The effects of stress on attention, and their electroencephalogram patterns

The effects of stress on attention, and

their electroencephalogram patterns

Martijn Keltjens (11209844) Universiteit van Amsterdam

Abstract

Train traveling can cause stress, but also needs a traveler to pay attention. How does stress effect attention? and how can stress resistance predict this effect? There are two forms of attention, top-down and bottom-up attention. Both these pathways work together when in a train station. Stress has effect via sympathetic nervous system and via the HPA-axis. In this exploratory NS pilot study low alpha (8-10 Hz) power was increased during attention when higher stress levels were given. This is against the expectations found in the literature. There was no correlation found between stress resistance and attention task performance. Therefore, from these contradicting results a lab research is proposed. In this proposed study stress would be induced by a socially evaluated cold pressor test (SECPT), and the stress levels controlled via salivary cortisol and alpha amylase samples, heart rate variability and EEG measurements. The attention tasks done will be an arithmetical attention task, a covert and overt attention task and the mirror drawing task. The mirror drawing task is also used to test the stress resistance. Expected is that a bigger alpha decrease is seen while the test is made for participants that are less stress resistant and a decrease in theta/beta ratio while doing the attention task. It is also expected that performance on the tasks are better done by participants that are more stress resistant. The findings from this proposed research study could give suggestion on how to order an optimal train station, for the best travel experience and the optimal amount of distracting advertisement.

Summary: A research proposal towards a better understanding of the effect of stress on attention. Stress

induction is done with SECPT. Stress levels are controlled with salivary samples, HRV and EEG measurement. Attention methods that have been used are a mirror drawing task and arithmetic attention task. Electroencephalogram is used to measure brainwaves, therefore the brain activity.

Keywords: Electroencephalography (EEG) study – Attention; Top-down & bottom-up – Stress – Alpha

power; Theta Beta Ratio – NS – arithmetical attention tasks – mirror drawing tasks – Posner’s attention tasks – Socially Evaluated Cold Pressor Test (SECPT)

Introduction

On a train station it is important to pay attention on the details of the travel information and on which platform the train will arrive. But travelling by train can be stressful. For example, catching the train on time, changing of platform (or train) and mainly arriving on time for a meeting. In addition there are a lot of distractions on a train station, such as the advertising signs of restaurants and shops. Thus, selective attention is important on a train station. Therefore it is important to acquire more knowledge about the effect of stress on attention and the mechanisms in which humans process information during traveling by train under stress.

The environment of a train station has a vast amount of stimuli; travel information, advertisement and a lot of people. Selecting relevant information, while filtering out less relevant information grants us the possibility to respond fast and effective to environmental changes. Prioritizing of relevant information is referred to as selective attention (Katsuki & Constantinidis, 2014). There are two neural pathways for attention. The first one is bottom-up, or exogenous attention, where an external stimulus is changing the focus of attention, almost automatically. Bottom-up attention is stimulus driven, or distraction driven, and is projected from lower cognitive areas toward higher cognitive areas, ending up in the prefrontal cortex (PFC) (Buschman & Miller, 2007). The other form of attention is top-down attention, or endogenous, the attention is induced in higher cortical areas. Top down attention is regulated in the PFC, and projected to the subcortical (motor) areas. The prefrontal cortex inhibits external stimuli that are distracting, and only allowing the relevant information necessary for finishing a task to focus on (Magosso, et al., 2019). When traveling by train both the attention pathways are active and closely work together.

Although any type of stimulus can drive attention, in this research the focus is on visual stimuli. The visual system is mostly studied and is also most commonly used while traveling by train. Since most the travel information and advertisements are presented on screens and posters. The visual bottom-up pathway starts at the primary visual cortex (V1) in the occipital lobe. From the V1 there are two main projections: a ventral pathway, which identifies and recognizes what objects are, and a dorsal pathway, which locates and recognizes where objects are. What is done with the visual information is processed in the PFC (Mishkin, et al., 1983). Both the attention pathways are regulated by the prefrontal cortex. The bottom-up stimulus can shift the attention. If distractions are not desired, they are inhibited from prefrontal cortex, via the top-down pathway. Both pathways are constantly active and work closely together to keep the right focus for the right situation (Katsuki & Constantinidis, 2012).

