Effects of Caffeine on EEG Theta/Beta Ratio and Cognitive Control

The Effects of a Single Administration of a Moderate Dose of Caffeine on Cognitive Control and Spontaneous EEG Theta/Beta Ratio

Cognitive control driven by the prefrontal cortex (PFC) of the brain is thought to be important for goal-directed control over stimulus-driven processes. EEG-measured spontaneous theta/beta ratio (TBR) may potentially be used as an electrophysiological marker for this PFC-mediated cognitive control. In the present study the investigators further examine TBR as an electrophysiological marker for cognitive control, by administering caffeine to forty healthy female participants. After a first training session, participants will visit the lab twice in separate weeks, during which they will orally consume one capsule containing 200mg of caffeine, and one capsule containing a placebo substance (double-blind and random order of administration). EEG will be measured before and after capsule consumption, and cognitive control tasks will be administered after capsule consumption. Generally, the investigators expect that caffeine will decrease TBR and thereby increase cognitive control. Additionally, the investigators expect that individual differences in baseline frontal (reflected by TBR) and central dopaminergic (reflected by spontaneously-measured eye-blink rates; EBR) activity will moderate the relationship between caffeine and cognitive control. Furthermore, the effects of caffeine on specifically hypervigilance for threatening distractors (taking into account a possible moderating role of trait anxiety) were examined as a separate research question.

BACKGROUND AND RESEARCH AIMS

Cognitive control is an essential part of cognitive functioning, allowing top-down, goal-driven control over stimulus-driven, automatic processes. The prefrontal cortex (PFC) of the brain likely plays an important role in many aspects of cognitive control, including goal-driven attention, working memory, and emotion regulation. Previous EEG studies examining cognitive control have identified a potential electrophysiological marker for PFC-mediated cognitive control. These studies have found that the ratio between EEG-measured theta and beta brain oscillations is elevated in children with ADHD (who generally experience poor cognitive control), and in unselected adults with poor attentional control. This theta/beta ratio (TBR) is thought to reflect the extent to which prefrontal brain activity exhibits regulatory control over more posterior and subcortical brain areas, which has made this measure useful in the diagnosis and treatment of disorders related to poor cognitive control such as ADHD.

In the present study, the investigators propose a new method to further validate TBR as an electrophysiological marker for cognitive control, by administering caffeine to participants. Caffeine is widely-known to improve various aspects of cognitive control, which would be expected to go alongside a decrease in TBR.

Importantly, cognitive control benefits from both the maintenance of task-relevant representations and the flexible updating of these representations. These cognitive control functions may rely on somewhat different areas of the brain. Namely, whereas maintenance of task-relevant representations may be more dependent on PFC functioning, the flexible updating of these representations may rely more on dopaminergic activity in the striatum. Theta/beta ratio as a marker of PFC activity may thus specifically reflect task-relevant maintenance of information, and contrasting this EEG measure to a measure of striatal dopamine activity may shed more light on the nature of TBR. In the present study, the investigators will do so by measuring spontaneous eye-blink rates (EBR), which have shown to be sensitive to dopaminergic striatal activity.

The first aim of the present study will be to further validate theta/beta ratio (TBR) as an electrophysiological marker for PFC-mediated cognitive control by administering caffeine to healthy participants. The investigators expect that (1a) theta/beta ratio will be inversely related to cognitive control performance (attentional focus, attentional shift, working memory; measured with TACT and N-back; possibly mainly related to TACT focus performance), (1b) caffeine will increase cognitive control performance, (1c) caffeine will decrease theta/beta ratio, and that (1d) these beneficial effects of caffeine on cognitive performance will be mediated by this decrease in theta/beta ratio.

The second aim of this study will be to examine whether TBR and EBR differentially moderate the effects of caffeine on cognitive control. Namely, as these measures likely reflect activity of different brain areas, they may predict differential (possibly inverted U-shaped) effects of caffeine depending on their baseline activity. For example, increasing striatal dopamine through caffeine consumption may only be beneficial when initial striatal dopaminergic activity (as reflected by baseline EBR) is not too high in the first place, because when it is, this may then result in an increase in distractibility, rather than improved cognitive flexibility. As such, the investigators expect that (2a) baseline TBR will moderate the effects of caffeine on cognitive control performance (possibly mainly TACT focus performance), and that (2b) baseline EBR will moderate the effects of caffeine on cognitive control functioning (possibly mainly TACT shifting performance).

