Transcranial Direct Current Stimulation (tDCS) as an intervention to improve empathic abilities and reduce violent behavior in forensic offenders: study protocol for a randomized controlled trail.

Background Recent studies show that changes in one of the brain areas related to empathic abilities (i.e. the Ventromedial Prefrontal Cortex (vmPFC)) plays an important role in violent behavior in abusers of alcohol and cocaine. According to the models of James Blair, empathy is a potential inhibitor of violent behavior. Individuals with less empathic abilities may be less susceptible and motivated to inhibit violent behavior, which causes a higher risk of violence. Recent neuroscientific research shows that modulating (stimulation or inhibition) certain brain areas could be a promising new intervention for substance abuse and to reduce violent behavior, such as the neurostimulation technique Transcranial Direct Current Stimulation (tDCS). This study aims to investigate tDCS as an intervention to increase empathic abilities and reduce violent behavior in forensic substance use offenders. Methods/design A total sample of 50 male forensic substance abuse patients (25 active + 25 sham stimulation) will be tested in a double-blind placebo-controlled study, from which half of the patients will receive an active stimulation + treatment as usual (TAU) and the other half will receive a sham stimulation (placebo) + TAU. The patients in the active condition will receive multichannel tDCS stimulation targeting bilateral vmPFC two times a day for 20 minutes for five consecutive days. Before and after the stimulation period, the patients will complete self-report measurements, perform the Point Subtraction Aggression Paradigm (PSAP) and a passive viewing empathy task. Resting state electroencephalography (rsEEG) will be measured before and after the treatment period. A follow-up will be conducted after six months. Primary outcome is to investigate multichannel tDCS as a new intervention to increase empathic abilities and wit that reduce violent behavior in offenders with substance abuse problems. In addition, it will be studied whether electrophysiological responses in the brain are affected by the tDCS intervention. Lastly the effects of tDCS on reducing craving will be investigated. Discussion This study protocol describes the design of an intervention using multichannel tDCS targeting the vmPFC. This study will explore the opportunities to introduce a new intervention to improve empathic abilities and reduce violence in forensic substance use offenders. In addition, this


Abstract
Background Recent studies show that changes in one of the brain areas related to empathic abilities (i.e. the Ventromedial Prefrontal Cortex (vmPFC)) plays an important role in violent behavior in abusers of alcohol and cocaine. According to the models of James Blair, empathy is a potential inhibitor of violent behavior. Individuals with less empathic abilities may be less susceptible and motivated to inhibit violent behavior, which causes a higher risk of violence. Recent neuroscientific research shows that modulating (stimulation or inhibition) certain brain areas could be a promising new intervention for substance abuse and to reduce violent behavior, such as the neurostimulation technique Transcranial Direct Current Stimulation (tDCS). This study aims to investigate tDCS as an intervention to increase empathic abilities and reduce violent behavior in forensic substance use offenders.
Methods/design A total sample of 50 male forensic substance abuse patients (25 active + 25 sham stimulation) will be tested in a double-blind placebo-controlled study, from which half of the patients will receive an active stimulation + treatment as usual (TAU) and the other half will receive a sham stimulation (placebo) + TAU. The patients in the active condition will receive multichannel tDCS stimulation targeting bilateral vmPFC two times a day for 20 minutes for five consecutive days. Before and after the stimulation period, the patients will complete self-report measurements, perform the Point Subtraction Aggression Paradigm (PSAP) and a passive viewing empathy task. Resting state electroencephalography (rsEEG) will be measured before and after the treatment period. A follow-up will be conducted after six months. Primary outcome is to investigate multichannel tDCS as a new intervention to increase empathic abilities and wit that reduce violent behavior in offenders with substance abuse problems. In addition, it will be studied whether electrophysiological responses in the brain are affected by the tDCS intervention. Lastly the effects of tDCS on reducing craving will be investigated.
Discussion This study protocol describes the design of an intervention using multichannel tDCS targeting the vmPFC. This study will explore the opportunities to introduce a new intervention to improve empathic abilities and reduce violence in forensic substance use offenders. In addition, this 3 study may help to reduce craving. The study will contribute to knowledge about increasing empathic abilities and the functioning of the vmPFC. Specifically, this study may give insight in how to implement the tDCS intervention in the setting of daily clinical practice for this complex, multiple problem target group and with that contribute to reduction of recidivism.
Trial registration Dutch Trial Register, identifier: NTR7701. Registered on 12 January 2019; prospectively registered before the recruitment phase. https://www.trialregister.nl/trial/7459 Background Repeatedly using substances has been found to lead to neuro-adaptations in the ventral striatum and ventral tegmental areas and with that decreased dopamine secretion [1]. Decreased dopamine secretion leads to a higher craving for substances and increased saliency for addictive cues [2][3][4].
Impaired functioning of the Dorsolateral Prefrontal Cortex (DLPFC) in patients with substance abuse disorders, are underlying to a diminished cognitive and inhibitory control and increase the tendency to relapse and maintain in addictive behaviors [5][6][7][8][9][10][11][12][13]. Recent studies show that changes in the brain areas related to less empathic ability (ea. the Ventromedial Prefrontal Cortex (vmPFC)) in abusers of alcohol and cocaine plays an important role in violent behavior [14], [15]. Preller and colleagues found in their study [15] that cocaine users show deficits in emotional empathy and that substance abuse patients are less emotionally responsive to the emotions of other individuals and their mental state.
Empathy is crucial for social enhancement, social interactions and relationships, for our emotional and social life [16]. A deficit in empathic ability could lead to antisocial and deviant behavior and with that a higher risk of aggression [15]. Aggression is stated here as behavior that is mostly defined by any behavior that is intended to harm someone who is motivated to avoid being harmed [17]. As said before, antisocial behavior, especially aggression, is associated with dysfunctions in the prefrontal cortex [18][19][20][21][22][23][24]. In addition, research has highlighted the importance of executive functions; the 'higher' cognitive functions that are controlled by the prefrontal cortex in control of aggression [25][26][27][28]. According to the models of James Blair [29], [30], [28] violent behavior is inhibited by empathy. Individuals with less empathic abilities may be less susceptible and 4 motivated to inhibited violent behavior, which causes a higher risk of violence.

