientists the enhancement of working memory is a subject of great interest. In addition to approaches of psychotherapy and pharmacotherapy for enhancing WM, in recent years, non-invasive brain stimulation techniques are a rising topic of study.
In particular, transcranial direct current stimulation (tDCS) has recently been used to explore the influence of DLPFC activity on WM performance. tDCS consists of applying a weak, direct electric
current that flows from the anode to the cathode. These electrodes are placed over the scalp with the goal of, respectively, increasing and decreasing cortical excitability (Brunoni, Russowsky & Vanderhasselt, 2014). tDCS has the advantage of being safe, easily applied, and reliably blinded by sham tDCS in clinical trial settings (Gandiga, Hummel & Cohen,2006). Several studies have shown that tDCS is an effective method for manipulating the excitability of the human cortex (Nitsche & Paulus, 2000).
There are several reason for choosing the DLPFC as a candidate for neuro-stimulation. Neurons in the DLPFC show sustained activity during delay of VWM tasks in monkeys (Miller, Erickson & Desimone, 1996). It is also shown in monkeys that Lesions of the DLPFC greatly impair WM performance. Recently, transcranial magnetic stimulation (TMS) studies have confirmed that the DLPFC is crucial for WM maintenance in humans as well (Oliveri et al, 2001; Turatto, Sandrini & Miniussi, 2004; Sligte, Wokke, Tesselaar, Scholte & Lamme, 2011)
Since discovering the link between WM and the DLPFC several studies have used tDCS to increase DLPFC activity, and thereby improve WM. These studies suggest that although tDCS may not be useful in localizing specific brain-function, tDCS can be especially well suited for more practical applications such as the enhancement of WM (Brunoni et al, 2014; Fregni et al, 2005; Ohn et al, 2008; javadi & walsh, 2011). tDCS may improve working memory in clinical populations as well. It has been found that tDCS over the DLPFC may exert a beneficial effect on WM in both patients with Parkinson’s disease and stroke patients. (Boggio et al, 2006; Jo et al, 2009).
It seems therefore that unilateral anodal tCDS stimulation over the left DLPFC may enhance WM. However, in previous tDCS studies, little attention is paid to the right DLPFC in VWM
performance. While the right DLPFC is shown to be critically involved in a large range of working memory dimensions (Paulesu et al, 1993; Salmon et al, 1996; Zimmer, 2008).
Seen the consensus in the literature of the importance of the right DLPFC in working memory (Paulesu et al, 1993; Salmon et al, 1996; Zimmer, 2008). And in particular the finding that VWM gets impaired when TMS is applied on the right DLPFC, suggests a causal role for the DLPFC on
VWM and the right DLPFC in particular (Sligte et al, 2011). The question is raised if VWM can be enhanced by boosting neural activity in the right DLPFC.
In summary, this study sought to investigate whether anodal tDCS of the right DLPFC is associated with a change in VWM task performance (change-detection task) when compared to a sham tDCS condition. If the right DLPFC is indeed critically involved in VWM, and an increase of neural activity is important for this process, it is expected that the stimulation would improve task performance.
Method Participants
Sixteen right-handed healthy students of which 11 females, and five males participated in the experiment. The mean of age and standard deviation was; 20.4 (SD=1.7). Subjects gave their informed consent which was approved by the ethics committee of the department of Psychology of the University of Amsterdam. None of the subjects had a history of neurological diseases, and they all had normal vision. For participation subjects received either no compensation or research participation points required for their graduation. All subjects reported transient burning and/or itching sensations during the active and sham stimulation sessions, but these sensations subsided quickly, usually within a minute. Overall, subjects reported that they tolerated the stimulation well and did not report discomfort. Although subject number four explicitly reported that the itching sensation was persistent, and that he could not perform well because of it. This was during the second block (during stimulation) of his third session. In addition, during the first session the VWM scores of block three (after stimulation) session one of the same subject were not saved due to technical difficulties, this data was excluded from analysis. The rest of his data was not excluded from analysis because it did not show any abnormalities.