Stress influences attention, and as said before travelling by train can induce stress, because travelling by train is a time- and location based action (Booth & Sharma, 2009). For example, if a traveler misses the train he/she has to wait longer and might miss their appointment. Stress is a combination of cognitive (short-term) and hormonal (long-term) reactions (Lemmers & Anna, 2017, see Appendix I). The classic endocrinal stress circuit is named the Limbic-Hypothalamus-Pituitary-Adrenal axis (LHPA-axis), where stressful stimuli are projected from the limbic area to the hypothalamus. The hypothalamus releases the neuroendocrine hormones cortisol releasing hormone (CRH) and Vasopressin (AVP). These are released into the pituitary and bind to receptors. CRH and AVP facilitate the release of adrenocorticotropic hormone (ACTH) from the pituitary (Herman & Cullinan, 1997). ACTH on his turn releases glucocorticoids and mineralocorticoids, like cortisol, from the Adrenal gland. Cortisol is the “Stress hormone”. It affects the mineralocorticoid and glucocorticoid receptors in the PFC (Erickson, et al., 2003). The other stress pathway is via the sympathetic nervous system, which releases catecholamine form the adrenal medulla. The catecholamines, dopamine and noradrenaline, resulting in decreased firing rate in the PFC. Noradrenaline does this via projections ascending from the locus coeruleus (LC) towards the PFC. Dopamine from the ventral tegmental area (VTA) projecting to the PFC (Sänger, et al., 2014). The LC is the most stress-sensitive nucleus in the brain (Herman & Cullinan, 1997). The LC is also involved in the inhibition of spontaneous orienting to distracting stimuli, thus uninhibited activation of the bottom-up attention pathway (Sänger, et al., 2014).

Thus acute stress might activate the bottom-up attention pathway, and therefore ensures more distractions.

Both catecholamines and cortisol have effect on the PFC, thus it is interesting to look at the effect of stress on attention in the PFC. More, the relationship between stress and attention is far from understood (Larra et al., 2016). It is known that stress can improve attentional performance but also

decrease attentional performance (Sänger, et al., 2014). For example Eysenck et al. (2007) stated that during anxiety, and thus stress, an increased stimulus driven attention allows for better reaction on salient and threat related information. On the other hand it disrupts the balance of top-down and bottom-up attention and taxes scarce executive resources, resulting in a decrease in attention control and suboptimal cognitive executive performance (buschman & miller, 2007). Therefore further research of the effect of stress on attention is necessary. Resulting in the research question that is addressed in this proposal: what is the effect of stress on the performance in basic attention tasks? Can this effect be predicted with a stress resistance task? In the pilot study two tasks were done to perform attention, a arithmetic attention task and a mirror drawing task. In the proposed research there will be two attention tasks added, the visual covert and overt attention tasks created by Posner (1980).

Since it is known that stress has an effect on attention, why would the a stress resistance task be able to predict this effect. Someone who is stress resistance could be less affected under stress, or should be less stressed from a stressful experience (Westman, 1990). The mirror drawing task is an attention task were top-down attention is needed, but induces stress as well. The mirror drawing task is a task were participants have to draw inside the lines of a pentagram, under time pressure, with the mouse moving the inversed. Thus the mirror drawing task induces acute stress, because the participants have to draw as precise as possible, as fast possible (Homma, 2005). Since it induces stress and perform attention therefore it can give a value for stress resistance. So if a participant performs good on a mirror drawing test the participant is probably more stress resistance.

A reliable method to examine attention and stress in the brain is by recording specific patterns of the Electroencephalography (EEG). EEG is a method to measure brain waves. These waves are initiated via electric pulses by simultaneously firing neurons. These electric pulses are measured outside of the skull with electrodes, with a higher local density of pulses resulting in higher measuered local voltages (Blinowska & Durka, 2006). Specific frequency bandwidths are delta (1-3 Hz), theta (4-7 Hz), alpha (8-12 Hz), beta (13-28 Hz) and gamma (28-100 Hz). Patterns in these bandwidths are markers and can show different functions of the brain. For attention there are some alterations in the brain waves. Magosso et al. (2019) found out that a decrease in alpha power improves the information processing in the corresponding brain networks. For selective attention the corresponding brain network is PFC. Next to the alpha band a change in the theta/beta ratio (or slow wave/ fast wave ratio) is seen. These slow wave (SW) frequencies, theta and delta, originate from the subcortical (limbic) brain areas (Schutter, et al., 2006). In this paper delta is excluded. The fast wave (FW) frequencies (beta frequency) originate from the cortical brain areas, opposed to the slow wave. SW frequencies originate from subcortical areas. So during selective attention it is suggested that a decreased theta/beta ratio is observed and therefore one of the hallmarks to examine selective attention (Putman, et al., 2014). So in this research it is looked at

Stress as well can be measured with EEG. According to Alonso et al. (2015) during stress there is a decrease of high alpha power and an increase of high beta. Tran et al. (2007) found, just like Alonso, that stress decreases in alpha power during stress compared to the baseline. Tran et al, did not find other differences in brain EEG pattern under stress.