The Role of the Prefrontal Cortex
Several studies show evidence that impaired prefrontal cortex areas lead to the emergence of aggressive behavior. Most notably are the impairments in the vmPFC (e.g. emotion regulation, moral decision-making), and the DLPFC (e.g. disinhibition and impulsiveness), which is associated with aggression and violent behavior [25][26][27][28]. These dysfunctional prefrontal cortex areas are inducing psychopathic traits such as blunted emotions and a lack of empathy [31], impaired perspective taking with increased egocentrism and rigidity [32], [19], [33]. Research found that children with psychopathic traits showed abnormal activity in the ventromedial prefrontal cortex during a response reversal task in comparison to children without these traits (Finger et al., 2008). In addition, Raine and colleagues [34,35] demonstrated the relationship between violence and empathy and argued that abnormal brain structures of the vmPFC cause the most extreme form of empathic inability named 'the psychopathic predatory violence'. This psychopathic predatory violence correlates with the lack of empathy, impaired moral judgment [34] and hyper metabolism [35], which is associated with increased aggressive impulses. Figure 1 shows the hypothesized relationship between the impaired brain areas and the subsequent deficits in empathy and violent behavior.
Although, the DLPFC modulates cognitive control and response inhibition which are associated with aggression and violent behavior [25][26][27][28], in this study we focus on the vmPFC.

Ventromedial Prefrontal Cortex (vmPFC)
The vmPFC is particularly relevant for empathic abilities and antisocial behavior [36], [29], [37]. Studies using functional neuroimaging (i.e. fMRI) have shown that neural activity in the vmPFC predicts empathic abilities and altruistic motivation [38], [39]. Another longitudinal study [19] demonstrated that lesions damaging the vmPFC, occurring in the first sixteen months of a human life, result in lifelong psychopathic antisocial traits and will also lead to impaired social and moral 5 reasoning. In addition, damage to the vmPFC is associated with poor decision-making in antisocial behavior [40], [41].
Furthermore, a recent study [42] combining tDCS with fMRI demonstrated that anodal tDCS placed on the forehead led to increased vmPFC activity and decreased negative emotions. Taken together, these studies suggest a potential link between vmPFC functioning and anger regulation.