Materials
Direct current stimulation
Direct current was transferred with a Direct Current-Stimulator Plus (neuroConn GmbH, Ilmenau, Germany) This device was capable of delivering the anodal/cathodal direct current and also the sham stimulation required for the study. To stimulate the DLPFC, saline-soaked electrodes (3 X 3 cm) were placed over the scalp in the following manner. The anode was placed over F4 above the right DLPFC as determined by the international 10-20 electrode-placement system. The reference electrode (5 X 7 cm) was placed over the left forehead. Stimulation sessions took 20 minutes. At the beginning of the session a constant current fade-in took place until reaching 1 mA within 60 seconds, and fade-out in 60 seconds at the end of the session. For sham stimulation the electrodes were placed in the same manner. However, to give the subjects the impression that they were stimulated and blind them for the stimulation condition, only the constant current fade-in and fade-out took place.
VWM assessment
In this study a slightly modified version of the change-detection paradigm described by Sligte et al. (2011) was used. Stimuli were generated using the Presentation 18.1 software package
(Albany, California, USA), and were presented on a computer display, from a viewing distance of 57 cm.
One trial sequence is shown in figure 1. Throughout the whole experiment a black fixation cross was present which only turned green for 0.5 sec to indicate the start of the trial. Subjects were instructed to maintain fixation during every trial. Black rectangles were presented by the memory and test arrays and could be oriented horizontally, vertically, 45◦ to vertical or 135◦ to vertical and were presented against a white background. All the orientations were randomized and shown equally often.
Every trial was structured as follows. After the fixation cross turned green for 0.5 sec indicating the start of a trial, the memory display was shown for 0,25 sec for which they were instructed to remember the orientations to their best ability. Next, a blank retention interval was shown for 1 sec, after which the test display appeared, showing one orientation for 3 sec. During the test display, subjects had to respond by pressing a button indicating whether the orientation, shown on the test display, was the same as the one held in memory, or whether the orientation had changed. In 50% of the trials the orientation in the test display was the same as the orientation in the memory display (no-change trial), and in 50% of the trials this orientation changed 90◦ (change trial). When a right answer was given, a single (beeping) sound was presented. When a wrong answer was given, no sound was presented. This sound was included with the purpose of giving subjects feedback on their task-performance.
Each trial lasted 5.75 seconds. Each block consisted of 160 trials, with 90 seconds break every 53 trials. The total block duration was therefore given by 160 x 5.75 (trial time) + 3 X 90 (break time) = 19 minutes / 83 seconds.
Figure 1. This figure displays one trial of the change-detection task.
Exclusion criteria
For screening a survey was used which was included in the informed consent form. For safety reasons participants were excluded based on a history of neurological injuries and/or disorders, left handedness, pregnancy, colorblindness, medication use, multiple sclerosis, pacemaker, and skin disorders.
Procedure
Each subject underwent three sessions with at least a 72 hour intervals to minimize confounds and the effects of previous runs. During the first introduction session, subjects were briefed about the procedure, received instructions, were screened for exclusion factors, and gave their informed consent. If all the requirements were met and informed consent was given, subjects underwent a tDCS test-run to experience tDCS first hand, and to determine their tolerance. During this test-run subjects experienced a fade-in and fade-out of 1-mA direct current stimulation. All subjects were explained that they were always allowed to stop without reason and further consequences. After this subjects could decide to participate or not. When subjects wanted to participate, they received 1 hour of training in the working memory task, during which 3 blocks of 160 trials were performed. During the experimental sessions, subjects performed three blocks of 160 trials of the working memory task during both active and sham stimulation. The order of active and sham stimulation was counterbalanced between subjects, and the application was carried out double-blind. Prior to every experimental session, tDCS-equipment was attached to the head of the subjects. It was only during the second block of trials that the tDCS-equipment was turned on, and active stimulation took place. After the experiment subjects were debriefed and received their compensation.
Data Analysis
All analyses were performed using the SPSS 20.0 statistical software package (Chicago, Ilinois, USA). Effects tDCS were assessed with a 2 x 3 repeated-measures analysis of variance (ANOVA) with stimulation condition (active or sham stimulation) and time (before, during, and after stimulation) as within-subjects factors. The outcome measure was the percentage of correct scores on the VWM task. Separate two-tailed paired-samples t tests were done to further explore time effects on the outcome measure. Statistical significance refers to a two-tailed P- value <0.05.