People that travel with the train are mostly more stressed then in normal circumstances, due to unpredictable events and the chance of missing a train. There is not much empirical research done on stations (Cox, et al., 2006). For organizations like Nederlandse Spoorwegen (NS) and Prorail it might give an insight in how people think and how people keep their attention in stations, while under stress. So train companies can create the most effective train station for the best travel experience and optimal advertisement on the stations. Therefore an exploratory pilot study was performed to examine the effects of stress on attention, this will be described next.

Introduction NS study

Ahead of this proposal an exploratory research was performed. The data was gathered in 2019, on Utrecht Centraal in the stationslab. The study was done by the University of Amsterdam in collaboration with the Nederlandse Spoorwegen (NS) and Prorail. The research was led by A. B. Mulder. The goal of this research was to evaluate how the train travelers would react on different kinds of sounds and perform in an attention task on train stations. The sounds given to the participants was nature sound, easy to listen to music and sounds from the surrounding. Afterwards the subjects had to make an

attention task, consisting of simple arithmetical tasks, finish number series and give three rhyme words. In addition, standard information was asked through a questionnaire and a stress resistance task was done. The research question in the pilot study was how people on train stations perform on attention tasks. In this pilot study there was not focused on the part which audio track was played to the participants. Attention was assessed with both EEG, since this is a reliable method for measuring attention, and the results in performance of attention tasks, looking at speed and percentage correct. The questions, that form the attention task, consist of three components. The first component is the visual component, where the arithmetical operation, number series or the starting rhyme word are shown; The second component is the computational component, where the questions are mentally solved and the last component is the motor component, typing the thought up answer. These tasks are comparable to the study of Magosso et al. (2019) where selective attention was researched with an arithmetical attention task. The data received for this task is analyzed on a couple of criteria. The EEG data was analyzed for specific markers, and the percentage correct answered questions and the response time for these questions. The expectations are that during stress and attention alpha power is decreased and theta/beta ratio is smaller. Because both stress as attention decrease alpha power and increase beta power. Also expected is that theta power decreases during attention. Since stress is a positive stimulator on short term attention tasks (Vedhara, et al., 2000). It is hypothesized that stress will have a positive effect on task making. Especially the reaction time (RT). Such as the task given in this research. Next to the attention task a stress resistance task is done. This is a task that participants will do better if they handle stress better. Stress resistance means working under pressure/ or stress. In this research the stress resistance task is a mirror drawing task. If the participants can perform better under stress, they will make less mistakes or finish the task faster.

Method NS study

The participants were all train travelers on Utrecht Centraal. In total there were 25 voluntary participants that did both the mirror drawing task (starting with 118 participants) and the attention tasks (starting 39 participants). During the attention task EEG data was collected. The 25 subjects were between 9 and 82 years old (mean age =38.4, SD = 19.69 years). There were 11 man and 13 women and one participant who did not specified its gender. All participants were included, also the one who did not give a gender, because this exploratory pilot study was used to create an overall view of the population. The audio track played was for 9 participants music, for 12 participants nature sounds and for 4 participants sound of the surroundings. From the 25 subjects 10 were secluded because of too many artifacts. All participants gave informed consent and all data was analyzed and reported anonymously.

Experimental tasks

The subjects took the tests in the Stationslab, a little room built inside a hall of Utrecht Centraal (see Appendix 2). First the participants filled in a questionnaire of general personal information. The people had to say how stressed they felt, between 1 and 5. Where 1 meant not stressed and 5 very stressed. The participants, while waiting, were presented with either 90 seconds of music, nature sounds or sound from their surroundings. The music played to the participants was typical elevator music, with slow and monotonous sounds, which was easy to listen to. The nature track consisted of sounds form tree leaves in the wind, birds chirping and a slowly streaming river. For the surrounding group no track was played and they heard the general sound on the station, people, trains and conductors. This was done for the original reason of this pilot study, to see the effect of different sounds on performance during attention task, but that is not the focus of this study.