Transcranial Direct Current Stimulation (tDCS)
tDCS is a non-invasive neuromodulation technique that modulates the brain region of interest by increasing or decreasing neuronal excitability through constant, low direct current electrodes. tDCS has been proven to be an effective intervention to modify brain activity [43], [44], [45] and has been investigated in many different disorders [46][47][48].
In this current study, the effectiveness of tDCS as a new intervention to increase empathy and reduce violent behavior in substance abuse offenders will be investigated. Through modulating (stimulating and inhibiting) certain areas of the brain, tDCS causes a change in the function of the brain, due to an increase in susceptibility to generate and facilitate brain related electrical impulses.
Several studies found that through anodal stimulation of the PFC with tDCS, emotional processes can be influenced. tDCS can modulate emotional pain [62,63] and enhance pain empathy [63]. Interestingly, studies [65][66][67] suggest that anodal tDCS of the PFC can enhance empathy and increase the feeling of morality. Although there seems to be a clear association between empathy and vmPFC function, only a few studies have investigated modulation of the vmPFC. Abend and colleagues [68] show in their study combining tDCS and fMRI that stimulating with tDCS demonstrated increased emotion-related activation in the vmPFC. Two studies show that anodal stimulation of the (right) vmPFC increases empathic ability and morality [67,68]. Furthermore, the study of Gilam and 6 colleagues [69] used tDCS simultaneously with fMRI and demonstrates that activity of the vmPFC was increased during active compared to sham stimulation.
In addition to empathy, also antisocial behavior, including aggression, are associated with dysfunctions in the prefrontal cortex [18][19][20][21][22][23][24]. Research shows that studies targeting the PFC with anodal tDCS can reduce social exclusion and the aggressive behavior that emerges from this exclusion [70], can reduce unprovoked aggressive behavior [71] and reduce the intentions of aggressive behavior [72,73]. Based on the earlier mentioned theory of Blair, linking impaired empathic abilities to violence, and the model demonstrated (Figure 1), it could be proposed that the vmPFC plays a crucial role in modulating both empathic abilities and thereby also in violent behavior.
Previous mentioned study of Gilam and colleagues [69] also demonstrates the relationship between the vmPFC and violent behavior and showed a decrease in violent behavior in the first session after active tDCS.
In addition to aberrant functioning of specific brain areas, recent research [74,75] also shows that functional connectivity of brain areas is also affected in individuals with less empathic abilities. Therefore, it is important for the evaluation of the effectiveness of a tDCS intervention, to determine how the resting-state connectivity of the patients, measured with resting state electroencephalography (rsEEG), changes from pre-to post-test to the intervention. Prior research demonstrates that tDCS can alter functional connectivity and that connectivity change is related to the treatment response (i.e. effectiveness) [76]. To optimize the intervention and increase current focality we use a newer development of the original. This technique, called high-definition tDCS (HD-tDCS or multichannel tDCS) uses gel-based electrodes similar to those used in EEG [77] and is more precise in targeting the brain area of interest.
Therefore, in this study we will investigate the idea that multichannel tDCS applied over the vmPFC will increase empathic abilities and subsequently reduce violence risk.

Aims of the study
The aim of this study is to investigate the effectiveness of tDCS in increasing empathy and reducing violent behavior in offenders with substance abuse problems.

7
The following research questions will be addressed in this study:

1.
Does stimulation with multichannel tDCS targeting the vmPFC increase empathic abilities in forensic substance abuse patients during an empathy task from pre-to post intervention?

2.
Does stimulation with multichannel tDCS targeting the vmPFC reduce aggressive responding and violence risk in forensic substance abuse patients from pre-to post intervention?
3. Does multichannel tDCS targeting the vmPFC reduce craving in forensic substance abuse patients?

4.
Does multichannel tDCS targeting the vmPFC affect electrophysiological responding in the brain measured through EEG from pre-to post intervention?
Outcomes will contribute to the development of more effective diagnostics and treatment of substance use patients. Furthermore, these data will potentially contribute to improving treatment through increasing understanding of specific targets for treatment interventions. In addition, outcomes may provide better insight in the functioning of the vmPFC and the relationship between empathy and violence risk.

Methods/design
Sample A total sample of 50 male forensic substance abuse patients (25 active + 25 sham stimulation) will be tested in a double-blind placebo-controlled study, from which half of the patients will receive an active stimulation+ treatment as usual (TAU) and the other half will receive a sham stimulation (placebo) + TAU.