The third aim of the present study will be to examine the effects of caffeine on emotional processing, and the role of TBR herein. Importantly, caffeine has traditionally been thought to increase threat-related processing and anxiety, especially in higher doses. However, an increasing number of animal studies are showing that small doses of caffeine may actually be anxiolytic. In part, this may be due to an increase in cognitive and attentional control capabilities. Namely, although the relation is complex, anxiety and attentional control influence each other mutually. Anxiety may increase vigilance for threat-related processing, whereas attentional control capabilities may protect against this. An optimal dose of caffeine could theoretically increase attentional control to the extent that this reduces over-processing of threat-related information. However, the dose-dependent effects of dopamine, nor-adrenaline, and caffeine predict that this optimal dosage depends on basal levels of these transmitters and therefore possibly on anxiety which is associated with increased dopamine and nor-adrenaline activity. The potential usefulness of moderate dosages of caffeine as a prophylactic drug for (attentional symptoms of) anxiety then may depend critically on such a possible moderating influence of anxiety. In the present study, the investigators examined this relation by taking into account the moderating effects of trait anxiety in the effects of caffeine on emotional processing. The investigators expect that (3a) higher TBR will be related to increased interference on the EST, (3b) a moderate dose of caffeine will decrease emotional interference on the EST, (3c) the effects of caffeine on emotional interference will be mediated by a change in TBR, and that (3d) the effects of caffeine on emotional interference will be moderated by individual differences in trait anxiety.

Trait anxiety will be measured via self-report using the STAI-t, which will be added as a covariate in the analyses in which the relation between TBR and AC is examined. Finally, in this study the investigators will explore several other possible confounding factors in the relationship between caffeine consumption and cognitive control, including self-report measures of trait and state attentional control, anxiety, sleepiness, and ADHD symptoms.

STUDY DESIGN

A double-blind, randomized, placebo-controlled, cross-over trial. After a first training session, participants will visit the lab twice, separated by approximately one week. In these two lab sessions, participants will orally consume one capsule containing 200mg of pure caffeine (a similar amount of caffeine is found in about 2 cups of coffee) and one capsule containing a placebo filler substance (randomized and double-blind administration). Blinding and randomization of capsule allocation will be done by the pharmacy of the Leiden University Medical Center (LUMC).

PARTICIPANTS

Forty healthy female volunteers (aged 18-26 years) will participate in this study. Participants will be fairly naïve to caffeine consumption (drinking ≤100mg of caffeine, or about one cup of coffee, per day), and will be asked to abstain from caffeine and alcohol consumption for 12h before the start of the lab sessions. Participants will be students recruited at Leiden University through various advertisements (e.g., at the university campus). Participants who have withdrawn from the study will, when possible, be replaced by another randomly selected participant. For such replacement participants, the LUMC pharmacist has prepared blinded capsules containing placebo and caffeine in the same order as the original participant that they are replacing.

PROCEDURES

After online screening procedures, eligible participants will be tested at the same time on three separate days. Test sessions will take place between 10:00 and 17:00, lasting approximately 2 hours per session. Participants will be asked to abstain from alcohol and caffeine consumption for 12 hours before the start of the experiment. On the first testing (training) day, participants will complete questionnaires assessing demographics, trait levels of attentional control (ACS), trait anxiety levels (STAI trait subscale), habitual sleepiness (ESS), and ADHD symptoms (ADHD Rating Scale-IV). Furthermore, resting-state EEG (8 minutes in alternating 1-minute blocks of eyes open/closed) and spontaneous EBR (one eyes-open block of 4 minutes) will be measured. In addition, participants will be familiarized with the cognitive tasks (TACT; N-Back; EST). Participants will complete these measures to reduce novelty or learning effects in subsequent test sessions. On each day of testing, state questionnaires (current sleepiness, KSS; attentional control, VAS; and anxiety, STAI state subscale) will be administered between each 4min (EBR) and 8min (EEG) baseline measure, and at the end of the day (i.e., administered twice on the first training day, and three times on the day 2/3 consumption days).