Next, the subjects had to do a attention task, consisting of some basic mathematic calculations (10), finish number series (5) and give three rhyme words (5), while being measured with EEG. These attention task consists of three components, the visual component, where the stimulus is shown on a computer screen, the computational component, where the questions are mentally solved, and the motor component, typing the answer. Performance of the questions were scored by two factors, the number of correctly answered questions and the reaction time needed for the questions. The attention task consisted of 20 questions, each worth 0.05 points so that a perfect score was 1.00. If the answer was partly wrong, no points were given. The reaction time was averaged per question and then normalized

to 1, since some questions would take longer than others. Then normalized results were averaged over all subjects.

For the whole experimental timeline see figure 1. The subjects had one EEG electrode directly placed on the forehead, used to scan brain oscillations from the PFC region. A ground electrode was placed on the forehead next to the measuring electrode and a reference electrode was placed on the right earlobe. There was no heart rate electrode used.

In addition to the attention task, the 25 participants also completed a mirror drawing task. The mirror drawing task is to test the stress resistance. The task is done on the computer. The participants had to draw a line between two boundaries of a pentagram. Their task was to stay between the lines. The computer mouse was mirrored, so the mouse on the screen went the other way as the mouse in the subjects hand. The task was valued on crossing of the lines, how far over the lines and the time needed to finish the task.

Figure 1. Diagram of the structure of the experiments. First an introduction was done, with a general survey and explanation of the tests. Than the attention task was done or the mirror drawing task, some participants did both tasks. The attention task was measured with EEG, so during the instructions EEG was installed, than participants had to close their eyes and listen to a 90 second audio track. Finishing with the attention task. The stress resistance task was done by mirror drawing task.

EEG analysis

For the EEG analysis the EEG electrode was placed on the forehead, so gave a representation of the PFC. The EEG data during the attention task was selected. The data was filtered between 0.5 to 48 Hz, to remove the alternating current switch frequency. The filtered data was cut in time frames of 5 second. Depending on the time needed for the attention task, since not all participants took the same time for the attention task, all 5 second trials possible during the attention task were used. The deletion of trials was done visually, based on the criteria for artifacts. The artifact criteria for removing trials were: if the amplitude of the oscillations exceeded 100 µV twice or more, or 200 µV once, or is the amplitude exceeded the average +2 standard deviations (SD) twice. Furthermore, If there were more than 3 eye-blinks or eye movements the trial was also deleted. These criteria were less strict then normal, because the data set had so much inaccuracies. Leading to the data of 15 participants. After artifact removal the EEG data was analyzed. Attention and stress can be characterized in EEG by analyzing the power (µ𝑉2_{/𝐻𝑧) of the bandwidths alpha, beta and the theta/beta ratio (see Introduction). The average power}

from these bandwidths during the attention task were calculated. The power of bandwidths theta (4-7 Hz), low alpha (8-10 Hz), high alpha (10-12 Hz), low beta (13-20 Hz), high beta (20.5-28 Hz) were calculated during the attention tasks. The delta waves (1-3 Hz) were not included, since the delta waves had too much noise.

Groups & Statistics

In this research the participants were divided on their performance during the attention task, their stress levels and the audio track the participants listened to, creating three different distributions. The performance was based on their reaction time and percentage correctly answered. This is shown in figure 2, it can be seen that this division is based on normalized reaction time, since all the participants answered above 50%. The spread of the percentage correct was not so big with 13 out of the 15 participants scoring between 0.85 and 1. All participants scored above the 0.5, which would be a small chance if everything was guessed, therefore it can be concluded that the subjects participated actively. Thus paying attention. Also a group of two participants is not preferred. Resulting that the percentage correct is not included. Creating two groups based on the normalized reaction time shown in figure 2. The reaction time of the questions is normalized by averaging per question and then per participant. This division in performance was done to see if there were any differences in brain oscillations characterizing for attention and stress, when the subjects performed faster. To reach this goal a two-sample T-test was used with reaction time as group division (figure 2).

Figure 2. The performance of the attention task. The x-axis represents the correct answers, on a value of 0.05 per correct answer. The Y-axis gives the normalized reaction time (RT) averaged over all the questions. The area in red, would be excluded from the data set, since these did not participated well. Creating two groups RT>1 and RT<1.

Next to the performance division the participants were divided by their stress levels given in the questionnaire. This questionnaire was given to the participants at the beginning of the experiment. Stress levels given in the questionnaire were between 1 and 5, with 1 not stressed and 5 very stressed. The highest stress level given was 3. So for each given stress level a group was made, three in total. The goal of the data analysis is to explore group differences in alpha suppression, theta/beta ratio and RT (performance), between the three stress level groups. This was done by a unpaired ANOVA was done between the three stress level groups.