Inclusion
In order to be eligible to participate in this study, a subject must meet all of the following criteria: male, age 18-60, good understanding of the Dutch language, diagnosed with an alcohol and/or cocaine SUD according to the DSM-5. The patients have to be abstinent and have an index offense in 8 violence category listed in the HKT-R. These criteria have been selected due to the fact that male subjects will show more aggression on the PSAP task and show more violent behavior in general than females.

Exclusion
Subjects meeting any of the following criteria will be excluded from participation in this study: major neurological conditions (e.g. traumatic brain injury), or major mental disorders (i.e. major depression, psychotic symptoms). Also patients taking antipsychotic or other severe medication cannot participate in the study.

Setting
The study will be carried out in two forensic institutions. The sample will be recruited from the Forensiche Verslavings Kliniek (FVK), the forensic addiction clinic of Bouman, Antes. The institution is located in Rotterdam, The Netherlands. The sample size is based on other studies, we will operate according to the 'Evidence-based guidelines on the therapeutic use of transcranial direct current stimulation (tDCS)' published in 2017 [78], and seen as achievable, due to the fact that the applicant is currently obtaining research at FVK Bouman Antes, which will make the inclusion of the patients more feasible.

Procedure /Design
In this double blind, placebo-controlled study, a total of 50 male participants between the age of 18-60 will be randomly assigned to either the active condition, or the sham condition. Eligible participants will be given written and oral information about the study and will be invited to participate. After informed consent they will participate at the forensic clinic where they are admitted and all the data collected will be anonymous linked to their participantnumber.
During the first session the following self-reported measures will be assessed. Psychopathic traits will be assessed using the short form of the Self-Report Psychopathy Scale Short-Form (SRP-SF) [79] aggression will be measured using the Dutch Reactive and Proactive Aggression Questionnaire 9 (RPQ), [80] empathy will be measured using the Dutch Interpersonal Reactivity Index (IRI) [81].
Alexithymia is a supposed cause of disruption in the accurate representation interpretation of affective states, causing empathy deficits. Therefore, alexithymia will be assessed using the Toronto Alexithymia Scale (TAS-20) [82]. Risky, impulsive and Self-destructive behavior is measured with the Dutch version of the Risky, Impulsive, and Self-destructive behavior Questionnaire (RISQ) [83].
To assess impulsive behaviour the Behavioral Impulsivity Scale (BISS-11) [84] [87] will be used. The craving for alcohol will be measured using the Obsessive Compulsive Drinking Scale (OCDS) [88] In addition; EEG will be measured during a resting-state task and during a Passive-viewing Empathy task [73]. Following the participants will perform a rating viewing Empathy task [73]. To measure behavioral indices of the level of aggression the patients will perform in the Point Subtraction Aggression Paradigm (PSAP) [89] will be used.
After this, a series of ten tDCS sessions (two per day for five consecutive days) will follow.
After every tDCS session the participants will fill in a questionnaire that assesses side effects. On Monday, after the last modulation on Friday, the patients will receive a post-intervention in which the rsEEG and the tasks (empathy task and PSAP task) again will be conducted. As a follow up, risk assessment scores of the participant will be obtained after six months to see whether risk reduction is maintained together with one last EEG session with the PSAP task.
Overview procedure 1). Inclusion of patients after informed consent.
2). Collecting information from patients regarding diagnosis, substance abuse, demographic information and HKT-R.
3). Pre-intervention in which the patients will receive a resting-state EEG (rsEEG) and perform the aggression task (PSAP) and the empathy task (victims of aggression). In addition, the patients will fill in the self-report questionnaires. 4). Intervention multichannel tDCS+TAU or Sham+TAU: twenty minutes of anodal stimulation of the vmPFC and cathodal stimulation of the left supraorbital area with multichannel tDCS, two times daily for a period of 5 consecutive days. 5). Post-intervention where the patients receive a rsEEG and will perform the aggression task (PSAP) and the empathy task (victims of aggression), and fill out the self-report questionnaires again. 6). Outcome variables: The results on the aggression task and empathy task as well as the score on the HKT-R and the self-report questionnaires will be lower with respect to the results on the pre-test.
Furthermore, it will be investigated whether the functional connectivity (EEG) has changed at the post-test in contrast to the pre-test and if this differs within the two conditions (tDCS vs. Sham). 7). A follow-up after six months where the patients will perform the aggression task and the empathy task one more time and the HKT-R will be administered.
See Figure 2 and 3 for a detailed overview over the procedure of the trial.
Multiple studies [90][91][92][93] have concluded that activating the brain state of the area of interest during stimulation, increases the effect of the modulation and contribute to optimize the intervention. An interesting study of Nissim and colleagues [94] demonstrated with fMRI that the optimal gains from using tDCS can be realized by simultaneously using behavior and modulation for stimulating neural networks. Therefore, to optimize our intervention and increase the activity of the vmPFC and the empathic abilities the participants will be occupied with tasks and movies that trigger these brain states. On the first day, the participants will complete the Multifaceted Empathy Task (MET) [94] and the Reading the Mind in the Eyes test (rMEt) [95], to actively enhance their perspective taking and emotion recognition. When finished, they will watch the movie 'Wonder' (2017), for the first part of the stimulation week. For the remaining sessions they will watch the movie 'I am Sam' (2001). The two movies are enhancing the empathic abilities and perspective taking, and therefore will contribute to the optimization of brain state.