Results
Two subjects had to quit the experiment due to planning issues. Sixteen subjects completed the entire experiment.
To analyze VWM performance, average VWM scores for every block were calculated, and the influence of stimulation condition (active and sham) and time (before, during, and after stimulation) on VWM scores were verified by means of a two-way repeated-measures ANOVA. Results of ANOVA revealed a non-significant effect of stimulation condition, F (1,15) = 0,78, p = .390, and a significant effect of time, F (2,30) = 5,52, p<0,05, see figure 2. No significant interaction effect was found between stimulation condition and time, F (2,30) = 0,53, p = .591. Means and standard deviations of the VWM scores are shown in Table 1. In sum, this analysis showed that active stimulation did not improve VWM performance compared to sham stimulation, but confirmed that time effects influenced the results.
To further analyze time effects, average VWM scores for every block were compared by means of separate paired t tests. Results of t tests revealed no significant difference in performances before active stimulation and during active stimulation, t (15) = -1,772, p= .097, or during active stimulation and after active stimulation, t (15) = -0,365, p= .720. However, performance improved significantly during sham stimulation compared to before sham stimulation, t (15) = -2,557, p<0.05, but not during sham stimulation compared to after sham stimulation, t (15) = 0,588, p= .565. See Figure 3. These comparisons failed to uncover a significant effect of tDCS. However, VWM performance did significantly improve during sham stimulation compared to before sham stimulation.
To test for learning effects, two paired samples t tests were conducted to compare VWM performances in the third block of the first session, with the performances before active stimulation, and the performances before sham stimulation. Results revealed no learning effects, t (14) = -0,160, p= .875 and t (14) = -0,028, p= .978.
60 61 62 63 64 65 66 67 68 69 70
Before During After
Co rre ct re sp on se (%) Time
Table 1. This table displays all the Paired t tests with corresponding mean, standard deviation, t-value, and p-value.
M (SD) t-value p-value
Training block 3 – Before tDCS -0,21 (5,05) -0,16 .875 Training block 3 – During Sham -0,42 (5,73) -0,028 .978
Before Sham – During Sham -4,80 (7,52) -2,557 .022*
During Sham – After Sham 1,09 (7,44) 0,588 .560
Before tDCS – During tDCS -2,34 (5,29) -1,772 .097
During tDCS – After tDCS -0,70 (7,69) -0,365 .720
Figure 2. This graph depicts the effect of time on VWM performance. The three time points displayed are before, during, and after both active and sham stimulation. VWM performance is displayed as the correct responses made in percentages on the change-detection task. The mean response is displayed for all three sessions and all participants. Error bars are included (standard error of the mean).
54 56 58 60 62 64 66 68 70 72
Before During After
Co rre ct re sp on se (%) Time Training tDCS Sham Figure 3. This graph depicts the correct responses made in percentages on the change-detection task for both the active and sham stimulation conditions. Correct responses made during the training session are also included. The three time points displayed are before, during, and after stimulation. For the training session the before, during, and after time points refer to the first, second, and third block of the training session. The mean response is displayed for all three sessions, and all participants. Error bars are included (standard error of the mean).
Discussion
The DLPFC is strongly associated with VWM performance (Paulesu et al, 1993; Salmon et al, 1996; Zimmer, 2008). Application of TMS over the right DLPFC impaired VWM performance,
suggesting a causal role for the right DLPFC on VWM (Sligte et al, 2011). This study sought to investigate whether anodal tDCS applied over the right DLPFC is associated with enhanced VWM performance compared to sham stimulation. It was expected that application of anodal tDCS over the DLPFC would result in an enhancement of VWM. In contrast with our hypothesis, results suggest that anodal tDCS applied over the right DLPFC for 20 minutes at an intensity of 1 mA was not
associated with enhanced VWM performance compared to sham stimulation. Even though this study failed to show such an effect, results did reveal a significant positive effect of sham stimulation on VWM performance, which indicates a placebo effect.