The second question asked was if stress resistance could indicate how stress would affect the performance of the participants. Since the mirror drawing task is a stress resistance task, therefore if a person is more stress resistant they will perform better on the mirror drawing task. The mirror drawing task was analyzed with a two-sample T-test where the groups were divided by performance based on reaction time. Since the performance was valued by reaction time.

Finally the participants were divided by the audio track they heard creating three groups: music, nature and surrounding, this was done to see if a specific audio track affected the attention visualized in EEG oscillations as described earlier. Therefore would interfere with the main question, because of multiple manipulations in the experimental design. This was also done by a one-way ANOVA for each bandwidth (Theta, alpha low and high, beta low and high), with audio track (music, nature and environment) as between subject factor, to see if there was any differences between the music the people listened to.

Results of NS study

In figure 3 a visualization of the different brain oscillations compared to the reaction time is shown. There was no significance found when the EEG power was analyzed for the reaction time groups. Although the high beta waves (20.5 – 28 Hz) in the slow reaction time group had a higher mean for power than those in the fast reaction time group. The difference in high beta power was not significant (p = 0.1867, t=1.394, df=13). Since there was no significant difference in brain oscillations between the fast reaction group and slow reaction group, it is suggested that the performance of an attention task do not differ the brain oscillations. This could be explained because all participants were doing the task, therefore were all performing attention.

Figure 3. EEG power for slow and fast reaction time. Power is given in 𝜇𝑉2_{/𝐻𝑧 . Graph a shows the power for low alpha} (8-10 Hz) oscillations for the fast and slow reaction time group. Graph b shows the high alpha (10-12 Hz) power for the slow and fast reaction time. Graph c shows the theta/beta ratio for slow and fast reaction time. Graph d shows the low beta (13-20 Hz) power for slow and fast reaction time. Graph e shows the high beta (20.5-28 Hz) power. Graph f shows the theta (4-7 Hz) power for the slow and fast reaction time.

Stress levels

The participants gave an average of 1.67 out of 5 (SD = 0.7) on their stress level question. Thus overall the participants were not very stressed. The three stress groups were compared with the alpha power (high and low), the theta/beta ratio and the RT. This resulted in a significant difference between the stress groups and the low alpha oscillations (8-10Hz) (p = 0.0351, F(2, 12) = 4.487). The alpha power increases gradually in power for higher stress levels, with stress level three having the highest power. The mean gets higher when the stress level rises, this is against the expectations. The expectations were that the alpha power would decrease when attention was performed under stress.

Also it can be seen in figure 4that the RT in stress level three is higher than the other two stress levels (f(2,12) = 1.306, P = 0.3069). This might be understood since stressed people tend to work faster (Eysenck et al., 2007). For the high alpha power (10-12 Hz) a reduction in spreading is seen (f(2,12) = 0.0435, p = 0.9575). Furthermore there was no significance nor interesting trends found in theta/beta ratio(f(2,12) = 0.04189, P = 0.6672). This could be due to higher stress levels have lesser people, thus these groups are not proportionally distributed. The group with stress level three had only 2 participants opposed to the stress level group 1, that had 8 participants.

Figure 4. The EEG power for three stress levels. Power given in 𝜇𝑉2_{/𝐻𝑧. In graph a the low alpha (8-10 Hz) power is shown} for the three stress levels, this shows a rise in alpha power when stress levels are higher. Graph b shows the high alpha (10-12 Hz) power for the three stress levels. Graph c shows the theta/beta ratio for the three given stress levels. Graph d shows the performance in reaction time for the three reaction times

The attention task performance and mirror drawing task performance

In the mirror drawing task there was no significance found when the subjects were divided in slow or fast reaction time (p-value = 0.5089, t=0.6691, df=28). In figure 5 the time is plotted against errors. Errors are the crossings of the outer lines. In figure 5 it is seen that, as well as the statistical analysis, no difference in performance on the mirror drawing task is seen, based on the performance in the attention task. It would be expected that participants that scored fast and with little errors on the mirror drawing task also had a fast performance on the attention task. So a clustering in the left corner would be expected with fast reaction time and few crossings, and a clustering of the reaction time above 1 were much time elapsed and with a lot of crossings. Therefore a stress resistance task does not indicate the effects stress can have on attention task. Although there might be different factors that have to be taken in advantage.

Figure 5. Mirror drawing task. in this graph the X-axis shows the time needed for the participants to finish the task. The Y-axis shows the amount of crossings. The blue dots represent the fast performance group and the red squares represent the slow performance group on the attention task.