Primary Objective
The primary objective is to investigate tDCS as a new intervention to increase empathy and reduce violent behavior in offenders with substance abuse problems. This will be measured with the results of the empathy and aggression task from pre-to posttest. In addition, it will be studied whether electrophysiological responses in the brain are affected by tDCS intervention, this will be demonstrated through comparing the EEG from pre-to posttest. A mixed design will be carried out to test whether ERP amplitudes pre-post intervention differs between the active-and sham group.

Secondary Objective(s)
The secondary objective is to reduce craving in offenders with substance abuse problems. This will be assessed using the four self-report measurements about alcohol-and drug craving, and to compare the results from pre-to posttest. Other study parameters are the results from the self-report questionnaires. These will be compared with the other variables mentioned before and also compared from pre-to posttest.

Outcomes
To test the effect of the intervention on aggression the PSAP will be used. As a standard for effectiveness of the reduction of violent behavior next to the PSAP, the results of a violence risk assessment instrument (HKT-R) [96] will be assessed before the intervention starts. The HKT-R is one of the mandatory risk-assessment tools in forensic institutions.
To measure functional changes in brain function during empathic induction caused by the intervention, rsEEG will be used. The expectation is that patients who receive tDCS intervention will show higher event-related potential (ERP) towards the pictures of the victims after the intervention and compared patients who have received the sham-condition.
After six months, as a standard for effectiveness of the reduction of violent behavior, the results of the aggression task (PSAP) and the risk-assessment tool (HKT-R) will be compared to the results previous to the intervention. The HKT-R risk level will be obtained to see whether there is a longitudinal effect of tDCS on aggression and risk reduction. In addition to check whether the effect of the tDCS intervention is long lasting, the participants will perform the self-report questionnaires, empathy task and the aggression task once more.

12
The recruitment will be active within the forensic institutions. Patients will be actively approached by the PhD student, or can sign up through a form at the department in which they receive their treatment. Research assistants will assist with the assessments. The entire research team will be trained extensively before they have an active role in the research.
After a screening for exclusion criteria, the patients will receive more detailed information about the study and can decide to participate definitive. An informed consent will be signed prior to the actual test day.

Experimental
The patients participating in the active, experimental condition start on Friday with completing the before mentioned self-report questionnaires. Also, EEG will be measured during a resting-state task (rsEEG) and during the passive viewing empathy task. After disconnecting the EEG device, participants will perform the empathy rating task and the PSAP. On Monday, the participants will start with the tDCS intervention; they will receive two times 20-minute tDCS stimulation each day. There

Sham
The patients participating in the sham condition will follow exactly the same routine as the participants in the active condition except they will receive 2 mA during 30 seconds instead of the 20 minute stimulation. This has proven to be effective for blinding as participants habituate to the sensation of stimulation within seconds of current initiation [97].