The second testing day will be approximately one week after the first testing day. Prior to capsule consumption, EEG and EBR will be measured. Participants will then receive a capsule containing either caffeine (200mg) or placebo (double-blind, randomized administration). After a 30 minute break (as caffeine takes about half an hour to become active), EBR and EEG will be measured again. Finally, participants will complete the same cognitive tasks they completed on the first day.

On the third testing day the testing protocol of the second day will be repeated, with the exception that the other, remaining caffeine (200mg) or placebo capsule will be administered.

COGNITIVE CONTROL TASKS

All cognitive tasks (TACT; N-Back; Emotional Stroop Task) are programmed and completed on a computer using E-Prime (v2.0). The cognitive control tasks are described below.

1. Two-Factor Attentional Control Task (TACT): The TACT is a global/local congruency task which aims to measure attentional focusing and attentional shifting. In this task, participants are shown a large arrow that consists of several smaller arrows pointing in the same (congruent) or opposite (incongruent) direction. Participants are instructed to respond to the direction of the large arrow or smaller arrows as fast as possible without making too many mistakes. They can do so by pressing the left or right arrow key on a keyboard, using the index and middle fingers of their dominant hand. Prior to the appearance of an arrow, participants see a grey screen for 500ms after which the word "Large" in the colour blue or "Small" in the colour yellow appears for 500ms, which indicates which arrow participants have to respond to. After a 250ms grey screen the arrow appears in the middle of the screen, which disappears after 2000ms or when a response is given. On each trial, the arrow appears in a slightly different position to ensure that participants look at the arrow rather than fixate on a few pixels on the screen (which in theory could be sufficient to identify the direction of the arrows).

   In the attentional focus part of the task, participants complete two blocks of 52 trials each. In one block participants only respond to the direction of the large arrow, and in the other block participants only respond to the direction of the smaller arrows. TACT focus performance is measured by subtracting RTs on congruent trials (i.e., trials in which no inhibition of distraction of non-attended arrows is necessary) from RTs on incongruent trials (i.e., trials in which participants have to inhibit their response to the distraction of non-attended arrows). As such, smaller differences in RTs indicate better focus performance. In the attentional shift part of the task, participants complete two blocks of 120 trials, separated by a one-minute break, in which the direction of the arrow to which participants have to respond changes every six trials (i.e., 40 shifts in total). TACT shift performance is calculated by subtracting RTs on the last two trials before an attentional shift from RTs on the first two trials after an attentional shift (i.e., trials in which participants shifted their attention from looking at larger to smaller arrows or vice versa). As such, smaller differences in RTs indicate better shifting performance. The two trials before and after a shift are always incongruent to increase the difficulty (and thus likely the shift costs) of this task. The order of focus blocks (i.e., large arrow or smaller arrows first), and the order of first completing focus or shift blocks, are counterbalanced over participants. Trials in which participants give incorrect responses, or improbably slow or fast responses, are not further analysed.

2. N-Back Task: The N-back task is a task that assesses different working memory processes. In the N-Back task, participants are sequentially shown letters. Participants have to indicate whether the current letter matches the letter shown n places before, using the 1 ("same letter") or 2 ("different letter") on a keyboard. The 1-back and 3-back tasks were administered, as caffeine likely has the biggest influence on 3-back task performance, and may not greatly differentially influence 1-back and 2-back task performance. For each n-back condition, 75 letters are shown in total, of which 25 (33%) are hits (i.e., trials which were shown n places back). The first three trials of each block are never targets. Each letter is shown for 500ms, which is followed by a 1500ms blank screen. Participants can respond to the letter in this entire 2000ms interval. Prior to each n-back condition, participants receive instructions and examples of the task (under supervision of the researcher on the training day), and complete 21 training trials (7 hit trials), to make sure that they understand the task. The order of 1-back and 3-back completion will be counterbalanced over participants. Performance is measured by examining accuracy scores (% correct), and by examining RTs on correctly-answered trials.