EEG and sound

The power bands of the audio track groups were analyzed and compared with an ANOVA. Figure 6 shows the power plots of the different bandwidths theta, high and low alpha and high and low beta. Overall there was no significant difference found, therefore the audio track played to the subjects does not interfere with the rest of the research. Some small differences can be visually seen in figure 6. In theta (4-7 Hz) the means are not far from each other and no difference was found (f(2,12) = 0.1807, P = 0.8371). For both low and high alpha powers the nature group has the lowest mean power and surrounding and music both have a higher mean power (alpha (8-10Hz) = f(2,11)=1.033, P = 0.3879, alpha (10-12Hz) = f(2,11) = 1.285, P = 0.3151). In the high beta (20.5-28 Hz) a difference in means is

visible, although there is no significance (F (2, 11) = 1.598, P = 0.2460). In low beta (13-20 Hz) similar differences are seen as at high beta, but less powerful (f(2,11) = 2.107, P = 0.1680).

Figure 6. The EEG power for three audio tracks played. Power given in 𝜇𝑉2_{/𝐻𝑧. The three audio groups are nature sounds,} sound from the surrounding and elevator music. In graph a the mean and SD for theta (4-7 Hz) power is showed for the audio groups. In graph b the mean and SD for low alpha (8-10 Hz) power is showed for the three audio groups. In graph c the mean and SD for high alpha (10-12 Hz) power is showed for the three audio groups. In graph d the mean and SD for low beta (12-20 Hz) is showed for the three audio groups. In graph e the mean and SD for high beta ((12-20.5-28 Hz) is showed for the three audio groups.

Discussion NS study & research proposal

If higher stress, and/or selective attention was felt, a decrease in alpha power was expected. An opposite result was that low alpha power increased significantly when higher stress levels were measured. The other markers for stress or attention, a decrease in theta/beta ratio or a rise in beta power, were not measured for both the reaction time, nor the stress level given. Further there was no difference in reaction time during the attention task, when looked at different stress levels. So for all results the EEG data conflicted with the expectations from the NS study. So it can be concluded that after the NS study it is still not clear how stress affects attention on train stations. The data had a lot of artifacts, probably since it is very hard to create clean data on a train station, since it is not a lab environment. People will have a lot of distractions and will move around more easy outside a lab environment.

Therefore another research is proposed, a follow up study with the same research question. How does stress effect the attention during attention tasks? And can this be predicted by a stress resistance task? So the effect of stress on attention control is better understood. To receive more clean data the research should take place in a lab so all manipulations can be controlled. Also more participants would be preferable to give more power to the statistical analysis. In addition to the pilot study the follow up experiment would consist of a couple of different factors.

First factor is to induce the stress, since the participant had to give their stress level on a train station. Since the participants had given an average stress level of 1,67 out of 5, it is preferable to induce more stress. The highest stress level given was 3. It is understandable that train travelers who felt a stress level 5 would not participate in this study. They would be in a hurry to catch their train or prefer to wait at the right platform. Therefore it can be concluded that the stress factor was hard to examine and no stress level 5 participants would participate in another station study. So stress would be induced with a socially evaluated cold pressor task (SECPT). This is a method were the subjects put their hand in a cold bucket between 0-3°C, this induces physical stress. But in addition to the cold pressor task the SECPT also ads a social-evaluative element, this is done by constantly filming the subjects. The social element increases the cortisol level significant (Schwalbe et al., 2008). The average duration of the

SECPT effect is about 30 minutes. The control group will put their hands in a bucket around 37 °C and will not be filmed. This will be done to create a stressful environment that is similar to travel stress.

Next factor that would be done in the proposed research is the control of stress. In the NS study the people had to tell how stressed they were, this is a subjective observation. Since some people would feel stressed earlier than others, this could interfere with the real results. But since the experiment is also about stress resistance the subjective stress levels are still asked. The level of stress would be controlled. To control the amount of stress that someone experiences, salivary cortisol and alpha amylase samples would be taken. High levels of cortisol and alpha amylase are markers for stress (Sänger, et al., 2014). Next to the salivary measures, a constant heart rate measurement would be done, to analyze the heart rate variability. The hearth rate variability is the time between heart pulses. Associated with stress reactions an increase in low frequency (LF, centred around 0.1 Hz) heart rate variability and/or an decrease of high frequency (HF, 0.12 or 0.15 – 0.4 Hz) power, thus an increase in LF/HF ratio would be seen (Berntson & Cacioppo, 2004). As last stress could be controlled by analyzing the alpha and beta power before and after stress induction. These control systems would be compared to before the stress induction.