Blinding
Participants and experimenters are blind to the tDCS condition allocation. An external researcher is 13 the only one who knows which participant number corresponds with each condition. The principal experimenter, neither the patients nor research assistants know which condition is being executed.
The trial established procedures to maintain separation in knowledge between the head researcher and the principal experimenter.

tDCS
For the intervention we will operate according to the 'Evidence-based guidelines on the therapeutic use of transcranial direct current stimulation (tDCS)' [78]. Multichannel tDCS will be administered with a CE-certified neurostimulator (StarStim-8, NeuroElectrics) following the protocol [79] the device is preprogrammed for stimulation with 2mA during 20 minutes (experimental condition), or with 2 mA during 30 seconds (sham condition). We program the device for each participant, ensuring that the participant and experimenter are blinded for experimental condition. The experimental treatment is a 20 minutes tDCS session, two times a day for five consecutive days (ten sessions per participant).
Multichannel tDCS stimulation will be applied over the vmPFC. Sham condition is the same as the experimental condition with the exception that there will only be a ramp up of the electrical stimulation to mimic the sensation of the stimulation.

Passive-viewing Empathy task [73]
To measure empathic abilities and how it changes between pre-and post-intervention, patients will participate in a passive-viewing empathy task. This measurement is based on previous research [73]. pictures in total will be presented, for 6 seconds with intervals of 1.8 seconds between the pictures, randomly. Participants are instructed to look at each picture passively, because then the automatic neural responding in the brain can be determined [73].

Rating Empathy task [73]
Following the Passive Viewing Empathy task, the pictures displaying the neutral and aggressive

The Point Subtraction Aggression Paradigm (PSAP) [89]
Although multiple paradigms are used to measure aggression on an experimental manner in the labs, the PSAP-task is known to be one of the best-validated instruments [98]. During the PSAP-task participants are playing a game against a (fictive) opponent. The goal of the game is to earn more 15 points than your opponent. Participants are offered 3 choices: 1). Participants can earn points by pressing a 100 times on the '1' button on the keyboard. 2). Stealing points from your opponent by pressing ten times on the '2' button or 3). Guarding your own points as an escape so the opponent cannot steal from you by pressing ten times on the '3' button. When a participant picks option number 2 (aggressive response), the points that they steal will not add up at their own score, but only will subtracted from the score of the opponent. If the (fictive) opponent is stealing points from the participants it will be shown in red letters on the screen, in this way the participant will be 'provoked'.
If this provoking leads to the participant pressing the '2' (stealing) button more frequently, this can be seen as reactive aggression. When the participant is not provoked by this event, the aggression can be seen as proactive aggression. In this study the e-prime version of the PSAP consists of three sessions of each twelve minutes. The outcome of the PSAP, the aggressive response, is the amount of '2' responses of the participant. When the '2' response is a result of the provocation (the fictive opponent stealing points), then this indicates a reactive aggression response. If the '2' response is not a reaction on the provocation, than the reaction will be seen as proactive aggression response.
Research [99] concludes that the PSAP has more ecological validity then for example the Taylor Aggression Paradigm (TAP) [100]. This is due to the following advantages: First, the PSAP task offers the participant an option to actively withdraw (pressing the '3' key), something that is not included in other paradigm tasks. Secondly, in the PSAP task the aggressive response (pressing the '2' button is not receiving points) and therefore more ecological valid then other tasks that do not have this option. An aggressive reaction is a cost-benefit consideration, so choosing an option to steal points that will not add up to your own score of winning could be considered as a costly option.

Self-Report Psychopathy Scale (SRP-S) [79]
In order to assess psychopathic traits we use the Dutch short version of the SRP-SF. The SRP-SF consists of a subset of 29 of the 64 original items and is a self-report questionnaire in which participants are asked to rate statements with a 5-point Likert scale. The questionnaire consists of four subscales: Interpersonal manipulation (manipulation and deception), Callous affect (lack of empathy or regret), Instable lifestyle (impulsivity and sensation seeking behavior) and Criminal behavior (delinquency and criminal behavior).

Reactive and Proactive Aggression Questionnaire (RPQ) [80]
In order to assess aggression we use the Dutch version of the RPQ. The RPQ is a 23-item self-report questionnaire in which the participant has to give a rating on how often this behavior has occurred in the past on a 3-point scale ("never", "sometimes" or "often"). Next to an overall total aggression score, the test provides two separate measures to proactive aggression (deliberate and planned aggression) and reactive aggression (aggression as a reaction on an unplanned circumstance).

Interpersonal Reactivity Index (IRI) [81]
In order to assess empathy we use the Dutch version of the IRI, this is a commonly used self-report instrument designed to assess empathic tendencies. The IRI consists of four separate subscales: Perspective Taking (PT), Fantasy (FS), Empathic Concern (EC), and Personal Distress (PD).

