The following section reflects the content of a presentation given at the Project Proposal meeting at the Donders Center for Cognitive NeuroImaging in Nijmegen on 24-1-2019. The powerpoint slides from that meeting are added as supplement. # Title Improving emotional-action control in socially anxious individuals using dual-site phase-amplitude coupled tACS # Have any data been collected for this study already? Data collection is completed. For this study, we will recruit a sample of highly socially anxious students from the Radboud University, Nijmegen. The Liebowitz Social Anxiety Scale (LSAS) will be used as a pre-screening tool and we will only include participants with a total score (fear/anxiety and avoidance sub-scales combined) above 30 (Mennin et al., 2002). A minimum sample size of 45* participants is required to replicate the three-way interaction between tACS-dose (PFC BOLD stimulation vs sham), Congruency (congruent vs incongruent) and Stimulation condition (in-phase vs anti-phase) reported in Bramson et al., 2020 (based on RM-ANOVA - eta2 = 0.19, 80% power). However, based on the following considerations we have decided to increase the sample size in the present study. First, we expect the current sample (as compared to the sample of Bramson et al., 2020) to show a larger variation in the behavioral effects of the task, due to the inclusion of both sexes and of highly anxious individuals (see Behavioral hypotheses 2). Second, due to the increased heterogeneity of our sample and anxiety-related differences, we expect more variation in brain stimulation efficacy. Third, we anticipate the precision of effect size estimates to show regression towards the mean (Wilson et al., 2020). Given these considerations, we have decided to increase our planned sample size to 50 participants. This sample size allows us to detect effects larger than p2 = 0.17. In case we do not find statistically reliable effects with this sample size we will conclude that the effect size of our intervention is likely too small to be clinically relevant. *The minimum sample size changed from 36 to 45 participants because of an error in the initial power calculation. However, this change did not affect our planned sample size of 50 participants. # Have data-analyses been performed already? No. # Study summary. In this study we apply theta-gamma phase-amplitude coupled tACS over the right anterior prefrontal- and left sensorimotor cortex to improve neuronal coupling between these two regions. Improving effective communication between these regions by synchronizing the phase of frontal theta rhythm to SMC gamma amplitude has been shown to improve emotional action control in healthy participants (Bramson et al., 2020a). This study is a direct follow-up on this previous study, with the aims of replicating these earlier results and extending the results towards a more clinically relevant sample (male and female participants with high social-anxiety). # Design Participants perform a social approach-avoidance (AA) task in the MRI. Trials start with the presentation of a white fixation cross presented in the center of a black screen for 1000 ms. After fixation, a face is presented for 100 ms, after which the subject has 2000 ms to respond by pushing the joystick away, or pulling it towards themselves. Participants are instructed to push the joystick towards or away from themselves for happy or angry faces respectively (congruent block), or towards/away from themselves for angry/happy faces (incongruent block), figure 1A. Several studies have shown that this type of control over social-emotional action tendencies depends on aPFC regulation of downstream areas such as the amygdala and motor cortex (Volman et al., 2011; Bramson et al., 2018, 2020b). During the AA task participants receive online, block-wise, transcranial alternating current stimulation (tACS) to the anterior prefrontal- and motor cortex using a ring-electrode setup. This tACS manipulation consists of a theta-band (6 Hz) sine-wave presented over frontal- and a gamma-band (75Hz) sine wave over the motor cortex. The gamma-band sine wave will be tapered with a 6Hz sine wave with power increases locked either to the peaks (in-phase) or troughs (anti-phase) of the frontal theta-band signal. We also included a sham condition with ramp-up and ramp-down stimulation prior to the first trial of these blocks. This manipulation is identical to (Bramson et al., 2020a). Participants will be screened before inclusion and should score a total score of higher than 30 on Liebowitz Social Anxiety Scale (LSAS) (Mennin et al., 2002). # Primary hypotheses and outcome measures Behavioral hypotheses: Given that congruency effects on this task are known to differ between healthy and socially anxious individuals (Heuer et al., 2007), which might influence the direction of the stimulation effects, we will base the analysis on baseline congruency effects, see below. There are two options: 1) If congruency effects on reaction times and error rates in the sham condition (no-stimulation) are not statistically different from the congruency effects in the sham condition in Bramson et al., 2020 we will perform the exact same analyses as reported in Bramson et al., 2020. Namely, a ‘tACS dose’ estimate will be estimated from right PFC† for the contrast stimulation versus sham. This dose response measure will be used as a regressor in a Bayesian mixed effects model testing whether correct responses depend on congruency (congruent vs incongruent) * stimulation condition (in-phase vs anti-phase) * tACS dose. Given earlier studies we mainly expect effects on error rates (Volman et al., 2011; Bramson et al., 2020a), but we will perform the same analyses for reaction times. We expect decreased congruency effects in the in-phase condition for those participants that show a stronger response in PFC. For the anti-phase condition we might find the opposite effect but that will likely be less statistically reliable (see Bramson et al., 2020). We will test whether a possible lack of effect in anti-phase condition is related to compensatory mechanisms in contralateral PFC. †The tACS-dose measure extraction has been focused on dlPFC as evidence has emerged suggesting that high-anxious individuals recruit dlPFC for solving emotional action control (Bramson et al., 2023), and considering that the current distribution evoked by the dual-tACS montage of this study includes dlPFC. Additionally, we will test whether improvement in the in-phase condition is related to social-anxiety score, testing the possibility that those participants with higher anxiety score are less proficient in controlling emotional actions at baseline (higher congruency effect in sham), and can therefore benefit more from in-phase stimulation as compared to participants with lower (relative) anxiety. We will also explore how baseline emotional-action control (congruency effects in RT and % correct) is related to social-anxiety scores. 