Another factor that will be added, to the addition attention task and mirror drawing task used in the NS study, are the visual attention tasks introduced by Posner (1980). This will be done to see how the stress affects attention on multiple factors of attention. So attention will be analyzed per pathway. Posner et al. created two attention tasks for covert and overt attention. These are visual tasks specific for top-down or bottom-up attention. Together with the arithmetical attention task and the mirror drawing task, this would be comparable to the attention needed on a train station. The visual-computational-motor pathway, as described in the introduction, will be active in all these tasks. Therefore there is an interest in the visual cortex were stimuli first enter the brain. Also the mirror drawing task will be done to answer the second research question of the effect of stress on attention can be controlled/predicted by a stress resistance task. Stress resistance is a term on how people react to stress. In this research proposal the mirror drawing task is a test used to indicate the stress resistance. All these attention tasks are done to realize similar attention as needed in a train station.

To acquire the EEG data of the brain areas interested in, a brain cap should be used with 9 electrodes. During the tasks, shown on the computer, a view of the PFC, posterior parietal cortex (PPC) and the visual cortex (occipital cortex) will be needed. Since both hemispheres would need to be recorded. For the heart rate variability an standard electrocardiogram (ECG) would be made, and thus a standard electrode placement method would be followed for this. The last part added is a baseline, this will be done to see how changes in attention or stress are compared to a resting state of the brain. This concept is widely used in EEG research.

Experimental design

First the participants get an little introduction about the EEG method, the SECPT and the goal of the experiment. then a survey is filled in by the participants, in this survey some general information is asked, as well as informed consent and the stress level. afterwards the participant is the installed behind the computer (±50 cm) with the EEG brain cap. When the brain cap is installed and the subject is placed behind the computer a mirror drawing task is performed by the participants as well as a baseline, at this moment stress is not yet induced. Than two sessions of SECPTs are performed. These sessions consist of a SECPT, a salivary control and three attention tasks. The three attention tasks, arithmetical attention task and Posner’s covert and overt attention tasks, will be given in random order so the duration effect of the SECPT will be irrelevant. For the first round also a baseline is done after the SECPT so the difference in rest without stress and rest with stress can be analyzed. Between the two sessions a short brake is given, so the participants can get relaxed after half an hour of doing tests. After the second SECPT again a mirror drawing task is done, since there is an improvement in performance when the mirror drawing task is done multiple times. Finally, the EEG brain cap is removed and an evaluation with the researcher is done. This is done to see if there are no mistakes made. This method is preferable over the method of the NS study, since the stress can be controlled, the lab environment gives less artefacts and more EEG electrodes are added therefore a more complete view of the systems are visualized. See the experimental timeline and the electrode placement on the brain cap in Appendix 3Appendix 2

The obtained results give insight in how attention react on stress. For train companies this could give useful information on how to order train station, so the travelers will have the best travel experience. For example if the results suggest that stress increases the accessibility for distractions. Then there should be less advertisement in a train station.

Not only for train companies this would be useful information, but also for student, teachers and all parties that deal with stress and attention in their lives. stress and attention have big role in human processes. Therefore more knowledge how stress affects attention and every aspect of attention differentiated is needed.

Literature

1. Alonso, J. F., Romero, S., Ballester, M. R., Antonijoan, R. M., & Mañanas, M. A. (2015). Stress assessment based on EEG univariate features and functional connectivity

measures. Physiological measurement, 36(7), 1351.

2. Berntson, G. G., & Cacioppo, J. T. (2004). Heart rate variability: Stress and psychiatric conditions. Dynamic electrocardiography, 57-64

3. Blinowska, K., & Durka, P. (2006). Electroencephalography (eeg). Wiley encyclopedia of biomedical engineering.

4. Booth, R., & Sharma, D. (2009). Stress reduces attention to irrelevant information: Evidence from the Stroop task. Motivation and Emotion, 33(4), 412.

5. Buschman, T. J., & Miller, E. K. (2007). Top-down versus bottom-up control of attention in the prefrontal and posterior parietal cortices. science, 315(5820), 1860-1862.

6. Cox, T., Houdmont, J., & Griffiths, A. (2006). Rail passenger crowding, stress, health and safety in Britain. Transportation Research Part A: Policy and Practice, 40(3), 244-258.