Toronto Alexithymia Scale (TAS-20) [82]
In order to asses alexithymia we use the Dutch version of the TAS-20. The TAS-20 is a self-report scale comprised of 20 items. Each item is rated on a five point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). The TAS20 is a reliable and valid measure of emotion processing in adults that includes a total score and three subscales: Difficulty Identifying Feelings (DIF), Difficulty Describing Feelings (DDF), and Externally-Oriented Thinking (EOT).

Risky, Impulsive, and Self-Destructive Behavior Questionnaire (RISQ) [83]
In order to assess risky and impulsive behavior we use the Dutch version of the RISQ. The RISQ is a 38-item self-report questionnaire-based measure, assessing eight domain-specific factors (measuring drug use, aggression, self-harm, gambling, risky sexual behavior, impulsive eating, heavy alcohol use, and reckless behavior).

Behavioral Impulsivity Scale (BIS-11) [84]
In order to assess the personality and behavioral construct of impulsivity we use the Dutch version of the BIS-11. The BIS-11 is a questionnaire which consists of one of the three second-order facets of impulsivity. The 30-item self-report questionnaire consists of six subscales: attention, (i.e. focusing on current tasks), cognitive instability,(i.e. intruding thoughts), motor impulsiveness, (i.e. acting quickly), perseverance (stable lifestyle), cognitive complexity, (i.e. enjoys mental challenges) and self-control, (i.e. plans and thinks deliberatively).

Alcohol Use Disorders Identification Test (AUDIT) [85]
In order to assess alcohol use we use the Dutch version of the AUDIT. The 10-item AUDIT includes questions to assess alcohol intake (questions 1-3), alcohol dependence (questions 4-6) and alcoholrelated problems (questions 7-10). Questions 1-8 are scored from 0 to 4, questions 9 and 10 are scored 0, 2 or 4, resulting in a maximum AUDIT score of 40.

Drug Use Disorder Identification Test (DUDIT) [86]
In order to assess drug use we use the Dutch version of the DUDIT. The DUDIT is an 11-item screening instrument to assess non-alcohol drug use patterns and various drug-related problems. The items range from on a 5-point Likert scale for the first nine items, the last two items are scored on a 3-point scale. The higher scores suggest more severe drug problems.

Obsessive Compulsive Drug Use Scale (OC-DUS-version Cocaine) [87]
In order to assess drug craving we use the Dutch version of the OC-DUS. The OC-DUS version Cocaine is a 13-item self-report questionnaire that assesses the inability to control or resist cocainerelated thoughts and behaviors, frequency and impact of thoughts and impulses related to cocaine use, and the degree of interference.

Obsessive Compulsive Drinking Scale (OCDS) [88]
In order to assess alcohol craving we use the Dutch version of the OCDS. The 22-item Dutch version of the OCDS was developed to reflect obsession and compulsivity related to craving and drinking behavior. The OCDS has been shown to be specific for the obsessive and compulsive characteristics of drinking-related thought, urges to drink, and the ability to resist those thoughts and urges in alcoholabusing and alcohol-dependent population.

Statistical analysis
Different General Linear Models (GLM) in SPSS will be carried out to analyze the main parameters.
For instance, analyzing the empathy and aggression outcome (empathy ratings and b responses on PSAP task), the outcome variables will be handled as the dependent variables. In addition, the group variable (active vs. sham) will be used as the independent between-group variables, whereas the prepost time will be included as a within subject independent variable. A mixed design will be carried out in which the pre-post intervention of the experimental group will be compared with the pre-post intervention aggression outcomes of the sham-group of patients. The mixed design will be carried out for the PSAP outcomes, empathy ratings, self-report questionnaires and violence risk with the HKT-R instrument.

Power analysis
The sample size is based on the primary outcome and other studies from the previous mentioned guidelines [77] and seen as achievable.