Previous studies exploring tDCS and WM have shown that various types of anodal tDCS could enhance WM performance (Brunoni et al, 2014; Fregni et al, 2005; Ohn et al, 2008; javadi & walsh, 2011). However, methodological considerations should be made. These studies have
produced diverse behavioral effects that could have been due to the use of different methodologies. Therefore, direct comparisons cannot be made.
First, afore mentioned studies were often conducted single-blinded or not included a sham stimulation control condition (Brunoni et al, 2014), while the current study was conducted in a sham-controlled, double blind manner, thus controlling for observer bias and expectancy effects.
Second, performance was measured with a different behavioral task. In this experiment a change detection paradigm was used to measure VWM performance, whilst the before mentioned studies all used N-back WM tasks to measure WM performance.
Third, but most importantly, in this study tDCS and VWM were explored by stimulating the right DLPFC, whilst the others did so by stimulating the left DLPFC. It should be noted that there is little evidence up to date that explains whether the effect of tDCS in working memory works through modulating the excitability of the DLPFC, and if so, under what possible mechanisms. Keshvari, Pouretemad and Ekhtiari, (2013) found that the effect of tDCS on WM performance is dependent of electrode positioning, and that bilateral stimulation of the DLPFC has a negative effect on WM performance. These findings indicate that the type of electrode positioning may differentially affect WM operations. Therefore, it could be that we incorrectly stated our hypothesis. It was
hypothesized that an increase in neural activity would lead to an increase in VWM performance, but since the general lack of knowledge about the functionality of the DLPFC region, it is possible that a heightening of neural activity in the right DLPFC modulates VWM performance in other ways than was expected.
tDCS seems to be a promising tool for the enhancement of WM. However, more research using different stimulation/inhibition electrode positioning is needed to further distinguish the functional roles of the left and right DLPFC areas in relation to WM. This study adds to an important body of research investigating these roles.
This study found that performances improved significantly when subjects were sham stimulated compared to the first block in which they did not receive any stimulation. This could indicate a placebo effect since subjects were aware of the fact that they were being stimulated from the moment the tDCS device was turned on, only during stimulation sessions. The afore mentioned studies have often not included a sham stimulation condition, and were conducted single blinded. Therefore, it would be possible that they overlooked placebo effects because subjects could have known if they were stimulated or not due to observer effects. To investigate placebo effects in future tDCS research, it may be advantageous to include a condition where tDCS and VWM are investigated in a sham-controlled double-blind manner, and also include a condition where it is done in a single blind sham-controlled manner. If both conditions are included, and a placebo effect is found in the double-blind condition but not in the single-blind condition, it would be an indication that the placebo effect was the consequence of using a double blind design. When a placebo effect is found in both the single blind and double blind conditions, it would be an indication that it would not matter whether a double blind or single blind design is used, because the placebo effect is not a result of either one.
A double-blind experimental design could potentially be a simple and straightforward measure to control for placebo effects. Therefore, to further investigate placebo effects in tDCS research, and potentially control for these effects , future research should consider to both include a single blind and a double-blind experimental condition.
Limitations of this study include the following. First, because of a small sample size (16 subjects), the ANOVA might have lacked enough power to reliably detect a difference between the conditions. Therefore, a larger sample size would have made it more likely to find an effect of tDCS.
Second, the experiment was conducted in a room where subjects were possibly exposed to distracting factors. Since the room was located on a university campus, noises and voices from outside could have been distracting. Also, experimenters were situated in the same room as where the subjects were participating in the experiment. This could have been another distracting factor since experimenters were sometimes talking or occasionally leaving the room to get something, or to go to the toilet. Therefore, participants should be conducting their tasks in a separate room, and test areas should be sufficiently isolated from noise.
In conclusion, this study failed to show that anodal tDCS applied over the right DLPFC was associated with enhanced VWM performance. However, it did show a significant effect of sham stimulation on VWM performance, indicating a placebo effect. Future studies should continue to investigate the functional differences between the left and the right DLPFC in WM. To further explore placebo effects in tDCS research and possibly control for them, the relation between tDCS and WM should be investigated in a sham-controlled double blind manner.