3. Emotional Stroop Task (EST): The EST is a variant on the traditional Stroop task (in which participants have to name the colour of a word printed in a different colour name, e.g. name the colour of the word "Red" written in blue ink). In the EST, emotionally-laden pictures are shown, on which a coloured square is superimposed. Participants have to respond to the colour of the square, and as such have to inhibit the distracting influence of the pictures. Pictures are chosen from the IAPS database and are selected based on their balance and arousal ratings to represent three picture conditions: neutral, negative, and positive. For each of the picture condition, 4 pictures (each time with the square in a different location: top or bottom x left or right) are shown 8 times (i.e., 32 trials per condition, 96 trials in total). The colour of the square is randomly chosen from three possible options (red, yellow, or blue) on each trial. During each trial, first, a picture is shown for 200ms, after which a coloured square is superimposed on this picture in one of four possible locations. Participants then get 1800ms to respond to the colour of the square, which they can do using coloured buttons of a standard response box. Then, a 2000ms grey screen is shown, after which the next picture appears. Prior to the actual task, participant complete 24 practice trials in which neutral pictures are shown. Performance on the Emotional Stroop Task is measured by examining RTs on the negative and positive condition trials, compared to RTs on neutral condition trials, on correctly-answered trials.

QUESTIONNAIRES

All trait questionnaires (ACS; STAI-t; ADHD Rating Scale-IV; ESS) are programmed and completed on a computer using E-Prime (v2.0). State questionnaires (STAI-s; VAS AC-s; KSS) are completed using pen and paper. The questionnaires are described below.

1. Attentional Control Scale (ACS): A modified Dutch translation of the ACS will be administered to participants. The ACS aims to measure trait attentional control, as assessed with 20 statements measuring attentional shift (e.g., "I can quickly switch from one task to another"), attentional focus (e.g., "My concentration is good even if there is music in the room around me") and flexible thought (e.g., "I have a hard time coming up with new ideas quickly"). Participants have to indicate to what extent the statements reflect their attentional control on a 4-point scale, ranging from 1 ("almost never") to 4 ("always").
2. State-Trait Anxiety Index (trait subscale; STAI-t): The trait scale of the Dutch STAI consists of 20 items measuring trait anxiety (e.g., "I worry too much over something that really doesn't matter"). Participants can respond to these items using a 4-point rating scale, ranging from 1 ("almost never") to 4 ("almost always").
3. ADHD Rating Scale-IV: A modified Dutch translation of the ADHD Rating Scale-IV will be administered to participants. This questionnaire consists of 23 items measuring ADHD symptoms of hyperactivity/impulsivity (e.g., "I feel restless") and inattention (e.g., "I am quickly distracted"). Participants are asked to rate the extent to which the item applied to them during the past 6 months, on a scale of 0 ("Never or rarely") to 3 ("Very often").
4. Epworth Sleepiness Scale (ESS): The ESS is used as a measure of habitual sleepiness. On the ESS, participants are asked to rate how likely it is that they would doze off in eight usual and recent situations (e.g., "While watching television"), using a scale of 0 ("Would never doze off") to 3 ("High chance of dozing").
5. State-Trait Anxiety Index (state subscale; STAI-s): The state scale of the Dutch STAI consists of 20 items measuring how anxious participants feel at this moment (e.g., "I feel calm", "I am worrying"). Participants can respond to these items using a 4-point rating scale, ranging from 1 ("Not at all") to 4 ("A lot").
6. Karolinska Sleepiness Scale (KSS): The KSS is a 9-point VAS assessing subjective, instantaneous sleepiness. Participants are asked to report how alert or sleepy they are on a scale of 1 ("Very alert") to 9 ("Very sleepy (fighting sleep)").
7. VAS AC-s: The VAS attentional control scale consists of six items assessing participants' current feelings of attentional control (e.g. "I have trouble concentrating"). Participants have to indicate to what extent the item reflects their current attentional control by marking 100mm lines, anchored with "Not at all" and "A lot" on the left and right sides.