Discussion
This study protocol describes the design of an intervention with multichannel tDCS targeting bilateral vmPFC next to TAU, in comparison with sham condition and TAU. This study will explore the opportunities to introduce a new intervention to improve empathy, reduce violence and reduce craving in substance use offenders.
The present study has several strengths. First, to our knowledge only a few studies focus on increasing empathic abilities [62][63][64][65][66][67][68] or modulate externalizing behavior [69][70][71][72], but none of the studies focus on the implication of tDCS in a forensic sample. These individuals are in need of effective care, and by modulating the brain activity this can be a first step towards a new treatment program.
Moreover, we will also include the participants who drop out during intervention in our five-day intervention program. This will give insight into the consequences of the formal procedures preceding intervention and may eventually enable us to describe profiles of the intervention with tDCS in treatment success and failure.
Finally, this study will contribute to development of a cheaper and less invasive treatment for substance abuse. As mentioned before, problems that substance abuse patients cause are an enormous burden to the community (financial and safety). Previous research has found that substance abuse, especially alcohol and cocaine, are related to (violent) criminal behavior. Money invested in treatment may lead to a large reduction in the costs associated with substance abuse.
Nevertheless, current interventions seem insufficient in treatment of substance abuse in forensic mental health care and are not sufficient enough to reduce violence risk, 66% of the patients reoffend.
Despite the strengths of this current study, several limitations may threaten the quality of our study. The greatest challenge will be to have an adequate response rate at baseline and at the end of the five-day treatment. Our target group will be conducted in a forensic institution and is participation is voluntary, so the participants can decide to quit at any time. The population has an extensive history of treatment and criminal justice, so they might have negative experiences with treatment or research studies. Moreover, the participants already have filled in numerous questionnaires in their time spend in institutions, that they may have become tired or suspicious about the purpose of the proposed study. To keep the nonresponse rate as low as possible, the researches conducting this study are trained to motivate and encourage the participants to complete the program. The 20 interaction will be transparent, attentive and flexible and the researchers will be prepared to give clear information about the purpose, goal and aim of the current study at any time. Furthermore, a treatment that uses neurostimulation and EEG might seem scary and new to the participants as compared to TAU, so extra explanation might be needed to insure the participants that it is safe. So, to recruit as many participants as we aim to in this study it will require a high degree of motivation.
Finally, an active control group (i.e., sham modulation + TAU) might lead to smaller effect sizes of active tDCS than when the tDCS treatment would be compared to no treatment or a wait-list group. However, considering the seriousness of the problems of the forensic substance use offenders it would not be ethical to let them wait for an intervention and it remains important to get a sense of the effects of placebo versus the effects of active stimulation.
In conclusion, with the present study design we are able to explore what the added effectiveness is of tDCS + TAU in comparison to sham + TAU, which could provide valuable information for institutions, researchers, psychologists and the professionals in the field of criminal justice. In addition, this study may help to reduce craving in forensic substance use offenders and with that reduce not only violent behavior but also recidivism. The study will contribute to knowledge about increasing empathic abilities and the functioning of the vmPFC.
Specifically, this study may give insight into implementing tDCS on the vmPFC in the TAU and the setting of daily clinical practice for this complex, multiple problem target group.

Trail status
The current study started February 2019. The data collection will run until the 50 participants all completed the five-day program. After six-months a follow up will be conducted to test whether the effects of the intervention are still present. Until then, the intervention effects are unknown.

Consent for publication
Not applicable.

Availability of data and materials
Not applicable.

Competing interests
The authors declare that they have no competing interests.

Funding
This research project is funded by de Kwaliteit Forensische Zorg (KFZ). KFZ is an organization which allows professionals to cooperate to optimize and increase the quality of forensic care.
In addition Stichting Volksbond Rotterdam and Stichting Koningsheide gave funding for this study.

23
Stichting Volksbond Rotterdam (SVR) initiates and finances innovative research of socially relevant cases of substance abuse and the background of behavior and genetically expression of the addiction.
Stichting Koningsheide supports original and promising projects aimed towards forensic psychiatry.
The financers are not involved in the design of the study nor the drafting of the manuscript.
Furthermore, the financers are not and will not be involved in the subsequent process of data collection, analysis and interpretation.

Authors' contributions
CS and JD are the principal investigators. JD obtained funding for the study. CS designed the trial in collaboration with JD. JD and AW designed the stimulation model for the tDCS.
CS coordinates the recruitment of the participants and the data collection during the study. CS drafted the manuscript with important contributions from JD, IF and AW.
All authors participated in the revision of the manuscript and approved the final version.

Figures
34 Figure 1 Relationship between the impaired brain areas, impaired processes and behavior. The highlighted brain area, impaired process and behavior are discussed in this paper. *DLPFC= Dorsolateral Prefrontal Cortex, vmPFC= Ventromedial Prefrontal Cortex