2) Previous work has shown that socially anxious participants might show differences in congruency effects on the AA task used in this experiment as compared to healthy controls. For instance, socially anxious participants might show increased avoidance tendencies for happy faces (Heuer et al., 2007). This can result in reduced congruency effects at baseline and thereby obscure potential improvements in congruency due to in-phase stimulation. If this is the case, as witnessed by differences in congruency effects between the socially anxious sample and the sample from Bramson et al., 2020 in the sham condition, we will adapt our analyses as follows: Effects of in-phase vs anti-phase stimulation on behavior will be separately determined within the happy and angry conditions. We expect that if in-phase stimulation increases control over emotional action tendencies, avoidance towards happy faces will be reduced, as witnessed in decreased error rates and reaction times. Additional descriptives To ease interpretation with respect to potential clinical relevance of our manipulation we will report statistics describing individual response rates and magnitude. For this we will use person centered effect size (i.e. how many of the participants show the expected effect: e.g. lower congruency in in-phase versus anti-phase or sham; Grice et al., 2020). We will also report on the magnitude of these effects in % change to sham. fMRI We will use BOLD contrasts both as a dose-response estimate (i.e. how do participants differ in their response to the stimulation) and as an index of changed connectivity between PFC and motor cortex. This will be assessed through use of the GLM and dynamic causal modelling (DCM). DCM estimates of changed effective connectivity due to tACS will be regressed against behavioral effects to confirm earlier findings relating tACS-induced connectivity changes to behavioral effects. Furthermore, we will attempt to replicate relationships between tACS-induced changes in effective connectivity and structural connections between aPFC and SMC to confirm explorative analyses reported in Bramson et al. 2020. # Secondary hypotheses and outcome measures In the previous study, we measured GABA+/Glx using magnetic resonance spectroscopy (MRS) to assess possible individual differences in response to tACS due to baseline inhibition in the aPFC and M1. The previous study showed a statistical trend in the five-way interaction: behavioral congruency (congruent vs incongruent)*stimulation condition (in-phase vs anti-phase) * tACS dose * GABA/Glx aPFC * GABA/Glx SMC. We will test this five way interaction in this new sample. This will show whether tACS effects depend on baseline excitation/inhibition ratio in stimulated areas (e.g. Khatoun et al., 2017). Furthermore, we aim to replicate findings showing a relationship between GABA/Glx ratio in aPFC and SMC and behavioral congruency irrespective of stimulation effects (Bramson et al., in preparation). Here, we predict that participants with more excitable aPFC will be better in controlling emotional action tendencies (reduced congruency effect in % correct). This will be assessed using Bayesian mixed effect model testing a three way interaction: Congruency (congruent vs congruent) * aPFC GABA/Glx ratio * SMC GABA/Glx ratio. We will explore potential relations between baseline GABA/Glx ratio’s in these regions and estimates of social anxiety. We will try to replicate the dependence of behavioral congruency effects on structural connections between amygdala and aPFC via the amygdalofugal pathway. # Additional descriptive variables measured • State trait anxiety inventory (STAI) ref • Testosterone and cortisol will be assessed pre and post task using saliva samples. # References Bramson B, den Ouden HE, Toni I, Roelofs K (2020a) Improving emotional-action control by targeting long-range phase-amplitude neuronal coupling. Elife 9. Bramson B, Folloni D, Verhagen L, Hartogsveld B, Mars RB, Toni I, Roelofs K (2020b) Human lateral Frontal Pole contributes to control over emotional approach-avoidance actions. J Neurosci:2019–2048 Available at: http://www.jneurosci.org/content/early/2020/02/06/JNEUROSCI.2048-19.2020.abstract. Bramson B, Jensen O, Toni I, Roelofs K (2018) Cortical oscillatory mechanisms supporting the control of human social-emotional actions. J Neurosci:3317–3382. Bramson B, van Nuland A, Toni I, Roelofs K (in prep). Neurochemical contributions to emotional action control. Grice JW, Medellin E, Jones I, Horvath S, McDaniel H, O’lansen C, Baker M (2020) Persons as Effect Sizes. Adv Methods Pract Psychol Sci:251524592092298. Heuer K, Rinck M, Becker ES (2007) Avoidance of emotional facial expressions in social anxiety: The Approach–Avoidance Task. Behav Res Ther 45:2990–3001. Khatoun A, Asamoah B, Mc Laughlin M (2017) Simultaneously excitatory and inhibitory effects of transcranial alternating current stimulation revealed using selective pulse-train stimulation in the rat motor cortex. J Neurosci 37:9389–9402. Mennin DS, Fresco DM, Heimberg RG, Schneier FR, Davies SO, Liebowitz MR (2002) Screening for social anxiety disorder in the clinical setting: Using the Liebowitz Social Anxiety Scale. J Anxiety Disord 16:661–673. Volman I, Roelofs K, Koch S, Verhagen L, Toni I (2011) Anterior prefrontal cortex inhibition impairs control over social emotional actions. Curr Biol 21:1766–1770 Available at: http://dx.doi.org/10.1016/j.cub.2011.08.050. Wilson BM, Harris CR, Wixted JT (2020) Science is not a signal detection problem. Proc Natl Acad Sci U S A 117:5559–5567.