EBR AND EEG MEASUREMENT

1. Spontaneous eye-blink rates (EBR): Spontaneous EBR will be collected in a 4 min. eyes-open block using vertical EOG, which records voltage differences between two electrodes placed above and below the eye. In this 4 min. interval, participants will be asked to look at a central point in the room in a relaxed state. Individual EBR will be calculated by dividing the total number of eye blinks during the 4 min. interval by 4.
2. Spontaneous EEG: Spontaneous EEG will be recorded for 8 min. continuously in alternating 1-min blocks of eyes open/eyes closed recording. EEG recordings will be acquired across the scalp using 16 electrodes, positioned according to international 10/20 system, using the ActiveTwo BioSemi system. Electrodes will be placed at the left and right mastoids for offline re-referencing of the scalp signals to the average of the mastoid signals. To measure eye and blink movements, electrooculogram (EOG) electrodes will be placed above and below the left eye (vertical EOG) and external canthi of each eye (horizontal EOG). Offline data processing will be done in Brain Vision Analyzer V2.02. A 0.1-Hz high-pass filter, 100-Hz low-pass filter, and 50-Hz notch filter will be applied. Data will be analyzed in four-second segments. The data will be automatically corrected for ocular artifacts and segments containing remaining artifacts will be removed. A fast Fourier transformation (with a resolution of 0.25 Hz, using a hamming window of 10%) will be applied to calculate area power density for the beta (13-30 Hz), and theta (4-7 Hz) frequency bands. The power densities for the three frontal electrodes will be averaged into measures for frontal beta and frontal theta power density, and used to calculate theta/beta ratios, which will be log-transformed in case of non-normality.

DATA PROCESSING

Outlier removal for each cognitive task will be done on a trial-to-trial basis (excluding improbably fast and/or slow responses) for each participant. First a rough RT removal will be applied, after which a more sophisticated removal of outlying trials will be conducted. Based on previous studies with these tasks, the following trials will be removed:

1. TACT: trials faster than 200ms or slower than 1200ms, after which trials completed faster or slower than 2.5 SDs of the mean RT for that task will be removed.
2. N-Back: trials faster than 300ms will be removed.
3. EST: trials faster than 300ms or slower than 1200ms, after which trials completed faster or slower than 2.5 SDs of the mean RT for that task will be removed.

There is some evidence of the stability of TBR and EBR over time, which complies with the notion that these markers reflect stable, trait-like neural processes. However, the influence of state-dependent circumstances on these measures are still not fully understood. In the present study, baseline TBR and EBR will be measured on each day of testing, but only one of those TBR and one of those EBR baseline measures will be used as trait-like neural marker in the moderation analyses. The selection of which of these baseline measures will be used in the moderation analyses, is based on the following rules:

1. If the investigators find that TBR or EBR baseline measures on day 1 are highly similar (mean comparison and/or correlation) to their corresponding baseline measures on day 2 and day 3, the day 1 baseline measures will be used in the moderation analyses.
2. If the investigators find that TBR or EBR baseline measures on day 1 differ greatly from their corresponding baseline measures on day 2 OR day 3, the day 1 baseline measures will be used in the moderation analyses.
3. If the investigators find that TBR or EBR baseline measures on day 1, day 2, and day 3, all differ greatly from each other, the day 1 baseline measures will be used in the moderation analyses.
4. If the investigators find that TBR or EBR baseline measures on day 2 and day 3 are highly similar, but differ greatly from their corresponding baseline measures on day 1, the averaged baseline measures of day 2 and day 3 will be used in the moderation analyses.

Inclusion and exclusion criteria will be assessed via self-report.

Inclusion Criteria:

• Female
• Age 18-26 years
• Consuming less than 100mg of caffeine (about one cup of coffee) per day on average (evaluated using pre-defined caffeine contents, e.g.: coffee: 85mg/cup; cappuccino: 100mg/cup; tea: 40mg/cup; cola: 18mg/cup; energy drink: 80mg/can)
• Fluent in Dutch language

Exclusion Criteria:

• Severe physical or psychological morbidity that would adversely affect participation
• Habitual smoking
• Use of psychopharmaceuticals