7. Erickson, K., Drevets, W., & Schulkin, J. (2003). Glucocorticoid regulation of diverse cognitive functions in normal and pathological emotional states. Neuroscience & Biobehavioral Reviews, 27(3), 233-246.

8. Eysenck, M. W., Derakshan, N., Santos, R., & Calvo, M. G. (2007). Anxiety and cognitive performance: attentional control theory. Emotion, 7(2), 336.

9. Herman, J. P., & Cullinan, W. E. (1997). Neurocircuitry of stress: central control of the hypothalamo–pituitary–adrenocortical axis. Trends in neurosciences, 20(2), 78-84.

10. Homma, S. (2005). The effects of stress in response to mirror drawing test trials on the electrogastrogram, heart rate and respiratory rate of human subjects. Journal of Smooth Muscle Research, 41(4), 221-233.

11. Katsuki, F., & Constantinidis, C. (2012). Unique and shared roles of the posterior parietal and dorsolateral prefrontal cortex in cognitive functions. Frontiers in integrative neuroscience, 6, 17.

12. Katsuki, F., & Constantinidis, C. (2014). Bottom-up and top-down attention: different processes and overlapping neural systems. The Neuroscientist, 20(5), 509-521.

13. Magosso, E., Ricci, G., & Ursino, M. (2019). Modulation of brain alpha rhythm and heart rate variability by attention-related mechanisms. AIMS Neuroscience, 6(1), 1.

14. Mishkin, M., Ungerleider, L. G., & Macko, K. A. (1983). Object vision and spatial vision: two cortical pathways. Trends in neurosciences, 6, 414-417.

15. Larra, M. F., Pramme, L., Schächinger, H., & Frings, C. (2016). Stress and selective attention: Immediate and delayed stress effects on inhibition of return. Brain and cognition, 108, 66-72.

16. Lemmers, S., & Anna, M. (2017). Stress, life history and dental development: a histological study of mandrills (Mandrillus sphinx) (Doctoral dissertation, Durham University).

17. Posner, M. I. (1980). Orienting of attention. Quarterly journal of experimental psychology, 32(1), 3-25.

18. Putman, P., Verkuil, B., Arias-Garcia, E., Pantazi, I., & van Schie, C. (2014). EEG theta/beta ratio as a potential biomarker for attentional control and resilience against deleterious effects of stress on attention. Cognitive, Affective, & Behavioral Neuroscience, 14(2), 782-791

19. Schutter, D. J., Leitner, C., Kenemans, J. L., & van Honk, J. (2006). Electrophysiological correlates of cortico-subcortical interaction: A cross-frequency spectral EEG

analysis. Clinical Neurophysiology, 117(2), 381-387.

20. Sänger, J., Bechtold, L., Schoofs, D., Blaszkewicz, M., & Wascher, E. (2014). The influence of acute stress on attention mechanisms and its electrophysiological correlates. Frontiers in behavioral neuroscience, 8, 353.

21. Schwabe, L., Haddad, L., & Schachinger, H. (2008). HPA axis activation by a socially evaluated cold-pressor test. Psychoneuroendocrinology, 33(6), 890-895.

22. Tran, Y., Thuraisingham, R. A., Wijesuriya, N., Nguyen, H. T., & Craig, A. (2007, May). Detecting neural changes during stress and fatigue effectively: a comparison of spectral

analysis and sample entropy. In 2007 3rd International IEEE/EMBS Conference on Neural Engineering (pp. 350-353). IEEE.

23. Vedhara, K., Hyde, J., Gilchrist, I. D., Tytherleigh, M., & Plummer, S. (2000). Acute stress, memory, attention and cortisol. Psychoneuroendocrinology, 25(6), 535-549.

24. Westman, M. (1990). The relationship between stress and performance: The moderating effect of hardiness. Human performance, 3(3), 141-155.

Appendix

Appendix 2

Stationslab: janpauldewildt. (6.10.2018). Tweet text [Hoe weet je brein waar je moet zijn? Kom proefjes doen in het NS Stationslab, het pop-up laboratorium van NS & de UVA. Dit weekend van de wetenschap op Utrecht CS.]. Retrieved from

https://twitter.com/janpauldewildt/status/1048526276808646657/photo/1

Appendix 3

Experimental timeline:

The electrodes placed in the brain cap

Appendix 4

Grantt chart of the experiment in weeks

0 5 10 15 20 25 30 35 40 45

Recruitment of subjects

Writing introduction & method

Conducting experiments

Preprocessing raw data

Statistical analyses