References
Baddeley, A. D., & Hitch, G. (1974). Working memory. Psychology of learning and motivation, 8, 47- 89.
Boggio, P. S., Ferrucci, R., Rigonatti, S. P., Covre, P., Nitsche, M., Pascual-Leone, A., & Fregni, F. (2006). Effects of transcranial direct current stimulation on working memory in patients with Parkinson's disease. Journal of the neurological sciences, 249(1), 31-38.
Brunoni, Russowsky, A. & Vanderhasselt, M,. (2014). Working memory improvement
with non-invasive brain stimulation of the dorsolateral prefrontal cortex: A systematic review and meta-analysis. Brain and Cognition, 86, 1-9.
Daskalakis, Z. J., Farzan, F., Barr, M. S., Rusjan, P. M., Favalli, G., Levinson, A. J., & Fitzgerald, P. B. (2008). Evaluating the relationship between long interval cortical inhibition, working memory and gamma band activity in the dorsolateral prefrontal cortex. Clinical EEG and neuroscience, 39(3), 150-155.
Gandiga P., Hummel F., & Cohen L., (2006) Transcranial DC stimulation (tDCS): a tool for double-blind sham- controlled clinical studies in brain stimulation. Clinical Neurophysiology ;117(4):845– 50.
Javadi, A. H., & Walsh, V. (2012). Transcranial direct current stimulation (tDCS) of the left
dorsolateral prefrontal cortex modulates declarative memory.Brain stimulation, 5(3), 231- 241.
memory of stroke patients using tDCS. American Journal of Physical Medicine & Rehabilitation, 88(5), 404-409.
Keshvari, F., Pouretemad, H. R., & Ekhtiari, H. (2013). The Polarity-Dependent Effects of the Bilateral Brain Stimulation on Working Memory. Basic and clinical neuroscience, 4(3), 224.
Miller, E. K., Erickson, C. A., & Desimone, R. (1996). Neural mechanisms of visual working memory in prefrontal cortex of the macaque. The Journal of Neuroscience, 16(16), 5154-5167.
Nitsche, M. A., & Paulus, W. (2000). Excitability changes induced in the human motor cortex by weak transcranial direct current stimulation. The Journal of physiology, 527(3), 633-639.
Oliveri, M., Turriziani, P., Carlesimo, G. A., Koch, G., Tomaiuolo, F., Panella, M., & Caltagirone, C. (2001). Parieto-frontal interactions in visual-object and visual-spatial working memory: evidence from transcranial magnetic stimulation. Cerebral Cortex, 11(7), 606-618.
Ohn, S. H., Park, C. I., Yoo, W. K., Ko, M. H., Choi, K. P., Kim, G. M., ... & Kim, Y. H. (2008). Time- dependent effect of transcranial direct current stimulation on the enhancement of working memory. Neuroreport, 19(1), 43-47.
Paulesu, E., Frith, C. D., Bench, C. J., Bottini, G., Grasby, P. G., FrackowiakR. S. J. (1993). Functional anatomy of working memory: the visuospatial 'sketchpad'. Journal of Cerebral Blood Flow Metabolism. 1,552- 557.
Salmon, E., Van der Linden, M,. Collette, F,.Delfiore, G,.Maquet, P,.Degueldre, C. Luxen, A., Franck, G. (1996), Regional brain activity during working memory tasks. Brain. 119, 1617- 1625.
Sligte, I. G., Wokke, M. E., Tesselaar, J. P., Scholte, H. S., & Lamme, V. A. (2011). Magnetic stimulation of the dorsolateral prefrontal cortex dissociates fragile visual short-term memory from visual working memory. Neuropsychologia, 49(6), 1578-1588.
Turatto, M., Sandrini, M., & Miniussi, C. (2004). The role of the right dorsolateral prefrontal cortex in visual change awareness. Neuroreport, 15(16), 2549-2552.
Zimmer, H. (2008). Visual and spatial working memory: From boxes to networks. Neuroscience and Biobehavioral. Reviews. 32, 1373-1395.
